oo¢ Weel pT. ee Shi ty Reiveny tet eh ofegat tebe ve reeeir Penpals © & weet ; Ateynst a or Stren bolo atee # y gee Type ae at ’ Ls dane te Go tenhgae ae bee ase dedeiel of «ey eta reee fr err tr ty gene sai setiied 9 wie ' saites vives Nete dy Vine) tO! ‘ mq J be te Ls NC Ue ent Ly ria? Wet ety naw ae tebe b reatale ape ite a eed a et om be ort) a) Ay ned Sad eat “44 a Peerryerre rT seat Qh reper wptehetel gate 4 \ngae behele | Ne ga Ms A Gre 014,504 . ae bol Bt: aE see hs “Hy! a hedin iat Avehs? ‘ rey an Ri wteG degre woeuatte te nets ay) oe @ seta athe aceeete pie at ‘ chy ve ‘ by sitvepet tiinet) ab he wha lete ety) * ey Herre passhyt eee Ved was eters Pi ieteh ‘arty ie re Pee | Teraenats diatitieee +0" ay \ ferry tees "9 '' errerete “ry ‘ “i Merete ”) rane \ pty Paras ‘ teaere Parerit) Steet ar , re : ie p/ Cone be Ae bem Bel 05 dee a > i r 4 . sean ss : atthoety i : ; @ , myth opois thi verted “eo waa i aearte c teh Teste Whyte vai eile f abe wed gh MM tinea ne. We bred gt Fe eta grey etree hata Sete . oe Re ihe fat) . ANA ¥ eer tr tt tee rs 1 rie rem part rr Nines ne he “ani it aad Od bi wadiae 6 ee heared) te yr hie eer Fy be } te at eid nant tees hey bs ed wg a daub dB it ay et Yaaadny . ri NAnRG teh st “ thy « t ' ms "i oa oe i) Anh of (ata ety ta be! ’ whitey feetnty abi rd 9 Sp84 | Vedvdewey ti Arka bres ae hint Oy att ty CGP ies baad 8 Ohad ying Cyr ater pe tedy abate tyye ett ‘ nee ach O45") e%: Aysassigienets bee Diy er eh Vsatselis Pia ae swes dene eae & Prey ed cert : A pee ewes ad gal bted. 4450 te Miltwisiechoet ‘ee hy . bia ; 4 aot y a) et i Baas) Letk 1 bee ° ‘i $ feo ee tee Tet aed ahs Ac 19a) ¢ cephha A be%p-Y thMg' Th aihart athena eel ey wheter te) VO 8PSt tone eee eh AP a | “4 15h re tee SPEDE Cd 3.5. be ag aan tate 2 ats Sotene ga)t © AP iF sdabsoa! the Ae hye Ants LH Nt Oe My isan | aera de te le [hei dbe bh ee Acs oi N , i Uh Mphetdael sated hi ded , Hy Haein La ye teh at shih Foe gater yeh iia. se4 vie iv prt a “4 ye ta he ge ’ Rie ht 4 5 bh wath, er we beria’ hy is pe Me etbeth. F Olt eT ESOP ITTY! FES Lat bee a satires ot Gaines 5 ag’ hats ] +f Sy yatetal WE aS ay PL ERIa ESS ia phases v4 i as ghebe te aees a yer rr roan’ Vine A, “4 Py Stag he ya ees ee ’ ' Hane baa +} 485 bal aitas rewyar hry 2g Mie} 2 rhe + paecnebe taht el bry * Ria be date DS satay c nae tye eh eh ab mt ian ic Ver he gern? ae. haw wey Wer seh Hyer Wiel, Ped to8 Nib «he Wega 02 Ps, yee Fi of Ph eel hee eteteiate aie MOL Mh bet tbe ” bihiee * ty Hy! 4 jet hetly etal a He haty teens ede \hee ih “s) fry) oe tet he Ge gees + Ne ROR EY Eee ayia a Pirate i ty wor dshe' H ie LO ity va be th) a hat Yor lettre rat one aah prvabenl t( MEO teat} Mois ahh ie As Seeder it oat Wh Laois Sopra tat fe he ie hy M 4 } yO j04 ny ? vende le! hey PLS Vr) ae ka eae! He Oi: ‘9 dak Ti at > be +c Sk Teg bet Pete ede Py ict ars 7 tem ate s0e* wa ie sreegete 4 eh teers preted Fie Paha ehat yes ‘ stat us ia atnas 4 tay . 1h) es Spite Ketel ete Stari ha HNg9 ol as bade Bit Wie or LE wee ba) L® * richie ae apn) veges ie M ea ber ph poke ivy 2 patere te lade nad vies ay af) " ng atlah bg rity ’ Peg fh tats oo a 4 ‘ dart? hawt | hei serie ie 4g ae 4s 7 : ‘ ae M4) ; ano pe aa pe tehad ane dy pene eh bade p hob ae 4 A | hytiw i ra fie b “i Wad sthe ba 7 bp Ratitetet ciate ir F aa dpaabeoney Leta tatalta yi Theite: s) »4elta 44 A da hy gta sabe tele dy ve ragty 445 Hie ayhe dele ai plete (7 Soke e tabe Ae ldats te Meta taire Va) ates age Rta cad wis SAE dia! an caplet Piatt 7p Frethe®, i" tetas His a it oni tyarte > yetetetideds f- r " Rm} eas iat ety La Ob eka od ‘ eh sas oielatitat a tytide ibe ab Wee ces deh atey Ha Pal gh: eq! LLG Poe Pt uae Fs its ‘at fh , Ni ma? } my 4] pieaat Wrap atte on © ose it si taktach sca wl iepe: ity say ¥ its dab a a teteiet 40 a4 ror Ade) pense bee by “yay iu ‘ tes » a hPa rited ian oat 4 oredr tat pidiguarat ; 4 etek dinye yah y dedere +4 ule ” + Ae bey Ay taeenieny eliteted RUT Tetele'a's? hale tebe he tet aster eet Tie. h ab 4) ss Baits Sain ASE AL eb Hatatpteten. os Hi; ": Ue Hehedeta ted aie ar tt 4%a Boral 3° ? TEP Re Daten) Neat eCH PLC: 5 EE Ped On y Pete UPL SSar 44 ur» : oy otras nsiatg eles ; a7 Vit 4d sisted ate" Aya) Oot ae ce hire w me) nyt } Tite } : “ ate 4 ) a ah aj oot ti te { pee " \Ne" * > . " rf ‘ Bath tradgitet ° ? aust ! 17 , a A i bp aA Ns ittab a ie aa $2 pelea teTeaet) vane teks Peli Ne noe wise pws ce mepina eee Perot sir iat Seah ee Fe ay Pedy 2etn! j ia} babes | ta, fie oe oo) ph ieee reg? nny ee oe ae} st) is ha bees Pidse welt atte raya) OOP cut is) +d bons \ed Lara, pekee Le tbe ty amare dy aR ta , ue hi peariee Abr Ye i He H Pa art aa 4 digi). Ha eine]. ta) 1) Uy betas is yree ie ee nee bebe een tere 9 9) vad ‘ Peserpisigdele eye) rile aye he ps 4 reer Alani qe eenet Mi cu 2 at chy WoleBary Pasko kbs 1304 shi tye spits Vode ie oo eds hs Teater eyee Pree 1 re ik e4 , hg Hie alee heed savtnbad 4 Bi gh Yio! pa eare eG a F pa sy ui peta i bi: a eee Sond bittpeht dpi si pio. Mae ope gts G8 any? Le bad ‘ @ pans 6: biti ie 4 ty aes odd eled ¢ ve i He ip debt pbup i tes rire Ny bg neues ; He penh bei einin ie) ty ale es eeresiinits Saab rier Le | « Ite non a ¥e3 9) say) es at Petts ” Lew te ‘bee yeths MG ee ie) ig re 85 yeb A wie gi aew A058" HOt Wivereere diaper rurise li. bbe Surrey aes Fyre be pal epee oh (PRE ANY he hs abt ane Fever nye » | 4) ih haga +4 i lodst she wi Vs cehavieadem mee Deiat ae caeibn Wh oteee Prerer st AS pepo tnt elt * 2). Gaede WU tee wm » Pet Ph AA ALL J | | fd Go a ‘ wad) | a) 7 at “lh AA Awvescueeeeey GAY ~~ A WS] LE ia © US Ny an ower) ; Oe Tea eee vive YW YVCGSS j vy Sav VET e) “WANG, vy Vweseoecess AY \ | Ne) et, te AS \es \ wae), ~7 WWVegieee ww fe DESI SUS NSS SIM jl Ned WIIG ~ I AGS lad A Y|\e, NY; LA oe we | yw MoO oS SR vuvuv | ve L | we y w4¥} esevy sous ied yy AEC CUMS Sed al w Siw “dA Wye oY wivainiarccvonant k Weieneensest tS WS WV SEC ry uous MAS vl Ue Vy 1S | WAS ESS Si MoUs IA GUY eve MI UU UE yy eeWeeuvove CRANES ssc tease or You vunasvesvoss Wevvvelluysuuuoryyy Vive eres REL od Pe be wt Wee: ISSN GNA! te LI MtAMMOIIYYSSY Vig Fn Va" uv PISS CS very ye weve ve Wiees' MUG IGS SII Y 0 ) \e\e ANA Sa tenaneteerrnecee Ye weeuey \ Le ae WA UY III Ow | h\ed \ Wy \ YI \e YOUVE ore | wr hd) Suewece ad vy \ ) \ } \_4 1 Nea | | \ i i\ | 4 ! | } ) | | aa | A) i It } f \ Ai | i } | | ji | { } i he ef 4 } } f | ii i} \ | I | | | | q | \ SIA y “aneiete oor weyyyyy Woy ieee Wuyi eeveeve \us ee a ag ag wuY vane ( Ve VOWS voces wy vreayuooanoelee ee, Serco dee acccw-sorcoecy iva avE VO ON PPO NNN A) | LY, veyruY Vv ew vy mana DIAS Viiv Wii Me CUM MALS vy vyveriweys | MM cd uly Wi wwe este ~ cS | L LY SSSI Nd AAA dS j we | Ak gy" OT de MESSE LA seca eensee tN Wah ee LUE e SSSI IVa © ee eae Nd ele el Od epveustUeTt SY Be Gr Uy MIMIAWAA wie Vue eeryvuvuvuy¥™ MUE ES oe be dat Nh hhh Oh ey Fd ed et ene jest ew’ New ww | ee | Nee |e GGG VS SS Newt Nees Ld S i Sis MINI AWN ha fhe AIG SS EEMEE whl 4 NN Ne |e “OO || Neve , | PAA I GO AAD) SSS ee CALS whl A ele =the hea ‘ ‘ ‘a ye , * or arf j , pe, my ‘se et Nf M w, we ; < t a ; rs . ry I yaa a ry 4 ‘ ¥ hy Ps i ' 4 7 = r, . ¢ a - ai ‘ —— \ ¥ . - i ‘ yee ee 3 e , ay ‘ 4 ~ . = 4 E 5 > by ~ ie = Fen e . , : ~ 2 vi & ANNALS OF THE SOUTH AFRICAN MUSEUM VOLUME XXXII ANNALS OF THE SOUTH AFRICAN MUSEUM VODEOME XX KIT PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM BY NEILL AND CO., LTD., 212 CAUSEWAYSIDE, EDINBURGH. 1935-1940. TRUSTEES OF THE SOUTH AFRICAN MUSEUM. J. G. VAN DER Horst, Esa. W. J. THorNeE, Esq. Prof. D. L. ScHoutz, D.Sc. FERDINAND Bosmay, Esa. Prof. M. R. Drennan, M.A., M.B., Ch.B., F.R.C.S., F.R.S.S.Afr. SCIENTIFIC STAFF OF THE SOUTH AFRICAN MUSEUM. Epwin LEonarp Git, D.Sc., Director. KepreL Harcourt BarnarpD, M.A., D.Sc., F.L.S., Assistant Director; in Charge of Fish and Marine Invertebrates. ALBERT JOHN HeEsszE, B.Sc., Ph.D., Assistant in charge of the Entomological Department. Lizuwe Dirk Boonstra, D.Sc., Assistant in Charge of the Palaeontological Department. Miss G. Joyce Lewis, B.A., Assistant in Charge of the Botanical Department. A. J. H. Goopwin, M.A., Honorary Keeper of the Ethnological and Archaeological Collections. Miss E. Marcarert SHaw, B.A., Assistant in Charge of the Ethnological Collections. LIST OF CONTRIBUTORS. K. H. BARnarp. Further Notes on South African Marine Fishes . Notes on Dragon-flies (Odonata) of the S.W. Cape, with Deseliptions of the Nymphs, and of New Species Contributions to the Crustacean Fauna of South ance XI. mites Additions to the Tanaidacea, Isopoda, and Amphipoda, together with Keys for the Identification of the hitherto recorded Marine and Fresh-water Species . Additional Records and Desernubus of New Snecies of South Athan Alder-flies, May-flies, Caddis-flies, Stone-flies, and Dragon-flies D. CARTER. See G. C. Shortridge. B. DE MEILLON. A New Jenkinshelea (Dipt. Ceratopogonidae) from Southern Rhodesia C. G. C. Dickson. Notes on the Early Stages of Phasis felthami Trim., a Lycaenid Butterfly from the Cape Peninsula, and a List of some recently determined Food-plants of some other South African Butterflies. S. G. HEpINc. A New Dendrochirote Holothurian from South Africa, with some Remarks on the Rhopalodinidae Ke EE ny, New Acridiidae from South Africa R. F. LAWRENCE. A New Species of Plume-footed Mite from South Africa The Genus Selenops (Araneae) in South Africa G. A. MavRoMoustTaKIs. Report of some Anthidiine Bees (Apoidea) in the South African Museum . Descriptions of New African Anthidiinae (Apoidea) T. MORTENSEN. A New Giant Sea-star, Withrodia gigas n. sp., from South Africa J. R. Norman. The European and South African Sea Breams of the Genus Spondy- liosoma and Related Genera; with Notes on JDichistius and Tripterodon PAGE 4] 169 381 609 261 545 31 List of Contributors. Vil W. L. Scumirt. PAGE A New Species of Emerita (Crustacea) from South Africa. , 5 20 G. C. SHORTRIDGE. A New Genus and New Species and Subspecies of Mammals from Little Namaqualand and the North West Cape Province; and a New Subspecies of Gerbillus paeba from the Eastern Cape Province . 281 J. R. LE B. Tomi. Reports on the Marine Mollusca in the Collections of the South African Museum. X. Family Verticordiidae . : ‘ ; 3 see C. J. VAN DER Horst. On a New South African Species of Balanoglossus, and a Comparison between it and Balanoglossus capensis 3 : ; x, 69 The Enteropneusta from Inyack Island, Delagoa Bay : : . 293 K. W. VERHOEFF. South African Geophilomorpha (Myriopoda) : 5 : : Sa PA! A New Genus of Colobognatha from the Cape Peninsula : So NE New Colobognatha from South Africa ; ‘ i : « L2b LIST OF NEW FAMILY AND GENERIC NAMES INTRODUCED IN THIS VOLUME. PAGE Aetiasis n. g. (Pisces, Lutianidae), Barnard, 1937 : : p ‘ : 59 Bureniidae n. f. (Myriopoda), Verhoeff, 1937. : : : E Le) ie Cryptochloris n. g. (Mammalia, Chrysochloridae), Shortridge, 1938 . 284 Cylichnogaster n. g. (Myriopoda, Bureniidae), Verhoeff, 1937 : : ied 1:7) Frontifissia n. g. (Insecta, Acridiidae), Key, 1937. ; : : . 144 Hirudisomidae n. f. (Myriopoda), Verhoeff, 1937 : : : : -- Ae Polyamblyodon n. g. (Pisces, Sparidae), Norman, 1935 ; : : so hel Rhopalodinopsis n. g. (Echinodermata, Rhopalodinidae), Heding, 1937 = Vell Rhynchomecogaster n. g. (Myriopoda, Bureniidae), Verhoeff, 1937. 3 E28 Stathmos n. g. (Crustacea, Sphaeromidae), Barnard, 1940 . ‘ : 2 425 DATE LIST OF PLATES. . Mithrodia gigas n. sp. . Pachymetopon canescens n. sp. . Emerita austroafricana n. sp. Emerita emeritus (Linn.). . Rhopalodinopsis capensis n. g., n. sp. . Rhopalodinopsis capensis n. g., n. sp. . Cetorhinus maximus (Gun.). . Cetorhinus maximus (Gun.). Quinquarius capensis (C. & V.). . South African Geophilomorpha. . South African Geophilomorpha. . Cylichnogaster lawrencei n. g., n. sp. . South African Colobognatha. . Betiscoides meridionalis Sjost. . Betiscoides sjostedti n. sp. . Betiscoides parva n. sp. . Frontifissia elegans n. g., n. sp. . Pyrgomorphella species. . Phasis felthami Trim. . Phasis felthami Trim. OF ISSUE OF THE PARTS. Part 1, June 1935. Part 2, March 1937. Part 3, July 1937. Part 4, July 1938. Part 5, February 1940. Part 6, August 1940. Vili Acentrella Achilophilus Adenophlebia Adicella Aega Aeschna . Aetiasis Afronurus Alepocephalus Allocnemis Allorchestes Ampelisca Ampithoé Anax : Anguilla . Anilocra . Anthidiellum Anthidium Aora ‘ Apanthura Aphanicerca Aphanicercella Aphilodon Aprionyx Apseudes . Arcturella Argyrobothrus . Artopoles Aspidopleres Astronesthes Austrocaenis Austrocloeon Austrofilius Baetis Balanoglossus Balistes Barbarochthon Barnardichthys Bathyamaryllis Bathyergus INDEX OF GENERA. 265, 69, PAGE 625 102 630 652 401 230 59 638 46 203 475 44] 480 233 49 404 266 663 478 387 657 Beryx Betiscoides Borostomias Burenia Calliopiella Cantharus Cantharusa Cantherines Caranthus Castanophlebia Centroptilum Ceradocus Cerapus Ceratogomphus Ceriagrion Cetorhinus Cheimacheramus Cheiriphotis Chelidonichthys Cheumatopsyche Chimarrha Chironesimus Chlorocypha Chlorolestes Chloroniella Chrysochloris Cirolana Cleantis Clinus Cloeon Corophium Coryphaesopia Crocothemis Crunobiodes Crunoeciella Cryptochloris Cylichnogaster Cymadusa Cymodoce Cyphocaris Cyproidea 135 Desmonemoura Dianthidium Dichistius Dipseudopsis Dipterodon Disparoneura Dynamenella Dyschimus Eatonica Echiostoma EKenomus . Elasmopus Elassoneuria Elattoneura Emerita Enallagma Engidotea Ephemerellina . Eriopisella Erythraeus Ktelis : EKuleptorhamphus Euryporeia Eurydice . Eurystheus Eurythenes EKurytion . Eusalpa EKustomias Exanthura Excirolana Exeuthyplocia . Exosphaeroma . Frontifissia Fuyangia Gallea Gammarus Gerbillus Gitanopsis Gliriscus . Glossobalanus Glossocephalus . Glyptidotea Gnathia Gnatholana Gobius Goerodes . Gonorhynchus . Gonostoma Index of Genera. PAGE 658 265° 21 654 21 205 418 640 144 443 462 290 444 286 316 485 429 382 391 62 639 46 46 Grandidierella . Grubia ; Gymnocrotaphus Halicordia Haliophasma Haplostomias Harriotta Helothemis Homilia Hoplopleon Hyale ; i “ Hydropsychodes Idiacanthias Iphimedia Ischnura . Ischyrocerus Jaera Jaeropsis . Jenkinshelea Katius Lakota Lampanyctus Lembos Leptecho . Leptocerina Leptocerus Leptonema Leptosialis Lestes Inbellago . Liljeborgia Lithogloea Loweina . Lutianus Lysianassa Macromia Macronema Macropisthopus Macroscelides PAGE 478 480 20 23 382 47 46 252 649 444 475 653 46 450 222 48] 434 434 261 440 44] 50 478 647 647 643 654 613 187 183 450 - _ 635 50 58 439 234 654 480 283 Maera Maurolicus Megaluropus Melamphaes Melita Mesogomphus Mesoschendyla . Microlysias Mithrodia Muraenichthys . Myspoleo . Neoscorpis Neostomias Nerocila . Ochlesis Oecetis Opsodentex Orchomenella Orthetrum Oxycephalus Pachymetopon . Paduniella Palinnotus Palpopleura Panoploea Pantala Paramoera Paramelita Parelasmopus Parharpinia Parhyale Paridotea Parisocladus Parorchestia Petromyscus Petroplax Petrothrincus Phasis Philacroterium . Photis Phreatoicus Phronima Phronimella Phylluropus Platychauliodes Platymops Podocerus Poemys N Index of Genera. 444 248 . 450 260, 660 452 462 461 442 472 428 418 473 288 641 643 545 99 479 434 484 484 453 610 282 483 287 Polyamblyodon Polycheria Polygonarea Polymitarcys Potamoryza Pontharpinia Pontogeloides Presba Proboloides Proteles Protophoxus Pseudaeginella Pseudaesopia Pseudagrion Pseudocloeon Pseudoleptocerus Pseudomacromia Pseudosphaeroma Ptychodera Pyrgomorphella Quinquarius Rhabdosoma Rhoizema : Rhopalodinopsis Rhynchomecogaster . Saccoglossus Scotophilus Seyracepon Selenops . . Simocantharus . Sphaeroma Spondyliosoma . Stathmos Stegostoma Stenetrium Stomacontion Streetsia . Sympetrum Synidotea Syscenus Taeniochauliodes Talitriator Talitroides 464 464 xXli Index of Genera. PAGE PAGE Talitrus . : : ‘ . 463 Vv Talorchestia . é ; . 468 Tanais . : , : » B83 Vibtha\ « : : : . 483 Taterona . . ; é 28a : Temnophlias . : : . 446 W Thylakion : : , Pe 13)7 | . Trachinotus . : ; . 54 | Willeyia . 5 ‘ : . 206 Tricorythus : : 3 ~/ 626 Tripterodon. c 3 ti 4 Ne Trithemis ‘ : : “2265 Tylos ; : : : . 438 | Yarella . ; ‘ Os U Z Urda : é é : _" 391 | Zuzara: . 2 ; . 418 ANNALS.” > SOUTH AFRICAN MUSEUM PART I, containing :— 1. A new Giant Sea-star, Mithrodia gigas n. sp., from South Africa.—By Tu. Mortensen, University Zoological Museum, Copenhagen. (With Plate I and 1 Text- figure.) 2. The European and South African Sea Breams of the Genus | Spondyliosoma and Related Genera; with Notes on Dichistius and Tripterodon.—By J. R. Norman, —- ae a Department of Zoology, British Museum (Natural History). (With Plate II and 6 Text-figures.) ee f f, ; (i. * fy 4 1] 2 ‘_ (ee PEL 13: a3 VON EGS Ac ly’ k lian { ISSUED JUNE 1935. PRICE 2s. 6d. ~ PRINTED FOR THE . TRUSTEES OF THE SOUTH AFRICAN MUSEUM,. CAPE .TOWN BY NEILL AND GO., LTD., 212 CAUSEWAYSIDE, EDINBURGH ANNALS OF THE SOUTH AFRICAN MUSEUM VOLUME XXXII. 1. A new Giant Sea-star, Mithrodia gigas n. sp., from South Africa.— By Tu. Mortensen, University Zoological Museum, Copenhagen. (With Plate I and 1 Text-figure.) SOME months ago I received from Dr. K. H. Barnard, South African Museum, Cape Town, photographs of a very large sea-star which he had received from Mr. Bell Marley, Durban, who suggested that it might be a Mithrodia. The photographs left no doubt of its being really a Muthrodia, and a new species, markedly different from the three species of that genus known till now. Dr. Barnard then sent me the specimen, asking me to describe it. On account of its being by far the largest of any specimen of Mithrodia hitherto recorded, and, indeed, one of the largest sea-stars known, I name it Mithrodia gigas nu. sp. Rays five, of somewhat unequal length. The longest ray is 330 mm., another ray is 310 mm., and the three others 300 mm. long. The total diameter thus is some 600 mm. Diameter of disk c. 90 mm. The rays of the dried specimen are c. 50-60 mm. broad, but clearly they are not so broad in life. The specimen evidently has been dried directly from the sea, not first fixed in alcohol or formalin, and therefore the dorsal wall of the rays has sunk in and the rays become flattened. Presumably, the rays will be cylindrical in life as in the other Mithrodia species, and their diameter not more than c. 30-40 mm. They are scarcely constricted at the base. The interbrachial angles appear to be somewhat rounded. The disk is rather closely set with low, almost spherical knobs, the largest of which slightly exceed 2 mm. in diameter; on their VOr. XXxil, PART I. i! 2 Annals of the South African Museum. upper surface they are densely set with low, rounded prominences, producing a somewhat scaly appearance. These knobs continue a little more sparsely on the dorsal side of the rays in their basal half, becoming then much more scarce in the distal part, disappearing almost completely, until at the tip of the rays they are replaced by large, spherical knobs of up to 5 mm. diameter (fig. 1). These large knobs are likewise covered with small, scaly prominences. There are no large spines on the dorsal side of disk and rays. Fic. 1.—Withrodia gigas, n. sp. Tip of ray, natural size. The dorsal skeleton consists of low ridges arranged so as to form more or less distinct stars, in the centre of which the knobs are placed. The whole dorsal side has evidently been covered by a thick skin, in which are imbedded numerous small spinelets, the ridges of the skeleton thus having a very finely granular, almost velvety appearance. Also the larger knobs have been covered with such a rather thick skin. On the sides of the rays the knobs increase in number, but no larger spines are found here either, and only close to the ambulacral furrow the knobs gradually grow a little larger, passing very evenly into the outer adambulacral spines. The ambulacral, or furrow spines are only four in number; the q | ; | : : : : | : | —— ss OO A New Giant Sea-star, Mithrodia gigas n. sp., from South Africa. 3 two median ones are somewhat stouter and longer (4-5 mm.) than the two lateral ones, the latter being, however, not rudimentary. The inner adambulacral spines are 5-6 mm. long, stout, club-shaped, and of the usual scaly-granular appearance, the side turning towards the ambulacral furrow being partly more smooth. They form a perfectly regular series, one to each adambulacral plate. The outer series of adambulacral spines is, on the whole, very inconspicuous, these spines passing very gradually into the general covering of the oral side of the rays. No ventro-lateral series of spines or larger knobs observable. At the tip of the rays the knobs grow larger, globular, but not so large as those of the dorsal side. The inner adambulacral spines do not thus transform. As on the dorsal side a thick skin covers the body skeleton and the knobs and spines also on the ventral side. Even the ambulacral spines are covered by this thick skin, so that—on the present, dried specimen—the number of these spines can only be ascertained on cleaning away the skin (by means of hypochlorite of sodium). A very extraordinary fact is that in this skin covering the ambulacral spines are imbedded a number of small, sharp granules or spinelets, so that even the ambulacral spines get a scaly appearance like the larger spines and knobs. The oral edges, or jaws, are rounded; there is no special oral armature, only the usual ambulacral and adambulacral spines, which are not larger here than along the ambulacral furrow. The madreporite is small, partly covered by the knobs. It is a little nearer the edge than the centre of the disk. The colour of the live specimen was, according to Mr. Bell Marley, purplish pink, the tip of the rays more cinnamon red; below pale yellowish; the ambulacral feet white. The dried specimen has lost the colour completely. The specimen was caught on a fish hook, having taken to the bait, fouling itself in the line. It was taken off Point Morgan, Hast London, on fishing grounds in 25-30 fathoms, on stony ground. This is another highly interesting addition to the rich endemic fauna of the seas along the South African Hast Coast. Mr. Bell Marley is to be sincerely congratulated on having secured this, and so many other treasures of that fauna. But it is not enough to leave it to the casual catches on fish hooks or by trawlers (a good deal of interesting forms have been saved by Captain Pace). A real scientific investigation of this sea is greatly needed. It will be sure to yield most important results. The little we know already shows t Annals of the South African Museum. this area to be one of the richest and most characteristic zoogeo- graphical regions of the world. _ The present species differs strikingly from the three other species of the genus Mithrodia hitherto known, M. clavigera (Lamarck) of the Indo-Pacific, M. bradleyz Verrill of the American West Coast and Hawaii, and M. fisheri Holly of Hawaii. It would seem the nearest related to M. bradleyi, which also lacks the large spines on the dorsal surface (in the adult); but it differs from all of them in the total absence of long spines, in having only four ambulacral spines (in the other species c. 6-12) and these even covered with granular spinelets imbedded and lying loosely in the thick skin enclosing the ambulacral spines. Finally, the large spherical knobs on the point of the rays is a quite unique feature. - partes s = 5 3 ~ Q > &S x yy ~ al 5 Ann, 8. Afr. Mus., Vol. XXXII. 2. The European and South African Sea Breams of the Genus Spondylio- soma and Related Genera; with Notes on Dichistius and Trip- terodon.—By J. R. Norman, Department of Zoology, British Museum (Natural History). (With Plate II and 6 Text-figures.) HAVING encountered some difficulty in identifying one or two speci- mens of Sea Breams allied to Spondyliosoma collected by the “Discovery” Expedition at the Cape, I sought the opinion of Dr. K. H. Barnard, Assistant Director of the South African Museum. He was good enough to suggest that I should undertake a revision of these fishes, and very kindly offered, not only to submit all the material in the South African Museum, including the types of the species described by Gilchrist and Thompson, but also to hand over to me some notes he had already prepared with a view to eventual publication. I take this opportunity of offering Dr. Barnard my sincerest thanks for his generosity, and of expressing my appreciation to the authorities of the South African Museum for permission to study this valuable material, which, with the specimens in the collection of the British Museum (Natural History), has enabled me to clear up a number of points concerning this group of fishes. My thanks are also due to Mr. A. Fraser-Brunner, for his assistance in procuring for me a series of specimens of Spondyliosoma cantharus from the Mediterranean and from the British coast, and for several helpful suggestions; and to Dr. C. Tate Regan for valuable advice given during the progress of the work. The names adopted for the new species of Pachymetopon are those which appear in Dr. Barnard’s MSS. SPONDYLIOSOMA, Cantor. Cantharus (non Bolten, 1798; Montfort, 1808), Cuvier, 1817, R. Anim., vol. 11, p. 278 [Sparus cantharus, Linnaeus]. Spondyliosoma, Cantor, 1850, J. Asiat. Soc. Bengal, vol. xvi (1849), p- 1032 [Sparus cantharus, Linnaeus—a substitute for Cantharus, preoccupied]. Caranthus, Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, p. 720 [Sparus cantharus, Linnaeus—a substitute for Cantharus, preoccupied]. ccpaply art ei 6 Annals of the South African Museum. Cantharusa, Strand, 1928, Arch. Naturgesch., vol. xcii, A. 8, p. 54 [Sparus cantharus, Linnaeus—a substitute for Cantharus, pre- occupied]. . Spondyliosoma (part), Fowler, 1933, Bull. U.S. Nat. Mus., 100 (12), p. 182. Body ovate covered with rather small, finely ctenoid scales; many scales, especially on hinder part of body, lobate in centre of free margin; tubules of lateral line short, bifurcated posteriorly. Prae- orbital narrow, its lower edge generally more or less notched. Posterior nostril slit-like. Teeth in front of jaws in broad bands, that become narrower laterally, arranged in 5 to 7 irregular rows; teeth of the outermost row largest, compressed, narrow, lanceolate; no canines; those of the innermost row mostly obtuse, molariform. Cheek and opercular bones scaled; interorbital region and flange of praeoperculum naked. Dorsal with 11 spines; soft rays naked, but with a low scaly sheath at the base of the fin. Anal with 3 spines. Two species: one from the Mediterranean and eastern Atlantic; the other from the coast of south-east Africa, Madagascar, and Mauritius. Synopsis of the Species. I. 64 to 72 scales in lateral line, 8 to 10 from origin of dorsal fin to lateral line 1. cantharus. II. 80 to 92 scales in lateral line, 14 or 15 from origin of dorsal fin to lateral line 2. emarginatum. 1. Spondylhosoma cantharus (Linnaeus). Sparus cantharus, Linnaeus, 1758, Syst. Nat., ed. 10, p. 280.* Sparus brama, Bloch, 1791, Nat. ausl. Fische, vol. v, p. 77. Sparus lineatus, Montagu, 1818, Mem. Werner, N.H. Soc., vol. 1 (2), p- 451, pl. xxi. Sparus vetula, Couch, 1823, Tr. Linn. Soc. London, vol. xiv (1), p. 79. Cantharus tanuda, Risso, 1826, H.N. Europe, vol. 11, p. 366. Pagrus lineatus, Fleming, 1828, Hist. Brit. Anim., p. 211. Cantharus vulgaris, Cuvier and Valenciennes, 1830, H.N. Poiss., vol. vi, Dao loy plrelxe Cantharus brama, Cuvier and Valenciennes, 1830, t.c. p. 328; Giinther, 1859, Cat. Fish., vol. i, p. 416; Moreau, 1881, H.N. Poiss. France, vol. ili, p. 52; Carus, 1889-93, Prodr. Faun. Medit., vol. 11, p. 626. * T have not attempted to give a full list of references under each name, but have merely indicated the principal synonyms and combinations of generic and trivial names. The European and South African Sea Breams. 7 . ’ of Pb et eae eeeee fee Cie LY. MAA) ' . : ei. yt : Vitae es My \ Jd brreece y) QP precece fd den te Sage 2 989-0 eles a teas : Vege aie veo eae cpener we LNs be ar St re, we AOS Dy ORL CEP) Tyee pe . A eit i SLUet ie ta et yee) . AOR CET ee yan fear ye SUP Nee ey tyne: artes ete ee pe WAS) Ie ae ee RS en gee , *. ‘oes Beg) on Sea Behe COMO IIL”. HPS e8\*e 8cR le] ee) Vie) Ol8 swine thea ee o « 2.0. . OG 8 Cia ae Bk Re) 8 8 Row) 8 ee 8 -e FP oe eased esponnes eee sooonese ee e oto wee aie Wear ee ee ee Fic. 1s.—Spondyliosoma cantharus. x}. 9. 8 Annals of the South African Museum. Cantharus orbicularis, Cuvier and Valenciennes, 1830, t.c. p. 331; Ginther, 1859, f.c. p. 416; Morean, 1881, te. p. 52,9 Carne 1889-93, t.c. p. 626. Cantharus griseus, Cuvier and Valenciennes, 1830, t.c. p. 8333; Lowe, 1839, Tr. Zool. Soc. London, vol. 1, p. 178; Yarrell, 1859, Brit. Fish., ed. 3, vol. 1, p. 165, fig.; Moreau, 1881, ¢.c. p. 49. *Cantharus senegalensis, Cuvier and Valenciennes, 1830, t.c. p. 337. Cantharus lineatus, Thompson, 1846, Ann. Mag. N.H. (2), vol. xviii, p. 313; Giinther, 1859, t.c. p. 413; Steindachner, 1867, Sitzungs- ber. Akad. Wien, vol. lvi (1), p. 649; Day, 1880-84, Fish. Britain, p. 26, pl. ix; Carus, 1889-93, t.c. p. 625; Smitt, 1893, Scand. Fish., vol. i, p. 54, fig. 14; Pellegrin, 1914, Ann. Inst. océan., vol. vi, p. 51; Le Gall, 1931, in Joubin, Faune Ichth. Atlant. Nord, No. vi, fig. Cantharus linner, Malm, 1877, Goteborgs Bohus. Faun., pp. 97, 384. Spondyliosoma cantharus, Jordan and Fesler, 1893, Rep. U.S. Com. Fish., vol. xvi (1889-91), p. 530; Buen, 1926, Cat. ict. Médit. Espah. Marruecos, p. 141. Spondyliosoma orbiculare, Jordan and Fesler, 1893, t.c. p. 530. Spondyliosoma brama, Fage, 1907, Arch. Zool. exp. gén. (4), vol. vii, p. 73. Caranthus lineatus, Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, p. 722. Depth of body 2 to 22 in the length, length of head 3 to 33. Profile more or less straight to occiput, thence moderately convex to origin of dorsal fin. Snout as long as or longer than eye, diameter of which is 3 (young) to 44 in length of head, 1 to 12 in interorbital width, and twice or nearly twice depth of praeorbital. Lower edge of praeorbital usually more or less notched, but sometimes nearly straight. 38 to 48 teeth in outer row of upper jaw, 42 to 52 in outer row of lower jaw. 14 to 16 gill-rakers on lower part of anterior arch. 6 or 7 series of scales on cheek; 64 to 72 scales in lateral line, 8 to 10 from origin of dorsal fin to lateral line; scales on upper surface of head extending forward to a point above middle of eye. Dorsal XI 12; 4th to 6th spines longest, length 13 to 24 in that of head; first soft ray not or only very little longer than last spine. Anal III 10; 2nd spine a little shorter than 3rd and 12 to more than twice as long as first; 3rd spine about 2 length of longest dorsal spine. Pectoral with 15 or 16 (occasionally 17) rays, extending to or a little beyond vent or not quite as far, length equal to or rather longer than that of head. Pelvic not reaching vent. Coloration variable; generally silvery grey, with numerous narrow, dark The European and South African Sea Breams. 9 longitudinal lines on the side, mainly below the lateral line; male with about 6 rather indistinct dark cross-bars on the sides, which are more clearly defined in the young; sometimes some narrow bars between the broader ones; male sometimes with irregular dark patches on head; dorsal and anal fins greyish in the female, more or less spotted and blotched with dusky; in the male these fins are much darker, the dorsal sometimes being quite black; caudal variegated in young, greyish in the adult female, more or less blackish in the male; pectoral pale in the female, dusky in the male; pelvic blackish or brownish in both sexes. Described from numerous examples, 100 to 390 mm. in total length, from the English Channel, Lisbon, Majorca, Monaco, Naples, Malta, Propontis, Madeira, Mogador, and the Cape Verde Islands. Hab. : Coasts of south-western Europe, from southern Scandinavia to the eastern Mediterranean; coasts of northern and western Africa, southwards to Angola. The marked sexual dimorphism in this species does not appear to have been previously recognised, but there is little doubt that the two forms respectively named S. cantharus (=lineatus, griseus) and S. orbicularis, both of which have received distinct local names in parts of the Mediterranean, represent the male and female of the same species. In addition to the differences in coloration, which are very marked in the living fish, comparison of specimens of equal size shows that the females have a constantly deeper body, more oblique anterior profile, and a somewhat shorter pectoral fin than the males. Mr. Fraser-Brunner informs me that his experience with Black Bream on the south coast of England suggests that the two sexes shoal separately. 2. Spondyliosoma emarginatum (Cuvier and Valenciennes). Cantharus emarginatus, Cuvier and Valenciennes, 1830, H.N. Poiss., vol. vi, p. 338; Giimther, 1859, Cat. Fish., vol. 1, p. 416; Kner, 1865, Reise “‘ Novara,” Zool., vol. i, 5. Fische, p. 73. Cantharus microlepis, Gilchrist and Thompson, 1909, Ann. 8. Afr. Mus., vol. vi, p. 231. Scatharus graecus, Clark, 1915, Sci. Res. “Scotia,” vol. iv, p. 396. Pagellus microlepis, Regan, 1921, Ann. Mag. Nat. Hist. (9), vol. vu, p: 419. Caranthus emarginatus, Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, pate2. Caranthus microlepis, Barnard, 1927, t.c. p. 723. ys tA 10 Annals of the South African Museum. Pachymetopon grande, Fowler, 1929, Ann. Natal Mus., vol. vi, p. 259; Fowler, 1933, Bull. U.S. Nat. Mus., 100 (12), p. 214. Spondyliosoma microlepis, Fowler, 1933, t.c. p. 183. Spondyliosoma emarginata, Fowler, 1933, t.c. p. 183. Depth of body 2} to 22 in the length, length of head 3 to 34. Profile in smaller specimens nearly evenly convex, but with a slight emargination above eyes, in larger specimens nearly straight to above eyes, thence convex to origin of dorsal. Snout as long as or a little shorter than eye, diameter of which is 3 (young) to nearly 4 in length of head, 1 to 14 in interorbital width, and 2 to 22 times depth of . iy en i ers 3 aera TP PIO lula telat ars rea eila fee e OOD = CAO ORAS Coo = . OCC 9 Sh TORSO eras LS evar POF OTe RO OE, Pe, QO e . . . elie SeCisleliny a ae . . a. ° a acd AOS ODO WOnG aS Ss 2 . . ste a? of re. ae : . evece ) .6 ae SO eZ ° e wes e oe +e 5s a ’ pa ies arin eis 13 . , ° . ’ A e site . . ’ e . *. . «, Me. Caeser . OB CUA aps bpd i ia OCS eaeeke) ° a es avae ’ 2 eis gue ee® ge . -® eee eee ees? oo Fie. 2.—Spondyliosoma emarginatum. x About vilro . 6: praeorbital. Lower edge of praeorbital with a notch which is much shallower in the young. 38 to 50 teeth in outer row of upper jaw, 42 to 54 in outer row of lower jaw. 15 to 17 gill-rakers on lower part of anterior arch. 8 series of scales on cheek; 80 to 92 scales in lateral line, 14 or 15 between origin of dorsal fin and lateral line; scales on upper surface of head extending forward to a point behind middle of eye. Dorsal XI 11-13; 4th or 5th spines longest, length 14 to nearly 3 (generally about twice) in that of head; first soft ray a little longer than last spine. Anal III 10; 2nd spine shorter than Srd and twice or more than twice as long as first; 3rd spine 2 to 3 length of longest dorsal spine. Pectoral with 15 or 16 rays, extending to a little beyond vent, length about equal to that of head. Pelvic The European and South African Sea Breams. i not or scarcely reaching vent. Greyish or brownish, with traces of dark longitudinal lines on the sides; male with a more or less distinct dark bar between the eyes, and with a dark patch on the suborbitals and on the flange of the praeoperculum; dorsal, anal, and pelvic fins blackish in the male, greyish or dusky in the female; male (?) with a dark spot in the axil of the pectoral. Described from 13 examples, 105 to 300 mm. in total length, from Table Bay, Simon’s Bay, Cape St. Blaize, False Bay, and coast of Natal; including the types of Cantharus microlepis and Pagellus microlepis. Hab.: Coast of south-east Africa, from Saldanha Bay to Natal; Mauritius; Madagascar. This species is clearly related to S. cantharus, but the scales are smaller and those on the posterior part of the body more distinctly lobate or even pointed in the centre of the free margin. The two sexes appear to differ somewhat in coloration, but, owing to the small amount of material and the difficulty of sexing many of the specimens, I am unable to say whether there are other differences as in the preceding species. It seems certain that S. microlepis is identical with S. emarginatus: in 4 examples of the former, including the types of Regan’s and Gilchrist and Thompson’s species, I count 83 to 92 scales in the lateral line, and in 9 examples of the latter, 80 to 86. PACHYMETOPON, Giinther. Pachymetopon, Giinther, 1859, Cat. Fish., vol. i, p. 424 [Pachymetopon grande, Giinther|. Caranthus (part), Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, p. 720. Simocantharus, Fowler, 1933, Bull. U.S. Nat. Mus., 100 (12), pp. 182, 185 [Cantharus aeneus, Gilchrist and Thompson]. Spondyliosoma (part), Fowler, 1933, t.c. p. 182. Close to Spondyliosoma, but with a deeper praeorbital, the lower edge of which is nearly straight. Teeth nearly all compressed, in 4 or 5 (occasionally 6) rows, those of the outermost row broader and fewer than in Spondyliosoma, those of the innermost row not molari- form. Dorsal with 10 or 11 spines; soft dorsal and anal densely scaled on basal third of fin, but without sheath. Five species from South Africa. 12 Annals of the South African Museum. Synopsis of the Species. I. 26 to 36 teeth in outer row of upper, 36 to 44 in outer row of lower jaw; depth of praeorbital usually less than diameter of eye; 13 to 16 gill-rakers on lower part of anterior arch. A. Flange of praeoperculum not scaled; 60 to 66 scales in lateral line; scales on upper surface of head extending forward to a point behind level of middle of eye; pectoral as long as or shorter than head. 1. (28) 30 to 36 teeth in outer row of upper, 40 to 44 in outer row of lower jaw; Ist dorsal spine 4 toZeye . . LL. bleehs: 2. 26 to 28 teeth in outer row of upper, 36 in outer row of lower jaw; lst dorsal spine about #eye . : 2. canescens. B. Flange of praeoperculum partly scaled; 80 to 86 scales in lateral line; scales on upper surface of head extending forward to above level of anterior part of eye; pectoral 11 to 14 times as long as head 3. aeneum. II. 18 to 22 teeth in outer row of upper, 22 in outer row of lower jaw; depth of praeorbital about equal to diameter of eye; 10 or 11 gill-rakers on lower part of anterior arch. A. Depth about 24 in length; anal III 10; pectoral nearly 1} times as long as head : ; : : : ‘ . 4. grande. B. Depth about 13 in length; anal III 11; pectoral about 14 times as long as head. : : : L ; : 5. glaucum. 1. Pachymetopon blocht (Cuvier and Valenciennes). ¢Sparus brama, Bloch, 1791, Nat. ausl. Fische, pl. cclxix. Cantharus blochi, Cuvier and Valenciennes, 1830, H.N. Poiss., vol. vi, p. 339; Giinther, 1859, Cat. Fish., vol. i, p. 416; Kner, 1865, Reise “‘ Novara,” Zool., vol. 1, 5. Fische, p. 74. Cantharus castelnaui, Bleeker, 1860, Nat. Tijdschr. Ned. Ind., vol. xxi, p- 99. Caranthus blochi, Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, p. 721. Spondyliosoma blochi, Fowler, 1933, Bull. U.S. Nat. Mus., 100 (12), pe caz Depth of body 24 to 24 in the length, length of head 3 to 3}. Profile more or less evenly convex from snout to origin of dorsal. Snout as long as to twice as long as eye, diameter of which is 3 (young) to 54 in length of head, 1 to 24 in interorbital width, and 53 to 14 times depth of praeorbital. Lower edge of praeorbital without notch, the hinder part of the maxillary not concealed. (28) 30 to 36 teeth in outer row of upper jaw, 40 to 44 in outer row of lower jaw: teeth of outer row considerably larger than those of succeeding inner rows, their apices reaching a much higher level. 13 or 14 gill-rakers on lower part of anterior arch. 9 series of scales on cheek; 60 to 66 a pe OSS FN ae end pes A> BERS ee ASE naa aioe AEA AEE eae ey 3S: Fa Wee oT 3 ERR S| Se A ae see \ hfe \ho SO. é Te ae \- $350 a i) Fic. 3.—Pachymetopon blochi. a4 ) AS es TA FPPEPEED eran * See ae ek a 14 Annals of the South African Museum. scales in lateral line, 9 or 10 from origin of dorsal fin to lateral line; scales on upper surface of head extending forward to a point behind level of middle of eye; flange of praeoperculum not scaled; scales on caudal fin not extending to its posterior margin. Dorsal X-XI 11-12; length of first spine 4 to % diameter of eye; 4th or 4th and 5th spines longest, length about 3 (sometimes 4) in that of head; first soft ray much longer than last spine. Anal III 10; 2nd spine shorter than 3rd and 14 to twice as long as first; 3rd spine 2 to 2 length of longest dorsal spine. Pectoral with 17 or 18 rays, extending to vent or not quite as far, length equal to or less than that of head. Pelvic not reaching vent. Uniformly greyish or brownish, sometimes with bronze or bluish shades, sometimes darker sometimes paler; usually paler or silvery below. Described from 10 examples, 85 to 450 mm. in total length, from Saldanha Bay, Table Bay, and False Bay. ‘Hab. : South-west Africa. 2. Pachymetopon canescens, sp. n. (Plate II.) Close to the preceding species. Depth of body 24 in the length, length of head 3}. Snout 14 times eye, diameter of which is 4 in length of head, 14 in interorbital width, and 12 times depth of praeorbital. Praeorbital nearly completely concealing the maxillary. 26 to 28 teeth in outer row of upper jaw, 36 in outer row of lower jaw; teeth of inner series larger than in P. blochi, the apices of the teeth in all the rows reaching the same or nearly the same level, at least in lower jaw. 14 gill-rakers on lower part of anterior arch. 8 or 9 series of scales on cheek; about 65 scales in lateral line, 10 from origin of dorsal fin to lateral line. Dorsal X-XI 10-11; length of first spine about 4 diameter of eye; 4th spine longest, length 3 in that of head. Anal II] 9-10; 2nd spine shorter than 3rd and about 12 times as long as first; 3rd spine £ length of longest dorsal spine. Pectoral with 17 rays, extending to vent, length about equal to that of head. Pale yellowish brown, with a small dark patch below the eye, another larger one on the operculum, and with irregular dark areas on upper parts of sides and on upper surface of caudal peduncle; a number of minute dark dots scattered over head and body, which below the lateral line tend to be arranged in longitudinal rows; similar dots on the rays of the fins. | Described from a single example (holotype), 275 mm. in total The European and South African Sea Breams. 15 length, believed to be from False Bay: this is the specimen mentioned by Barnard in his monograph of South African Marine Fishes (p. 721). A second specimen (263 mm.), from Kalk Bay, False Bay, is in the South African Museum. 3. Pachymetopon aeneum (Gilchrist and Thompson). Cantharus aeneus, Gilchrist and Thompson, 1908, Ann. 8. Afr. Mus., vol. vi, p. 166. Cantharus natalensis, Gilchrist and Thompson, 1908, t.c. p. 167. Cantharus sumus, Gilchrist and Thompson, 1909, t.c. p. 231. Caranthus aeneus, Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, p. 724. Spondyliosoma aenea, Fowler, 1933, Bull. U.S. Nat. Mus., 100 (12), p- 185. Depth of body 2} to 24 in the length, length of head 24 to 23. Profile more or less evenly convex from snout to origin of dorsal, but there is a fairly prominent bulge in front of the eyes, below which the snout is concave. Snout longer than eye, diameter of which is 34 (young) to 44 in length of head, 14 to 1# in interorbital width, and 14 to 14 times depth of praeorbital. Lower edge of praeorbital straight or very little concave, the maxillary not entirely concealed. . 30 to 36 teeth in outer row of upper jaw, 36 to 40 in outer row of lower jaw; teeth of inner rows more or less chisel-shaped, usually with a single median point. 15 or 16 gill-rakers on lower part of anterior arch. About 10 series of scales on cheek; 80 to 86 scales in lateral line, 10 to 12 from origin of dorsal fin to lateral line; scales on upper surface of head extending forward to a point above level of anterior part of eye; flange of praeoperculum scaled; scales on caudal fin extending nearly to its posterior margin. Dorsal XI 11; 4th or 4th and 5th spines longest, length 2 to 24 in that of head; first soft ray longer than last spine. Anal III 10; 2nd and 3rd spines subequal or 2nd a little longer and 14 to 13 times as long as first; 3rd spine about 2 length of longest dorsal spine. Pectoral with 16 or 17 rays, extending to or beyond vent, length 14 to 1} times that of head. Pelvic not or scarcely reaching vent. Greyish or brownish above, silvery below, with dark longitudinal lines, especially below the lateral line; dorsal, anal, pelvics, and sometimes hinder part of caudal fin blackish or violaceous. Described from 7 examples, 225 to 430 mm. in total length, in- cluding the holotype of the species and the types of Cantharus natalensis and C. simus. 7 a bx ny x STILE 7 Ted 7 ZL IS LLL . ». > . “\; 4 aD y) 9 yoy - Sie y 23753": Ph hes ; g 4, 2. Lcd =~ and e D ~F ge 5 ) } ) ‘ ) ). . aed x Fie. 4.—Pachymetopon aeneum. The European and South African Sea Breams. 17 Hab.: Coast of Natal, at certain seasons extending as far west- wards as False Bay. Fowler (1933) makes this species the type of a new subgenus, Simocantharus. 4. Pachymetopon grande, Giinther. Pachymetopon grande, Giinther, 1859, Cat. Fish., vol. i, p. 424; Giinther, 1886, Ann. Mag. Nat. Hist. (5), vol. xviii, p. 367; Barnard, 1927, Ann. 8. Afr. Mus., vol. xxi, p. 636. Pachymetopon guentherz, Steindachner, 1869, Sitzungsber. Akad. Wiss. Wien, vol. lx (1), p. 135. ¢Pachymetopon gibbosus, Pellegrin, 1914, Bull. Soc. zool. France, vol. xxxix, p. 264. Depth of body about 25 in the length, length of head about 4. Profile convex from snout to origin of dorsal, with a bulge in front of the eye, more prominent in the larger specimen. Snout longer than eye, diameter of which is 4 in length of head, 1? in interorbital width, and equal to depth of praeorbital. Lower edge of praeorbital a little concave, the maxillary not entirely concealed. Apices of teeth truncate, rounded, or with a single median point; 18 to 20 teeth in outer row of upper jaw, about 22 in outer row of lower jaw. 10 or 11 gill-rakers on lower part of anterior arch. 7 series of scales on cheek; about 80 scales in lateral line, 11 from origin of dorsal fin to lateral line; scales on upper surface of head extending forward to a point above level of an- terior part of eye; flange of praeoperculum not scaled; scales on caudal fin extending nearly to its hinder margin. Dorsal XI 11; 4th to 6th spines longest, length 2 to 2§ in that of head; first soft ray longer than last spine. Anal III 10; 2nd and 3rd spines subequal and 1} to 12 times as long as first; 3rd spine about 3 as long as longest dorsal spine. Pectoral with 17 rays, extending to or nearly to origin of anal, length nearly 14 times that of head. Pelvic scarcely reaching vent. Brownish, with a number of narrow, dark longitudinal lines on the side below the lateral line; spinous dorsal black; soft dorsal and anal black- ish or violaceous; pectorals, pelvics, and hinder part of caudal dusky. Described from 2 examples, 420 and 520 mm. in total length, including the holotype of the species, a stuffed skin, 520 mm. long. Hab.: Coast of Natal; Cape of Good Hope; Madagascar (?). Steindachner’s description of P. guentherz differs a little from the above, but, as he does not state the size of his specimen, it is 1m- possible to say how far these differences are due to age. Giinther himself has suggested that guenthert and grande are identical. VODA Mx), PART I: 2 2 3 x About 3. Fic. 5.—Pachymetopon grande. atv SY 7 tay BY x About 2. Fic. 6.—Pachymetopon glaucum. 20 Annals of the South African Museum. 5. Pachymetopon glaucum, sp. n. Close to the preceding species. Depth of body 1% in the length, length of head 33. Diameter of eye 3% in length of head, 1% in inter- orbital width, and about equal to depth of praeorbital. Lower edge of praeorbital nearly straight; flange of praeoperculum not scaled. Apices of teeth of inner rows mostly truncate, but sometimes with a small median point or feebly crenulate; 22 teeth in outer row in both upper and lower jaws. 11 gill-rakers on lower part of anterior arch. 85 scales in lateral line, 10 or 11 from origin of dorsal fin to lateral line; caudal fin less densely scaled. Dorsal XI 11; 5th spine longest, length 2 in that of head. Anal III 11; 2nd and 3rd spines subequal, 13 times as long as first: 3rd spine about 3 as long as longest dorsal spine. Pectoral about 14 times as long as head. Dark grey above, with a bluish or violaceous tinge, silvery below; a number of narrow, dark longitudinal lines on the side, mainly below the lateral line: dorsal, anal, pelvics, and hinder part of caudal fin dark violaceous: upper part of pectoral dusky, lower part pale. Described from a single example (holotype), 310 mm. in total length, from East London. This was sent to the South African Museum by Mr. W. L. Wright, Hon. Secretary of the South African Sea-Anglers Association, in June 1933. According to him, this species, locally known as the Blue Fish,* attains a weight of 10 lb., but none of the local anglers have taken one under 21b. A prominent frontal gibbosity over the eye is sometimes developed. Hab.: South Africa. GYMNOCROTAPHUS, Giinther. Gymnocrotaphus, Giinther, 1859, Cat. Fish., vol. 1, p. 482 [Gymno- crotaphus curvidens, Giinther}. Close to Pachymetopon, but with no scales on the cheek. Teeth of innermost series more or less obtuse, but not molariform. A single species from South Africa. 1. Gymnocrotaphus curvidens, Giinther. Gymnocrotaphus curvidens, Giinther, 1859, Cat. Fish., vol. i, p. 432: Barnard, 1927, Ann. 8. Air. Mus., vol. xxi, p. 727, pl: xxix, fee; Fowler, 1933, Bull. U.S. Nat. Mus., 100 (12), p. 186. * This vernacular name is apparently applied indiscriminately to several species. Dr. Barnard informs me that examples of Pachymetopon grande, P. aeneum, Polyamblyodon germanus, as well as of another type of Percoid fish, were all sent to the South African Museum as “Blue Fish.” The European and South African Sea Breams. 21 Hab.: False Bay, extending along the coast to East London. The holotype is a stuffed specimen, 320 mm. in total length, from the “Cape of Good Hope”: I have also examined 8 examples in spirit, 175 to 300 mm. in length, from Kalk Bay, presented by the South African Museum in 1932. POLYAMBLYODON, gen. nov. Genotype—Pachymetopon germanum, Barnard. Close to Pachymetopon, differing in the form of the dentition. An outer row of strong, curved, compressed chisel-like teeth in each jaw, behind which is a broad band composed of 6 or 7 rows of small rounded molariform teeth. A single species from South Africa. 1. Polyamblyodon germanus (Barnard). Pachymetopon germanum, Barnard, 1934, Ann. Mag. Nat. Hist. (10), vol. xi, p. 231, fig. 2. Hab. : Coast of Natal. The holotype, 382 mm. in total length, in the British Museum collection, is believed to have come from Natal: a second specimen, 375 mm. long, from Durban, is in the South African Museum. THE SysTEMATIC PosITION OF DICHISTIUS AND TRIPTERODON. When he published his classification of the Percoid fishes, Regan * was able to examine only dried specimens of Pachymetopon and Dichistius (= Dipterodon, Cuvier nec Lacepede),t both of which he placed with doubt in the family Girellidae, a position which they occupy in Barnard’s monograph. The close relationship of Pachy- metopon to the Sparid genus Spondyliosoma has already been demon- strated, and the examination of a well-preserved specimen of Dichistius capensis leaves little doubt that this genus should be removed from the Girellidae and placed in the allied family Kypho- sidae. The principal characters distinguishing the Kyphosidae from the Girellidae are the exposed distal part of the maxillary; the scaly * 1913, Ann. Mag. Nat. Hist. (8), xii, p. 127. t+ Dipterodon, Cuvier, 1829, is preoccupied by Dzpterodon, Lacepéde, 1802 (type D. hexacanthus, Lacepéde)—a synonym of Apogon. Lacepéde’s genus includes species of Lutianus, Apogon, Aspro, Bairdiella, etc. Gill (1888) has proposed the name Dichistius as a substitute. Coracinus, Gronovius, 1763, is not accepted (Opinion 89). 22 Annals of the South African Museum. gill-membranes, joined to the isthmus; and the outer row of incisors in the jaws, implanted by horizontal roots, behind which is a series of small viliform teeth. In Dichistius the roots of the incisors are not conspicuous, and the inner series of teeth is represented by a few small conical teeth which are occasionally present, and are more or less concealed in fleshy pads.* Nevertheless, in other respects this fish appears to be a typical Kyphosid, and the definition of the family should be emended for its inclusion. Regan did not mention T'ripterodon in his classification, but, here again, at the time that his paper was published the British Museum possessed only the type of 7. orbis, a dried skin from Playfair’s Zanzibar collection. A much smaller example preserved in alcohol was received from Messrs. Marley and Robinson in 1919, but un- fortunately the mouth has been damaged and it is impossible to ascertain the form of the jaws. However, | feel certain that this fish cannot belong to the Sparidae, and its place is almost certainly either with the Girellidae or with the Ephippidae. The compressed, tricuspidate teeth are reminiscent of those of Grrella, but the general appearance of the fish is so like certain species of Ephippus and Chaeto- dipterus that it is difficult to believe that they are not related. Pending a study of its osteology, therefore, 7’rvpterodon may be placed — in the family Ephippidae. * Barnard, 1927, Ann. 8. Afr. Mus., xxi, p. 635. Ann. 8S. Afr. Mus., Vol. XX XII. Plate ITI. 2 xX 3. Pachymetapon canescens n. sp. Norman Neill & Co., Lid. baal on z ee “i oh rl ¥ 7% ox eye 7 ; Pa nna. ay My eee PARTS OF THE ANNALS PREVIOUSLY ISSUED— Vol. I1.—Part 1, out of print; Part 2, out of print; Part 3, out of saps eae Vol. I1.—Part 1, out of print; Part 2, 5/6; Part 3, out of print; Part 4, 3/-; Part 5, 1/6; Part 6, 3/-; Part 7, 1/6;°.Part 8, 3/-; eas 9, eae Part 10, 7/—; Part 1, 3/-; Index, "Title, etc. 16. SS Vol. I11.—Part 1, out of print; Part 2, 1/6;. Part 3, 5/6; Part 4, 3/-; Part oe 5/6; Part. 6, 7/-; Part I, 1/6 Part 8, 3+ Part 9, 1/6; basic Title, ete., 1/6. ez Vol. IV (containing Palaeontological papers published in ‘conjunction with the Geological Survey). 4 Part 1, 11/—; Part 2, 7/—; Part 3, 4/63 Part 4, 4/6; Part 5, ae 5 ie Part 6, 4/6; Part 7, 14)-; Part 8, 8/-. iy ea es Vol. V.—Part L, 4/6; Part 2, 8/6; Part 3; 2/6; Part 4, 1/6; Part Be spre Part 6, 5/~; Part: 7, 3/—;. Part 8, IB; Part 9, 4/6; Index, Title, es etc., 1/6. Li Vol. VI.—Part 1, 13/6;. Part -2, 4/6; Part 3, 3/6; Part 4; 30/53. Index, title, eee etc., 1 6. See Vol. VII (containing Palaeontological papeen published B conjunction with the — : Geological Survey).— Part 1, 3/-; Part 2, out of print; Part 3, 5c Part 4.8/1 Part 9, 5/6; Part 6, 1/6; Index, Title, etc., 1/6. ~ : Vol. VIII.—Complete, out of print. Index, Title, etc., 1/6. AO i 8 Vol. IX.—Part 1, 4/6; Part 2, 5/6; Part 3, 10/-; Part 4, 6/6; Sie 5, 3/05 ee ed ES Part 6, 11/=; Part 7, 9/—; Index, Title, etc., 1/6. Ses. Vol. X.—Part 1, 3/—; Part 2, 2/6; -Part s: 2); Part 4, 3/-; Part ‘5, 20/-; ries Part 6, 3/-; Part 7, 10/-; Part 8, 2/6; Part 9, 5/-; Part 10, 2/6; — ne Part 11, 20/-; ‘Part, 12, 7/-. Pome: Title, etc., contained. in S Part 11. Be ee Vol. XI.—Part 1, 3/6; Part,2, QI Part 3, 13/6; Part 4, 1/6; Part 5, ‘U/ss sn eke Part 6, 11/-; Index, Title, ete., and Plate Til, Bi. os ee ee Vol. XII (containing Palaeontological papers pees in conjunction with the 2 ch ie Geological Survey).— - age, Part 1,: 15/6; ~ Part, 2, 3/6: Part 3,.4/6; Part 4, 3/-; Part be Es Gia y? Part 6, 6/—; Part 7, 20/-; Part 8, 20/—; Index, Title, etc., 1/6. PE ee: Vol. XIIT.—Part 1, 6/-;. Part 2 “2/6; Part 3, 3/—; Part 4, 8/6; Part 5, 6s ate Part 6, 5/—; Part 7, 30/-; Part 8, i Index, Title, etc., 1/6. ec as ee Vol. XIV.—Part 1, 8/6; Part 2, 8/—; Part 3, 6/-; Part 4, 21/-; Pere, 5 Bie Aneto Part 6, 9/-; Index, Title, etc., 1/6. a eee Vol. XV.—Part I, 17/-;. Part 2, 17/-; Part 3, 14/-; Part 4, 12/6: Part 5, 5/65 Part 6, 3/6; Index, Title, etc., 1/6. ; _ “i Vol. XVI.—Part 1, 30/6; Part 2, 4/-; Part 3 (with Title, Index, ete. ), 25/-. CSE : : Vol. X VII.—Part 1, 12/-; Part 2, 9/6; Part 3, 3/—; Part-4, 17/-; Part 5, Hijcs ess . Part 6, 2/6; Index, Title, etc., 1/6. nee Vol. XVIII.—Part 1, 20/-; Part 2, 7/6; Part 3, 30/—-; Part 4, 12/6; Index, Title, etc., 1/6. Pe tgs ae Vol. XIX.—Part 1, 22/-; Part 2, 17/6; Part 3, 11/-; Part 4, 5/6; fader’ Title, a Sh ae etc., 1/6. vies Vol. XX.—Part 1, 8/6; Part 2, 12/6; Part 3, 4/-; Part ne aS Part 53: 4-5 a Part 6 (with Title, etc.), 4/6. Vol. XXI.—Part 1, 25/-; Part 2 (with Title, etc.), 30/-. Vol. XXII (containing Palaeontological papers published in conjunction with the Geological Survey).— PS Flee Part 1, 20/—; Part LA 10/—;. Part 3 (with Title, etc.), 3/6. _- Vol. X XITI.—Part 1, 12/6; Part 2. 8/-; Part 3 (with Index, Title, etc.), 9/-. Vol. XXTV.—Part 1, 10/-; Part 2, 2/6. fece te Vol. XXV.—Part 1, 12/6; Part 2, 10/-; Part 3 pee Index, Title, etc.), 10/-. Vol. XX VI.—Complete, 25/-. Vol. XX VII.—Complete, 25/-. Vol. XXVIII (containing Palaeontological papers published in | conjunction with the Geological Survey).— “ Part 1, 10/-; Part 2, 17/6; Part 3, 6/-; Part 4 (with Index, Title, etc.), 10/-. eee, Vol. XXIX.—Part 1, 20/-; Part 2 (with Index, Title, etc.), 20/—. Vol. XX X.—Part 1, 15/6; Part 2, 20/—;. Part 3, 8/-; Part 4, 9/-; Part 5, A. Vol. XX XI.—Part 1, 20/-; Part 2, 13/6. Vol. XX XII.—Part 1, 2/6. The Annals of the South African Museum will be tasied at irregular intervals, as matier for publication is available. Copies may be obtained from— Messrs. WHELDON & WESLEY, Lrtp., oe 2, 3, and 4 Artuur STREET, NEw OXFORD STREET, Lonvon, W 6, 235.0ree The LIBRARIAN, Souta Arrican MusEum, Carr Town. OF THE VOLUME XXXII. oe i, containing : — 3. aes on the Marine Mollusca in the Collections of the South | : African Museum. X. Family Verticordiidae.—By J. R. LE ‘B. Tomi, M.A. (With | Text-figure.) ss ae E A New Species of Emerita (Crustacea) from South Africa.—By = _ _Watpo L. Scumrrr, Curator of Marine Invertebrates, U.S. _ National Museum, Washington, D.C. (With Plate III.) . A New Dendrochirote Holothurian from South Africa, with some Remarks on the Rhopalodinidae——By 8. G. Hepine, Uni- versity Zoological Museum, Copenhagen. (With Plates SS1V5V, and 3 Text-figures.) 7 en Purther ‘Notes on South African Marine Fishes—By K. H. : ‘Barnarp, Assistant Director. (With Plates VI-VIII and 4 Text-figures.) a ‘z “On a New South African Species of Balanoglossus ‘nd a Comparison between it and Balanoglossus capensis — _ (Gilchrist).—By ©. J. van DER Horst, University of the - = Witwatersrand, (With 19 Text- Peutees), [iF en fs : G Ay ISSUED MARCH 1987. PRICE 7s. 6d. =o aS Ae PRINTED FOR THE : BY NEILL AND CO., 1D. — ate | CAUSEWAYSIDE, EDINBURGH ~ ( 23 ) 3. Reports on the Marine Mollusca in the Collections of the South African Museum. X. Family Verticordidae.—By J. R. LE B. Tomuin, M.A. (With 1 Text-figure.) Halicordia flexuosa (Verrill and Smith). Tuts lamellibranch species is very remarkable—not only for its general appearance, but for the wideness of its range. It was originally described by two joint authors as Mytilimeria flexuosa in the American Journal of Science (3), vol. xxi, p. 302, 1881, Fie. 1.—Halicordia flexuosa (Verrill and Smith). from a dead shell dredged off the New England coast in 571 metres. Subsequently it was dredged there alive in 137-1211 metres, and Verrill described both animal and shell in Tr. Connecticut Academy, vol. vi, p. 258, 1884. The type was also figured in vol. v of the same serial, on plate 58, fig. 38, 1882. Only two years after its discovery as above it was dredged by the “Talisman” at two stations on the West African coast, in 1495-2330 metres, and figured in the ‘‘Expéditions Scientifiques du Travailleur VOtle XXXIr, PART 2. 3 JUN 3 0 1937 24 Annals of the South African Museum. du Talisman,” Moll. Test., vol. x1, p. 210, pl. 10, f. 14-17, 1898, by Locard. In 1894 Dall minutely described and figured the anatomy of the species in Proc. U.S. Nat. Mus., vol. xvu, p. 697, pl. 23, figs. 1, 3, 5, 6, and pl. 24, fig. 3, from a specimen dredged off Georges Bank in the Gulf of Maine in 677 fathoms. This shell measured 45 by 39 mm.— more than half as large again as the holotype. In this paper Dall assigns the species to a new genus, Halicordia, and withdraws a previous reference of it to Vertecordia. Halicordia is placed near to Lyonsiella. We next hear of it in the Report of the Scientific Results of the Michael Sars North Atlantic Deep Sea Expedition (1910), vol. i, pt. 2, pl. 7, figs. 1-4, 1920, in which Grieg reports it from the west coast of Scotland and gives most admirable photographs. Two dead specimens were dredged by the Cape Government trawler s.s. “‘ Pieter Faure”’ in 1903 at the following localities: Cape Point N.E. by E., distant 36 miles, 650 fathoms, and Cape Point N. 81° E., distant 32 miles, 460-630 fathoms. The bottom at both localities was green mud. The first specimen (S.A.M., Reg. No. 150019) measures 54 mm. in breadth and 41 mm. in height, the second one (deposited in Brit. Mus.) measures 41 mm. either way. Grieg particularly remarks on the great variation in the dimensions of the species, the breadth varying from 83 to 103-2 per cent. of the height. ( 25 ) 4. A New Species of Emerita (Crustacea) from South Africa.—By Watpo L. Scumitt, Curator of Marine Invertebrates, U.S. National Museum, Washington, D.C. (With Plate III.) In the course of a revision of the genus Emerita (Sci. Survey Porto Rico and Virgin Ids., N.Y. Acad. Sci., vol. xv, pt. 2, p. 210 et seq., figs. 71-78) the regret was expressed that no South African specimens were at hand for examination. Recently the South African Museum has been so kind as to remedy this lack. In view of the number of species added to the genus from the Atlantic coast of America, it appeared likely that a second species might also occur in Indo- Pacific waters. Such, indeed, is the case, for these South African specimens from Durban Bay represent an undescribed species which may not inappropriately be known as : Emerita austroafricana, n. sp. 1910. Emerita emeritus Stebbing, Ann. 8. Afr. Mus., vol. vi, pt. 4, p- 366 (non Linn.). A strikingly different species, inasmuch as the distally sharply pointed dactyls of the first legs are laterally toothed or spined, perhaps better described as being armed with conspicuous marginal teeth tipped with tiny, movable, clear, corneous spines. These first dactyls in general outline are ovate, and, not counting the terminal or apical tooth or spine, are armed with two spines on the upper or outer margin and four on the inner or lower. This count holds for the specimen selected as the holotype, as well as for each of the other four specimens before me, all a part of the same lot, with the exception of the left first dactyl of one of the two largest and “soft” or recently moulted specimens, which carries five teeth on the lower margin instead of the usual four. Between these teeth or spines the margin of the dactyl is finely spinulose. The carapace is more or less cylindrical and dorsally conspicuously and transversely striate from behind the groove or impressed line, setting off the frontal region from the anterior portion of the carapace before the cervical groove to the posterior margin of the carapace. 26 Annals of the South African Museum. The greater part of the lateral or epimeral regions of the carapace is smooth and punctate. The transverse striae are well marked, rather widely spaced, and considerably interrupted. For the most part they are somewhat wavy, and toward their outer ends sparsely and scatteringly crenulate on their raised margins. In running a needle up the median line, or in a line parallel to it and not far removed from it on the dorsum of the carapace, one will cross parts or dash- sections of approximately 17 to 21 transverse lines or striae. The impressed line just before the hinder margin of the carapace and under the epimeron of the second abdominal somite to which I have elsewhere referred in characterising other representatives of this genus, is not in evidence in any of the five specimens of the present species before me. This line in other species is formed, as it were, vy the thickening of the hinder margin of the carapace, and, where it occurs, sets off this thickening from the surface of the carapace proper. Our species seems not to present such a thickening, and hence no impressed line of demarcation. I must needs here make note that a number of specimens of L. talpoida do show a suggestion, or even a slight ventral extension, of this line below the epimeron of the second abdominal somite contrary to an earlier observation (loc. cit.) that this was not the case. In shape, acuteness, and relative length the three projections of the front are much like those of H. emeritus (L.). The lobe at the antero-inner angle of the large and operculiform joint of the third or outer maxilliped is low-triangular, broad at base, and distally rounded. The prominently spined second joint of the antennal peduncle is quite smooth, and ornamented, if at all, with a few scattered double punctae. Near its ventral margin, dipping anteriorly and toward the under side, are two oblique incisions, the lower margins of which are densely piliferous and ridge-like. These ridges or incisions seem to be developed in all species of Emerita, and end, in the case of the more anterior of the two, at or on the third and lowermost of the three spines with which this joint of the antennal peduncle is distally armed. The uppermost, or dorsal, of the three spines arming this joint is quite long and slender, continuing the line of the upper margin of the joint more or less straight forward; the outer margin of the joint at the level of the median spine, as viewed from above, is about straight. Holotype.—An ovigerous female (U.S.N.M. Cat. No. 71446) from Durban Bay, Natal, measuring in length of carapace, including the A New Species of Emerita (Crustacea) from South Africa. 27 median rostral projection, 26 mm., greatest width 18 mm. With this specimen were four other females, of which three are also ovigerous. Of these last, two are larger and soft, one of these being the largest of the five specimens seen, 33 mm. long over the carapace and rostral projection. (Paratype: S.A. Mus. Cat. No. A 911.) Regarding the distribution of this species, Dr. Barnard writes : “The specimens come from Durban Bay, Natal, and, so far as I am aware, Hmerita has not been collected anywhere farther westwards along the coast, though it probably extends northwards all up the East African coast.” Stebbing (Ann. 8. Afr. Mus., vol. vi, pt. 4, p- 366) merely mentions: “A specimen sent by Mr. Quekett, from the Durban Museum, has a carapace measuring 30 mm. in length from the central tooth of the tridentate front to the hind margin.” Remarks.—In true E. emeritus from the Indo-Pacific (loc. cit.) the second joint of the antennal peduncle, on its outer surface, is ornamented with several oblique, somewhat irregular lines of crenulations (crenulated ridges running obliquely downward and back from the dorsal spine) and a number of very short crenulated ridges of similar inclination in a series. behind and on the upper, dorsal slope of the rise or swollen ridge, as it might well be called, which terminates anteriorly in the largest or major spine of the joint. The lower half of the joint is more or less smooth except for the two oblique piliferous incisions common to all members of the genus. The uppermost spine is short and stubby, and at extreme tip a little bent up, not particularly noticeably so, but very acute and spiniform. The outer margin of the joint at the level of the middle spine, as viewed from above, is very slightly concave at about the base of the spine. The second joint of the antennal peduncle of H. analoga is much hike that of H. emeritus, but in degree of ornamentation lies between the latter species and EH. austroafricana, though nearer the former; the outer margin of the joint in dorsal view is slightly concave. E. talpoida has this second joint much like that of #. analoga and E. emeritus in general shape; the outer margin, too, is slightly concave; the upper spine is short and stubby, and directed more straightly forward than in the other two. However, the ornamenta- tion is strikingly different, inasmuch as a crenulated ridge or line runs backward from both the uppermost and median spines to distinctly mark off a depressed area between them; these lines converge and meet at or a little behind the middle of the joint at that level. Above and behind these lines, and on the hinder portion 28 Annals of the South African Museum. of the joint below their level, there are some scattered, irregular, short lines of crenulations. The noticeably depressed area or sinus between the V or narrow U formed by the two crenulated lines running back from the spines is roughened a bit with a few more or less isolated, much abbreviated, little lines of crenulations. E. portoricensis much resembles £. talpoida in these several respects; the crenulated line running back from each of the two upper spines meets its fellow to form a more broadly U-shaped but less depressed area, also roughened within. In some specimens there are several broadly U-shaped lines of crenulations, the upper limb of the more prominent one running part way on to the base of the uppermost spine of the joint, while the lower limb is more or less continuous with a line of crenulations running out on to and forming a ridge or line of crenulations on the middle spine; both before and behind this major U are similar duplicate shallow U’s without the arms or lines of crenulations running up on the spines. The rest of the surface of the joint carries a few widely separated, short lines of crenulations. In these respects EL. benedicti is much like portoricensis. In the larger, older, and better-developed specimens of #. rathbunae there is a longitudinal, raised, crenulated line running back from the middle or larger spine of the joint to about half the length of the body of the joint; above this line the surface of the joint is con- spicuously roughened with short, wavy, coarsely crenulated lines, which also ornament the surface of the joint posterior to the hind end of the line on the middle spine, as well as, to a lesser degree, the surface below the line. The uppermost spine is somewhat abruptly and markedly bent upwards above the general trend of the upper margin of the joint; the outer margin of the joint at the level of the middle spine, as viewed from above, and the general trend of the spine tend to meet at quite a decided, though shallow, angle of 150° more or less. In £. brasiliensis a shallow U-shaped depressed area les behind the sinus between the upper and middle spines; below, this is bounded by a line or crenulated ridge made up of a successive series of short little vertical lines of crenulations, one close behind the other, along the line of the larger or middle spine of the joint; this shallow U-shaped area is bounded by one or more short crenulated lines, while farther back are a few more, except for which the surface of the joint is generally smooth and sparsely punctate; there is no ridge or crenulated line along the line of the upper spine of the joint. A New Species of Emerita (Crustacea) from South Africa. 29 This spine is of good size, directed quite straightly forward. The outer margin of the joint, as viewed from above, though markedly concave, is not angled as in #. rathbunae. In order to fit our new species into a recently published key to the species of Emerita (Sci. Survey Porto Rico and Virgin Ids., N.Y. Acad. Sci., vol. xv, pt. 2, p. 210 ef seq., figs. 71-78), the first major head A! and the first secondary head B! under it need to be revised as follows: A}. Dactyls of the first legs distally subacute or sharply pointed. Transverse striations on carapace often conspicuously crenulate, more or less continuous, and crossing dorsum for whole of its extent, except at times the frontal region. Bt. Dactyl broadly ovate, distally pointed, less than twice as long as greatest width. Surface of lateral, epimeral expansion of carapace smooth and punctate. The three projections of the front are all quite narrowly triangular and more or less sharply pointed, the median more so than the lateral. C!. Margins of dactyl armed with several conspicuous spines or teeth. No impressed line just before hinder margin of cara- pace either side of first abdominal somite. Transverse lines or striations on carapace little crenulate, for most part wavy and considerably interrupted . austroafricana Schmitt. (Type locality: Durban Bay, Natal, South Africa.) C?. Lateral margins of dactyl not armed with conspicuous teeth or spines. Impressed line just before hinder margin of carapace either side of first abdominal somite, not showing in lateral view below epimeron of second somite. Crenula- tions of transverse lines on hinder part of carapace numerous, crowded, and very rough to touch when finger is passed over carapace from front to back : . emeritus (Linn.). (EZ. asiatica of Milne Edwards and others.) (Type locality: “Habitat in Mari Indico.’’) EXPLANATION OF PLATE. Dorsal and lateral views of: Top: Emerita austroafricanan. sp. Type. xl. Bottom: Emerita emeritus (Linn.). x14. = ty / (repeal e . Ae; Ann. Schmitt. S. Afr. Mus., Vol. XXXII. EMERITA. Plate ITI. Neill & Co., Lid. La) : A 4 ; ns 5. > ( 31 ) 5. A New Dendrochirote Holothurian from South Africa, with some Remarks on the Rhopalodinidae.—By 8. G. Hepine, Uni- versity Zoological Museum, Copenhagen. (With Plates IV, V, and 3 Text-figures.) ile Rhopalodinopsis n.g. Diagnosis —Body composed of two parts, a lower spherical soft part containing the intestine, the gonads, and the respiratory trees; and an upper part, which is proboscis-shaped and hard, and contains the cloaca, the oesophagus, the ring-canal with the polian vesicle, and the stone-canal and the calcareous ring with the retractors. Mouth and anus are both placed at the end of the proboscis, but are clearly separated by the dorsal part of the dorsal interambulacrum. The tube-feet are confined to the radii of the sphere and the proximal part of the proboscis. Tentacles about twenty. Retractor muscles present and respiratory trees well developed. Tentacle-ampullae and rete mirabile lacking. Gonads in two tufts, opening on a conical papilla into the cloacal cavity, close to the anal opening. Anal opening supplied with five pairs of papillae, placed on a system of large perforate anal plates. Calcareous deposits of body-wall in two layers, an inner well-developed layer of perforate, rather solid plates, and an outer very sparse layer of rudimentary tables. Genotype.—Rhopalodinopsis capensis n.sp. Remarks.—The specimens at hand must without doubt be referred to the family Rhopalodinidae as characterised by Panning in 1932 and 1936, but they differ in several characters so definitely from the species described by Panning in his latter work that it seems most reasonable to establish a separate genus. One of the features which I consider as being of generic value in the classification of these forms is the arrangement of mouth and anus. These two openings are well separated in Rhopalodinopsis, and thereby form a real “dorsal side’”’ on the specimens, whereas in Rhopalodina they are set very close together and usually in such a way that the anal pore is placed on the tip of the trunk, and the mouth on the side of it, though quite close to the tip. Another 32 Annals of the South African Museum. feature available for separating the genera is the presence of well- developed tube-feet on the ventral pole of the body-sphere. In the species of Rhopalodina such well-developed ventral tube-feet (pedicels) are lacking, which has occasioned Semper’s statement that R. lageni- formis has 10 ambulacra. In the present genus the well-developed tube-feet clearly show the shape of the ambulacra, and that they are only 5 in number, of which the two dorsal ones do not reach the pole of the body-sphere (Plate IV and fig. la). Also the shape of the calcareous plates from the inner layer of the body-wall, and especially the shape of those from the sphere, indicates a generic difference between the present species and those of Rhopalodina. Rhopalodinopsis capensis u.sp. Locality. Kalk Bay, Cape Peninsula, low water (the specimens were found washed ashore after a storm, October 1931, and presented to the Museum by Mr. C. L. Biden). External Features.—The specimens (Plate IV) measure about 10 cm. in length. They are shaped as Bonellia with the proboscis stretched out. The ventral part of the specimens, the sphere, measures about 3 cm. in diameter and is quite soft to the touch, due to the relatively few and small plates in the body-wall. Its colour in alcohol is pale yellowish, and the body-wall itself is semi- transparent, so that the gonads and the intestine can be seen through it. The trunk is hard and white, due to the numerous large im- bricated plates. [The colour was the same when the specimens were first brought to the Museum.—ED.] The pedicels are placed in two rows in each ambulacrum, and reach from the ventral pole of the sphere to about 1 cm. on the base of the proboscis. They are largest on the ventral half of the sphere (Plate IV), and are rudimentary on the proboscis. Contrary to what is the case in Rhopalodina the pedicels are well developed on the ventral pole of the sphere (Plate IV, fig. 4 and fig. la), and it is easily seen that only the mid-ventral ambulacrum crosses the actual pole. The other ambulacra are only bent towards the pole, leaving narrow interambulacral spaces free. Mouth and anus are both placed on the dorsal side (v.e. the tip of the proboscis) (fig. 1b). There is a distinct interambulacral dorsal space developed, and here the large calcareous plates of the trunk are lacking and replaced by a few small perforated plates. The oral Opening is surrounded by some rather large scales, which, when the eee eee Or A New Dendrochirote Holothurian from South Africa. 33 tentacles are totally retracted, may close the opening. Around the anal opening there are ten surprisingly well-developed anal papillae, two in each ambulacrum. Anal teeth are lacking, but the anal papillae are at their base supplied with some large perforated plates, one for each papilla; these plates further lie in pairs on five other deeper-lying radial plates. Superficially we may thus speak about an anal calcareous ring, as did Semper. Further, care must be taken Fie. 1—Rhopalodinopsis capensis n.g., n.sp. (a) Ventral pole of body-sphere, showing arrangement of radii and tube-feet. (6) End of proboscis with oral (o.) and anal (a.) openings, showing the peculiar medio-dorsal part of the dorsal interambulacrum. (c) Transverse section of proboscis about 1 cm. from end, showing longitudinal muscles (l.m.), cloaca (cl.) gonoduct (g.) oesophagus (oe.) and cavity (l.) between anterior and posterior part of the folded inner layer of the medio-dorsal interambulacrum. (d) Medio-dorsal part of calcareous ring (c¢.r.), with ring-canal (r.c.), polian vesicles (p.v.), stone-canal (s.c.), and axial organ (a.0.). that the highly developed anal papillae with their fenestrated plates are not confounded with the tentacles and the calcareous ring. Anatomy.—The specimens were all washed ashore, so they are scarcely well enough preserved to allow a very detailed examination of their anatomy; in any case, not of that of the body-sphere. In the sphere the intestine, the respiratory trees, and the gonads are found, and in the proboscis the cloaca and the anterior body-end with calcareous ring, ring-canal, oesophagus, gonoduct, and retracted muscles. A transverse section of the proboscis about 1 cm. from its distal end shows the arrangement of some of these organs (fig. 1c). On the one side (i.e. that which represents the posterior end of the animal) there is a rather voluminous cloaca (cl) surrounded by five strong 34 Annals of the South African Museum. longitudinal muscles. On the other side (7.e. the anterior end) there are the oesophagus (oe) and the gonoduct (g), also surrounded by five longitudinal muscles. In the section here figured the oesophagus and the gonoduct are placed side by side, but farther towards the end of the proboscis the gonoduct bends behind the oesophagus and then runs through the two laminae seen in the figure, and opens on a large papilla into the cloaca. This is quite a unique place for the genital opening in Holothurians. Further, it appears from the section that the oesophageal part of the proboscis is separated from the cloacal part by a narrow cavity, a cavity the origin of which is at first sight difficult to ascertain. A closer examination, however, shows that the two laminae which separate the oesophageal and cloacal parts of the body-cavity are the muscular part of the dorsal interambulacrum which, forming a large fold, divides the coelomic cavity of the proboscis into two channels. Thus the cavity separating these two laminae is seen to be, not of coelomic origin, but merely a secondary cavity formed when the saddle-shaped external layer of the dorsal interambulacrum parted from the muscular inner layer and was drawn into the interior of the proboscis. The ring-canal (fig. 1d) is supplied with 2-3 polian vesicles, a short stone-canal with a rather large madreporite, and a long and well- developed “‘axial organ.””. The calcareous ring (fig. 2) consists of ten pieces, of which the radials have more or less rudimentary posterior prolongations. The interradials are all rather simple, triangular, thick pieces, usually with a deep incision in their posterior margin. The radials are of different shape, not only in different specimens, but even in the same specimen, and the mid-ventral radial is always less well-developed than the lateral ones. Anteriorly all the radials have more or less deep incisions, and in every case the larger part of these incisions form insertions for tentacles or incisions for the passage of the water-vascular system and the radial nerves. In spite of the fact that the pieces of the calcareous ring are rather thick, so thick that we should not call them incisions, but rather bowl- shaped depressions for the musculi retractor tentaculi, the limits between the calcified part of the ring and the surrounding connective tissue are often indistinct; for which reason it is impossible to ascer- tain the real number of tentacles when studying the shape of the calcareous ring. From the shape of the calcareous ring one would expect between 20 and 30 tentacles, and a careful dissection of the retracted muscles A New Dendrochirote Holothurian from South Africa. 35 showed that there were 22, but of these 2 may perhaps be merely torn-off pieces or folds of the introvert. In any case, this species has about 20 tentacles, and of these it seems (it is not quite certain) that 5 are placed in an inner circle. The tentacles themselves are all unbranched and supplied with numerous calcareous deposits. Calcareous Deposits—Similar to the species of Rhopalodina (as shown by Panning), the present species of Rhopalodinopsis has two me Gy e d f Fic. 2.—Rhopalodinopsis capensis n.g.,n.sp. (a—c) Calcareous ring of specimen I. -(a) Right ventral radial with additional interradials. (b) Mid-ventral radial ' with additional interradials. (c) Mid-dorsal interradial with additional radials. (d) Mid-ventral radial with additional interradials of specimen II. (e, f ) Pieces of calcareous ring of specimen II cleared in xylol and mounted in balsam: (e) Mid-ventral radial with left ventral radial and additional interradials. (jf) Mid-dorsal interradial and left dorsal interradial with left dorsal radial and fragments of right dorsal radial and left ventral radial. layers of calcareous deposits in the body-wall, a deeper layer of perforated plates and a superficial layer of tables, but here the outer layer is so reduced that it has not been possible to find more than two rudimentary tables. The deposits of the inner layer are very different in the two main parts of the body. In the proboscis they are large circular plates with round holes (Plate V, fig. 1). They are imbricated, and in their free parts have a number of large knobs on their external side. These knobs are smallest in the plates from the proximal part of the proboscis, and close to the free end of the proboscis they may be as high as is shown in the figure (Plate V, fig. la). The deposits from the wall of the sphere (Plate V, figs. 2-5) are not imbricated, 36 Annals of the South African Museum. being small thick bodies which are fairly uniformly dispersed. None of these plates, either from the proboscis or from the sphere, consists of more than one layer, which 1s often the case in the genus Rhopalodina. As stated above, the deposits of the superficial layer are nearly lacking. There are, however, two small spiny button-like deposits lying superficially above the large “‘fat’’ bodies in the wall of the sphere, deposits which do not leave much doubt that they are rudi- mentary tables (Plate V, figs. 6, 7). The tube-feet are without end-plates, but have small spiny rods (Plate V, figs. 14-18) in their walls. In the tentacles there are two different sorts of rods: relatively large perforated plates (Plate V, figs. 8-10) and rather characteristic small oval plates or bodies (Plate V, figs. 11-13). II. REMARKS ON THE Rhopalodinidae. The fact that the new species described above has about 20 tentacles appeared to be quite contrary to what is the case in the other species belonging to the Rhopalodimidae; Semper, as well as Panning, stating the normal number of tentacles in Rhopalodina to be 10. A com- parison of the calcareous ring of Rhopalodinopsis capensis with Semper’s figure of that of R. lageniformis showed, however, that there were reasons for supposing that Semper’s statement was not quite correct. I therefore wrote to Dr. Panning in Hamburg asking him to undertake a renewed examination of the tentacles of his specimens. As Dr. Panning had long ago finished his beautiful work on the genus Rhopalodina, and for the present had no time to work on these specimens, he kindly sent me some specimens and fragments of the two species R. lageniformis and gracilis, and suggested that I under- take the examination myself. This examination gave the following results: in the specimens of R. lagensformis (Nos. 4477 and 4479) the calcareous ring consists of ten pieces, of which all the radials have a distinct posterior bifurcating prolongation (fig. 3, a-d). The real shape of the pieces is very difficult to ascertain, as they are not only small, but also placed in such a way that quite different results are obtained when the piece under observa- tion is turned a little to one side or the other. Figures a—I in fig. 3 are all drawn, as far as possible, in such a position that the central piece in the figure is lying parallel to the surface of the alcohol in the dissecting dish. Only figures 6 and ¢ are drawn in a slightly A New Dendrochirote Holothurian from South Africa. 37 oblique position—in 6 turned slightly to the right, in ¢ slightly to the left. It is evident that if all the incisions in the anterior side of the calcareous ring, apart from the five radial incisions for the radial nerves and canals, are insertions for tentacles, which is normal in i . I Fie. 3.—Rhopalodina lageniformis Gray. (a-d) (Panning, No. 4477). (a) Mid- ventral radial with the two ventral interradials and parts of the two lateral radials. (6, c) Right ventral radial and interradial drawn in two slightly ~ different positions, showing how the contours of the pieces may vary in accordance with the position of the piece under examination. (d) Mid-dorsal interradial with the two dorsal radials and lateral interradials. (e, f ) (Panning, No. 4479). Mid-ventral radial with ventral interradials and mid-dorsal interradial with right dorsal radial, showing the variation of the shape of the calcareous ring in this species. Rhopalodina gracilis Panning. (g-l) (Panning, No. 4484). (g) Mid-dorsal inter-radial with the abnormal double right radial; the stippled line is the outline of the abnormally wide right retractor. (h) Right lateral interradial with a part of the right ventralradial. (7) Right ventral radial with additional interradials. (7) Mid-ventral radial with ventral interradials. (k) Left dorsal radial with mid-dorsal interradial. (l) Left lateral interradial with additional radials. most other Holothurians, we should expect about 20 tentacles, and dissection of specimen No. 4479 showed 18 tentacles, of which 5 were placed in an inner circle. After having done this as carefully as possible, I cleared the whole introvert in xylol and mounted it on a slide. Further examination showed that I had overlooked two small ventral, apparently regenerating, tentacles. The number of tentacles is thus in this specimen (i.e. R. lageniformis sens. str.) 20, 38 Annals of the South African Museum. and they are arranged with 15 in an outer and 5 in an inner circle. The other specimen of R. lageniformis, No. 4477, was unfortunately so badly preserved that I was unable to count the tentacles. The two other specimens at hand, Nos. 4484 and 5863, represent the species R. gracilis Panning. One of them, No. 5863, had the calcareous ring destroyed, and could not throw any light on either the shape of the ring itself or the number of tentacles. The other, No. 4484, showed a slight but apparently distinct difference from the calcareous ring of R. lageniformis. This may be a systematic character, but as the ring is evidently abnormal, having two right dorsal radials, we should not lay too much stress on the shape of it. The examination of the tentacles in the opened introvert, as well as in cleared preparation, showed that they were 18 in number. As far as it was possible for me to ascertain there were not more than 18 tentacles, and if the calcareous ring were not abnormal we should have here a valid character for separating the two species lageniformis and gracilis. As, however, the calcareous ring is abnormal, it is possible that the normal number of tentacles is 20 in gracilis also. The fact that the species of Rhopalodina and Rhopalodinopsis, 1.e. all the members of the family Rhopalodinidae, have not 10 (as pre- viously assumed) but about 20 tentacles, arranged in two circles with 5 in the inner and 12-15 in the outer circle, appears to be of the greatest value for the classification of the Dendrochirote Holothurians. Owing to the large number of tentacles it is now quite evident that Rhopalodina (and Rhopalodinopsis) cannot be placed close to Echino- cucumis and Sphaerothuria with Ypsilothuria (see Perrier, 1902), but that it is closely related to the Phyllophorinae. If all the characters are taken into consideration, there is no reason to maintain a separate family for these two genera, since they may be placed, without serious difficulty, close to or within the subfamily Phyllophorinae. However, so long as our knowledge of the classification of the Phyllophorinae, and indeed of the whole order Dendrochirota, is as unsatisfactory as it is at present, it is of no practical use to try to find the true natural - place for these odd forms. I prefer for the moment to maintain a separate taxonomic group for the two genera here under discussion, but I suppose it must be regarded as a subfamily, the Rhopalodinae, closely related to the Phyllophorinae. In reality the only important difference between the Rhopalodinae and the Phyllophorinae is the presence of large plates in the former, as I Jo not think that the peculiar body-shape is of any higher taxonomic value. I wish to express my sincerest thanks to the Director of the South A New Dendrochirote Holothurian from South Africa. 39 African Museum, who has given me the opportunity of studying these very interesting specimens, and to Dr. Panning, Hamburg, who has supplied me with material of Rhopalodina for examination of the calcareous ring and tentacles in that genus, and to Dr. Mortensen of Copenhagen, whose good advice and never-failing interest in my studies have been of invaluable help to me. LITERATURE. Gray, J. E., 1853. Description of Rhopalodina, a new genus of Echinodermata, Ann. Mag. Nat. Hist., Ser. 2, vol. ix. SEMPER, C., 1868. Reisen im Archipel der Philippinen, 2 Theil, Wiss. Res., Bd. 1, Holothurien. Lupwie, H., 1877. ‘‘Ueber Rhopalodina lageniformis Gray und die darauf gegriin- dete Klasse Diplostomidea Semper,”’ Zeitschr. Wiss. Zool., Bd. 29. PERRIER, E., 1886. Les Explorations sous-marines, Paris. THhEL, Hy., 1886. Holothurioidea II, in “‘Challenger” Report, vol. XIV, 2. Lupwic, H., 1889. ‘‘Nochmals iiber Rhopalodina lageniformis,” Zeitschr. wiss. Zool., Bd. 48. Stup#rR, TH., 1889. Die Forschungsreise 8.M.S. “‘Gazelle”’ in den Jahren 1874-76, 2 Theil, Zoologie und Geologie. Lupwie, H.,,1892. Die Seewalzen, in Bronn’s Klassen und Ordnungen, Bd. II, 3 Abtlg., Echinodermen | Buch. Goopricu, EH. S., 1900. Holothurioidea; in Lankaster: Treatise on Zoology, part 3; Bather, F. A., Echinoderma. PERRIER, R., 1902. Holothuries, in Expéditions scientifiques du “Travailleur”’ et du “Talisman.” Vaney, M. C., 1911. MHolothuries, in Mission Gruvel sur la céte occidentale d’ Afrique (1909-10); in Ann. Inst. Oceanogr., T. 2. Mortensen, Tu., 1927. Echinoderma; in Monod, Th., Contribution 4 l’étude de la Faune du Cameroun, Premiére Partie; in Faune des Colonies Frangaises, Bd. I. MortTENSEN, TH., 1927. Handbook of the Echinoderms of the British Isles, footnote, p. 395. PannineG, A., 1932. Echinoderma III. MHolothurioidea 3 (Fam. Rhopalodinidae); in Michaelsen, W., Beitrage zur Kenntniss der Meeresfauna Westafrikas, Bd. III. Panning, A., 1936. Die Gattung Rhopalodina. Verhand]. des Naturwiss. Vereins in Hamburg, Vierte Folge, Bd. V, Heft 1-4. EXPLANATION OF PLATES. Prats IV. Rhopalodinopsis capensis n. g., ND. Sp. Fics. 1-3. Three specimens, natural size. Fic. 4. The pole of specimen 3 slightly enlarged. VOL. XXXII, PART 2. 4 40 Annals of the South African Museum. PLATE V. Calcareous deposits of Rhopalodinopsis capensis n. g., n. sp. Fic. 1. Circular fenestrated plate from the proboscis. la. Part of circular plate from the end of proboscis, seen from the side, showing high spine-like tubercles. 2-5. Small perforate plates from the wall of the sphere. 6-7. Rudimentary tables from the outer layer of the wall of the sphere. 8-10. Perforated plates from tentacles (inner layer). 11-13. Oval minute bodies from tentacles (superficial layer). Fig. 11 with the same magnification as the other figures. 14-18. Rods from the tube-feet. Ann. 8, Afr. Mus., Vol. XXXII. Plate IV. RHOPALODINOPSIS CAPENSIS, n.g., n.sp. S. G. Heding, photo. Neill & Co., Ltd. Plate V. Ann. 8. Afr. Mus., Vol. XXXII. RHOPALODINOPSIS CAPENSIS, n.g., n.sp. Neill & Co., Lid. S. G. Heding, del, a el 6. Further Notes on South African Marine Fishes.—By K. H. Barnargp, Assistant Director. (With Plates VI-VIII and 4 Text-figures.) THE following notes are based on specimens added to the South African Museum collection in recent years. For several species a large number of new localities has been ascertained, thanks to several correspondents. These have not been included here, as the net result is to show that many of the so-called warmer water species extend much farther westwards than was hitherto thought to be the case. The occurrence of the Springer (Hlops saurus) in False Bay is perhaps the most noteworthy record. A list of papers published since 1927 is given, and some of the main changes in nomenclature, particularly in the names of the Flat-fishes, are incorporated. It is a pleasure to refer to the work which Dr. J. L. B. Smith, of Rhodes University College, Grahamstown, is doing in this country. Dr. Smith is favourably situated for investigating the stretch of coast between the Cape and Natal, in particular the Knysna and Port Alfred areas, about whose marine fauna we know so little at present, and he is making excellent use of his opportunities. The following papers, published since 1927, deal wholly or partly (changes in nomenclature, etc.) with South African marine fishes :— 1934. Barnard, K. H., Ann. Mag. Nat. Hist. (10), xi, pp. 228- 235, figs. 1935. af Ann. 8. Afr. Mus., xxx, pp. 645-658, text- figs. and pl. xxi. 1927. Chabanaud, P., Bull Inst. Océan. Monaco, No. 500, pp. 1-15. 1927. Hh Ann. Mag. Nat. Hist. (9), xx, pp. 523-527. 1928. ze Bull. Soc. Zool. France, lin, pp. 272-279. 1929): a Ann. Inst. Océan. Paris, n.s. vu, pp. 215- 260, text-figs. and plates. 1930. He Bull. Inst. Océan. Monaco, No. 550, pp. 1-23. 1934. Bull. Soc. Zool. France, lix, pp. 420-436. 1929. Fowler, W. H., Ann. Natal Mus., vi, pp. 245-264. 42 Annals of the South African Museum. 1931. Fowler, W. H., Proc. Ac. Nat. Sci. Philad., Ixxxi, pp. 245- 1931. 1933. 1934. 1934. 1934. 1935. 1928 1929: 249 (marine). Bull. US. Nat. Mus.; No. 100; sola Ibid., vol. 12. Proc. Ac. Nat. Sci. Philad., Ixxxv, pp. 233- 367, text-figs. (subfam. names, etc.) (Jan.). Ibid., Ixxxvi, pp. 405-514, text-figs. (Nov. 6). Ann. Natal Mus., vu, pp. 403-433. Proc. Ac. Nat. Sci. Philad., Ixxxvu, pp. 361- 408, text-figs. . Fowler, W. H., and Bean, B. A.,- Bull. U.S. Nat) Wise iio: 99 1930. 1929 - 1930. 1931. 1934. 1935. 1935. 1929 1930. 1929 1931 1933. 1933. 1934. 1934. 100, vol 7 ss Ibid., vol. 8. Me Ibid., vol. 10. . Norman, J. R., Ann. Mag. Nat. Hist. (10), iv, pp. 153-168. 29 99 29 ““Discovery”’ Reports, 11, pp. 261-370, text- figs. and pl. i. Ann. Mag. Nat. Hist. (10), vii, pp. 507— 516. Monograph of Flat-fishes, vol. 1, Brit. Mus. Ann. 8. Afr. Mus., xxxu, pp. 5-22, text-figs. and pl. u (Spondyliosoma). Proc. Zool. Soc. Lond., pp. 99-135, 18 text- figs. (Synodontidae). . Regan, C. T., and Trewavas, H., “Dana”? Oceanogr. Rep., 29 No. 5, pp. 12-30, text- figs. and plates. ee liid., No. 6) pps: text-figs. and plates. . Schnakenbeck, W., Mitt. Zool. Mus. Hamburg, xliv (1931), pp. 23-46. . Smith, J. L. B., Rec. Albany Mus., iv, pp. 145-160, text-figs. and pl. xvi. Tr. ‘Roy. Soc: 8. Afr.) xsappeiiasere pl. 1x (Myctophum). Ibid., xxi, pp. 129-150, pls. x-xii and 1 text- fig. (Hemirhamphidae). Ibid., xxii, pp. 83-87, pl. iv and 1 text-fig. (Pteroplataea). Ibid., xxi, pp. 89-100, pls. v, vw and 1 text-fig. Further Notes on South African Marine Fishes. 43 1934. Smith, J. L. B., [bid., xxii, p. 321-336, pls. xvi-xxiii and 1 text-fig. (Triglidae) (Dec.). 1935. fe Ann. 8. Afr. Mus., xxx, pp. 587-644, pls. Xv-xxii and 17 text-figs. (Wugilidae). 1935. “3 Ibid., iv, pp. 358-364, pls. xl—xlii (Aluteridae). 1935. - Rec. Albany Mus., iv, pp. 169-235, text- figs. and pls. xviii-xxiii. 1935. a Tr. Roy. Soc. S.-Afr., xxii, pp. 265-276, pls. xii—xvu (Dichistiidae). 1936. in Ibid.. xxiii, pp. 303-310, 1 text-fig. and pls. xxi—xxill (Tripterodon). 1936. a Ibid., xxiv, pp. 1-6, 2 text-figs. and pls. i, ii. 19350: . Ibid., xxiv, pp. 47-54, 2 text-figs. and pls. ii—v (Gobiorder). 1936. Svetovidov, A., Faune de |’Urss. Poissons, vi, 9, pp. 1-21, pls. 14, with resumé in English, pp. 22-24 (Triglidae). 1929. von Bonde, C., Fish. Mar. Biol. Surv., vii, Spec. Rep. 3, pp- 1-11, figs. (Heterosomata). 1933. - J. Comp. Neur., lviu, pp. 377-417 (cranial nerves, etc. (Pliotrema and Sphyrna). Fam. IsuRIDAE. Cetorhinus maximus (Gunner). (Plates VI, VII and text-fig. 1.) 1925. Barnard, Ann. 8. Afr. Mus., xxi, p. 34, pl. u, figs. 1, la. Two more specimens have been captured in South African waters. Like the first recorded specimen, which was caught in 1917 and mounted in the South African Museum, the second specimen was also a young individual. It measured 11 feet 6 inches, and was caught in Table Bay 23rd March 1930. It was blackish in colour, with small black spots on the sides; the lips were white, and there were white streaks below the snout and on the chin and throat. The third specimen to be recorded from South African waters was netted by Mr. Trauter in Hout Bay (west coast of Cape Peninsula) on 20th January 1935. It was a male measuring 28 feet in length, and was towed into Table Bay and exhibited for a few days on the Cape Town pier. The animal was then cut up and dumped in the sea, without advising 44 Annals of the South African Museum. the Museum. The head was presented to the Museum, but the chance of securing one of the claspers was lost. This is much to be regretted, as White (1930, Bull. Amer. Mus. Nat. Hist., lxi, p. 158) says the a 5 Fig. 1.—Cetorhinus maximus (Gunner). Dorsal view of rostral cartilages, with cross-sections of the ventral projection at a and b. examination of the clasper of an adult might help to clear up the relationship between Cetorhinus and Rhineodon. The following measurements were taken :— Tail hanging in water—not measured; ventral distances separating gill-shts not visible, and outer margin Ist dorsal curled up and not measured. Pit at root of caudal to end of base 2nd D. . f . 133: tehes Base of 2nd D. olf ees Front margin of 2nd Ds 14 inches: “hind “margin, 15 inches; lower margin 1 ne Front of 2nd D. to end of base Ist D. : f i ee Base of Ist D. . : ae Front margin of Ist D., 47 inches: lower margin : , OS Further Notes on South African Marine Fishes. 45 Depth of caudal peduncle . ; ; : . Ill inches Base of C. to end of base of anal 4 ; SOO hiae Base of A. : : : ; ’ Or ae Front margin of A., 13 inches: upper margin, 9 inches; lower margin : Sas oak Front of base of A. to angle of base of ventrals 36(+6) ,, Base of V. : : 7: aime Front margin of V., 27 inches: outer margin of V. sa 5 NR ae Clasper . 2 OME 5 V. to base of pectoral : 5 One Upper margin of P., 57 inches; outer margin of ey 52 inches; lower margin of P. : 1 6 aga Width i base of P. . : ‘ : : Meee Ist D. to snout. : 3 : ; ; ot ee Snout to Ist gill- elec 5 : ‘ {ty Veagere.: Ist gill-cleft to 5th cleft. b . ; Bl eyelet an Snout toeye . 2 . [Wve Kye diameter, vertical, 23 inches: horizontal ; 23, Hye to eye over curve ao heads): é : : 7,9) aa Hye to angle of mouth : : | Ores Base of D. to 5th gill-cleft (lower margin) gS oe See Nostril to eye . : tae 5th gill-cleft, dorsal to oe length j CF inHg Distances separating dorsal ends of gill- slits (over curve of back), Ist, 5 inches; 2nd, 9 inches; 3rd, 15 inches; 4th, 22 inches; 5th, 29 inches. The symphysis of the upper jaw at the vertical from eye. . Mucus Canals.—Snout with large mucous pores, mostly transverse, a faint indication of longitudinal arrangement where the rostral and prenasal canals might be expected to lie. No other canals visible externally. Rostral Cartilages—There is a triangular ventral projection, with a groove on its upper surface; and slender, gently curved, dorsal processes (fig. 1). Teeth.—In the upper jaw the bands of teeth are separated by a wide bare space. The teeth in the first two or three rows (nearest the middle line) are triangular with broad bases; the successive rows show a gradual transition to the more conical and terete form of tooth found throughout the rest of the dental band (Plates VI and VII). Fam. ORECTOLOBIDAR. Stegostoma tigrinum (Gmelin). — Algoa Bay (Port Elizabeth Museum); St. Francis Bay (C. L. Biden); Durban (H. W. Bell-Marley, presented to the South African Museum, 1935). 46 Annals of the South African Museum. Fam. HARRIOTTIDAE. Whether or not Rhinochimaera be regarded as a synonym of Harriotta, the latter is the earlier genus, and the family name should be in accordance. Harriotta punnata Schnakenbeck 1929. Walfish Bay. Fam. ALEPOCEPHALIDAE. Bathytroctes rostratus Brnrd. 1925, non Gnthr. = Alepocephalus barnardt Norman 1930, p. 270. Fam. GONORHYNCHIDAE. Dr. Chabanaud informs me that an examination of type specimens in the Paris Museum has shown that the Cape species is correctly termed gronovw C. and V. It is an endemic species and is not conspecific with Australasian and Japanese species. Fam. STOMIATIDAR. The correct date for Astronesthes boulengert Gilch. is 1902, not 1904 (see Monograph, p. 1032, footnote). Two specimens have been captured by trawlers belonging to Messrs. Irvin & Johnson, Cape Town, and presented to the South African Museum. Both were taken off Dassen Island, north of Table Bay, in 200-300 fathoms; one on 8th July 1925, the other 2nd July 1936. For Astronesthes capensis G. and von B. see Regan and Trewavas, 1929S pn 30: For Borostomias richardsom see Regan and Trewavas, 1929, p. 25. The South African examples should be re-examined to see whether they belong to Borostomias or Astronesthes. Idiacanthus ferox (Gnthr.) should read J. fasciola Peters, 1876. See Regan and Trewavas, 1930, p. 129. The same authors reduce Neostomias to the rank of a subgenus of Eustomias, loc. cit., p. 73. Throughout this paper the name of Gilchrist’s species is spelt “filofer” instead of filiferum. Gonostoma grandis Coll.=G. bathyphilum (Vaill.); see Norman, 1930, p. 285. The genus Yarella stands, but Y. africana G. and von B. becomes a synonym of Y. corythaeola (Alcock); see Norman, 1930, p. 289. Maurolicus pennants (Walb.) becomes M. muellert (Gmelin); see Norman, 1930, p. 298. Further Notes on South African Marine Fishes. 47 Gen. Haplostomias R. and T. 1930. Regan and Trewavas, loc. cit., p. 109. Separated from Melanostomias by having the teeth simple, not bicuspid. The specimen described below is a 2, and the question may be raised whether simple and bicuspid teeth are merely characteristic of the 2 and ¢ respectively. If so, the genus Haplostomias is un- necessary. Haplostomias tentaculatus R. and T. 1930. Regan and Trewavas, loc. cit., p. 109, pl. x1, fig. 1, and text- figs. 105a, 106a. Depth 54 length of head 7, in length of body. Hye subequal to snout, 6 in length of head. Teeth: about 18 in upper jaw, Ist and 3rd short, 2nd and 4th long, 4th longest, 5th—-7th subequal, 8th onwards small and decreasing in size posteriorly; 12 in lower jaw, 2nd, 4th and 5th smallest, 3rd, 7th and 8th largest, last 4 increasing slightly in length posteriorly (the two sides in both jaws not quite symmetrical); 1 on each vomer, and 4-5 on each palatine, 2 pairs on tongue; none of the teeth bicuspid. Gull-rakers 8-9 small single or double spines on anterior arch. D16. A118. P5. V7. Branchi- ostegals 10 or 11 (as nearly as can be counted). Photophores, except the suborbital one, very indistinct. Barbel about twice length of head, ending in an elongate ovate bulb, with the black stem continued as a filament half the length of the bulb. Length.—235 mm. Colour.—Black, the bulb at end of barbel dull orange. Locality — Off Table Bay, 185 fathoms. Distribution.—Caribbean Sea. The specimen is a 2 containing ripe or nearly ripe eggs, and was caught in a trawl about the middle of November 1928. In spite of certain small differences (proportions of body, relative sizes of teeth) this specimen is referred to tentaculatus. These differ- ences may be due to age, as the specimen is more than twice as large as the “Dana” specimens described by Regan and Trewavas. Gen. Echiostoma Lowe. 1843. Lowe, Proc. Zool. Soc. Lond., p. 87. 1883. Gill, Proc. U.S. Nat. Mus., vi, p. 256 (Hyperchoristus). 1895. Goode and Bean, Ocean. Ichth., p. 108. 1930. Regan and Trewavas, loc. cit., p. 116. 48 Annals of the South African Museum. Body elongate, compressed, naked. Eye moderate. Mouth large, straight. Teeth on jaws typically large, depressible; teeth also on vomer, palatine and tongue. Barbel well developed. Pectoral small, the uppermost ray elongate, free. Ventral short, far behind middle of body. Dorsal and anal opposite, far back. No adipose fin. Caudal forked. Pseudobranchiae absent. A large suborbital photo- phore and two lateral rows. Gill-rakers minute. The first record of this genus in South African waters. Echiostoma tanneri (Gill). 1883. Gill, loc. cit., p. 256 (Hyperchoristus tanneri). 1895. Goode and Bean, loc. cit., p. 109, fig. 130 (barbatum, non Lowe). 1927. Parr, Bull. Bingham Ocean. Coll., iii, p. 53, fig. 31 (barbatum, non Lowe). 1930. Regan and Trewavas, loc. cit., p. 117, fig. 113 (barbels). Depth of body equal to length of head, 64 (smaller)—74 (larger specimen) in length of body. Eye 14 in snout, 7 in length of head. Teeth: 12 plus a number of minute ones on hind part of maxilla, 2nd or 3rd, 4th and 5th largest; about 12 in lower jaw, 3rd, 6th and Tth largest (the two sides not quite symmetrical), the posterior teeth in a double or triple series; 1 on each vomer, about 7 on each palatine, 2 pairs and 2 small single teeth on tongue; most of the teeth minutely bicuspid. Gill-rakers: 7—9 pairs of minute spines on anterior arch. D12. A116. P1+3, the free ray 1#-2 times length of head, arising in front of (neither above nor below) the. bases of the 3 short rays. V 8. Branchiostegals 10 or 11 (as nearly as can be counted), with a series of 10 photophores (one between each pair of branchiostegals), very distinct in the smaller specimen, but only the hinder ones distinct in the larger specimen. Cuneiform postocular photophore 14 times the eye; 2-3 photophores on opercle; lower series of photophores beginning on isthmus: isthmus to pectoral 8+2, pectoral to ventral 26-27, ventral to beginning of anal 14-15+2 along anterior base of anal; lateral series opercle to ventral 24-25, ventral to beginning of anal 16, anal to caudal 12. Regan and Trewavas’ figure (of barbatum) shows the ventral series continuous along base of anal to caudal; here the ventral series has 2 photophores along anterior base of anal and then stops, while the lateral series curves evenly down to posterior end of anal and continues along lower side of caudal peduncle. Barbel rather shorter than head, ending in two bulbous Further Notes on South African Marine Fishes. 49 swellings, with several short filaments, corresponding with that of a fish 223 mm. in length figured by Regan and Trewavas. An anteorbital spine with 3-4 small denticles, a postorbital spine, 2 temporal spines, interorbital with 2 denticulate ridges beginning far apart at a level a little behind eye, and converging forwards, meeting in a single short spine at a level just in front of nostrils; a few other scattered denticles on upper part of head. All these spines more or less concealed by the loose skin, the interorbital crests very distinct in the larger specimen. A large mucus pore on each side of interorbital, just within (medial to) the denticulate ridge; large suborbital and temporal pores. Length.—200 mm. and 305 mm. Colour.—Dark slaty-black, with small black dots on head and similar dots in vertical band-like series along the body to tail (ef. Giinther’s figure of Opostomias on pl. lin, fig. 4. of Challenger Rep., xxi, and Parr’s figure 32 of H. ctenobarbus, and as described by Giinther for barbatum); each of these black dots (photophores, as described by Giinther) with a pale translucent centre; fins whitish, barbel greyish, postocular photophore pink, ventral and lateral series of photophores lilac or amethyst. Locality.— Offi Table Bay. Distribution.—Gulf of Mexico, Caribbean Sea. The smaller specimen was captured in the trawl at a depth of 185 fathoms in November 1928, together with the specimen of Haplo- stomias mentioned above, and the larger specimen in September 1935, depth probably also about 200 fathoms. The South African Museum is indebted to Messrs. Irvin & John- son for both these specimens, the larger of which is the largest specimen of this genus yet recorded. Fam. ANGUILLIDAE. Anguilla mossambica (Peters). 1925. Barnard, Ann. 8. Afr. Mus., xxi, p. 175 and 1927; «bhd., p- 1018. 1935. Id., Rep. S. Afr. Mus., 1934, p. 10. An elver of this species was caught in the estuary of the Uvongo River, near Port Shepstone, Natal, by Mr. L. A. Day of the Inland Fisheries Survey. The specimen was handed to Mr. A. C. Harrison, from whom I received it, without further data as to time of year when caught. 50 Annals of the South African Museum. It is 49 mm. in length, with 103 myomeres. It thus falls within the limits found for this species by the late Dr. J. Schmidt, viz. 100-105, as I am informed by Dr. .V. Taning (in litt. 9/2/35). Fam. MyRipae. Gen. Muraenichthys Blkr. 1853. Bleeker, Verh. Batav. Gen., xxv, p. 71. 1916. Weber and de Beaufort, Fish. Indo-Austr. Archip., 11, p. 274. Hlongate cylindrical, vermiform. Scales absent. Vent before middle of length. Dorsal arising far behind gill-openings. Dorsal, anal and caudal confluent. Pectorals absent. Snout somewhat pro- jecting. Nostrils on margin of upper lip, the anterior tubular, the posterior at base of a flap. Teeth on vomer and jaws. Tongue adnate. Lateral line present. Gill-openings small. Muraenichthys gymnotus Blkr. 1857. Bleeker, Act. Soc. Sc. Indo-neerl., ii, p. 90. 1864. 7 Atl. Ichthyol., iv, p. 33. 1871. Klunzinger, Verh. Zool. Bot. Ges. Wien., xxi, p. 608. 1916. Weber and de Beaufort, loc. cit., p. 277. 1934. Barnard, Ann. Mag. Nat. Hist. (10), xi, p. 230.. Fam. MycToPHIDAE. Loweina, a new subgeneric name to include Myctophum rarum. Fowler, Amer. Mus. Novitat., No. 162, p. 2, 1925. Lampanyctus warming (Liitken) = L. townsend: Hig. and Hig. 1889. Lampanyctus argenteus Gilch.=L. hectoris Guthr. 1876. Fam. HEMIRHAMPHIDAER. Euleptorhamphus longirostris (Cuv.). 1925. Barnard, Ann. 8. Afr. Mus., xxi, p. 264 (references). 1928. Fowler, Mem. B. P. Bishop Mus., x., p. 74. 1936. » Bull. Amer. Mus. Nat. Hist., lxx, p. 432 (quotes the South African record under name of E. viridis (van Hasselt)). A specimen of this genus, the second specimen to be recorded from South African waters, was brought in a perfectly fresh condition to the S. Afr. Museum in June 1936 by a fisherman, who stated that it had been caught somewhere “in Table Bay.” Further Notes on South African Marine Fishes. 51 Depth of body 20, length of head 24, length of lower jaw 34, in length of body. Eye 34 in distance from tip of snout to hind margin of operculum, subequal to interorbital width and to snout, 2 post- orbital part of head. Longitudinal length of preorbital scarcely more than 4 eye-diameter. D 23, arising slightly in advance of anal fin (above anterior margin of vent). A 22. V inserted at a distance in front of anal fin equal to the distance from anterior margin of nostril to hind margin of operculum; its length subequal to eye-diameter, and 74 in length of uppermost pectoral ray. P 9, the lowermost ray very short, uppermost ray not quite twice the distance from tip of snout to hind margin of operculum. Lower caudal lobe longer than the upper lobe. Scales nearly all lost. Bright silvery, the back duller, greenish-black, lower jaw bluish- black, fins pale, caudal with suffused hind margin. Length.—355 mm. to end of mid-caudal rays, or 375 mm. to end of lower caudal lobe. The tip of lower jaw appears to be not quite complete. Remarks.—The differences between this specimen and the Walfish Bay one described in 1925 are not to be taken as indicating specific differences, because the latter specimen is in a very poor condition, scarcely suitable for a critical study. On the contrary, it is a reason- able assumption that the two are conspecific. Jordan and Evermann (1905. Hawaiian Fishes) seem to be the only authors who have had adequate material on which to base a specific diagnosis. Whether van Hasselt’s name deserves priority I have no means of determining. Fam. CoRYPHAENOIDIDAR. Whitley (1931, Austr. Zool., vi, p. 334) proposes Fuyangia in place of Chalinura G. and B., preoccupied. Fam. BERYCIDAE. Beryx splendens Lowe. 1833. Lowe, Proc. Zool. Soc. London, p. 142. 1843-60. Id., Fishes of Madeira, p. 47, pl. viii. 1895. Goode and Bean, Ocean. Ichth., p. 176, fig. 197. 1924. Roule, Bull. Mus. Paris, No. 1, p. 73. 1934. Barnard, Ann. Mag. Nat. Hist. (10), xin, p. 230. Distinguished from the other South African species, B. longipinnis 52 Annals of the South African Museum. Brnrd., by the more oblong body, fin and scale formulae, and by the absence of the filamentous rays of the dorsal and ventral fins. Depth 23, length of head 3, in length of body. Eye 24 in length of head, twice length of snout, which is slightly longer than inter- orbital width. No spine on nasal or lower jaw; preorbital spine without accessory spine. D IV 13. A IV 27. V I11. Longest rays scarcely reaching to origin of anal. P 17, 2 length of head. Scales scabrous, with smooth median longitudinal groove; 1.1. 77, Ltr. 5. Length.—270 mm. Colour.—As in longipinnis. Locality.—45 miles off Table Bay. Distribution.—Madeira, West Indies, Japan. A single specimen was caught in the trawl in December 1932, and presented to the South African Museum by Captain Taylor. Fam. MELAMPHAIDAE. See Norman, 1929. Plectromus is a synonym of Melamphaes Guthr. 1864. M. mizolepis stands; coronatus, inadequately described by Gilchrist and von Bonde, is regarded as possibly the same as unicornis Gilb. 1905; and macrophthalmus Gilch.=megalops Liitken, 1877. Heterosomata. Norman (1931, p. 508) vindicates Pseudorhombus natalensis Gilch. as a species distinct from arsius (=russelli), and makes the follow- ing changes :— Laeops microphthalmus von Bonde, transferred to Arnoglossus. Lambdopsetta kitharae von Bonde, non Smith and Pope, redescribed as Laeops natalensis Norman. Barnardichthys Chabanaud 1927, new genus to include Solea fulvomarginata Blegr. S. quadriocellata von Bonde=Quenselia ocellata (Linn.); see Chabanaud (1930, p. 12). S. melanoptera Gilch. and S. capensis Gilch. =Synapturichthys kleini Bonap. subsp. variolosa Kner. 8S. alboguttata Fowler 1929 seems to belong here also; see Chabanaud (1930, p. 10). S. ampar=Pegusa lascaris (Risso). Synaptura barnardi Smith 1931. Gt. Fish Point. Close to marginata Bler. Achirus = Heteromycteris Kaup. Further Notes on South African Marine Fishes. : 53 Coryphaesopia Chab. 1930, new genus for A. cornuta Kaup. The South African form is described as a n. subsp. barnardi Chab. 19384. Pseudaesopia Chab. 1934, new subgen. of Zebrias for Aesopra regan (Gilch.). Revised nomenclature of South African Flat-fishes in the sequence adopted by Norman (Monogr. Flat-fishes, vol. i, 1934) (equivalents in Barnard, 1925, in brackets). Norman’s vol. 1, containing the Solerdae, not yet published. Psettodidae. Psettodes erumez (Bl. Schn.). Bothidae. Pseudorhombus arsius (Hamilton). (P. russellit.) a natalensis Gilch. (P. russellit.) Citharoides marcrolepis (Gilch.). (Paracitharus m.) Arnoglossus capensis Bler. 3, dalgleisht (v. Bonde). (Trichopsetta d.) » mocrophthalmus (v. Bonde). (Laeops m.) Engyprosopon grandisquama (Temm. and Schl.). (Scaeops g.) 5 natalensis Regan. (Crossorhombus dimorphus part, Juv.) Crossorhombus valde-rostratus (Alcock). (C. dimorphus part, adult.) Bothus pantherinus (Riippell). » mancus (Brouss.). » ovalis (Regan). (Platophrys circularis.) Chascanopsetta lugubris (Alcock). (C. gilchristi.) Laeops mgromaculatus v. Bonde. ,» natalensis Norman. (Lambdopsetta kitharae v. Bonde, non Smith and Pope.) » pectoralis (v. Bonde). (Lambdopsetta p.) Pleuronectidae. Poecilopsetta natalensis Norman. (Limanda beani v. Bonde, non Goode and Bean.) Marleyella bicolorata (v. Bonde). (Poecilopsetta b.) Paralichthodes algoensis Gilch. Samaris ornatus v. Bonde. » delagoensis v. Bonde. (Norman: “perhaps identical with cristatus.’’) 54 Annals of the South African Museum. Soleidae. Dicologlossa cuneata (Moreau). (Solea senegalensis, non Kaup.) Barnardichthys fulvomarginata (Gilch.). (Solea f.) Quenselia ocellata (Linn.) (Solea quadriocellata.) Synapturichthys kleont Bonap. variolosa Kner. (Solea melanoptera and capensis.) Pegusa lascaris (Risso). ? (Solea impar.) Solea bleekert Bler. », turbyner Gilch. Heteromycteris capensis (Kaup.) (Achirus c.) Pardachirus marmoratus (Lacep.). Synaptura marginata Bler. x barnardi Smith. (Close to marginatas.) Austroglossus pectoralis (Kaup). a mocroleps (Blkr.). Zebrias (Pseudoaesopia) regani (Gilch.). (Aesopia r.) Coryphaesopia cornuta (Kaup), subsp. barnardi Chab. (Aesopia c.) Paraplagusia marmorata (Blkr.).. Cynoglossus lida (Blkr.). durbanensis Regan. gilchristi Ogilby. Ls hunterr v. Bonde. Arelia attenuata (Gilch.). Areliscus marleyr (Regan). co ecaudatus (Gilch.). Trulla capensis Kaup. » microphthalmus (v. Bonde). Symphurus variegatus (Gilch.). strictus Gilbert. ocellatus v. Bonde. 39 99 99 Fam. CARANGIDAE. Trachinotus baillonia (Lacep.). 1934. Fowler, Proc. Ac. Nat. Sci. Philad., Ixxxvi, p. 452. Fowler accepts the synonymy of russelli and oblongus with Lacépéde’s species. Fowler’s remarks at the end of his description of this species are difficult to understand. I did not place Fowler’s 1919 description of Caestomorus glaucus Linn. under Trachynotus glaucus of Bloch, but under glaucus of Linnaeus, clearly stating “not glaucus Bloch,” Further Notes on South African Marine Fishes. 5D as reference to my monograph, p. 553, will show. C. glaucus Linn. was not later placed in a new genus Campogramma by Regan (1903, Ann. Mag. Nat. Hist. (7), xu, pp. 348 sqq.); in fact, neither glaucus Linn. nor glaucus Bloch were placed in this genus. Fam. HISTIOPTERIDAE. Quinquarius capensis (C. and V.). (Plate VIII.) 1927. Barnard, Ann. 8. Afr. Mus., xxxi, p. 623. 1935. Fowler, Proc. Ac. Nat. Sci. Philad., IxxxVii, p. 398, fig. 26. For the opportunity of including a record of this species in this paper I am indebted to Mr. R. Orpen, who forwarded a specimen picked up on the beach at Port Nolloth, Namaqualand, by Miss Irene Carstens in October 1935. Since the original description of a single specimen by Cuvier and Valenciennes in 1829, only two other specimens have been recorded. One was taken by the Government Survey vessel s.s. Pickle off the coast of Natal on 28th December 1920. No description of this speci- men was given, which is all the more unfortunate since the present specimen differs in a rather important feature from the original description. One cannot say whether Cuvier’s specimen or the present one is abnormal. The other specimen, described by Fowler, also came from Natal. The main features of the Port Nolloth specimen are as follows:— Total length 63 mm. Depth 14, length of head 23, in length of body (caudal excl.). Width of the flat space between the ventral fins a little more than 14 in the depth of body, subequal to the distance between the bases of ventral spines and the somewhat pro- jecting knob on the throat (Giinther, 1859, Cat. Fish. B.M., 1, p. 212, seems to have reckoned the narrowest width between the ventrals, and the longest medio-ventral distance extending to the angle of the branchiostegals in C. and V.’s figure), and a little more than the distance between ventral spine and vent. Hye equal to snout, 3 in length of head. Depth of caudal peduncle (between ends of bases of dorsal and anal fins) slightly greater than its length. A band (4-5 rows) of fine conical teeth on both jaws, the outer ones slightly larger, and a few more granular ones in an oval patch on vomer: apparently none on palatines. Pseudobranchiae well developed. Gill-rakers 6 on upper, 16 on lower part of anterior arch, the longest ones about 4 eye-diameter. The bony scute (lst suborbital) has a VOL. XXXII, PART 2. 5 56 Annals of the South African Museum. short acute forwardly directed spine on its anterior margin. The spine on the median occipital scute is directed backwards. D XIII 12, 1st spine short, $eye-diameter and 3} in 2nd spine, 4th and 5th longest, a trifle more than 2 eye-diameters, 13th spine a little shorter than the rays which are subequal to eye. A VI 7, Ist, 4th and 6th spines subequal, slightly less than eye, 5th shorter, 3rd a little longer, 2nd longest, 14 eye-diameter. V I 5, the spine a little over 2 eye-diameters. P 17, Ist ray short, 4th and 5th longest, 17th very short. C17 (+3), rounded-truncate. Lat. line 46-47 (35-36 to where the line becomes horizontal on. the caudal peduncle). Cheek-scales 5 vertical rows (across the deepest part of the patch). Body silvery with blackish-brown markings almost exactly as in C. and V.’s figure, but the dark irregular bands on the flanks broader; a faint yellowish tinge on hinder part of body and on caudal peduncle; the flat ventral surface silvery-grey, with a white band between the bases of the ventrals and a pale area around the vent; opercle and cheek silvery; spinous dorsal and ventrals dark; soft dorsal, caudal and anal faintly yellowish; pectoral transparent. The specimen thus follows the description and excellent figure given by the French authors, except as regards the extra spine in the dorsal and anal fins. This is an interesting feature in view of the number of spines found in allied species. Pentaceropsis recurvirostris : . DX-XT 14. A III 11 (total) D 24-25. A 14. Pseudopentaceros richardson . . (type) D XIV 10. A IV 8. D 24.. A 12. (juv.) D XIV 9-10. AIV 8-9. D 23-24. A 12-13. Quinquarius japonicus . : . DX 14. AVES: D 25. A 14. capensis. . (type) D XIT 12. AV 7. D2.’ AT (Port Nolloth) D XIII 12. A WET. D 25. A183. Quadrarius decacanthus : - DX 1s: A IV 8. D 23. AY: hendecacanthus . - D Xt 13-15. A IV 10-11. D 24-26. A 14-15. So far as concerns the dorsal and anal fins all these species are closely allied, and Jordan’s arrangement (1907) seems a little artificial. McCulloch has already (1915) made Quadrarius a synonym of Quin- quarius, as the only difference seems to be in the number of anal spines; and suggested that, though possibly the latter may be synonymous with Pseudopentaceros, there are other differential characters. Further Notes on South African Marine Fishes. 57 The shape of the body and the relative length to breadth of the space between the throat and the ventral fins seem to be good characters. McCulloch’s figure of the young of P. richardsoni (1923, Rec. Austral. Mus., xiv, p. 18, pl. iv, fig. 1) shows that the body shape does not alter much between the young and the adult. Such is also the case in QY. hendecacanthus McCull. (1915, ‘‘ Endeavour, ”’ Sei. Res., i, p. 144, pl. xxvi). But one feature does alter with age, viz. the gradual disappearance of the spines on the head. Con- sequently we can regard the type of @. capensis and the present specimen as juveniles, and we may expect the adult, when captured, to have more or less the same body-shape, but to have no spines on the scutes on the head. The 180 mm. specimen described by Fowler is interesting, as it has only 4 anal spines (as in Pseudopentaceros and Quadrarius) in association with 12 dorsal spines (as in the type of Q. capensis). There is also considerable resemblance between Fowler’s figure and McCulloch’s figure of the 70 mm. young P. richardsoni (loc. cit.), and I do not feel fully satisfied that Fowler’s specimen should not rather be identified as richardson. Smith’s figure of the 525 mm. type of the latter species may be possibly not quite accurate as to shape of body and head, if the figure, like the description, was taken from the dried skin. When more material is available it may be possible to decide the true relationship between P. richardson and Q.. capensis. In view of the excellence of the original figure there is no need to figure the present specimen. But as a matter of interest an illustra- tion from an early work on fishes is here reproduced. This is the frontispiece to “The Naturalist’s Library: Ichthyology, vol. i, by Sir William Jardine, Bart.”’, published in Edinburgh by W. H. Lizars & Stirling & Kenney in 1835, in which a very accurate copy of Cuvier and Valenciennes’ figure is engraved against a vignette of Table Bay. The following paragraph appears on page ix of the publishers’ advertisement :— ‘““We have introduced as back grounds, wherever they could be procured, the scenery which the fishes frequent. In many cases, the _ landscapes are representations of real views, which will be discovered upon examination of the plates; the figures upon which, . . . have been taken from the magnificent work upon this subject, by the Baron Cuvier and M. Valenciennes, whose representations are so perfect, as to leave little to be desired in the way of improvement.” In this instance the only “improvement” has been the addition of patches of pale pink on the pre- and post-orbital scutes, and of 58 Annals of the South African Museum. pale yellow on the opercle, below the pectoral fin, and on the flank (Plate VIII). Fam. LuTIANIDAE. Lutianus duodecimlineatus (Val.). 21802. Lacépeéde, Hist. Nat. Poiss., 11, pp. 430, 477, pl. xxi, fig. 2 (Labrus octovittatus). 1830. Cuvier and Valenciennes, Hist. Nat. Poiss., vi, p. 529. 1874.* Bleeker in Pollen and van Dam. Faune Madagasc., pt. 4, Poiss., p. 27, pl. ix, fig. 1 (coloured) (octovittatus, non Lacép.). 1927. Barnard, Ann..S. Afr. Mus., xx, p. 653, p. 27) We: sau(aeuee Sauvage). 1931. Fowler, Bull. U.S. Nat. Mus., No. 100, x1, p. 158. 1934. Id., Proc. Ac. Nat. Sci. Philad., Ixxxvi, p. 466, fig. 38. Fowler (1934) makes the following statement: “The colour pattern as described by Barnard . . . is therefore incorrect as my figure and description show.” This is a non sequitur. I have re-examined the specimens and can state that my description of the pattern is correct for the specimens in the South African Museum. Except for the absence of the common base of the 3rd and 4th streaks, Sauvage’s figure agrees. As a matter of fact, Fowler’s figure also agrees, except that in my specimens the streaks are no longer traceable on the head. It would perhaps have been clearer if after the word “behind” [the opercle] in my description the following words had been used: “ behind the head” or “‘behind a line drawn more or less vertically from the upper part of opercle.” Bleeker’s identification of this form as Lacépéde’s octovittatus is not accepted by Fowler (1931). Gen. Htelis Cuv. 1828. Cuvier and Valenciennes, Hist. Nat. Poiss., 1, p. 127. 1931. Fowler, Bull. U.S. Nat. Mus., No. 100, vol. xi, p. 193. Distinguished from Pristipomoides by the shortness of the hinder spines in the spinous dorsal fin, causing a concavity in the margin; the spinous and soft portions, however, are not separate. * The South African Museum copy has 1875 on the outer cover and 1878 on the title-page. The British Museum Library Catalogue gives 1874. Further Notes on South African Marine Fishes. 59 Etelis carbunculus Cuv. 1828. Cuvier, loc. cit., p. 127, pl. xvi. 1905. Jordan and Evermann, Bull. U.S. Fish. Comm., xxii, p. 242, pls. xviii and xxxviii (Htelis evurus). 1928. Fowler, Mem. B. P. Bishop Mus., x, p. 193, pl. xvu, fig. A. feat Id., loc. cit., p. 195. Depth 33, length of head 34, in length of body. Eye 44 in head, 14 in snout. Snout subequal to interorbital, 3in head. Least depth of preorbital half the (longitudinal) diameter of eye. Cardiform bands of teeth in both jaws, with well-spaced small canines in outer row; teeth on vomer and palatines, but none on tongue. Gill-rakers, 14 on lower part of anterior arch. D X11. Ist spine very short, 2} in eye, 6 in 2nd spine, 3rd spine slightly longer than 4th, nearly 2 in head, 10th spine 14 in eye; last ray prolonged, half as long again as penultimate ray. A III 8, last ray not prolonged. Caudal forked, but lobes not prolonged. Seales: 1.1. 51; ltr. 6 to base of Ist dorsal spine (including the uppermost narrow sheath-like scale), 13 to base of ventral fin, 16 to middle line of belly. Cheek scales 6-7. Axillary scale of ventral fin present, but more or less concealed under an ordinary body scale. 800 mm. As preserved, silvery, most of the scales on upper part of body with a rose or pale crimson centre. Locality—Ofi Bashee River mouth. Caught by Mr. H. Smedsvik, of Seafoods Successors, Durban; and forwarded for identification by KH. C. Chubb, Esq., Curator, Durban Museum, August 1936. Remarks.—I have little doubt that this specimen should be identified with Cuvier’s species, which is known from Mauritius, Réunion, Seychelles, Japan, and the Hawauan Islands (and the West Indies). It does not entirely agree with the published descriptions. It is a male. Aetiasis n. g. Teeth on jaws conical, the front ones moderately large, the hinder ones small, an irregular row of small teeth internal to the outer row in front; small conical teeth on vomer and palatines; tongue smooth. An inconspicuous flat spine on opercle. Preopercle entire. Top of head, cheeks, and maxilla scaly. Caudal fin, but not dorsal and anal, scaly. Dorsal not deeply notched, spines 12. Anal rays 8. Dorsal and anal rays not produced. Pectoral subequal to head, scarcely falcate, scaly at base. No transverse groove between inter- orbital and occipital. 60 Annals of the South African Museum. The fish for which this genus is proposed does not seem to fit in with any of the other genera of the family, though the complaint (auTLacis) is justified that there is already a large amount of synonymy among the Indo-Pacific species, which await a full and authoritative revision. Aetiasis cantharoides n. sp. (Text-fig. 2.) Depth about 3, length of head 332, in length (excl. caudal). Hye (long diameter) subequal to snout, 4 in length of head. Interorbital Fic. 2.—Aetiasis cantharoides n. g., n. sp. a, Head; 6, inner view of right dentary bone, cleaned to show teeth; c, ventral view of right premaxilla, cleaned; d, dorsal view of right dentary, cleaned; e, vomer and palatines; f, left upper pharyngeal, anterior end above; g, left lower pharyngeal, anterior end below (inner margin (median line) to left in both f and g). width slightly more than 14 times eye. Preorbital (opposite end of maxilla) + the long diameter of eye. Greatest depth of caudal peduncle (between ends of bases of dorsal and anal fins) subequal to length along its dorsal profile, 14 times length of its ventral profile, least depth (at base of caudal) twice in greatest depth. A very small notch near the angle of preopercle, which is not denticulate. One — short flat spine (not prominent) on opercle. D XII 7 (possibly XIII 6, the posterior spines and anterior rays Further Notes on South African Marine Fishes. 61 being broken), spines slender, Ist ? the long diameter of eye, 2nd about twice the Ist, 3rd longest, twice the long diameter of eye, following spines successively shorter, ultimate ray more slender and shorter than the penultimate ray, which is the longest ray and equal to the 4th spine. Rays not scaly. Apparently not notched between spinous and soft portions. A III 8, spines stouter than the dorsal spines, 1st 14 times in the 2nd, which is subequal to the long diameter of eye, and 12 times in 3rd, ultimate and penultimate rays subequal, not elongate, 14 times the eye diameter. Rays not scaly. P 17, subequal to length of head, scaly at base, the uppermost 2 rays simple. . V 15, with axillary and medio-ventral enlarged scales. Caudal forked, scaly. Maxilla without supplementary bone, exposed, extending to below anterior margin of eye, its posterior width subequal to diameter of pupil, with 2 series of scales. Lips and gums villous, the teeth in consequence being concealed, even the larger canines not being clearly visible. Teeth small, conical, in a single row laterally on lower and upper jaws, becoming a double row in front, and with an outer series of enlarged conical teeth in front, 6 in upper jaw, 10 in lower jaw. A few small conical teeth in an oval patch on vomer, and a single row (in some places anteriorly double) on each palatine. Tongue smooth. Upper and lower pharangeals with conical teeth in centre, passing into slender curved spiniform teeth around margins. Branchiostegals 6, the anterior 4 slender, the posterior 2 stout. Gill-rakers 19-20 on lower part, 6 on upper part, of anterior arch. Pseudobranchiae present. Scales minutely ctenoid. LJ. 72; ltr. 9 (1st dorsal spine to 1.1.), 18 (1.1. to spine of ventral fin). Seven series on cheek. About 24 predorsal scales beginning at about vertical from anterior third of eye. Opercle and subopercle scaly. Length.—400 mm. Colour (as preserved).—Brownish, silvery on cheeks, opercles, and belly. Locality.—Natal coast (one specimen secured by the kindness of Mr. C. L. Biden). This fish bears a strong resemblance to a species of Pachymetopon (olum Cantharus), e.g. P. aeneum. In fact, it was included in the material sent to Mr. Norman for his revision of the Spondyliosome- 62 Annals of the South African Museum. and allied fishes (Ann. 8. Afr. Mus., xxxii, p. 5, 1935), and I am indebted to Mr. Norman for suggestions as to its true position. Fam. SCORPIDIDAR. Neoscorpis Smith 1931, new genus for Scorpis lithophilus. Fam. SPARIDAE. Dentex macrophthalmus (B1.), Schnakenbeck 1929, p. 25, Walfish Bay. Fowler (1925, Amer. Mus. Novit., No. 162, p. 4) places this species in a new subgenus Opsodentex. In the same paper (p. 4) Fowler proposes a new subgenus Husalpa for Box salpa, but in 1934 (Proc. Ac. Sci. Philad., Ixxxvi, p. 472) records this species as Sarpa sarpa Linn. Revision of Spondyliosoma and Pachymetopon: see Norman (1935, Anna. Air: Mug. xxocis ame OEsqd.): Fam. GOBIIDAE. Gobius delagoae n. sp. (Text-fig. 3.) 3 Body elongate. Depth 6%, length of head 43, in length of body. Hye slightly greater than snout, 34 in length of head. Interorbital Fic. 3.—Gobius delagoae n. sp. very narrow, 4 in vertical diameter of eye. Maxilla extending to below anterior margin of eye. Teeth in narrow bands in both jaws, the outer ones somewhat enlarged, a single enlarged curved canine on each side of lower jaw. Tongue truncate in front. Gill-rakers reduced. No flaps on shoulder girdle. Further Notes on South African Marine Fishes. 63 Anterior nostril shortly tubular. No nuchal crest. Large pores; one medial to each pair of nostrils, 4 in a groove running from eye to upper end of gill-opening, one between preopercle and opercle. No pores on chin. Rows of minute papillae as shown in figure. DV1I+I 138, 1st spine half-length of head, 5th twice as long as Ist, produced in a filament. AI 13. P 19, upper rays not silk-like. Ventrals reaching to vent (almost to anal fin). Caudal lanceolate, half as long again as head, middle rays filamentous. Seales cycloid, those on hinder part of body feebly ctenoid. Many of the scales lost, apparently about 55 in longitudinal series, and 16 between dorsal and ventral fins. 75mm. Pale brown with darker mottling, somewhat vermiculate on head, blotchy on body; the body with oblique dark bands running from above downwards and forwards, the first band broad, behind pectoral, followed by 2 narrow bands, then a broad one ending below at vent, then one narrow, one broad, one narrow, the last between middle of soft dorsal and middle of anal; the narrow bands are composed of 2 dark stripes with pale interval, the broad ones of 3 dark stripes. Two dark stripes between eye and upper jaw; the dorsal with a few small black spots, anal dark grey, middle rays of the ventral and of caudal blackish. Locality.—Inyack Island, Delagoa Bay, buried in mud. - Professor C. J. van der Horst, Witwatersrand University, 1935. Fam. CLINIDAE. Fowler (1934, Proc. Acad. Nat. Sci. Philad., lxxxvi, pp. 505-507) records Clinus capensis, swperciliosus, cottoides, and anguillaris, with- out localities, but he says in his introduction (p. 405) that in such cases 1t is to be understood that the specimens came from Durban. I know, however, that Mr. C. L. Biden, when residing in the Cape Peninsula, collected and sent to Mr. Bell-Marley several “ klip-fishes,”’ and it seems not unlikely that the above records are based on these specimens, without definitely specified localities. The presence of these typically Cape species in Natal needs confirmation. Clinus rotundifrons n. sp. (Text-fig. 4.) Body strongly compressed, of similar build to C. fucorum. Depth 33, length of head 4, in length of body. Depth of caudal peduncle subequal to its length. Eye slightly greater than snout, 4 in length 64 Annals of the South African Museum. of head. Snout and interorbital subequal. Profile rather abruptly descending, snout blunt. Lower jaw not projecting, maxilla reaching vertical from posterior } of eye,.lips thin. No barbels. No supra- orbital tentacle. Anterior nostril a short tube, with a sub-bifid tentacle on its hind margin; posterior nostril fringed with 6-7 short Fie. 4.—Clinus rotundifrons n. sp. Head, with anterior nostril further enlarged. lobes. Numerous pores on upper and hinder parts of head, on suborbital and preopercle, and a row of 5 pores on each side of chin. Teeth in a narrow band (3-4 rows) in upper jaw, in a double row in lower jaw, a single transverse row on vomer. Gill-rakers few and feeble, about 5 filaments in each pseudobranch. A hooked process on inner margin of shoulder-girdle. DXXX 8, arising above preopercle, Ist spine a little over 3 times in length of head, 2nd and 3rd longer, 4th and 5th decreasing to 6th, which equals 1st, subsequent spines gradually increasing to the last, which equals 2nd; first 2 rays longer than last spine. AII 22. P13. Third ray of ventral distinct. Further Notes on South African Marine Fishes. 65 Scales minute, cycloid. Lateral line a single row of tubules, about 42. 95 mm. Colour (as preserved) uniform yellowish, centre of eye black. Locality Oudekraal (south of Camps Bay), west side of Cape Peninsula. Professor T. A. Stephenson, July 1934. Although only a single 2 specimen was captured, it seems distinct enough from all the other South African species to justify description. At first sight it looks very like an example of C. fucorum with abnormal blunt snout, but closer examination reveals a number of essential differences. Fam. TRIGLIDAE. Schnakenbeck (1929, pp. 27, 28) records two species of Trigla: lyra Linn. and hirundo Bl. from Walfish Bay. For revision of South African species see Smith (Tr. Roy. Soc. S. Afr., xxi, pp. 321-336, December 1934). Fowler (1934, November 6) describes Lepidotrigla stigmapteron (p. 487, fig. 46), which has the breast scaleless, and seems to be synonymous with natalensis. Gen. Chelidonichthys Kaup. 1927. Barnard, Ann. 8. Afr. Mus., xxi, p. 939. 1934. Smith, loc. cit., p. 328. Although neither Dr. Smith nor myself have compared South African examples of this genus with actual European examples, we are both agreed that in South African waters there are three well- distinguished species. Recently, however, Svetovidov (1936) has attempted to show that capensis, kumu, and queketti are all synonymous with the European lucerna ; queketti being small, and kumu middle-sized, examples. He distinguishes a subspecies kumu from the typical lucerna as follows :— No bluish-white spots on inner side of pectoral fin in adult. Atlantic coasts of Europe, South Africa, Mediterranean, Black Sea : i : . Llucerna. Several bluish-white spots on inner side of pectoral fin. Pacific coasts of Japan, Sea of Japan, Yellow Sea, coasts of Australia, New Zealand, and Tasmania lucerna kumu. To one who has examined a large number of South African specimens, and moreover has seen them landed alive on board trawlers, Svetovidov’s arguments are inadequate and unconvincing. They are based, as least in part, on a misreading of my descriptions; e.g. 66 Annals of the South African Museum. “coloration of the inner side of the pectoral fin in T. capensis and queketti is the same, at [sic=as] in TJ. lucerna.” But capensis has bluish-white spots on the pectoral in both young (200 mm.) and adult, and surely specimens 450-500 mm. in length can be regarded as adult. I have seen examples of capensis from 100-500 mm., kumu 75-300 mm., and quekettc 150-300 mm. in length. When freshly trawled there is no difficulty in distinguishing all three forms, queketti (all sizes) by the scaly breast (a feature ignored by Svetovidov), and the absence of spots on the pectoral, and kumu (all sizes) by the keeled preorbital. In the face of these facts it is a little difficult to see how growth-changes could account for the observed differences. Sexual differences in coloration have not been observed in the South African species. Fam. BALISTIDAE. Cantherines modestus (Guthr.). 1934. Fowler, Proc. Ac. Nat. Sci. Philad., Ixxxvi, p. 510. Fowler has the following paragraph: “Barnard distinguishes his C. arenaceus by the ‘Skin with distinct scales, each of which is granular or spinulose,’ and adds: ‘The character of the skin distinguishes [it] . . . from all the other South African species.’ It is thus evident that this character is also shared by C. modestoides.”’ As written, the sentence seems illogical. My key and description of modestoides (loc. cit., 1927, pp. 957, 958) state clearly that this species has a soft velvety skin. Balistes conspicillum Bloch Schn. 1865. Bleeker, Atlas Ichthyol., v, p. 116, pl. cexxi, fig. 2 (coloured). 1929. Gudger, Bull. Amer. Mus. Nat. Hist., lvim, fig. 3 (p. 499) (reproduction of figure from Pike’s “Mauritius Fishes’’). 1935. Smith, Rec. Albany Mus., iv, p. 232. Dr. C. J. van der Horst, of the Witwatersrand University, collected a specimen at Delagoa Bay. Further Notes on South African Marine Fishes. 67 EXPLANATION OF PLATES. Puate VI. Cetorhinus maximus (Gunner). View of centre of upper jaw and base of snout, after mounting. Prate VIL. Cetorhinus maximus (Gunner). Enlarged view of the beginnings of the dental bands, the figure on the left showing the right side of jaw, and vice versa. Pirate VIII. Quinquarius capensis (C. and V.). Reproduction of the frontispiece to “The Naturalist’s Library,” by Sir William Jardine, 1835. % ist: ’ Bt | : hae #i 7 ies ot Ai re) a Ann. S. Afr. Mus., Vol. XXXII. Plate Vie -- CETORHINUS MAXIMUS. Neill & Co., Ltd. Barnard. Ann. S$. Barnard. Afr. Mus., Vol. XXXII. CETORHINUS MAXIMUS. Plate VII. Neill & Co., Lid. eee LL ir Ann. §. Afr. Mus., Vol. XXXII. Plate VIII. THE 10 Ye. 2s FDINBURGIL FL LIANS SS TIRING SKENNEY, ON SAMUEL HIGHLEY. 22 FLEET STREET : AV CURIN TUNE Rv QUINQUARIUS CAPENSIS. Barnard. Neill & Co., Lid. oan ay ( 69 ) 7. On a New South African Species of Balanoglossus and a Com- parison between rt and Balanoglossus capensis (Gilchrist). By C. J. vAN DER Horst, University of the Witwatersrand. (With 19 Text-figures.) In July 1936 Professor T. A. Stephenson of the University of Cape Town collected some Enteropneusta near Port Elizabeth and sent them to me for identification. The three specimens proved to belong to the genus Balanoglossus, two species of which have been described from the 8. African coast by Gilchrist, viz. B. natalensis and B. capensis. The latter species, the anatomy of which has been described in more detail by von Bonde, occurs in a sexual and an asexual generation, known as B. capensis and B. proliferans (Gilchrist). B. natalensis is a large animal, as yet very imperfectly known, but as I collected some material at Delagoa Bay I hope to give a more accurate description of its anatomy soon. The Port Elizabeth specimens are, however, very small, and as B. capensis also belongs to the smaller species of this genus an accurate comparison of these specimens with B. capensis is necessary. This comparison reveals that the specimens from Port Elizabeth differ in several of their characters from B. capensis as well as from all other species of Balano- glossus; B. australiensis (Hill), however, seems to resemble them most closely. For this reason I am describing the specimens col- lected. by Professor Stephenson as a new species, and have pleasure in naming it Balanoglossus stephensoni n. sp. B. stephensoni was found underneath stones rather close inshore, half buried in muddy sand. For a species of Balanoglossus it is remarkable that it lives under stones; this is the usual habitat of Ptychodera rather than Balanoglossus. Hill mentions that B. australiensis is found in loose gravelly sand, especially under and around larger stones, but otherwise the different species of Balano- glossus make burrows in the sand more or less U-shaped, as described by Stiasny. According to von Bonde B. capensis also occurs generally in coarse sand or gravel. 70 Annals of the South African Museum. This species belongs to the smallest among Enteropneusts, and it is undoubtedly the smallest species of Balanoglossus. The total length of the specimen of fig. 1 is 38 mm. and of that of fig. 2 only 31:55 mm. B. proterogonius, the next largest in size, is about 9 cm. long and B. australiensis 10 cm. Even allowing for a considerable contraction in fixing, the living animal will have a total length of well under 10 cm., probably not more than 5 or 6 cm. In some species of Balanoglossus, as, for instance, B. numeensis, the proboscis is greatly reduced in size, whereas the collar is long ; but a long, well-developed proboscis may also occur in this genus. In B. stephensoni, as far as this can be judged from its contracted stage, the proboscis is of a medium size and the collar is rather short for this genus. As the dimensions of these parts depend largely on the degree of contraction in preserved material, they are not of great importance and may be gauged from figs. | and 2. The anterior part of the collar, as frequently occurs in preserved specimens, is undulated. The five epidermal zones, though present, cannot be seen externally except for the narrow ring along the posterior end of the collar, but even this ring zone cannot always be easily distinguished from the part in front of it. The branchio-genital region is 11-14 mm. long, of which no more than 3 mm. comprises the branchial region proper, so that this region is extremely short. The genital wings are demarcated from the body by a lateral groove (figs. 2 and 13). The posterior end of the collar is directly continuous with the genital wings. This, combined with the fact that the two wings adhered to each other along their free edge by the much swollen epidermis, made it very difficult to expose the branchial region. They also completely covered the anterior liver saccules. The wings are broadest near the posterior end of the branchial region, but on the whole, taking their entire length into consideration, there is not much variation in their breadth as far as the liver region. But in the anterior part of the liver region the wings decrease in size rather abruptly, soon disappearing altogether. In this respect this species obviously differs from B. capensis, in which species, as has been figured by von Bonde, the wings decrease very gradually in size from the posterior end of the branchial region, where they reach their maximum breadth. They extend far into the liver region, but as their disappearance is so gradual it is difficult to ascertain their posterior limit. But even more pronounced is the difference between B. stephensoni and such species as B. biminiensis, where the genital wings end suddenly in On a New South African Species of Balanoglossus. 71 front of the liver region, so that there is a transitional zone between the genital and the liver region. Fig. 1.—Balanoglossus stephensoni. Dorsal view. x4. Fie. 2.—Balanoglossus stephensoni. Side view. x4. The most obvious external character of B. stephensonc is the shortness of the liver region and the abrupt beginning and ending of the external liver saccules. Internally a few more small saccules can be seen both at the anterior and the posterior end of the row, but these do not bulge out on the dorsal surface. In a specimen of VO xxx, PART 2. 6 72 Annals of the South African Museum. B. capensis, 13 cm. long, the liver region measures about 3°5 cm., but as the saccules fade away very gradually at the posterior end it is difficult to give the exact length of the liver region. In Hill’s figure of B. australiensis the liver region is about one-fifth of the total length of the animal, and in this species also the saccules fade away at the posterior end of the row. In B. stephensoni, on the other hand, the liver region comprises only about one-tenth of the total length, but as all three specimens were abruptly bent in the liver region it is difficult to determine the measurements exactly. The number of saccules is accordingly small; there are 20 to 22 in a row, whereas in B. capensis about 75 liver saccules are visible in each row. In both species a few of the liver saccules—4 to 5 in B. stephensonm— have a dark brown colour, whereas those in front and behind are pure white. In general two kinds of liver saccules may be distin- guished in the Ptychoderidae, but in other species of this family all the frontal saccules differ in colour from the caudal ones, whereas in B. capensis and B. stephensona it is: only a few in the middle of the row that exhibit this colour variation. This also corresponds with the observations on their histological structure. The long abdominal region—2 cm. in the specimen of 38 mm.— is characterised by the usual absence of distinguishing features. The caudal region, 3°5 mm. in length, is not swollen, but differs from the abdominal region in the absence or scarcity of the cross-striations. The preserved specimens are of a uniform white colour, except for the few darker liver saccules. Professor Stephenson informs me that the animals, when alive, had almost the same colour, being rather more yellow. Internal Anatomy. Proboscis.—The epidermis and musculature of the proboscis do not show any features that might separate B. stephensoni from other species. Von Bonde has described these tissues fully in B. capensis. Concerning the ventral proboscis septum of B. capensis, my observations do not quite agree with von Bonde’s description. I find this septum greatly reduced in this species. Starting a short distance behind the top of the stomochord there is a thin layer of dorso- ventral muscle fibres just where the septum is to be expected, but no septum in the form of a limiting membrane is visible here. Some- what more caudally a thick blood-vessel, the ventral proboscis artery, runs from the ventral side of the stomochord to the epidermis, accompanied by some muscle fibres (fig. 10), and more caudally there Fic. 3.—Balanoglossus capensis. Cross-section of the proboscis organs. On a New South African Species of Balanoglossus. 73 are some more but thinner blood-vessels, without, however, any trace of a septum. It is only at the level of the ventral pouch of the stomochord that the septum appears, separating the two ventral coelomic blind-sacs; but here too it is not complete, as there is a large hole in it (fig. 3). Nevertheless the septum extends to the caudal end of the blind-sacs, which are not, however, deep, as they cease at the front end of the skeleton. i dill AN \\ s ao (Za 4, , YAS) _ Gin = & | iy LH y ay = pad ty ff Sano Feo Yh py i Fie. 4.—Balanoglossus stephensoni. section of the proboscis organs. lumen of stomochord ; x 66. cbs=central blood x 66. cl=central space ; cl=central lumen of stomochord; epd= epidermis; epv=efferent proboscis vessels; gl= caudal end of glomerulus; p=pericardium ; vll = ventro-lateral cavity of stomochord; vpc = ventral proboscis coelomic blind-sac. epd =epidermis ; efferent proboscis vessel; p=pericardium; pe= right and left part of dorsal proboscis coelom; pv = proboscis vein; vb = ventral blind-sac of stomochord ; vpc=ventral proboscis coelom. In B. stephensoni I find the septum better developed. In front of it and free from it is seen the ventral proboscis artery ; the septum itself begins a short distance behind the top of the stomochord, and from there it is complete and without a hole up to the posterior end of the ventral blind-sacs, that is near the frontal end of the skeleton (fig. 4). There are some smaller blood-vessels running through the septum. In contrast to B. capensis there are no dorso-ventral muscle fibres neither near the ventral artery nor at the sides of the septum, the latter being covered by a layer of coelomic epithelium only. The dorsal part of the proboscis coelom is as usual divided into 74 Annals of the South African Museum. right and left halves by the attachment of the pericardium to the basal membrane of the epidermis. According to von Bonde in one of his series the right dorsal proboscis coelom opens into the end-sac, and further by the proboscis pore to the exterior. This is unusual, as the right dorsal coelom nearly always splits up into smaller branches and ends blindly in the chondroid tissue, whereas the left opens to the exterior. This more usual condition was found by von Bonde in another specimen of B. capensis. In two series of sections of B. capensis which I studied the right coelomic sac was found to open to the exterior, the left splitting up in the chondroid tissue. It therefore seems that this is the normal condition in B. capensis, whereas in other Enteropneusta it is to be considered as abnormal. In B. stephensona the normal condition obtains, the left dorsal proboscis coelom opening to the exterior. But in both species the ectodermal pouch or end-sac with which the coelom communicates and which in its turn opens to the exterior by the proboscis pore shows a bilateral symmetry, to some extent at least. It always extends in an anterior direction at one side, mostly the left, to its communication with the proboscis coelom. In both species, however, I found in addition an anteriorly directed extension of the end-sac on the opposite side, but in B. capensis it is on the left side and in B. stephensoni on the right side ; in both species this extension ends blindly without communicating with the coelom. Hill sometimes found a similar condition in B. australiensis, and von Bonde describes this feature of the end-sac, called by him the proboscis canal, in one of his specimens. In between these two forward extensions of the end-sac lies the connection between the proboscis veins and the sinus venosus (fig. 5). In both species the end-sac is large; in B. capensis it is nearly square in cross-section, in B. stephensoni it 1s greatly elongated dorso- ventrally (figs. 6 and 7). The proboscis pore of B. capensis is small and round and slightly inclined towards the left side. In B. stephen- sont the pore is much larger and the end-sac extends for a short distance beyond the proboscis pore ; a feature not found in B. capensis. The stomochord of B. capensis is very compressed near its anterior end, as has been figured by von Bonde. Im the sections, where the pericardium appears, the dorsal part of the stomochord disappears and the stomochord assumes the usual circular form in the cross- section (fig. 10). There are small isolated cavities in this anterior part, the continuous main lumen appearing only a short distance in front of the ventral extension of the stomochord. Although the On a New South African Species of Balanoglossus. 75 stomochord of B. capensis has the usual swelling or extension at its ventral side there is no blind-sac of the main lumen extending into Ml Lp A aes ¢ S LS 2 yp. LU} a4 59 Cross-section of the proboscis neck. = Fie. 5.—Balanoglossus capensis. x66. cc=collar coelom; epv=efferent proboscis vessel; es=end-sac ; les=forward extension of end-sac at the left side; sk=skeleton; st= stomochord ; svu=sinus venosus. Cross-section of the proboscis neck. sk =skeleton ; Fic. 6.—Balanoglossus capensis. cc=collar coelom; epv=efferent proboscis vessel; es=end-sac ; x 66. phe=top of perihaemal cavity; pp=proboscis pore ; st=stomochord ; sv=sinus venosus. Fie. 7.—Balanoglossus stephensoni. Cross-section of the proboscis neck. x66. ct=chondroid keel of skeleton; epv=efferent proboscis es=end-sac; pp=proboscis pore; sk=skeleton; st=stomo- chord ; sv=sinus venosus. Instead of the vessel ; this swelling, a feature also found by von Bonde. one ventro-median blind-sac there are two ventro-lateral cavities. Usually, as in B. australiensis and B. clavigerus, these ventro-lateral 76 Annals of the South African Museum. cavities communicate with a ventro-median extension of the main lumen. In B. capensis, however, they remain separate as there is no ventro-median extension (fig. 3). More towards the posterior end of the swelling, however, the cavities shift in dorso-lateral direction, where they have the appearance of dorso-lateral cavities of the stomochord (fig. 8), and at the level of the caudal end of the swelling ‘e Gis > Fic. 8.—Balanoglossus capensis. Cross-section Fic. 9.—Balanoglossus capensis. Cross-section of the proboscis organs. x66. cbs=central blood near the anterior end of the proboscis neck. x 66. space; cl=central lumen of stomochord; dll= epd=epidermis; epv=efferent proboscis vessels ; dorso-lateral lumen of stomochord; epd=epi- p=pericardium; pc=dorsal coelomic cavities of dermis; epv=efferent proboscis vessel; p=peri- the proboscis; py=proboscis veins; sk =gskeleton; cardium; pc=dorsal proboscis coelom; sk= st =stomochord. skeleton ; st=stomochord. they communicate with the main lumen (fig. 9). In some species of Balanoglossus (B. carnosus, B. biminiensis, B. jamaicensis, B. numeensis, and B. gigas), which also in other respects are nearly related to each other, there are separate ventro-lateral and dorso- lateral cavities present in the stomochord ; in others (B. clavigerus, B. australiensis, B. stephensont) only the ventro-lateral cavities are found. B. capensis is intermediate between these groups, as the ventro- and dorso-lateral cavities intercommunicate on each side. In B. stephensoni the anterior end of the stomochord is slightly compressed and contains some small isolated cavities. The con- On a New South African Species of Balanoglossus. cei tinuous main lumen appears where the stomochord widens out in a ventral, and especially in a lateral, direction. At this place is found a large ventral blind-sac that is connected with the main lumen by a narrow canal (fig. 4). Where the ventral blind-sac disappears and the body of the stomochord passes over into the neck, the main central lumen widens out considerably. Asin B. capensis the anterior part of the neck is nearly round in cross-section (fig. 6), then it becomes broader (fig. 7). Somewhat more towards the posterior end, where the proboscis neck has already fused with the collar, the dorsal wall of the stomochord of B. stephensoni bulges out in between the perihaemal cavities, which thereby are separated from each other ventrally (fig. 12). Furthermore the lumen of the stomochord sends out a blind-sac into this bulge. This may be a peculiarity of the specimen that was studied in serial sections, but it is more likely that this feature is of specific importance because a similar blind-sac is found in B. misakiensis. In B. capensis it is absent. As the crura of the skeleton spread out nearly in the transverse plane in B. stephensoni, the stomochord becomes very broad just before it communicates with the buccal cavity. Anteriorly the skeleton starts in the cross-sections in the form of a U in both species, the anterior rim being of rather chondroid appearance in B. capensis (fig. 8). When the stomochord becomes smaller, owing to the disappearance of its ventral blind-sac, the skeleton becomes more of a solid structure (fig. 9) and the keel appears, which becomes more and more prominent until it soon surpasses the body in size. In B. stephensoni the anterior part of the keel is quite chondroid (fig. 7). In B. capensis the keel gradually diminishes in size towards its posterior end and disappears altogether a short distance in front of the splitting up of the body into the two erura. In B. stephensoni, on the other hand, the size of the keel is almost constant up to the region just in front of the splitting up of the body (fig. 12), and here the keel stops suddenly. In some species of Balanoglossus (B. biminiensis, B. carnosus, B. numeensis) a reduction of the skeleton can be observed, caused by other tissues penetrating into the skeleton and more or less breaking it up. In B. biminiensis there is even a connection between the right and left collar coelomic cavities piercing through the body of the skeleton. Although the skeleton of B. capensis cannot be considered as being reduced, a similar connection between the collar coelomic cavities was found here, locally separating the keel from the body of the skeleton (fig. 5). Von Bonde also seems to have 78 Annals of the South African Museum. observed this, as he mentions that the body, in the form of a rod, is completely separated from the keel. In B. capensis the crura of the skeleton reach to about one-third of the length of the collar and they surround the buccal cavity for about one-third of its circumference. In B. stephensoni the crura are even shorter and are lying almost completely in the transverse plane. The pericardium does not show any peculiarities in either species. In B. stephensoni its anterior end is at the same level as the top of the stomochord, and in B. capensis it is slightly behind this level, as was also observed by von Bonde. In B. stephensoni the central blood space projects as a free blind-sac into the anterior part of the pericardium, as was described by Hill in B. australiensis. I do not consider this, however, to be of specific importance ; it may be found in one specimen, and in another the blood space occupies its usual place in between the pericardium and the stomochord for the whole length of the pericardium. I was not able to find such a free projection of the blood space in B. capensis, but von Bonde describes it in the specimens of the species studied by him. The glomerulus is markedly different in the two species. In B. capensis the glomerulus appears in the cross-sections at the same place as the stomochord, so that the glomerulus does not project in front of the stomochord. But the right and the left parts of the glomerulus are connected here along the dorsal side of the stomo- chord in front of the pericardium, and when later the pericardium appears in the sections this connection is found at the dorsal side of the pericardium (fig. 10). Thus the anterior part of the peri- cardium is completely covered by the glomerulus. More caudally the glomerulus is broken up into two lateral parts owing to the disappearance of the dorsal connection. Here the glomerulus covers the ventral half of the lateral pericardial wall and extends only slightly over the stomochord. Its caudal end is found near the anterior end of the ventral swelling of the stomochord. In B. stephensonr, on the other hand, the glomerulus projects in front of both the stomochord and the pericardium, and it is already well developed when the top of the stomochord appears in the cross- sections. In this species the lateral parts of the glomerulus are connected along the ventral side of the stomochord and not along the dorsal side of the pericardium (fig. 11). When this ventral connection has disappeared from the sections the glomerulus almost On a New South African Species of Balanoglossus. 79 equally covers the pericardium and the stomochord. It disappears just in front of the ventral blind-sac of the stomochord. Although there is no dorsal connection of the two parts of the glomerulus there is a very small dorsal glomerulus at the side of the dorsal proboscis artery, quite isolated from the main glomerulus (fig. 11). iv ON Fie. 10. — Balanoglossus capensis. Fig. 11.—Balanoglossus stephensoni. Cross-section of the anterior part of the Cross-section of the anterior part of the proboscis organs. x66. gl=glomeru- proboscis organs. x66. bs=central lus; p=pericardium ; st=stomochord; blood space; dgl=dorsal glomerulus; vpa=ventral proboscis artery. gl=glomerulus ; p=pericardium ; st= stomochord. Collar.—In both species the epidermis of the collar shows the usual five zones, the second and fourth of which are composed of thin cells and stain more deeply, because the numerous slime glands occurring in the other zones are lacking in these two (fig. 12). The collar musculature also exhibits no features that might distinguish these species from others. Concerning the coelom von Bonde writes that “the dorsal and ventral septa dividing the collar coelom into two halves are well developed.’ I find, however, that the ventral mesentery in both B. capensis and B. stephensoni is greatly reduced. It is only present in the most posterior part of the collar near the collar-trunk-septum. For nearly the whole length of the collar the right and left coelomic cavities communicate freely with each other along the ventral side. The dorsal septum is better developed. In B. capensis it begins 80 Annals of the South African Museum. with the first and only dorsal nerve root and in B. stephensoni with the second root, from where it extends to the posterior end of the Fic. 12.—Balanoglossus stephensoni. Cross-section of the dorsal part of the collar near its anterior end. x66. cc=collar coelom; cm= circular musculature; dv=dorsal blood-vessel; elm=external longi- tudinal musculature; epv=efferent proboscis vessel; epl=first epi- dermal zone; ep2=second epidermal zone; ilm=internal longitudinal musculature ; nc=nerve cord; phc=perihaemal cavity ; sk=skeleton ; snv =supra-neural blood-vessel ; st=stomochord. collar. In B. capensis, however, there are some holes in this septum ; in B. stephensoni it is entire. The collar pores of both species exhibit the usual features. They have a longitudinal fold along their dorsal side, and the epithelium of this fold is hardly thinner than that of the ventral side of the pores. On a New South African Species of Balanoglossus. 81 The collar pores unite with the first branchial pouch near the external opening. In B. capensis the anterior end of the perihaemal cavities is found at the level of the proboscis pore (fig. 6) and in B. stephensoni just behind this pore. In both species there are some openings in the septum separating the perihaemal cavities ventrally to the dorsal blood-vessel. An anterior epidermal depression, connected with the anterior end of the collar nerve cord, is just indicated in B. capensis; in B. stephensoni it is entirely absent. In the latter species the posterior epidermal depression is also missing, whereas it is deep in B. capensis. The collar nerve cord is very flattened in both species. There is neither an anterior nor a posterior neuropore, as the central canal is quite reduced to a number of small cavities. In B. stephensoni the lateral cavities are small and isolated ; in B. capensis they have more the form of narrow elongated canals, of which a few occur behind each other throughout the whole length of the cord. This was also observed by von Bonde. A few isolated median cavities are found in both species. B. stephensona has three dorsal nerve roots, the first of which runs free through the coelom as the dorsal mesentery only begins with the second root. All the roots are found in the anterior half of the collar. In B. capensis von Bonde found one root in one specimen and two roots in another. In the specimens I have studied there is a single root, but its peripheral part divides, so that it has two connections with the epidermis. Trunk.—In all three specimens of B. stephensoni the epidermis of ~ the trunk, and especially that of the branchial region, is very much swollen, except along the dorsal and ventral nerve cords and around the branchial and genital pores (fig. 13). As the intestine is quite empty, the animals apparently were kept for some time in a disk with sea-water before being fixed, and I have noticed in other species of Balanoglossus that the animals in these circumstances, unprotected as they are by stones or sand, secrete an enormous amount of slime. B. stephensoni is characterised by its short and broad branchial tongues and septa; in B. capensis, as in other species of Balano- glossus, they are narrower. In B. capensis the dorsal or branchial part of the pharynx is larger than the ventral or digestive part, a feature also observed by von Bonde. This author, however, refers to this ventral part as the oesophagus, a term which is obviously inaccurate, for, as in Amphioxus, Tunicates, or Vertebrates, that part of the alimentary tract, which is pierced by the gill slits, is wil SS gasses i hhh; iM t fey t t Fic. 13.—Balanoglossus stephensoni. Cross-section of the branchial region, composed from a number of sections. x45. bp=branchial pore; brs=branchial sac; bs =branchial septum; bt=branchial tongue; bv=parabranchial blood-vessel; dn=dorsal nerve cord; dv=dorsal blood-vessel; ebr=epibranchial ridge; g=gonad; gp=genital pore; lr=parabranchial or limiting ridge; phl=branchial part of pharynx; ph2=digestive part of pharynx; ssk= septal skeletal bar; sv=septal blood-vessel ; tu=tongue blood-vessel; wn=ventral nerve cord; vv=ventral blood-vessel. On a New South African Species of Balanoglossus. 83 termed the pharynx, and this is followed by the oesophagus, which in Enteropneusta extends up to the liver region. With the exception of Schizocardium the gill slits of Enteropneusta are confined to the dorsal part of the alimentary canal, for which reason the ventral half is best designated the digestive part of the pharynx. Taking into account the many folds of the ventral wall, the two parts of the pharynx are nearly equal in size in B. stephensont. As was to be expected in such a small animal, the number of gill slits in B. stephensoni is very small; I counted only 30 of them. Next comes B. australiensis, where, according to Hill, the number of gill slits does not exceed 40; in B. capensis there are 52, which also is a small number for a species of Balanoglossus. In both B. capensis and B. stephensoni the first gill slits open to the exterior by separate pores ; in B. misakiensis the first four gills have a common opening. But B. stephensoni and B. capensis show a remarkable feature in connection with the first branchial pore. All these pores are found in a row at the medial side of the genital wings (fig. 13) as in other species of Balanoglossus, but the common opening of the collar canal and the first gill is found on the outer side of the genital wings (fig. 14). The wings, which are continuous with the dorsal wall of the collar, are only incipient in this region. The difference between the first and the second branchial pores is quite evident on comparing figs. 14 and 15. Even at the level of the second branchial pore the genital wings are better separated from the dorsal side of the body than they are at the first pore. In B. australensis the first branchial pore also occupies a similar position on the outer side © of the genital wing, though it is not mentioned in Hill’s description. On the other hand, in B. carnosus and B. biminiensis the first branchial pore is on the medial side of the genital wing. I could not verify its position in the other species of Balanoglossus owing to lack of material. The number of synapticula is 10 in B. stephensoni and 12 to 13 in B. capensis. Ventral blind pouches to the branchial sacs are absent in both species. In both species, owing to the fact that the epithelium on the inner side of the tongues attains about twice the thickness of that on the septa, the branchial tongues project slightly more into the pharyngeal cavity than the septa. Even so the . epithelium of the tongues is rather low compared with that of other Enteropneusta. In both species the cells of the epibranchial ridge, and therefore also those of the inner side of the tongues and septa, are all of the same character. 84 Annals of the South African Museum. The parabranchial ridges between the branchial and digestive parts of the pharynx are well pronounced in both species, and attain a thickness about twice that of the rest of the pharynx. In B. stephensoni it could clearly be seen that a side branch of the parabranchial blood-vessel ascends along the peripheral edge of the Soa “ a a Uy Hah NAS NX \ %) X " : : Ms iB . rs ae a ce : ebr Fic. 14.—Balanoglossus capensis. Cross-section through the first branchial pore. x66. 6pl1=first branchial pore ; bs =second branchial septum; 6sl=first branchial sac; 6s2=second branchial sac; 6i2= second branchial tongue; dn=dorsal nerve cord; dv=dorsal blood- vessel ; ebr=epibranchial ridge ; rc=trunk coelom; sy=synapticulum. branchial septum up to the dorsal end of the septum (figs. 13, 15). This vessel does not give off lateral branches, nor does it connect with the main dorsal vessel, but it breaks up into smaller vessels at the dorsal end of the septum. These smaller vessels could not be traced further; undoubtedly they enter the tongue and form a capillary network in the tongue, after which they unite again with another larger vessel that could be seen ascending just underneath the epithelium of the inner side of the tongue. This vessel joins the dorsal blood-vessel (fig. 13). The post-branchial canal, forming a short continuation of the branchial part of the pharynx, is very similar in the two species. ——— ————— ae ee eee. On a New South African Species of Balanoglossus. 85 The only difference is that in B. stephensoni the high epithelium of the post-branchial canal is divided into dorsal and lateral parts by two dorso-lateral grooves (fig. 16), whereas in B. capensis this epithelium is thrown into numerous irregular folds (fig. 17). R oy bee aS ACA = F: ae Zz EE INT Re eateoe: ¢ VA V€ ¢ Sw Sy Beg > vk 4 CREE a Ye + oe OEY rer SL Stet ce care cid, Sel ‘ aa5) / walt eS Sooeg say [ 5 Ue 2 . aN .: i y S77 4 Sa 5B, % -) LTE” OSs, v KS CSesct = D G = “TS Gi & oy nr xt s J ass secs nee cae 2 Fie. 15.—Balanoglossus capensis. Cross-section through the second branchial pore. x 66. bp2 =second branchial pore ; bs =third branchial septum ; bs2=second branchial sac ; 6s3 =third branchial sac ; bf =third branchial tongue; dn=dorsal nerve cord; dv=dorsal blood-vessel ; ebr =epibranchial ridge ; sv=septal blood-vessel; sy=synapticulum. both species an anteriorly directed blind-sac of the canal, which in other species is found lying dorsally to the branchial part of the pharynx, is absent. Maser found that it may be absent in young individuals of B. clavigerus and B. carnosus, whereas it is present ‘In older ones. The specimens of B. stephensoni and B. capensis, which I have studied, however, are not young individuals, as their gonads are fully developed. In both species the coelomic blind- pouches that usually project like a pair of valves into the oesophagus 86 Annals of the South African Museum. are also absent. Otherwise these valves seem to be a constant character of the Ptychoderidae, though they are missing in B. proterogonius also. The post-branchial canal ends abruptly in both species. In B. stephensoni the lateral septum could be followed for some considerable distance into the branchial region; it extends over 15 gill pores from the posterior end of the branchial region, 7.e. about over half of that region. On the other hand, in the larger B. capensis it could only be followed for a distance of 10 gill slits, 7.e. for not even one-fifth of the branchial region. As usual, the medial attachment of the branchial septum shifts along the last gill slit from the epidermis to the intestinal wall (figs. 16 and 17). In B. capensis the first genital pore is found between the fourth and the fifth branchial pores, but the corresponding gonad is very small and it is either sterile or very young. All the anterior gonads are small, although they may contain some yolk cells. They occupy only a small part of the genital wing and hardly extend into the body proper. Posteriorly the gonads gradually increase in size, and although the tops of the gonads are also sterile they contain ripe eggs. This indicates that the row of gonads is about to extend in an anterior direction by the addition of secondary gonads. The gonads are profusely branched, and except for the top they fill up the whole of the genital wing; ventrally they extend into the body cavity and reach as far as the ventral end of the gills (fig. 17). Although the gonads are much more branched than in B. clavigerus they have only the primary pore in the sublateral line; secondary pores which may occur either at the distal or medial side of this line are absent. Concerning the position of the most anterior of the gonads in B. stephensoni, these are even further forwards than in B. protero- gonius, which was so called by Belichov on account of the occurrence of gonads between the second and third gills. In B. stephensoni gonads are found immediately behind the collar-trunk-septum ; they can be seen in the same sections as the collar pores. These anterior gonads are small, but they show some branches and each opens by a separate pore in the niche between the genital wing and the body. Their pores are thus found in front of the second branchial pore. These pores may be considered as secondary ones unless it can be assumed that the sublateral line curves round in the niche. Further back in the row B. stephenson, like B. capensis, has only primary pores in the sublateral line. Behind these anterior small gonads an On a New South African Species of Balanoglossus. 87 increase in size soon is observed, and the gonads become profusely branched in both dorsal and ventral directions. In B. stephensona the peripheral part of the genital wing is quite empty ; on the ventral side the gonads penetrate into the body cavity to about the middle of the gills (fig. 13). As has been previously mentioned, a few of the liver saccules in am ~ > 2 : ~— pee Lo AW LS = zB Z —_— “ RSS . Sa Ark Q) vit. Wy y ™ ; meen S A) Uy} ae Hi ee « Ua ae i WSS \vv ~ vn Fic. 16.—Balanoglossus stephensoni Cross-section at the level of the post-branchial canal. x45. cv=commissural blood-vessel; dm=dorsal mesentery ; dn=dorsal nerve cord ; dv=dorsal blood-vessel; g=gonad ; ls=lateral septum; oes=oesophagus; pb=post-branchial canal; vn= ventral nerve cord ; vy=ventral blood-vessel. about the middle of the row are sharply defined by their dark colouration from the pure-white saccules in front of and behind them. In most Ptychoderidae all the anterior saccules differ from the posterior ones, sometimes, as in B. biminiensis, not only in colour but also in their form. In B. carnosus it was found that the two kinds of liver saccules were separated from each other by a well- pronounced swelling of the intestinal wall and the anterior saccules contain many glandular cells in their walls. In B. stephensoni the VOLS S&Xit, PART 2. i 88 Annals of the South African Museum. “ i LM Gis < He : tag ar ANY EA Za Gp ws : CASTER ASa | | : : EELS Ta MOT er Se ST UR hes IRS pe A srre, v D sail Ney Z Fic. 18.—Balanoglossus stephensoni. Longitudinal section of the hepatic region. x27. epd=epidermis; iw=ventral wall of intestine: lm =longitudinal musculature ; Isl, Js2, and Js3 =first, second, and third region of liver saccules; vn=ventral nerve cord; w=ventral blood- vessel. On a New South African Species of Balanoglossus. ASS » 2 eer Fic. 17.—Balanoglossus capensis. Cross-section at the level of the post-branchial canal. «45. cv=commissural blood-vessel; dn=dorsal nerve cord; dv=dorsal blood-vessel; gp=genital pore; /s=lateral septum; lv=lateral blood-vessel; oes =oesophagus ; pb=post-branchial canal; vm=ventral nerve cord ; w=ventral blood-vessel. Fic. 19.—Balanoglossus stephensoni. Cross-section of the pygochord. x 290. itw=intestinal wall; pg=pygochord ; vw=ventral blood-vessel. 89 90 Annals of the South African Museum. dark saccules could easily be identified in sections by numerous small dark granules in the liver cells. The nuclei of the liver cells occupy their usual position almost in the middle of the cell and the dark granules are found only at the side of the nucleus facing the intestinal lumen (fig. 18). Although there is a clear limit between the saccules with granular cells and those without, a swelling of the intestinal wall such as occurs in B. carnosus could not be found. All the saccules have the same form. At the anterior as well as at the posterior end of the row a few small saccules are found that do not cause an epidermal bulging, so that they are invisible from the outside. The ventral wall of the intestine in the liver region is thrown into numerous small transverse folds. As the epidermis in this region also shows similar folds, these may have been caused, at least partially, by the contraction of the ventral longitudinal muscles during fixation. The intestine in the abdominal region of B. stephensoni has the usual pair of ciliary bands, but I was unable to find a continuous line of epidermis without glandular cells, corresponding to the ciliary bands as described by Hill in B. austrahensis, although indications of it were seen here and there. Anteriorly the ciliary bands could be followed through the whole liver region, but at the anterior end of this region they become very indistinct and they do not extend in front of it; in B. australiensis, however, they extend for a con- siderable distance into the genital region. Posteriorly the ciliary bands stop at the caudal region. The anus is surrounded by a feebly developed sphincter. The pygochord of B. stephensoni extends over the whole length of the caudal region, and it is not separated from the intestinal wall. Anteriorly it starts as a small evagination of the wall of the intestine in the ventral midline. Growing larger it reaches the ventral blood- vessel, and more posteriorly, where this vessel ends, it comes into contact with the basal membrane of the epidermis (fig. 19). The pygochord has the same breadth throughout. Its cells, especially those along the dorsal side near the intestine, contain large vacuoles. Diagnosis of Balanoglossus capensis (Gilchrist). Total length up to 13 cm.; in a specimen of 13 cm. the length of the proboscis is 3-5 mm., that of the collar 3 mm., of the branchial region 7-5 mm., the total length of the genital wings is 37 mm. and of the liver region 35 mm.; the abdominal region is 68 mm. long. Genital wings fused with posterior rim of collar; they reach On a New South African Species of Balanoglossus. 91 their maximum breadth at the end of the branchial region, from where they gradually decrease in size ; they reach well into the liver region, where they fade away gradually. The liver saccules form a regular row of about 75 visible saccules on each side. The middle liver saccules are dark, the anterior and posterior ones light in colour. Right and left halves of the glomerulus connected along the dorsal side of the pericardium; glomerulus only slightly covering the stomochord ; ventral proboscis septum greatly reduced but extending to the end of the ventral coelomic blind-sacs; stomochord without ventro-median blind-sac, but with ventro-lateral blind-sacs which communicate with dorso-lateral blind-sacs; only the latter are in connection with the central lumen. Keel of skeleton well developed, gradually diminishing in size towards its posterior end ; crura short, surrounding the buccal cavity for about one-third of its circumference. In the majority the right dorsal coelomic cavity of the proboscis opens into a large end-sac, which has a blind forward extension on the left side as well. Proboscis pore small and round. Nerve cord without central lumen, with one or two dorsal nerve roots. Ventral collar mesentery only present near hind end of collar. Anterior branchial sacs not connected with each other, each opens to the exterior by a separate pore ; first branchial pore at the outer side of the genital wings, all the others at the medial side. Branchial sacs without ventral blind-sacs. 12 to 13 synapticula, about 50 gills. Post-branchial canal without anteriorly directed blind-sac ; no valves projecting into the oesophagus. The first genital pore between the 4th and the 5th branchial pores; except near the anterior end of the row the gonads are large and much branched ; no secondary genital pores. Locality.—False Bay, South Africa. Diagnosis of Balanoglossus stephensoni un. sp. Total length up to 4 cm.; in a specimen with a length of 38 mm. the proboscis is 1-5 mm. long, the collar nearly 2 mm., the branchial region 3 mm.; the total length of the genital wings is 12-5 mm. ; the abdominal region is 20 mm. long. Genital wings fused with posterior rim of the collar; they reach their maximum breadth near the hind end of the branchial region, this breadth remaining practically constant up to the beginning of the liver region. In the anterior part of the liver region the genital wings decrease abruptly in size. The short liver region begins and ends abruptly. The 92 Annals of the South African Museum. liver saccules form a regular row on each side and there are about 20 saccules visible in each row. The middle liver saccules are dark, those in front and behind light in colour. Anteriorly the right and left halves of the glomerulus are connected with each other along the ventral side of the stomochord. Glomerulus covering the pericardium and the stomochord about equally. There is a very small dorsal glomerulus present. Ventral proboscis septum extends from shortly behind the top of the stomochord to the end of the ventral coelomic blind-sacs. Stomochord with large ventro- median blind-sac, which communicates with the central lumen; no dorso-lateral blind-sacs. Keel of the skeleton well developed, its size is almost constant throughout its whole length ; it stops suddenly just in front of the bifurcation of the body into the crura. Crura of skeleton short, situated nearly in the transverse plane. The left dorsal coelomic cavity of the proboscis opens into a large end-sac, which has a blind forward extension on the right side as well. Proboscis pore large and elongated. Nerve cord without central lumen, with three dorsal nerve roots. Ventral collar mesentery only present near the hind end of the collar. Hach anterior branchial sac opens to the exterior by a separate pore ; first branchial pore at the outer side of the genital wings, all the others at the medial side. Branchial sacs without ventral blind-sacs. 10 synapticula, about 30 gills. Post-branchial canal without anteriorly directed blind-sac ; no valves projecting into the oesophagus. The first genital pore between the first and second branchial pores. Gonads large and much branched ; no secondary genital pores. Locality.—Port Elizabeth, South Africa. Type in South African Museum, Cape Town. ——————<— = eee re On a New South African Species of Balanoglossus. 93 LITERATURE. BeticHoy, D. V., “Contributions to the Systematics of Enteropneusta,’”’ Proc. 3rd Congress of the Russian Zoologists, Anatomists, and Histologists in Leningrad, December 14-20, 1927, 1928. (Russian.) Bonne, C. von, “‘ The Morphology of Balanoglossus capensis (Gilchrist), a Species of Enteropneusta from False Bay,” Trans. Roy. Soc. S. Afr., vol. xxii, 1934. Gicurist, J. D. F., “On Two New Species of Ptychodera (P. proliferans and P. natalensis),’’ Ann. 8. Afr. Mus., vol. vi, 1908. ——., *“‘ New Forms of Hemichordata from South Africa,’ Trans. S. African Phil. Soc., vol. xvii, 1908. ——, “A Form of Dimorphism and Asexual Reproduction in Ptychodera capensis (Hemichordata),”’ Journ. Linn. Soc. Zool., vol. xxxv, 1923. Hit, J. P., ““ On a New Species of Enteropneusta (Ptychodera australiensis) from the Coast of New South Wales,’ Proc. Linn. Soc. N.S.W., 2nd ser., vol. x, 1894. Horst, C. J. VAN DER, “ Observations on Some Enteropneusta,” Vidensk. Medd. fra Dansk naturh. Foren., vol. lxxxvii, 1930. ——, “On some Enteropneusta,”’ Annals Transvaal Museum, vol. xiv, 1932. ——,, ‘‘ Die Enteropneusten aus den Meeren der U.S.S.R.,” Zool. Jahrb. Abt. Anat., Bd. 58, 1934. Maser, O., ‘“ Uber Balanoglossus carnosus (Willey) und eine ihm nahestehende, neue Art von Neucaledonien,”’ Zool. Jahrb. Abt. Anat., Bd. 33, 1913. “\ nee Pell er fe Sa wid avon iin ee mone yt = halt A ty a SE. AY ‘ PIE a tar Ree ts c= =k hae PRE Fr vaue a ah Lie st ¢ i jp Ae a CON Date ‘i @ ’ & 4 > « ) , > ' : iv m j r ‘ ai " jpn 1p ita * Lian Fs 1a 4) ioe j Bat at wa as oth fps, ae A | I 7 tS & na Tene ee en " Wana “4 cle Sis ae ee + of may ; Labs Dy pai Nt yi a JP hs a, “a P) a sy Mu f ae - 4 as Co \ 4) os ie ee ) PARTS OF THE ANNALS PREVIOUSLY ISSL ED— Vol. I.—Part 1, out of print; Part 2, out ‘of ‘print : ' print Vol. I1.—Part 1, out of print; Part 2, 5/6; est our oak ‘Part ¢ Part 5, 1/6; Part 6, 3/-; Part 7, 1/6; Part 8, 3 4 Part 10, 7/-; Part ul, 3/-; Index, "Title, et tc., 1/6. Vol. I1IJ.—Part 1, out of print; Part 2, 1/6; Part 3, 5/6; Part 4, 5/6; Part 6, 7/-; Part 7, 1/6; fet 8 a goes fies Title, etc., 1/6. % Geological Survey).— a Sa oes Part 1, 11/-; Part 2, ee ‘Part 3, or Part 4, 4/65, 5 Patt 5 Part 6, 4/6; Part 7, 14/-; Part 8, 8/-. : Vol. V.—Part 1, 4/6; Part 2, 8/6; Part 3, "2/65, Pak 4, 1/6; Pasta Part 6, 5/-; Part a; ah: Part 8, 4/6; Part 9, SISt: ‘Inde: etc., 1/6. : Sat ec Vol. VI.—Part 1, 13/6; Part 2, 4/6; Part 3, 3/6; Part 4, 80/5 Index, Pi etc., 1/6, aes Vol. VII (containing Palaeontological papers published in ‘conjunction Geological Survey).— Part 1, 3/-; Part 2, out of print; Part Ey By Part 4, 8 . 5/6; Part 6, 1/6; Index, Title, etc., 1/6. a - Vol. VIII.—Complete, out of print. Index, Title, etc., 1/6. - Vol. [X.—Part 1, 4/6; Part 2, 5/6; Part 3, 10/-; ’ Part 4, , 6/6; ‘Rast 6, 3/6; Part 6, 11/-; Part 7, 9/-; Index, Title, etc., 1/6. Vol. X.—Part 1, 3/-;. Part 2, 2/6; Part 3, 2/-;- Part 4, 3/-3 ‘Part 5, | Part 6, 3/-; Part 7, 10/-; Part 8, 2/6; Part 9, 5/-; Part 10, 2/6; _ “Part 11, 20/-; Part 12,7 ee Complete. Title, ete., contained oa >. Part-11. 2 ae Vol. XI.—Part 1, 3/6; Part 2, 2/-; Part 3) 13/6; ‘Part 4, 1/6; Part 5: Wes - Part 6, 11/-; ‘Index, Title, ete., and Plate Ii, 3/-. : Vol. XII (containing Palaeontological papers published in conjunetion with t h Geological Survey).— eee Part 1, 15/6; Part 2, 3/6; Part 3, 4/65 Part 4, 3/-; Part 5, Part 6, 6/—; Part 7, 20/-; Part 8, 20/-; Index, Title, ete., ue Vol. XIII.—Part 1, 6/-; Part 2, 2/6; Part zy 3/3 Part 4, 8/6; Part 5, », 1/6 ~ Part 6, 5/—; Part 7, 30/-; Part 8, Y=; Index, Title, ete. ie ee Vol. XIV.—Part 1, 8/6; Part 2, 8/—; Part 3, 6/-; Part 4, 21/-; 5, d/- Part 6, 9/-; Index, Title, ete., 1/6. Vol. XV. —Part 1, 17/-; Part 2, via Part 3, 14/-; Part 4, 12/65 Part Part 6, 3/6; Index, Title, ete., 1/6. — ee Vol. XVI.—Part 1, 30/6; Part 2, 4/-; Part 3 (with Title, Tree -ete. ), 28h Vol. XVIT.—Part ‘A, 12/-; Part 2, 9/6; Part 3, 3h3 ‘Part 4, 17/-; Part 5, Part 6, 2/6; Index, Title, 6164310; is Vol. XVIII. = 1, 20/-; Part 2, 7/6; Part 3, 30/-; ‘Part 4, 12/65 Index, , 1/6.” : oe 5 Vol. XIX. eet 1, 22/-; ‘Part 2, 17/65 Part 3, Ls ‘Part 4, 5/6; Index, Title, etc., 1/6." Vol. XX.—Part 1, 8/6; Part 2, 12/6; Part 3, 4/-; Part 4, ates Part Be Ans - Part 6 (with Title, etc.), 4/6. eens es Vol. XXI.—Part 1, 25/—; Part 2 (with Title, oie. )» 30/-. = Vol. XXIT (containing Palaeontological Jeet pels in conjunction with the ne Geological Survey).— nee a Part 1, 20/-; Part 2, 10/-; Part 3 (with Title, etc.), 3/6. geen Vol. XXITI.—Part 1, 12/6; Part 2, 8/-; Part 3 (with Index, Title, etc.), eee : Vol. XXIV.—Part 1, 10/-; Part 2, 2/6; Part 3, 4/6; Part 4, 11/-. Vol. XXV.—Part 1, 12/6; Part 2, 10/-; Part 3 (with Index, Title, etc. ) 10/-. Vol. XXVI. —Complete, 25/-. Vol. XX VII.—Complete, 25/-. Vol. XXVIII (containing Palaeontological papers s published in sanjundtion with the Geological Survey).— Part 1, 10/-; Part 2, 17/65 Part 3, 6h Part 4 (with Index, Tile, etc.), 10/-. : = Vol. XXTX.—Part 1, 20/-; Part 2 (with Index, Title, ete. ), 20/—. nat a Vol. XX X.—Part 1, 15/6; Part 2, 20/-;. Part 3, 8/-; Part 4, 9/-; Part 5, of. we Vol. XX XI (containing Palaeontological papers published in ee wes the re Geological Survey).—. = si Part 1, 20/-; Part 2, 13/6; Part 3,17/6. > == = oa a Vol. XXXIT.—Part 1, 2/6; Part 2, 7/6. a Ae. The Annals of the South African Museum etl be issued ¢ at reslar intervals, as matter for publication is available. - — = Copies may be tad ae Ee Ae ae = es Messrs. WHELDON & WESLEY, Lrp., eee Sree TS 2, 3, and 4 ArTHuR Street, NEw OxrorpD ee Loxpox, Ww C. 25 OTe SS The LIBRARIAN, Soura Arrican Mosxtm, CaPE Town. Re ‘ ce v : an = | = ots: ; ; 2? bi: er 2 Pan ees ates tf a ee z 7 7 ; 2) “ Z . Sy ea rt tr? # or 7 f vay 3 - Bete: er Janwars "SOUTH AFRICAN MUSEUM PART III, containing: — : 8. South African Geophilomorpha (Myriopoda).—By Dr. Karu W. Vernoerr, Pasing, Munich. (With Plates [IX and X.) 9. A New Genus of Colobognatha from the Cape Peninsula.—By eer _Dr.- Karun W. Vernorrr, Pasing, Munich. (With sees we == Plate XI) | —— —_—«*10. New Colobognatha from South Africa.—By Dr. Karu W. pee VerHoerf, Pasing, Munich. (With Plate XII.) : 11. New Acridudae from South Africa.—By K. H. L. Key, M.Se., ~ Ph.D.; D.1.C., Canberra, Australia. (With Plates XIII-— BA ei XVII and 2 Text-figures.) 12. Notes on Dragon-flies (Odonata) of the S.W. Cape, with ae _ Deseriptions of the oY ymphs, and of New Species.—By eee KH Barnarp, D.Sc., Assistant Director. (With Set i ee _ 32 Text-figures.) se 13, A New Jenkinshelea (Dipt. Ceratopogonidae) from Southern bee _Rhodesia.—By Bora DE MeILton, D.Sc., F.R.E.S. (South te mae eae eae African Institute for Medical Research, Johannesburg). = Sg eee (With 1 Text-figure.) - Be ee 1k, Report on Some Anthidiine Bees (Apoidea) in the South African z agen ee, Museum.—By G. A. Mavromoustakis, Limassol, Cyprus. eS 15, A New Species of Plume-footed Mite from South Africa.—By : ee ee R. F. Lawrence, Ph.D., Director, Natal Museum, Pieter- 5 z : maritzburg. (With 4 Text- figures. ) : | eee JULY 1987. PRICE 17s. Be: . PRINTED FOR THE si TRUSTEES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN BY NEILL AND CO., LTD., 212 CAUSEWAYSIDE, EDINBURGH, 4 ( 95 ) 8. South African Geophilomorpha (Myriopoda).—By Dr. Kari W. VERHOEFF, Pasing, Munich. , (With Plates [IX and X.) SUBSEQUENT to Attems’s monograph on the Myriopoda of South Africa which appeared in 1929 in the Annals of the South African Museum, the following species of Geophilomorpha have been made known from South Africa :— Aspidopleres intercalatus Porat. Mesoschendyla monopora Attems. - caledonica Attems. Schendylurus australis Silvestri. - polypus Attems. Ballophilus braunsi Silvestri. Purcellinus robustus Attems. Geoperingueyia conjungens Attems. Achilophilus monoporus Attems. Eurytion dolichocephalus Attems. trichopus Attems. aporopus Attems. badiceps Attems. dentatus Attems. ie sabulosus Attems. Polygonarea kraepelini Silvestri. oligopus Attems. at monospathis Attems. Brachygonarea apora Attems. (Philacroterium cribellatum Attems) _ =Aphilodon cribellatum (Att.). (Philacroterium pauperum Attems) =@ of Aphilodon weberv. Aphilodon weber Silvestri. bP) 3) 99 be) 99 To the Director, and to Dr. R. F. Lawrence, formerly Assistant at the South African Museum, Cape Town, I here take the opportunity VOL. XXXal, PART’ 3. 8 ocT 41987 96 Annals of the South African Museum. of tendering my best thanks for sending me the series of Geophilo- morpha from the Cape Province with which this paper deals. It contains 14 species of which 10 are new to Science, a fact which proves that the extent of the Geophilomorphid fauna of South Africa has by no means as yet been determined. The following is the list of the forms under consideration :— Polygonarea porosa 0. sp. Re litoralis n. sp. Eurytion dentatus Attems. re lawrencet n. sp. *e trichopus Attems. Ps gracillimus un. sp. i kalaharinus elongatus n. subsp. i brevis n. sp. Achilophilus pachypus nu. sp. Aphilodon weberr Silvestri. 3 caffrarvum n. sp. - porosum XN. sp. Mesoschendyla cribrifera n. sp. Aspidopleres intercalatus Porat. Of the 22 species given by Attems and the 14 which now lie before me, 4 species are contained in both lists. It must also be taken into account that my investigations have demonstrated that the genus Philacrotervum erected by Attems is untenable, as it is identical with Aphilodon. Attems separated Philacrotervum with a two-jointed tarsus of the terminal legs from Aphilodon, as the latter has only a single-jointed tarsus in the terminal leg. I have been able to show that this dimorphism of the terminal legs is a sexual character, as the male has thickened terminal legs with a single tarsal segment, while the female has slender terminal legs with a normal two-jointed tarsus. Philacrotervum pauperum Att. must be dropped, as it repre- sents the female of Aphilodon weber: Silv., and P. cribellatum thus becomes Avphilodon cribellatum Att. From this it follows that Attems’s list comprises 21 and not 22 species, and with the new forms contributed by myself no less than 31 forms are now known to occur in South Africa. As my 10 new species are not known from any other countries and of Attems’s list only 3 are known from other regions, and those 3 from South West Africa (Aspidopleres intercalatus, Meso- schendyla monopora, and Polygonarea kraepelint), it follows that all South African Geophilomorpha (Myriopoda). Sil the 31 Geophilomorphid species are endemic in South Africa and 28 of them are known only from the Cape Province and Natal. The genus Aphilodon requires special attention from the viewpoint of comparative morphology. In his work on the Geophilomorpha in “Das Tierreich,” Lief. 52, 1929, Attems (on p. 157) makes the Aphilodontinae a subfamily of the Geophilidae, and says of it: “‘ Tro- chanteropraefemur und Femur der Kieferfiisse verwachsen, sodass das Telopodit dreigliedrig erscheint.” By this character the Aphilo- dontinae can be readily distinguished from all other Geophilidae. In contrast to the above statement Attems writes (on p. 12) that the telopodite of the maxillipedes in the Aphilodontinae is “sogar nur dreigliedrig, indem Femur und Tibia verschmelzen.” He therefore says that one of the segments of the maxillipede has disappeared, in the first place by the fusion of the femur and tibia, and then again by the fusion of trochanteropraefemur and femur, so that there is no basis for either one or other of these hypotheses. Actually there is nothing to be seen in his figures of the maxillipedes of the Aphilo- dontinae to support either view (e.g. his fig. 39 on p. 191, in “The Myriopoda of South Africa ’’). In the circumstances I was determined, with Aphilodontine material before me, not to let slip the opportunity of settling this question. Previously there was, besides the views put forward by Attems, a third possibility, namely, that of the two short intersegments of the maxillipedes, one had become altogether atrophied or suppressed, ' as can be observed in the Chilopoda in general where successive reductions of both the intersegments can be observed, and in many Geophilids, e.g. the subgenus Onychopogaster of the genus Geophilus, where these intersegments have actually been strongly reduced. My investigations on the maxillipedes of Aphilodon soon proved that in this genus there is actually a fusion of femur and trochantero- praefemur, as can be seen from figs. 4 and 8. In Aphilodon weberi and caffrarium there is a more or less long suture in front of the more basal of the two tubercles which occur on the inner margin of the maxillipede telopodite which I at once recognised as the vestigial traces of an articulation between trochanteropraefemur and femur; this suture extends further above than it does below towards the outer side. If there was still any doubt remaining this would be dispelled by Aphilodon porosum (fig. 4), in which species the boundary between trochanteropraefemur and femur has remained almost complete; above and below and just at the outer end (y) a small portion is missing. On the inner side behind the tubercle there is 98 Annals of the South African Museum. even a small intersegmental membrane representing the remains of a true articulation. From this it will be seen that the femur is not always absent in Aphilodon, but that it is sometimes reduced to a more or less imperfectly defined terminal portion of the trochanteropraefemur. Thus all the segments of a normal walking leg are recognisable in the maxillipedes of Aphilodon, the trochanter being indicated by the well-known “Trochanterkerbe”’ (tk), which also exhibits an abbreviated marginal suture, while the coalescence of tarsus and ungulum to form a tarsungulum (tau) is denoted by at any rate a difference of chitinisation and pigmentation. This leads me finally to the conclusion that too much systematic weight has been attributed to the alleged 3-jointed telopodite of the maxillipede, at least when it has been made the basis of family distinctions, as Silvestri has done. The Aphilodontinae without doubt constitute a natural group, but seeing that they have lost the labrum and both pairs of maxillae are reduced, their natural relationships are not with the true Geophilidae but as a subfamily of the Scolio- planidae (Dignathodontinae). I must here refer to an important characteristic of the Aphilo- dontinae, as it has hitherto been quite unknown, viz. the structure of the poison glands of the maxillipede. As is well known, the body of the poison glands in most Geophilidae lie on the inner side of the maxillipedes, so that the vesicles of the poison gland are found on the inside of one or two of the segments of the telopodite; in Achilo- philus, for instance (fig. 10), where four segments of the telopodite are found close together, they are situated at the outer extremity of femur and tibia. In Achilophilus, however, there is nothing to be seen of a poison vesicle in the neighbourhood of the maxillipedes. In an unmacerated specimen of A. caffrarium I could follow the fine poison canal throughout the whole maxillipede telopodite, and even further backwards in the region of the coxosternum of the maxillipede between muscles. Further along the body, however, the faint outline of the canal could no longer be seen. On the other hand, I have observed the presence of a poison vesicle in the neighbour- hood of the 18th leg segment in Aphilodon porosum, which was about half the length of the 18th sternite, truncate anteriorly and posteriorly, and about as wide as the trochanter of a leg. I noticed this vesicle only on the one side, on the other it became lost. That it really is a vesicle of the poison glands is proved by its well-known structure, namely, the penetration of the walls of the vesicle by large numbers South African Geophilomorpha (Myriopoda). 99 of fine close-set pores, by means of which the canals of the individual cells of the gland open. That the poison glands of the maxillipedes are situated in the body instead of the maxillipedes has hitherto only been observed once in the Geophilidae, namely, in the genus Chaetechelyne. On this question I wrote in 1902 in my work on Chilopoda (Bronn’s Klass. u. Ordn. des Tierreichs, Lief. 63-65, p. 35) the following :— “A noteworthy exception (from the usual position of the poison glands) is found in the genus Chaetechelyne where the glands have left the maxillipede segment and retreated to the 12th to 18th segment of the body (Duboscq).” This exception in the case of Chaetechelyne is of all the more interest as this genus belongs to the Scolioplanidae, and thus both the excep- tions with regard to the maxillipedes belong to the same family, a position which is therefore in accord with my previously stated views on the systematic status of the Aphilodontinae. Finally, I should like to observe that the “absence of ventral pores,” which as a common character has hitherto united members of the A philodontinae, and therefore also the absence of glands in the region of the sternites, has been invalidated by my discovery of a species, Aphilodon porosum, in which a number of sternites in the posterior part of the body have two large porose areas situated one behind the other (fig. 6). Gen. APHILODON Silv. (=Aphilodon 3 +Philacrotervum 2 Attems.) I have been able to show that Aphilodon and Philacroterium are generically identical, and, as mentioned above, merely represent different sexes of the same form in the case of weberi Silv. and also caffrarium, while I have only seen the 2 of porosum. Attems has recognised 5 species of Aphilodon in the old sense, of which, however, only one, weber Silv., inhabits South Africa, while the four others from South America certainly need re-examination. The South African species can be distinguished as follows :— (a) The 8 sternites anterior to the terminal segment with a large porose area which is usually divided, one area lying behind the other (fig. 6). Claw of the maxillipede with an inner tooth basally (fig. 4). Coxopleurae of terminal legs with numerous scattered pores (fig. 5). 75 pairs of legs porosum N. sp. (6) All sternites entirely without pores. . : : 4 : ‘ 0, de 100 Annals of the South African Museum. (c) Claw of maxillipede with a small basal tooth. Coxopleurae of terminal legs with numerous scattered pores. 55-71 pairs of legs? cribellatum Attems. (I do not consider this species valid, and regard it as merely representing larger individuals of weberi Silv. in which a small tooth is present at the base of maxillipede claw. It is also noteworthy that Attems must have seen a 3 of his cribellatum, as he says, ‘“‘ Terminal legs of ¢ moderately incrassate and densely covered with hairs ventrally.’’ Here either the 3 possesses an abnormal 2-jointed tarsus of the terminal legs or Attems has been mistaken.) (dq) Claw of maxillipede without a tooth (fig. 8) . , A : Re Mi 2 (e) Coxopleurae of terminal legs with 5-8 large pores (fig. 9) of which almost all lie opposite the semi-circular sternite and only one is occasionally situated more posteriorly. The femoral tubercle of the maxillipede is well de- veloped, that of the praefemur, on the other hand, rudimentary (fig. 8). Ultimate segment of terminal legs of 2 only 3-4 as long as the pen- ultimate segment. 49-55 pairs of legs . : ; caffrarium Nn. sp. (f) Domaine of the terminal segment with numerous pores which are, however, on an average smaller and far more scattered (fig. 7), its more trapeziform sternite broadly truncate posteriorly. The praefemur of the maxillipede with an inner tubercle, but little smaller than that of the femur. Ter- minal legs of 2 with the two last segments of equal length. 45-55 pairs of legs : ‘ : 2 ‘ : ‘ s : weberi Silv. Aphilodon porosum n. sp. (Figs. 4-6.) ° from 64 mm. with 75 pairs of legs. Most of the characters are the same in all species of the genus Aphilodon so that I shall give only those of diagnostic value. In the maxillipedes the upper coxal plates are fused posteriorly in the middle, this fusion being reinforced in the middle line by a strong rib enclosing a brown stripe (which is absent in other species with a rib). Clypeus with close reticulation but without bristles (both the other species have them). Instead of a labrum there are two undifferenti- ated transverse plates divided in the middle by weak reticulation. The fulcra are transverse and anteriorly are elongated outwards and forwards as lobes. The almost triangular basal joints of the antennae overlap each other only slightly in the middle (more markedly in the other species). The maxillipedes, which have already been dealt with above, have four distinctly projecting tubercular teeth on the inner side (fig. 4); at the same time they are longer than in the other species, which is evident from the fact that the trochantero- South African Geophilomorpha (Myriopoda). 101 praefemur is as long on its outer side as its basal width (in the other species, fig. 8, they are considerably shorter). The maxillipedes are accordingly as long as the head. On the anterior leg-bearing sternites there is a median strip of reticulation in the posterior and middle thirds, whereas the whole of the anterior third is taken up with reticulation in which the pores appear like islands. Of the porose areas which appear on the 8 sternites anterior to the terminal legs, those on the 5 most posterior sternites are divided into two large rhomboidal areas, the one situated close behind the other (fig. 6); those on the sixth sternite from the end unite to form a single very large area, on the seventh from the end they are again separate, while on the eighth from the end only the posterior area is present, being the same size as that of the follow- ing sternite, while the anterior area is reduced toa few pores. From the ninth last sternite anteriorly, there is no trace of pores. Otherwise there are a few pores on the procoxa of almost all the legs, these being most conspicuous in the posterior part of the body in front of the terminal segment. Terminal legs of the 2 slender, with scattered, fairly short bristles, the last segment 2—? as long as the penultimate, the claws well developed. Sternite of last segment (fig. 5) broadly truncate behind, the pores of the coxopleurae numerous but all confined to the under side, their openings nearly always free, a few being hidden by the sternite. Locality.— This species, the largest of the genus and with the most numerous segments, is known to me only by a 2 from Oudebosch, © River Zonder End, Cape Province. Aphilidon caffrarvum x. sp. (Figs. 8 and 9.) 2, ¢ 28-30 mm. with 49-55 pairs of legs. To distinguish this species from the closely related weberz I here supplement the key given above with the following description:— On the anterior sternites the reticulation forms a longitudinal strip 1-4 of their width, which widens only near the anterior margin; the bristles are short and scattered. (In weberi the anterior sternites have irregularly disposed reticulation, while the bristles are longer and less numerous.) Most of the pores are situated on the coxopleurae of the terminal segment (fig. 9), being larger than the posterior stigmata, while they are so close together that the distance between 102 Annals of the South African Museum. them is often less than the width of the pores themselves (in weber, on the other hand, fig. 7, the largest pores are hardly larger than the posterior stigmata, while they are so dispersed that the distance between them is always much greater than the width of the pores themselves). The claws of the terminal legs are rudimentary and blunt in the ¢ (in webert they are also very small but clearly defined and sharply pointed); the terminal segment in the ¢ is about 12 as long as wide in both species. Localhity.—This species is found at Hogsback, Amatola Mts., and Grahamstown, both in the Eastern Cape Province. Aphilodon weberi Silv. is the commonest species of the genus, and seems to be found throughout the greater part of the Cape Province, viz. River Zonder End, Table Mountain (Cape Peninsula), Ladismith, and Leliefontein (Namaqualand). Gen. ACHILOPHILUS Attems. The genus Achilophilus is known only from the Cape Province by a single species monoporus Att. Achilophilus pachypus MgSO (Figs. 10 and 11.) The form before me corresponds so closely with Attems’s descrip- tion that I can confine myself to merely a statement of the diagnostic characters. monoporus Att. 38 mm. in length, 49-55 pairs of legs. Clypeus anteriorly with area. Head not + longer than wide. ‘“‘Maxillipedes do not reach the frontal margin.” Antennae touching at their bases. Telopodite of first maxillae with small externallobes. Clypeus anteriorly with some bristles. Anterior sternites with ill-defined porose areas which tend more and more to move to the sides. Fulcra T-shaped, widening anteriorly. Trochanter of terminal legs as long as wide. pachypus n. sp. 22 mm. in length, 53 pairs of legs. Clypeus entirely without area. Head more than 14 times as long as wide. Claws of maxillipedes reaching to beyond the head (fig. 10). Antennae not touching. ‘Telopodite of the first maxilla entirely without outer lobes. Clypeus smooth, without bristles an- teriorly. Anterior sternites in and posterior to the middle with a few scattered pores. Fulcra bent to form a blunt angle and transversely situated. Trochanter of the terminal legs much wider than long. As a supplement to the description of pachypus I note the following: The claws of the maxillipedes (fig. 10) exceed the head anteriorly, and their bases are situated only a little posterior to the anterior ( South African Geophilomorpha (Myriopoda). 103 margin of the head. The sutures of the pleurocoxae run somewhat obliquely. Claws of the maxillipedes with a small basal tooth, the structures otherwise without teeth. The small abbreviated poison vesicles lie on the outer extremity of the two intersegments, and the femur is inserted decidedly obliquely. Labrum atrophied. Coxo- sternum of the second maxilla with a median suture. The coxo- pleurae of the terminal segment have a gland on each side (fig. 11), the pore of which lies above the lateral margin of the broad trapezoid sternite, just as in monoporus. The coxal parts project below, posteriorly, and on the inner side as a rounded pad, and above this I made out a gland; this also exhibits an inner vesicle, which does not, however, open to the exterior. Whether this gland is as yet unde- veloped (like that of monoporus) must be decided by further investiga- tion. Some glands without a recognisable opening (fig. 11) are also /present in the segments of the telopodite, a condition which I have also observed in some other Geophilomorpha, e.g. Schendylidae. Anal glands absent. Antennae 3 times as long as the head, with sparse, short hairs, the ultimate segment 23 times as long as the penultimate one. Locality.—I have seen only two specimens from Prince Albert, Cape Province. Gen. Eurytion Attems. The genus Hurytion, of which Attems in his monograph on the Geophilomorpha recognised 13 species, has been hitherto represented in South Africa by 7 species, of which 6 are found in the Cape Province, kalaharinus being confined to South West Africa. Hurytion is by far the richest in species of the Geophilomorphid genera, a fact confirmed by the series before me which contains 6 forms of Eurytion. The characters used by Attems in his key are for the most part beyond criticism. I have, however, to reconsider his so-called “‘ Aussentaster”’ of the first maxillae. This expression must certainly be rejected as these structures have absolutely nothing to do with organs of touch (y, fig. 1), but are, on the other hand, very pale and delicate lobes,* and even then as macerated specimens they are often so extremely transparent that one is in great doubt as to their actual nature. Further, on account of their delicacy, the lateral lobes can be easily folded over and in this way overlooked. On these grounds I have in the following key, which includes all South African forms, completely ignored the lateral lobes (‘“‘Taster’’) of the first maxillae. * The term ‘‘Taster”’ could at most only be applied to the telopodite (t, fig. 1). 104 Annals of the South African Museum. Another misleading character is that of the fringes of the labrum, as these are also of a very pale delicate nature, so that in macerated specimens they may be quite indistinguishable. All, or at any rate the majority, of the pores of the coxopleurae of the terminal segment are scattered and have free openings (figs. 20 and 23). Key to the South African Species of Eurytion. (a) Coxopleurae of the terminal segment entirely without pores, 57-61 pairs of legs : j : : é : : . 1. aporopus Attems. (6) Coxopleurae of the terminal legs with scattered pores . - : . ¢, a. (c) Trochanteropraefemur of maxillipedes with 2 stout teeth on the inner side Cui. (d) Trochanteropraefemur of maxillipedes with 1 or without teeth on the inner side 1, k. (e) ue pores of the coxopleurae arranged in the form of a horseshoe of which the ventral portion abuts on the lateral margin of the sternite, the dorsal portion being dispersed above, while some pores lying between these two groups are anteriorly situated.* Claws of terminal legs normal. Porose areas of the sternites longish oval. 71-81 pairs of legs 2. dentatus Attems. (f) The pores of the coxopleurae dispersed ventrally . é : . A Pea (g) Claws of terminal legs unusually small (fig. 19), hardly half as long as the width of the second tarsus. Median piece of the labrum transversely elongate, not toothed (fig. 18). Clypeus with close hexagonal reticulation but no bristles, a rather ill-defined area. 59 pairs of legs © 3. gracillimus n. sp. (h) Claws of terminal legs normal, thus as long or almost as long as the width of second tarsus (fig. 21). * The 7-8 pores of the coxopleurae situated in two rows next to the sternite. Porose area of the anterior sternites round. Area of the clypeus longish oval with 3-5 bristles, behind these 2-6 bristles. Median portion of the clypeus almost cordiform, not wider than long. 69-77 pairs of legs : ‘ 4, sabulosus Attems. ** The 9-10 pores of the coxopleurae (fig. 20) not all situated close to the sternite, half of them lying more laterally. Porose areas on ster- nites 2-14 transversely oval. Clypeal area rounded and with a bristle. Median piece of the labrum transversely elongate, more than twice as broad as long. 47 pairs of legs . 5. brevis n. sp. * Attems’s account (loc. cit., p. 261), in which the coxopleurae of dentatus “ are provided with 2 groups of larger pores, the one consisting of a strip of 7-8 opening beneath the margin of the sternite, the other of 5-6 dorsally beneath the margin of the tergite,”’ is not sufficiently exact and refers to immature individuals. South African Geophilomorpha (Myriopoda). 105 (7) The pores of the anterior sternites form a transverse dispersed strip posterior to the middle of the sternite, and in addition there is a small group of pores on each side anterior to the middle. Trochanteropraefemur of the maxillipedes entirely toothless. Coxopleurae of the terminal segments with 15-16 dispersed ventral and lateral pores. 39-49 pairs of legs lawrencei 0. sp. (k) The pores of the anterior sternites forming a compact rounded to transversely ovalarea. Clypeus witharea . : : : : ; l,m. (1) Trochanteropraefemur of the maxillipedes with a tooth or tubercle on the inner side. Anal glands absent . : : ; : : 5 N, O. (m) Trochanteropraefemur entirely without a tooth or tubercle . : Ds q.- (n) Porose area of the sternite, beginning from the second, transversely oval. Pores of the coxopleurae dispersed over the greater part of the ventral surface almost as far as the posterior margin of the coxae. 71-75 pairs of legs : : é ‘ ‘ é : 7. dolichocephalus Att. (0) Porose area of sternites round to slightly transverse ovate. * Pores confined to the anterior half of the coxopleurae, not passing beyond a line touching the hind margin of the sternite. Tibia and both the tarsal segments of the terminal legs much slenderer than the preceding segments. Sternite of the terminal segment with its sides convex. Areas of sternites round. 61 pairs of legs 8. kalaharinus Att. 3 ** Pores of the coxopleurae opening partly above, partly below ; some of them, however, situated posteriorly to a line touching the hind margin of the sternite (fig. 23). Only the two tarsal segments of the terminal legs much slenderer than the remaining segments. Sternite of the terminal segment with straight sides, being thus trapeziform. Areas of sternites slightly transverse ovate. 83-87 pairs of legs b : ; 9. kalaharinus elongatus n. subsp. (p) The porose areas situated posterior to the middle beginning from the second sternite, entirely round. Coxosternum of first maxillae without lateral lobes. Anal glands absent. Pores of the coxopleurae numerous, ven- trally and dorsally. 81-89 pairs of legs ; . 10. badiceps Attems. (q) The porose areas situated posterior to the middle beginning from the second sternite and bounded anteriorly by an arcuate furrow while posteriorly they form approximately a right angle, thus resembling a quadrants. Coxosternum of the first maxilla with lateral lobes. Anal glands present. Pores of the coxopleurae numerous, ventral and pleural. 65-69 pair. of legs eee : : : : : 11. trichopus Attems. Eurytion lawrencei n. sp. (Figs. 16 and 17.) 2 24-32 mm., with 39-49 pairs of legs. Antennae with segments 2-5 with 3 whorls of hairs. Labrum with a bluntly triangular 106 Annals of the South African Museum. median piece (fig. 17) about as long as wide, and weakly striated posteriorly; | have not observed a fringe on the labrum. Head-plate with sparse bristles; clypeus anteriorly without an area and without bristles, with dense mosaic cell structure; posteriorly, on each side in front of the labrum a smooth window-like area (fe, fig. 17). Coxo- sternum of the first maxilla without, basal segment of the telopodite with, a short outer lobe; maxillipedes with a small tooth only at its base. Vesicle of poison gland twice as long as wide, situated at the base of the claw. First sternite with 6 pores. The 2-11 anterior sternites on each side with a transverse strip of well-separated pores posterior to the middle, well separated from the sides of the sternites (fig. 16), while anteriorly another small group of pores can be seen on each side; otherwise all these sternal pores are not very noticeable, but in their arrangement they differ from those of all other South African species. On the sternites following, the pores are entirely absent. Two pairs of larger bristles are present on the most anterior and most posterior sternites, while the sternites between them are provided with a few very short bristles. Sternite of terminal segment trapeziform but arcuate posteriorly. The 14-15 pores of the coxopleurae open freely and are scattered ventrally and at the sides, most of them being as large as the posterior stigmata, but 4 on each side distinctly smaller. Terminal legs of 2 with a few bristles, the claws stout, the proximal tarsal segment little shorter than the terminal one. Genital sternite with 2 trans- verse rows of bristles. Anal pores large. Locality.—This species, named in honour of Dr. R. F. Lawrence, is represented only by a 9 from River Zonder End, Cape Province. Eurytion brevis un. sp. (Figs. 20 and 21.) Immature example 19-26 mm. in length, with 39-47 pairs of legs. In general most closely resembling sabulosus Att., from which it differs in its smaller size, in having fewer pairs of legs, in the more scattered coxopleural pores (fig. 20), in the median piece of the labrum being 4 as wide as the lateral pieces (in sabulosus it is hardly + as wide as these), and in the homogeneity of the clypeal reticulation, which contains neither an area nor bristles; furthermore, the porose areas beginning from the second segment are transversely oval. Trochanteropraefemur of the maxillipedes with 2 sharp stout teeth on the inner side, a similar one at the base of the claw, while both South African Geophilomorpha (Myriopoda). 107 intersegments are smooth; anterior margin of coxosternum with 2 blunt tubercles. Segments 1-5 of the antennae with 1-2 whorls of hairs (it is doubtful if the first maxillae have lateral lobes). First sternite without pores, the transversely oval porose areas which are distinct up to the thirteenth sternite consist in the second sternite of 3-4 pores in a longitudinal row, 6-7 in a transverse row; fourth sternite similarly with 4 and 10 pores. These areas are situated behind the middle and occupy the middle third of the segment in a transverse direction. The 9-10 coxopleural glands have free openings by means of scattered pores (fig. 20). Locality.—The specimens before me I take to be half grown, but the species is nevertheless quite distinct from the other known forms. Table Mountain, Cape Town. Eurytion kalaharinus elongatus nu. subsp. (Figs. 22 and 23.) 2 47-56 mm. With 83-87 pairs of legs. Median piece of the labrum weak (fig. 22), hardly as wide as long, and without noticeable serrations on the hind margin. (In dolicho- cephalus Att., though it is not larger than elongatus, the median piece of the labrum is distinctly dentate posteriorly and the lateral pieces have long fringes.) The fringes of the lateral pieces fairly long, but diminishing fairly suddenly mesially. Maxillipedes quite similar to dolichocephalus, the claws therefore with a sharp triangular tooth and the trochanteropraefemur with a small blunt inner tooth. Poison vesicles hardly longer than wide, situated on the outer side of the tibia. Clypeus with a large area bearing 2 pores, a few bristles at the sides only. First sternite without pores. From the second sternite onwards there is short transversely oval to round porose area which on sternite 5 is not quite as wide as its distance from the lateral margins. These sternal pores are divided into two fairly large groups in the posterior half of the body, which again converge on the two sternites before the terminal segment. The sternites are practically smooth, large bristles being absent on nearly all of them. The numerous scattered and freely opening pores of the coxopleurae lie partly above and partly below (fig. 23) their sternite trapezoid. EL. dolichocephalus, kalaharinus, and elongatus are all closely related, and further material will show whether or not their position as given 108 Annals of the South African Museum. above is valid. The 4 99 before me are from Kakamas (Orange River) and Garies (Namaqualand). Eurytion gracillumus n. sp. (Figs. 18 and 19.) Subadult form 28 mm., with 59 pairs of legs. Trochanteropraefemur of maxillipedes with 2 blunt teeth, claw with a sharp basal tooth, intersegments without teeth, anterior margin of coxosternum with 2 blunt projections bearing a small bristle. Antennal segments 1-5 each with 1-2 whorls of hairs. Median piece of the labrum transverse, not toothed; the whole labrum with extremely weak fringes (fig. 18). First maxilla entirely without lateral lobes. Clypeus with comparatively large hexagonal reticulation, without area and without bristles. Anterior sternites in and posterior to the middle with a very small group of a few ill-defined pores. Body very slender. Only one pore opening freely in the middle of the under- side of the coxopleurae of the terminal segment; posterior to this are 1-2 incompletely developed glands without a pore opening. The species is distinguished by the unusually small size of the claws of the terminal legs (fig. 19), both tarsi of which are more slender than the tibia. : Although I have only seen a subadult example of this form, I must regard it as new; even if we disregard the number of the coxal glands, which as we know increase during epimorphosis, it differs from its nearest ally, #. brevis, in several respects. Locality.—Prince Albert, Cape Province. Eurytion trichopus Attems. (Figs. 24 and 25.) S$ 52 mm. with 65 pairs of legs. 950 mm. with 71 pairs of legs. The author writes of this species on p. 262 of his monograph, “first maxillae with 2 pairs of ‘Aussentastern’ both finely spined, those of the synocoxite short triangular, those of the first telopodite segment longer and slenderer.” I have already mentioned the incorrect use of the term “‘ Taster,’ but the expression “‘spined”’ is also misleading, as by spines we mean strong setae or bristles, while actually no bristles are present on these lateral lobes and no nerve fibres enter them, which shows how little morphological meaning this expression has. In actual fact there are on the lateral lobes of all Geophilomorpha South African Geophilomorpha (Myriopoda). 109 very minute prickles which apparently represent minute chitinous hairs. . trichopus has only short and extremely pale lateral lobes on the first maxilla. On the coxopleurae there are, according to Attems, “numerous large pores”; his fig. 227 shows these as being of different size, but he does not say whether they also occur dorsally. I therefore state that the pores of the coxopleurae are ventral, pleural, and dorsal in the g and 9. The posterior portion of the last sternite and the postero-ventral region of the coxae are so thickly covered with short bristles that the comparatively large sockets of these bristles resemble a porose area. This area of bristles decreases posteriorly throughout the whole of the ventral region in the 9, in the ¢ it widens on the inner side near the margin of the sternite; in the $ the coxopleurae exceed the hind margin of the sternite by } their length, in the 2 by only t. Terminal segment of last legs about $ as long as the penultimate in ¢ and 9; the sternal porose areas are larger than those shown in Attems’s fig. 226, e.g. almost twice as wide as their distance from the lateral margins in sternite 6. The first sternite with more than 20 pores in the ¢ and 8 in the 9. Im sternite 22 the area is broader and shorter, and on 23 it begins to divide into two lateral groups. | Locality.—I have identified this species from River Zonder End, and Table Mountain, Cape Town. Eurytion dentatus seems to be the commonest species of the genus. I have examples from Garies, Leliefontein (Namaqualand); Hout Bay, Smitswinkel Bay (Cape Peninsula); Swellendam; Grahamstown. 71-79 pairs of legs. Gen. PoLYGONAREA Attems. Three species have thus far been known of this genus, which is endemic in South Africa. Hurytion and Polygonarea are extremely similar. I should thus like to indicate that Polygonarea, besides the lateral longitudinal grooves of the coxosternum of the second maxilla (which are present in all Chilenophilenae), differs further from Eurytion in the glands of the coxopleurae of the terminal segment; the pores of these glands are not only larger and on an average fewer in number, but almost all of their openings are hidden. In addition there is a process on the inner side of the second maxillar coxa. 110 Annals of the South African Museum. The following is a key to the known species of Polygonarea:— (a) The 2-6 antennal joints densely covered with strong bristles which form 4-5 irregular whorls (fig. 15). Middle piece of the labrum distinctly toothed. Anal pores large. Trochanteropraefemur of the maxillipedes with a strong blunt tubercle on the inner sides. Clypeus without bristles. Head-plate anteriorly and anterior to the middle with a reticulated trans- verse band, between these bands on each side about 15 strong bristles. 51-53 pairs of legs ‘ : . litoralis n. sp. (6) The 2-6 antennal joints with only 1-2 ptisile whole (fig. 12), or if in joints 2-4 other bristles are present between the two whorls, they are either isolated or very small . ; 3 «(G54 (c) The 3-6 antennal joints with only one eel bristle Who storniton 2-18 and the 10 last sternites with a median, rounded, porose area, in addition a small group of pores on each side anteriorly. 51-57 pairs of legs monospathis Attems. (d) The 3-6 antennal joints with 2 bristle whorls. Most of the sternites with a transverse band of pores posterior to the middle which is more or less interrupted in the middle of the body and divided into two groups, in addition a small group of pores on each side anteriorly . : Pt (e) Middle piece of the labrum smooth, not toothed. Anal pores present. (Clypeus?) Trochanteropraefemur of the maxillipedes with a small, blunt tubercle. First sternite with 2 small pore groups. 61-65 pairs of legs kraepelini Silv. (f) Middle piece of the labrum toothed (fig. 13) . : , : , gy h. (g) Anal pores present. Trochanteropraefemur of the maxillipedes on the inner side entirely without a tubercle. First sternite with a transverse band of scattered pores. Clypeus with a dense mosaic reticulation, unspined except in the rounded area, head-plate anterior and posterior to the middle with a transverse band of reticulation, between these 2-3 long bristles laterally. Claws of the second maxilla 2-2 as long as the inner side of the middle joint of the telopodite. 51-59 pairs of legs . porosa nN. sp. (h) Analdriisen fehlen, Trochanteropraefemur der Kieferfiisse innen mit stumpfem Hacker . : : ; « Vee (i) Sternite 1 with 2 small pore groups, ener with a , renee area without bristles posterior to it. Coxae of the second maxillae with a process on their inner side which reaches the middle of the praefemur (as in porosa) not longer than basally broad. Claws as long as the inner side of the middle joint of the telopodite . . 5. porosa nodulifera n. subsp. (k) Sternite 1 without porose area. Clypeus with a transversely oval area and posterior to it a transverse row of 4-6 bristles. Coxae of the second maxillae with an inner process which reaches the inner apex of the pre- femur (or a very little below it) and is longer than basally broad. Claws of second maxillae only half as long as the inner side of the middle joint of the telopodite. 51-51 pairs of legs . : : oligopus Attems. The three forms porosa, porosa nodulifera, and oligopus are perhaps all races of one species, a question which must be decided by further investigation. | : weasel hnads South A frican Geophilomorpha (Myriopoda). ala Polygonarea litoralis n. sp. (Fig. 15.) 2 34-48 mm., with 51 or 53 pairs of legs. This species can be distinguished from all others by the short and dense bristles of the head-plate and antennae (fig. 15). ? Clypeus with dense reticulation but entirely without bristles, the area also only with fine punctuations. Middle piece of the labrum weakly toothed. Maxillpedes with a stout tubercle on trochantero- praefemur, a pointed tooth at the base of the claws. The porose areas of the sternites resemble those of most other species in having posteriorly a transverse, in the middle a more or less interrupted zone of pores, and anteriorly behind the anterior angles two rounded groups. Anterior sternites with extremely short weak bristles. Legs partly clothed with long bristles, the terminal legs of the 9 with its last segment little shorter than the penultimate one, with stout claws which are, however, exceeded by 4 long bristles. Anal glands present. Locality.—Only the 9 of this species is known from Hout Bay, Cape Peninsula, and from Keurbooms River near Knysna. Polygonarea porosa n. sp. (Figs. 12-14.) ~Qandg 29-52 mm., with 51-59 pairs of legs. Process on the inner side of the coxae of the second maxillae at most reaching the middle of the praefemur. Sternite I on each side with a small transverse porose area. Coxopleurae of the terminal legs with 4—8 large concealed pores on each side. Anal glands present. Attems portrays in his fig. 267 of P. monospathis a projecting lobe densely covered with short fine hairs on the inferior side of the coxae of the terminal legs, which he also gives as a generic character, assuming that it occurs in all species of Polygonarea. In the form before me the 2 and ¢ is provided on the posterior margin of the coxopleurae with a more or less projecting area clothed with short hairs. I have, however, not observed an outwardly projecting lobe. The tooth-like appearance of the outer extremity of the first and second telopodite joints of the second maxilla as shown by Attems in fig. 266 for P. monospathis is not present in the forms seen by me. _ Locahty.—Grahamstown and Hermanus. Perhaps these localities represent two races, as I observed differences in the forms of the VOU. XXXI1, PART .o: 9 112 Annals of the South African Museum. poison sacs. To decide this point, however, more specimens are required. Polygonarea porosa nodulifera n. subsp. 23-24 mm. 45 or 47 pairs of legs. Claws of the maxillipedes serrated on the inner side. Coxopleurae of terminal legs with 5+5 large concealed pores. Other than the characters already given in the key there is little of importance to describe. Locality.—Hogsback near Alice, Eastern Cape Province. Gen. MESOSCHENDYLA Attems. Of the three known forms of Mesoschendyla two are found in South Africa and one in Java, so that it is very probable that other species will be found in the intervening regions. The new species is closely related to monopora Att., but can be distinguished from it as follows :— monopora Att. 2 with 59-69 pairs, ¢ with 57-59 pairs of legs. Anterior margin of the coxo- sternum of the maxillipedes incised. Coxosternum of first maxillipede with- out, telopodite with lateral lobes. Claws of the second maxilla serrated. Sternites 2-30 (32) with round to transversely oval porose areas. Ven- tral surface of body well provided with hairs, dorsal surface almost smooth. Terminal legs well provided with hairs. Terminal joint in the § almost half as long as the penultimate joint. The two joints of the gonopods about equal in length. cribrifera n. sp. Q and ¢ with 55 pairs of legs. An- terior margin of coxosternum of maxillipedes with a median incision. Coxosternum and telopodite of first maxillae with lateral lobes (fig. 1). Claws of the second maxillae below with striated lamellae. Sternites 2—25 usually with a triangular, anteriorly truncated, posteriorly rounded, porose area (fig. 3). Tergites with two inter- calary tergites provided with a trans- verse row of fine hairs. Sternites sparsely covered with fine hairs. Hairs of the terminal legs so minute that they appear almost smooth. Terminal joint of § 2% as long as the penultimate, comparatively longer than in monopora. The basal of the two joints of the gonopods about twice as long as the terminal one in the ¢ (fig. 2). M. cribrifera n. sp. (Figs. 1-3.) g and? 34mm. Body slender and pale. The teeth of the labrum cannot be counted with precision as they are indistinct in the middle. Antennae with reticulated structure only on the first joint; the fine hairs commence on the first joint, ee ee ee ee ee eee eee a — | ee a a South African Geophilomorpha (Myriopoda). 113 there being no stouter hairs; the last segment with sensory pits on each side behind the middle. The claws of the second maxillae widening on the inner side into an extremely delicate lamella which appears to have very fine striations (fig. 1). It gives the impression that the minute teeth which form a serrated edge in related species are here fused to form a lamella, as the outer side is smooth and simple. There is a prebasal plate behind the head. The porose areas in sternites 2-25 become progressively less distinct in sternites 20-25 owing to their delicacy. The two isolated pores of the coxo- pleurae of the terminal legs are situated exactly as figured by Attems in fig. 108 for M. monopora, and the sternite has an identical shape. In the joints of the telopodite there are numbers of gland cells without recognisable pores. Locality. — Kamieskroon, Leliefontein, and Garies, all in Namaqualand. Gen. ASPIDOPLERES Porat. Aspidopleres intercalatus Porat. The 99 before me, 74-84 mm. in length and with 107 pairs of legs, agree very closely with Attems’ description of specimens from South West Africa in “Das Tierreich,” Lief 52, 1929, p.115. They were collected at the Aughrabies Falls and Kakamas, both on the Orange River. REVIEW OF THE GEOGRAPHIC DISTRIBUTION OF THE SoutH AFRICAN GEOPHILOMORPHA. Up to the present 13 genera are known from South Africa. These can be divided into two groups, the one consisting of endemic genera, the other of genera occurring outside of South Africa as well. Seven genera can be regarded as endemic, viz. Achilophilus, Aspido- pleres, Brachygonarea, Diphtherogaster, Geoperingueyia, Polygonarea, and Purcellinus, each of which, with the exception of Achilophilus and Polygonarea, are represented by a single species. Of the six remaining genera which are found outside the South African region, Orphnaeus (O. brevilobatus Newp.) is the only one which is widely distributed (and that only in the Tropics), and is thus without special interest here, seeing, moreover, that Orphnaeus is the only genus of the Geophilomorpha appearing in South Africa (including the Transvaal) which does not possess an endemic species in this region. All the other South African species are thus endemic, a proof of the spatial and climatic isolation of South Africa which has endured through vast periods of time. 114 Annals of the South African Museum. With regard to Mesoschendyla, besides the three species from South Africa, only one is known from Java, which differs so strongly from the others that 1t may prove on closer investigation to represent another genus. Aphilodon, besides South Africa, occurs only in South America. Schendylurus is found in South, West, and North Africa and in South America. Ballophilus is represented in South and Central Africa as well as in the Indo-Australian region; Hurytion, as the genus with the widest distribution, is most strongly represented in South Africa, but two species are found in Chile and three in Australia. Although we might be inclined to regard the South African fauna as having its closest relationships with that of Central and North Africa, judging from the present land distribution, this is not supported by the distribution of the Geophilomorpha, of which South and Central Africa have only two genera in common, South and North Africa only one. On the other hand, three genera are found both in South Africa and South America, while the similarity between South Africa and the Indo-Malayan region is expressed by having three genera in common. There is therefore an undeniable relation- ship between the regions of the Southern Hemisphere (South Africa, South America, and Australia), which may be explained by ocean currents, or sunken land masses, or by both. The Geophilomorpha as animals without a calcareous skeleton can easily float on water, and as at the same time they are well adapted for attaching themselves to objects and for concealment in narrow holes and fissures, it is not improbable that, in spite of the enormous distances between South America, South Africa, and Australia, in the course of long periods of time they were occasionally transported by trees drifting on ocean currents from one of these regions to another. The North African deserts represent such a formidable barrier that only a single genus, Schendylurus, has been able to overcome it. With Europe, Asia, and North America, South Africa has neither a genus nor a species in common (the widespread tropical Orphnaeus brevilabiatus excepted). It is especially remarkable that the Himantarudae, though very well represented in North Africa, are completely absent in South Africa. The Oryidae are evidence of the very old connection between Africa and South America, as they are practically found in these two continents alone, where they are represented by several genera; these genera are, however, with the exception of Orphnaeus, different. In a few cases members of the Oryidae have reached India, but with a ee ee ee ——— South African Geophilomorpha (Myriopoda). 115 these exceptions they are absent from Asia, Europe, North America, and the Malayo-Australian region. When all the distributional evidence is taken into consideration, there can be no doubt that the Ethiopian Geophilomorpha are most closely allied to those of South ‘America. EXPLANATION OF PLATES IX, X. FIGS. 1-3. M esoschendyla cribrifera n. sp. 1. First and second maxillae, seen from above. p, pore of the maxillary glands on the coxosternum of the second maxilla. x 220 2. Gonopods (gp) and penis of § (p), seen from below. x 220. 3. Sternite of seventh segment. x 125. 4-6. Aphilodon porosum n. sp. — 4. Telopodite of maxillipede, seen from below. x, boundary between the tarsus and ungulum of the tarsungulum (tau); y, termina- tion of the suture between femur (fe) and praefemur (prf), tk, notch of trochanter. x 125. 5. Sternite, coxopleura, and trochanter of terminal segment, seen from below. x 125. 6. The third sternite from the posterior end of body. x 125. 7. Aphilodon weberi Silv. 9, Sternite, coxopleura, and trochanter of the terminal segment, seen from below. x 125. 8,9. Aphilodon caffrarium n. sp. 8. Telopodite of maxillipede, seen from below. x 125. 9. Sternite, coxopleura, and trochanter of terminal segment in the 4, seen from below. x 125. 10. Telopodite of maxillipede, seen from below. x 220. 11. Sternite, coxopleura, trochanter, and praefemur of the terminal segment, the glands indicated by dotted lines. x 220. 12-14. Polygonarea poyosa n. sp. 12. The two basal antennal joints. x 125. 13. Labrum seen from below. x 340. 14. Sternite, coxapleura, and trochanter of the terminal segment; pores of the coxal glands indicated by dotted lines. x 125. 15. Polygonarea litoralis n. sp. The two basal antennal joints. x 56. 16,17. Eurytion lawrencei n. sp. 16. Fifth sternite with pore groups. x 125. 17. Labrum (la) and the portion of the clypeus anterior to it with window- like smooth areas (fe). x 125. 18,19. Eurytion gracillimus n. sp. 19. Labrum seen from below. x 220. 19. The two tarsal joints of the terminal leg. x 125. 20,21. Hurytion brevisn. sp. Q. 20. Sternite and coxopleura of the terminal segment, seen from below. x 125. 21. The two tarsal joints of the terminal legs. x 125. 116 Annals of the South African Museum. ; 22,23. Hurytion kalaharinus elongatus n. subsp. 22. Labrum seen from below. x 220. 23. Sternite and coxopleura of terminal segment, seen from below. x 125; below and at the side, the receptacula of two coxal glands. x 220. 24,25. Eurytion trichopus Att. 24. Labrum seen from below, anterior to it a portion of the cell structure of the clypeus. x 340. 25. Area of the clypeus with a portion of the cell structure. x 340. Ann. S. Afr. Mus., Vol. XX XII. . Plate IX. <0. 66 e- “= Sene.6 2 Mi ° 38 Cheba ral . e's ev. e.e 2.6 tek a e ° # Sie ©,% 5 Pe id na . Lane iaten oo Mae ee eve ° ° Sine Me |e Sal SY an ove "ese. 2 eee of e- 2 axe t ee ° ens Neill & Co., Ltd. K. W. Verhoeff. Ann. S. Afr. Mus., Vol. XX XII. Plate X. K. W. Verhoeff. Neill d Co., Lid. ’ “Vy - ‘ ¥ ' , 7% © 2 we &) vr ' a ¥ ‘3 * , hie | a R + Ae : s * . wns - . - 7 ‘ ~ = ” - = = } - - = r v (We) 9. A New Genus of Colobognatha from the Cape Peninsula. By Dr. Karyt W. VERHOEFF, Pasing, Munich. (With Plate XI.) In 1929 C. Attems made the first comprehensive study of South African Diplopods in his “Myriopoda of South Africa” (Ann. 8. Afr. Mus., vol. xxvi), in which it may be noted that only a single Colobog- nathid has been recorded from this region, viz. the Polyzoniid Burema nasuta, Attems. This form its author rightly placed in his subfamily Hirudisominae, as it possesses both of the two most im- - portant characters of this subfamily: firstly, the vasa differentia opening on the coxae of the second pair of legs in the 3; secondly, the position of the pore remote from the suture. Seeing that both these characters, as is clear from figs. 5 and 6, also hold good for the new genus Cylichnogaster, they can henceforward only be used in connection with the Hirudisominae. In the key (loc. cit., p. 417) which Attems has given for the three genera of Huirudisominae, Hirudisoma, Burenia, and Rhinotus, he makes in the first place the following distinctions: la. Two or three ocelli on each side, anal segment completely concealed by the prae-anal segment and not visible from above : Hirudisoma Carl. 1b. One ocellus on each side, anal segment partly visible from above Burenia and Rhinotus. Generally speaking, not much weight can be attached to the differ- ence in the number of eyes, as here a mistake may easily occur on account of the deep pigment and the indistinctly differentiated ocelli. With regard to the distinction dealing with the anal region, however, this must receive a more precise interpretation, as the expressions “not visible” and “partly visible,’ when applied to the new genus and to my fig. 4, give rise to justifiable doubt. Thus in C. lawrencei, though the anal segment is indeed beneath the prae-anal one, its posterior apex protrudes a little beyond it, and is on this account more or less visible from above. Seeing that Cylichnogaster has 2+2 ocelli, I should be compelled on this character to place it with Hirudisoma, alternatively in consideration of the prae-anal segment, with Burena. If, on the other hand, we consider not the variable 118 Annals of the South African Museum. amount of the prae-anal segment visible from above, but the different positions of the anal valves at the dorsal posterior border of the prae-anal segment, we have far more precise criteria to apply to the relative positions of the anal and prae-anal segments. Thus in Burenia apparently the whole, but actually only the posterior halves, of the anal valves lie posterior to the prae-anal segment, while in Hirudisoma they lie well in front of the posterior margin of this segment.* As this latter applies almost equally well to Cylichnogaster, there is no doubt that it is, of the two genera, more closely related to Hirudisoma than to Burenia. As a survey of the three genera I give the following key:— A. The anal segment f with its posterior half posterior to the hind margin of the prae-anal segment. Head with one rounded ocellus on each side. Trunk of 3 with 64-88 segments; the trunk cannot be enrolled. Sixth antennal segment distinctly narrowed from base to apex. Sternites of second pair of legs in § with a triangular process between the coxae. Collum with several rows of bristles, its postero-lateral angles bluntly rounded. An- terior gonopods 6-jointed, the post-femur being followed by a fused tibio-tarsus { , : : : : . Burena. B. The anal segment lying wholly in front of the hind margin of the prae-anal segment. Head on each side with two or three ocelli situated in a large elongate pigmented area (fig. 1). Sternites of the second pair of legs in ¢ truncate . ‘ ; ; ; : ; ; ‘ ‘4 6 ameitd: C. Hind margin of the prae-anal segment a considerable distance behind the pos- terior margin of the anal segment. The trunk can be only partially flexed, not enrolled, the tergites are less curved. Head but little elongated anteriorly so that it forms an approximately right-angled triangle in front of the antennae. Anterior legs of g with a foliiform structure instead of a claw é : : : ‘ ‘ é : . Hirudisoma. D. The hind margin of the prae-anal segment lies almost exactly in a line with that of the anal segment, so that the posterior apices of the anal valves (accord- ing to the position in which the telson is viewed) show a little or not at all beyond the hind margin of the prae-anal segment (fig. 4). The trunk, which in the § consists of only 20-26 segments, can be completely enrolled, and the tergites are thus strongly curved. Head more decidedly elongated anteriorly so that in front of the antennae it appears slender, the apex bluntly pointed (fig. 1). Anterior legs of § with normal claws, these, * See also the ninth paper of my “‘ Beitrage zur Kenntnis palaarctischer Myriopo- den,” 1899, Archiv. f. Naturgesch., Berlin, Bd. 1, H. 3, Taf. xix, fig. 1. _ + Attems’s drawing of the telson of Burenia in fig. 424 of his “Myriopoda of South Africa” is not wholly clear, as the hind margin of the prae-anal is not put in. He has indicated the prae-anal segment with an # and the anal valves with an av. { Compare, however, what I have said below about the two different interpreta- tions of the gonopod articulation. A New Genus of Colobognatha from the Cape Peninsula. 119 however, with an additional slender lappet (fig. 2). Sides of sixth antennal segment almost parallel. Anterior gonopods 7-jointed, the tibia and tarsus remaining completely separate . ‘ : . Cylichnogaster. With regard to the telson I should like to remark as follows: In all Hirudisominae the anal segment is sunk into the ventral surface of the prae-anal segment, so that the anterior ventral boundary between the segments is indicated by a semicircular line, while the dorsal hind margin always forms a transverse one. The anal valves lie ventrally far down this semicircle. The deeper the anal segment is sunk in these forms the more degenerate is the anal tergite. In Hirudisoma it is therefore rudimentary, in Cylichnogaster it forms a small strip, while it is most strongly developed in Burenia; the latter is, however, not correctly portrayed in Attems’s fig. 424, as the transverse groove in the middle is not shown. Cylichnogaster lawrencei n. sp. $6: length 32-5 mm. @: length 33-51 mm., juvenile 34 mm. Both ¢ and 2 with 20-26 somites, juvenile with 18. Tergites dark brown, smooth, shiny, strongly curved. In adaptation to the animal’s ability to enroll, the body is more strongly narrowed, both anteriorly and posteriorly, than in the allied genera. A further adaptation appears to me to be the strong thickening of the sternites and pleurites and the excavation of the pleurites, as seen in fig. 3. While in the allied genera there are indentations between the tergites which give the sides of the body a serrated appearance, in Cylichno- gaster, on account of the sloping sides of the tergites, caused by ‘a considerable reduction in their anterior width (fig. 5), there are still deeper incisions between the paratergites, by means of which the telescoping of these tergites is brought about during enrolment. The head is only covered by the collum in its posterior fourth (fig. 1), but in enrolment it is pressed strongly downwards and backwards towards the venter. There is throughout a connective relation between the sternites and pleurites, which takes the form of a longi- tudinal suture between them. The pleurites have only a membranous connection with the tergites, so that in a preparation they can easily be separated from each other. Dorso-ventral muscles connect the tergites on each side with the sides of the sternite and assist in raising the ventral surface.* The two sternites belonging to each pair of * It should be noticed that in enrolment the tergites are to a large extent forced away from each other, the paratergites on the contrary towards each other. 120 Annals of the South African Museum. pleurites are almost equally broad, but otherwise very differently formed, the anterior being only about as long as the posterior. All sternites lie overlapping one another, the hind margin of the anterior covering the front margin of the posterior one; the anterior angles of the anterior sternite have a membranous connection with the pleurites, the anterior angles of the posterior sternite are triangular, narrowly produced anteriorly to fit into a small cavity in the anterior one. The coxae in the middle of the body are separated from each other by $ their own width and implanted in a curved slit on the hind margin of the sternite. Farther forward along the body the coxae draw nearer to each other (fig. 6), but still remain separate through- out. The small coxal sacs begin on the third pair of legs. There is a small lappet close to the claw of the anterior pair of legs of the 3, which is more than twice as long as the claw (figs. 6 and 7), while next to it there is in most legs a long bristle. Head with sparse bristles (fig. 1), two of which, characterised by unusual length, are situated anteriorly between the anterior ocelli of the eyes. The antennary pits are distant from each other about twice the breadth of the basal segment of the antennae, and are somewhat obliquely inserted, their anterior end being directed inwards. In the same way the eyes are also directed obliquely, being placed in two pigmented areas which anteriorly are about as far from each other as the length of the pigmented area itself. Hach pigmented area contains two ocelli, a small round one anteriorly, a large oval one posteriorly; these ocelli are only discernible when the incidence of the light under the microscope is favourable. The antennae consist of six very distinct stout joints, with a seventh, small and disc-like, appearing at the apex of the sixth joint, while from this again pro- trude the four well-known olfactory cones; the antennae are thus altogether 8-jointed. On the distal outer ends of joints 5, 6, and 7 are several sensory rods, thus agreeing with Burenia, according to Attems’s fig. 413; these rods are, however, easily seen only on the fifth jot, while on the sixth and seventh they are often indistinct owing to pigment and bristles. The spination of the tergites is in general very sparse and weak and almost always confined to the sides and hind margins (figs. 2 and 5); superficially most of the ter- gites appear to be quite smooth, but under higher power 2-3 very fine hairs can be seen on their rounded posterior angles. Stouter and sometimes longer bristles are found only on the telson (fig. 4). Four long bristles project posteriorly apparently from the hind margin of the anal segment, but actually from the hind margin of the prae- —_—- A New Genus of Colobognatha from the Cape Peninsula. 121 anal tergite. The pores (po, fig. 5) open on the paratergites, where they are equidistant from the anterior, posterior, and lateral margins. Attems has shown in the walking leg sternites of Burenia (though not in those of the anterior and posterior gonopods) a bluntly triangular projecting lappet between the coxae, while in Cylichnogaster the margin of the sternite between the coxal (figs. 3, 6, 8, and 10) is always completely truncate without a projection, and this posterior truncated portion usually bears two bristles para-medially, while the sternites are otherwise smooth in contrast to the outer edges of the pleurites, which are more or less fringed with hairs (PI. XI, fig. 6). The walking legs are broad where their coxae join the sternite, but they terminate with a very slender tarsus. In the second pair of legs in the ¢ (fig. 6), the medially fringed tarsus is also much more slender than the tibia, differing in this from Burenia, where the seg- ments, according to Attems, are almost equally broad. On cursory inspection the anterior and posterior gonopods of Burema and Cylichnogaster seem very alike, but on closer examination considerable differences and also some difficulties appear. While the gonopods of Burenia have been described as having 6 segments, I found them to consist of 7 in Cylichnogaster, a difference which I have attributed in my key to the separation of the fused tibia and tarsus. This difference can, however, be explained in another way. In the gonopods of both genera, one of the segments of the telopodite (fe, fig. 8) is distinguished by its outstanding size, and broadly ‘triangular shape, caused by its strong external curvature. As can be seen from fig. 8, the gonopods of Cylichnogaster have the same - articulation as the walking legs, except that a normal claw is no longer present. The large triangular segment therefore corresponds to the femur; between it and the coxa are two other segments narrowed on the inner side—a large praefemur, and a smaller trochanter. Distally to the femur are three more segments differing little in size, viz. post-femur, tibia, and tarsus; the last named is produced into such a striking tooth or claw-like process, bent strongly backwards and basalwards (figs. 8 and 9), that we may regard it as a coalescence of the tarsus and claw, 7.e. a tarsungulum. When we apply to these gonopods the interpretation given by Attems in his fig. 423 of the gonopods of Burenia, and at the same time consider the interpretation of the tibio-tarsus as expressed in my key, it appears that on the one hand there are, between the large triangular segment and the tibio-tarsus, two segments in Burenia, as opposed to only one in Cylichnogaster; whilst, on the other hand, there are, between the large 122 Annals of the South African Museum. triangular segment and the coxa, two segments in Cylichnogaster but only one in Burenia. There are two possible explanations: Either, A. The large triangular segment (fig. 8) is the femur in both genera. In this case the trochanter has disappeared in Burenia and the praefemur has been reduced to a small outer triangle, while the terminal segment is not the tibio-tarsus (as it is treated in my key), but a tarsungulum, as in the other genus (Cylichnogaster) ; Or, B. The large triangular segment is only the femur in Cylichno- gaster, while in Burenza it is the praefemur; then the small inter- mediate segment developed only on the outside is the trochanter, the femur being, however, reduced and pushed more inward. The femur is followed by the post-femur and tibio-tarsus, conforming with the interpretation expressed in my key. Which of these two views is the correct one will be finally decided by making use of the character of the musculature. The muscles of the gonopods of Cylichnogaster I have indicated in fig. 8, but unfortunately nothing is known of those of Burenia. As can be seen from fig. 8, there are in the telopodite of the gonopods of Cylichnogaster two muscle bundles, one behind the other, agreeing with those of the posterior gonopods (fig. 10). If, however, we compare these muscle bundles with the musculature of the walking legs of Polyzoniidae, e.g. with figs. 3 and 4 of my third paper on the legs of tracheates (Sitz. Ber. d. Ges. natur- forsch. Fr., Berlin, 1903, No. 2, Taf. 1), we readily see that in them, exactly as in the gonopods, the main line of division in the musculature of the telopodite les between the prae-femur and femur, which confirms my interpretation of the articulation of the gonopods as shown in fig. 8. The decision as to whether the largest segment of the gono- pod telopodite is femur or prae-femur in Burenia, therefore, depends entirely upon the position of the division between the two muscle bundles of the telopodite. At the extremities of the gonopods of Burenia there occur, according to Attems (figs. 414 and 423), an inwardly projecting lappet at the base of the terminal segment (lo). This lappet is completely absent in Cylichnogaster. If Attems’s fig. 423 is correct then the gonopods in Burenia should be much more deeply sunk in the sternite than in Cylichnogaster; moreover, the hind margin of the sternite in Cylichnogaster is truncated, in Burenza, on the other hand, broadly concave. The posterior gonopods of Cylichnogaster (fig. 10) strongly resemble those of Burenia which Attems has described as 7-jointed, and are likewise much less deeply sunk in the sternite on the one hand and more strongly bent towards the anterior on the other. The 7 segments . 4 : A New Genus of Colobognatha from the Cape Peninsula. 123 of Burenia are composed of coxa, trochanter, prae-femur, femur, post-femur, tibia, tarsungulum, so that there is an almost normal articulation. Although the curved blade-like terminal portion of the posterior gonopods is practically similar in the two genera, except for the slender apex which is simple in Cylichnogaster, one cannot speak of a normal articulation in the latter genus. The four distal segments are very similar, the penultimate one in Burenia being, however, twice as broad as long and in Cylichnogaster longer than broad; in the proximal half, however, z.e. in the neighbourhood of the basal muscle bundles, the articulation has become obliterated; a trochanter is altogether absent, and the boundary between coxa and prae-femur is only incompletely developed, and that on the outer side. Occurrence. I am indebted to Dr. R. F. Lawrence, formerly Assistant in Charge of Arachnida and Myriopoda at the South African Museum, Cape Town, for the specimens 3 $g and numerous 99 which form the basis of this paper; the new form is named in his honour. They were discovered by him beneath damp rotting logs in one of the valleys of the mountain slopes near Chapman’s Peak, Cape Peninsula. Note.—From the point of view of distribution this Polyzoniid is of extraordinary interest for the reason that it is decidedly more closely allied to the Mediterranean genus Hirudisoma than to the South African Burenia, which is evidence that, at an earlier period of the earth’s history, there was an interchange of forms between North and South Africa. It is highly probable that still more repre- sentatives of the Colobognatha may eventually be found in South Africa. EXPLANATION OF FIGURES. (Plate XI, with Figures 1-10.) Cylichnogaster lawrencet nov. gen. et nov. sp. FIG. 1. Head, right antenna, and anterior region of collum, seen from above. x 125. Ze Collum. x 125? 3. A segment from the middle of the body in transverse section; tergite above, sternite (with coxa) below, and pleurite. x56. . Anal segment and the hind margin of the prae-anal segment (pr) seen from below. x220. a, anal valves; y, dorsal portion of the anal segment. . Lateral portion of two tergites from the middle of the body; n, suture; 1, longitudinal muscles; po, pore. x 125. i OL 124 Annals of the South African Museum. 6. Second pair of legs of, with sternite (b), pleurite (pl), pseudopenis (p). x 125. 7. Apex of second leg with claw and lappet. x 220. 8. Right gonopod with sternite (v) and the most medial scelerite of the pleurite (pl), seen from in front; co, coxa; tr, tronchanter; prf, praefemur; fe, femur; ps, post-femur; ti, tibia; ta, tarsus. x 220. 9. Apex of gonopod, seen from behind. x 220. 10. Posterior gonopod with its sternite, seen from below; v, anterior; h, posterior margin of the sternite; st, stigma; e, terminal piece. x 220. Ann §. Afr. Mus., Vol. XXXII. KW. Verhoeff. Plate XI. Neill & Co., Ltd. is, a’ 7 oy or Veh pr a es in oe ar , ? ; j oo asl 7 oy ee Bh 1% ? - . ¢ vat ad Ln Pe ' * | ’ os de : e MN 7 m = ! ‘ ) } \ ‘ ° ; ‘ \ 3 x t 4 Ms r be ( 125.) 10. New Colobognatha from South Africa. By Dr. Kart W. VeRHOoEFF, Pasing, Munich. (With Plate XII.) INTRODUCTORY REMARKS. Untit a few years ago not a single member of the Colobognatha was known from South Africa, and Attems in his Myriopoda of South Africa was the first to record a representative of this order, with its peculiarly modified mouthparts, in Burenia nasuta Att. I had therefore great pleasure in receiving from Dr. R. F. Lawrence, formerly of the South African Museum, Cape Town, an interesting series of South African Myriopods, among which I found three new species of Colobognatha, belonging to the genera Burenia, Rhyncho- mecogaster n. gen., and to the genus Cylichnogaster which has been described in a previous paper. Four species of Colobognatha belong- ing to three genera are thus now known from South Africa, and it can be safely predicted that many more new forms of this order await discovery, especially from the mountainous regions of South Africa. A study of the South African Colobognatha led me to question whether the definition of the Hirudisominae, as given by Attems in the work cited above and in Kiikenthal’s Handbuch der Zoologie, Bd. 4, 1926, p. 213, is a correct one? His key to the genera of the Hirudisominae is as follows:— la. Two or three ocelli on each side. Anal segment completely concealed by the preanal segment and not visible from above . Hirudisoma Cav. ‘1b. One ocellus on each side. Anal segment partly visible from above. yee 2a. Pleurites connected with the tergites by membranes. Anterior gonopod 4-jointed, posterior gonopod 3-5 jointed 3 : Rhinotus Cook. 2b. Pleurites nearly coalescent with the tergites : 3. 3a. Anterior gonopod 6-jointed, posterior gonopod iajotutes ue ieee wits a bristle only on the third joint : : : é Burema Attems. 36. Anterior gonopod 3-jointed, posterior gonopod 5-jointed, both pairs with groups of bristles bearing accessory spicules . : Siphonotus Brdél. THE TELSON OF THE COLOBOGNATHA. A complete reinvestigation of the telson of Hirudisoma as compared with those of other genera is very necessary, seeing that, on the one 126 Annals of the South African Museum. hand, this structure has not been clearly enough defined; on the other, the distinctions expressed in Attems’ key under la and 1) give rise to serious doubts. Furthermore, the difference in the number of ocelli is an untenable character and should certainly not be used in generic distinctions. In the telson of the Diplopods we have to distinguish between an anal and a preanal segment (figs. 1 and 2). As Attems speaks only of an anal segment it is doubtful if he means thereby the true anal segment or the whole telson. The distinctions in the anal segment expressed by Attems under la and 16 are made clear in fig. 1 for Hirudisoma, and fig. 2 for the other genera. In fig. 1 the anal segment is thus seen to be shifted far forward, while in fig. 2 it projects. somewhat with its posterior extremity. Kven so the distinctions between Hwrudisoma and the other genera have not been sufficiently defined. If we consider the Diplopoda in general, there is no doubt that a telson clearly visible from above represents a primitive condition as contrasted with one which has been shifted forwards, because the simple position of the somites, one behind the other, constitutes the most primitive condition that we can conceive. Thus fig. 2 shows the primitive structure of the telson in Rhynchomegaster (and Burenia), while in Hirudisoma it is indubitably of a secondary nature. We have thus in figs. 1 and 2 three segments posterior to the last leg-bearing somite, viz. a posterior legless segment (ul), a preanal segment (pr), and an anal segment consisting of two lateral anal valves (an). In Burenva and its allies (fig. 2) only the ventral part of the legless segment is visible from below, but in Hirudisoma (fig. 1) the telson (pr+an) has moved so far forward that a section of the dorsal part of the legless segment is visible behind it. The extreme hind margin of the body is thus part of the last segment in Hirudisoma, while in the genera represented by fig. 2 it is formed by the telson. In the latter the anal valves are placed so far back that they are situated entirely behind the legless segment, being at the same time only partly enclosed by the preanal segment, while they are also much shorter than the section of the preanal segment anterior to them. The anal valves of Hirudisoma, on the other hand, have not only drawn nearer the legless segment (fig. 1), but they are so largely enclosed by the preanal segment that they are longer than the section of the preanal segment in front of them, in which the ventral displacement in an anterior direction is especially marked. } ; | a ee, ee a eee New Colobognatha from South Africa. 127 From these conditions it appears that the distinctions between Hirudisoma and the other genera are more far-reaching and com- plicated than those used in the above-mentioned key of Attems. This difference in the structure of the telson runs parallel with another in respect of the mouthparts, up to now unknown. Hirudi- soma, on account of its mouthparts and especially the fairly well- developed wedge-shaped mandibles, is allied to Polyzoniwm. In Burema and its allies, on the other hand, there are no such wedge- shaped mandibles, showing that degeneration has already gone a long way. : : Similarly, the prolongation of the anterior part of the head in Buremia and its allies is far more marked than in Hirudisoma, indica- tive of an approach to the Siphonophoridae. The conclusion I am forced to draw from all these differences is that Burenia and its allies form a distinct family standing midway between Hirudisoma and the Siphonophoridae. Huirudisoma is, however, so widely different from the Polyzonudae that it represents a separate family between the Polyzoniidae and the Burenudae. The following are the points of difference between the two families :— A. The telson is so completely sunk in the preceding segment that it is not visible from above (fig. 1). The anal valves are so sunk in the preanal segment that they are almost encircled by it. The dorsal posterior margin of the preanal segment lies far in front of that of the preceding segment, the telson being less than half as wide as the latter. Head but little projecting so that anteriorly it forms a right angle or almost a right angle. The mandibles are developed as biting organs in the form of little wedges like those of Polyzonium. Family Hirudosomidae n. fam. (for Hirudosoma). B. The telson is not completely sunk in the preceding segment (fig. 2), the hinder end being thus to a large extent visible from above. The anal valves are only slightly sunk in the preanal segment and are situated so far back that they are much shorter than the ventro-preanal portion in front of them. The dorsal hinder margin of the preanal segment lies behind that of the preceding segment. The telson is much more than half as wide as the latter. Head strongly projecting anteriorly so that its apex forms less than half a right angle, thus appearing beak-like. The mandibles are not modified to form wedge-shaped biting organs. Family Bureniidae n. fam. (containing the genera Burenia, Rhyncho- mecogaster n. gen., Cylichnogaster, Rhinotus, and Siphonotus). The family Bureniidae can be divided into the following subfamilies :— a. Body short and broad, able to enroll. Trunk with 20-26 segments. Pleurites transversely rectangular 1. Subfamily Cylichnogastrinae (for Cylichnogaster mihi). ROL. SMX. PART 3. 10 ey 128 Annals of the South African Museum. b. Body longer and more slender, with 35-90 segments, never able to enroll. Pleurites of more or less quadrate form (figs. 3 and 4) 2. Subfamily Bureniinae (containing the remaining genera). The two genera Rhinotus (from Sierra Leone, Madagascar, Seychelles, and the Malayan-Australian region) and Siphonotus (from South America and the Malayan-Australian region) have 3-4 jointed anterior gonopods and 3-5 jointed posterior gonopods in common as distinguished from the South African genera Burenia and Rhynchomecogaster; they require, however, further investigation with regard#to other characters. Rhynchomecogaster n. gen. The genus is most closely allied to Burenia, from which, however, it can be distinguished as follows :— Rhynchomecogaster mihi. Metatergites clothed with such numerous and long hairs that they have a silky sheen. Collum provided with such numerous and long hairs (fig. 5) that 7-8 rows can be distinguished, which are, however, not all regular. Pleurites (fig. 3) with hairs on their outer posterior sides and also partly on the inner ones. Head acuminate anteriorly and without sensitive rods at its apex. One simple bristle near the claw of the anterior pair of legs in the g. Terminal segment of the anterior gonopods (fig. 6) drawn out into a recurved hook which reaches to beyond the femur. Terminal segment on the outer side without a lobe. Pre- femur of the anterior gonopod plate- like, well defined on the inner and outer sides (fig. 6). Burenia Attems. Metatergites sparsely clothed at the sides with short hairs so that the back appears smooth and shiny. Collum also with hairs mainly at the sides, but these so few and small that it appears hairless. Pleurites with only a few weak hairs at the sides on the outer and posterior portions (fig. 4). Head truncate anteriorly, usually with 2+2 very small sensory rods.* A small thickened tongue-like bristle near the claw of the anterior pair of legs in the 3 (cp. Attems’ figs. 416, 419, and 420). Terminal segments of the anterior gonopod produced into a process which at most only reaches to the femur. On the outer side of the terminal segment a marginal lobe (cp. fig. 9).7 Pre- femur of the anterior gonopod cunei- form, its inner surface so much con- tracted as to be indistinct (prf., fig. 8). CoMPARATIVE ANATOMY OF THE ANTERIOR AND PostTERIOR GONOPODS. It is easy for errors to. arise respecting the segmentation of the anterior and posterior gonopods of the Colobognatha, as the bound- * Attems made these much too large in his fig. 412. + Attems denoted the marginal lobe with “‘lo”’ in his fig. 414. New Colobognatha from South Africa. 129 aries of the segments are often difficult to make out, while sometimes they actually have partly disappeared (fig. 9)- This being so, it is important to study the relations of the muscles, because a correct idea of the segmentation can be gained from the well-defined points of insertion of these muscles. Regarding their segmentation, the anterior and posterior gonopods are homodynamous throughout— that is, of equal value—apart from the differences of the segments. Both pairs of appendages are 6-jointed in the South African Burenicdae. The muscles of the gonopods constitute two segments, one behind the other, which do not overlap each other, 1.e. a basal and an apical. The apical muscles always end basally on the inner side of a terminal joint which is drawn out distally into a process. Whether this process constitutes merely a claw or a tarsungulum is a point which cannot be decided with certainty. On account of the small change which had taken place in the anterior gonopods of the Colobognatha in contrast to the walking legs (in comparison with other Proter- andria), it is probable that they have retained their original postfemur. Then the segments marked (a) in figs. 6 and 9 would have to be regarded as the postfemur, (0) as the tibia, and (c) as the tarsungulum, and the same with the posterior gonopods (fig. 10). It is noteworthy that the muscle segments of the anterior gonopods have apparently different relations in Burenia and Rhynchomeccgaster, for although 5 segments follow after the coxa (co) in both genera, the basal muscle segment in Rhynchomecogaster traverses two joints, the terminal muscle segment three joints; in Burena the basal muscle segment traverses only one, the terminal four, joints. As the gonopod segmentation in Rhynchomecogaster is much more clearly defined than in Burenia, the former genus must be regarded as deciding the issue. There are also in these gonopods four distinct telopodite segments between coxa and terminal joint, as in the walking legs. In Burenia, on the other hand, I observed only one well-defined segmental boundary, the others being more or less indistinct and imperfect. This applies to the species before me, B. attemsi n. sp., while Attems, in his fig. 423 of B. nasuta, gives six distinctly separate segments, of which the second, the prefemur, appears as a wedge-shaped segment developed only on the outer side. The muscles of Burenia and Rhynchomecogaster gracile (fig.6) accord- ing to their positions thus show a difference of one segment as has been figured for Burenia attemsw in fig. 9. There is in Burenia attemsu, however, a fine striated band behind the coxa, marked (x) in fig. 9, which can be considered as a very poorly developed prefemur, 130 Annals of the South African Museum. and in this case the position of the gonopod muscles is the same in the two genera. The segment (prf) would then be the femur, (fe) the postfemur. For this interpretation I have used another system of lettering bracketed beside the previous notation. The faint line (y) (fig. 9) would not then define a segment seeing that it does not pass right across the gonopod. In both interpretations of the anterior gonopods of Burenia attemsit there is agreement with B. nasuta in that the prefemur is a wedge- shaped segment diminishing on its inner side, in contrast to Rhynchomeccgaster. The position of Burenia attemsw regarding the weak or incomplete formation of boundaries between the telopodite segments is note- worthy, as it shows that too much weight cannot be attributed to the number of gonopod segments alone in regard to genera, and I must draw attention to this fact in connection with unsatisfactory characters of the genera Rhinotus and Srphonotus. The physiological significance of the posterior gonopods is as yet unknown, and in this connection I refer to fig. 9 in which the slender terminal segment of the posterior gonopod can be seen hooked into a groove in the terminal segment of the anterior gonopod. What effect is produced by the mechanical coupling of the anterior and posterior gonopods is difficult to say when we know of no analogous occurrences in other Diplopod groups. It reminds me especially of the flagella in Julidae, the physiological significance of which I have repeatedly discussed. In my opinion the posterior gonopods are organs of stimulation whose terminal segments, by being inserted in the apical segment of the anterior gonopods, are guided to their correct position for stimulating the female in copulation. Rhynchomecogaster gracile n. sp. 3d 163-19 mm., with 48-56 segments. 2 20-215 mm., with 54-55 segments. The body appears on the whole to be greyish yellow without noticeable pigmentation. The narrowly produced fore part of the head gives the appearance of a concavity at the sides. Fore part of the head with 2 rows of bristles on each side below, the outer constituting a laterally project- ing fringe of which 3-4 are outstanding in respect of their length and curvature. Forehead with fairly long scattered hairs above; anteriorly on the inner side a long bristle between the ocelli. The New Colobognatha from South Africa. 131 antennae, which are provided with long hairs, are sunk into a circular basal lobe, the distal margin of the fourth segment reaching almost to the apex of the head. The transversely oval collum (fig. 5) has been already mentioned. Between the antennae 2+2 ocelli, the dense pigment round them forming an oblique ellipse on each side. The coxae of the walking legs, which carry the well-known small coxal sacs, are close to each other, remaining, however, separate in the median line, while between them at their bases is a small sternal horseshoe-shaped lobe, serving as a buffer (fig. 2). Walking legs with long bristles, prefemur and postfemur about as long as wide, femur and tibia on the other hand much longer than wide. Of the two smooth sternites of each segment the anterior one is rounded behind, the posterior one truncated behind, the postero- lateral angles, however, projecting backwards. The dense hairs of the tergites have already been mentioned. They are entirely confined to the metatergites, which are twice as long as the prozonites and divided from them by a fine transverse suture. Most of the tergites have five transverse rows of long bristles, and just anterior to their hind margins there are two fine transverse striations. There is a continuation of the transverse suture on the pleurites (fig. 3), the metapleurites being twice as long as the pro- pleurites, bristles being found only on the metapleurites. On their inner inferior margins the pleurites form a blunt angle in the middle, while the anterior margin is not only enlarged, but forms above on the outer side a knob-like projection (fig. 2). The pores of the tergites situated more or less as in Burenia. The anal valves of the telson quite hairless, while the preanal segment (fig. 2) has bristles above as well as at the sides. The anal valves agree with Burenia in having their hinder extremities reaching to a little beyond the preanal segment, and the posterior margin of the latter in the middle forms a bluntly rounded projec- tion. Inthese characters of the telson Rhynchomecogaster and Burenia differ from Rhinotus and Siphonotus in which the telson is completely rounded posteriorly and the anal valves end a little in front of the posterior margin of the preanal segment (cp. my paper on Diplopoda in Mjébergs Australischer Expedition, Arkiv for Zoologie, Stockholm, 1924, Bd. 16, N. 5, Taf. V, Abb. 93 and 100, Rhinotus mjébergi and Siphonotus latus Verh.). The anterior gonopods of Rhynchomecogaster have been mentioned above. The following may be added (fig. 6). 132 Annals of the South African Museum. The 6 segments described are all clearly separated from one another and the hairs are of considerable length, longer than in Burenia. On the outer extremity of the coxa there is a demarcated triangular field; this is either a membrane or the indication of a trochanter. The terminal segment is drawn out into a powerful dagger-like process which has an oblique strip bordering it below at its base and a small notch on its outer side. The posterior gonopods (fig. 7) thus correspond fully with the anterior gonopods in their segmentation. Only on the prefemur are there two fine hairs. The prefemur is not so broad as that of Burenia, and the same applies to the two segments proximal to the terminal piece. Locality.—The numerous specimens before me have the sexes in the proportion of ¢:2=1:3. Most of the specimens are from Noordhoek, Cape Peninsula, the remainder being from Hermanus, River Zonder End, and Bredasdorp. From these last three locali- ties, [ must, however, add, I have seen only females. Burema attemsw n. sp. $ 11 mm., with 52 segments. 9 16 mm., with 53 segments. (The 3 of B. nasuta Att. has 64-88 segments.) Dorsal surface either a uniform whitish yellow or with 1-2 small transverse greyish-brown stripes on the tergites. 2+2 ocelli which le obliquely close behind each other (according to Attems nasuta has only 1+1 ocelli, but this is open to doubt as it is easy to overlook the second ocellus on account of the centration of pigment in the ocellar area). As far as the characters given by Attems go, nasuta agrees with attemsu, but important differences exist in the anterior and posterior gonopods which have to a certain extent already been discussed. The boundaries of the gonopod segments are sometimes indistinct or incomplete, while in nasuta they are well defined. The hairs in attemsi are much shorter than in nasuta, this being especially notice- able in the terminal segments. The process of the terminal segment in nasuta is distinctly stronger, projects far out when seen in profile, and is obliquely truncate at its apex. In attemsw, on the other hand, the process of the terminal segment (fig. 8) is not only much shorter, so that seen in profile it hardly projects at all, but it terminates in two small teeth which are opposed to each other, the outer one being provided with a further accessory nodule. The prefemur of the posterior gonopods is provided with a bristle, while in attemsu New Colobognatha from South Africa. 133 the posterior gonopods are altogether muticous (fig. 10). The femur of nasuta is much wider than long, while in attemsiz it is a little longer than wide. Attems figured a trochanter between the coxa and prefemur for nasuta of which I have seen no sign in attemsiz. Locality—Among several female and immature specimens from Hermanus there was only 1 3. Other individuals from River Zonder End and French Hoek. Attems described his nasuta from Knysna, Cape Province. EXPLANATION OF PLATE XII. FIG. 1. Hirudisoma carniolense Verh. Posterior end of body seen from below. an, anal valves; pr, preanal segment; ul and 1, the two preceding segments; m, detractor muscles of the telson. x 125. 2, 3. Rhynchomecogaster n. g. gracile n. sp. 2. Posterior end of body seen from below (only the coxae and pre- femora of the last pair of legs drawn in). x 125. 3. Pleurites of the gonopod segment. v, anterior, h, posterior, u, inferior, 0, superior margin; su, suture; z, knob-like projection. x T25: 4. Burenia attemsii n. sp. Pleurite of the gonopod segment. x 125. 5-7. Rhynchomecogaster gracile n. sp. 5. Collum seen from above. x 125. 6. Anterior gonopod seen from in front. fe,femur. x 125. 7. Posterior gonopod seen from in front. x 220. 8-10. Burenia attemsii n. sp. 8. Anterior gonopod with its sternite (vv) seen from in front. prf, prefemur; fe, femur; m1,m*, muscles. x 125. 9. Anterior and posterior gonopods seen from behind. The hooked terminal rod of the posterior gonopod is inserted in the terminal joint of the anterior gonopod. co, coxa; vv, anterior, vh posterior sternites. x 220. 10. Posterior gonopod seen from behind. co, coxa; pri, prefemur; fe, femur; a, b, c, the three terminal segments. x 220. a Set. Th a ve! 7. ae i Aa peti - La ays ys abe D5 : Ann. S. Afr. Mus., Vol. XX XII. | Plate XII. K. W. Verhoeff. Neill & Co., Lid. 7% as clea 11. New Acridudae from South Africa.—By K. H. L. Key, M.S8c., Ph.D., D.I.C., Canberra, Australia. (With Plates XIJI-XVII and 2 Text-figures.) Most of the material which forms the basis of the present paper was collected by myself during the years 1928-1931. In the case of Betiscoides meridionalis Sjést., my own material was supplemented by specimens from the collection of the South African Museum, from which the unique type of Pyrgomorphella rugosa was also obtained. The two specimens of P. variegata are from the British Museum collection. J have to thank the Director and Dr. A. J. Hesse of the South African Museum and Dr. B. P. Uvarov of the British Natural History Museum for permitting me to use this material, and for putting at my disposal all the facilities of the two museums. The Type of Pyrgomorphella rugosa has been deposited in the South African Museum, and the Type and allotype of P. variegata in the British Museum. All the other Types are in the British Museum, and paratypes in both museums. CATANTOPINAKH. Genus BETISCOIDES Sjést. (Ark. f. Zool. 15, No. 22, 1923, p. 12.) The material before me includes Sjostedt’s B. meridionalis and two new species. These enable us to see the genus in a new light, and Sjostedt’s description is shown to require modification. In particular, the differences mentioned by him between the genera Betiscoides and Betisca are seen, with one exception, to be inconstant and unreliable, while two characteristic differences are not mentioned by him. I have therefore undertaken a redescription of the genus Betiscoides and of B. meridionalis in the light of the new material, transferring a number of characters from the generic description to the species. Body slender or very slender, smooth, apterous. Antennae more or less tapering, evenly punctured. 136 Annals of the South African Museum. Head from above conical; fastigium of vertex produced, poorly or not at all marginated; its surface strongly and evenly convex sideways. A faint median carina is discernible at any rate at the extreme apex of the fastigium. Head in profile nearly straight above; face nearly straight and strongly sloping. Frontal ridge between the antennae compressed, not sulcate, but becoming sulcate lower down; margins of ridge more or less diverging towards the clypeus, either not reaching the clypeus or obsolescent by the time it is reached. Hyes longitudinal, not very prominent from above. Median ocellus very small; lateral ocelli absent. Pronotum cylindrical, with no true carinae; anterior margin slightly convex, posterior slightly concave. Sides of pronotum flat; lower margin straight or slightly incurved; anterior margin sloping forward, straight or more or less incurved in the lower part; posterior margin nearly straight in the lower part, more or less sloping forward in the upper part. Sulci of pronotum represented only by a short shallow sulcus somewhat behind the middle of the lobes, which does not reach the disk nor the lower margin of the lobe, and one shortly behind the anterior margin of the lobe, which also does not reach the disk. Prosternal tubercle more or less compressed in a longitudinal direction. Meso- and metanota and tergites of abdomen traversed by a very fine median longitudinal carina, which may often be traced even on the pronotum as a very fine sulcus or carina. Tympanum absent. Anterior and middle legs short. Hind femora smooth, almost devoid of hairs, both upper and lower margins widely rounded; no median apical spine. Inner genicular lobes widely rounded, outer somewhat produced and pointed. Male.— Cerci very small, straight, conical. Supra-anal plate roughly diamond-shaped; the plate is traversed by a curved trans- verse sulcus, the convex aspect of the curve being directed anteriorly ; the basal part of the plate thus divided bears a longitudinal depres- sion. Subgenital plate elongate, convolute, conical. - Female.—Genital valves more or less recurved at the tips, not toothed, provided with hairs on the inside. Supra-anal plate longer than in the male, especially in the apical part; transverse sulcus not so distinct. Subgenital plate flattened, with an apical median triangular projection. Genotype : Betiscovdes meridionalis Sjést. New Acridudae from South Africa. 137 All known members of the genus are confined to patches of short reeds of the family Restionaceae. The British Museum collection contains a series of specimens of the genus Betisca from various localities in Australia, and these permitted a careful comparison to be made with the Betiscoides series. Two characters considered to be characteristic of Betisca were found to be variable. These were (1) the lateral depressions of the fastigium, and (2) the hind margin of the pronotum, which is sometimes as much incurved as in Betiscoides. On the other hand, a very characteristic feature of Betisca is the enlargement of the galeae of the maxillae, which are turned upwards, almost completely cover- ing the labrum, their points reaching to about the clypeus. In this respect Betisca resembles Acanthoxia, and differs completely from Betiscoides, in which the galeae are normal. Further, the apex of the fastigium is always sulcate in Betisca, carinate in Betiscoides. The characters of the prosternal tubercle, the galea of the maxilla, and the apex of the fastigium thus readily enable the two genera to be separated. Betiscoides meridionalis Sjést. (Loc. cit. 1923, p. 14, pl. 2, figs. 8, 9.) Male (Plate XIII, A, B, C, E.).—Body very slender and elongate. Antennae 24-jointed; distinctly triangular in cross-section; distinctly biserrate; tapering gradually to a fine point; in length subequal to the head and pronotum together. Head very elongate, about 34 times its width at the occiput; fastigium of vertex nearly twice the length of an eye; its margin takes the form of a broad, flat, very slightly raised portion of the head extending forwards from each eye, and narrowing, until at the base of the antennae it has become fairly sharp, but little raised; margins straight, converging very gradually to the apex; median carinula faintly discernible only near the apex of the fastigium. Head in profile nearly straight above, face nearly straight and very strongly sloping. Frontal ridge between the antennae projecting somewhat, and then bending upward to meet the fastigium, so that - in profile the head appears to be more or less obliquely truncate; margins of frontal ridge only slightly diverging and obsolescent shortly below the ocellus; lateral facial carinae ill-defined, and obsolescent above about the middle of the eyes. Eyes elongate-oval, almost oblong, about 12 times as long as their maximum width; 138 Annals of the South African Museum. their surface almost flat; the lower margin straight, and the upper margin almost so; both anterior and posterior ends widely rounded, the latter somewhat wider than the former. Hind margin of pronotum noticeably concave. Sides of pronotum with the lower margin almost straight; anterior margin sloping well forward, incurved; posterior margin straight below, sloping forward above; anterior lower angle slightly more than 90°, widely rounded; posterior lower angle 90°, rounded. Prosternal tubercle strongly laminate, with its length (measured along the cephalo-caudal axis) many times its width in the middle, where it is very thin, its anterior end strongly swollen and rather more downwardly projecting than the rest of the tubercle, so that the lower surface of the tubercle is not quite level, but slopes down gently from the anterior to the posterior end; posterior end very slightly swollen. Hind legs only just reaching the tip of the abdomen. Hind femora about seven times as long as their maximum breadth. Hind tibiae with 11 outer and 15 inner spines. Hind tarsi rather less than a third the length of the tibiae. | Apex of abdomen slightly upturned. Supra-anal plate with the basal part one-third to one-half the length of the apical part; the longitudinal depression well marked, straight, and narrow; apical part roughly triangular, but with the apex much attenuated and the sides somewhat incurved; a shallow depression in the apical angle. ; Subgenital plate extremely elongate and acutely pointed, about three times as long (measured from tip of anal plate to tip of subgenital plate) as the supra-anal plate. Lower margin slightly convex at the base; upper margin nearly straight. General coloration purple-brown, somewhat darker at the sides than above. A pale yellow stripe extends from the base of each eye across the cheeks, along the lower margin of the pronotum, and across the pleurae to the base of the hind femur. Head, eyes, antennae, and abdomen purplish brown, subgenital plate somewhat paler than the rest and greenish at the tip; fore and middle tibiae and tarsi greenish; hind femora greenish brown outside, pale green inside; hind tibiae pale brownish above, dark underneath; spines and spurs pale with black tips. Female.—Larger than the male. Antennae not quite as long as the head and pronotum together; head about three times as long as its width at the occiput; head in profile very slightly concave above. New Acridiidae from South Africa. 139 Hyes, relatively to the size of the head, smaller than in the male. Lower margin of side of pronotum distinctly incurved; anterior margin also strongly incurved; anterior and posterior angles each about 90° and widely rounded. Hind tibiae with 11-12 outer and 15-16 inner spines. Genital valves very feebly curved. General coloration pale green, faintly tinged with brown on the margins of the fastigium, bases of antennae, abdomen, and knees. Hyes dark; fore and middle legs and tip of abdomen bright green; hind legs green, tibial spines and spurs black-tipped. No trace of a pale stripe extending back from the eye. Measurements : Male. Female. mm. mm. Body . ; : : : ss (65) 49-5 Pronotum . : : ; A 3°0 3°8 Hind femur. ; : AMULOS 12-5 Antenna . ‘ ? : b-hkh-@ 10-5 Head . : : eee aaa seas 7:5 Subgenital plate . aC) Described from 28 males and 18 females. The two specimens selected for description are both from the Cape Flats, Cape Peninsula, December 1930 (K. Key). The remainder are from the following localities: Table Mountain, Cape Peninsula, 3500 ft. (16 males, 4 females); “Cape Town” (3 females); other parts of the Cape Peninsula (6 males, 5 females); Elgin, Cape Province (2 males); near Hermanus, Caledon Division, Cape Province (1 male); Paarl Road, Cape Province (1 male); Tradouw Pass, Swellendam District, Cape Province (1 male, 2 females); Great Winterhoek, Tulbagh, Cape Province, 4500 ft. (1 female); Steenbras River, Sir Lowry Pass, Cape Province (2 females). 3 Dates, where given, fall within the period November to May. In form the species is very variable. The specimens from the Cape Flats represent the extreme of elongation, the head being narrow and long, especially the fastigium, and the antennae and subgenital plate also very long.. The specimens from Table Mountain (Plate XIII, D) and all the other mountainous localities are more robust, the head being broader, and the eyes, owing to the relatively shorter fastigium, set further forward. The two extremes are united, however, by a complete series of forms, those from Hermanus, Elgin, and Paarl Road being intermediate conditions. There is also a great variation in size. In colour the males are relatively constant, being always dark 140 Annals of the South African Museum. with a pronounced lateral stripe backwards from the eye, and the hind femora usually green or greenish. The females, however, are extremely variable, all kinds of colour combinations being found, and many, like the specimen described, are of a more or less uniform colour throughout. The lateral stripe may be very well marked, but is usually much less so than in the male, and often quite absent. Betiscoides sjostedti sp. n. Male (Type) (Plate XIV).—Much less elongate and more robust than the genotype. Antennae 21-jointed, not biserrate, slightly flattened above, hardly tapering, not finely pointed, in length subequal to the head. Head from above about 12 as long as its breadth at the occiput; fastigium about # the length of an eye; the margins well raised but obtuse, narrowing from the anterior margin of the eye forwards, straight; apex of fastigium acute; median carinula discernible on the fastigium. Head in profile very slightly convex above, face nearly straight. Frontal ridge between the antennae hardly projecting at all, the apex of the head appearing sharply pointed in profile; margins of ridge fairly distinct in the sulcate portion and for a short distance below the ocellus, faintly discernible right to the clypeus. Lateral facial carinae straight, distinct throughout. Hyes ovate, about 13 times as long as their maximal width, the surface somewhat convex; both margins somewhat curved, the upper more so than the lower. Hind margin of pronotum very slightly concave. Sides of pronotum with the lower margin straight, dipping slightly in front; anterior margin sloping forward, somewhat incurved in the lower part; posterior margin practically straight; anterior and posterior lower angles about 90°, rounded. Prosternal tubercle slightly laminate, about twice as long as broad, its margins and angles rounded; the anterior end little broader than the posterior; lower surface level. Hind legs reaching the tip of abdomen or just beyond. Hind femora fairly robust, about 44 times as long as their maximal width; hind tibiae with 10-11 outer and 11-12 inner spines. Hind tarsi about a third the length of the tibiae. Apex of abdomen slightly upturned. Supra-anal plate with the basal part about half the length of the apical part; the longitudinal depression indistinct; the apical part of the plate shaped like an New Acridudae from South Africa. 141 equilateral triangle with the base curved and the sides straight; apical angle rounded; no depression in the apical angle. Subgenital plate (measured from the apex of the supra-anal plate) about equal in length to the anal plate; comparatively bluntly pointed; lower margin slightly convex, upper margin straight, convex at the base; apex rounded. General coloration deep purple-brown, slightly paler above than at the sides. Head, antennae, fore and middle legs the same colour as the rest of the body. Eyes very dark. Hind femora very dark, both inside and outside, except the carinae and knees, which are fairly pale. Hind tibiae very dark on the lower side, spines and spurs black-tipped. Subgenital plate not as dark as the rest of the abdomen. Lateral stripe present but not very clear. Female (Allotype).—Larger than the male. Fastigium about as long as an eye. Sides of pronotum with the lower margin straight; anterior margin sloping forward, only very shghtly incurved; posterior margin slightly incurved, sloping forward in the upper part. Anterior lower angle more than 90°, posterior lower angle 90°, both rounded. Hind tibiae with 9-10 outer spines, 11-12 inner. Genital valves more strongly curved than in the genotype. General coloration similar to, but slightly less dark than, that of the male. Hind femora, both outside and inside, paler than the rest of the body; hind tibiae darker apically than basally. Measurements : Male. Female. mm. mm. Body . : : : 5 22-5 29-8 Pronotum . : : : 2:7 3°8 Hind femur . 5 ; xD anes 9-5 Antenna : : . a B45D)7 4-5 Head . ; : 5 i 4-0 4-5 Subgenital plate . ee Described from 6 males and 3 females from Table Mountain, December 1929-30 (K. Key). There is practically no variation in form in these specimens. All are of a more or less uniformly dark colour, except one male, which is olive-green above. The species occurs on the top of Table Mountain in the same patches of reeds as B. meridionalis, yet no intermediate forms occur. 142 Annals of the South African Museum. Betiscoides parva sp. n. In form resembling B. sjostedtz much more closely than the genotype, but readily separable from both the other species on account of its much smaller size, as well as by other good characters. Male (Type) (Plate XV).—Antennae 21-jointed, practically fili- form, sightly tapering, reaching nearly to the hind margin of the pronotum. Head from above about 13 as long as its breadth at the occiput; fastigium about half the length of an eye; margins less obtuse than in B. sjostedti, somewhat curved; apex not as acute as in B. syostedtv; median carinula discernible on the fastigium. Head in profile slightly convex above, face nearly straight. Margins of fastigium in profile somewhat arched, so that the apex of the head in profile is fairly widely rounded. Frontal ridge between the antennae somewhat projecting; margins of ridge fairly distinct in the sulcate portion and for a short distance below the ocellus, discernible right to the clypeus. Lateral facial carinae straight, distinct throughout. Hyes large, fairly prominent, ovate, about 14 times as long as their maximal width, their surface strongly convex; both margins curved, the upper strongly so. A low, obtuse, pale, callous ridge starts on the upper margin of each eye, follows the margin round for a short distance, and then traverses the side of the occiput in a longitudinal direction, and continues as a shghtly incurved lateral carina along the pronotum, losing itself after a while in the abdomen. Hind margin of pronotum very slightly concave. Sides of pronotum with the lower margin straight; both anterior and posterior margins sloping forward and slightly incurved in the lower part; anterior lower angle slightly more than 90°, rounded; posterior lower angle 90°, rounded. Prosternal tubercle slightly laminate, 2-3 times as long as broad, its margins and angles rounded; somewhat broader and considerably ‘ higher at the anterior end, so that it slopes strongly downwards from the anterior to the posterior end. Hind legs reaching beyond the tip of the abdomen. Hind femora fairly robust, about four times as long as their maximal width; hind tibiae with 9 outer and 11 inner spines. Hind tarsi nearly half the length of the tibiae. Apex of abdomen strongly upturned, making an angle of about 90° with the long axis. Supra-anal plate with the basal part more than half the length of the apical part; the longitudinal depression short New Aeridudae from South Africa. 143 and broad, but deep; the apical part of the plate with the base curved and the sides straight; apical angle rounded; no depression in the © apical angle. Subgenital plate (measured as in the previous species) subequal to the anal plate, comparatively bluntly pointed, with the apex rounded; both upper and lower margins slightly convex. General, coloration plum, with the upper surface of the head, thorax, and abdomen, within the callous carinae, pale and silvery. A darker longitudinal stripe extends from the front of the head back- wards along the central ridge of the body. The lateral stripe back- wards from the base of the eye is silvery, and extends a short way round the lower hind corner of the eye as a somewhat raised callous ridge. The stripe does not reach the hind femur. Antennae plum; eyes pale; hind femora very pale; hind tibiae somewhat pale above, dark beneath. Female (Allotype).—Larger than the male. Antennae shorter than those of the male, reaching only to about the middle of the pronotum. Head in profile nearly straight above; face straight. Hyes relatively smaller than in the male. Lower margin of sides of pronotum incurved in the anterior part; anterior and posterior margins almost straight; anterior angle about 90°, posterior a little more than 90°; angles rounded. Prosternal tubercle shorter and somewhat less sloping than in the male. Hind tibiae with 10 outer and 13 inner spines. Genital valves more strongly curved than in the genotype. Sides of body deep plum, top of body pale yellow-green, with only a trace of the darker central stripe; antennae plum; cheeks yellow- green, frontal ridge plum, rest of face yellow-green tinged with plum; eyes brownish; fore and middle legs green in the femur, becoming plum in the tibia and tarsus. Hind femora yellow-green, paler inside than outside; hind tibiae green at the base, becoming plum at the apex; tarsi plum; spines and spurs black-tipped; genital valves green. Measurements : Male. Female. mm. mm. Body . : : , er 15-0 18:5 Pronotum . 4 ’ t 1:5 2-0 Hind femur. : Eg ih eyahy 6-5 Antenna : : : : 4-3 4-0 Head . : ; : Ue ee 3:0 Subgenital plate . OT hee eS Von. XxKi, PART 3. te: 144 Annals of the South African Museum. Described from 9 males and 8 females collected by the author, the Type and Allotype from the Cape Peninsula, Cape Province, December 1930; 1 male paratype from near Hermanus, Cape Province, January 1931; the rest from Cape Peninsula, 1928-30. The male from Hermanus has a very slightly longer head and subgenital plate, but on the whole the species is fairly constant in its morphological features. In colour both males and females are exceedingly variable, but the females show no tendency to a uniform colour as in B. meridionalis. A fairly homogeneous series of specimens collected by myself from Ceres, Cape Province, resemble B. sjostedti, but are more elongated. Other isolated specimens from various localities appear to differ sightly from the three species described above. The genus has every appearance of being a difficult one, and much more extensive collecting will be necessary before the status of these specimens can be determined. FRONTIFISSIA Nov. gen. This interesting new genus belongs to the group including the three genera Kabulia,* Ramme; LHurynotacris,, Ramme; and Lyrotylus,t Uv. The genotype is smaller than any of the other species of the group, and differs in its vivid coloration, but especially in the structure of the fastigium, which is unlike any of the others, but nearest to Hurynotacris. The pronotum and femora, on the other hand, are least like Hurynotacris. The front of the head is nearest to Kabulia. Frontifissia is the first representative of this peculiar group to be found in Africa. Fairly small, but robust (Plate XVI and text-figs. 1 and 2); body pilose, especially in the male. Antennae fairly short and thick, filiform. Head from above short, little exserted; both occiput and fastigium punctured; fastigium flat, moderately sloping; more so in the male than in the female; distinctly marginated, roughly diamond-shaped, transverse; the lateral angles acute, the anterior margin with a V-shaped median emargination produced by the deep sulcus of the frontal ridge cutting into the apex of the fastigium, which is not roundly continuous with the frontal ridge, but meets it at an obtuse, but quite obvious, * Deutsch. Entomol. Ztschr., pp. 299-302, 1928. + Mitt. Zool. Mus. Berlin, xvi, p. 935, 1931. { Journ. Bombay Nat. Hist. Soc., xxix, No. 3, 1923. New Acridudae from South Africa. 145 angle; temporal foveolae absent, median carina distinctly traceable right along the head, or obsolete. Face in profile somewhat sloping, frontal ridge straight or very slightly incurved, except at the base, where it bends up to meet the fastigium. Frontal ridge broad, more than half the breadth of the lateral frontal areas, strongly concave, with the margins well raised, especially above the ocellus, subparallel below the ocellus, more or less widening at the ocellus or just above it, and gradually converging towards the fastigium; frontal ridge, where it meets the fastigium, about half its breadth below the ocellus. Lateral facial keels well raised, subparallel to the frontal ridge, straight, or more or less curved forwards, reaching the fastigium. Subocular sulcus more or less distinct, straight or some- what curved. Cheeks and front punctured. LEHyes large, prominent from above; separated, in the male, by a distance subequal to the breadth of the frontal ridge at the ocellus, in the female by a greater distance; viewed laterally, more or less oval, the anterior margin tend- ing to be straight, especially in the female; viewed from in front, curved. Pronotum tectiform, widening from anterior to posterior end, slightly in the male, strongly in the female; anterior margin convex, posterior margin widely emarginate; median carina low, thick; lateral carinae obsolescent or indicated on each side by two very slightly raised regions only: an anterior one, not reaching the first transverse sulcus; and a posterior one, sloping backwards and inwards between the second and third sulci. In the female these areas may be broad and callous; first transverse sulcus obsolescent in the female, at any rate on the disk, bending horizontally forward on the lobe; second and third transverse sulci nearly straight on the disk, bending somewhat forwards on the lateral lobes, the third completely intersecting the median carina, the second not. Meta- zona about one-third the length of the prozona (measured along the median carina). Anterior margin of lateral pronotal lobes slightly forward-sloping, slightly incurved; lower margin curved, first sloping downward, then level; hind margin straight as far as the shoulder, sloping at an angle of about 45°, thickened, raised, and callous; anterior lower angle obtuse, rounded; posterior lower angle obtuse, very widely rounded. There is a well-marked vertical sulcus shortly behind the anterior margin of the lateral lobe, which follows the line of the anterior margin, not reaching the disk. Whole of pronotum coarsely punctured, especially on the disk. Part of mesonotum visible behind pronotum; mesonotum and meta- notum similar to the abdominal terga, but more coarsely punctured. 146 Annals of the South African Museum. Prosternal tubercle straight, conical. Mesosternal lobes with the inner margin straight, inner angles obtuse, rounded; mesosternal interspace in the male subquadrate, subequal to the lobes in width; in the female slightly transverse, slightly wider than the lobes. Metasternal lobes with the inner margin straight, inner angles very obtuse, rounded; metasternal interspace trapezoidal, in the male a quarter or less of the width of the lobes, in the female about half the width of the lobes and narrower than the mesosternal interspace. Elytra lateral, lobe-like, covering well-developed tympana. Wings absent. Anterior and middle tibiae with a few spines. Pulvillilarge. Hind femora just reaching, or not quite reaching, the tip of the abdomen in the female, reaching well beyond it in the male; the upper and lower margins slightly, and about equally, curved; upper median carina not very pronounced, serrulate; lateral carina strong, slightly serrulate; externo-median area flat; outer and inner genicular lobes similar, rounded apically. Hind tibiae, both from above and to a lesser degree from the side, curved, with a swelling near the base; tibial spines not constant, about 7-8 outside, 9-11 inside; inner apical spine present, outer absent. Inner spurs much longer than the outer. Abdomen only very lightly punctured, with a broad, low, callous, median keel. Male.—Supra-anal plate slightly transverse; divided, somewhat anterior to the middle, into an apical and a basal part by a well- raised transverse carina; the basal part traversed longitudinally by a deep broad sulcus with its margins well raised, which partly inter- rupts the transverse carina, and is continued as a fine indistinct sulcus in the apical part of the plate; apical part more or less depressed on either side of the median sulcus; apical angles rounded; apical margin with a slight median emargination. Cerci long, laterally compressed, strongly attenuate apically, the apical third, seen from above, bent slightly inwards, reaching the apex of the supra-anal plate or just beyond. Subgenital plate short, conical. Female.—Supra-anal plate about twice as long as broad, the apical part shorter than the basal, and indistinctly separated from it by a fine transverse sulcus. Longitudinal sulcus well marked throughout, though more so in the basal than in the apical part. Genital valves long, strongly curved, not denticulate. Cerci short, very broad at the base, attenuate apically. Genotype : Frontifissia elegans sp. n. - New Acridudae from South Africa. 147 Frontifissia elegans sp. n. Female (Type) (Plate XVI and Text-fig. 1).— When fresh a remarkably striking insect. General coloration deep green, with numerous silvery-white spots and markings. Head green, with several indistinct whitish markings on the face; median carina on the occiput white; eyes and antennae purple- brown; a white stripe sloping down from the eye across the cheek, and one passing straight back from the eye to connect with a short white line on the shoulder of the pronotum. ae Pronotum green, with the median line white, broadening in the neighbourhood of the second transverse sulcus; a large oblique white mark on the shoulder, extending from just behind the second sulcus to just behind the third sulcus, and below the large mark, on the lobe just behind the second sulcus, a smaller white spot; a white line, situated on the lateral lobe at a level slightly above the base of the clypeus, connects the second sulcus to the anterior margin; region of the lobe between the anterior margin and the second sulcus blackish for a short distance above this line; sloping hind margin of pronotal lobes bearing a white stripe, bounded dorsally by a dark purplish stripe of about the same width. Mesonotum without markings, except for the pale median carina. Metanotum with two white patches in the same relative positions as those occurring on the abdominal terga. Elytra green, darker at the base than at the apex; lower margin with a white stripe, upper margin white at the base. Anterior and middle legs green. An irregular white mark on the pleuron just above the insertion of the middle coxa, and another in the corresponding position above the hind coxa. Hind legs pale green throughout (the interno-median area of the femur is discoloured); a row of black dots, which are of the nature of obtuse teeth, on the upper external and internal carinae, but not on the lower internal and external carinae, which are whitish, especially the former. Abdomen dark green, the median keel dirty white; two rows of large white spots on each side of the abdomen, each row consisting of one spot on each tergite; the upper row is in line with the large oblique spot on the shoulder of the pronotum; the lower row lies Hie:-1. 148 Annals of the South African Museum. just above the lower margins of the tergites; on the under side of the abdomen there are also two rows of white spots running along the sides of the sternites.. The spots on the sternites, and those of the lower row on the tergites, are triangular in shape, the apex being directed posteriorly. Sternum pale green, the lateral margins white; metasternal interspace, and a short line forwards from it, and also a line forwards from the mesosternal interspace, white. Genital valves pale green, the tips black. Male (Allotype) (Text-fig. 2)—About half the size of the female and much less striking. General coloration green. Top of head green, except the fastigium, which is dark; frontal ridge and lateral frontal areas red-brown; cheeks green; antennae deep purple-brown; eyes buff; clypeus and mouth parts deep purple-brown. Pro- notum green, with median carina and lower and hind margins of lateral lobes buff, and with a bright maroon mark on the shoulder just inside the hind margin. Abdomen green, with a median buff stripe bounded on either side by a row of small black dots; hind margins of tergites buffish. Forelegs brown- purple with a tinge of green, especially on the tibiae; middle legs green with a tinge of brown, especially on the femora, and the tarsi brown-purple. LHlytra about twice as long as their width at the middle, lower margin nearly straight, upper margin curved; green with a tinge of purple and with the lower margin pale. Hind femora green, except in the in- ternal and external upper areas, which are brownish. Upper external, upper internal, and lower external carinae with a number of black dots, which are of the nature of obtuse teeth, and are most pronounced on the upper external carina. Hind tibiae green at the very base, red-purple in the rest, with the tarsi purplish tinged with green. Sternum deep brown-purple, under side of abdomen pale brown. Que Uh Measurements : Male. Female. mm. mm. Body . ; ap) Os 23-0 Pronotum . t : f 3°2 6-0 EKlytron ue 115) 2°5 Hind femur ; ; : 8-0 12-5 Antenna ; , : : 5.0 6:0 New Acridudae from South Africa. 149 Described from 13 males, 7 mature females, and 5 female nymphs; the Type and Allotype from the Cape Peninsula, Cape Province, 1930 (K. Key); the other male paratypes from the Cape Peninsula, 1 female paratype from near Hermanus, January 1931; the rest from Cape Peninsula, 1930; 2 nymphs from Paarl Road, January 1931; and 3 nymphs from Cape Peninsula, 1928-30. (All collected by the author.) The species is fairly constant in its characters, the chief differences in the males being the degree of invasion by purple-brown, none of the male paratypes having any on the face, though all have it on the under side. The female from near Hermanus is distinctly smaller than the other mature females, but it is only recently mature, the abdomen being very contracted and shrunken. It also differs to some extent in the markings, these being dirty white, and the median stripe on the pronotum being present only near the centre. PYRGOMORPHINAE. ‘Genus PyRGOMORPHELLA Bolivar. (Bol. Soc. Esp. Hist. Nat., iv, p. 457, 1904.) The genus Pyrgomorphella was founded by Bolivar in 1904, to include the species Pyrgomorpha granosa Stal, from Beyrout, and Pyrgomorpha serbica Brunn., from Serbia, as well as three new species from Africa and Madagascar. Since that date no further species have been described. The material before me includes five very distinct new species, all from Africa, bringing the total number of known species up to ten. The subjoined key enables these species to be readily separated. Of the earlier species, I have examined the Type and female paratype of P. serbica and a good series of P. granosa from Jerusalem (Brit. Mus. Coll.), but as regards Bolivar’s three species I have had to rely, in compiling the key, upon his own key and descriptions. P. sphenarioides is included, although as a completely apterous insect it should perhaps be placed in a separate genus. Key to the Species. 1. Apterous : : : : : : : . P. sphenarioides Bol. 1.1 Brachypterous. 2. Elytra hardly exceeding anterior margin of metanotum. 3. Elytra rounded apically. : : . PP. madecassa Bol. 3.3. Elytra acuminate apically. : : P. carinata Bol. 150 Annals of the South African Museum. 2.2. Elytra extending over the whole of the metanotum or beyond. 3. Definite tympanum absent. 4, Elytra dorsal, the inner margins overlapping. 5. Elytra about twice as long as their maximal width, covering about half the first abdominal segment. Hind margin of pronotum not excised : , ; P. rugosa sp. n. 4.4. Elytra dorso-lateral, well separated. 5. Elytra about twice as long as their maximal width, just covering the metanotum, sepa- rated at the base by a distance subequal to their length. Hind margin of pronotum widely angulately excised, not quite reach- ing the metanotum. Wings absent P. capensis sp. n. 5.5. Elytra less than twice as long as their maximal width, covering part of the first abdominal segment, separated at the base by a distance equal to about half their length. Hind margin of pronotum with small, sharp, median emargination—covering half the metanotum. Minute scarlet wings present P. senecionicola sp. n. 3.3. Definite tympanum present. 4, Elytra narrow (about four times their maximal width), diverging outwards and downwards from their point of origin. Scarlet wings absent. Insects not green. 5. Front nearly straight. Anterior lateral carinae of pronotum running back to join the posterior lateral carinae. Hind margin of pronotum widely angulately excised P. granosa Stal. 5.5. Front considerably concave. Anterior and posterior lateral carinae completely separate. Hind margin of pronotum with a very small median emargination . P. variegata sp. n. 4.4, Elytra broad (about twice their maximal width), growing straight backwards. Small scarlet wings present. Insects usually green. 5. Sulcus of frontal ridge widening considerably before meeting fastigium, and continuous on to it. Elytra covering part or whole of second abdominal segment, the lower margin rather sharply upturned about one-third from the apex. Hind margin of pronotum very slightly indented medianly. Pronotal carinae sharp ‘ . LP. serbica Brunn. New Acridudae from South Africa. 151 5.5. Suleus of frontal ridge narrowing uniformly, extremely fine anterior to antennal bases. Elytra not extending on to second abdominal segment, lower margin not sharply upturned. Hind margin of pronotum uniformly convex. Pronotal carinae poorly developed P. rubripennulis sp. n. Pyrgomorphella capensis sp. n. Male (Type) (Plate XVII, E and F).—Body slender, fusiform, finely punctuated, slightly hairy, especially on the under side. Antennae subequal to head and pronotum, 16-jointed, the third joint subequal to the fourth, the last two joints indistinctly separated, the last joint straight on the outside, sloping or incurved on the inside. Section between fifth and sixth joints triangular, the upper inner and outer lower faces being flat and broad, the inner lower narrower and slightly rounded. Head from above conical, about twice as long as its width at the occiput. Occiput with a fine median carina originating at the fastigium and reaching almost to the pronotum. Fastigium as long as broad or very slightly longer, varying somewhat in shape among individuals. The foveolae in contact along a length equal to the greatest width of a foveola, distinctly but not strongly marginated, and but little depressed. Head from the side slightly and evenly convex above, the face strongly sloping, slightly and evenly concave (though the degree of concavity varies slightly in individuals), bend- ing upwards between the antennae to meet the fastigium. Frontal ridge straight, fine, even, sulcate throughout, very nearly reaching the clypeus. Median ocellus extremely minute, evanescent. Lateral facial carinae faint, extending to the base of the antennae. A small tubercle reminiscent of an ocellus situated between the lateral facial carina and the lower angle of the eye. The true lateral ocelli very small, each situated directly in front of the eye, between the upper anterior angle and the base of the antenna, and at the base of the margin of the fastigium. Hyes from above and in front prominent, from the side more or less oval, but the hind end more or less obliquely truncate. A single row of white callosities stretching from the eye to the lower margin of the lateral pronotal lobes. Pronotum on top slightly convex. Anterior margin slightly emarginate, posterior margin strongly angulately emarginate. Hind margin not quite reaching the metanotum. Median carina fine but distinct, intersected by the second and third transverse sulci. The 152 Annals of the South African Museum. first sulcus is represented only by a slight transverse furrow not reaching the lateral carinae. The second sulcus exactly bisects the median carina, and is slightly curved forwards at the point of inter- section. Lateral carinae almost, though not entirely, obsolete, even in the prozona, and obliquely interrupted in the manner character- istic of the genus. They reach their greatest development in the metazona, where they are indicated by a line of two or more very small white callous tubercles, and a purplish mark. Lateral lobes with the inferior margin slightly sinuate, raised, and callous; anterior margin sloping forward, a very indistinct sulcus just behind it; hind angle about 90°, hind margin incurved; a broad callous rugose stripe immediately above the inferior margin, and continuous with the line of tubercles from the eye. The second sulcus distinct right to the callous stripe, the third sulcus not, though in some individuals it is. Surface of lateral lobes flat. Mesonotum just visible behind the pronotum. Metanotum resembling an abdominal tergum. Mesosternal lobes subquadrate. Mesosternal interspace sub- quadrate, slightly wider than the lobes. Metasternal interspace not as wide as the mesosternal interspace, small, slightly transverse, oval, depressed. Prosternal tubercle blunt, broad. Elytra about twice as long as their greatest width, the inner margins subparallel, separated at the base by a distance subequal to their length, with a few longitudinal rows of fine pits, just covering the metanotum. Wings absent. | Abdomen with a low dorso-median carina, obsolescent in places. Tympanum absent. Fore and middle tibiae with outer and inner spines. Hind femora with the upper outer carina and the oblique carinae of the outer area not very strongly developed. Hind tibiae with 8-9 outer and 10 inner spines (11 in two male paratypes). A small apical outer and a large apical inner spine present. General colour yellow-green. Antennae (except the two basal joints, which are green) and apex of fastigium purple-brown. LHyes pale brown. Row of callosities from eye, white. Pro- and meso- notum with a faint purple-brown line along the median carina, and continued along the abdomen, widening in the first abdominal tergum. Lower margin of pronotal lobes with a white band sur- mounted by a fine dark line. Hind portions of lateral carina with white callosities bordered below by a fine dark line. Under side pale. Fore and middle legs brownish. Sides of abdomen with a black New Acridudae from South Africa. 153 spot on the anterior margin of each segment. Hind femora green, the knees pink at the apex. Hind tibia blue-green, the apex and the feet purplish. Female (Allotype).—Larger than the male, body more fusiform. Antennae and eyes relatively smaller. Lateral facial carina further from the eye than in the male owing to the smaller size of the eye. Line of callosities from the eye, and also the callous border to the lateral pronotal lobes, less well developed than in the male. Median ‘ pronotal carina with a fine sulcus. LElytra separated at the base by a distance greater than the length of an elytron. Mesosternal interspace rectangular, transverse, about twice as wide as long. Metasternal interspace elongated transversely, about four times as wide as long, equal in width to the mesosternal interspace. Genital valves slender, gently curved, the upper pair roughly serrate in the basal half. Hind tibia with 10 outer and 10-11 inner spines. (In other paratypes there are 9 outer and 10-11 inner spines.) Both outer and inner apical spines present, the former small. General coloration brown. Antennae purple-brown. Head dark brown. Line of callosities below the eye and on the pronotal lobes less vividly white than in the male. Abdomen brown with a dark brown stripe on the sides, especially in front. Under side pale. Fore and middle legs dark brown. Hind femora more or less mottled with brown and dark brown. Hind tibiae almost black apically. Measurements : Male. Female. mm. mm, Body ; : Ed la (as: 15-7 Pronotum . : ’ 2-0 3°0 Hind femur . i ; 6-5 8-0 Antenna é : : 4-75 5-0 Head Mae : : 2:5 3°0 Described from 4 males and 16 females from the Cape Flats, Cape Peninsula (Key, December 1930). The colour pattern exhibited by the Type apparently occurs only among the males, though other patterns are equally common. Both among males and females, however, coloration is very variable, ranging from pale grey, through all shades of grey and brown, to grey-green and blue-green. All the specimens were collected from a mass of low-growing silvery- leaved herbs of the family Compositae, which presumably forms their food-supply. 154 Annals of the South African Museum. Pyrgomorphella senecionicola sp. n. Larger than the previous species, and differing from it mainly in the pronotum from above, especially the posterior margin, the elytra, and the typical female coloration. Male (Type) (Plate XVII, D an G).—Body fusiform, less slender and slightly more rugulose than in the previous species, slightly hairy. Antennae subequal to head and pronotum, 16-jointed, the third joint much longer than the fourth, but showing traces of division (more pronounced in a few of the paratypes), the last two joints indistinctly separated, the last joint straight on the outside, sloping on the inside, not so pointed as in P. capensis. Section between fourth and fifth joints triangular, the upper inner and lower outer faces being flat and broad, the lower inner narrower and slightly rounded. Head from above conical, very nearly twice as long as its width at the occiput. Occiput with a fine median carina originating at the fastigium and reaching the pronotum. Occiput somewhat rugose, the rugosities tending to become transverse on the vertex. Fastigium as long as broad, varying somewhat in shape among individuals. The foveolae in contact along a length equal to the greatest width of a foveola, more or less distinctly, but not strongly marginated, and but little depressed. Head from the side slightly and evenly convex above, the face strongly sloping, slightly and evenly concave, bending slightly upwards to meet the fastigium. Frontal ridge fine, straight, even, sulcate throughout, very nearly reaching the clypeus. Median ocellus small, but not as degenerate asin P. capensis. Lateral facial carinae faint, extending to the base of the antennae. A small tubercle reminiscent of an ocellus situated between the lateral facial carina and the lower angle of the eye. The true lateral ocelli each very small, situated directly in front of the eye, between the upper anterior angle and the base of the antenna, and at the base of the margin of the fastigium. Eyes from above and in front prominent, from the side more or less oval, but the hind end more or less obliquely truncate. A single row of white callosities stretching from the eye to the lower margin of the lateral pronotal lobes. Pronotum on top slightly convex. Anterior margin slightly emarginate, posterior margin in general contour slightly convex, but with a small, sharp, median emargination. Hind margin covering half the metanotum. Median carina very faint, but discernible, New Acridudae from South Africa. 155 intersected by the second and third transverse sulci. The first sulcus is represented only by a slight transverse furrow not reaching the lateral carinae. The second transverse sulcus is situated slightly but definitely in front of the mid-point of the median carinae, and bends slightly forwards at the point of intersection. Lateral carinae almost, though not entirely, absent in the prozona, and obliquely interrupted in the manner characteristic of the genus. In the metazona they are indicated by a line of two or more white callous tubercles and a purplish mark. The two lines of tubercles diverge slightly forwards. Lateral lobes with the surface very slightly impressed; the inferior margin slightly sinuate, raised, and callous; anterior margin sloping forward, a very indistinct sulcus just behind it, hind angle about 90°, hind margin incurved. A callous rugose stripe, narrower than in the previous species, immediately above the inferior margin and continuous with the line of tubercles from the eye. Both transverse sulci distinct down to this stripe. Mesonotum not visible behind pronotum; hind half of metanotum visible only. Mesosternal lobes with the inner hind angle rounded, somewhat less than 90°, the interspace slightly transverse, trapezoidal, about as wide as the lobes. Metasternal interspace oval, transverse, about twice as wide as long, nearly as wide as the mesosternal interspace. Elytra larger than in P. capensis, less than twice as long as their greatest width, the inner margins converging towards the base, separated at the base by a distance about half their length, with many longitudinal rows of pits, covering part of the first abdominal segment. Minute scarlet wings, incapable of being expanded, present beneath the elytral pads. Abdomen with a low dorso-median carina, callous in some parts, obsolescent in others. Tympanum absent. Fore and middle tibiae with inner and outer spines. Hind femora with the oblique carinae on the externo-median area not strongly developed. Hind tibiae with 8-9 outer and 11 inner spines (8-10 outer and 9-11 inner in the male paratypes). A small apical outer and a large apical inner spine present. General coloration green and brown. Antennae purple-brown except the two basal joints, which are green. Head brownish on the vertex, green elsewhere. Eyes brown. Lateral row of tubercles from the eye yellowish white. Pronotum pale brown on top, brownish green on the lateral lobes. Pleura greenish brown. Dorsal surface of abdomen brown, the apex and sides greenish brown. Fore and 156 Annals of the South African Museum. middle legs brown. Hind femur green, the apex of the knee brown. Hind tibia blue-green for about the basal two-thirds, the apical third and the tarsus pink. Under side pale brown. Female (Allotype).—Larger than the male, the body more fusiform, more rugose. Antennae and eyes relatively smaller than in the male. Antennae 15-jointed (16-jointed in one paratype). Head from above relatively shorter, the median carina less distinct than in the male and much less distinct than in the female P. capensis. Median ocellus obsolete (present in the other paratypes). Lateral facial carina further from the eye than in the male. The row of callous tubercles below the eye narrower, less distinct, the tubercles reduced in number and in size, hardly larger than the other tubercles on the head. The stripe on the pronotal lobe similarly reduced. Median carina of pronotum obsolete except in the hindmost part of the meta- zona, where it is just discernible. Lateral pronotal carina very blunt—really only an angular bend in the plane of the pronotum— and with not more than one extremely small white tubercle (in some paratypes tubercles are quite absent). Surface of lateral lobes slightly more impressed. Mesosternal interspace rectangular, about twice as wide as long, and twice as wide as the lobes. Metasternal interspace strongly transverse, about five times as wide as long, and equal in width to the mesosternal interspace. Hlytra separated at the base by a distance slightly less than the length of an elytron. Genital valves more robust than in P. capensis, strongly curved, the upper pair roughly serrate in the basal half. Hind tibia with 10 outer and 11 inner spines (8-9 outer and 10-11 inner among other para- types). Both outer and inner apical spines present. Antennae pale pinkish. Eyes pale brown. Callous stripe below eye and on inferior margin of pronotal lobe indistinct, yellowish white. The rest uniformly green. Measurements : Male. Female. mm. mms. Body . ; : : 1 13-0 19-25 Pronotum . ; ; : 2°8 4-2 Hind femur . : ; ee 9-2 Antenna : . E ' 5-2 5-5 Head . § i ] : 2:7 3°25 Described from 7 males and 5 females, all from near Hermanus, in the Caledon Division, Cape Province (Key, January 1931). The males all possess essentially the same coloration, differing only in the extent New Acridudae from South Africa. 157 to which brown invades green, while the pink apex to the hind tibia may be absent. All the females possess the same coloration except one, which is uniformly buff. No approach is made to the degree of variability exhibited in both males and females of P. capensis. Pyrgomorphella rubripennulis sp. n. Larger than either of the last two species, and differing from them mainly in the pronotum and elytra, and in the possession of a tympanum. Male (Type) (Plate XVII, C and H).—Body fusiform, slightly hairy, more rugulose than in the previous species. Antennae subequal to head and pronotum together. Third and fourth joints imperfectly separated (quite separate in some paratypes). Counting these joints as separate there are 17 joints in all. The last joint with the outer margin straight, the inner sloping. Section between fifth and sixth joints more or less triangular, the one long side (representing the upper inner face) straight, the other long side (representing the lower outer face) curved, and the short side some- what curved. Head from above conical, less than twice as long as its width at the occiput. Occiput with a fine median carina originating at the fastigium and reaching the pronotum. Occiput somewhat rugose, more so than in the preceding species, the rugosities tending to become transverse on the vertex. Fastigium as long as broad, the apex more rounded than in the previous species (though in one paratype it is indistinguishable from P. senecionicola). The foveolae in con- tact along a length equal to the greatest width of a foveola, clearly marginated, little depressed. Head from the side very slightly convex—almost straight—on top, the face strongly sloping, slightly and evenly concave, bending slightly upwards to meet the fastigium. Frontal ridge straight, fine, even, sulcate throughout, very nearly reaching the clypeus. Median ocellus small. Lateral facial carinae clearly marked, extending to the base of the antennae. A small tubercle reminiscent of an ocellus situated between the lateral facial carina and the lower angle of the eye. The lateral ocelli small, situated directly in front of the eye between the upper anterior angle and the base of the antenna, and at the base of the margin of the fastigium. Hyes from above and in front prominent, from the side more or less oval, but the hind margin more or less obliquely truncate. A band of callous tubercles which may contain more than one row stretching from the eye to the lower margin of the pronotal lobes. 158 Annals of the South African Museum. Pronotum from above slightly convex. Anterior margin nearly straight, posterior margin very obtusely angulate, the apex rounded, without any emargination (one male paratype has a slight median indentation). Hind margin covering about half the metanotum. Median carina very faint, in some places obsolescent, in others very finely sulcate, intersected by the second and third transverse sulci. The first sulcus is represented only by a slight transverse furrow not reaching the lateral carinae. The second transverse sulcus is situated in front of the middle, and has a forward bend in the middle. Lateral carinae very blunt and poorly developed, being little more than angular bends in the surface plane of the pronotum; inter- rupted obliquely in the manner characteristic of the genus; the hind portion with only one small white tubercle (two in some paratypes). Lateral lobes with the surface very slightly impressed, the inferior margin sinuate, raised, and callous, anterior margin sloping forward, a very indistinct sulcus just behind it, hind angle about 90°, hind margin incurved. A callous band of tubercles immediately above the inferior margin and continuous with the band of tubercles from the eye. Transverse sulci extending down to this band. Mesonotum not visible behind pronotum. Hind part of metanotum only visible. Mesosternal lobes with the inner angle rounded, somewhat less than 90°, the interspace trapezoidal, not transverse, about as wide as the lobes. Metasternal interspace oval, transverse, about twice as wide as long, nearly as wide as the mesosternal interspace. Elytra larger than in the previous species, about twice as long as their greatest width, the inner margins almost in contact at the base; with rudimentary longitudinal veins separated by rows of pits, extending just beyond the hind margin of the first abdominal segment (in one paratype the elytra just fail to reach the hind margin). Elytra capable of being raised when the insect is disturbed to exhibit the minute scarlet wings, just capable of being expanded. Integu- ment under and immediately around the wings (but covered by the elytra) scarlet. Abdomen with a low dorso-median carina. The infero-apical part of the elytron covers a functional quadrant-shaped tympanum. Fore and middle tibiae with outer and inner spines. Hind femora with the oblique carinae on the externo-median area not strongly developed. One hind tibia with 10 spines on the outside and 10 on the inside, and both apical spines present, the other with 8 outer spines, the apical one represented only by a small tubercle, and 11 New Acridudae from South Africa. 159 inner spines (in the paratypes there are 9-10 outer spines and 10-11 inner; in one paratype there are 8 outer spines on both tibiae, the apical one in each case being only a small tubercle). When present, the outer apical spine is very small. General coloration green invaded by purple-brown. Antennae dark purple-brown, except the basal two joints, which are some- what greenish. Head on top mainly purple-brown, especially on the fastigium. Rest of head buff-green. Eyes pale brown. Pronotum purple-brown in the prozona, brownish green in the metazona. Lateral lobes of pronotum brownish green. The stripe below the eye and on the lower margin of the pronotum yellow. Pleura, abdomen, and fore and middle legs brownish green. LHlytra purple- brown on top, greenish near the outer margin. Hind legs dull green, apex of knees pinkish. Under side buff-green. Female (Allotype). — Larger than the male, the body more fusiform. Antennae relatively slightly shorter than in the male, more ensi- form at the base, and in that region also slightly serrate. Third and fourth joints imperfectly separated. Counting them as separate there are 18 joints in all, the last two imperfectly separated. (In some paratypes the third and fourth joints are quite separate, in others quite united, so that the antenna has only 17 joints.) Head from above relatively shorter than the male. Lateral facial carinae further separated from the eyes than in the male on account of the relatively smaller size of the eyes. Row of tubercles below the eye less well developed than in the male; the band of the inferior margin of the pronotal lobes nothing more than a yellow stripe, because the tubercles are no larger than those on the rest of the pronotum. Pronotum from above more convex than in the male. Hind margins very slightly incurved before they meet at the rounded apex. Lateral pronotal carina obsolescent in the prozona, where it is indicated by a row of extremely small callous tubercles (better developed in some female paratypes), similar to the male in the metazona. Surface of the lateral lobes slightly more impressed than in the male. Meso- sternal lobes with the inner hind angle rounded, 90° (less than 90° in some paratypes), the interspace rectangular (trapezoidal in some paratypes), transverse, nearly twice as wide as the lobes and as its own length. Metasternal interspace transverse, about 6-8 times as wide as long, equal in width to the mesosternal interspace. Elytra more widely separated at the base than in the male, but not separated by a distance greater than half the width of an elytron at the base. VOL, XXXII, PART 3. 12 160 Annals of the South African Museum. Hind tibiae with 10 outer and 11 inner spines, both apical spines present, but the outer one very small. (In other female paratypes there are 8-10 outer and 10-11 inner spines.) Genital valves more robust than in the two previous species, the upper pair more abruptly curved at the apex. General coloration uniformly green. Antennae, eyes, and meta- notum grey. Apex of fastigium and fore and middle legs with a trace of grey. Sides of abdomen with the lower margins of the tergites buff. Subocular and lateral pronotal stripes yellow. Infero-external carina of hind femur yellow. Measurements: Male. Female. mm. mm. Body , ‘ : a5 95) 21-5 Pronotum : : , 4-0 5-25 Hind femur ; 4 : 9-25 11-0 Antenna es : : ; 6:25 6:8 Head , : A i 2-8. 4-0 Described from 7 males and 6 females, all from Grahamstown, Eastern Cape Province (Key, May 1931). Some of the specimens are somewhat discoloured. The males vary in the extent to which purple-brown has invaded green. One male paratype is of exactly the same colour as the allotype, while one is uniformly brown. The females vary to a less extent. Both sexes vary considerably in size, shape of fastigium, and degree of concavity of face. The external lower carina of the hind femur may be either yellow (as described for the female allotype) or normal in both sexes. The three species capensis, senecionicola, and rubripennulis form (in the order mentioned) an interesting series which shows a steady gradation from one species to the next in several characters. The direction of the gradation corresponds with the distribution of the species geographically. Thus, starting with capensis at Cape Town, and passing east along the coast, we come to senecionicola at Her- manus, about 40 miles away as the crow flies, and then to rubr- pennulis at Grahamstown, a further 430 miles east. The following characters show a regular increase from capensis to rubripennulis: size, relative length of elytra, degree of development of wings and tympanum; and the degree of emargination of the hind margin of the pronotum shows a regular decrease. New Acridudae from South Africa. 161 Pyrgomorphella variegata sp. n. Male (Type) (Plate XVII, A and L).—Body fusiform, rugulose, very slightly hairy on the legs and under side. | Antennae in length subequal to head and pronotum together, 17-jointed, the third joint not perfectly separated from the fourth, the last joint straight on the outside, sloping on the inside; somewhat flattened and serrate, especially near the base, triangular in section, the long sides of the triangle representing the upper inner and lower outer faces. Head from above conical, about twice as long as its width at the occiput. Occiput with a fine median carina, originating at the fas- tigium and reaching the pronotum. Fastigium slightly longer than wide, slightly convex, the apex widely rounded. Foveolae in contact along a length equal to the greatest width of a foveola, not impressed. The carinae which, at the base, form the inner margin of the foveolae are not in contact until the very apex of the fastigium, owing to the fact that they invade the foveolae and no longer form the true inner margins of the foveolae. Head from the side evenly convex above, the face strongly sloping, considerably concave. Frontal ridge narrow, very indistinct in the lower part, not quite reaching the clypeus, sulcate throughout, narrowing gradually towards the fas- tigiuum. Median ocellus minute. Lateral ocelli small, situated just in front of the eye, and between the eye and the base of the antenna. Lateral facial carina faint in the middle, more distinct above and below. A small low tubercle reminiscent of an ocellus situated between the lateral carina and the lower angle of the eye. Hyes from above and in front prominent, from the side more or less oval, but the hind end more or less obliquely truncate. A fairly broad band of more than one row of raised callosities stretching from below the eye to the base of the lateral pronotal lobe. Surface of the gena above and below the subocular stripe smooth and shiny. Pronotum on top more or less flat. Anterior margin nearly straight, posterior margin shghtly convex, with a shght median emargination. Hind margin not quite reaching the metanotum. Median carina low and fine but distinct, intersected by three transverse sulci. The first is poorly developed and only just reaches the lateral pronotal carinae; the second is well developed and cuts deeply into the lateral carinae; the third is also well developed, but cuts the lateral carinae less deeply. The second sulcus cuts the median carina slightly in front of the mid-point, and is slightly bent forward at the point of 162 Annals of the South African Museum. intersection. The lateral carinae are quite well developed, and obliquely interrupted. The anterior portions of the carinae do not extend further back than the second transverse sulcus, and diverge strongly forwards. The posterior portions diverge forwards at an angle subequal to that of the anterior portions, and extend obliquely downwards and forwards into the lateral lobes as far as the anterior margin, though they are less well developed at the very front. At the second sulcus the anterior portion of the lateral carina and the posterior portion are separated, along the line of the sulcus, by a distance subequal to the distance between the second and third sulci. In front of the second sulcus, running right between the anterior and produced posterior portions of the lateral carina, is another short low carina formed by a row of tubercles; this carina bends upwards in front to meet the anterior portion of the lateral carina. Lateral pronotal lobes with the lower margin somewhat sinuate, raised, and callous; the anterior margin sloping forward, the posterior margin incurved; hind angle about 90°, rounded. Surface of the lateral lobes impressed. A broad, raised, yellowish, callous stripe traversed by irregular sulci extends just above the lower margin and is continuous with the subocular stripe; it is less well developed in the region behind the third sulcus. Shortly behind the anterior margin of the lateral lobe is a transverse sulcus. The second and third sulci reach the callous stripe. The surface of the lobe in front of the third sulcus and above the callous stripe is smooth and shiny, that behind the third sulcus lightly rugose like the disk. | Mesonotum just visible behind the pronotum. Metanotum smoother than the abdominal segments. Mesosternal lobes and interspace subquadrate and subequal. Inner hind angles of meso- sternal lobes not rounded. Metasternal interspace more or less oval, but the anterior margin straight; about twice as wide as long, not quite as wide as the mesosternal interspace. Elytra 3-4 times as long as their maximal width, diverging outwards and downwards from their point of origin, the outer margin nearly straight, the inner slightly curved, the apex rounded; just failing to reach the hind margin of the first abdominal segment, separated at the base by a distance subequal to their maximal width. Abdomen without a dorso-median carina. A small. tympanum present just underneath the apex of the elytron. Fore and middle tibiae with outer and inner spines. One hind leg missing, the other detached. Hind femur with the upper outer carina and the oblique carinae on the externo-median area well New Acridudae from South Africa. 163 developed. Hind tibia with 8 outer and 11 inner spines. An apical inner, but no apical outer spine present. : General coloration brown. The Type is clearly a good deal faded. Antennae grey, somewhat paler near the outer margin, and the distal margin of each joint pale, so that the antenna appears striated. Basal three or four joints brown. Head brown, somewhat darker along the hind part of the median carina, and above the subocular stripe, which is yellowish. Eyes reddish brown. Pronotum discoloured on the lobes and the anterior part of the disk; brown, darker on the shiny part above the stripe on the inferior margin of the lobes. The stripe discoloured, probably yellowish in life. The pleura are very characteristically coloured. Beginning at the suture separating the anterior half of the mesopleuron from the mesosternum, and working backwards, we have first a dark brown region; at about the lower hind angle of the pronotal lobe there arises a pale band running parallel to the pleural sutures, to the coxa; behind this comes a dark band, and behind that, just in front of the mesopleural suture, another pale band; the hind half of the meso- pleuron—that is, behind the afore-mentioned suture—is mainly dark; then comes the suture separating the meso- and metapleura, and behind that the front of the metapleuron, which is pale, then the metapleural suture, followed by the hind half of the metapleuron, which is dark save for the infero-posterior process, which is pale. Abdomen brown with a narrow dorso-median pale stripe. The sides with a pale stripe along the lower margins of the terga, sur- mounted on each segment by a small dark mark. LElytra and legs brown. Female (Allotype).— Larger than the male, the body more fusiform, more rugose. Antennae with the joints relatively longer than in the male, especially the last one, less flattened and serrate, the one 15- the other 16- jointed. Head from above less than twice as long as its width at the occiput. Median carina very faint. Foveolae slightly impressed. A slight transverse depression on the vertex shortly behind the apex of the eyes. Head in profile nearly straight above. Median ocellus not as small, nor the frontal ridge so indistinct below the ocellus, as in the male. LHyes relatively smaller than in the male. Pronotum on top with a few slight depressions, especially one on either side of the metazona. Hind margin covering part of the metanotum. Trans- verse sulci less well developed than in the male. Lateral pronotal 164 Annals of the South African Museum. carinae subparallel. Mesonotum not visible behind the pronotum. Mesosternal lobes and interspace trapezoidal, the inner hind angles of the lobes considerably less than 90°, the interspace not quite twice as wide as long. Metasternal interspace as wide as the meso- sternal, about four times as wide as long. LElytra not quite as attenu- ate as in the male, about three times as long as their maximal width, just surpassing the hind margin of the first abdominal segment, separated at the base by a distance nearly twice their maximal width. Genital valves strongly curved, black-tipped, the upper pair coarsely serrate above, slightly hairy below. Coloration essentially similar to the male, but unfaded. Antennae purple-brown, striated as in the male. Head pale in front. Sub- ocular stripe yellowish, bordered above and below by a very dark, nearly black band. Head and pronotum on top purple-brown, but the heads of the numerous small rugosities yellowish, so that a paler brown effect is produced. The dorsal surface of the abdomen is darker, because there are fewer rugosities. A very dark dorso- median stripe on the head. Eyes purple-brown. Shiny part of lateral pronotal lobe deep purple-brown. Pleura as in male, except that the lower anterior part of the metapleuron is dark purple- brown. Under side mottled purple-brown and yellowish. Fore and middle legs purple-brown. Hind legs missing. Measurements : Male. Female. mm. mm. Body : ; : pila) 28:0 Pronotum . 5 5 : 4-0 5-5 Hind femur 5 , ; 9-5 ae Antenna . } : 5 Bs 8-25 Head , : é “i eed 3°8 Described from one male (Type) from Narosura, Masai Reserve, Kenya (Capt. A. O. Luckman, February 1914, 6500 feet), and one female from El donyo eb Viru, Kenya (C. 8. Betton, 1902, May—July). Though the Type and allotype differ in several points, especially the top of the head, these are almost certainly sexual differences. In any case, the shape of the head varies considerably among indi- viduals in the other species. Pyrgomorphella rugosa sp. n. Male (Type) (Plate XVII, B and K).—Body slightly hairy. Antennae slightly shorter than head and pronotum together, 15-jointed, the third joint imperfectly separated from the fourth New Acridudae from South Africa. 165 and itself showing a very faint sign of a dividing line, the last joint straight on the outside, curved on the inside; slightly wider and flattened at the base, evenly tapering, triangular in section, the long sides of the triangle representing the upper inner and lower outer faces. Head from above slightly conical, about twice as long as its width at the occiput. Occiput with a fine, well-raised, median carina originating a short distance behind the fastigium, and obsolete shortly behind the eyes. Fastigium about as long as wide, well marginated, the apex widely rounded; foveolae in contact along a length equal to the greatest width of a foveola, slightly impressed. Vertex with a few irregular rugosities. Head in profile strongly convex above, the bend being mainly between the eyes; the face strongly sloping, irregularly concave, suddenly impressed a short distance below the ocellus. Frontal ridge wider than in all the previous species, distinctly marginated throughout, not quite reaching the clypeus, widened at the ocellus, suddenly depressed shortly below the ocellus, strongly compressed between and above the antennae, but sulcate throughout. Median and lateral ocelli well developed. Lateral ocelli situated level with the front of the eye, and between the eye and the base of the antenna. Lateral facial carinae distinct throughout, well developed in the region of the eyes, less so lower down. A small tubercle reminiscent of an ocellus situated between the lateral carina and the lower angle of the eye. Face and gena with a few small rugosities. Eyes from above and in front very prominent, from the side nearly circular. No subocular stripe. Pronotum on top slightly convex, coarsely rugose. Anterior margin slightly and widely emarginate. Posterior margin slightly convex, not emarginate, not reaching the hind margin of the meso- notum. Median carina well raised, but, like all the carinae, widely interrupted and broken up by the transverse sulci, including the first. The first sulcus is very faint and stretches only as far as the front portions of the lateral carinae. The second and third sulci are well developed, widely interrupting the median and lateral carinae. Both bend forward somewhat at their points of inter- section with the median carina. The second sulcus intersects the median carina at its mid-point or very slightly in front. The lateral carinae are much broken up, both by the transverse sulci and apart from them, and their course is further obscured by the presence of other rugosities and short carinae on the pronotum and its lobes. 166 Annals of the South African Museum. Kach lateral carina is obliquely divided into an interior and posterior portion. The anterior portions can be traced back as far as the second transverse sulcus, and diverge forwards. The posterior portions are most clearly indicated in the metazona; they are produced forwards and downwards on to the lateral lobes, becoming obsolescent towards the anterior margin. Lateral pronotal lobes with the surface slightly impressed, the lower margin strongly sinuate, raised, and callous; the anterior margin sloping forward, the posterior margin incurved; hind angle obliquely truncate. Shortly behind the anterior margin is an indistinct short sulcus. The second transverse sulcus extends on to the lateral lobe and nearly reaches the lower margin; the third sulcus does not extend so far. Surface of lobe strongly punctured in the metazona. Mesosternal interspace trapezoidal, wider than the lobes. Inner hind angles of mesosternal lobes rounded, less than 90°. Metasternal interspace somewhat depressed, oval, slightly wider than long, narrower than the mesosternal interspace. EKlytra more or less oval, about twice as long as their maximal width, the outer margin straight for the basal two-thirds, bending inwards for the apical third; inner margins straight except at the very base, overlapping right to the apex, but leaving a small triangular portion of the mesonotum showing between them at the very base. EHlytra covering the whole of the metanotum, no part of which is visible, and half the first abdominal segment. Abdomen with a low dorso-median carina. Tympanum in- dicated only by a small depression behind the first abdominal spiracle. Anterior and middle femora well developed, the tibiae spined. Hind legs missing. General coloration grey-brown, the apices of the larger rugosities dark. Antennae grey, in the basal part paler at the margins than in the centre; there are also paler regions at about joints 9-11 and joint 14. Foveolae of fastigium also greyish. Eyes reddish brown with dark streaks. Face pale brown. Frontal ridge, below the ocellus, at the point where it is suddenly depressed, with the margins black; margins below that point also black in parts. Lower margin of pronotal lobes with a broad pale stripe interrupted at the second sulcus. Pronotal lobes darkish except in the metazona. Abdomen dark in the region of the first spiracle, the lower margin of each tergite with a slightly oblique black mark. Under side buff. New Acridudae from South Africa. 167 Measurements: Male mm, Body : ! : 2) TGA Pronotum . : 2 os GeO) Hind femur ; , ase Antenna . 2 : ; 5-0 Head ; : : ; 3:0 Described from one male only in the South African Museum’s collection, from Dunbrody, Uitenhage Division, March 1912. EXPLANATION OF PLATES XIII-XVII. XIII. Betiscoides meridionalis Sjést. XIV. Betiscoides sjostedts sp. n. XV. Betiscoides parva sp. n. XVI. Frontifissia elegans sp. n. XVII. Pyrgomorphella capensis sp. n., figs. E, F. $5 senecionicola sp. n., figs. D, G. by rubripennulis sp. n, figs. C, H. “A variegata sp. n., figs. A, L. oe rugosa sp. n., figs. B, K. ee iste lay IN aon io) ee at fetcyee’ ‘agg Aah aes Oth. is iis - ? 2 a ff ee ‘ (i > 4 Plate XIII. Amn. o. Air. Mus., Vol. XX XII. ‘AVAIIGIXYOV NVOIHAVY HLOOS Neill & Co., Ltd. K. HH. L. Key. Ann. S. Afr. Mus., Vol. X X XII. K. H. L. Key. Plate XIV. Neill & Co., Lid. SOUTH AFRICAN ACRIDIIDAE. Plate XV. Ann. S. Afr. Mus., Vol. XX XII. Neill & Co., Lid. K. H. L. Key. Plate XVI. Ann. 8. Afr. Mus., Vol. XX XII. teleeewemee SrOOooconnrrrrrrrrr, 0., Lid. Neill & C TO Key: oo hrc (i mae ach oY eT rome Be Plate XVII. = Vols XX XI. Afr. Mus. Ann. S. Neill & Co., Lid. K. H. L. Key. a” eau ( 169 ) 12. Notes on Dragon-flies (Odonata) of the S.W. Cape, with Descriptions’ of the Nymphs, and of New Species.—By K. H. Barnarp, D.Sc., Assistant. Director. , (With 32 Text-figures.) _ Tuts is the tenth report on the Fauna of the Mountain Ranges of the Cape Province, for the investigation of which I have received grants from the Royal Society of South Africa and the Research Grant Board.* My thanks and acknowledgments are herewith tendered to these bodies. Both from a general faunistic point of view, and also in connection with the food of trout and other fishes, dragon-flies are of considerable importance. During the course of my researches on the Mountain Fauna much material and many observations have been collected. Additional occurrences and localities of several species since the publication of Ris’ monograph of the South African Odonata (1921, Ann. 8. Afr. Mus., vol. xviii) have accumulated and are worthy of record. The main purpose of this paper is to describe the nymphs (larvae, or nalads) of the Cape species, only one of which was available for ‘inclusion in Ris’ work. Only those species occurring in the 8.W. Cape, roughly within a radius of 200 miles from Cape Town, have been included, as I have had no opportunities of investigating the life-histories of the more tropical forms. The fauna-list for this area includes 39 species and 2 varieties, of which 4 species and: 2 varieties are described as new. The nymphs of 24 of these have been discovered; in terms of genera, the nymphs of only 4 genera out of 22 remain to be discovered (viz. Lestes, Ceriagrion, Palpopleura, Sympetrum). * Previous reports: 1. “Freshwater Crustacea,” Trans. Roy. Soc. 8. Afr., vol. xiv, 1927. 2. “Colophon (Coleoptera),”’ sbid., vol. xviii, 1929. 3. “‘ Alder-flies,”’ ibid., vol. xix, 1931. 4. “May-flies,” cbid., vol. xx, 1932. 5. “Terrestrial Isopoda (Woodlice),”” Ann. S. Afr. Mus., vol. xxx, 1932. 6. “Further New Species of Colophon,” Stylops, vol. i, pt. 8, 1932. 7. ““A New Corduline Dragonfly,” zbid., vol. ii, pt 7, 1933. 8. ‘‘Caddis-flies,” Trans. Roy. Soc. S. Afr., vol. xxi, 1934. 9. “‘Stone-flies,’? Ann. S. Afr. Mus., vol. xxx, 1934. SS ee 170 Annals of the South African Museum. In 1929 Brain (Insect Pests and their Control in South Africa, pp. 159-160) gave a short key to 15 genera of dragon-flies found in the winter-rainfall area. Following the localities given in Ris, he did not include any Gomphines or Cordulines, or the genera Chloro- cypha (olim Lrbellago) and Elattoneura (olim Disparoneura). As regards the imagos, particular attention has been paid to the genitalia of both sexes, and for the sake of comparison several figures of species occurring outside the 8.W. Cape area are given. This aspect of dragon-fly taxonomy is one to which far too little attention has been paid in the past. In so many instances where the “g genitalia” are figured, it is found that actually only the secondary or accessory structures are figured; and discussions on synonymy take no account of the penis, which, as the sequel will show, often offers valuable specific characters. Once again I have to record the great assistance rendered by Mr. H. G. Wood, who has discovered several new localities for species not hitherto believed to occur so far to the south-west, and who has very generously provided motor transport on our joint expeditions. Mr. A. C. Harrison, Mr. F. G. Chaplin (of the Jonkershoek Trout Hatchery), and Mr. J. C. Dendy have taken an interest in the work from the angler’s point of view, and have been of very material assistance. Mr. Harrison was the first to discover a Corduline in South Africa (Stylops, vol. ii, 1933), and has bred several of the nymphs recorded in the present paper. My colleagues Dr. A. J. Hesse and Mr. C. W. Thorne have also rendered much assistance. Mr. Harrison has contributed the following paragraphs on Economics, for which I express my thanks. ECONOMICS. Dragon-flies may take some part in the control of aquatic insects which are directly harmful to man—such as mosquitoes, or even of orchard pests in the vicinity of streams and dams. Their importance to inland fisheries depends to some extent upon the food requirements of the fish under consideration. Apart from vertebrates, dragon-flies are the dominant insectivores in many waters, and thus their position in the economy of trout waters is a doubtful one. Trout grow well and attain high condition upon a regular diet of the smaller aquatic insects—such as may-flies, caddises, and chironomids, and dragon-flies must be considered as direct competitors for such food. In stony mountain streams, dragon-fly larvae are usually well concealed from the trout, either on or beneath 4 SS ee ee eee ee ee —- ss sok - PS ee eS ee ee ee a a ee Notes on Dragon-flies (Odonata) of the S.W. Cape. 171 the substratum (Aeschnine, Gomphine, and Libelluline nymphs) or in vegetation (damsel-fly nymphs); and they are not very common in numbers in trout stomachs although they are sometimes prominent in bulk, except when floods have disturbed them from their usual haunts. At such times they occur more numerously along with other large food items—such as alder-fly larvae and river crabs. In the slower reaches of trout streams, the larvae may play a very useful part in the diet of the larger trout. On 9th May 1936 Mr. Robert Murdock, of Capetown, caught a rainbow trout of 11 lb. in the lower Eerste River at Faure, C.P. Its stomach contained 115 Mesogomphus nymphs (averaging about ? in. long), 10 Aeschna nymphs (4 in. to 1% in. long), and 10 Libelluline nymphs (about 4 in. long, and probably Trithemis sp.); and no other food could be recognised. Another trout of similar size caught at the same time contained a smaller amount of the same larvae. Aeschna and Anax nymphs are a direct menace to trout alevins and young fry. Adult dragon-flies are not usually common in trout stomachs, but on one occasion an adult of the large species Anax speratus, the “Great Red Dragon-fly,” was found in the stomach of a brown trout of 1 lb. caught in the Witte River, Wellington, during the summer peak of this species. (Several long Mermis-like parasitic Nemotodes were found loose in the same stomach. Tillyard (1917, p. 331) mentions a Filaria found by Selys.) Dragon-flies are of value as a fish-food in enclosed waters devoted to large-mouth black bass, as, on the whole, these fish require larger food-items than do trout. Libelluline dragon-fly larvae have been found to be a staple aquatic insect food taken by the bass in Paarde Vlei Lake, Somerset West, and some stomachs have contained from 50 to 100 individuals forming in bulk a good meal. The annual visit of the ‘Migratory Dragon-fly,” Sympetrum fonscolombe: Selys, to Cape waters has been noticed to induce surface feeding by black bass and to improve the fly-fishing. Black bass have been seen to leap right out of the water to take these insects as they flew close to the surface, particularly when the females were going through the actions of ovipositing or flying 7 copula with the males. TERMINOLOGY OF WING VEINS. Although Ris retained the old system, he was evidently open to receive the new interpretations of Lameere and Tillyard, based on a study of fossil wings. It might be thought convenient to adopt in Eee ee Sy > 172 Annals of the South African Museum. this paper the same terminology as in Ris’ monograph of the South African Dragon-flies, but one must keep pace with scientific progress. Tillyard’s notation, as expressed in his Insects of Australia and New Zealand (1926), is therefore adopted here, and the following table will show the chief alterations. Tillyard’s Biology of Dragon-flies (1917) is such a useful book that the preliminary new notation there used is also included. Ris 1921. Tillyard 1917. Tillyard 1926. R R 1 M,_3 M,_3 Rs M, M, KR, — Mya IR, M, M, Re Rs Ms TR, um, ms wi" 4 as Cu Cu+A Cu, +1A Cu, Cu, Cu, Cu, Cu, 1A AF Ab A’ Cugq Ac Ac The Radial and Median Supplements (Rspl, Mspl) may remain the same, though the latter should really be MAspl. Cu,, except in a single fossil form, is regarded as absent in all fossil and recent Odonata (Tillyard, loc. cit., 1926, p. 68, footnote). GENITALIA. For purposes of the systematics of this paper, attention had already been paid to the penis, and certain structural peculiarities had been noted, when Mr. Harrison gave me a pair of Ischnura senegalensis preserved in alcohol, which had remained in copula. The study of this pair enabled one to observe the interaction of the several parts of the g copulatory appendages and the 2 valve and stylets. The grasping of the 2 prothorax by the ¢ claspers is well known, but a figure (10) is given of this specimen showing how the Q pro- thoracic ridge is held between the lower and upper 3 appendages, while the projecting tergite of the 10th segment presses down on the 2 mesothorax (cf. Kennedy, Proc. U.S. Nat. Mus., vol. xlix, p. 296, fig, (89,1915). The relation of the parts of the 2nd segment ¢ and the genital segments of the 2 is shown in fig. le. The amplexus is actually closer, but for the sake of clearness in the figure the sexes have been pulled slightly apart. The bursa copulatrix or vagina (6) (cf. fig. 1 d) Notes on Dragon-flies (Odonata) of the S.W. Cape. 173 opens between the bases of the stylets on the 8th segment. Both pairs of stylets lie in the anterior pocket of the genital fossa of the $, where they are held by the anterior lamina and its posterior processes (Tillyard, Biol. Dragonfl., p. 218). Kennedy (loc. cit., p. 296) says that the anterior laminae fold inwards and that the ovipositor sheath of 2 does not fit into the depression on 2nd segment of ¢. Kennedy examined dried material. In the present instance the S$ anterior laminae undoubtedly clasp and enclose the ovipositor sheath of 2 (genital valves). The penis is inserted to the end of the bursa or vagina, with the apical fork recurved and the two spines pointing distally (for a more detailed figure of the penis of this species, see fig. 19). In all the Zygoptera studied in this paper, the penis is strongly curved like a fish-hook, and for the most part is heavily chitinised (see figs. 4, 9, etc.). It is attached at two points (fulcra) in the median line to the membranous and elastic wall of the genital fossa (fig. lc, h). From the upper (anterior) point of attachment a thickening of the wall of the fossa forms a chitinous strut (st.) on either side, which runs downwards and backwards, with an elbow- like articulation, to the posterior hamule. Splaying or an outward lateral movement of the hamules will exert a downward pull on the penis, which will thus be exserted from the fossa to enter the bursa of the @. The heavy chitinisation of the basal portion is continued on the anterior and ventral edges of the penis, but the inner and dorsal surface forms a groove (fig. 1h, gr.) bordered by a membranous flange. Useful specific characters may often be found in the sculpturing of the flanges and the presence of spines on the chitinous framework (see figs. 11-13). Kennedy (Ohio Journ. Sci., vol. xxi, No. 1, 1920) has already figured the penes of four South African species (3 genera), but as his purpose was to show the general relationships of the Zygopterous genera his figures are too small for details. The apical portion is more or less distinctly hinged. The chitinous framework of the basal part is continued as a pair of more or less strongly chitinised plates, and the groove formed by the flanges is continued on to a soft membranous flap of variable extent. Sub- apical, exsertile appendages (distal hook) of manifold shape occur (figs. 9, etc.), which are of specific or even generic importance. These appendages may be for the purpose of retaining the penis in the bursa, while the sperms or sperm-packets pass down the groove formed by the flanges. How the sperms are propelled is a matter 174 Annals of the South African Museum. Fic. 1.—Ischnura senegalensis. a, § and 2 in copula. 6b, apex of abdomen ¢ clasping prothorax 9. c, optical section of ¢ and Q genitalia in copula. d, penis g in bursa, ventral surface of latter uppermost. e, ventral view showing the 2 oviducts entering the bursa. ff, g, accessory sacs from other specimens. j, ventral view of median chitinous plate (cf. h, 7). Phaon iridipennis. h, sagittal section of 2nd abdominal sternite g, showing genital fossa, penis, and associated structures, and appendage of 3rd sternite. Chlorolestes conspicua. 1, ventral view of median chitinous plate. (acc.s. accessory sac (with or without saccule). a.l., anterior lamina. a.p., anterior pocket. 06., bursa copulatrix or vagina. g., 8th ganglion. g.f., genital fossa. gr., groove between membranous flaps on penis. m.c.:p., median chitinous plate. o.d., oviduct. p.h., posterior hamule. p.l., posterior lobe. sp., spine. st., strut.) | | Notes on Dragon-flies (Odonata) of the S.W. Cape. 175 for conjecture. Possibly the curious structure on the anterior margin of the 3rd sternite may help to perform this function. It would certainly seem to serve the purpose of retaining the ues in the genital fossa prior to intromission. This vesicular process of the 3rd sternite appears to be homologous with the “vesicle of the penis” (Tillyard) of the Anisoptera. It is heavily chitinised along the sides and basal ventral surface, leaving the whole of the upper (inner) and the lower apical surfaces mem- branous, and the membranous portion is apparently distensible. In ventral view the membranous area has the appearance of a window (fig. 9). It is filled with a homogeneous tissue, without tracheae, and there is no opening either to the exterior or into the body-cavity. In this latter respect it differs from its supposed homologue in the Anisoptera. In repose the apex of the penis is tucked in between this vesicle and the 2nd sternite. The posterior part of the 2nd sternite is occupied by a chitinous plate (fig. 1, m.c.p.), with rather indefinite boundaries in Phaon, but in Chlorolestes conspicua definitely shield-shaped with a minute tubercle, and in Ischnura tricuspid. Chitinous strips flank the median plate on either side and continue to the posterior hamules. Anterior to the plate the sternite is membranous and bears a flexible spiniform process (Phaon) (fig. 1, h, sp.) or a small knob with 2-3 setules (Chlorolestes conspicua). In one very important point the above description of the Zygopteran penis differs from that given by Tillyard (loc. cit., p. 217), viz. the lack of communication between the penis and the body-cavity. In fact, it would seem that every statement in the last paragraph on p. 217 is erroneous, except that the penis occupies the same position as the Anisopteran sheath, and that it lacks muscles and tracheae. Apparently Tillyard relied on statements by early writers and on “chitin preparations,”’ which latter would be flattened by the cover- slip in mounting, thus giving a distorted conception of the structure. When we examine the bursa copulatrix of the 2 we find that the impression created by Tillyard’s description (loc. cit., p. 221), viz that the possession of two accessory sacs is a feature common to all dragon-flies, must also be qualified. Here again Tillyard relied on early authorities, and happened to study an example of the Anisoptera, which confirmed the earlier work. In all the Zygoptera here examined the bursa has only one accessory sac.* * This discrepancy was also noted by George (Quart. J. Microsc. Sci., ns., vol. Ixxii, 1928, pp. 469-470) when studying Agrion. VOL. XXXII, PART 3. 13 176 Annals of the South African Museum. The bursa is a fairly elongate tube, whose diameter in the various genera can be more or less correlated with the robustness or bulk of the subapical appendages on the penis of the respective males (ct. Phaon, figs. 1, h and 2, 6, with Ischnura, fig. 1, d). Its lateral walls are more or less chitinised, chiefly towards the inner end, sometimes (Phaon, Chlorocypha) forming rather definitely demarcated Fic. 2.—Bursa copulatrix (vagina) and accessory sac 2 of Zygoptera. a, Chloro- cypha caligata. 6b, Phaon iridipennis. c, Lestes virgatus and plagiatus. d, e, f, Chlorolestes conspicua, with variation in shape of apex of accessory sac. g, C. longicauda, accessory sac. h, C. peringueyi and nylephtha. 1, j, Elationeura mutata and frenulata, respectively. k, Pseudagrion caffrum. (In a-d, h-j the opening of the united oviducts is dotted. Variation in apex of accessory sac indicated by dotted line in a, b.) chitinous plates, and as a rule with a patch of minute granules on the internal surface (fig. 2). The oviducts (o.d.) enter the bursa by a common opening on the ventral surface. Between this opening and the dorsal accessory sac is a small (Ischnura, Enallagma, Pseudagrion, Allocnemis), or very small (Lestes), chitinous patch; in Elattoneura and Chlorolestes, where the opening of the accessory sac into the bursa is large, there ae Notes on Dragon-flies (Odonata) of the S.W. Cape. ii is a feebly chitinised patch on either side, which in Phaon and Chlorocypha is scarcely traceable. The accessory sac (acc.s.) in Ischnura, Enallagma, Elattoneura, Pseudagrion, Allocnemis, Phaon, and certain species of Chlorolestes has its own accessory saccule attached to its posterior surface by a longer or shorter duct. Both the sac and its saccule are distensible, and vary in size in different individuals of the same species. When largely distended the contents seem to be of a homogeneous glutinous constitution, not spermatozoa. In certain species of Chlorolestes, viz. conspicua, longicauda, fasciata, umbrata, the saccule is bifurcate; in two other species, peringueyi and nylephtha, there is no saccule at all. It is noteworthy that this difference is correlated with certain differences in the § appendages (see infra, p. 191). In Lestes virgatus and plagiatus there is also no saccule; and Chlorocypha, in which the whole bursa closely resembles that of Phaon, has only a slight distension in the position of the saccule (the same shape was found in two specimens). Thus it will be seen that there is considerable variety in the bursa and that it is worthy of further study. The very curious differentia- tion in Chlorolestes and its correlation with features of the ¢ is especially intriguing. The technique here used involved cutting out only the 8th sternite and underlying tissues, and no attempt was made to examine the accessory glands in the 9th segment noted by George (loc. cit., 1928, p. 469). Turning to the Anisoptera we find that the 2 bursa possesses two accessory sacs, but that there may be considerable variations in shape; and that previous accounts of the structure of the penis cannot be confirmed. Both Ingenitzky (Zool. Anz., vol. xvi, p. 405, 1893 *) and Tillyard (loc. cit., p. 217) maintain that the lumen of the penis opens to the exterior by a pore, on the lst joint according to the former, on the 2nd according to the latter. If there were an opening one would expect, after treatment with KOH, to be able to press out the softened and disconnected contents (muscles, tracheae) through this opening. Williamson (Occ. Papers, Mus. Zool. Ann Arbor, Michigan, No. 80, 1920, explanation to fig. 8 on pl. 1) says there is an opening on the 1st joint and also one on the (presumably) 3rd, and that in relaxing the specimen air and fluids were discharged through the opening * T have not seen the paper in Russian (Warsaw Univ. Bericht, No. 1, 1893) where his results are set forth in greater detail. 178 Annals of the South African Museum. on the Ist joint. As he used dried material the possibility of cracking and accidental pricking by dissecting needles is not excluded. For my part, in all the dozens of penes examined for purposes of this paper, including both dried and alcoholic material, I have not been able to squeeze out any contents or to satisfy myself of the existence of any natural opening. On the other hand, there is an opening at the base of the “bulbus”’ (Ingenitzky) or “vesicle” (Tillyard) into the body-cavity, through which tracheae, etc., enter the penis, though Ingenitzky (loc. cit., p. 406) says the bulbus is completely shut off from the body-cavity. The apex of the Ist joint on its dorsal surface, and the dorsal surface of the 2nd joint are much less heavily chitinised than the rest of these joints, being sometimes quite membranous, and this gives the impression of an opening. The dorsal surface of the 2nd joint is more or less grooved (as Ingenitzky correctly states), leading on to the apical joint with its manifold varieties of lobes and processes. Instead of the bulbous Ist joint (or vesicle) forming a reservoir for the sperms as Ingenitzky maintained, it would seem that the sperms or sperm-packets prior to copulation are merely held in position between the penis, the sheath (Ingenitzky: ligula), and the sternite of the 2nd segment;. during copulation they could be pressed along the channel formed by the groove on the 2nd joint of the penis and the opposed sheath, while the apical lobes of the penis distend the vagina. For determination of the actual method, however, one needs a pair preserved in copula. As mentioned -above, the 2 bursa possesses a pair of accessory sacs. Figures are given here of the bursa in most of the genera studied for this paper. In most cases the bursa is shorter than in the Zygoptera, but is fairly long in the Aeschnines (Anaz, Aeschna) and the Corduline Presba. In the Libellulines there is a dorso- lateral pair of more or less well-chitinised areas, with or without a median connecting band, and anteriorly to them a median projecting chitinous plate. The latter varies in shape in different genera, and apparently serves as a muscle attachment. In the closely allied genera Trithemis (arteriosa and risi) and Helothemis (dorsalis typical and the var. mentioned infra, p. 255) this plate is of the same shape. This plate is absent in Presba, Mesogomphus, and Ceratogom- phus, and in the Aeschines, and in the latter the paired chitinous areas are also absent. In the Gomphines there is a long chitinous spiniform process at the anterior end of the bursa, dividing posteriorly into a fork. An Notes on Dragon-flies (Odonata) of the S.W. Cape. L79 : Fic. 3.—Bursa copulatrix and accessory sacs 2 of Anisoptera. a, Anax imperator mauricianus and Aeschna minuscula. 6, Presba venator, with dorsal view of chitinous plate. The genital valve at upper left-hand corner is abnormally shaped (cf. fig. 1, ¢ in original description. Stylops, 1933). c, Mesogomphus cognatus and Ceratogomphus pictus, in latter the anterior chitinous process is longer. d, Pseudomacromia torrida, with dorsal view of chitinous plates. e, Orthetrum capense, with dorsal view of chitinous plates. jf, Crocothemis erythraea, with dorsal view of chitinous plates. g, Helothemis dorsalis, Trithemis risi and arteriosa, with dorsal view of chitinous plates. The paired plates at base of fan-shaped plate absent in typical H. dorsalis. h, Sympetrum fonscolombei, with dorsal view of chitinous plate. 7, Palpopleura jucunda, with dorsal view of chitinous plate and accessory sac. j, Palpopleura lucia, dorsal view of chitinous plate and accessory sac. (The opening of the united oviducts indicated by dotted lines in a-2.) ath mteie ee EE ee ah. wee 180 Annals of the South African Museum.” indication of such a process, but not chitinised or forked, is seen in Presba. In the Aeschnines there is a median chitinous patch in a corresponding position at the end of the bursa, and a median chitinous rod-like thickening of the ventral wall of the bursa. The accessory sacs in the Libellulines are mostly elongate and slender tubes, transversely wrinkled as if indicating a capacity for considerable distension. Viewed dorsally they spread out laterally and then curve posteriorly ike two horns. In one example of Pseudomacromia and six Crocothemis they were large and of consider- able diameter. In Palpopleura they are saccoid in shape; in yucunda (2 specimens examined) they are all small compared with a median prolongation of the bursa, but in lucia (also 2 specimens examined) they are very large and the bursa has no median prolongation. In the Gomphines the accessory sacs unite and enter the bursa by a common tube of rather large calibre. In Presba and the Aeschnines they appear as accessory appendages to a large sac, which in the Aeschnines is enormous; when fully distended this sac must almost fill the whole of the 8th segment. | From the foregoing remarks it will be realised that the genital structures in both Zygoptera and Anisoptera are in need of much fuller and detailed comparative study. I have not had recourse to. serial sectioning, and have confined my study to such genera as are found in the 8.W. Cape. Key to the genera of the S.W. Cape region: Imagos. Zygoptera.. I. Numerous ante-nodal cross-veins. Clypeus (epistome) enlarged and pro- jecting. Tibiae of ¢ flattened : 3 ‘ ‘ . Chlorocypha. II. Only 2 ante-nodal cross-veins. A. Pterostigma long, covering 2 or more cells. Superior appendages of ¢ forcipate. 1. Origin of R,,; nearer to arculus than to nodus : . Lestes. 2. Origin of R,,,; at nodus, or nearer to nodus than to arculus Chlorolestes. B. Pterostigma short, covering 1 cell. Superior appendages of g not forcipate. 1. Quadrilateral rectangular. a. IA extending several cells beyond nodus. R,,,; at nodus, IR, 1 cell distal. Pterostigma cream, end of abdomen orange. : : ‘ : ; . Allocnemis. b. IA absent, A’ stopping at cross-vein which continues distal end of quadrilateral. IR, at nodus, R,,; 1 cell proximal. Pterostigma brown. @ with 2 pairs of prothoracic stylets Elattoneura. a > 4 Notes on Dragon-flies (Odonata) of the S.W. Cape. 181 2. Quadrilateral oblique, anal distal angle acute. a. Origin of A’ at Ac or very slightly proximal. 2 without spine on 8th sternite. i. Colour scheme mostly blue and black. No transverse ridge at frons. 5 Sipe appendages of ¢ usually bifurcate . : ‘ . Pseudagrion. ii. Colour scheme orange or peal A transverse ridge at frons. Superior appendages of ¢ short, entire Ceriagrion. 6. Origin of A’ considerably proximal to Ac. 2 with spine on 8th sternite. i. Rg in fore-wing at 5th or 6th, in hind-wing at 4th or 5th, post-nodal cross-vein. Pterostigma of ¢ alike in fore- and hind-wings . : é . Enallagma. ii. Rg at 4th and 3rd cross-veins respectively. Pterostigma in fore-wing of ¢ black and blue . ; Ischnura. Anisoptera. I. Lateral lobes of labium about equal to median lobe. Triangles similar in both wings, their long axes in long axis of wing. Ante-nodal cross-veins in costal and subcostal areas not coinciding. A. Eyes widely separated. Inferior 2) eae 3 bifid. 1. No anal loop 5 : j : Mesojgomphus. 2. A distinct anal loop . : Ceratogomphus. B. Eyes contiguous. Inferior se nendaw 3 at bifid. 1. IR, forked proximal to pterostigma, more than 2 rows of cells between its branches ’ 3 : Aeschna. 2. IR, forked distal to middle of Se ey only 2 rows of cells between its branches : . ¥ Anax. II. Lateral lobes of labium very large, contiguous in edt line; median lobe very small. Triangles dissimilar, long axis transverse in fore-wing, longitudinal in hind-wing. Ante-nodal cross-veins coinciding. A. Anal border of hind-wing excised in g. Auricles on 2nd abdominal segment §. ‘Triangle of fore-wing not very narrow. 1. Anal loop short, 3 cells in width . ? ‘ : Macromia. 2. Anal loop longer, 2 cells in width . : é . Presba. B. Anal border rounded in both sexes. No auricles. Teanslen in ae -wing narrow. 1. Arculus at or distal to 2nd Ang. More than 10 Angq in fore-wing, the last one complete. : : : ; Orthetrum. 2. Arculus proximal to 2nd Anq. a. Last Anq in fore-wing complete . 3 . Helothemis. 6. Last Ang incomplete. i. Costal margin in fore-wing indented. Wings with black and yellow markings ; . Palpopleura. ii. Costal margin not indented. a. Number of Ang 64-73 . : . Sympetrum. fp. Number of Anq 93-133. 182 Annals of the South African Museum. § Triangles at about same level in fore- and hind-wing. * MA and Cu, in fore-wing divergent. One row of cellsinIR,-Rspl : . Crocothemis. ** MA and Cu, parallel or slightly convergent. + Anal loop short. Medium-sized species (hind- wing not exceeding 30 mm.) . Trithemis. tt Anal loop long. Large species (hind - wing 40 mm.) : : . Pseudomacromia. §§ Triangle in fore-wing considerably distal to level of triangle in hind-wing . ; 4 Pantala. Key to genera of the S.W. Cape region: Nymphs. Zygoptera.—Body slender. 3 (or 2) lamellate or triquetral gills at end of abdomen. Mask flat. I. Two triquetral caudal gills . : ; ‘ : . Chlorocypha. II. Three caudal gills, lamellate but sometimes more or Hess triquetral. A. Median lobe of mask with median cleft. Gills lamellate, broadly oval Chlorolestes. - B. Median lobe of mask not cleft. 1. Gills simple, triquetral . 5 : : . Allocnemis. 2. Gills nodate, lamellate, elongate. One ee seta, 3 lateral Pseudagrion. 3. Gills lamellate or feebly triquetral, lanceolate, subnodate or simple. a. Mental setae, 4-5; lateral, 6 f inal t Ischnura. 6. One mental seta, 3 lateral setae . : . Elattoneura. Anisoptera.—Body robust. Abdomen ending in an anal pyramid composed of 5 appendages. I. Mask flat. A. Only 4 antennal joints (4th often ea Median Jobe of mask not cleft. Legs more or less adapted for burrowing, fore and mid tarsi only 2-jointed. 1. Abdomen relatively narrow. Legs short ’ . Mesogomphus. 2. Abdomen relatively broad. Legs longer : Ceratogomphus. B. Antennae 7-jointed. All tarsi 3-jointed. Median lobe of mask cleft. 1. Eyes large, hemispherical. Mask not extending beyond mid coxae Aeschna. 2. Eyes very large, flattened dorsally. Mask extending to hind coxae Anax. II. Mask spoon-shaped, with broad lateral lobes. A. Lateral lobes of mask with deep indents on distal margin. Mental and lateral setae very stout. 1. Body flattened, abdomen broadly oval. Eyes very prominent. Macromia. 2. Body not flattened, abdomen elongate oval. a. Abdomen segments 8 and 9 without strong lateralspines Presba. b. Abdomen segments 8 and 9 with strong lateral spines Pantala. Notes on Dragon-flies (Odonata) of the S.W. Cape. 183 B. Lateral lobes with indents very shallow or obsolete. 1. Eyes small. Lateral margins of head behind eyes parallel. Legs short. Body strongly setose, opaque : : Orthetrum. 2. Eyes prominent. Lateral margins behind eyes convergent. Legs moderately long. Body sparsely setose or nearly glabrous, more less mottled and semitransparent. a. Abdomen dorsally not keeled 4 : . Crocothemis. 6. Abdomen medio-dorsally keeled and toothed. i. Segments 8 and 9 ending in small points. Medio-dorsal fa elothemis. / | Lrithemis. ii. Segments 8 and 9 ending in strong spines. Medio- dorsal teeth small. a. 14-16 mm. . : : Z . Sympetrum. fp. 24-26 mm. . . : . Pseudomacromia. teeth large Fam. LIBELLAGINIDAE. 1917. Tillyard, Biology of Dragon-flies, p. 274 (Calopterygidae part). 1926. - Insects Austr. N. Zeal., p. 80. 1934. Fraser, Fauna Ind. Odonata, vol. 1, p. 1 (Agridae part). Imago.—Wings seldom petiolate. Nodus distant from base of wing. Numerous ante-nodal cross-veins (Anq). Quadrilateral rect- angular or nearly so, long, crossed by 1 or more cross-veins. Nymph.—First antennal joint elongate. Guzzard nearly always with 16 folds. Mask with median lobe entire or more or less incised; mental and lateral setae absent. Caudal gills saccoid or triquetral, the median one sometimes short, lamellar, or obsolete. Lateral abdominal gills sometimes present. Subfam. LIBELLAGINAE. 1934. Fraser, loc. cit., p. 2. Gen. Chlorocypha. 1921. Ris, Ann. S. Afr. Mus., vol. xviu, p. 261 (Libellago). 1928. Fraser, J. Bombay Nat. Hist. Soc., vol. xxxu, p. 684. 1934. Id., loc. cit., p. 5d. 1936. Longfield, Tr. R. Entom. Soc., vol. lxxxv, p. 467. Imago.—Wings petiolate, hyaline in both sexes. Quadrilateral with only 1 cross-vein. Pterostigma long and narrow in both wings of both sexes. Ante- and post-clypeus (epistome) enlarged and projecting. Sectors of arculus arising separately. MA curved but not zigzagged until at level of or distal to nodus. 1A zigzagged almost from its origin. All tibiae in ¢$ flattened and dilated (but not in all species). 184 Annals of the South African Museum. Nymph.—First antennal joint very elongate, at least twice as long as 2nd, 5 flagellar joints. Mask flat, with median lobe narrowly and not very deeply incised. No lateral abdominal gills. Lateral \ Fic. 4.—Chlorocypha caligata (Selys). Imago: a, dorso-lateral view of right inferior appendage ¢, the upper left-hand hollow being position of insertion of the superior appendage. 6, penis, with sculpturing of the membranous portion further enlarged. c, ventral view of apical lobe of penis. d, genital valve °, with margin further enlarged. e, 3 major and 2 minor folds of the gizzard. Nymph: f, head with basal joints of antenna. g, dorsal view of XIth abdominal segment with bases of the 2 caudal gills, left cercoid omitted. h, lateral view of XIth segment with rudimentary median gill. i, ventral view of XIth sternite, with base of 1 gill. J, outer view of caudal gill. &, portion of mask. caudal gills triquetral, median gill reduced to a mere conical rudiment. Cercoids prominent. Gizzard with 16 folds. IXth sternite without processes (Anlagen of genital valves). Remarks.—Ris stated that a new generic name would be necessary Notes on Dragon-flies (Odonata) of the S.W. Cape. 185 for the African species, and in 1928 Fraser proposed Chlorocypha. In 1934 he includes the two Asiatic species while expressing doubt as to their being really congeneric. The Indian wittata has MA zigzagged almost from its origin, and quadrilateral with 3 cross-veins (Fraser, loc. cit., p. 56, fig. 19). If future students consider it advisable to separate them, a further new name will be required for these two species, leaving Chlorocypha solely for the African species. Karsch (Berl. Ent. Zeitschr., vol. xxxvii, 1893) described two forms of nymphs from Togoland (figs. 10 and 11), both of which he was inclined to assign to “Libellago.” Tillyard (loc. cit., p. 275) and Ris (loc. cit., p. 262) have accepted only one of these (fig. 11) as being in all probability the nymph of “Libellago.” This is confirmed by Fraser’s discovery of the Indian nymphs, and by the South African nymph described below. In the Togoland nymph only 1 of the triquetral lateral gills was present, and Karsch consequently did not appreciate the fact that the median gill was not merely missing in his specimen, but normally quite rudimentary. In 1928 Fraser (loc. cit., p. 691, pl. 11) described three species of Libellagine nymphs: two of Rhinocypha and one of Libellago (sensu stricto). These agree in all essentials with the present nymph. The XIth tergite is present as a short conical or spiniform appendix dorsalis in Rhinocypha, but does not appear in the figure of the Inbellago nymph. In the latter the cercoids are represented, but not in the two figures of Rhinocypha. The gizzard has 16 folds, each bearing 4-5 teeth. Fraser’s accounts of the antennae, and his figures are conflicting (see 1928, loc. cit., p. 691, pl. 11; and 1934, - loc. cit., pp. 4 and 5). Chlorocypha caligata (Selys). 1908. Ris in Schultze’s Reise, vol. i, p. 307. 1921. Id., loc. ctt., p. 262, fig. 6; and pl. vi, fig. 6. Imago.—All tibiae of 3 dilated. A figure of the lower appendages of g is given here because the teeth (4—6) on the inner upper margin are indistinctly represented in Ris’ figure. Penis as in fig. 4, 8, c. Gizzard with 16 folds, 8 major (of which 4 are shorter than the other 4) and 8 minor, each fold with 2 series of recurved spine-like teeth. Nymph.—Up to 18-19 mm. plus gills 6-5-7 mm. (only empty shucks examined). First joint of antennae equal to anterior width of mask, 24 times the length of 2nd joint, the 5 flagellar joints de- creasing in length, the 5th minute. Mask extending back to bases 186 Annals of the South African Museum. of fore-legs; median lobe moderately convex, crenulate, with short cleft, lateral margins with strong spines each on a scute-like base; inner margin of lateral lobes feebly crenulate. A small conical tubercle on each side of prothorax, a little above insertions of legs. Median gill (XIth tergite) rudimentary, merely a short concial process. Lateral gills elongate (24 times width of last abdominal segment), terete, triquetral, all three edges armed with strong spines and setae. The two halves of the XIth sternite (to which the lateral gills are attached) are produced in triangular lamellar processes. IXth sternite without any trace of the valvules or ovipositor sur- rounding the genital pores of imago. Gizzard? Localities. —In addition to the localities mentioned in Ris, I have seen examples from Port St. Johns, Pondoland; and from two localities in the 8.W. Cape: Palmiet River, near Kleinmond, Southern Hottentots Holland Mts. (H. G. W., December 1931, 1933; K. H. B. and H. G. W., December 1934), and Tradouw Pass, near Swellendam (A. J. Hesse, November 1925). Remarks.—There are no differences in the 3 appendages or penis between the Cape specimens and those from Zululand identified by Ris. The discovery of the Palmiet River colony and of the very interesting nymph is due to Mr. H. G. Wood, who in 1933 obtained two empty shucks. In December 1934 Mr. Wood and myself visited the locality with a view to obtaining living nymphs. The search, however, was unsuccessful, although all possible habitats were examined, not forgetting Fraser’s remarks (loc. cit., 1928, p. 691, and 1934, p. 4) on the habits of the Indian species. I have therefore figured a portion of the gizzard of the imago. Tillyard (1917, loc. cit., p. 107) says that in the imago the gizzard and its dentition is considerably reduced, though least in the Calo- pterygidae and Agrionidae. As the dentition in the imago C. caligata is by no means feeble, that of the nymph may be expected to be somewhat more powerful. Probably it will be found to be easily distinguished from that of the Indian species of Rhinocypha, which Fraser says has only 4-5 teeth on each fold. Apparently no trace of the gizzard remains in the nymph-shuck after emergence of the imago. The gizzard belongs to the fore-gut and is presumably drawn out through the mouth of the emerging imago, while the hind-gut is drawn out through the vent. The latter portion of the digestive canal remains in the nymph-shuck, and in the Anisoptera even the delicate “branchial basket” can be dis- sected out. But even in robust species of Anisoptera (e.g. Orthetrum, SE ee i te Notes on Dragon-flies (Odonata) of the S.W. Cape. 187 Pseudomacromia, Presba) I have not found the remains of the fore-gut including the gizzard. It would be interesting to know what becomes of it, as it is the one character lacking in the nymph-shuck which prevents one from giving a satisfactory diagnosis of the nymph based on an empty shuck. Probably the examination of a series of examples of nymphs fixed immediately before emergence of the imago, and during the early phases of emergence, would provide a solution of the problem. The imagos settle on rocks in mid-stream, with the wings folded over the abdomen. Pairing or oviposition has not been observed. Fam. LESTIDAE. 1917. Villyard, loc. cit., p. 276. 1933. Fraser, Fauna Ind. Odonata. I., p. 18 (Coenagridae part). Imago.—Wings petiolate. Only 2 ante-nodal cross-veins (Anq). Pterostigma large, covering 2 or more cells. Short intercalary veins between distal ends of longitudinal veins. Quadrilateral with anal distal angle very acute. Superior appendages of ¢ forcipate. Nymph.—Slender. Legs long. Caudal gills in form of simple lamellae. Gizzard with 8 major folds, each bearing a few large teeth. Subfam. LESTINAE. 1917. Tillyard, loc. cit., p. 277. fesa- Wraser, loc: cit/, p. 21. - Imago.—Arculus midway between base of wing and nodus. Sectors of arculus arising high up near R. Origin of R,,; much nearer to arculus than to nodus. Origin of A’ at Ac. Nymph.—Mask elongate, with incised median lobe; lateral lobes irregularly cleft; mental and lateral setae present. Caudal gills with sécondary tracheae at right angles to main axis. Fig. 5. Gen. Lestes Leach. Lestes plagiatus (Burm.) (upper) : . and virgatus (Burm.) (lower). 1971. Ris, loc. cit., P- 268. Apex of penis, with ventral 95a. Eraser, loc. cit., p. 29. view of linguiform process of former. It is not proposed to deal with the South African species of Lestes as only one species has yet been found in the 8.W. Cape. Kennedy, however, has proposed to 188 Annals of the South African Museum. separate L. virgatus under the generic name Africalestes (1920, Ohio Journ. Sci., vol. xxi, p. 84), and so a brief discussion of its validity may be included, together with figures of the penis. Africalestes is defined as having the venation of Chalcolestes (LZ. viridus), 1.e. differing from Lestes in that the upper half of the arculus equals the lower half; IR, is nearly straight throughout its length; the penis has a “strap-like inner fold as in Ceylonolestes”’ (the latter n.g. has “penis with a spiral strap on the terminal lobe’’). The equal division of the arculus by the sectors is a character in conflict with Tillyard’s definition of the Lestinae, and, moreover, all the South African Museum specimens of virgatus and plagiatus have the sectors arising high up near R. In some specimens of virgatus IR, is nearly straight, but in others distinctly zigzag; intermediate stages are found, and the wings on opposite sides of the same individual are not always symmetrical in this respect. Specimens from the same locality vary. The same variability is found also in plagiatus. Examination of the penes of 3 examples each of virgatus and plagiatus shows a strap-like projection in plagiatus but not in virgatus, though the latter was made the genotype. In view of the above conflicts I consider that the validity of Africalestes is very doubtful. The nymphs of the South African species are not known (for figure of a European species see Ris, loc. cit., pl. xu, fig. 1). Lestes plagiatus Burm. 1921. Ris; lecnet., p. 276, fig. 18; and pl. vi, fees: Locality.— Knysna (H. G. W., January 1936, 1 3). This is the first record of a species of Lestes so far to the south-west in the Cape, the nearest hitherto being Dunbrody (Uitenhage district). The penis agrees with those of the 3 specimens (identified by Ris) mentioned above. The apical lobe is spatulate, concave dorsally, and the two heavily chitinised areas at its base are connected by a less strongly chitinised horseshoe-shaped band which forms a slight ridge ventrally. Subfam. SYNLESTINAE. 1917.. Tillyard, loc. cit., p. 277. 1933. Fraser, loc. cit., p. 19. Imago.—Arculus nearer nodus than to base of wing. Sectors of arculus arising a little above half-way along arculus. Origin of Notes on Dragon-flies (Odonata) of the S.W. Cape. 189 R,,; at or slightly proximal to subnodus (usually). Origin of A’ distal to Ac. Cu, at its origin from quadrilateral strongly curved towards costa. Nymph.—Mask with incised median lobe; lateral lobes narrow, cleft into 2 teeth, movable hook long and slender; no lateral or mental setae. Pedicel of antenna elongate. Caudal gills with secondary tracheae oblique to main axis. Remarks.—This subfamily is only found in South Africa and Australasia, and according to Tillyard is a survival of the palae- genic fauna. Chlorolestes is characteristic of the Cape mountain ranges, although extending to Natal, Zululand, and the Transvaal. Gen. Chlorolestes Selys. 1862. Selys, Bull. Ac. Roy. Belg., ser. 2, vol. xiv, p. 33. 1920. Kennedy, Ohio Journ. Sci., vol. xxi, p. 84 (Huchlorolestes). 1920) Ris, loc. cit., p. 279. 1933. Fraser, loc. cit., pp. 19, 20. Imago.—Wings petiolate to level of quadrilateral (discoidal cell). IR, and R,,, arising nearer to nodus than to arculus. Nervures at end of wing strongly curved downwards towards hind margin. $ claspers with or without a basal tooth; inferior appendages entire or bifid; penis with or without an exsertile distal hook. Posterior hamules well developed. Accessory sac of 2 bursa copulatrix with a bifurcate saccule, or without a saccule (see p. 177). Nymph.—First joint of antenna long and slender, but not as long as 2nd. Cercoids acute, conspicuous. Pedicel of caudal gills flattened, forming part of the gill lamella, gills not caducous. Gizzard with 8 major folds, no minor folds. Remarks.—In the thoracic coloration all the species are much alike, except peringueyi. In venation there are some useful specific differences. All species except uwmbrata have 2 or 3 rows of cells in the area between IR,—R, (at level of pterostigma). All species have a single row of cells between IR,—R,,, except conspicua, which has 2rows. ‘Two pairs of veins, R, and IRs, R,,,; and MA, are subparallel in umbrata, peringueyi, and nylephtha, but divergent in the other species. The following revised key has been based mainly on the venation and the ¢ genitalia. The two species peringueyr and nylephtha, form a group sharply marked off from the rest. Another group is formed by fasciata, tessellata, and longicauda. 190 Annals of the South African Museum. Kennedy (loc. ci., p. 84) has proposed the genus Huchlorolestes to include fasciata, tessellata, and longicauda, based on the shape of the distal hook of the penis, and the position of Ac in relation to the first Ang. There is undoubtedly a clear distinction between this group and conspicua, but it does not seem to call for generic separation because, in my opinion, the difference between the conspicua and fasciata groups are less striking than those between the conspicua-fasciata group and the peringueyi group. mea tot a oN Cos SSG Cu2 MA Fic. 6.—Fore-wings of Chlorolestes conspicua 3 (upper) and peringueyi (lower). The position of R,,; is surely more important than the position of Ac, which is never quite constant even in the same species (see e.g. conspicua and peringueyr). With the position of R,,, are cor- related the characters of the g claspers, the penis, and the accessory structures of the 2 bursa copulatrix. The present grouping of the species was worked out, before seeing Kennedy’s paper, and there seemed no occasion to multiply names. Since, however, the fasciata group has been named, there is no avoiding a name (Ecchlorolestes nov.) for the more important peringueyr group. The interrelationships of the species are far better visualised by keeping all of them in one genus. In addition to that of conspicua, the nymphs of wmbrata, longicauda, peringueyt, and nylephtha have been correlated. That of umbrata does not attain so large a size as that of conspicua, but otherwise there are no tangible specific differences between any of the species. The acute points to the lateral abdominal keels are less well developed in longicauda than in other species, but as in these latter there is often variation, this character should not be used. Notes on Dragon-flies (Odonata) of the S.W. Cape. “PSI A noticeable feature of the gills in comparison with those of Agrionid nymphs is the flattening of the “pedicel,” so as to form part of the gill lamella. At first sight the gill appears to be 2-jointed or nodate. But these basal portions correspond with the chitinous rings which support the gills in Agrionid nymphs, and are really the remnants of the XIth segment (see Tillyard, 1917, Biol. Dragonfl., Fic. 7.—Chlorolestes conspicua Selys. Imago: a, 9 ovipositing in stem of reed. b, reed stem cut open to show 3 ova. c, genital valve 9, with margin further enlarged. Nymph: d, Xth abdominal segment with XIth tergite and median gill, showing suture (s). e, anterior margin of median lobe of mask. ff, 3 major folds of gizzard. g, genital valve 9. fig. 39, H, J.). In Agrionid nymphs the gills break off very easily at the junction between gill and basal supporting ring, whereas in Chlorolestes the gills are not caducous. Key to the species of Chlorolestes. I. R,,; at subnodus. Claspers (superior appendages) $ without either basal teeth, or lobes on inner margin. Penis with a distal hook. Accessory sac of 2 bursa with bifurcate saccule (fig. 2d). Chlorolestes (sensu stricto). A. Inferior appendages ¢ not bifid. Distal hook of penis apically acute. Pterostigma unicolorous. 1. Larger. Hind-wing 32-37 mm. Wings never banded. Yellow stripe on mesepisternum entirely dorsal to humeral suture. Veins R,-IR, and R,,;-MA evenly diverging conspicua Selys. 2. Smaller. Hind-wing 22-25 mm. Wings when fully mature banded. Yellow stripe on mesepisternum embracing the humeral suture. R,;-IR;, and likewise R,,;-MA, subparallel (except near margin) . : : ; . umbrata Selys. VOL. XXXII, PART 3. 14 192 Annals of the South African Museum. B. Inferior appendages ¢ bifid. Distal hook of penis with spatulate apex. Pterostigma bicolorous. Ac distal to 1st Ang . . Euchlorolestes Kenn. 1. Metepimeron with dark (metallic) stripe adjoining the 2nd lateral suture . ; . fasciata Burm. 2. Metepimeron wholly pale sath billy indiietinat rides of dark patches. a. Wings banded. ; : ‘ . . tessellata Burm. b. Wings hyaline. : : : longi Burm. II. Ry,; proximal to subnodus. Claspers ¢ with basal tooth. Penis without distal hook. Accessory sac of 2 bursa without a saccule (fig. 2, h). R,-IR,; and R,,;-MA subparallel ee near margin). Inferior appendages 3 not bifid . : ; : . Lcchlorolestes nov. A. Clasper with distinct lobe distally on inner margin. Anteclypeus pale cream. Nospotonmetanotum . ‘ peringueyt Ris. B. Clasper without distal lobe, merely bhidened: Anteclypeus metallic green. A blue spot on metanotum : : nylephtha n.sp. Chlorolestes conspicua vA 1839. Burmeister, Handb. Entom., vol. ii, p. 823 (longicaudum part). 1862. Selys, Bull. Ac. Roy. Belg., ser. 2, vol. xiv, p. 34. 1921. Ris, loc. cit., p. 280, fig. 15; and pl. vin, figs. 1, 2. 1921. Barnard in Ris, loc. cit., p. 445, fig. (oviposition and nymph). Imago.—Ac at about level of Ist wee (sightly proximal, at, or shightly distal). Origin of A’ at level of proximal side of Q (mis- printed “9” in Ris). Ry,,. at subnodus; IR, 1 cell distal. Ptero- stigma covering 3 (or 4) cells. At level of pterostigma 3 rows of cells in area IR,—R,, 3 rows in R,-IRs, these 2 veins diverging; 2 rows in IR,—R,,,, usually only 1 row from middle of pterostigma, but the double row sometimes continued almost to margin; 2 rows in R,,;-MA beginning at level of origin of IR,, usually changing to 3 rows near margin, these 2 veins evenly diverging. In MA-Cu, there is a single row of cells, but sometimes an incomplete double row, or in 8 up to 12 double rows. Genitalia 3.—Posterior hamules with antero-distal corners rect- angular. Penis with short, strongly curved hook distally, which apically narrows rather abruptly to an acute point. Claspers without either basal teeth, or lobes on inner margin. Inferior appendages with the chitinised portions forming single claw-like processes. Abdomen (excl. claspers)—¢ 47-51, 2 41-49 mm. MHind-wing, $32-33, 2 33-37 mm. Notes on Dragon-flies (Odonata) of the S.W. Cape. 193 Labrum and occiput pale yellowish or ochreous, the upper portion of latter metallic blue-green, often blackish in 2. Genae and bases of mandibles yellow. Anteclypeus pale ochreous, postclypeus dark brown. Labrum black. Frons and vertex metallic blue-green, darker and duller or often blackish brown in 9; the vertex separated from upper metallic portion of occiput by an ochreous or brown interocular band. Prothorax metallic green (dark) in 3g, duller and dark brown in 9, with a deep yellow round spot on each side. Mese- pisternum metallic peacock-green, often darker and more bronzy in 9, with a chrome-yellow stripe near (but wholly dorsal to) the humeral suture, extending 3—4 length of mesepisternum. Mese- pimeron metallic green except the antero-ventral portion. Mete- pisternum chrome-yellow, not extending quite to the 2nd lateral suture, along which is a metallic green stripe, more extensive and often brown in 9. Lower portion of metepimeron and metasterna pale yellowish or whitish, with white pruinosity in J. Abdomen as described by Ris, but bright metallic green in g, duller and more bronzy in 9; segment 1 laterally, segments 9 and 10 dorsally with pale bluish white pruinosity in g. Wings hyaline. Pterostigma russet in 3g, paler ochreous in &. Nymph.—To the description already given, may be added: Ist joint of antenna slender, 4—2 length of 2nd, 5 flagellar joints; margin of median lobe of mask finely setulose laterally, becoming spinulose medianly, inner margin of lateral lobe very feebly and inconspicuously crenulate; lateral abdominal keels ending in short acute points on the 6th (5th—7th)-9th segments; teeth on the major folds of gizzard all of the same size, alternating 3, 4, 3, 4, etc. Brownish, somewhat mottled, especially the gills, which usually have a dark transverse band. Oviposition.—The puncturing of twigs on trees overhanging streams has been recorded (Barnard, loc. cit., p. 445). The name of the tree, however, should be Cunoma. On Table Mt. the females have been observed ovipositing in the stems of rushes (Hlegia juncea). While Ovipositing the female is not held by the male, who may not even be in attendance. On one occasion a female was seen to interrupt egg-laying in order to pair with a male; the act lasted about two minutes, when the sexes separated, and the female continued egg- laying. The position of the female is shown diagrammatically in fig. 7, a; the wings are held horizontally at right angles to the body, and the abdomen forms a double right-angled bend at the 4th segment. 194 Annals of the South African Museum. Habits —This species inhabits wooded and open ravines in the S.W. Cape mountains, where it is common and widely distributed. On Table Mt. the imagos start emerging about the end of November or beginning of December, and they can be found on the wing throughout the summer ‘up to the end of April. One or two can occasionally be found in the first week in May, but by the middle of this month they have all disappeared. One specimen was seen on 21st August 1933 and was apparently freshly emerged. They settle on bushes and rushes, never far from the stream. The nymphs live openly on the bottom of streams or crawling about on rocks or submerged vegetation. They are very carni- vorous; a full-grown one has been observed to seize a tadpole almost as long as itself and hold on to it until its struggles ceased. Before emergence the nymph crawls up, by preference, rocks projecting out of the water. Localities.—Has been observed over the whole area in question, as far north as the Cedar Mts. (Clanwilliam), and as far east as the Zwartberg Pass in the Zwartberg Range (Prince Albert), and the Robinson Pass in the Outeniqua Range. Altitude from about 500 ft. to 5000 ft., but the character of the stream is more important than actual altitude. Remarks.—This is the only species which has a double row of cells between IR, and R,,; (not reckoning the 2 or 3 double cells immedi- ately within the margin). The area IR,—R, is broad, and sexual dimorphism occurs in the area MA-Cu,, where there is usually an extensive series of double cells in the @. Chlorolestes umbrata Selys. 1862. Selys, loc. cit., p. 37. 1921. Ris, loc. cit., p. 283. Imago.—Ac about at (or slightly proximal or slightly distal) level of lst Ang. Origin of A’ distal to proximal end of Q. Ry,5 at subnodus, IR, 2-3 cells distal. Pterostigma covering 2-3 cells. One row of cells in area IR,—Rs, a short series (2-3) of double or triple cells at margin; 1 row in R,-IR;, these 2 veins sub- parallel, very slightly diverging near margin where there are 2 rows of cells; 1 row in each of the areas IR;—-R,,;, R,,;-MA, and MA-Cu,, veins R,,, and MA subparallel except at margin. Genitalia $.—Posterior hamules apically narrowed. Penis with short, strongly curved distal hook, apically with acute point. Process Notes on Dragon-flies (Odonata) of the S.W. Cape. 195 of sternite III shorter than in other species (fig. 9r). Claspers without either basal teeth, or lobes on inner margin. — Inferior appendages with the chitinised portions forming single claw-like processes. Abdomen (excluding claspers)—j 34-37, 9 31-32 mm. Hind- wing, ¢ 22-25, 2 21-23 mm. Labrum black. Occiput bronzy black. Labium, genae, and bases of mandibles pale ochreous. Anteclypeus dark brown. Postclypeus, frons, and vertex metallic blue-green. Prothorax bronzy black, with a pale ochreous bar on each side. Mesepisternum bronzy greenish black, with bluish-white pruinosity in mature specimens ‘with banded wings, with a lemon-yellow stripe which at about 2 the length crosses the humeral suture on to the postero- dorsal end of mesepimeron. Remainder of mesepimeron except antero-inferior corner dull bronzy green. A lemon-yellow stripe along Ist lateral suture. A dark bronzy cuneiform stripe on metepisternum. Rest of metepisternum and metepimeron pale lemon-yellow or cream, with whitish pruinosity when mature; a dark stripe on postero-inferior margin of metepimeron. Meta- sterna cream, with dark streak anterolaterally. Legs testaceous, dark on outer edges. Abdomen metallic bronzy or bluish green, with pale narrow bands antero-laterally on segments 3-7; segments 9 and 10, and sometimes posterior half of 8 also, with bluish white pruinosity dorsally. Wings hyaline when freshly emerged; when mature banded, whitish pruinose from nodus to 6-8 post-nodal cross-vein in fore-wing, to 5-7 in hind-wing, thence brownish to pterostigma, or to middle of pterostigma. Pterostigma russet, unicolorous. @Q similar, but duller, the metallic portions more bronzy than green. Nymph.—Resembling that of conspicua except in its smaller size. Habits.—Occurring in the same habitats and with the same habits as conspicua, but apparently much more local. Localities —Table Mt., Cape Town (H. G. W. and K. H. B., March); Groot Drakenstein (K.H.B., March 1931); Kogel Berg, Hottentots Holland Mts. (K. H. B., January 1923); Palmiet River (H. G. W., January 1937); du Toit’s Kloof, Rawsonville (K. H. B., March 1932, and H. G. W., March, April 1934); Wellington Mts. (H. G. W., January 1934); Waaihoek Mts., Goudini (K. H. B., March—April 1934); Buffelshoek, Hex River Mts., Worcester Distr. (K. H. B., April 1936); Knysna (R. M. Lightfoot, March 1892, 2 $3); George (H. G. W., January 1936, gs); Coldstream (H. G. W., January 1936, 13). 196 Annals of the South African Museum. Remarks.—I have to thank Mr. D. E. Kimmins and Miss Longfield for comparing Cape specimens with the ¢ specimen in the British Museum referred to by Ris. In their opinion the specimens are certainly conspecific. The inferior appendages of the B.M. 3g have only a single point, corresponding with Selys’ description. Evidently Ris made a slip when he said (p. 284) the appendages were “‘of the same type as in the three following species”’ (i.e. fasciata, tessellata, longicauda), unless he intended to refer only to the superior ap- pendages. This is the smallest species of the genus. Apart from the non- bifid inferior appendages, it is easily separated from fasciata and tessellata, both of which develop bands on the wings, by the single row of cells in the area IR,—R,, and the subparallel pairs of veins R,, IR, and R,,;, MA, and also the position of origin of IR;. The penis is like that of conspicua. Chlorolestes fasciata Burm. 1839. Burmeister, Handb. Entom., vol. ii, p. 822. 1862. Selys, loc. cat., p. 36. 1920. Kennedy, Ohio. Journ. Sci., vol. xxi, p. 84 (Huchlorolestes f.). 1921. Ris, loc. cit., p. 284, fig. 17; and pl. vu, fig. 4. Imago.—Ac slightly distal to 1st Ang. Origin of A’ distal to proximal side of Q. R,,; at subnodus, IR, 1 cell distal. Ptero- stigma covering 3-4 cells. Two rows of cells in area IR,—-R,, 3 beyond pterostigma; 3 rows in R,-IR,, these 2 veins diverging evenly; 1 row in each of the areas IR, R,,, and MA-Cu,; 2 rows in area R,,,—-MA at level of origin of IR,, 3 rows distal thereto, these 2 veins diverging evenly. Genitalia $.—Posterior hamules narrowing apically. Penis with rather long, gently curved distal hook, cultrate in shape, the dorsal edge being thin, apex spatulate or with disc-like expansion; the distal corners of the two flaps, between which the hook lies, are sharp. Claspers without either basal teeth, or lobes on inner margin. Inferior appendages with the chitinised portion forming a bifid process. Abdomen.—é 32-38, 2 32 mm. Hind-wing, 5 25-28, 2 26 mm. Labium, genae, bases of mandibles, and anteclypeus pale ochreous. Labrum, occiput, postclypeus, frons and vertex metallic blue-green. Prothorax metallic blue-green with a yellow bar on each side. Mesepisternum metallic bronzy green (duller in 9), with a narrow yellow stripe at the humeral suture but not extending quite to Notes on Dragon-flies (Odonata) of the S.W. Cape. 197 dorsal end. Mesepimeron metallic green above, yellow below, with a metallic green or dark bar on ventral margin. Metasterna yellow, with small median line at hind end. Legs ochreous, darker and some- what metallic greenish on outer sides. Abdomen as described by Ris. Wings hyaline or banded; in the latter case whitish pruinose from Q or nodus to 5th-—6th postnodal cross-veins, thence brown- ish to a short distance proximal from _ ptero- stigma. Pterostigma bicolorous, blackish proximally, ferruginous distally. Localities (see Ris).— This species has not been found in the 8.W. Cape, except the one record © “ from “Albert Distr.” (= Prince Albert.) Remarks.—The metallic a f stripe along the upper fy. 8.—Chlorolestes, diagrams of thoracic patterns. half of the metepisternum a, conspicua. b, umbrata. c, fasciata. d, tes- : F sellata and longicauda. e, nylephtha n.sp. is the most conspicuous f, peringueyi. Cross shading represents difference separating this metallic green or bronzy; dots, brown; un- 5 h d d, ll : 4 iti ibe species from tessellata and Se a re, OF WEES longicauda. All three species form a closely allied group distinguished from the other species by the apex of the distal hook on the penis, the bifid inferior appendages of 3, and the venation. Chlorolestes tessellata Burm. 1839. Burmeister, loc. cit., p. 822. 1862. Selys, loc. cot., p. 35. ALS His, toe, et., p. 286, pl. vu, fig. 5. Imago.—Venation and 3 genitalia as in fasciata, except that the posterior hamules are slightly clavate, and the distal corners of the two flaps, between which the distal hook of the penis lies, are rounded. 198 Annals of the South African Museum. Abdomen.—41-44 mm. Hind-wing, 29-31 mm. Coloration as in fasciata, but the mesepisternal yellow stripe s Fic. 9.—Chlorolestes, 3 genitalia. a, b, c, d, e, penes of conspicua, umbrata (apex in lateral and ventral view), fasciata, tessellata, and peringuey respectively. f, 9, h, 1, posterior hamule (anterior margin to left) of conspicua, tessellata, peringueyi, and nylephtha respectively. 9, k, 1, m,n, ventral view of inferior appendage of conspicua, fasciata, tessellata, peringueyt, and nylephtha respec- tively. o, p, ventral and lateral views of upper and lower appendages of umbrata. q, dorsal view of upper appendage of nylephtha. r, s, ventral view of process of 3rd sternite of wmbrata and tessellata (and all other species) respectively. broader and extending the whole length, and also below the humeral suture on to the mesepimeron; whole of the metepisternum and Notes on Dragon-flies (Odonata) of the S.W. Cape. 199 metepimeron yellowish with only a faint and indefinite suffusion near the dorsal ends of these two plates. Localities.—See Ris, and add, Port St. Johns, Pondoland. Remarks.—l have seen one § from Grahamstown, April 1892, which is not the specimen seen by Ris as the date is different. The two old specimens mentioned by Ris are no longer in the South African Museum, but there are two from Port St. Johns, named by Ris. Some specimens from the 8.W. Cape may belong to- this species, but as they are all unbanded they are listed under longicauda. Chlorolestes longicauda Burm. 1839. Burmeister, loc. cit., p. 823. 1862. Selys, loc. cit., p. 35. 1921. Ris, loc. cit., p. 287, fig. 18; and pl. vu, fig. 6. Imago.—Venation, ¢ genitalia, and coloration as in tessellata, except that the wings apparently always remain hyaline. Abdomen.—(Transvaal and Zululand specimens) g 43-45, 9 38-45 mm. (Cape specimens) ¢ 35-39, 2 38 mm. Hind-wing (Transv., Zululd.) g 29-34, 2 33-35 mm. (Cape) 3g 27-28, © 31 mm. A pale cobalt-blue spot (often not conspicuous) between bases of hind-wings. Pterostigma in both sexes bicolorous, black (or very dark brown) proximally, cream distally (Cape specimens). - Nymph.—Resembling that of conspicua. Localities.—See Ris, and add: St. Mathews, King Wiliams Town Distr. (R. M. Lightfoot, 1894, 1 2); and the following Cape localities: Keurbooms River, Plettenberg Bay (K. H. B., January 1931); Seven Weeks Poort, Ladismith (K. H. B. and H. G. W., February 1932); Robinson Pass, Outeniqua Range (K. H. B. and H. G. W., February 1932); George (H. G. W., January 1936). Remarks.—Ris says this may be a geographical race or subspecies of tessellata. In view of the Western Province records it can scarcely be regarded asa race. The Western Province specimens are appreci- ably smaller than those from Zululand and the Transvaal. Chlorolestes peringueyi Ris. 1921. Ris, loc. ct., p. 282, fig. 16; and pl. vu, fig. 3. Imago.—Ac at level of 1st Anq (or slightly proximal or slightly distal). Origin of A’ at level of proximal side of Q. R,,; proximal to subnodus, IR, at subnodus. Pterostigma covering 2-23 cells. 200 Annals of the South African Museum. IR, strongly curved towards costa at level of pterostigma. Three rows of cells in area IR,-R,; 2 rows (distally) in R,-IRs;, these 2 veins subparallel except at margin; 1 row in each of the areas TR,-Ry,5, Ry,;-MA, and MA-Cu,, veins MA and R,,; subparallel except near margin. Genitaha §.—Posterior hamules apically narrowed. Penis without a distal hook. Claspers with basal teeth, and subterminal lobes on inner margin. Inferior appendages with the chitinised portions forming single processes. ‘ Abdomen.—38-39 mm. Hind-wings, 27-29 mm. Labium, genae, bases of mandibles, and anteclypeus pale buff or cream. Occiput black, becoming metallic greenish ventrally. Labrum, postclypeus, frons, and vertex metallic peacock-green in 3, duller in 9. Prothorax dull coppery or bronzy green, with 2 pale buff or cream round spots anteriorly. Mesepisternum dull metallic coppery or bronzy green, with 3 pale buff or cream spots near humeral suture, the inferior one sometimes divided into 2. Mesepimeron dull metallic coppery or bronzy with cream-coloured irregular stripe bordering Ist lateral suture. Metepisternum cream with irregular dark brown behind the spiracle, more or less confluent with a dark line along the posterior half of Ist lateral suture. Metepimeron cream with dark marks anteriorly and postero-dorsally. Metasterna pale cream or whitish with dark marks. Abdomen as described by Ris, except that “reddish brown” should be pale buff or cream (in living or fresh specimens); segments 9 and 10, sometimes also hinder half of 8, dorsally with pale bluish white pruinosity in g. Wings hyaline. Pterostigma dark brown in 3, paler in 9. Nymph.—Resembling that of conspicua. Habits.—This species inhabits open stream-beds, where the flies settle on the rocks with their wings spread out flat and closely adpressed to the surface of the rock. Their non-metallic (compared with the brilliance of e.g. conspicua) coloration harmonises very closely with the rocks, which, whether they be granite or sandstone, are of a more or less greyish colour. Only very rarely have the males been observed to settle on bushes; on the other hand, the females have frequently been seen in such situations, presumably with a view to oviposition, though the act in this species has not actually been observed. The resting position at night has not been observed. Localities.—Ceres (R. M. Lightfoot, April 1913); Waaihoek Kloof, Goudini (K. H. B., March, April 1928); Bain’s Kloof (east side) K. H. B., Ist May 1933); Jonkershoek, Stellenbosch (H. G. W., Notes on Dargon-flies (Odonata) of the S.W. Cape. 201 April 1931); Breede River near Mostertshoek (K. H. B., April 1933); du Toits Kloof, Rawsonville (H. G. W. and C. W. T., April 1934); Zanddrift and Buffelshoek Kloofs, Hex River Mts., Worcester Distr. (K. H. B., April 1930 and April 1936). Remarks.—In common with umbrata and nylephtha this species has the 2 pairs of veins R,-IR,, and R,,;-MA subparallel. The strong curve of IR, distally is characteristic. Also the thoracic colour scheme is quite different from that of the other species. Chlorolestes nylephtha u.sp. Imago.—Ac slightly distal to level of Ist Ang. Origin of , F ‘ » = f a f 4 \ f ' : ‘ ’ ( 265 ) 14. Report on Some Anthidiine Bees (Apoidea) in the South African Museum.—By G. A. Mavromovustakis, Limassol, Cyprus. THE present paper is based on a collection of Anthidiine bees loaned by the South African Museum, through the kindness of the Director and of Dr. A. J. Hesse. The type of the new species is placed in the South African Museum. Anthidium tuberculiferum Brauns. NAMAQUALAND. Bowesdorp, 1 3, November 1931; Klip Vlei, Garies, 2 $$, November 1931. This species was described by Brauns from Willowmore; it is a member of Anthidium (s. str.). Anthidium platyscepum Mavromoustakis. S.W. Arrica. Kaross, 1 g¢, February 1925. Anthidium poecilodontum Mavromoustakis. NAMAQUALAND. Bowesdorp, 2 92 1 3, November 1931. The type locality of this species is Matjesfontein. Anthidium capicola Brauns. Cape Province. Olifants River Valley, Clanwilliam, 8 92 2 33. Similar to topotypical specimens in my collection, but differs as follows :— First abdominal tergite black; tergites 2 to 4 with a short linear, central, pale yellow stripe interrupted in middle. Dianthidium bruneipes (Friese). Female.—Length 11:5 mm. Black; lower margin of clypeus crenulate; pulvilli present; second recurrent nervure out of second transverse cubital nervure; scutellum normal and rounded; apical half of abdominal tergites 1 to 5 yellowish brown; last abdominal tergite black, apical margin rounded and sightly emarginate in middle. 266 Annals of the South African Museum. Male.—Length 11-5 mm. Black; similar to the female; clypeus yellow; face laterally, clypeus, supraclypeal area and between antennae with somewhat dense and shining white hairs; cheeks with shining white hairs; vertex and occiput with pale yellowish brown hairs. Thorax with pale yellowish brown hairs above and shining white hairs at sides; wings clouded; sixth abdominal tergite with apical margin produced in middle, and with a short and stout spine at sides; seventh tergite with a curved spine laterally, and between these lateral spines tri- angularly produced in a long spine (the spine rounded at the apex and longer than the lateral ones); ventral segments black, apical margins of sternites 1 to 4 yellowish brown; fifth ventral segment with a very short tubercle in middle of apical margin; sixth ventral segment polished and shining, base concave except laterally, apical margin truncate. NAMAQUALAND. Kamieskroon, 1 91 3, September 1930. _ This species was described by Friese in the female sex (Zool. Jahrb., xxxv, Abt. f. Syst., p. 596), and it is a member of Dianthidium of the group of Dianthidium rufocaudatum (Friese). Anthidiellum polyochrum nu. sp. Female.—Length hardly 5 mm. Black with cream-coloured markings; clypeus shining, densely punctured, cream-coloured; lateral marks on face nearly reaching level of antennal insertions and a mark above each eye, cream-coloured; a longitudinal cream-coloured stripe on sides of supra-clypeal area and on each inner side of antennal insertions ; mandibles black with a basal cream-coloured rounded spot; antennae black brown; vertex and occiput moderately shining ; clypeus, sides of face and cheeks with some very short and sparse shining white hairs. Thorax black; mesonotum strongly and densely punctured, dullish, and with a cream- coloured mark on each side near tegulae; tubercles sharply pointed, cream-coloured; scutellum projecting, rounded at sides, apical margin truncate in middle and without emargination; apical margin of scutellum cream-coloured and broadly interrupted by black in middle; tegulae very finely punctured, deep brown, with subhyaline margin except behind; wings clear; second recurrent nervure out of second transverse cubital nervure; thorax with some very short and sparse white hairs above and shining white hairs at sides; femora and tibiae black; apex of femora narrowly yellowish brown; anterior tibiae with Anthidune Bees (Apoidea) in the South African Museum. 267 the basal cream-coloured stripe nearly reaching apex; middle and hind tibiae with a basal cream-coloured stripe above; anterior tarsi yellow brown; middle tarsi with nearly black brown basitarsi, small joints yellow brown; hind tarsi with basitarsi cream-coloured, small joints brown, last ones yellow brown; legs with shining white hairs; hind basitarsi with short and dense golden hairs on inner side; spurs yellowish; pulvilli present. Abdomen black and shining ; first and second tergites finely punctured; third tergite finely and somewhat sparsely punctured; first tergite with a lateral cream-coloured mark; second tergite with a longer lateral cream-coloured mark; tergites 3 to 5 on each side of middle with a short basal cream-coloured stripe, that on fifth longer; sixth tergite cream-coloured, sides basally black; third tergite with a narrow cream-coloured mark at sides; sixth tergite with very short white hairs; ventral scopa very light golden white. S.W. Arrica. Kaross, near Franzfontein Kaokoveld, 1 9 (type), February 1925. Related to Anthidium absonulum Ckll., but the latter has the abdominal markings lemon yellow, the punctures of abdominal tergites strong, and the tegulae black. This small species is a member of the Anthidiellum zebra (Friese) group. r ut ( 269 ) 15. A New Species of Plume-footed Mite from South Africa.—By R. F. Lawrence, Ph.D., Director, Natal Museum, Pieter- maritzburg. | (With 4 Text-figures.) In December 1936, Dr. A. J. Hesse of the South African Museum, Cape Town, sent me for identification two small mites collected by himself and Mr. C. Thorne at Kamieskroon and Bitterfontein in Namaqualand. These appeared at first sight to be EHrythraeus plumipes L. Koch, but for various reasons which appear below, it has been decided to describe them as a separate species, Hrythraeus namaquensis. This Namaqualand form in its appearance and habits very closely resembles Hrythraeus plumipes L. Koch, a widespread species or group of allied species, characterised by a peculiar feathery tuft of modified hairs on the tibia of the last pair of legs. Dr. M. André has cleared up the synonymy (1, 2), and given an account of the geographical dis- tribution and habits (8) of H. plumipes in a recent series of very helpful papers. According to him the distribution of the species is as follows: Jersey, France, Switzerland, Spain, Northern Africa, Corfu, Russian Armenia, and Turcomania. The Namaqualand form of Hrythraeus described in this paper is of unusual interest; though many Trombidiform mites have been collected in various parts of South Africa, none resembling EL. plumipes of the European and Mediterranean regions have hitherto been discovered. The South African Museum has organised a number of expeditions to Namaqualand during recent years; on at least four occasions large collections of invertebrates, including. Acari, have been made at Kamieskroon and Bitterfontein. On two of these occasions I accompanied Dr. Hesse and Mr. Thorne and collected numerous Trombidiform mites. More Acarine material has therefore been taken in Namaqualand than in most other regions of South Africa, so that this form of Erythraeus must, to say the least, be un- common in South Africa. The discoverers, who have done intensive collecting in South Africa for many years, stated that they had never before seen an Arachnid anything like it. 270 Annals of the South African Museum. HABITS. Dr. Hesse has sent me an account of his observations on the mite r which can be summar- a ised as follows: In agree- ment with the statements of various authors, it is a diurnal animal running over the bare sand. The specimens were captured during the hottest time of the day, one at noon, the other in the early afternoon; they were found on reddish drift- sand near a dried river bed, and did not seem to be associated with . any special plant or type of vegetation, though a species of Mesembryan- : themum was the common- est plant in the vicinity. _“Ttdoes not run straight, but more or less in circles.” With regard to its running powers, Dr. | Hesse says “it runs like | a small Solpuga but even | quicker,’ and “it runs so quickly that the eye has difficulty in following it.” With regard to the | position of the posterior legs he says, “the two feathery processes are kept upright,” while the specimen caught by Mr. Thorne “had the last > pair of legs projecting straight backwards while running.” In both cases there was an extremely rapid alternate x ~ eS VOY 5 Sree SS Lies > Lz aN SSS SS Tet | SSP ey SS LL wg thle LP Iternately (see text). The arrows indicate the arc through which the legs move a ae! Via. 1.—LHrythraeus namaquensis n.sp. Horizontal and erect positions of the posterior legs while running. A New Species of Plume-footed Mite from South Africa. 271 movement or vibration of these legs, as illustrated diagram- matically in fig. 1, adapted from a sketch supplied by Dr. Hesse. This account of the positions assumed by the last legs agrees with the observations of various Kuropean authors which have been ably summarised by M. André (8). Quoting Dr. André, the posterior legs, according to Gessner (in 4), are trailed like caudal appendages, while Luff (in 5), Tragardh (6), and Birula (7) describe them as being carried erect in the air; this attitude is figured in M. André’s paper on the distribution of E. plumipes (8). J. Millot (in 8) states they are directed backwards at an angle of 45° to the body, while Cambridge (8) describes them as being moved rapidly, the feathery tufts resembling the dance of minute flies above the mite. Birula and Tragardh also mention the ‘to and fro’ movements of the legs while they were held aloft. All these attitudes, then, are substantially the same as reported in the case of LH. namaquensis, and probably represent different phases of the same movement of the legs. The Function of the Modified Posterior Legs in Erythraeus. With regard to the function of these curiously modified appendages nothing is known. Lucas (9) has compared them to the tufts of silky hairs on the legs of certain longicorn beetles. In advancing a tentative and quite hypothetical explanation of the use to which these special- ised legs are put, four possibilities may be considered: (1) a mechanical function aiding locomotion; (2) a sensory function; (3) a defence mechanism; (4) a sex character. (1) A Mechanical Function Arding Locomotion.—Tragardh (6) is inclined to regard them as balancing organs which enable the mite to shift its centre of gravity while in motion. It seems also admissible to suggest that these feathered legs may be used simply as sails to take advantage of a following wind, the legs being lowered when the wind is in an adverse direction, or when the mite has arrived at its temporary goal; the movements of the legs may be used to test whether the wind is favourable. This would partly explain the extraordinary speed which the mite exhibits in running from place to place. It seems difficult otherwise to account for such exceptional running perform- ances, carried out with only three pairs of legs which are by no means strongly constructed. (2) A Sensory Function.—Tragardh (6), as an alternative to his suggestion that the legs may serve as balancers, supposes that they may be tactile organs for apprehending movements of the air, enabling the animal to take refuge from strong winds. Dr. Hesse also con- VOL. XXXII, PART 3. 19 272 Annals of the South African Museum. siders that “‘they may be connected with a sense which enables the mite to become aware of external sensations connected with weather, such as wind movements, etc.” He says further, “it is possible that the mites appear only just after or just before rain. The one specimen was taken a day before thundery weather set in, the other a day or two after a storm had passed over the spot. The ground was still damp.” In this connection it may be noted that the flattened hairs which form the feathery tufts of the posterior legs are covered with large numbers of fine setae (fig. 2, f, g), a fact also noted by Tragardh (6) in specimens captured in the desert near Cairo. (3) A Defence Mechanism.—Tragardh (6) does not think that this modification can be interpreted as a defence mechanism on account of its delicate structure. In this connection one fact should be mentioned which does not seem to have attracted comment from any previous authors. In both specimens sent me, both of the posterior legs had become detached, all four appendages having broken off at the same point, the junction of the trochanter and the trochantin; the remaining legs of both specimens were all am situ. Dr. Hesse also remarks of the posterior legs, “they are easily detachable, for the one specimen (captured by Mr. Thorne) lost one of its appendages very easily.” This fact points to a specially weakened joint at which the posterior legs can be easily thrown off. Though this may not necessarily be considered a defence or escape mechanism, as the legs may be discarded for other reasons, it is difficult to avoid the suspicion that natural enemies such as sand-living lizards would be more attracted by, and more lable to seize, the plumed legs waving aloft than the mite itself. Desert-living lizards, which are mainly insect feeders, would very readily snap at a small fast-moving object, and might in this case easily mistake the vibrating tufts of hair for a hovering insect. Other Arthropoda, like the centipede Cryptops, lose the last pair of legs very easily; in members of this genus the gonopods almost inevitably break off in captured specimens, always at the same joint. (4) A Sex Character.—There seems to be some justification for in- ferring that these feathered legs in Hrythraeus plumipes and E£. nama- quensis may have a sexual significance. A very similar condition is found in many diurnal Solifuges, a group which is most usually found in desert and sandy localities resembling those in which the plume- footed mite has been captured. In the genus Solpuga, a distinct “mane” of long silky hairs often adorns the distal segments of the fourth leg, very conspicuous in South African species of diurnal habits A New Species of Plume-footed Mite from South Africa. 273 such as S. derbiana and S. chelicornis (11, figs. 2, 3, pl. ii); though quite noticeable in the female this “‘mane”’ is more strongly developed in the male, These Solifuges, when seen darting hither and thither during the hottest time of the day, also have a rapid and erratic gait. The writer (12, p. 262) has recorded his impressions of the males of Solpuga sericea in South West Africa as follows: ‘The erect mane of the hind legs, together with their erratic gait, gives them a striking resemblance to a fluffy wind-blown seed.” Of these four possible explanations, we have no information with regard to (4); of the others (1) and (3) appear to me to have more in their favour than (2); it is not impossible to suppose that these specialised appendages may have a mechanical function while also serving as a defence or escape mechanism. If this view is taken, the plumed feather-like legs would in the first place have been evolved as a purely mechanical device for taking advantage of currents of wind; such a device would be of great value in accelerating the speed of a small light animal in which the legs are comparatively short and weak. If he were to judge from preserved specimens only, a morphologist would hardly guess that this small mite could rival the speed and activity of the Solpuga. In the second place, the habit of holding the conspicuously feathered legs erect would become a danger to their owner on account of their attraction for predaceous animals. To meet this danger the joint between the two trochanter segments, which allows the leg to be readily detached, would have been developed as a secondary escape mechanism. The Namaqualand form is described below in some detail, on account of its close resemblance to the European form, and owing to the confusion which has arisen from a lack of accurate figures and descrip- tions by the older authors. Famity ERYTHRAEIDAE Oudemans. Gen. Hrythraeus Latr. 1806. Erythraeus namaquensis n. sp. Type, 1 specimen collected by Mr. C. W. Thorne and Dr. A. J. Hesse at Kamieskroon, Namaqualand (November 1936). Colour.—Body in general light reddish-brown, dark reddish or wine- red in living specimens. The feathery tuft of hairs on the tibia of the fourth leg blackish, individual hairs seen under the microscope yellow- brown, the smaller tuft of hairs on the dorsal apex of the patella entirely white, contrasting strongly with those of the tibia. 274 Annals of the South African Museum. Body.—Body longer than broad, its anterior and posterior margins truncate, widest between the anterior and posterior pairs of legs. General body surface on the dorsum and sides with small flattened pear-shaped hairs, ventral surface between the legs with normal slender and pointed hairs. The pear-shaped hairs of the dorsal surface seen under high power of the microscope (fig. 2, a) leaf-like, with spicules on their dorsal surface and a main central darker axis which is seen better when focussing a little downwards from the dorsal surface; under surface entirely smooth; these modified hairs usually curved, sometimes strongly so, the convexity directed upwards. Hairs of dorsal surface very numerous, not arranged in rows but evenly spaced, 40-60 across the body, giving the skin a granulose appearance. Two sessile eyes on each side, posterior to the middle of the crista, anterior eye twice as large as posterior one; posterior enlargement of crista with a pair of fine hairs of insignificant size, the anterior enlargement with a row of 4—6 flattened and spiculated hairs on each side, the proximal ones broad and short, the distal ones narrower and longer; in addition a pair of fine smooth hairs in the middle, anterior to the lateral flattened hairs. Mouthparts.—Palpi as in fig. 2, b, the second joint with some enlarged and flattened serrated hairs along its dorsal surface. Maxillary lobes (fig. 2, b, c) more or less immovably attached to each other along the middle line; they can, however, be easily separated. Hach internal lobe ending anteriorly in a corolla-like suctorial organ composed of two or more concentrically arranged rings of transparent, membranous, modified hair-like structures resembling the petals of a flower (fig. 2, c). Kach circle of hairs entire except for an opening in its inner side which is continuous with a straight groove passing down the inner side of the maxillary lobe, in which the mandible lies. Seen from above in an undissected specimen, the closely contiguous rami of the mandibles project slightly from the suctorial organ, the greater part of which, while enclosing them, is situated laterally to them. Mandibles styliform, seen from below (fig. 2, d), showing a portion of the two main tracheal branches piercing the muscles attaching the two halves of the mandible. Legs.—All legs 7-jointed, I and II subequal, III longer than I and II, IV much longer than III. Coxa I largest of the four, its posterior apex forming an angle of 45°, the inner margins of the opposing coxae vertical and parallel. All legs covered evenly with hairs, those of the proximal segments short, fairly broad, blunt, and with distinct spicules, those of the distal segments longer, narrower, sharply . A New Species of Plume-footed Mite from South Africa. 275 pointed, and with indistinct spicules. All these hairs with a median groove. eC SOT (| (6) pedipalp and maxillary lobe; (c) apex of maxillary lobe, enlarged; (d) mandibles; (e) tibia and tarsus, leg III; (f) modified hair, ventral surface of tibia IV; (g) the same, dorsal surface of tibia IV. Tibiae of legs I-III on their ventral surfaces with thickened, modi- fied, spine-like hairs, fleshy pink in colour and provided with fine accessory filaments (fig. 2,e); these hairs larger and more distinct in 276 Annals of the South African Museum. II than in I, and more so in III than in II, I with 12-14 on the distal two-thirds of its ventral surface, II similarly with 14-16, III with 16-18 (fig. 2,e). All tarsi, especially the first, swollen and short, I about two-thirds, II half, III less than half as long as the preceding tibial segment; tarsi I-III with a pad of numerous short sinuous hairs on their ventral surfaces, these hairs provided with lateral filaments longer than those of the remaining hairs of the legs. Claws of all tarsi small and simple. Leg IV as in fig. 3, about 34 times as long as body, tibia very long and slender, provided along its entire length, except for a small basal Fie. 3.—Hrythraeus namaquensis n. sp. Patella, tibia, and tarsus of leg IV. portion, with modified hairs; these hairs much more numerous in the basal than in the apical half, extremely flattened and leaf-like, semi-transparent, and thickly covered with fine setae. The modified hairs of the dorsal surface shorter, broader, and not so sharply pointed (fig. 2, g) as those of the ventral surface (fig. 2,f). Tarsus (fig. 4, b) incrassate, flattened, and sharply truncate apically, many times shorter than the tibia, with a few more or less flattened hairs on its dorsal surface. Dorsal apex of patella with a small cluster of snow-white hairs, similar in shape but a little shorter than those on the dorsal surface of the tibia (fig. 4, a). Genital Opening.—In the smaller of the two specimens the genital Opening is opposite the junction of coxae III and IV; it consists of an oval longitudinal area enclosing a pair of smaller, dark, chitinous plates, the whole area about as long as the distance between the first pair of coxae. Anal opening situated considerably posterior to coxa IV and not far from the posterior margin of the body. It is also oval in shape and considerably smaller than the genital opening. A New Species of Plume-footed Mite from South Africa. 277 Dimensions.—Length of body 1-6, breadth 1-2, leg III 2-2, leg IV 6mm. Dr. Hesse and Mr. Thorne also collected a second larger specimen at Knegsvlakte, Bitterfontein, Namaqualand. This specimen is about 2mm. in length; legs, [ 2-7, If 2-6; III 3, 1V 7-5 mm. Erythraeus namaquensis seems to resemble most closely the form described by Tragardh under the name of Lucasiella plumipes from EAA {LL se. Fic. 4.—Erythraeus namaquensis n. sp. (a) apex of patella and base of tibia, leg IV; (6) apex of tibia, and tarsus, leg IV. the neighbourhood of Cairo (6). Both forms have spines on the under surface of tibia I and II; these are, however, also present on the femora of legs II and III in Tragardh’s specimens, while absent in the Namaqualand ones. The two forms also agree in the general shape and structure of the tibial hairs of leg IV, which Tragardh describes as “mit schwarzbraunen, langen, lanzettf6rmigen, fein behaarten Haaren besetzt.’”? Though the differences between the two forms are very small, it seems probable that minor distinctions 278 Annals of the South African Museum. will be found to exist when specimens from the two localities can be compared. Tragardh’s form is again undoubtedly co-specific with Lucas’s types and probably also with the specimens taken by various collectors in the Mediterranean region (Malaga and Corfu). The appearance of the fourth leg in #. namaquensis is, however, rather different from that of Lucas’s specimens judging by the figure given by André (1, p. 382, P.iv). The modified hairs in this figure are longer, narrower, and more sinuous, there is no tuft of hairs at the apex of the patella, and the tarsus is quite devoid of flattened hairs. Still more apparent is the difference between the South African and Swiss forms figured by Haller and reproduced by André (1, p. 382, 1). Apart from the improbability of identical forms of Erythraeus occurring in such widely separated and climatically different regions as Namaqualand and Switzerland, the South African species is obviously distinct from the Swiss form, if Haller’s figure has any semblance of accuracy. The more obvious differences are that in the latter the dorsal hairs of the body are much fewer and larger; the pair of hairs on the posterior enlargement of the crista are far longer and project beyond the sides of the body (in #. namaquensis they are minute); the greater size of the mouthparts. Haller’s specimens are also probably larger in body size, and the hairs of the tibial tuft of leg IV appear to be longer and different in shape. I have not been able to see Birula’s description and figures of his Macropus plumifer, which according to André is also synonymous with #. plumipes. It would seem that all forms recently grouped by André (1, 2, 8) under the name of Erythraeus plumipes L. Koch, and including the one described in this paper, are very closely allied. They all have the same flattened papillate hairs on the dorsal surface of the body, and a cluster of peculiarly modified hairs on the tibia of the fourth leg. It is, however, quite possible that, when actual specimens from the various regions where they have been collected are compared, the South-Russian, Swiss, Mediterranean, and Namaqualand forms may be found to be distinct. They will, however, have to be separated on other characters and for different reasons than those given by N. Banks (10); the differences may be so slight that no more than a recognition of varieties or local races of the one species will be justified. The characters which appear to be most convenient for establishing such differences are (1) the number of dorsal papillate hairs, (2) the mouth- parts, (3) the size and shape of the modified hairs of the fourth tibia. A New Species of Plume-footed Mite from South Africa. 279 The features which distinguish Hrythraeus namaquensis from other members of the plumipes group are (1) the mouthparts, (2) the apical tuft of hairs on the patella of leg IV, (3) the shape and size of the flattened hairs of tibia IV. I wish to tender my best thanks to Dr. Hesse and Mr. Thorne for sending me these interesting mites and for the observations they have made. REFERENCES. (1) AnpR&, M. Bull. Mus. nat. hist. nat., vol. xxxiii, p. 380, text-figs. 1-3, 1927. (2) AnpRE, M. Bull. Mus. nat. hist. nat., 2° ser., Tome III, p. 351, text-figs. 1-3, 1931. (3) AnpRE, M. Assoc. Fr. Avance. des Sciences, Nancy, p. 457, 1 text-fig., 1931. (4) Hatter, G. Beitr. Kennt. Schweiz. Milbenfauna. Vier. Jahrs. Naturf. Gesell. Zurich, vol. xxx, p. 82, pl. i. fig. 2, 1885. , (5) GzorGze,G.F. Science Gossip, vol. ii, p. 153, 1896. (6) TracaRrpH, I. Res. Swed. Zool. Exped. Egypt and White Nile. N. 20, p. 68, 1905. (7) Brruta, A. Horae Soc. Entom. Rossicae, vol. xxvii, p. 387, pl. vii, figs. 1-9, 1893. (8) CamBRIDGE, O. P. P.Z.S. London, p. 939, pl. lv, 1897. (9) Lucas, H. Ann. Soc. Ent. Fr., 4 ser., iv, p. 206, 1864. (10) Banks, N. The Canadian Entomologist, vol. xxxii, N. 2, p. 32, 1900. (11) Hewirt, J. Ann. Transvaal Museum, vol. vii, pt. 1, pp. 1-76, pl. i—viii, 1919. (12) Lawrence, R. F. Ann. S. Afr. Museum, vol. xxv, pt. 2, pp. 217-312, 1928. VOheekextls PART 3. 20 iY Lt " ¥ Ae 4 m "i Pik Seis noes “4 GOS soe ot 0 De ay IM <\ } Ne i ren , f , meee? 2 CSCS no re eg) dss ae A SAE ERY ORE chs eT } tbe wads f 4a) Pad era) Bids) its i Aibde) dy ae tae i aN nee) .! ; Ae ; ive 2 f : ; ier bins sano vith Ci ce Oe (itp. hn f ‘ 5 ~ 1 ” f i. ps e - | | 4 x a > e : . ~ ; ; ‘ ‘ [= ~ * . PARTS OF THE ANNALS: ge ISSUED Reh Vol. I.—Part 1, out of print; Part 2, out of print; Part 3, out se fe Vol, I1.—Part 1, out of print; Part 2, 5/6; Part 3, out of print; Part 4 Part 5, 1/6; Part 6, 3/-;. Part 7, 1/6; Part 8, 3/—s Fart 2 Part 10, 7/-; Part ll, 3/-; Index, "Title, etc., 1/6. ‘v2 Vol. I1I1.—Part 1, out of print; Part 2, 1/6; Part 3, 5/6; Part 4, 3/3 5/6; Part 6, 7/-; Part 7, 1/6; Part 8, 3/-; Part 9, 1/6; 3 Title, etc., 1/6. : Vol. IV (containing ’Palacontological papers published in conjunction ke Geological Survey).— - Part 1, 11/-; Part 2, 7/-; Part 3, 4/6; Part 4, 416; Part 5 Part 6, 4/6; Part 7, 14/—; Part 8, 8/-. | Vol. V.—Part L, 4/6; Part °2, 8/6; Part 3 2/6; Part 4, 1/6; Part 5, ‘9/3 Part 6, 5/-; Part 7, 3/-; Part 8, 4/6; Part 9, 4/6; ‘Index, Title, etc., 1/6. ee Vol. VI. Pars 1, 13/6; Part 2, 4/6;° Part 3, 3/6; Part 4, 30/-; Index, Title, ee etc., 1/6. tee Vol. VII (containing Palaeontological papers published in conjtian ee with the wees Geological Survey).— eee Part 1, 3/-; Part 2, out of print; Part 3, 5/-; Part 4, 8/3 ‘Part 6, SE oe 5/6; Part 6, 1/6; Index, Title, etc., 1/6. ee _ Vol. VIII.—Complete, out of print. Index, Title, etc., 1/6. 2 = Vol. 1X.—Part_l, 4/6; Part 2, 5/6; Part 3, °10/—-; Part 4, 6/6: Part 5, 3165 Part 6, ll/—; Part.7, 9/-; Index, Title, et¢.,.1/6, cf ~~ Vol. X.—Part 1, 3/-; Part 2, 2/6; Part 3, 2/-; Part 4, 3/-; Part 5, 20/-3 ay Part 6, 3/-; Part 7, 10/-; Part 8, 2/6; Part 9, 5/-; Part 10, 2/65 Part 11, 20/-; Part 12, 7/|-. Complete. Title, etc, contained ine : Part 11. Spee Mea Vol. XI.—Part 1, 3/6; Part 2, 2h; oe 3, 13/6; Part 4, 1/6; Part 5, Tf- = Part 6, 11/-; Index, Title, ete., and Plate ITI, 3/-. i = Vol. XII (containing Palaeontological Papers published in conjunetion with ‘the = Geological Survey).— Part I, 15/6; Part 2,-3/6; Part 3, 4/6; Part 4, 3/-; Part 5, a5 - Part 6, 6/-;. Part 7, 20/-; Part 8, 20/-; Index, Title, ete., 1/6.5; = : Vol. XIII.—Part L, 6/-;. Part 2, 2/6; Part 3; 3/-; Part 4, 8/6; Part os 1/85 Part 6, 5/-; Part 7, 30/—; Part 8, 1/-; Index, Title, ete., 6s : Vol. XIV.—Part 1, 8/6; Part 2, 8/-; Part 3, 6/-; Part 4, 21/-; Part Es ae Part 6, 9/—; Index, Title, etc., 1/6. —- : Vol. XV.—Part 1, 17/-; Part 2, 17/-; Part 3,14/-; Part 4, 12/6; Part By 5103. Part 6, 3/6; Index, Title, ete., 1/6. es Vol. XVI.—Part 1, 30/6; Part 2, 4/-; Part 3 (with Title, inded: etc.), 25. See ee Vol. XVII.—Part 1, 12/—;_ Part 2, 9/6; Part 3, 3/-; Part 4, L/-5 Part. 5, ce os Part 6, 2/6; Index, Title, etc., 1/6. Vol. XVIII.—Part 1, 20/-; Part 2, 7/6; Part 3, 30/-; Part 4, 12/6; Sadex Title, etc., 1/6. EE Vol. XIX.—Part I, 22/-; Part 2, 17/6; Part 3, 11/-; Part 4, 5/6; Index, Title, : etc., 1/6. se Vol. XX.—Part 1, 8/6; Part 2, 12/6; Part 3, 4/-; Part 4, 10/-; Part 5, sos Bee Part 6 (with Title, etc.), 4/6. : pele Vol. XXI.—Part 1, 25/+; Part 2 (with Title, etc.), 30/-. : eae Vol. XXIT (containing Palaeontological papers published in coujnntebtn: with the s = 3 Geological Survey).— es a Part 1, 20/-; Part 2, 10/—; Part 3 (with Title, etc.), 3/6. ae vee Vol. X XIIT.—Part 1, 12/6; Part 2, 8/-; Part 3 (with Index, Title, etc.), ao as Vol. XXIV.—Part i: 10/-; Part 2, 2/6; Part 3, 4/6; Part 4, 11/-. a aes ~ - Vol. XXV.—Part 1, 12/6; Part 2, 2 Part 3 (with Index, Eye ete. )» 10 Pee eee Vol. XXVI. —Complete, 25/-. . . Vol. XX VII.—Complete, 25/-. eee Vol. XXVIII (containing Palaeontological papers published in conjunction. valle the ee Geological Survey).— ee Part 1, 10/-;. Part 2, 1/6; Part 3, 6/—; Part 4 (with Index, So etc.), 10/-. ; . Vol. XXITX.—Part 1, 20/-; Part 2 (arith Index, Title, etc.), 20/-. ae Vol. XXX.—Part 1, 15/6; Part 2; 20/-; Part 3, 8/-; Part 4, 9/-; Part 5, $e acaba Sagar Vol. XX XI (containing Palaeontological papers ‘published in conjaueues ‘with the kb? Fs Geological Survey).— 23S ieee eae Part 1, 20/-; Part 2. 13/6; Part 3, 17/6. Vol. XX XII.—Part 1, 2/6; Part 2, 7/6; Part 3, 17j-. The Annals of the South African Museum will be issued at irregular intervals, SS as maiter for ss eae is available. | Ss t P Copies may be obtained from— Messrs. WHELDON & WESLEY, Lrp., Sy tae 2, 3, and 4 ArtHuR STREET, New Oxrorp Saas oe W.C. 2; or, BS The LIBRARIAN, Souta Arrican Musrum, Care Town.. poate Moss ANN ALGO ANTS 29 k SS Wa ry 7) > SAQA Lat > a OF THE meee A eae ~SpOUTH AFRICAN MUSEUM VOLUME XXXII, PART IV, containing :— 16. A new Genus and new Species and Subspecies of Mammals from Little Namaqualand and the North-West Cape Province ; and a new Subspecies of Gerbillus paeba from the Eastern Cape Provinces— By G. C. Snorrripvce, Director, Kaffrarian Mugeum, King William’s Town. (In collabora- tion with DonaLp Carrer, Assistant Curator of Mammals American Museum of Natural History.) ISSUHD JULY 1938. PRICK 1s. 4 . PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN BY NEILL AND CO., LTD., 212 CAUSEWAYSIDE, EDINBURGH. A 4) ‘ > 2 ( 281 ) 16. A new Genus and new Species and Subspecies of Mammals from Inttle Namaqualand and the North-West Cape Province ; and a new Subspecies of Gerbillus paeba from the Eastern Cape Province—By G. C. SHoRTRIDGE, Director, Kaffrarian Museum, King William’s Town. (In collaboration with DonaLD Carter, Assistant Curator of Mammals, American Museum of Natural History.) THis is the first report on the Cape Museums’ Mammal Survey of the Cape Province. Owing to generous financial co-operation on. the parts of the Museum of Comparative Zoology (at Harvard University), the South African Museum, the Natal Museum, the Cape Provincial Museums (at Grahamstown, Kimberley, Port Elizabeth, King William’s Town, and Hast London), the Research Grant Board of the Union of South Africa, and Dr. H. Merensky, it has been possible to undertake, by means of two expeditions, a systematic mammal survey of (1) the whole of Little Namaqualand, and (2) the northern part of the Western Cape Province. The preliminary scientific results of this mammal survey of the Cape Province include the determination of a new genus of Golden Moles, and the discovery of no less than three genera of mammals, Platymops, Gliriscus, and Petromyscus, that were not known previously to occur in the Cape Province. An unexpected discovery is that of an alpine form of Bathyergus, a genus hitherto thought to be restricted to coastal sand-dunes, on the highest plateau of the Kamiesberg. Besides the new forms herein described, further additions to the Cape Mammal Fauna include two South-West African Bats, Cistugo seabrae and typical Platymops haagneri; whilst the re-discovery in the Cape Province of the rare Eptesicus megalurus, after nearly a hundred years, is of considerable zoological interest. No Mammal Survey of Little Namaqualand had been carried out since the Rudd Expedition in 1903, when Mr. C. H. B. Grant collected in the region between Port Nolloth and Klipfontein in the north- western part of that territory. Memorable results of the Rudd exploration included the discoveries of Bathyerqus janetta and Herpestes ruddi, both of which were figured VOL. XXXII, PART 4. 21 = ope QG 1938 YAWN \ 282 ‘Annals of the South African Museum. in the Proceedings of the Zoological Society of London (1904, vol. i, pl. vi). The number of mammals already collected totals just over 3700 specimens, and full reports on the results of the expeditions, together with field notes and short topographical details of the camping- places, will be published in due course. Fam. VESPERTILIONIDAE. Scotophilus angusticeps sp. n. Two specimens 99 from near Citrusdal. General colour above drabby umber-brown (rather as in EHptesicus capensis), without gloss or sheen; below light drabby brown, centre of abdominal and anal regions strongly washed with rufous, throat tinged with sienna-brown; base of fur, above and below, dark seal- brown; ears and flying membranes dusky black. Type (in the Kafirarian Museum).—Adult female, original number 1937, collected 28th December 1937 at Hex River Estate, 10 miles north of Citrusdal, N.W. Cape Province. Co-type.—Adult female, original number 3308, collected 3rd January 1938. . Dimensions of Type and Co-type.—H. and b., 77 (74); tl., 58 (57); hf., 12 (11-5); ear, 18 (17-5); forearm, 53 mm. Skulls —Total length, 20; basal length, 17 (17-2); zygomatic width, 13:5; width of brain case, 10 (9-5); interorbital constriction, 4-5; length of palate, 9; upper dental series (from front of canine), 7 (7-5); lower dental series (from front of canine), 7-5 (8); mandible, 16 (15-5); height of brain case, 7-5 mm. Skull flatter, less massive and with a markedly narrower zygomatic width than in S. nigrita dingaani or S. n. herero; cranially S. n. herero is in some respects intermediate between S. n. dingaani and S. angusticeps, although much more closely resembling the former. Fam. MOLossIDAE. Platymops (Sauromys) haagneri umbratus subsp. n. One specimen from Kliphuis, northern spur of the Cedarberg. A Cape subspecies of P. haagneri, distinguished by its dark coloration. Colour above smoky seal-brown, a shade paler than in Nyctonomous bocagei; under surface smoky-brown, several shades lighter than the Mammals from Litile Namaqualand and N -W. Cape Province. 283 upper surface; ears and flying membranes almost black—as opposed to comparatively pale horn-brown in typical haagnerv. In typical haagneri (series from Goodhouse, lower Orange River, examined, and a single specimen from Berseba, Great Namaqualand) the colour above is drabby brown, paler and less cinereous than the under surface of umbratus, and pale buffy, approaching whitish, below. Type (in the Kafirarian Museum).—Adult male, original number 2827, collected 7th December 1937 at Kliphuis, Pakhuis Pass, 11 miles N.E. of Clanwilliam, N.W. Cape Province. Dimensions of Type.—H. and b., 61; tl., 31; hf., 8-5; ear, 15; forearm, 37 mm. Skull.—Basal length, 16-2; zygomatic width, 10; width of brain case, 9; interorbital constriction, 3-7; length of palate, 5-5; upper dental series (from front of canine), 6; mandible, 12-5 mm. Genus new to the Cape Province. Fam. MACROSCELIDIDAE. P Macroscelides typicus isabellinus subsp. n. One specimen from Port Nolloth (near township). A markedly pale form of Macroscelides typicus. Upper parts pale ashy-buff with only the faintest peppering caused by dusky tips to the hairs; underparts white; tail buffy, bristles on terminal half black above and below; hands and feet pale buff; ears blackish, as in the other local race of typicus. Type (in the Kaffrarian Museum).—Adult male, original number 1821, collected 3rd February 1937 at Port Nolloth (close to the sea). Dimensions of Type.—H. and b., 114; tl., 122; hf. (s.u.), 36-5; ear, 25 mm. Skull.—Greatest median length, 33; basal length, 27-5; zygomatic width, 20-5; width of brain case, 15:5; interorbital constriction, 6; width across inflations, 18; length of nasals, 11; upper dental series (including incisors), 16; lower dental series (including incisors), 15; mandible, 25 mm. I am regarding this specimen as representing a desert subspecies, like the pallid coastal races of other mammals peculiar to the South- West African Namib. Four other specimens from the Port Nolloth hinteriand (only 10-15 miles inland) are indistinguishable from specimens provision- ally referred to P. t. typicus from other parts of Little Namaqualand. 284 Annals of the South African Museum. Fam. CHRYSOCHLORIDAE. Cryptochloris gen. nov. Type Cryptochloris zyli sp. n. Size and shape of body foreshortened as in Hremitalpa, but with relatively short, iridescent fur. Claws of forefoot well developed and equal respectively in size to those of Hremitalpa. Skull, with 40 teeth in all, approaching that of Chrysochloris, but interorbital region more expanded. Cryptochloris zyli sp. n. A small rather dark coloured Golden Mole, similar in form to Eremitalpa granti, but with relatively short fur, as in Chrysochloris. One specimen from near Lamberts Bay. General colour above drabby lead colour washed with an inky violet iridescence. Underparts drabby lead colour, hardly paler than above. Base of fur dark slate throughout. A slight admixture of pale hairs between the whitish-buff face markings—which are well defined, as in Chrysochloris asiatica. Type (in the Kaffrarian Museum).—Adult male, original number 3477, collected 13th January 1937 at Compagnies Drift, 10 miles inland from Lamberts Bay, North-Western Cape Province. Dimensions of Type.—H. and b., 82; hf., 12 mm. Skull.—Greatest length, 22; basal length, 17-5; greatest breadth, 15-5; interorbital breadth, 8-5; greatest height, 11; dental series (front of incisors to back of posterior molar), 10; palate (across posterior molars), 8 mm. Chrysochloris wintont Broom should apparently be assigned to this genus. I have named this very distinct Golden Mole, the type of a new genus, in honour of Mr. Gideon van Zyl, Compagnies Drift, to whom I am indebted for much assistance during my visit to Lamberts Bay. Chrysochloris concolor sp. n. Two specimens g@ from Nieuwoudtville and Traveller’s Rest (15 miles N.E. of Clanwilliam). A silvery drab Golden Mole, much resembling pale individuals in a series of C. namaquensis from the Kamiesberg, but without any - trace of greenish or violet reflections. Mammals from Litile Namaqualand and N.-W. Cape Province. 285 Upper parts pale drab with a satiny sheen; underparts pale drab washed with greyish white; base of fur dusky slate throughout; whitish buff cheek markings like those of C. asiatica, but with a more profuse admixture of whitish hairs above snout. Type (in the Kaffrarian Museum).—Adult male, original number 2500, collected 10th November 1937, 3 miles west of Nieuwoudtville, N.W. Cape Province. Dimensions of Type.—H. and b., 110; hf., 13 mm. Skull.—Greatest length, 23-5; basal length, 18; greatest breadth, 17-5; greatest height, 12; interorbital breadth, 8-5; palate (across posterior molars), 9; dental series (front of incisors to back of posterior molar), 10:5 mm. Fam. PROTELIDAE Proteles cristatus canescens subsp. n. Specimens from Hselfontein, Witwater, Henriet, and Port Nolloth (Little Namaqualand); Kliphuis and Klaver (N.W. Cape Province); and from Oas and Sandfontein (South-West Africa—Gobabis Dist.). A long-haired, uniformly dark grey race of P. cristatus, heavily grizzled throughout. General colour above iron-grey, individual hairs strongly annulated; flanks similar in colour to dorsal crest, which tends to become tipped with black on the rump only; black bands on flanks relatively ill defined, owing to a profuse overlapping of long grizzled-grey hairs; buttocks faintly washed with brownish rufous; head dusky, speckled with buff; blackish region above almost bare muzzle three times as wide as in typical cristatus (specimens from Kafiraria). Underparts buffy, tinged with rufous; tail, individual hairs grey at base, tipped with black; end of tail heavily blackened; forefeet black above, the black extending outwardly 6 inches along the forearm; hindfeet black above, thighs mottled (not striped) with black; ears, outside black, narrowly but conspicuously edged with buff, inside whitish. Type (in the Kaffrarian Museum).—Adult male, original number 1048, collected 11th December 1936 at Eselfontein (Kamiesberg), Little Namaqualand. Dimensions of Type.—H. and b., 700; tl., 270; hf. 154; ear, 103 mm. Skull.—Total length, 136-5; basal length, 130-5; palate, 71-5; nasals, 37; zygomatic width, 85; width of brain case, 48; greatest length of mandible, 101 mm. 286 Annals of the South African Museum. In a specimen from Witwater (Kamiesberg) with worn fur and only scattered long hairs on the flanks the side stripes contrast almost as in typical cristatus, and the dorsal crest is tipped with black throughout its length. In two specimens, from Port Nolloth and Eenriet, the hairs of the dorsal crest are everywhere uniform grizzled grey like the flanks. In two newly born puppies, from Sandfontein and Klaver, the dorsal lines and tails are entirely black, all individual hairs being black from the base. The approximate range of P. c. canescens appears to be the Western Cape Province and South-West Africa—to at least as far north as the Tropic of Capricorn. Two specimens from the Kaokoveld (the extreme north-west of S.W. Africa), referred to P. c. harrisoni of Angola, approach typical cristatus in general coloration, the flanks being rich buffy, contrasting with the grizzled dorsal crest. Fam. MUSCARDINIDAE. Gliriscus rupicola australis subsp. n. Two specimens $Q from Kenriet; a third specimen (in alc.) since received from Port Nolloth. A southern subspecies of G. rupicola with a uniformly whitish tail— below as well as above. General coloration above a shade paler than in typical rupicola, this being especially marked on the forehead; under surface also paler, the ends of the hairs being more profusely white; tail paler throughout, the individual hairs, below as well as above, more uniformly tipped with white than in either of the two South-West African subspecies; hands and feet white; colour of ears and ocular markings as in typical rupicola and r. montosus. Type (in the Kaffrarian Museum).—Adult female, original number 1522, collected 19th January 1937 at Eenriet, Little Namaqualand. Co-type.—Adult male, original number 1320, collected 12th January 1937. Dimensions of Type—H. and b., 115; tl., 86 (105 in co-type) hf(su); 22> ear, 19 am, Skull.—Total length, 32; basal length, 24; zygomatic width, 17-5; width of brain case, 15; interorbital constriction, 5; length of palate, 9-5; upper dental series, 4; lower dental series, 3-7; mandible (excluding incisors), 15 mm. Mammals from Little Namaqualand and N.-W. Cape Province. 287 G. r. montosus from Brukaros Mountain, Great Namaqualand, the darkest of the three races, with drabby feet, separates geographically this slightly larger-eared Cape form from G. r. rupicola of Damaraland. Genus new to the Cape Province. Fam. MuRIDAR. Taterona brantst namaquensis subsp. n. Three specimens 3g 2 from Goodhouse, south bank of the lower Orange River. A desert-coloured race of Taterona brantsi. T. b. namaquensis from the north of Little Namaqualand agrees closely in colour with T. 6. perpallida from the Kalahari and Ngami- land, but the pale, crescentic, somewhat shadowy ocular markings are almost if not entirely absent. The uniform coloration of the head thus serves to distinguish namaquensis from the geographically distant perpallida. General colour above pale sandy buff, slightly grizzled, below white; tail pale speckled sandy buff above, white below; in two out of three specimens (including the type) the terminal third of the tail is white above; ears sandy buff, not differing in shade from the head; hands and feet buff-white. Type (in the Kafirarian Museum).—Adult female, original number 1088, collected 23rd December 1936 at Goodhouse (Raman’s Drift), Lower Orange River. Dimensions of Type.—H. and b., 126; tl. 167; hf. (s.u.), 34; ear, 21 mm. Skull.—Total length, 36; basal length, 28; length of palate, 15; zygomatic width, 18-5; length of nasals, 15; upper molar and premolar series, 5-5; lower molar and premolar series, 6; mandible (exclusive of incisors), 20 mm. The range of T. b. namaquensis is separated from that of T. 0b. perpallida by the whole of Great Namaqualand, in which no species of Taterona has yet been discovered. Poemys melanotis insignis subsp. n. Specimens from Hselfontein. A small, silvery grey, narrowly striped form of P. melanotis without any trace of ocular markings, but with a dark frontal patch, almost as well defined as in P. nigrifrons. 288 Annals of the South African Museum. General colour above pale drab grey, closely approximating in shade to that of P. m. capensis (typically from Wolseley, C.P.): black dorsal stripe narrow, extending as far forward as in a specimen I refer to P. m. melanotis from the Pirie Forest; a conspicuous dusky black frontal patch; ocular markings obsolete; underparts drabby white; ears dusky, not so dark as in typical melanotis; tail drab grey above and below; hands and feet white. Type.—Adult female, original number 843, collected 27th November 1936 at Eselfontein, Kamiesberg. Dimensions of Type.—H. and b., 70; tl., 75; hf. (s.u.), 16-5; ear, 17-5 mm. Skull.—Greatest length, 20-5; basilar length, 16; zygomatic width, 10:5; width of brain case, 9:5; interorbital constriction, 3-5; length of upper molars, 3-3 mm. This Namaqualand form of Poemys melanotis may prove to be restricted in range to the Kamiesberg. Petromyscus barbourt sp. n. Specimens from Witwater, Platbakkies, Eselfontein, and Henriet. A small drab grey Petromyscus, at once distinguishable from all other members of the genus by its bicoloured tail. General colour above grizzled drab grey (about as in Mus musculus) ; underparts white; tail markedly bicoloured, drab grey above, white below; hands and feet white; ears drab grey. Mammae 4 (ing.); no pectoral mammae. Type (in the Kafirarian Museum).—Adult male, original number ~ 253, collected 20th October 1936 at Witwater (Kamiesberg), Little Namaqualand. 3 Co-type. Adult female, original number 146, collected 11th October 1936. Dimensions of Type.—H. and b., 78; tl., 80; hf. (s.u.), 19; ear, 14 mm. Skull.—Total length, 23; basal length, 17-3; zygomatic width, 11 approximately (of co-type 11); width of brain case, 11; inter- orbital constriction, 4; length of palate, 9-2; upper molar and pre- molar series, 3-5; lower molar and premolar series, 3:5; mandible (exclusive of incisors), 11-5 mm. This rather short-tailed species of Petromyscus, the most southern representative of the genus, approaches the relatively long-tailed P. shortridgei, the largest and most northern representative, most Mammals from Little Namaqualand and N.-W. Cape Province. 289 closely in colour; but it is drabby cinereous instead of umber brown. It may further be noted that in P. shortridge: the two pectoral mamme are also usually but not invariably absent, whereas they are present in all females of the P. collinus group collected, as well as in the only known female of P. monticularis. P. monticularis, which occurs in association with P. collinus bruchus on Brukaros Mountain (Great Namaqualand), may be distinguished from all of the other species by its short ears and relatively thick tail. Genus new to the Cape Province. The genus Petromyscus and all of its previously known species and subspecies were discovered by myself in South West Africa.* To these may now be added Petromyscus barbouri and P. collinus capensis, both from Little Namaqualand. I am naming the above strikingly distinct new species in honour of Dr. Thomas Barbour, Director of the Museum of Comparative Zoology, Harvard University, who is so generously contributing towards and associating himself with the Cape Museums Mammal Survey of the Cape Province. Petromyscus collinus capensis subsp. n. Two specimens $3 from Goodhouse, northern Little Namaqualand. A brownish-chestnut form of P. collinus, approaching the typical Damaraland subspecies in colour, but separated from it by the darker P. collinus bruchus of Great Namaqualand. Characterised by a markedly hairy tail, the individual bristles being twice the length of those in collinus and bruchus. General colour above brownish chestnut; underparts white; tail pale drabby, slightly darker above and still more so terminally (above and below) owing to a thick pencilling of relatively long dusky bristles; ears dusky brown, not tinged with rufous as in collinus and bruchus; hands and feet buffy white. Type (in the Kafirarian Museum).—Original number 1223, collected 5th January 1937. at Goodhouse (Raman’s Drift), lower Orange River—south bank. Dimensions of Type and Co-type.—H. and b., 87 (90); tl., 99 (98); maeiseue). 17 (17); ear, 16-5 (17) mm. Skulls —Total length, 26-5 (26); basal length, 20 (19); zygomatic * P. shortridgei, typically from the Kaokoveld, has since been found to extend into 8.-W. Angola. 290 Annals of the South African Museum. ~ width, 12 (12-5); width of brain case, 13-5 (12); interorbital con- striction, 4-5; length of palate, 11-5 (10-5); upper molar and premolar series, 3-5; lower molar and premolar series, 3-5; mandible (exclusive of incisors), 13 mm. The Orange River Valley presumably constitutes the southern limit of the range of P. collinus and its subspecies, the genus being represented farther south in Little Namaqualand by P. barbour. Fam. BATHYERGIDAE. Bathyergus janetta inselbergensis subsp. n. Specimens from Hselfontein, Kamiesberg. This mountain race of Bathyergus janetta differs from the typical coastal subspecies in the general colour above being silvery buff instead of drab grey, and thus the seal-brown dorsal and occipital area, together with the dark forearms and thighs, contrasts much more sharply with the flanks, which have an almost frosty appearance. The tail hairs (above) are slaty black basally instead of pale brown, and bufi-white terminally; the hairs on the forefeet are whiter, less soiled looking, than in typical janetta. Type (in the Kafirarian Museum).—Adult female, original number 977, collected 4th December 1936 at Hselfontein, Kamiesberg (altitude 4400 ft. approx.). Dimensions of Type.—H. and b., 230; tl., 44; hf., 45 mm. Of a large male (without skull), h. and b., 280; tl., 52; hf., 51 mm. Skull.—Total length, 50-5; basal length, 44; zygomatic width, 33; width of brain case, 19-5; interorbital constriction, 9; upper molar and premolar series, 10; lower molar and premolar series, 10-5; mandible (excluding incisors), 39-5 mm. The discovery of Bathyergus in the Kamiesbergen at an altitude of between 4000 and 5000 ft. was unexpectedly interesting, previously known forms being only known to occur in low-lying coastal sand- plains. Fam. Muripas. Gerbillus paeba exilis subsp. n. Three specimens $¢? from Alexandria District (coast), Cape Province. A remarkably pallid race of G. paeba with fawn-white flanks, white ocular rings and a white tail tip. General colour above pale rufous fawn; flanks still paler, shading Mammals from Little Namaqualand and N.-W. Cape Province 291 to white; cheeks and ring above eye white; underparts white; tail very pale fawn above, white below, tail tuft white; hands and feet white; ears pale fawn. In the type and co-type there is no trace of dusky tips to the hairs anywhere, but in a third specimen (original No. 102X—in the Albany Museum) the dorsal region has a smoky appearance due to a slight darkening of the ends of the hairs. Type (in the Albany Museum).—Adult male, original number 100X, collected 27th May 1937 at Paardevlei, Alexandria Dist. (coast), by F. and W. Pannell. . Co-type (in the Kaffrarian Museum, presented by the Albany Museum).—Adult female, original number 101X, collected 13th March 1934 at Sundays River Mouth, Alexandria Dist., by O. West. Dimensions of Type and Co-type.—H. and b., 84 (80); tl., 108 (-); hf. (c.u.) 27 (26) ; ear, 15 (12%) mm. Skulls.—Total length, 29-5 (30); basal length, 21 (23); length of palate, 12 (11:5); zygomatic width, 13-5 (—); interorbital constriction, 5 (5-5); width of brain case, 13 (13-5); length of nasals, 11-5 (12); upper molar series, 4:5; lower molar series, 4:5; mandible, 14-5 (15) mm. This eastern subspecies of G. paeba, from a region far from where any other form of Gerbillus is known to occur, approaches most closely in colour some bright orange-rufous specimens in a series of G. p. brooms from Little Namaqualand, but it may at once be distinguished by its almost white flanks, and by the white on the cheeks extending in a broad ring round the eye. Dr. J. Hewitt, Director of the Albany Museum, who has kindly given me permission to describe this gerbil, records that it was discovered amongst drift-wood just above high-water mark on the east side of the Sundays River Mouth and also on adjacent sand-hills at Paardevlei, Alexandria District. Genus new to the Hastern Cape Province. ba A iS a : Bi ieee Pel es A ee rr oe ; in oft by ce arN ix aes a ” : PARTS OF THE ANNALS ‘PREVIOUSLY Issv1 Baa ee Vol. I.—Part 1, out of print; Part 2, out of print; aa out of print. Vol. I1.—Part 1, out of print; Part 2, 5/6; Part 3: out of print; Part 4, “Part 5, 1/6; Part 6, 3/-; Part 7, 1/6; Part 8, 3/-7. Part, 9; Part 10, 7/-; Part 11, ‘sh Index, "Title, ete., ‘1/6. sc Vol. I11.—Part* 1, -2/6; — Part 2, 1/6; Part 3, 5/6; Part’ 4, 3f-3 Pe 5/6; Part 6,7 /-3 ‘Part’ 7, 1/6; Part 8, 38/-3 Part ‘9, 16; "3 : Title, ete., 1/6. Vol. IV (contaming ‘Palaeontological papers published in conjunction Geological Survey).— Part 1, 11/-; Part 2, 7/5 ‘Part 3, 4/6; Part: 4, 4/65 Part Part 6, 4/6; Part 7, 14j=3 Part 8, 8/-. Vol. V.—Part i: 4/6; Part 2, 8/6; Part os 2/6; Part 4, 1/6; Part’ Part 6, 5/—; Part 7, Bi Part 8, 4/6; Part 9, 4/6; Index etc., 1/6. Vol: VI.—Part , 13/6; Part 2, 4/65. Part. ee 3/6; Part 4, 30/5. Index, etc., 1 /6. ee Vol. Vil (containing Palaeontological papers published in conjunction w Geological Survey).— aa Part.1, 3/—;. Part 2, out of print. Part 3, 5/-; Part 1 a ‘Part Pi 5/6; Part 6, 1/6; Index, Title, etc., 1/6. > Vol. VIII,—Complete, out of print.. Index, Title, ete., 1/6. Vol. 1X.—Part 1, 4/6; Part 2, 5/6; Part 3, 10/-; Part 4, 6/6; Part 6 3/6: Part 6, li/-3 Part 7, 9/-;- Index, Title, etc., 1/6. Vol. X.—Part 1, 3/-; Part 2, 2/6; Part: 3, 2/-; Part 4, Bie ‘Part, Bi oe * Part 6, 3/-; Part 7,.10/-; Part 8, 2/6; Part 9, 5/-; Part 10; 265 ; Part 11, 20/-; Part 12, see Complete. “Title, pate contained . Part 11. 5 Vol. XI.—Part 1, 3/6; Part 2, 2/-; Part 3, 13/6; Part 4, 1/6; Dag: 5: Ws Part. 6, 11/-; Part 7, including = Index, Title, ete. Appendix and Plate TLL, Bn 5 Vol. XII (containing Palacontological. papers published in conjunction with t Geological Survey).— ~ eer Part 1, 15/6; Daee 2, 3/6; Part 3, 4/6; Part 4, os ay Wea Part 6, 6/=;. Part 7, 20/3 Part 8, 20/-3 Index, Title, ebe.5 1/0." ee Vol. XIII.—Part 1, 6/-; Part 2, 2/6; Part sp 3/-; Part 4, 8/6; Part. By Part 6, 5/—;, Part-7, 30/-; Part 8, He: Index, Title, etc., 1/6. Vol. XIV.—Part 1, 8/6; Part 2, 8/-; Part & 6/-; Part 4, mule. ‘Part 5, 5} Part 6, 9/—; Index, Title, etc., 1/6. . - Vol. XV.—Part 1, 17/-; Part 2, 17/-; Part 3, 14/-; Part 4, “12/6; Part 5, 5/85 %e Part 6, 3/6; Index, Title, etc., DG) oN as Vol. XVI.—Part 1, 30/6; Part 2, 4/-; Part 3 (with Title, Thad: etc. \s 25). : Vol. XVII.—Part ie 12/—; Part 2, 9/6; Part 3, 3/-; Part 4, 15/3 Part 5, | - Part 6, 2/6; Index, Title, etc., 1/6. . Cm XVIIL. cee 1, 20/-; Part 2, 7/6; Part 3, 30/-;. Part 4, 12/6; Index, Title, , 1/6. Vol. KIX Park 1, 22/-;, Part 2, 17/6; Part Ds Hs Part 4, 5/65 Index, Tit 5 etc., 1/6. : oe, ae - Vol. XX.—Part 1, 8/6; Part 2, 12/6; Part 3 4/—; Part 4, 10/5 ‘Part 5 4p Part 6 ‘(with Title, etc, ), 4/6. : ie tae coe ae Vol. XXI.—Part 1, 25/-; Part 2 (with Title, etc.), 30/-. ae ~ ; Vol. XXII (containing Palaeontological papers eiblished. in conjunction with the Geological Survey).— Sees eae Part 1, 20/—; Part 2, 10/-; Part 3 (with Title, etc.), 3/6. — Se aS Aka Vol. XXITI.—Part 1, 12/6; Part 2, 8/-; Part 3 (with Index, Title, etc.), ne Vol. XXIV.—Part 1, 10/-; Part 2, 2/6; Part 3, 4/6; Part 4, 11/-; Part 5, 16/-. as Vol. XXV.—Part ie 12/6; Part 2, 10/—; Part 3 (with Index, Title, ete.), 10. ie ep Vol. XX VI.—Complete, 25/-. . ‘ Vol. XX VII.—Complete, 25/-. . Set. Sie ie = Vol. XXVIII (containing Palaeontological papers published in conjunction with the e) Geological Survey).— Part 1, -10/-; Part 2, 17/6; Part 3, 6/-; Part 4 (pte Tndlex, ‘Mitte, etc.), 10/-. : te: : Vol. XXIX. —Part 1, 20/-; Part 2 (with eee Title, etc.) » 20/-. : Vol. XX X.—Part 1, 15/6; Part 2, 20/-; Part 3, 8/-; Path 4, 9/-; Part 5; ce ae ee Sate Vol. XX XI (containing Palaeontological papers published in cone ‘with the eae Geological Survey).— Pin ese an Ose ice ers Part 1, 20/—; Part 2, 13/6; Part 3, 17/6. —. Wey rg cree eee Vol. XXXITI.—Part 1, 2/6; Part 2, 7/6; Part 3, Te; ‘Part 4, to: Es as : Vol. XXXIV. —Complete, 40/-. The Annals of the South African Museum will be. augue at “irregular intereals, as matter for publication is available. * Copies may be obtained from— Messrs. WHELDON & WESLEY, Lro., : Peg 2, 3, and 4 ArrHuR Street, New Oxrorp Sea Tomeoe W. 0. 25 or, The LIBRARIAN, Sourn AFRICAN Muszum, Care Town. = i ANNALS ! OF THE SOUTH AFRICAN MUSEUM VOLUME XXXII | Al PART V, contarning :-— 11. The Enteropneusta from Inyack Island, Delagoa Bay.—By ©. J. VAN DER Horst, University of the Witwatersrand, Johannesburg. (With 70 Text-figures.) 18. Contributions to the Crustacean Fauna of South Africa. XII. Further Additions to the Tanaidacea, Isopoda, and Amphipoda, together with Keys for the Identification of the hitherto Recorded Marine and Fresh-water Species.—By K. H. Barnarp, D.Sc., F.L.S., Assistant Director. (With 35 Text-figures. ) | ISSUED FEBRUARY 1940. PRICE 20s. . PRINTED FOR THE TRUSTEES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN BY NEILL AND CO., LID., 212 CAUSHWAYSIDE, EDINBURGH ( 293 ) 17. The Enteropneusta from Inyack Island, Delagoa Bay.—By C. J. VAN DER Horst, University of the Witwatersrand, Johannesburg. (With 70 Text-figures.) CONTENTS. PAGE PAGE Saccoglossus inhacensis Kapelus . 293 | Balanoglossus hydrocephalus n. sp. ps ie: ; and studiosorum n. sp. . . 335 Beg sclegoonstsn.'sp. Nag Ptychodera flava Eschscholtz San) Glossobalanus alatus n. sp. . . 316 | Literature ; é i Se) THE island of Inyack, in Delagoa Bay, Portuguese Hast Africa, is an ideal place for collecting marine animals that live burrowing in the soil in the tidal zone. The eastern side of the island is exposed to the Indian Ocean, but along its northern, western, and southern sides it faces the Bay, which is here very shallow. As the difference between high and low tides can be over 3 metres, large flats and sand-banks are exposed at low tide. In the more sheltered bays these flats have a muddy soil; in places where a coral reef extends in front of the flats the mud is mixed with coral debris, but in the more exposed parts, where either the tidal stream or the action of the waves is stronger, large banks of pure sand occur. This is a place where one can expect Hnteropneusta, and they are found here in great abundance. During the last few years I have visited Inyack repeatedly and have collected a great number of Hnteropneusta. Up till now I found six different species, only one of which had pre- viously been described. Another species was recently described by my student, Mrs. Kapelus, and in the following pages the description of the other four species is presented. SACCOGLOSSUS INHACENSIS Kapelus. The small enteropneust, Saccoglossus inhacensis, described by Mrs. Kapelus, occurs in countless numbers on the muddy flats west of the island. In nearly every spadeful of sandy mud one can find one or more specimens. In a letter to “Nature” I described the characteristic burrow made by this animal in the mud; and now, after some further observations, VOL. XXXII, PART 5. 22 JUL yo te 294 Annals of the South African Museum. I am able to supplement the previous remarks (fig. 1). At the place where the animal occurs most abundantly the black mud is covered by a layer, about 1 cm. in thickness, of yellow sand, and as this sand does not hold together so well as the mud, I was at first unable to detect the tube of the animal in the sand. But later I found it extending from the surface to a depth of 4 to 7 cm. The upper part | | of the burrow is irregularly coiled, but deeper down it takes the form of a regular spiral, con- sisting of up to eight turns. The direction of the main axis of the spiral is variable; in the major- ity it was found to be approxi- mately vertical, but it may even be horizontal. In the dark mud this burrow is very conspicuous because it is lined by a thin layer of clear sand. This in turn is covered by a layer of slime, giving the inner surface a smooth and shiny appearance. As the sand is also cemented together by the slime the spirals are rather persistent; at least as many were found that had been vacated by the animals as inhabited ones. In the letter to “Nature” I stated that no indication of the presence of the animal was found on the surface during low tide. Subsequently I found that this statement was wrong. Like other enteropneusts, Saccoglossus inhacensis also makes heaps of coiled castings round the entrance of its burrow. But these heaps are very small, and the coils seem to be not so persistent as in Balanoglossus. Only when there is no disturbance at all the heaps, which are about 1 cm. wide and 2 mm. high, will preserve their typical form and then they cover the surface in great numbers at low tide. In order to make these heaps of castings, the animals must protrude their hind- ends, where the anus is, above the surface of the sand. Nevertheless, when one finds the animals in their burrow by digging, they invariably take up a position with the proboscis and anterior end of the body in Fic. 1.—Burrow of Saccoglossus inhacensis Kapelus. Nat. size. The Enteropneusta from Inyack Island, Delagoa Bay. 295 the irregularly formed upper part of the burrow, and the abdominal region is found in the spiral. Therefore the animal must be able to turn round in its burrow. In miocene strata from different localities (Switzerland, Bavaria, Maryland, California, Mexico) fossilised internal casts in the form of spirals have been described under the name of Xenoheliz. Several suggestions have been made about the origin of these spirals. Mans- field was of the opinion that they owe their origin to some marine plant; Heer supposes them to be the burrows of the mussel Mactrina. Abel, in his book “ Vorzeitliche Lebensspuren,” expresses the opinion that these spirals are the burrows of small crabs. Along tropical shores a small crab, Mictyris,* may occur in countless numbers, and at the approach of danger these crabs disappear in the mud by digging a burrow in the form of a spiral. However, these burrows can be only of a very temporary nature, and as Abel wrote me later he has given up this opinion. The stratum in which these spirals are found was formed in a shallow sea with a very level shore, where mangroves grew in great abundance. It is exactly in such a place that the spirals of Saccoglossus inhacensis are found, and as these spirals are rather persistent, even when they are uninhabited, I venture to suggest that Xenoheliz is the fossilised burrow of an enteropneust. WILLEYIA DELAGOENSIS 2. Sp. In November 1935, during a short stay on Inyack Island, I secured a single specimen of a pure white enteropneust on the flats at the west side of the island. It proved to belong to the genus Willeyza, established by Punnett for an enteropneust from Zanzibar. This genus is more or less intermediate between Glandiceps and Spengelia; it agrees with Spengelia in having peripharyngeal cavities and with Glandiceps in the absence of synapticula. As Punnett’s description is not as extensive and accurate as it might have been, Spengel has expressed his doubts about the validity of the genus. However, I can corroborate Punnett’s observations by the study of this specimen, and this leaves me in no doubt regarding the validity of the genus Willeyia. In addition to some minor characters, the only specimen that was collected at Inyack differs in one important feature from Willeyia bisulcata Punnett, i.e. in the extension of the gonads into the branchial region, a character which Punnett thought might be * Mictyris has not been recorded from South African shores, but the closely allied Dotilla is very common.—ED. Annals of the South African Museum. of generic importance. It is for this reason that I am describing this specimen as belonging to a new species of the genus Willeyia. 296 EKaternal Features. In its external appearance W. delagoensis closely agrees with W. brsulcata according to Punnett’s description (fig. 2). The pro- boscis is very elongated; in the fixed animal it has a length of 15 mm. i i Y We re ) SCR SEN Sa LE LA ] M left) and ventral (right) Fic. 2.—Willeyia delagoensis n. sp. Dorsal ( views. x2. and its greatest width is 5 mm.; in the living animal the proboscis It also shows the longitudinal grooves was considerably longer. along the dorsal and ventral mid-lines. The collar is about 5 mm. long The Enteropneusta from Inyack Island, Delagoa Bay. 297 and is broader behind than in front. The proboscis is flattened, but the collar and the trunk, at least as far as the latter is present, are nearly cylindrical. There is no trace of genital wings or folds, not even behind the branchial region. It is difficult to determine the length of the branchial region owing to the fact that the minute branchial pores open into a narrow and often deep groove, which fades away gradually in the anterior part of the genital region. By counting the sections it was possible to estimate that the branchial region is about 3cm. long. The specimen was broken off in the genital region, so that it remains uncertain whether Wealleyia has liver saccules or not. Punnett’s specimen was also incomplete. The animal, when alive, was of a pure white colour throughout. Internal Anatomy. Proboscis.—The epidermis of the proboscis consists of elongated cells. I can confirm Punnett’s statement that the epidermis is almost destitute of glands, particularly as special glandular cells seem to be absent. A great number of the ordinary cells are swollen in their peripheral part, exhibiting the features of slime cells. It seems likely that all epidermal cells can secrete slime. As in W. bisulcata, elongated nuclei are arranged in a distinct row in the middle of the cells, and another layer of round nuclei is found in the basal part of the epidermis; the latter are evidently the nuclei of the nerve cells. In W. bisulcata the circular muscle layer is about half as thick as the nerve layer; in W. delagoensis, however, these layers are of about the same thickness, except in the grooves where the nerve-fibre layer is better developed (fig. 3). Sensory cells could not be found in the epidermis of these grooves. The muscle fibres are neither arranged radially nor in concentric layers; they are only more crowded in the central part of the pro- boscis than near the epidermis. It is well known that these longi- tudinal muscle fibres are attached by both ends to the basal membrane of the epidermis. Whereas the anterior end of each fibre pierces the circular muscle and attaches itself to the basal membrane independent of other fibres, one finds that in the posterior part of the proboscis the longitudinal muscle fibres collect in bundles, where they pierce the circular muscle layer and are attached to the basal membrane. In cross-section it seems therefore that the circular muscle layer is 298 Annals of the South African Museum. interrupted; in reality this muscle layer will have the appearance of a network. The basal membrane is produced into an inwardly directed point or crest where the longitudinal muscle fibres are attached to it (fig. 4). The proboscis coelomic cavity extends nearly to the top of the proboscis. Over its whole length it is separated into a right and a left part by a kind of dorso-ventral septum formed of connective tissue and muscle fibres (fig. 3). Near the anterior end of the glomerulus and again just in front of it there are a few openings in this septum by means of which the right and the left coelomic cavities intercom- municate. Punnett figures the right and the left coelomic parts as of equal size in W. bisulcata, but in W. delagoensis the right part is wider than the left, though taken together the two parts form a symmetrical figure; the septum between the two has evidently shifted towards the left side (fig. 3). -Owing to the fact that this septum consists of fibres only and not of a limiting membrane, this shifting does not seem to be of great importance and the difference in size of the coelomic parts may not be of specific value. The dorso-ventral muscle fibres go by way of this septum from the dorsal to the ventral mid-line, so that they deviate somewhat from the straight course. These fibres are thicker than those of the longitudinal or circular musculature. They do not form a continuous layer as they run separately; there may be one or more fibres visible in a cross-section or none at all. At the dorsal side of the septum the fibres are arranged in two longitudinal rows. Ventrally there is usually one row, but if there are two rows these lie very close together, except in the posterior part of the proboscis, just i. ont of the glomerulus, where the two rows are more distant from each other and can be easily discerned at the ventral side of the septum. From the anterior to the caudal part of the proboscis the coelomic cavity gradually increases in width. This cavity is surrounded by a distinct epithelium which, as Punnett also mentions, is considerably thickened on the septum. It is somewhat lower in the right and the left angles of the cavities and it is very low along the dorsal and ventral border, although here also it is clearly visible (fig. 3). This epithelium does not seem to be of an excretory nature, as may be the case in some other Enteropneusta. The ventral proboscis septum, as in Glandiceps, reaches to the top of the stomochord but leaves the vermiform process free. Its attachment to the basal membrane of the epidermis, however, does not reach so far forward, as a result of which the free edge of the “£1941” TeIyUOA “MA “SUTOA srosoqoud ‘a ‘wingdoes ‘s ‘ornpeTnosnut Suid roponu ‘wu ‘zoXe] orqy-oAdou “U “ornzeTNOSsnUr jeurpnyisuoy “7 JO UOTPOOS-SsOID B JO WRT “P “PE X “Ud ‘“S{[oo OATOU JO ‘Kroqie Testop ‘yp *APLAwO oTUTO[IOD “Od ‘stutmopide jo ouvIqULoUr "G1Z x ‘stosoqoad oy} ‘gues10 [e1yU0d oY JO JUOLF UL grosoqord yo woroos-ss01y “g “d yeseq “wg su sisuaobnjop vihantM—'P “& “SOL vt) Win, i), i Bye oS iw. . * ad oF un yee fee ; OF Lay AAAI: a SO ean er - elias b os o/h, Rd: ar UME pe . re ® frtesg ¢ f ° “jy. “ en ~ fe, hs 5. _ Se. AY Sy . . . se = Ate ian 300 Annals of the South African Museum. septum runs from the top of the stomochord to the epidermis in a ventro-caudal direction. Probably on account of this the ventral Fic. 5.—Willeyia delagoensis n. sp. Cross-section of anterior part of the central organs of proboscis. x55. au, auricle of pericardium. c, coelomic cavity. epd, epidermis. gl, clomerulus. lm, longitudinal musculature. m, dorso-ventral muscle fibres. nl, nerve-fibre layer. rm, ring musculature. st, stomochord. va, ventral proboscis artery. vp, vermiform process. vs, ventral proboscis septum. proboscis artery is not found in the septum; it runs from the top of the stomochord freely through the coelom and directly to the epi- dermis in a ventral direction (fig. 5). According to Punnett this vessel, called by him the ventral recurrent vessel of the proboscis, The Enteropneusta from Inyack Island, Delagoa Bay. 301 runs in the upper edge of the septum in W. bisulcata, and according to his figure the ventral septum is exceedingly long in this species, reaching to nearly the top of the proboscis. The ventral coelomic blind-pouches are separated from each other throughout their whole length by this septum; they have the form of narrow finger-like canals, the epithelium of which is higher on the dorsal than on the ventral side (fig. 8). Their caudal end is at the same level as the caudal extremity of the ventral blind-sac of the stomochord; here near their posterior end the ventral pouches are quite embedded in the skeleton (fig. 6). The right dorsal coelomic pouch ends blindly, but the left is con- tinuous with a median end-sac (fig. 6) which opens by a left proboscis pore to the exterior (fig. 7). The proboscis pore has the form of an elongated fissure, the caudal end of which opens into the anterior neuropore. The end-sac does not extend beyond the caudal end of the pore and is surrounded by a thick limiting membrane, which probably contains blood-vessels (fig. 7). It is impossible for the right and the left halves of the glomerulus to fuse in front of the pericardium or the stomochord because of the presence of the dorso-ventral septum. The glomerulus reaches as far in a frontal direction as the vermiform process, and that is not far in front of the stomochord. It covers the stomochord to the same extent as the pericardium and is found at the lateral side of these structures only; it even remains at some considerable distance from the dorsal and ventral mid-line. There is a triangular mass of cells extending for a considerable length along the dorsal side of the pericardium, but this mass does not seem to be a dorsal glomerulus as no blood-vessels could be detected init; otherwise a dorsal glomerulus seems to be present in all Spengeliidae. On the coelomic side the glomerulus is covered by a rather thick layer of cells (fig. 5). Immediately caudal to the mass of cells mentioned above, the free dorsal wall of the pericardium is thrown into irregular folds, which continue to where the pericardium attaches itself to the basal mem- brane of the epidermis, and only then the pericardium assumes its usual triangular form. Anteriorly the pericardium has two auricles as in other Spengeliidae. These auricles, however, are small; they do not reach to the top of the stomochord, but they are, as usual, surrounded by the anterior portion of the glomerulus (fig. 5). Anteriorly the pericardium is quite filled with a mass of cells; the middle portion, on the other hand, is quite empty except for a layer of cells along the wall; and in the posterior part, where the | ' | 302 Annals of the South African Museum. <\"ti i SGmISS. f < | Ly Dy Miran: LA EPL NRARSSN NS LOT TARR SAN ns Fics. 6-9.—Wiulleyia delagoensis n. sp. 6. Cross-section of proboscis neck. 7. Cross-section of proboscis neck caudal to that of fig. 6. 8. Cross-section of proboscis neck frontal to that of fig. 6 (epidermis being much damaged is not drawn in). 9. Dorso-median portion of a cross-section of anterior part of collar. All x 45. bl, central blood sinus. cc, collar coelom. dbv, dorsal blood-vessel. dpc, dorso-lateral pouches of proboscis coelom. epv, efferent proboscis vessel. es, end-sac. k, keel of skeleton. Jp, lateral blind-pouch of stomochord. mc, nerve cord. , pericardium. pp, proboscis pore. sk, skeleton. st, stomochord. sv, sinus venosus. v, proboscis vein. vbp, ventral blind-pouch of stomochord. vbs, ventral blind-sacs of proboscis coelom. The Enteropneusta from Inyack Island, Delagoa Bay. 303 pericardium is connected with the epidermis, strong fibres are seen connecting the two side walls. Near the posterior end of the pericardium the two proboscis veins unite at the right-hand side of the end-sac of the proboscis pore and then communicate with the wide sinus venosus (fig. 6). This com- munication pushes the posterior wall of the pericardium slightly forward so that in the sections the pericardium is visible at both sides of the blood-vessel. As in all Spengeliidae, the stomochord is prolonged into a vermiform process (fig. 5), but in W. delagoensis this process is extremely short, being only 80 » in extent. In W. bisulcata its length is about two- thirds of the length of the rest of the stomochord. The anterior part of the stomochord is flattened; in cross-section it is oval in shape (fig. 5). Farther caudally it becomes more quadrangular. The lateral blind-pouches protrude somewhat in an anterior direction before they connect with the main middle part of the stomochord. The lumen, which is continuous throughout and begins near the top of the stomochord (fig. 5), becomes very wide and broad where the lateral blind-pouches connect with the middle part. The rather regular epithelial wall of the stomochord is very distinct in this region (fig. 8). More caudally, in the proboscis neck, the stomochord becomes first narrower but higher, then it is divided by the skeleton into a dorsal part, which is the neck of the stomochord, and a ventral blind-pouch, which is quite embedded in the skeleton (fig. 6). Except for the fact that ventrally an upgrowth of the skeleton causes an inpushing in its wall, this portion of the neck is nearly circular in cross-section (fig. 6), but farther caudally the neck becomes more depressed (figs. 7 and 9). In the neck part of the stomochord the epithelium contains numerous glandular cells, but in the body such cells could not be found. Near the level of the middle of the lateral blind-pouches the limiting membrane at the sides of the stomochord increases in thickness, and this constitutes the most anterior end of the proboscis skeleton (fig. 8). The limiting membrane also becomes slightly thicker ventrally to the stomochord and the coelomic blind-sacs. More posteriorly these two lateral portions of the cup-shaped anterior end of the skeleton are united across the mid-line by the dorsal side of the cup, by which, as previously mentioned, the neck of the stomochord is separated from its ventral blind-pouch (fig. 6). Furthermore, the ventral wall of the cup has also thickened here considerably. At this point a dorso-median ridge of the skeleton cuts into the ventral wall oa ES A a ~ remap amen ae ee ne eee Se 304 Annals of the South African Museum. of the stomochordal neck. In the dorsal part of the proboscis neck as well the limiting membranes have become very thick and the whole may be considered as one skeletal mass in which the efferent proboscis vessels, the dorso-lateral coelomic sacs, the posterior end of the pericardium, the sinus venosus, and the end-sac are embedded (fig. 6). On the other hand, chondroid tissue, otherwise so well developed in the Spengeliidae, is hardly present in W. delagoensis. Caudally to the ventral blind-pouch of the stomochord, where the real skeleton becomes a solid structure, the more dorsally situated limiting membranes become thinner again. Anteriorly, 7.e. directly behind the anterior cup, the body of the skeleton cannot be distinguished from the keel, but near the anterior end of the collar coelomic cavities the skeleton is constricted by these cavities, so that body and keel may be easily distinguished here (fig. 7). The keel has here a pair of wings dorso-laterally; it is large and broad though its dorso-ventral diameter is longer than its horizontal diameter. At the level where the proboscis neck is attached to the collar, the connection between the body and the keel of the skeleton becomes broader again (fig. 16); here the crura also become visible in the middle of the skeleton, thus the fused crura are surrounded by secondary layers of skeletal material on all sides. The skeleton now increases considerably in breadth, and the keel is much broader than high (fig. 9). The crura reach to about half the length of the collar, and their extremities are found in the ventro-lateral angles of the buccal cavity, so that they surround this cavity for about two-thirds of its circumference. Collar.—The epidermal zones of the collar are quite distinct and well differentiated from each other. The epidermis of the anterior zone 1s rather low and it shows the usual nerve-fibre layer adjoining the basal membrane; this is followed by a layer of round nuclei of the nerve cells (fig. 10). Externally to this is a dense and thick layer of elongated nuclei, which apparently belongs to the undifferentiated epidermis cells; this layer reaches nearly to half the thickness of the epidermis. In the peripheral part of the epidermis are found some scattered, nearly round nuclei, which may belong to the glandular cells. There are many glandular cells in the epidermis and they seem to extend over the whole thickness of this layer. They are not swollen over their entire length, but the swollen part of the cells may be found at any level in the epidermis. In some patches the swollen part of most if not all glandular cells is found near the surface, in others near the base of the epidermis. In the latter case the oval The Enteropneusta from Inyack Island, Delagoa Bay. 305 nuclei of the epidermis cells are displaced and found higher up in the cells. The greater part of the epidermis of the collar consists of a very high epithelium formed by narrow, elongated cells (fig. 11). Here Wal ivhole r Tah ear ———_._ »! > q < “ oe 5 A! 4 Nerd i 5 r | i Ri | 8 E } i bis > v7 = } ll A ! ¢ | ke | pein Hi, ie : pans, a f « a i) 9) Sea Sh z et } B yi Zz if Hy i i bhgip r4 ahi bit NM J nA » ‘ | 4 4 , = 12 il Fies. 10-12.—Willeyia delagoensis n. sp. 10. Epidermis of the anterior zone of collar. 11. Epidermis of second zone of collar. 12. Epidermis of third zone of collar. All x 220. again are found the round nuclei of the nerve cells adjoining the nerve- fibre layer. A few very elongated nuclei are found scattered over the greater part of the epidermis, but these nuclei form a distinct layer at about one-quarter of the total thickness of the epidermis from the surface. In this part of the epidermis there are many glandular cells, which are of two types. Some are as long as the epidermis is thick, their content is granular and stains with haematoxylin. The 306 Annals of the South African Museum. other type is found only near the surface, at least the part of the cell containing the glandular product; this product is homogeneous and stains with eosin. This zone is separated from the next one by a narrow deep groove in which the cells are similar to those just described except that they are not so elongated. The next, rather narrow epidermal zone consists of a very charac- teristic epithelium (fig. 12). It is somewhat lower than that of the preceding zone, although the nerve-fibre layer is thicker. Next to the fibre layer are found the round nuclei of the nerve cells. The oval nuclei of the elongated epidermal cells are arranged in two distinct rows; one is found slightly external to the nerve-cell nuclei and the other higher up in the cells. Between these two rows of nuclei the cells are coloured uniformly blue by haematoxylin. External to the peripheral layer of nuclei and all starting at exactly the same level, all cells are filled with a fine granular substance that stains with eosin. The epidermis in this region therefore gives the impression that it is formed by two layers of cells. The cilia of this zone are very short, much shorter than in the two preceding zones. The zone is sharply delimited from the next one. The posterior part of the collar, just in front of the circular nerve, is covered by a low epithelium containing many glandular cells and not differing from the epidermis of the trunk. The external longitudinal musculature of the collar is well developed in the anterior half of the collar. It fills about half of the coelomic cavity between the epidermis and the gut (fig. 13). Its fibres, however, do not cross the body cavity in order to become attached to the wall of the gut at their caudal end. They remain in the same position near the epidermis throughout their whole length, and caudally they are fixed to the basal membrane of the epidermis. The external longitudinal musculature ends abruptly in about the middle of the collar (figs. 14 and 15). According to Punnett there is no layer of circular muscles outside the longitudinal musculature. This is usually the case in Enterop- neusta, as the external circular muscle of the collar is found inside the external longitudinal musculature. This external ring-muscle layer is very well developed in Glandiceps, but it is reduced in its extent in Spengelia. In Willeyia I find this layer very poorly developed and it is often interrupted (fig. 13). Furthermore, it is restricted to the anterior part of the collar. The internal longitudinal musculature consists of coarser fibres than the external. It is, as in other Spengeliidae, fan-shaped. P . . : . The Enteropneusta from Inyack Island, Delagoa Bay. 307 Starting anteriorly at the dorsal side of the gut, where it is attached to the folds of the limiting membrane in which the efferent proboscis vessels run (figs. 9 and 13), its fibres gradually spread out caudally Fic. 13.—Willeyia delagoensisn.sp. Cross-section of anterior part of collar. x 35. be, buccal cavity. dbv, dorsal blood-vessel. dm, dorsal mesentery. elm, external longitudinal musculature. epv, efferent proboscis vessel. erm, external ring musculature. ilm, internal longitudinal musculature. irm, internal ring musculature. nc, nerve cord. ph, top of perihaemal cavity. sk, skeleton. st, stomochord. wm, ventral mesentery. towards the ventral side, reaching the ventral mid-line about in the middle of the collar near the posterior end of the external longitudinal musculature (fig. 15). In the posterior half of the collar the internal longitudinal musculature is very strong, filling nearly the whole body cavity. It is divided up in irregular bundles by strands of radial 308 Annals of the South African Museum. muscle fibres (fig. 15). In the anterior part of the collar there is a strong internal circular muscle layer round the buccal cavity (fig. 13). The peripharyngeal cavities with their circular muscle fibres begin in the anterior part of the collar near the dorsal side of the buccal cavity, and as these cavities extend caudally they gradually surround a greater part of the buccal cavity and replace the internal ring musculature of the collar. The peripharyngeal cavities are rather irregular, because the limiting membranes at their outside and inside often coalesce locally, thereby interrupting the ring musculature in the cavities (fig. 14). On the other hand, the ventral end of the peripharyngeal cavities is not always closed, so that there are com- munications between these cavities, which in reality form part of the trunk coelom and the collar coelom. The ring muscles of the peripharyngeal cavities are then directly continuous with the internal ring musculature of the collar (fig. 15). In the anterior part of the collar the ventral boundary of the peripharyngeal cavities follows, as usual, the crura of the skeleton. From there the fold of the limiting membrane, in which the efferent proboscis vessel runs, extends somewhat more ventrally. The peripharyngeal cavities are closed off ventrally up to the end of the crura, and it is only beyond the crura that these communications occur. At the dorsal side the peripharyngeal cavities do not extend below the perihaemal cavities; they do not even reach the latter cavities. Therefore, though there are ring-muscle fibres in the perihaemal cavities, there is nevertheless an interruption in the whole ring- muscle layer round the buccal cavity. At their anterior end the perihaemal cavities diverge, so that here only the dorsal blood-vessel separates the collar nerve cord from the stomochord (fig. 13). The perihaemal cavities are also rather short, as they do not reach the anterior end of the collar (fig. 16). The septum between the right and the left perihaemal cavities is entire throughout their whole length. The dorsal and ventral mesenteries stretch uninterruptedly through- out the whole length of the collar. In this species the ventral mesentery reaches the anterior end of the collar, whereas, according to Punnett, it is lacking in W. bisulcata. At the front end of the collar nerve cord there is a deep anterior epidermal depression (the “‘ vordere Vorhéhle”’ of Spengel). In Spen- gelia alba and Gilandiceps hacksi there exists besides this anterior depression another inpouching of the epidermis just dorsal to the depression (‘‘ Epidermisgrube ” of Spengel). Willey interpreted the ae eed ll ee ee ss — NAY ieee a OE ANG URNA K aie Do abs a) ef io | LUN AP ii Late a ASICS iN TT 15 Fies. 14, 15.—Willeyia delagoensis n. sp. 14. Ventro-lateral part of a cross- section of the collar. 15. The same, a few sections behind that shown in fig. 14. Both x 52. be, buccal cavity. elm, external longitudinal musculature. epv, efferent proboscis vessel. ilm, internal longitudinal musculature. im, internal ring musculature. pp, peripharyngeal cavity. wm, ventral mesentery. WO. XXXIT, PART 5D. 23 310 Annals of the South African Museum. depression and the pouch as the same structure, which he called anterior neuropore. Spengel has pointed out that they are different structures, but he was of the opinion that the depression could not be called anterior neuropore, because its wall has an epidermal structure. However, it must be noted that the dorsal wall of the collar cord itself, if a central canal is present, may have an epidermal character, and as epidermis and collar cord are continuous, it is often very difficult to determine where the one stops and the other begins. Therefore I think it best to call the opening of the depression the neuropore; in a general way it corresponds with the neuropore of vertebrates. The depression itself is then the central canal that is present in the anterior part of the collar only. Thus, in addition to the neuropore, there may be an epidermal pouch dorsal to it, or the neuropore may be closed and the pouch present, or the pouch may be absent as in most Enteropneusta. In his fig. 10 on plate xli Punnett draws the neuropore only, but in his fig. 15 on plate xli the infolding of the epidermis is partially separated from the nerve cord and lying at its dorsal side. From this figure and fig. 16 one gets the impression that both pouch and neuropore are present in W. bisulcata, that the two are not completely separated from each other, and that the pouch is the deeper of the two. However this may be, in W. delagoensis there is no pouch dorsal to the collar cord, but as a canal penetrates from the anterior end deep into the collar cord there is an open anterior neuropore (figs. 9, 13, and 16). A similar canal penetrating into the collar cord at its posterior end is hardly present; one may therefore conclude that the posterior neuropore is closed. The nerve cord shows only a few medullary cavities; in W. bisulcata it is quite solid. Punnett mentions a well-marked ridge projecting up towards the dorsal mesentery in the hinder portion of the nerve cord in W. bisulcata. No trace of such a ridge could be found in W. delagoensis. Punnett’s statement that there are oesophageal nerves connecting the cord with the epithelium of the buccal cavity, called by him oesophagus, ~ seems doubtful to me, as nothing resembling these nerves could be found either in W. delagoensis or any other enteropneust. The collar canals fuse near their anterior ends with the first branchial sac, which here is already reduced to the size of a canal. Therefore what seems to be the collar canal is partly branchial pore, the inner wall of this canal being formed by the epithelium of the branchial sac (fig. 17). The dorsal or outer wall of the canal, which is slightly thinner than the side walls, shows a deep, inwardly directed fold. The Enteropneusta from Inyack Island, Delagoa Bay. 311 The two lateral walls are very thick, and as in Glandiceps talaboti and a few other Enteropneusta the cells are so numerous and narrow that their nuclei, instead of being arranged in a single layer in the middle of the cells, form a dense mass occupying nearly the whole thickness of the epithelium (fig. 17). Fies. 16, 17.—Willeyia delagoensis n. sp. 16. Dorso-median portion of a cross- section of collar slightly in front of that shown in fig. 9. 17. Cross-section of the collar canal. Both x 45. bsl, wall of first branchial sac. bs2, second branchial sac. ce (fig. 16), central canal of collar nerve cord. cc (fig. 17), collar coelom. cp, collar canal. dbv, dorsal blood-vessel. dm, dorsal mesentery. elm, external longitudinal musculature. epd, epidermis. epv, efferent proboscis vessel. erm, external ring musculature. im, internal longitudinal musculature. nc, collar nerve cord. sk, skeleton. st, stomochord. ic, trunk coelom. The inner surface of the buccal epithelium is very irregular owing to the presence of numerous small folds penetrating into this epi- thelium (figs. 18, 14, and 15). There are numerous small glandular cells, which occur only in these folds and are absent from the free surface. They secrete a slimy substance in the form of long threads. The threads of the separate cells coalesce into thicker strands, which emerge from the folds into the buccal cavity. The whole is reminis- cent of a byssus gland of a Lamellibranchiate. Trunk.—The first few branchial sacs are situated far in front of SS e——eEEEOEoEeEOEOEOEEeEeEeEeEeEeEeeeeeeEeEeEe een eeeeeaayanDn”DmwUEYUvrEEECrmOOOO ———EEe LT eeeeeeeEeEEeEeEeEeEeEeeeEeEeEeEeEEE————E——EEE —- 312 Annals of the South African Museum. their external apertures, so that these sacs are drawn out into long canals leading to the pores (fig. 17). The total number of gills is 144 on each side of the body; Punnett, unfortunately, does not give the number of gillsin W. bisulcata. The latter species is characterised by the very short branchial bars, so that the branchial part of the pharynx is much smaller in the cross-section than the digestive part. Although the tongues are also short in W. delagoensis, the two parts of the pharynx are of about equal size, the branchial part being even a little larger than the digestive part. There is no indication even of parabranchial ridges separating the two parts of the pharynx. According to Punnett the first three gill pouches of W. bisulcata are dorsally confluent, forming a chamber with a single external pore into which opens the collar pore. A similar fusion of a few of the anterior gill pouches is known in some other Enteropneusta. In W. delagoensis, however, the anterior gill pouches have separate pores, though the first and second pores touch each other. These pores open into a very deep and narrow sublateral groove. In the middle portion of the branchial region the groove becomes shallower (fig. 18), but in the posterior part it is very deep again (fig. 19). This accounts for one’s inability to see the small branchial pores when studying the external features of the animal. The absence of synapticula has been already mentioned as a char- acter of the genus Willeyva. The tongues are short and very thick, so that they fill up nearly the whole of the branchial sacs (fig. 18). This is due to the fact that the ciliary epithelium covering the anterior and posterior surfaces of the tongues and septa is very high. It has the same character as the epithelium of the collar canals, consisting of very high and narrow cells, the nuclei of which fill up nearly the whole thickness of the epithelium instead of forming a single row in the middle. In Glandiceps talaboti the collar canals have a similar epithelium, but the gills have the low ciliated epithelium as in most Hnteropneusta. On the other hand, Harrimania kupfferi and a few other species belonging to the Harrimaniidae have high epithelium like W. delagoensis both in the collar canals and the gills. The nature of this epithelium in W. bisulcata is not mentioned by Punnett. The back of the septa, facing the pharyngeal lumen, is covered by a thin epithelium, while that on the back of the tongues is higher. Following the last gill there are two more rudimentary gills, each consisting of a small sac and tongue only but having no opening to the exterior (fig. 19). | The Enteropneusta from Inyack Island, Delagoa Bay. 3138 The digestive part of the pharynx is covered by a very high epithelium, the inner surface of which shows many irregular folds (fig. 18). In the anterior part of the oesophagus the epithelium is much lower and without folds (fig. 19). The longitudinal musculature of the branchial region as well as that of the genital region, as far as this could be studied in this speci- men, is very well developed. It fills nearly the whole of the body cavity, leaving only a narrow space around the intestine. There is no special thickening of this musculature near the ventral mid-line; it was only found that the fibres may be more crowded there (fig. 18). An internal ring muscle around the ventral pharynx is well developed. The ventral nerve cord projects very far inwards, a fact which has also been observed by Punnett. According to Punnett the branchial region of W. bisulcata is probably devoid of gonads, or if gonads occur here then they are found only in the most posterior portion. In this respect W. delagoensis shows a marked difference from W. bisulcata. In the former species the first gonopore is found near the 34th branchial pore, which means that the row of gonads starts at about a quarter of the total length of the branchial region away from the collar. Punnett’s specimen of W. bisulcata was a male; the only specimen of W. delagoensis available is a female, but as the extent of the gonads is not known to show sexual dimorphism in any other enteropneust it seems most unlikely that such a difference should occur here. _ The first gonads are small and have a well-pronounced gonoduct, opening in the sublateral groove at the ventral side of the branchial pores (fig. 18). Caudally they increase in size, thus gradually envelop- ing the intestine as they extend towards the ventral side (fig. 19). The specimen was broken off in the anterior part of the genital region, and, as yet, the gonads had not increased sufficiently in size to cause the body wall to protrude in the form of genital folds. As in W. bisulcata, there are no median gonads, although the gonads project forward around the posterior gill slit so as to appear on its medial side (fig. 19). Diagnosis of Willeyia delagoensis n. sp. In the fixed state the proboscis is 15 mm. long and its greatest width is 5 mm., the collar is 5 mm. long and broader behind than in front, the branchial region is about 3 cm. long (the only specimen collected is broken off near the frontal end of the genital region). ay an poe oe Read : cx ~ BRASS ESSN ~ N Sew: WWE \ a 19 Fias. 18, 19.— Willeyia delagoensisn. sp. 18. Cross-section of the branchial region, composed of several sections, in order to show a tongue at one side and a septum at the other; in reality two tongues are opposite one another. The section is not through the middle of the tongue, showing the tongue coelom, but through the anterior or posterior end of the tongue and the septum in order to show the thick layer of crowded nuclei. x20. 19. Cross-section of the posterior end of the branchial region. x 34. bp, branchial pore. 6s, branchial sac. bt, tongue. dm, dorsal mesentery. dn, dorsal nerve cord. dv, dorsal blood-vessel. ebr, epibranchial ridge. g, gonad. gd, gonoduct. gp, genital pore. I/m, longitudinal musculature. oes, oesophagus. ph, pharynx. rm, ring musculature. slg, sublateral groove. ss, septal skeletal bar. ts, tongue skeletal bar. vn, ventral nerve cord. wv, ventral blood-vessel. The Enteropneusta from Inyack Island, Delagoa Bay. 315 The proboscis is flattened, with a distinct groove in dorsal and ventral mid-lines. The collar and the branchial region are cylindrical, no trace of genital wings. The living animal is of a pure white colour. Proboscis.—No special glandular cells in the epidermis of the proboscis. Circular muscle layer of the proboscis about as thick as the nerve-fibre layer; longitudinal muscle fibres do not show any special arrangement. The proboscis coelomic cavity extends nearly to the top of the proboscis and is separated into right and left parts of unequal size by a dorso-ventral septum formed of connective tissue and muscle fibres. Ventral proboscis septum does not extend farther forward than the top of the stomochord. Ventral proboscis artery free from ventral septum. Ventral coelomic blind-sacs separated from each other throughout their whole length. Left dorsal coelomic sac communicates with a median end-sac; proboscis pore on the left side. Right and left half of the glomerulus not connected in front of the pericardium or stomochord nor at their dorsal or ventral sides; frontal end of glomerulus at both sides surrounding the auricles of the pericardium. No dorsal glomerulus. Pericardium with two short auricles. Vermiform process of stomochord very short. Stomo- chord with wide lumen, extending to near the top, very wide where the lateral pouches connect with central part. Hardly any chondroid tissue, but skeleton very well developed, with deep anterior cup; just behind the cup the body of the skeleton is nearly separated from the keel by the collar coelomic cavities, farther caudally the skeleton becomes broader, and keel and body are indistinguishable. Crura reach to about half the length of the collar and surround the buccal cavity for about two-thirds of its circumference. } Collar.—First zone of collar epidermis low, second zone with very thick epithelium and small superficial glandular cells, third zone with thick epithelium and a distinct superficial layer of glands, fourth zone low lke trunk epithelium. External longitudinal musculature in anterior half of collar only, attached to the epidermis at both ends. Thin external circular muscle layer, internal circular muscle layer well developed ventrally to the peripharyngeal cavities; these cavities do not reach the perihaemal cavities, the latter with a layer of circular muscle fibres. The perihaemal cavities quite sepa- rated from each other, their anterior ends diverge and do not reach the anterior end of the collar. Dorsal and ventral mesenteries complete. Central canal in anterior end of collar cord only; in the rest a few medullary cavities. No epidermal pouch dorsal to the anterior neuropore. No dorsal ridge on the nerve cord. Collar 316 Annals of the South African Museum. canals with thick epithelium that is almost entirely filled by the crowded nuclei. Trunk.—Number of gills 144. Branchial part of the pharynx about equal in size to digestive part. No parabranchial ridges. The first gills open to the exterior by separate pores, though the first and second pores touch each other. Sublateral groove narrow and deep in anterior and posterior part of branchial region, shallower in middle portion. No synapticula. Ciliated epithelium of tongues and septa very high, with thick and dense layer of nuclei. Epithelium of digestive part of the pharynx very high, much higher than that of anterior portion of oesophagus. Row of gonads begins at one-quarter the length of the branchial region. No median gonads. GLOSSOBALANUS ALATUS Na. sp. External Features. In the material collected at Inyack in July 1934 was found a single specimen of a Glossobalanus, broken off near the hind end of the genital region (fig. 22). Two other specimens were found in November 1935, one of which was broken off shortly behind the liver region (fig. 20) and the other in the genital region (fig. 21). The specimens 1 and 3 were cut into serial sections. On account of its rather broad genital pleura it might be thought to be a species of Balanoglossus, but its internal anatomy clearly shows it to be a Glossobalanus. Gl. marginatus seems to have even broader genital wings as can be seen from Meek’s figures, but other- wise the genital pleura are not very pronounced and set off from the body in the species of Glossobalanus. In Gl. mortensent and Gl. ruficollis the pleura are absent altogether. These genital wings are well set off from the body itself (fig. 22); they start shortly behind the collar and reach their maximum width near the end of the branchial region (figs. 36 and 37). In the genital region they show hardly any decrease in size and they end well in front of the liver region, so that, as in some species of Balanoglossus, a transitional region is formed between the genital and hepatic regions. In most species of Glossobalanus the branchial region is short compared with that of Balanoglossus. In Gl. alatus it is extremely short, measuring only 3:3 mm. of a total length of about 28 mm. for the whole branchio-genital region, or 34-5 mm. for the distance between the collar and the liver region (fig. 20). This is relatively and also absolutely shorter than in any other species of The Enteropneusta from Inyack Island, Delagoa Bay. 317 Glossobalanus, with the exception of Gil. ruficollis, in which much larger animal the branchial region measures about 8 mm. in a full- Rees ie rf Wood es WW wis OSs \ a hey. pie ‘ 1 CS ia, a ifesy Seeeuee 22 Figs. 20-22.—Glossobalanus alatus n. sp. 20. Dorsal view, x3. 21. Dorsal 22. Lateral view, x 6. erown specimen; the specimens of Glossobalanus alatus are full grown also, as is shown by the well-developed gonads. Moreover, 318 Annals of the South African Museum. Spengel has shown that during growth the branchial region becomes relatively shorter in comparison with the genital region, but even in the largest specimens of Gl. minutus Spengel found the branchial region to have a length of one-fifth of the whole branchio-genital region; here in Gl. alatus it is only one-tenth. The epithelial ridges of the trunk are not so pronounced as in other species, e.g. Gl. marginatus. The proboscis is 2 mm. long and 3 mm. broad; the collar is 3 mm. long by a greatest width of 3-8 mm. The total length of the genital pleura is 28 mm. and of the transitional region 6-5 mm. The liver region is very short, 3-7 mm. only, and well delimited both at the anterior and posterior ends. Only 14 liver saccules could be seen from the outside, and they are all subequal in size. Over the whole length of the trunk, as far as could be seen from the specimens, a regular row of small depressions is clearly visible at both sides of the ventral nerve cord. Internal Anatomy. Proboscis.—The epidermis of the proboscis is of uniform thickness, no longitudinal striae, as shown in Gl. mortenseni, being visible either on the surface or in cross-sections. The nuclei are scattered in the basal half of the epidermis outside the nerve-fibre layer; a few nuclei may be found in the peripheral part (fig. 23). Many small glandular cells occur in the peripheral part of the epidermis between the ordinary epithelial cells. A thin circular muscle layer, about half as thick as the nerve-fibre layer, is found underneath the basal membrane of the epidermis. Though no membrane could be detected along the inner side of the circular musculature as in Gl. crozeri, there are in this place many cells which, however, do not form a well-defined epithelium. In Gl. mortensena both the membrane and the cells between the circular and the longitudinal musculature are absent. The longitudinal muscle fibres are accumulated in the central part round the proboscis cavity; in the peripheral part a delicate network of connective tissue is found with only a few isolated muscle fibres. The longitudinal muscle fibres show a tendency to form bundles, and these bundles again are more or less radially arranged, especially in the caudal part of the proboscis near the central organs (fig. 23). A dorso-ventral muscle plate, found in other species, is absent in Gl. alatus. The central proboscis cavity is small and is confined only to the posterior part of the proboscis round the central organs. It is also aN 5 WR t jit NS SY 25 26 Fics. 23-26.—Glossobalanus alatus n. sp. 23. Dorso-median segment of a cross- section of proboscis. 24. Cross-section of proboscis neck. 25. The same, a few sections caudal to that of fig. 24. 26. Cross-section of proboscis neck where it fuses with the collar. All x 52. bl, central blood space. cc, collar coelom. cf, connective tissue. dls, dorso-lateral blind-pouch of stomochord. epd, epidermis. epv, efferent proboscis vessel. es, end-sac. gl, anterior extremity of glomerulus. 4k, keel of skeleton. Im, longitudinal musculature. , posterior end of pericardium. pc, posterior end of right dorso-lateral proboscis coelom. pp, proboscis pore. rm, ring musculature. sk, skeleton. st (figs. 24, 25), main lumen of stomo- chord. s¢ (fig. 26), neck of stomochord. sv, sinus venosus. v, proboscis vein. vbs, ventral proboscis blind-sac. vls, ventro-lateral blind-pouch of stomochord. 320 Annals of the South African Museum. very narrow, at the dorsal side it is even quite suppressed, the glomerulus and the pericardium being here in direct contact with the muscular and connective tissues.surrounding the cavity (figs. 23 and 27). The central cavity is therefore crescent-shaped in cross-section. The ventral proboscis septum is very short. It connects with the basal membrane of the epidermis only at the level of the lateral blind-pouches of the stomochord; its attachment to the stomochord extends slightly more frontally. The ventral proboscis artery runs along its anterior edge (figs. 28, 29, and 30). The ventral coelomic blind-sacs of specimen I| are very small, being found in only 7 sections each 10 » thick. When they have disappeared in the cross-sections one finds in the same position between the stomochord and the epidermis a group of cells without a lumen. These cells are con- tinuous with the wall of the blind-pouches. A few sections more caudally, at the anterior end of the skeleton, there appears a median lumen in this group of cells that becomes rather wide and can be considered as a reappearance of the ventral blind-sacs (fig. 24). A few sections farther on, this lumen is divided into two parts by the keel of the skeleton (fig. 25), and when the keel becomes larger this continuation of the ventral blind-sacs ends (fig. 26). In specimen 3, however, an interruption of the ventral coelomic blind-sacs does not occur; after fusing with each other, they continue backwards between the body and the keel of the skeleton and disappear at the level where the keel increases in size. The right dorsal coelomic sac ends blindly as usual, only the left being in communication with the rather irregular, triangular, median end-sac. The proboscis pore is also median (figs. 24, 25, and 26). After the two proboscis veins have fused with each other, they run along the right-hand side of the end-sac and unite with the sinus venosus along the caudal end of the pericardium (figs. 24 and 25). The sinus venosus exhibits no particular features. The anterior end of the stomochord is, except for some protrusions into the central blood space, nearly cylindrical at first (fig. 27); then the dorsal side becomes flatter, so that the stomochord is here broader than high (figs. 28, 29, and 30). The lumen begins as a small cavity near the top and is continuous throughout; in specimen 3 it is a straight canal, but in specimen 1 it is often twisted, so that it may be cut three times in one section. Two ventro-lateral blind-pouches of the lumen occur in the ventral dilation of the stomochord (fig. 24). In specimen 1 each of these blind-pouches has a separate narrow sonnection with the main lumen that has here increased considerably The Enteropneusta from Inyack Island, Delagoa Bay. 321 in size; they are also connected with each other by an equally narrow canal (fig. 25). Asin specimen 3, these narrow canals are interrupted, the pouches are here not connected with each other nor with the main lumen. Besides the ventro-lateral blind-pouches, the stomo- chord of specimen 1 has also dorso-lateral blind-pouches, which otherwise are found only in some species of Balanoglossus (fig. 24). They have a common communication with the central lumen (fig. 25). Caudal to the level of fig. 25 the lumen in the dorsal part of the stomochord disappears, leaving only the main central lumen in the body. The tissue of this dorsal part remains, however, and is con- tinuous with the neck of the stomochord. A new lumen appears here which is connected with the main lumen of the body by a narrow canal (fig. 26). The dorso-lateral blind-sacs are altogether absent in specimen 3. There are hardly any glandular cells in the stomochord except near its opening into the buccal cavity. The pericardium reaches as far anteriorly as the stomochord in specimen 1; in specimen 3 the stomochord extends in front of the pericardium. In the anterior part of the pericardium there is a thick layer of muscle fibres along its ventral wall (fig. 27), but even before the pericardium has become attached to the epidermis these fibres have disappeared (fig. 29). The pericardium of specimen 1 shows a remarkable peculiarity that has never before been recorded in any other enteropneust. Alwaysin Enteropneusta the pericardium is a completely closed vesicle, and no conclusion can be drawn as yet as to whether it originates from the coelom or not. It is therefore of interest that the pericardium of this specimen communicates with — the proboscis coelom. In its anterior part the pericardium is closed off completely from the coelom, and the limiting membrane between the two is uninterrupted (fig. 27). But where the pericardium wall begins to extend towards the dorsal side in order to attach itself to the basal membrane of the epidermis, there appears a large opening in the dorsal wall (fig. 28), which opening extends along the right wall of the pericardium when the latter is in contact with the epidermis (fig. 29). Further caudally the pericardium again is completely closed off from the proboscis coelom (fig. 30). Through this opening the connective tissue and even the longitudinal muscle fibres of the proboscis coelom enter the pericardium. Such an opening has never been described in any other enteropneust before; it may be quite accidental in this specimen, but it is certainly not an artefact. The glomerulus covers and surrounds the anterior ends of both the pericardium and the stomochord. The glomerulus soon becomes 322 Annals of the South African Museum. smaller and then disappears at the dorso-median side (fig. 27). At this point it reaches nearly to the mid-ventral line, but as the stomo- chord is small and the pericardium extends far ventrally, the greater part of the glomerulus covers the sides of the pericardium (figs. 27 and 28). The anterior end of the skeleton is U-shaped in cross-section and surrounds the ventral dilation of the stomochord (fig. 24). It becomes thicker when this dilation gets smaller (figs. 25 and 26), until it becomes more of a solid mass (fig. 31). But also in this region it seems to be composed of two symmetrical halves due to the presence of a dorso- median groove and a deep narrow fissure extending from this groove deep into the body of the skeleton. In specimen 1 it is only in a few sections, near the posterior end of the keel, that the body appears as a real unit. The keel begins shortly behind the anterior end of the skeleton (fig. 25); it is small at first but then suddenly increases in.size and it extends over nearly the whole length of the body; caudally it ends rather abruptly. A very characteristic feature is the separation of the keel from the primary body of the skeleton nearly throughout its whole length. Only with the secondary body, formed by the fusion of the crura, is the keel firmly united. Also, extensions from the collar coelom and side branches from the efferent proboscis vessels penetrate in between the body and the keel (fig. 26). Coelomic tissue also invades the sides of the body itself, thus giving it the appearance of chondroid tissue (fig. 31), which otherwise is only poorly developed. In specimen 1 the secondary body, formed by the union of the crura, begins immediately behind the posterior end of the keel, but in the older specimen 3 the secondary body is much longer and the keel extends along it over a considerable distance. The free crura are short and thick, and extend straight in ventro- lateral direction from the body. They surround the buccal cavity for about half its circumference. Collar.—The five epidermal zones of the collar are well demarcated (fig. 82). The first, rather broad, zone is well supplied with large vacuoles. The nuclei are found just outside the nerve-fibre layer, below the vacuoles, but as many vacuoles reach as far down as the nerve-fibre layer, the nuclear layer is rather irregular. A few scattered nuclei may be seen in the more peripheral part of the epidermis. There are small, short glandular cells filled with a granular substance near the surface, and some narrow elongated glandular cells extending deeper down in the epidermis, the content of which is darkly stained and homogeneous. beets e Pigs tacit ap a Gl We bese os Regs PANG. Me aitigs SU Cag wy Hak Cir = = A, 29 Fics. 27-30.—Glossobalanus alatus n. sp. 27. Cross-section through anterior part of proboscis organs. 28. The same, caudal to that of fig. 27. 29. The same, caudal to that of fig. 28. 30. The same, caudal to that of fig. 29. All x 100. bl, central blood space. epd, basal membrane of epidermis. epv, efferent proboscis vessel. gl, glomerulus. , pericardium. pc, proboscis coelom. st, stomochord. va, ventral proboscis artery. vs, ventral proboscis septum. 324 Annals of the South African Museum. The second zone is narrower than the first, but the epithelium is about twice the thickness of that of the first zone. It consists of very elongated, non-vacuolated cells, the oval nuclei of which are situated in a row about in the middle of the cells. A layer of SOGGY Fic. 31.—Glossobalanus alatus n. sp. Every fifth cross-section, each 10 w thick, of the proboscis skeleton. x55. round nuclei, apparently belonging to the nervous cells, les just outside the nerve-fibre layer. Very small glandular cells with a granular content are found just beneath the outer surface of the epithelium. | The third zone is a little broader and as thick as the second, and has a few vacuoles in the peripheral part of the epithelium. The whole basal half is filled with nuclei, and judging by the number of The Enteropneusta from Inyack Island, Delagoa Bay. 325 nuclei the cells must be very numerous and consequently narrow. This is the main difference between the second and the third zones. There are a few small glandular cells near the free surface. The fourth zone, forming the circular groove, is consequently narrow and consists of a lowepithelium. It is a striking fact that the nerve- fibre layer in this zone is very thin. Just above this layer is a single row of nuclei, and a few more nuclei are scattered throughout the epithelium. The cells must be rather broad here. There are large vacuoles and long thin glandular cells that reach to near the basal membrane. The cilia of this zone are much shorter than those of the other zones. The fifth and last zone is very similar to the first. The epithelium, which is very vacuolated throughout its whole thickness, contains small glandular cells near the surface. The majority of the nuclei are found in the basal part underneath the vacuoles, only a few scattered nuclei being observed nearer the surface. As in other species of Enteropneusta, the dorsal mesentery of the collar begins at the first or the second dorsal nerve root and is complete from there to the posterior end of the collar. No trace of this mesentery could be found in front of the first nerve root. A ventral mesentery, on the other hand, is almost entirely missing, it being only present over a short distance in the posterior end of the collar. Furthermore, there are a few connections between the ventro-median blood-vessels running along the intestinal wall and along the inner side of the epidermis. The collar canals have the usual form with only the dorsal wall folded inwards. As in Gl. mortensent the medullary cord has a continuous central canal, but Gl. alatus differs from that species in the absence of an anterior epidermal pouch dorsal to the anterior neuropore. There are two dorsal nerve roots in specimen 1 and four in specimen 3, and although the central canal sends a branch into these roots, they are solid for the greater part of their length (fig. 33). The supraneural blood-vessels are quite conspicuous. The perihaemal cavities begin at the level of the proboscis pores. Their anterior ends are adjacent to each other except where they are separated by the dorsal blood-vessel. The mesentery separating the two perihaemal cavities from each other is complete over the whole length of the collar. The epithelium of the buccal cavity shows a nearly uninterrupted layer of glandular cells near the free surface. Their contents are VOL. wes Whi < Fe .> \ 4 o7 } P A\ NINO SRA, Wine “a, = < Fics. 58, 59.—Balanoglossus studiosorum n. sp. 58. Cross-section of the dorso- median part of the collar at level where the proboscis neck has fused with the collar. 59. Cross-section of the dorso-median part of the collar shortly in front of the opening of the stomochord into the buccal cavity. Both x 21. bp, ventro-lateral blind-pouch of stomochord. dnr, dorsal nerve root. dv, dorsal vessel. ecm, external circular musculature. elm, external longi- tudinal musculature. env, epineural vessel. ep, end-plate of skeleton. epd, epidermis. ilm, internal longitudinal musculature. hk, keel of skeleton. mc, nerve cord. ph, perihaemal cavities. pp, peripharyngeal cavity. sk (fig. 58), secondary body of skeleton. sk (fig. 59), crus of skeleton. st, stomochord. vbs, ventral coelomic blind-sac of proboscis. VOR. KXMIT, PART 5. 26 358 Annals of the South African Museum. Collar.—The epidermis of the collar shows the different zones as usually present, but I do not find them well differentiated from each other in either species. Concerning the musculature a the collar there is a great difference between B. hydrocephalus and B. studiosorum, as could be expected in accordance with the difference in size and function of this part of the body. In the former species the muscles show the arrangement usually found in the Ptychoderidae. Along the outside of the collar, just under the basal membrane of the epidermis, we find the outer longitudinal muscles, and inside them, at least in the anterior part of the collar, is the layer of the external circular musculature (fig. 60). Both layers, the longitudinal and the circular musculature, are well developed and of about equal thickness in the anterior part of the collar that encloses the proboscis neck; farther caudally the circular musculature disappears altogether and the external longitudinal muscle layer becomes very thin, much thinner than the internal musculature along the wall of the buccal cavity. The latter is, as usual, interrupted along the dorsal side of the collar nerve cord; the muscles of the perihaemal cavities at the ventral side of the nerve cord here supplement the internal longitudinal musculature (fig. 54). Dorsally the internal longitudinal muscle fibres are attached to the proboscis skeleton, and here the collar coelom sends forward two finger-like extensions, filled with longitudinal muscle fibres, that penetrate deeply into the proboscis neck (fig. 51). To the proboscis skeleton are also attached the muscle fibres that spread out along the anterior funnel-like surface of the collar and here form an internal longitudinal muscle layer. This layer is only thin in B. hydrocephalus, at most as thick as either the external longitudinal or external circular layer at the same level. An internal circular layer along the anterior surface of the collar could not be found. The radial muscle fibres, traversing the coelomic cavity between the outside of the collar and the funnel-like anterior surface or, more posteriorly, between the epidermis and the wall of the buccal cavity, are not very numerous and rather thin in B. hydrocephalus. In B. studiosorum the anterior part of the collar in front of the connection between the proboscis and the collar has not the form of a funnel but more that of a hollow cylinder surrounding the proboscis. The external longitudinal and circular muscle layers are slightly stronger than in B. hydrocephalus (fig. 61). The internal longitudinai musculature, however, is very strong, not only in the posterior part of the collar along the buccal cavity but also in the anterior cylinder- The Enteropneusta from Inyack Island, Delagoa Bay. 359 like part. Here the internal longitudinal musculature forms a layer at least as thick as the external longitudinal and circular together. Usually in Enteropneusta the internal longitudinal muscle fibres stop at the vascular fold surrounding the buccal cavity, and the internal fibres along the anterior funnel-like wall of the collar form another set, spreading out from the proboscis skeleton. Butin B. studiosorum the internal longitudinal musculature does not stop at the vascular fold, but is continuous with the layer along the inside of the cylinder. Only at the dorsal side the broad thin crura of the skeleton intersect this muscle layer, and the fibres are attached to both sides of the erura. Finger-like extensions of the collar coelom into the proboscis neck, in which otherwise a great part of the muscle fibres arise, are absent in B. studiosorum, not only because there is no real proboscis neck, but also because of the continuity of the internal longitudinal musculature in front of and behind the attachment of the proboscis. As in B. numeensis and B. aurantiacus, longitudinal muscle fibres as well as radial fibres are also found dorsally to the nerve cord, though the layer here is not very thick. The radial musculature between the inner and the outer wall of the cylinder is much stronger in B. studiosorum than in B. hydrocephalus. Not only are the fibres thicker but they are also more numerous. In the peripheral half of their course they form irregular bundles and there are open spaces between the bundles, whereas in the central half they run independently and are embedded in connective tissue. These open spaces, filled with coelomic fluid which probably is just water, apparently make the action of the musculature easier. In B. hydro- cephalus the whole coelomic cavity is filled with loose connective tissue (fig. 60). The action of the radial musculature seems to be antagonistic to both the circular and longitudinal musculature. Upon contraction of the radial fibres the outer and inner walls of the cylinder approach each other, so that the cylinder either has to become longer or get a greater diameter. The first will happen if the circular musculature is contracted and the longitudinal relaxed, and the second if the longitudinal is contracted and the circular relaxed. As in many Ptychoderidae the dorsal mesentery of B. hydrocephalus is absent in the anterior part of the collar and starts only at the back of the first dorsal nerve root. But whereas in one specimen it 1s complete from there to the end of the collar, in the other sectioned specimen it stops again some way behind the last nerve root and is reduced then to a broader or narrower vascular fold along the dorsal 360 Annals of the South African Museum. side of the collar nerve cord, and only over a short distance near the hind end of the collar is it complete again. In this same specimen also the ventral mesentery is represented by a broad vascular fold hanging down from the wall of the buccal cavity, and it is only in the posterior part of the collar that this fold attaches itself to the basal membrane of the epidermis, thereby forming a complete ventral mesentery. In the other specimen the vascular fold extends over a shorter distance and therefore the ventral mesentery is longer. In B. studiosorum the dorsal mesentery is greatly reduced. In the long anterior part of the collar it is absent altogether, neither does it start at the back of one of the dorsal roots. There is a vascular fold, containing the epineural vessel, caudal to the last nerve root, and this vessel gives off a few side branches that run free through the coelomic cavity to the epidermis; only near the posterior end of the collar a dorsal mesentery is present. In the ventral mid-line a broad and complicated vascular fold hangs down from the wall of the buccal cavity, but it connects with the basal membrane of the epidermis only near the posterior end of the collar, thereby establishing a short ventral mesentery. The tops of the perihaemal cavities, applied to each other, are found at the level or slightly in front of the proboscis pore in B. hydrocephalus (fig. 51). Except for some small openings ventral to the dorsal vessel the two cavities are quite separated from each other (fig. 55). In B. studtosorum the perihaemal cavities also reach the level of the proboscis pore, but on account of the absence of a pro- boscis neck, the pore is situated far forwards in respect to the other proboscis organs; the perihaemal cavities extend in frontal direction quite a distance along the dorsal side of the pericardium and therefore well in front of the sinus venosus (fig. 52). There is hardly a sinus venosus in this species; the dorsal blood-vessel of the collar is directly continuous with the dorsal vein of the proboscis that is situated - between the anterior ends of the perihaemal cavities and splits up into two dorso-lateral veins in front of these cavities. Near the posterior end of the pericardium this dorso-median vessel gives off two branches that connect along the sides of the pericardium with the central blood space. As in B. hydrocephalus the perihaemal cavities of B. studiosorum are only connected with each other along a narrow opening at the ventral side of the dorsal vessel (fig. 58). In B. hydrocephalus the peripharyngeal cavities, surrounding the buccal cavity, are quite separated from the perihaemal cavities : : BD» y a K\ 7 itty AW ow oO ce) coca ot Y AN WS sm N ( Tanner A oO cr wn 63 Fie. 60.—Balanoglossus hydrocephalus n. sp. Part of a cross-section through the x 50. Fics. 61-63.—Balanoglossus studiosorum n. sp. x 50. 62. Cross-section of the collar anterior region of the collar. 61. Part of a cross-section through the anterior region of the collar. nerve cord and surrounding structures. x36. 63. Median section of the proboscis and collar. x 3. be, buccal cavity. bm, basal membrane of the epidermis. cc, collar coelomic cavity. cts, collar-trunk-septum. dv, dorsal vessel. ecm, external circular musculature. elm, external longitudinal musculature. env, epineural vessel. g, branchial basket pulled high up into the collar. «im, internal longitudinal musculature. nc, nerve cord. p, pericardium. pc, proboscis coelom. ph, perihaemal cavity. pp, peripharyngeal cavity. rm, radial muscle fibres. sk, skeleton. st, stomochord. 362 Annals of the South African Museum. (fig. 54). They extend to the dorsal mid-line along the ventral side of the perihaemal cavities, and here a distinct limiting membrane can be seen between the peripharyngeal and perihaemal cavities. In B. studiosorum, on the other hand, the peripharyngeal cavities have fused with the perihaemal cavities (fig. 62). The circular musculature of the former can be seen running along the ventral wall of the latter. In one of the two sectioned specimens the peripharyngeal cavities widen out along the lateral sides of the perihaemal cavities, and some longitudinal muscle fibres are found here in the peripharyngeal cavities. A distinct limiting membrane separates the cavities here, but this membrane does not reach the ventral wall. The collar pores of both species have the usual appearance found in the Ptychoderidae, so that their wall is not very thick and the nuclei in the cells form approximately a single row. The dorsal fold of the pore is apparent in both species, only in B. hydrocephalus this fold is much deeper than in B. studiosorum. An anterior epidermic pouch, dorsal to the place of the anterior neuropore, is quite distinct, although shallow, in B. hydrocephalus (fig. 55), but in B. studiosorum there is no trace of such a pouch. As in nearly all species of Balanoglossus a central canal in the collar nerve cord is absent, but there are medullary cavities, and these are wider and more numerous in B. hydrocephalus than in B. studiosorum. Also at the anterior end of the cord there is no trace of a canal, and therefore also the anterior neuropore is absent. A central canal with a posterior neuropore is present in both sectioned specimens of B. studiosorum and in one of B. hydrocephalus; in the other specimen of the latter species the nerve cord is solid up to its posterior end. This canal is present before there is any indication of the splitting up of the collar nerve cord into the dorsal nerve of the trunk and the circular nerves, therefore it cannot be considered to be a simple epidermic depression. In the two specimens of B. hydrocephalus there are 3 and 4 dorsal nerve roots, all of which are rather thin. The fourth root of the second specimen is very far back near the caudal end of the collar. In one specimen of B. studiosorum I find 2 nerve roots, the anterior of which is very thick and has a wide central canal that ends blindly at both sides. In the other specimen there are 3 nerve roots, the first and last of which also have a similar central canal. The first nerve root is situated far anteriorly, at the anterior end of the nerve cord. In a species like B. hydrocephalus with a short collar, the anterior surface of which is funnel-shaped, there is no doubt about the position The Enteropneusta from Inyack Island, Delagoa Bay. 368 of the mouth opening. It is found at the bottom of the funnel, where the proboscis neck is fixed to the collar or more exactly where the stomochord opens into the buccal cavity. Here the epidermis of the anterior surface of the collar is continued into the wall of the buccal cavity, and there is a sudden change in the character of the epithelium. Whereas the epidermis is very thin and the nuclei of the cells are scattered over the whole thickness of the epithelium, the wall of the buccal cavity is very thick and the nuclei are found in a thin layer near the surface; a few very small glandular cells, stained very darkly by haematoxylin, are found near the surface (fig. 54). In B. studi- osorum, on the other hand, the proboscis is connected with the collar about in the middle of the latter, and the part of the collar in front of this connection is not funnel-shaped but cylindrical (fig. 63). In general form, therefore, there is hardly any difference between the part of the collar in front of and behind the connection, and it would at first sight be possible to maintain that the mouth opening is situated at the anterior end of the collar. But in B. studiosorum there is a sudden change in the character of the epithelium at the level of the opening of the stomochord, and this clearly indicates the position of the mouth opening. Not only is the wall of the buccal cavity again much thicker than the epidermis at the inside of the anterior end of the cylinder, but the nerve-fibre layer, which as usual is rather thick in the epidermis of the collar, suddenly becomes very thin just where the buccal epithelium begins. _ Trunk.—In many Enteropneusta, e.g. B. stephensoni, there is a more or less deep groove along the ventral and dorsal mid-line in the branchial region, and the dorsal and ventral nerve cords are found under these grooves. In B. hydrocephalus there is such a groove in the dorsal mid-line, but the nerve cord, with the cell layer covering it, is at least as thick as the epidermis alongside of it (fig. 64). In the ventral mid-line there is not even a groove at all, and the nerve cord protrudes. On the other hand, the ventral nerve cord of B. studiosorum is found in a median groove and, together with the covering cell layer, is much thinner than the adjacent epidermis (fig. 65). There is also a shallow groove in the dorsal mid-line in this species, but the nerve cord here is not thinner than the epidermis at its sides, and as the glandular cells in the epidermis are swollen considerably, the nerve cord may even protrude here, when the glandular cells are not hypertrophied. The epidermis of the branchial region contains many glandular cellsin both species, some of which are darkly stained by haematoxylin 364 Annals of the South African Museum. CON NES Cf eens y Va RS / oe ue HN Ny ANAT Fic. 64.—Balanoglossus hydrocephalus n. sp. Cross-section of the branchial region, partly (branchial basket and gonad at side) composed from a number of sections. x 12. bp, branchial pore. bs, branchial sac. bt, branchial tongue. dg, dorsal branch of gonad. dn, dorsal nerve cord. dph, dorsal pharynx. dv, dorsal vessel. er, epibranchial ridge. gp, genital pore. lg, lateral branch of gonad. Im, longitudinal musculature. Js, lateral septum. br, parabranchial ridge. ppbv, parabranchial vessel. s, branchial septum. vg, ventral branch of gonad. vn, ventral nerve cord. wph, ventral pharynx. vv, ventral vessel. _——~ The Enteropneusta from Inyack Island, Delagoa Bay. 365 and are found in the external one- or two-thirds of the epidermis only. In B. studiosorum especially these cells form a nearly continuous layer in the epidermis. In this species deep slits occur in the epidermis running at regular intervals from fronto-dorsally to caudo-ventrally Fic. 65.—Balanoglossus studiosorum n. sp. Cross-section of the branchial region, partly (branchial basket, blind-sac at right side, and gonad at left side of figure) composed from a number of sections. x 8. bl, blind-sac of branchial sac. bp, branchial pore. 6s, branchial sac. bt, branchial tongue. cm, circular musculature. dn, dorsal nerve cord. dph, dorsal pharynx. dv, dorsal vessel. er, epibranchial ridge. gp, genital pore. lg, lateral gonad. Im, longitudinal musculature. br, parabranchial ridge. s, branchial septum. vg, ventral branch of gonad. wn, ventral nerve cord. vph, ventral pharynx. vv, ventral vessel. (figs. 65 and 67). Darkly stained glandular cells are found along these slits, and these reach the basal membrane because the nerve-fibre layer underneath the slits is reduced to a minimum. These slits are so deep that the basal membrane of the epidermis assumes a wavy course. There are also many eosinophil cells in both species, and these are found throughout the whole thickness of the epidermis. 366 Annals of the South African Museum. The dorsal side of the body and the inside of the genital wings are even richer in glandular cells than the outside of the wings and the ventral side of the body, and at the dorsal side the cells stained with haematoxylin extend over the whole thickness of the epidermis. The external circular muscle layer is better developed in B. studi- osorum than in B. hydrocephalus. It is not only slightly thicker in the former species, but it is also continuous and quite distinct along the inside of the genital wings, whereas in B. hydrocephalus it becomes so thin that only here and there can a few fibres be distin- guished. The longitudinal musculature also is better developed in B. studiosorum than in B. hydrocephalus, especially on the ventral side of the body and the outside of the wings. On the dorsal side of the body of B. studiosorum the layer of longitudinal muscle fibres reaches about half the thickness of that of the ventral side of the body, whereas in B. hydrocephalus it is of about equal thickness dorsally and ventrally. As usual, the longitudinal musculature is poorly developed on the inside of the wings in both species. The muscle fibres in B. studzosorum are thinner but by far more numerous than in B. hydrocephalus. : The dorsal mesentery presumably is absent in the greater part of the branchial region in both species. As, however, the whole branchial region was not sectioned, this cannot be said with certainty, but near the hind end of the branchial region the dorsal mesentery is complete. The dorsal vessel in the branchial region runs along the inside of the epidermis, and the connections between this vessel and the branchial vessels go freely through the coelomic cavity. The ventral mesentery is complete in B. hydrocephalus, but in B. studiosorum it is not complete throughout; sometimes it is attached to the alimentary canal only, leaving an opening between the right and left coelomic cavities near the epidermis, and sometimes it is attached to the basal membrane of the epidermis only and free from the alimentary canal. In B. studiosorum the ventral, digestive part of the pharynx is wider than the dorsal branchial part (fig. 65). The wall of the ventral pharynx is very thick throughout, and much thicker than the para- branchial or limiting ridges between the dorsal and ventral parts. Therefore the limiting ridges are not very conspicuous in this species. On the other hand, in B. hydrocephalus the limiting ridges are well pronounced (fig. 64). In the smaller of the two specimens of this species the dorsal pharynx is about twice the size of the ventral, but in the larger specimen the ventral part has at least the same dimensions The Enteropneusta from Inyack Island, Delagoa Bay. 367 as the dorsal (fig. 64). The wall of the ventral pharynx exhibits a noteworthy peculiarity in B. hydrocephalus. According to a figure of B. clavigerus, given by Spengel, the dorso-lateral walls of the digestive pharynx are much thinner than the ventro-lateral walls, the whole of this part of the alimentary canal being about quad- rangular, and the transition between the thicker and thinner epithelium is very sudden at the lateral edges of the canal. In his description of the species Spengel does not mention this difference in thickness. A similar condition occurs in B. hydrocephalus, in which the ventral pharynx is also nearly quadrangular in cross-section. But B. hydro- cephalus differs from B. clavigerus in that the ventro-lateral walls are thin and the dorso-lateral thick (fig. 66). In the ventral mid-line is a groove and here the wall is thicker than the ventro-lateral wall. This difference in thickness was especially clear in the small specimen; in the bigger one the dorso-lateral walls are also rather thin except near the limiting ridge and along the lateral edges (fig. 64). At regular distances a narrow strip of thicker epithelium crosses the thin ventro-lateral wall between the lateral edge and the ventral mid-line. This thicker epithelium does not protrude into the lumen of the pharynx but into the coelomic cavity; it resembles the beams supporting a floor. In the larger specimen it appeared that these strips are in reality deep and very narrow grooves, but as they run in the transverse plane it is difficult to see this in transverse sections. In both species the branchial pores are too small to be visible externally, therefore it was not possible to count the number of gills directly. But by counting the number of pores in the serial sections it was found that in B. studiosorum there occur 32 pores over a length of 4-5 mm. of the branchial region. In the larger specimen of B. hydrocephalus 21 pores were counted in a length of 4 mm., but in the smaller specimen the gills are by far more crowded as 21 pores were counted in only 1-65 mm. This cannot be due to contraction only, although it may have been possible that the smaller specimen was more contracted than the larger. Apparently the branchial region extends during growth, not only by an increase in the number of gills at the posterior end of the row, but also by an increase in size of the individual gills, which results in an increase in the distance between the succeeding pores. The estimation of the total number of gills from the numbers given above is rather un- reliable, but itis the only method of approach. In the well-preserved and large specimen of B. hydrocephalus the branchial region has a total length of 58 mm., and with 21 pores in 4 mm. this would give 368 Annals of the South African Museum. a total number of about 300 gills. In B. studiosorum it becomes even more difficult, because, as previously mentioned, it is hardly possible to distinguish the posterior end of the branchial region in this species. In a large specimen, in which this could be done, this region was only 22 mm. long, and that would give a total number of about 160 gills. The epibranchial ridge of B. hydrocephalus is made up of a high, convex epithelium (fig. 64). In the middle of the ridge is a number of small glandular cells stained by haematoxylin; in the large specimen this row of glandular cells is broader than in the small one. This central strip with glandular cells is flanked by rows of undifferentiated high epithelial cells, and here the nuclei are situated in a regular row near the free ends of the cells, with some scattered nuclei in the basal half of the epithelium. In B. studiosorum the epibranchial ridge is much thinner than in B. hydrocephalus and it is hardly convex (fig. 65); glandular cells occur over the whole breadth of the ridge, and all nuclei are scattered. As usual, the tongue protrudes farther into the pharyngeal cavity than the septum. The septa are narrow, and the epithelium on their back is lower than that on the backs of the tongues; in B. hydro- cephalus the thickness of this epithelium on the tongues reaches even twice the thickness of that on the septa; in B. studiosorum the difference isless. In conformity with the structure of the epibranchial ridge, glandular cells are found over the whole breadth of the back of the tongue in B. studiosorum, whereas in B. hydrocephalus there are two rows of glandular cells separated from each other by a central row of undifferentiated cells. On the narrow backs of the septa a few glandular cells are also found. The number of synapticula is the same in both species; it is about 20. The branchial sacs of B. hydrocephalus are rather spacious, and although the wall between these sacs and the coelomic cavity is very much folded, ventral blind-sacs to the branchial sacs are absent (fig. 64). In B. studiosorum, as in B. numeensis and some other species, these blind-sacs are well developed and reach the ventral wall of the digestive part of the pharynx (fig. 65). Also, at the dorsal side of the branchial sacs blind-sacs are formed that extend into the base of the genital wings, lying there in between or external to the gonads (fig. 68). Small glandular cells occur in the thin walls of the branchial sacs of both species; in B. hydrocephalus these glandular cells appear to be more numerous than in B. studiosorum. The 369 The Enteropneusta from Inyack Island, Delagoa Bay. ‘[OSSOA [RIZUOA “Aa ‘xudreyd peryuoa “yda ‘peuos Jo yoursq [vryuoa “ha “ows-pul[q [Bestop oy} pue UOIsue}xo [e107R[-OSIOp OY} W9oArjod uotoouuo0d [edjzUOA “92 “OATVA [BI}UOA ‘a -‘aSpla perpouvaqeaed “gd *vanydes yeroyey ‘S7 ‘gA0019 osaoAsueT) ‘46 “[OSSeA [esIOp ‘ap ‘xudieyd JerjUoA JO UOISU04xXO ['B107e]-OSLOp ‘yp + “peuos jo youemq jesaop ‘bp ‘assoa [einssTuULoo ‘aga ‘onsuo4 TeIpouriq 4qQ ‘oes [VTYOUBIG ‘sq ‘*peueo [erpourrqysod jo ovs-pul[q [es1op 19 "6X ‘po}jturO 918 SSUIM [eyIUOD oY} SUOTFISUBIY [ejtues-oryouvsq oY} JO [OAS] OY} 7B uorqyoos-ssorg, “ds “U wnsosoLpnys snssopbounjog— L9 “SI ‘gp x ‘*xudseyd jo qaed [eIyUIA OY} JO WOTPDOS-SSO1) ‘ds ‘u snpoydaooouphy snssopbouvjog—'99 “OT 99 \ Stg&N~N GN UR pe. Ni ‘ a ; . if wy, fons oe 370 Annals of the South African Museum. branchial pores of B. hydrocephalus are very small, and a short canal, piercing through the longitudinal musculature, connects the branchial sac with the pore. In B. studiosorum such a canal is absent; the sac itself touches the epidermic basal membrane directly, and the Paes on the whole are wider than in B. hydrocephalus. In B. studiosorum the first two gills, with the collar canal, open to the exterior by acommon pore. The peripheral part of the septum between the first two gills is missing, and in this way these two sacs communicate with each other. Similarly in B. misakiensis the first four gill-sacs are connected and have a common pore. In B. hydro- cephalus, as in most species of this genus, the first branchial sacs do not communicate with each other and each sac has its own pore. As in B. carnosus, the first pore in both species is situated at the inside of the incipient genital wing, whereas in B. capensis and some other species this pore is at the outside of the wing and all the sub- sequent pores are at the inside. The postbranchial canal exhibits some features of interest, all the more so as the two species differ from each other in this respect. In B. hydrocephalus the branchial pores are at the bottom of a deep groove, except at the posterior end of the branchial region, where the pores are higher up at the medial side of the groove. This branchial groove comes to a sudden end; it is even slightly prolonged at its caudal end as a small blind-sac, in which the last gill opens. Near the posterior end of the row the gills gradually become smaller, as a result of which the epibranchial and parabranchial ridges approach each other and finally fuse behind the last gill-pore. As the para- branchial ridges are lying close to each other in the branchial region, a narrow median slit-like extension of the wide ventral pharynx is the result of the fusion of the three ridges, much in the same way as described and figured by Maser for B. clavigerus. This narrow dorso- median part of the gut corresponds to the postbranchial canal of other Ptychoderidae. The wall of this postbranchial canal retains the same structure as the parabranchial ridges and is therefore much thicker than the wall of the ventral pharynx or oesophagus. In B. clavigerus the postbranchial canal has a small blind-sac dorsal to the posterior gills; in other Ptychoderidae, as in most species of Glossobalanus, this blind-sac is much larger than in B. clavigerus. In B. hydrocephalus there is no trace of such a blind-sac. In the posterior part of the branchial region the lateral edges of the ventral pharynx turn in a dorsal direction and form two narrow dorsal extensions of this part of the gut as is shown in all Ptychoderidae. The Enteropneusta from Inyack Island, Delagoa Bay. 371 These extensions become higher and higher, then their tops turn suddenly in medial direction and fuse with the dorsal end of the post- branchial canal. In this way there is formed on each side a coelomic blind-sac that extends into the oesophagus and that probably acts as a valve in the alimentary canal. The commissural vessels exhibit nothing of particular interest; they are simple transverse connections between the lateral and dorsal blood-vessels, running in the lateral septum and over a short distance, dorsal to the postbranchial canal, free through the coelom. In B. studiosorum the hind end of the branchial region and the postbranchial canal are more complicated than in B. hydrocephalus. Further, the deep branchial groove does not gradually become shallower towards the posterior end of the branchial region, but it is suddenly separated from the surface and from there it continues in a caudal direction over a rather long distance in the form of a blind-sac. Quite a number of gills open into this blind-sac. In fig. 67, where some of the last gills are still visible, the blind-sacs have already disappeared; farther frontally they le between the gills and the dorso-lateral extensions of the ventral pharynx. The postbranchial canal has a well-developed blind-sac, lying dorsal to the posterior gills and extending slightly beyond the caudal end of the branchial region. This blind-sac is clearly defined from the epibranchial and parabranchial ridges, because in the latter small glandular cells are found near the surface of the epithelium, whereas the epithelium of the blind-sac is darkly stained by numerous very narrow glandular cells extending over its whole thickness. The dorso-lateral extensions of the ventral pharynx form ridges increasing in height as in B. hydrocephalus; finally, they nearly reach the basal membrane of the epidermis. Already in the posterior end of the branchial region the dorsal blind-sac makes a connection with the two dorso-lateral extensions; this connection is, however, made at about the middle of the height of the extensions, so that a considerable part extends freely farther dorsally. By these connections there is formed a pair of coelomic blind-sacs, in which a part of the posterior gills is still visible (fig. 67). Just behind the branchial region the dorsal ends of the extensions turn to the mid-line and fuse with the top of the dorsal blind-sac, so that another pair of coelomic blind-sacs, lying dorsally to the first, is formed. Both pairs of coelomic blind- sacs protrude backwards into the oesophagus, but as they are small and the oesophagus here is wide, they can hardly act as valves. In B. hydrocephalus these valves really block up the whole lumen of the 372 Annals of the South African Museum. oesophagus. Hach commissural vessel makes two connections with the dorsal blood-vessel. | As only one specimen of each species was cut into serial sections at the level of the postbranchial canal, it is difficult to be certain about the specific value of the form of the postbranchial canal and the valves. B. hydrocephalus is almost in complete agreement with B. clavigerus in this respect, and there seems to be no doubt that the form of the postbranchial canal, as described above, is characteristic for the species. The two pairs of coelomic blind-sacs, as found in B. studiosorum, are not known to exist in any other species. But the animal is quite symmetrical, which makes it at least likely that the two pairs of blind-sacs are a normal feature and of specific value. The lateral septum of B. hydrocephalus extends very far forwards into the branchial region (fig. 64). In most species of Balanoglossus the lateral septum extends only over a very short distance into the branchial region; for instance, in B. capensis 1t stops already at the level of the 10th gill-slit from behind. In B. stephensoni, with its very short branchial region, the lateral septum reaches about the middle of that region. But in B. hydrocephalus this septum extends over nearly the whole length of the branchial region. As previously stated, there may be 300 gills in this species, and I found the anterior end of the dorsal coelomic chamber and therefore of the lateral septum at the level of the 13th branchial pore. How far the septum extends into the branchial region of B. studiosorum has not been ascertained. In the two series of sections of the anterior part of the body, which extends a good distance into the branchial region, no lateral septum is present, but it was found over the whole length of a series of the branchio-genital transitional zone. The row of gonads in B. hydrocephalus begins already a short distance behind the collar, the first genital pore being found near the 8th branchial pore. The gonads of the branchial region are very slender, having much the same form as those of B. clavigerus; they do not nearly fill the open coelomic space of the genital wings (fig. 64). The lateral lobe of the gonad nearly reaches the edge of the genital wing, therefore the peripheral part of the wing is rigid and not turned in as in B. stephensona and especially in Glossobalanus marginatus, as described by Meek. The ventral genital lobe goes down into the body cavity to about half of the height of the branchial basket. A distinct though narrow lumen could be seen in the ovary of the larger specimen, containing many ripe eggs. In the small specimen only a few egg-cells were present, but a great amount of yolk filled The Enteropneusta from Inyack Island, Delagoa Bay. 3738 the whole gonads, so that no lumen was visible. In the posterior part of the branchial region the gonads are much larger and fill up the whole coelomic space in the wings. The gonads are also lobed here, whereas more anteriorly they consist of simple lateral and ventral branches. But secondary genital pores, as are found in some other species of Balanoglossus, are absent in B. hydrocephalus. In most species of Balanoglossus dorsal branches of the gonads, lying in the dorsal coelomic chamber, occur only in the genital region and in the most posterior part of the branchial region. But just as the lateral septum extends so far forward in B. hydrocephalus, the gonads have dorsal branches nearly from the anterior end of the row (fig. 64). The first dorsal branch, although small, was found at the level of the 25th branchial pore. As the genital pores are rather high up on the inside of the genital wings and consequently the lateral septum is rather broad and the dorsal coelomic chamber rather wide, these dorsal branches of the gonads soon reach a considerable size. The gonads have here three branches, a dorsal, a lateral, and a ventral. In other species of Balanoglossus these three branches can be found in the genital region. But in this region only two branches, the dorsal and the lateral, occur in B. hydrocephalus. In cross-sections the lateral septum from its insertion to the intestinal wall can be seen going in lateral direction, and then, closely applied to the inside of the external longitudinal musculature, it enters the wing and follows this course up to the level of the genital pore. Here it turns to the inner side of the wing, where it is inserted along the row of genital pores. Therefore branches of the gonads are found only peripherally to the genital pores and in the dorsal coelomic chamber. The row of gonads begins in B. studiosorum at the same level as in B. hydrocephalus, i.e. near the 8th or 9th branchial pore. A very small gonad was observed in front of the first genital pore in one specimen, but this gonad does not seem to open to the exterior, at least no pore could be found. The ovaries of the branchial region are much stouter than in B. hydrocephalus; there is hardly any room left in the coelomic cavity between the gonads and the longitudinal musculature (fig. 65). But the testes are more slender and repeatedly branched, whereas the ovaries form only blunt lobes. Although the gonads do not quite reach the peripheral end of the genital wings, the top of the wing, nevertheless, is not turned inwards like a loose flap; because of the thickness of the longitudinal muscle layer this part of the wing is rigid. In a ventral direction the gonads reach nearly to the level of the digestive part of the pharynx. But in the frontal end VOL. XXXII, PART. 5. 27 374 Annals of the South African Museum. of the row the ventral branch of the gonad is small and short, as its place is occupied by the blind-sacs of the gills (fig. 65). In the female I find the genital pore much nearer the base of the wing than in the male or than in B. hydrocephalus. Secondary genital pores could not be found in the region of the branchio-genital transi- tion. In conformity with most species of Balanoglossus, but distinct from B. hydrocephalus, the gonads have well-developed ventral branches here; the lateral septum could be followed from its insertion in the intestinal wall peripherally, where it lies between the ventral and dorsal branches of the gonads. When the testes are quite ripe and full of spermatozoa they exhibit a remarkable form in B. studiosorum (figs. 68 and 69). They are profusely branched, and each branch or duct ends in a swollen vesicle filled with spermatozoa. These end vesicles are all turned towards the medial side of the wing, the ducts are along the lateral side applied to the external longitudinal muscle layer along the lateral wall of the genital wing. The yolk cells are well separated from the primary genital cells; the former are found in the ducts and the part of the wall of the end vesicle near the ducts, the rest of the wall of the end vesicles being formed by genital cells. In B. misakiensis a similar clear separation between the yolk and the genital cells was found, but here the yolk cells are away from the ducts, and the genital cells form the wall at the side of the ducts. As I wanted to make out what causes the regular dark rings that make the abdominal region of B. hydrocephalus so conspicuous, a part of this region with the caudal region was cut into serial sections. It was found that the epidermis shows here alternating regions that are quite different from each other (fig. 70). In some parts the epidermis is very thin and composed of almost cubical cells. Suddenly these very thin stretches become continuous with parts in which the epidermis is much higher, about five times as high as the thin parts. In these thick regions of the epidermis the nuclei form a dense regular layer in the middle of the cells, and many glandular cells are found in the outer half of the epidermis. The circular musculature under the basal membrane of the epidermis is quite distinct and about as thick as the thin regions of the epidermis; it is continuous and of equal thickness over its whole course. The longitudinal musculature exhibits a remarkable arrangement. It forms a rather thick layer underneath the thick epidermal parts, but it is absent underneath the thin epidermis. It does not even quite reach the edges of the thick epidermal zones. The dark rings, visible externally, are most oo RAW cA Nir OSES: oe tM co ey Ub U4 ny He msscnn . ) i hy, iar ein sanytttens Dine “J HB peer a iis ~ I wy" it "(Nl NN \ IK \\ N NNN ‘AN \ y y MLTR Figs. 68, 69.—Balanoglossus studiosorum n. sp. 68. Cross-section of a genital region. x 30. circular musculature. epithelium. gp, genital pore. sp, spermatozoa. Pp wing of male. x10. 69. Part of testis of fig. 68 further enlarged. x 50. bl, blind-sacs of the branchial sacs. bp, branchial pore. cg, ciliary groove. dv, Fic. 70.—Balanoglossus hydrocephalus n. sp. Cross-section of the abdominal cm, dorsal vessel. g, testis. ge, germinal lm, longitudinal musculature. pg, pygochord. vv, ventral vessel. y, yolk cells. 376 Annals of the South African Museum. likely formed by the thick epidermis with its many glandular cells; therefore the longitudinal musculature is broken up into rings, which are separated from each other by stretches without these muscle fibres. A dorsal mesentery is entirely absent in the caudal end of the animal, and the place of the ventral mesentery is occupied by the pygochord. So the latter structure is not confined to the caudal region, but is also found over some length in the abdominal region. It is not separated from the intestinal wall by a limiting membrane, and it exhibits much the same form as that described by Dakin in Ptychodera flava (Pt. pelsarti) but differs in that it is not swollen ventrally. The pygochord does not show any interruptions, and it is only absent at the extreme caudal end of the animal. A ciliated groove is conspicuous at one side of the intestine only, occurring for some distance on the right and for some distance on the left side. On the other side it is only indicated by a thickening of the intestinal wall, where the free surface of the epithelium is undulated. This thicker undulating epithelium extends into the caudal region, and finally near the anus it forms the whole wall of the intestine. Diagnosis of Balanoglossus hydrocephalus. Total length about 30 cm. Proboscis conical, long, up to 20 mm., collar less than half the length of the proboscis, and at most as long as it is broad; branchial region and genital region of about equal length, 30-40 mm.; liver region long, as long as or longer than branchial region. Posterior end of branchial basket well indicated. Genital wings fused with posterior rim of collar, wings narrow at their anterior end, so that their free edges are separated by a short distance from each other; they reach their maximum breadth near the hind end of the branchial region, and decrease gradually from there; they end abruptly at the beginning of the liver region; no transitional zone present. The liver region is well delimited anteriorly; the saccules in two regular rows, each saccule extends over the whole breadth of the row. At the anterior end the liver saccules soon reach their maximum size and in the first half of the region they hardly decrease in size; in the second half they gradually become smaller, but they are closely applied to each other; itis often hard to determine the hind end of the liver region; there may be about 200 liver saccules. Abdominal region with regular dark rings. Proboscis.—Circular muscle of equal thickness as nerve-fibre layer; The Enteropneusta from Inyack Island, Delagoa Bay. 377 it does not form a sphincter at the base of the proboscis. Longitudinal musculature strongly developed, fibres more numerous in peripheral than in central part of proboscis. Dorso-ventral muscle-plate well developed, but not extending in front of the central proboscis organs. Large open proboscis cavity in posterior two-thirds of proboscis. Ventral septum extends to end of coelomic blind-sacs. Only left dorsal coelom opens by proboscis pore. No central lumen in body of stomochord ; large ventro-lateral blind-pouches not connected with each other; no dorso-lateral blind-pouches. Glomerulus large, extend- ing in front of stomochord and caudally to level of ventro-lateral blind-pouches; the two halves separated from each other by the dorso-ventral muscle-plate; dorsal glomerulus present. Skeleton well developed, with long and prominent keel and thick end-plate. Collar.—External longitudinal and circular musculature of about equal thickness. Longitudinal musculature along anterior surface of collar thin, and circular muscles here absent. Radial muscle fibres thin and not very numerous. Coelomic cavity almost entirely filled with connective tissue. Dorsal mesentery starts at first dorsal nerve root; from there it may be complete or not to the end of the collar. Ventral mesentery not complete, of varying length. Perihaemal cavities reaching to proboscis pore, almost entirely separated from each other. Peripharyngeal cavities separated from perihaemal cavities. Collar pores with thin epithelium and deep dorsal fold. Nerve cord with anterior epidermic pouch, no anterior neuropore, posterior neuropore may be present, with numerous medullary cavities and 3 or 4 dorsal nerve roots. Trunk.—No grooves in dorsal and ventral mid-line of branchial region. Hxternal circular muscle layer thin, hardly present on inside of genital wings. Longitudinal musculature on dorsal side of body of about equal thickness as that of the ventral side; muscle fibres thick and not very numerous. Dorsal mesentery absent in anterior part of branchial region, present in posterior part; ventral mesentery complete. Parabranchial ridges well pronounced; ventral pharynx smaller than or equal to dorsal pharynx; ventral pharynx with thin wall and thick cross-bars formed by deep grooves along ventro-lateral sides. Number of gills probably about 300; 20 synapticula. Epi- branchial ridge high and convex with a strip of glandular cells in the middle. Tongues more protruding than septa; no ventral blind-sacs to the branchial sacs; branchial pores very small; first branchial sacs open to the exterior independently of each other. Postbranchial canal without dorsal blind-sac. Lateral septum extends nearly over eee 378 Annals of the South African Museum. whole branchial region, begins near 13th branchial pore. First genital pore near 8th branchial pore; lateral lobe of gonads nearly reaches the edge of the genital wing; no secondary genital pores; in the branchial region the gonads have dorsal, ventral, and lateral branches; no lateral branches in genital region. Diagnosis of Balanoglossus studiosorum. Total length about 30 cm. Proboscis short, extending a few mm. only from the collar; collar long, up to 20 mm., and about twice as long as broad; branchial region short, 22 mm., but hind end of branchial region hardly distinguishable; whole branchio-genital region about 150 mm. long; liver region short, 10-20 mm. long. Genital wings fused with posterior rim of collar; they are here already so broad that their free edges touch each other; they increase rapidly in breadth, keep the same breadth over the greatest length, and become narrower only near their hind end; wings end abruptly just at begin- ning of liver region; no transitional region present. Liver saccules of irregular form; anterior end of liver region well defined; posteriorly the broad row of saccules ends suddenly but may be followed by a row of small, regularly arranged saccules that gradually become smaller. | Proboscis.—Circular musculature very thin, about one-third of nerve- fibre layer; it does not form a sphincter at the base of the proboscis. Longitudinal musculature poorly developed, concentrated along central core of connective tissue. Dorso-ventral muscle-plate almost entirely absent. Proboscis cavity greatly reduced, only in ventral blind-sacs and near proboscis pore. Ventral septum short and not reaching the end of the ventral blind-sacs; the latter large and extending beyond the end-plate of the skeleton; only the left dorsal coelom opens by proboscis pore. No central lumen in body of stomo- chord; ventral blind-pouches small, not connected with each other; dorso-lateral blind-pouches may be present; neck short and of irregular form. Glomerulus small; no dorsal glomerulus. Skeleton poorly developed; end-plate thin; body absent or very small; keel short but prominent; crura broad, thin, anteriorly concave. Collar.—External longitudinal and circular musculature well developed and of equal thickness. Internal longitudinal musculature very strong, at least as thick as external longitudinal and circular together in anterior part of collar. Radial muscle fibres numerous and thick. Coelomic cavity rather open; connective tissue only in The Enteropneusta from Inyack Island, Delagoa Bay. 379 central part. Dorsal and ventral mesenteries only present near posterior end of collar. Perihaemal cavities reaching to proboscis pore, almost completely separated from each other. Peripharyngeal cavities not separated from perihaemal cavities. Collar pores with thin epithelium and small dorsal fold. Nerve cord without anterior epidermic pouch; without anterior but with posterior neuropore; with few medullary cavities and 2-3 dorsal nerve roots. Trunk.—A groove in ventral mid-line of branchial region present, but hardly in dorsal mid-line. External circular muscle layer well developed also on the inside of the wings. Longitudinal musculature on the dorsal side of the body of about half the thickness of that of the ventral side; muscle fibres thin and very numerous. Dorsal mesentery absent in anterior part of branchial region, present in posterior part; ventral mesentery not complete everywhere. Para- branchial ridges thin; ventral pharynx wider than dorsal pharynx; ventral pharynx with very thick wall. Number of gills probably about 160; 20 synapticula. Epibranchial ridge thin and rather flat, with glandular cells over whole breadth. ‘Tongues more protruding than septa; branchial sacs with large ventral blind-sacs, extending also dorsally into base of genital wings; branchial pores rather small; first 2 gills with common pore. Postbranchial canal with dorsal blind-sac; 2 pairs of valves formed by postbranchial canal. First genital pore near 8th or 9th branchial pore. Testes much branched, ovaries lobed. Lateral branch of gonad does not quite reach the top of the genital wings; no secondary genital pores; gonads with dorsal, lateral, and ventral branches in the genital region. PTYCHODERA FLAVA KEschscholtz. Ptychodera flava, ubiquitous in the tropical Indo-Pacific region, occurs also near Inyack Island, and this is the most south-westerly locality in which it has been found up till now. It occurs on a flat, to the south-west of the island, that is dry at low tide. A great number of small specimens were procured here by turning over stones. LITERATURE. ABEL, O., Vorzeitliche Lebensspuren. Jena, 1935. Daxin, W. J., ““A New Species of Enteropneusta, Ptychodera pelsarti, from the Abrolhos Island,”’ Journ. Linn. Soc. Zool., vol. xxxiii, 1916. Gitcurist, J. D. F., ““On Two New Species of Ptychodera (P. proliferans and P. natalensis),’”? Ann. S. Afr. Mus., vol. vi, 1908. 380 Annals of the South African Museum. Hitz, J. P., ““On a New Species of Enteropneusta (Ptychodera australiensis) from the Coast of New South Wales,”’ Proc. Linn. Soc. N.S.W., 2nd ser., vol. x, 1894. ——, “The Enteropneusta of Funafuti,’”» Mem. Australian Mus., vol. iii, 1897. Horst, C. J. vAN DER, “‘ West-Indische Enteropneusten,” Bijdragen tot de Dier- kunde, vol. xxiii, 1924. . ——, ‘“‘Quelques observations anatomiques sur l’entéropneust Glandiceps talaboti (Marion),” Bull. Soc. Sc. Nat. du Maroc., vol. ix, 1929. ——, *‘Observations on Some Enteropneusta,’’ Vidensk. Medd. fra Dansk naturh. Foren., vol. lxxxvii, 1930. ——, “‘The Burrow of an Enteropneust,’”’ Nature, vol. cxxxiv, 1934. ——., “Ona New South African Species of Balanoglossus and a Comparison between it and Balanoglossus capensis (Gilchrist),’”’ Ann. S. Afr. Mus., vol. xxxii, 1937. Kapewus, F. Froorxo, “The Anatomy of the Enteropneust Saccoglossus inhacensis sp. n., Ann. Natal Mus., vol. viii, 1936. Kowatevskry, A., ‘““Anatomie des Balanoglossus Delle Chiaje,”” Mém. de lAcad. Imp. des Sciences de St.-Pétersbourg, VII® ser., vol. x, 1866. Maser, O., “Uber Balanoglossus carnosus (Willey) und eine ihm nahestehende neue Art von Neucaledonien,”’ Zool. Jahrb. Abt. Anat., vol. xxxiii, 1913. Meer, A., “‘Glossobalanus marginatus : a New Species of Enteropneusta from the North Sea,”’ Quart. Journ. Micr. Sc., vol. Ixvi, 1922. Punnett, R. C., “The Enteropneusta,” The Fauna and Geography of the Maldive and Laccadive Archipelagoes, vol. ii, pt. 2. Cambridge, 1903. SPENGEL, J. W., “Die Enteropneusten des Golfes von Neapel,” Fauna und Flora des Golfes von Neapel, Monogr. 18, 1893. ——, “* Hine neue Enteropneustenart aus dem Golf von Neapel, nebst Beobachtungen tiber den postbranchialen Darm der Ptychoderiden,” Zool. Jahrb. Abth. Anat., vol. xx, 1904. ——, “‘ Studien iiber die Enteropneusten der Siboga-Expedition nebst Beobacht- ungen an verwandten Arten,” Uitkomsten Siboga-Expeditie, Monogr. 26. Leiden, 1907. Strasny, G., “Zur Kenntniss der Lebensweise von Balanoglossus clavigerus Delle Chiaje,”’ Zool. Anz., vol. xxxv, 1910. WiteEy, A., “Enteropneusta from the South Pacific, with Notes on the West Indian Species,”’ Willey’s Zoological Results, pt. iii. Cambridge, 1899. ( 381 ) 18. Contributions to the Crustacean Fauna of South Africa. XII. Further Additions to the Tanaidacea, Isopoda, and Amphipoda, together with Keys for the Identification of the hitherto Recorded Marine and Fresh-water Species—By K. H. Barnarp, D.Sc., F.L.S., Assistant Director. (With 35 Text-figures.) THIS paper contains records of new localities, and of species new to the fauna-list, and descriptions of new species. The new material in the South African Museum is due partly to collecting by members of the staff, and partly to correspondents. Professor T. A. Stephenson, of the University of Cape Town, in the course of ecological work has submitted many specimens, including the very interesting addition to the fauna-list of a species of the Amphipodan family Ochlesidae. Mr. H. W. Bell-Marley, of Durban, has added several species to the fauna-list. One of Professor Stephenson’s collecting localities was Port Nolloth, from which there are very few previous records. In addition to the records given below under various species, the following Isopods were also collected at this locality. Sphaeramene polytylotos, and Paridotea ungulata, rubra, reticulata, and fucicola. Although many more additions may confidently be expected, it would seem that keys to the identification of the recorded species might be of some value to South African students. An attempt has therefore been made to provide such means of identification. While the aim has been to make the keys as simple as possible, it should be emphasised that a modicum of knowledge of the chief forms of Isopods and Amphipods, and of the terminology employed in describing them, is necessary. TANAIDACEA. Fam. TANAIDAH. Tanais gracilis Heller. 1866. Heller, “Novara”? Exp., vol. 11, p. 133, pl. xu, fig. 3. 1905. Stebbing, Herdman’s Ceylon Pearl Fish. Rep. Suppl., 23, p. 3, pl. i, fig. D. 382 Annals of the South African Museum. 1914 (Feb.). Barnard, Ann. 8. Afr. Mus., vol. x, p. 198, pl. xvii, fig. A (spongicola). 1914 (July). Vanhéffen, D. Siidpol. Exp., vol. xv, p. 468, fig. 6. 1925. Barnard, Ann. S. Afr. Mus., vol. XX, p, sole Additional Localities.—(West coast) Lambert’s Bay (Professor Stephenson, 1938), Table Bay (K. H. B.); (south coast) Simon’s Bay (Vanh6ffen), Buffels Bay, False Bay (K. H. B.), Still Bay and East London (Prof. T. A. Stephenson). Distribution.—St. Paul and New Amsterdam, Ceylon. Fam. APSEUDIDAE. Apseudes austro-africana nom. nov. 1920. Barnard, Ann. 8S. Afr. Mus., vol. xvii, p. 322, pl. xv, fig. 2 (australis, non Haswell, 1881). ISOPODA. Fam. GNATHIIDAE. 1926. Monod, Les Gnathiidae, Mem. Soc. Sci. nat. Maroc., vol. xiii, pp. 1-668, 1 pl. and 277 text-figs. Gnathia cryptopais Brnrd. 1925. Barnard, Ann. Mag. Nat. Hist. (9), vol. xv, p. 417. 1926. Monod, loc. cit., p. 625. This species was originally likened to elongata (Kréyer) (syn. cerina Stimpson), which, as Monod shows, is very like antarctica (Studer). The type specimen is more slender than Monod’s figure of the latter species, but with only one specimen available a detailed comparison is not possible. Fam. ANTHURIDAE. 1925. Barnard, J. Linn. Soc. Lond., vol. xxxvi, p. 109, revision. Gen. Haliophasma Hasw. 1925. Barnard, Ann. S. Afr. Mus., vol. xxv, p. 385. A point to which perhaps more attention might be paid (in the whole family) is the junction of the telson with the 6th pleon segment. With one exception, in the species of this genus already described and in those to be described below, the junction between the two is Contributions to the Crustacean Fauna of South Africa. 383 always clearly marked by the hind margin of the 6th pleon segment forming a more or less pronounced ridge, the telson being at a lower level than that of the 6th pleon segment. In pseudocarinata, however, the telson and 6th pleon segment are completely fused, there is no transverse ridge, and the level of the telson is flush with that of the 6th pleon segment. Haliophasma hermani n. sp. (Fig. 1.) Integument not strongly indurated, not pitted. Dorso-lateral keels feeble, dorso-lateral grooves distinct. Medio-dorsal pits deep. Kyes well developed. Pleon segments 1-5 with indistinct sutures. Fic. 1.—Haliophasma hermanin. sp. a, hind margin of 6th pleon segment, with telson, and sagittal and cross-sections of latter. 6, peraeopod 1. c, outer ramus of uropod. Telson arising beneath the arcuate hind margin of 6th pleon segment; obovate, apex semicircularly rounded, upturned, ventral surface evenly but rather strongly convex, dorsal surface concave, with sharp medio-dorsal keel. Flagellum of antenna | a single joint, with minute setiferous apical rudiment of a 2nd joint; flagellum of antenna 24-jointed. Maxilliped 5-jointed. | Peraeopod 1, 4th joint very broad, 5th with lower apex slightly projecting, 6th not as wide proximally as the 4th, tapering distally, palm straight. Pleopod 1 not indurated, outer surface smooth. Uropod, inner ramus slightly longer than wide, apex rounded, not reaching to level of telsonic apex, outer ramus folding over telson, 384 Annals of the South African Museum. ovate, apically obliquely truncate, outer margin finely serrulate, with fringe of plumose setae. 20mm. Cream coloured, with faint greyish stippling in the hollows beneath the dorso-lateral margins into which peraeopods 2-7 can be folded, eyes black. Localhity.—Hermanus, from cavity in the root-stock of an Allopora coral (April 1932, 1 immature ? 9). | Remarks.—Kasily distinguished from the other species of the genus by the telson, but nearest to dakarensis in this respect. The hand of peraeopod 1 is unusually feeble. Haliophasma foveolata n. sp. (Fig. 2.) Integument strongly indurated, with numerous large foveolae. Dorso-lateral keels feeble, dorso-lateral grooves distinct. Medio- dorsal pits obsolete. Eyes well developed. Pleon segments 1-5 with sutures indistinct. Hind margin of 6th pleon segment distinct but not projecting strongly over base of telson. Telson broadly ovate, the distal half narrowing rather abruptly, apex narrowly rounded, dorsally with 3 longitudinal subparallel ridges which do not coalesce basally, the intervening grooves foveolate, ventral surface slightly concave. Flagellum of antenna 1 7-jointed, the distal joints with sensory filamentous setae; flagellum of antenna 2 6-jointed. Maxilliped 5-jointed, inner plates small but distinct. Peraeopod 1 robust, 3rd joint unusually long, lower apex of 5th sub- acute, 6th subtriangular, palm expanded at base; outer surface of 2nd, 3rd and 6th joints with scattered foveoles. Pleopod 1, outer surface of outer ramus with a medio-longitudinal groove, another groove on outer margin distally, between the two grooves a few large foveoles. Uropod, inner ramus subtriangular, longer than its basal width, not reaching telsonic apex, outer ramus folding over basal lateral portion of telson, ovate, with sinuous outer margin, apex acute, outer margin serrulate, with plumose setae. 12-5x1:5 mm. Creamy-white with faint grey speckling on dorsal surface, eyes black. Locality.—Port Elizabeth, amongst corallines, worm-tubes, etc., under stones (Prof. T. A. Stephenson, July 1936, 1 immature ? 9). Remarks.—Like the @ co-type of the Australian purpurea (Barnard, Contributions to the Crustacean Fauna of South Africa. 385 J. Linn. Soc., vol. xxxvi, p. 132), this specimen has the integument strongly pitted or foveolate. The telson agrees with that of purpurea in being noticeably wider in the proximal than in the distal half; but it agrees with that of the South African tricarinata in that the 3 dorsal keels do not coalesce basally. There is, however, a difference in the shape of the telson in cross-section: biconvex in tricarinata, concavo-convex in foveolata. Perhaps the wider body, the pitted integument, and the telson proximally widened and concavo-convex a Fig. 2.—Haliophasma foveolata n. sp. a, maxilliped. 6b, peraeopod 1. c, telson, with cross-section. in cross-section are female characters. [am inclined to suspect that this is so, and that this species is only the female of tricarinata. ‘A comparison with coronicauda (fig. 3, d) on the other hand shows that in cross-section the telson is nearly the same; there is even a very slight indication of a median ridge in coromcauda; but the thin tim of the telson shows no sudden narrowing as in the present specimen. In view of this perplexity, all the specimens of the three forms being immature, and until some definite evidence of sexual dimor- phism is forthcoming, it seems clearly advisable to institute a full specific name for this foveolate specimen. As coronicauda has not been figured previously, a figure of the telson is given here for comparison. Haliophasma pseudocarinata n. sp. (Fig. 3, a—c.) Integument strongly indurated, not pitted. Dorso-lateral keels feeble, dorso-lateral grooves distinct. Medio-dorsal pits shallow and ill-defined. Eyes well developed. Pleon segments 1-5 with 386 Annals of the South African Museum. sutures indistinct, but indicated ventro-laterally by pale (non- pigmented) lines. Telson fused with 6th pleon segment, without any transverse dividing ridge or suture, surface flush with that of 6th pleon segment; ovate, sides slightly sinuous, apex rounded; dorsally apparently smooth when viewed in liquid, but when dried there is seen a shallow median pit basally, a pair of shallow depressions laterally (into which Fic. 3.—Haliophasma pseudocarinata n. sp. a, maxilliped. 6, 3rd—5th pleon segments, fused 6th segment and telson, with cross-section of latter. c, ex- ternal view of outer ramus of uropod. Haliophasma coronicauda Brnrd. d, telson, with cross-section. the statocysts open), and three very feeble longitudinal keels, one median, and one on each side nearly parallel with the lateral margin, but gradually coalescing with the margin apically; ventral surface concave. Flagellum of antenna 1 8-jointed, of antenna 2 6-jointed. Mazxilli- ped 5-jointed. Peraeopod 1, 6th joint robust, proximally much wider than the preceding joints, subtriangular, palm straight, unguis short. Pleopod 1 feebly indurated, outer ramus with slight median and lateral grooves. Uropod, inner ramus ovate, apex extending slightly beyond apex of telson, outer ramus folding inwards, but so narrow as not to overlap on to telson when thus infolded, ovate, outer (upper) margin slightly sinuous, serrulate, with plumose setae. 18x2-3 mm. Creamy, the dorsal surface of head, peraeon, pleon, telson, and peduncle of uropods mottled with brown, eyes black. Contributions to the Crustacean Fauna of South Africa. 387 Locality.—Port Elizabeth, under stones (Prof. T. A. Stephenson, July 1936, 1 immature 9). Remarks.—The telson of this species is a good example of the danger of examining a specimen in liquid only; the sculpturing is so slight as to be easily overlooked. It bears a very strong resemblance to that of coronicauda (fig. 3, d), and this form may eventually prove to be the 2 of the last-mentioned species. The shape of the tail-fan, however, which can scarcely be said to form a cup owing to the narrowness of the outer rami of the uropods; and the complete fusion of the telson and 6th pleon segment seem to indicate the necessity of a separate specific name. Ezxanthura macrura Brurd. 19252 Barnard, loc. cit., p. 131. Additional Locality.Lambert’s Bay (Prof. T. A. Stephenson, 1938, 1,2). Apanthura sandalensis Stebb. 1925. Barnard, loc. cit., p. 141. Additional Locality East London (Prof. T. A. Stephenson, July 1937). Fam. EURYDICIDAE. 1930. Monod, Ann. Sci. Nat. Zool., ser. 10, vol. xii, pp. 129-183 (Cirolanidae). Gen. Hurydice Leach. 1914. Vanhéffen, Deutsch. Stidpol. Exp., vol. xv (Zool. vii), p- 505. 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 350 a. 1925.. Id., abid., vol. xx, p. 381. 1931. Nierstrasz, Siboga Exp. monogr., xxxii ¢, p. 147. Hurydice latistylis Vanhoffen (non Dana) is obviously not an Eurydice (which has the 5th pleon segment free laterally), and is perhaps an Argathona or a young Cirolana cranchii (syn. vicina Brurd.). Eurydice natalensis Vanhéffen also is clearly not a member of this genus, but belongs to Pontogelordes as I suggested in 1925, or to Excirolana (see infra). . This genus contains therefore only the one species in South Africa, viz. HL. longicornis (Studer). Gen. Hzcirolana Richardson. 1912. Richardson, Proc. U.S. Nat. Mus., vol. xl, p. 201. 1931.. Nierstrasz, loc. cit., p. 148. EE 388 Annals of the South African Museum. Excirolana natalensis (Vanh6ffen). (Fig. 4.) 1914. Vanhéffen, loc. cit., p. 506, fig. 42. 1925. Barnard, loc. cit., p. 381. 1931. Nierstrasz, loc. cit., p. 149. A single 9 resembles P. latipes very closely. It possesses the same clypeus and frontal lamina, bifoveolate telson with the median row Fic. 4.—Eacirolana natalensis (Vanh6ffen). a, 6, peraeopods 1 and 7, with ungues further enlarged. c, uropod. d, Ist free side-plate (on segment 2) of right side. of little pits, pleopods with accessory laminae, acorn-like Ist and 2nd joints of peduncle of antenna 1, 4-jointed peduncle of antenna 2, and free margins of 5th pleon segment. The differences, however, are as follows:— Antenna 1 extending to end of 3rd peraeon segment, antenna 2 to middle of 6th segment; joints of peduncle of antenna 2 not so broadly laminar, 4th joint distinctly longer than 3rd. Side-plates shallower, especially the anterior ones, side-plate 1 (z.e. Ist free plate, on segment 2) being twice as long as deep. Uropod, inner apex of peduncle not so quadrate, slightly tapering, inner ramus with small notch slightly proximal to middle of outer margin. Contributions to the Crustacean Fauna of South Africa. 389 Peraeopods not so robust (fig. 4, a, 6). Mandibular palp 3-jointed, slender, inserted behind level of molar process (as in P. latipes). 10x43 mm. Pale straw colour, a few black stellate specks on head, and a single series across each peraeon and pleon segment. Locality.—Plettenberg Bay (K. H. B., Jan. 1931. 192 washed up on ocean beach near mouth of Keurbooms River). Remarks.—Whether this specimen is really the same as Vanhoffen’s Port Natal (Durban) species is uncertain, because Vanh6ffen did not dissect the mouth-parts of his single specimen, and left several other features undescribed; unfortunately also his figure of the pigmenta- tion shows the hinder side-plates instead of the anterior ones. The antennae fit in with natalensis as regards the length, but the peduncle of antenna 2 in my specimen is more robust (but not so robust as in latipes); possibly Vanhéffen’s specimen was a 3, though in latipes there is no sexual difference in the antennae. The figure of natalensis shows the outer margin of inner ramus of uropod without any notch. Both Vanhdéffen’s and the present specimens have the telson shorter relatively to its breadth than in orentalis (Dana). Gen. Pontogeloides Brnrd. 1914. Barnard, loc. cit., p. 355 a. - 1930. Monod, loc. cit., pp. 174 sqq. (subgen. of Excirolana). ial id... Rev. Zool. Bot. Afric., vol. xxi, p. 3: 1931. Nierstrasz, loc. cit., p. 149 (subgen. of Hxcorolana). Nierstrasz included japonica Thielemann, 1910, because the figure shows a 2-jointed mandibular palp, though this is not mentioned in Thielemann’s text. The character of the peduncle of lst antenna, however, fits in with Hzcirolana, all three joints being subequal. Pontogeloides latipes Brurd. | (Fig. 5.) 1914. Barnard, loc.-cit., p. 356 a, pl. xxx, fig. C. 1930. Monod, loc. cit., p. 179, figs. 28 F, 31, 32. Wsliokd.; loc. cit., p.'d: The statement in my original description that the peduncle of antenna 2 is 5-jointed is incorrect; it is impossible to distinguish two fused joints in the first free joint, and thus in practice the peduncle must be regarded as 4-jointed. VOL. XXXII, PART 5D. 28 390 Annals of the South African Museum. I also overlooked the fact that the telson is bifoveolate, a character which makes the resemblance of H. carangis van Name to latipes all the greater. The whole integument is very sparsely pitted, but on a Fic. 5.—Pontogeloides latipes Brnrd. a, 1st free side-plate (on segment 2) of right side. 6, telson. c,d, e, f, g, pleopods 1-5 respectively. h, inner (dorsal) view of inner ramus of pleopod 1. 7, inner ramus of pleopod 2, ¢. the median region of the telson between the two foveae, a number of them form a single more or less regular medio-longitudinal line. The suture between the rostral point of the head and the anterior upturned end of the frontal lamina is distinct and on the dorsal surface. The pleopods are interesting (cf. Monod, 1930, loc. cit., fig. 31). The peduncle of all 5 pairs has a laminar expansion (? epipodite) of the outer margin (cf. Calman in Lankester’s Treatise Zool., 1909, fig. 127, Nerocila). Further, the inner ramus of all the pleopods Contributions to the Crustacean Fauna of South Africa. 391 ~ bears a lobe at its inner basal corner (cf. Calman’s figure), which lobe is usually folded back on the inner (7.e. the dorsal surface) of the ramus. The outer ramus of pleopod 1 also bears a lobe on its inner basal corner. All these lobes, especially that on outer ramus of pleopod 1, are larger in the 9 than in the g. The ¢ stylet on pleopod 2 is densely covered with minute adpressed spinules. Side-plates deep; side-plate 1 (on segment 2) a little longer than deep (if dissected; prior to dissection the anterior corner is concealed under peraeon segment 1, and the side-plate appears to be as deep as long). Monod (1924, Parasitolog. Maurit. Bull. Com. d’Et. Hist. Sc. Afr. Occid. Fr., vol. 1, p. 68, fig. D, and loc. cit., 1930, 1931) makes carangis van Name (1920, Bull. Amer. Mus., vol. xliu, p. 49, figs. 1-5) a synonym of this species, and gives its distribution as from Rio de Oro (Sahare Occidental) to South Africa. The Ist antennae are much longer in carangis than in latipes, and in my opinion the identity of the former with the latter requires confirmation from a comparison of actual specimens. Locality—Keurbooms River estuary, Plettenberg Bay (K. H. B., Jan. 1931, numerous specimens on the sandbanks in the estuary, collected at low tide). Gen. Gnatholana Brurd. 1920. Barnard, loc. cit., vol. xvui, p. 352. Monod in his monograph of the Gnathidae (1926, loc. cit., pp. 639 sqq.) has discussed the relationship of the fossil Urda, from the Jurassic of Solenhofen, to the Gnathidae and Cymothoids; and regards Gnatholana as a living representative of the Urdazdae. Without having seen any of the actual fossils of Urda (Rektur), or even the original descriptions (e.g. Kunth, 1870, Zeitsch. Deutsch. Geol. Ges., vol. xxii), I make no comment except that the figure of U. rostrata (Monod, loc. cit., fig. 274 after Kunth) shows many re- semblances to the Cymothoids, especially some juveniles, and none to Gnatholana. And without throwing the slightest doubt on the interpretation of the falcate processes in front of the head as being really mandibles, one may remark that some juvenile Cymothoids have falcate dactyli and that the first pair of legs might have been pushed forward prior to fossilisation. One assumes, however, that a fossil specimen clearly showing the ventral surface has been examined. Nevertheless I feel some doubts as to the propriety of 392 Annals of the South African Museum. uniting the living Gnatholana with the fossil Urda in one and the same family. * Gen. Oinslana Leach. 1925. Hale, Trans. Roy. Soc. 8. Austr., vol. xlix, p. 129. 1930. Monod, loc. cit., pp. 130 sqq., 141, 142. 1931. Nierstrasz, loc. cit., pp. 149 sqq. 1935. Barnard, Rec. Ind. Mus., vol. xxxvu, p. 308. One character, to which insufficient attention has been paid, is the form of the penial processes or papillae. Wanhoffen (1914, loc. cit.), ‘ a i ‘ b i Fic. 6.—Cirolana, penial processes. a, venusticauda Stebb. 6, cranchit Leach (vicina Brnrd.). c, theleceps n. sp. d, cingulata. e, meinerti Brnrd. f, sulcata Hansen. g, palifrons Brnrd. however, is an exception among authors, and has given figures for all the species described by him, except hirtvpes, of which species apparently he had no gg. The reason for this exception may be due to the fact that in this species there are no upstanding papillae, the vasa deferentia opening by pores flush with the surface of the sternum (Barnard, 1935, loc. cat., p. 309). As far as ¢ material is available, the South African species fall into the following groups (fig. 6):— hirtipes. vorilis. undulata. No papillae, pores flush with surface Ps, | littoralis. [pleonastoca not 8. Afr.]. No papillae, pores flush with surface but : venusticauda and var. a very slight transverse ridge sumplex. cranchw (vicina Brnrd.). theleceps. [welleyz not S. Afr.]. Papillae very short, mammilliform Contributions to the Crustacean Fauna of South Africa. 393 mernertt. Papillae well developed, at least twice as iets ulata. long as broad ; son ie ifrons. sulcata. rugicauda. Clearly this character is worth incorporating in the description of a species. Cirolana virilis n. sp. (Ries Ts d:) Superficially resembling hirtipes. Antenna 2 extending back to beginning of 4th peraeon segment. Side-plates with oblique ridges slightly more distinct than is normal in hirtipes, thus differing from borealis (where they are obsolete). Postero-inferior corner of side- plate 4 (on 4th peraeon segment) quadrate. On the head the im- pressed punctate line is confined to the hind margins of the oblong eyes and is not continuous across the median area. Frontal lamina and peraeopods as in hirtipes. Vasa deferentia opening by pores flush with the surface. Apex of telson with 16 spines in addition to the plumose setae. Stylet on pleopod 2, g, very stout and strongly curved, even more so than in borealis (see Hansen, 1890, Cirolanidae, pl. 1, fig. 1, £), and absolutely different from the slender, straight stylet of hirtipes (Hansen, pl. i, fig. 2, f). 13-5 mm. Locality.—33° 59’ 8., 25° 43’ E. (Algoa Bay), 33 fathoms (s.s. “Pieter Faure,” Nov. 1898, 1 3). Cirolana natalensis n. sp. In general appearance similar to hortvpes. Eyes subrotund. A small rostral point almost meeting the frontal lamina and separating the bases of Ist antennae. An obscure impressed line on hind margin of eyes, but not continued across the median area. Side- plates 1-4 with postero-inferior corners distinctly rounded. Telson about as long as its basal width, apex somewhat pointed (cf. neglecta, Hansen, loc. cit., pl. i, fig. 3, a), with 12 spines among the plumose setae. Antenna 2 extending back to end of peraeon segment 3. Frontal lamina as in hirtipes. Peraeopods 1-3 more slender than in hirtipes, with a less con- Spicuous spine on outer apex of 4th joint, and more slender spines 394 Annals of the South African Museum. on inner margins of 4th-6th joints. Peraeopods 4-7 more robust than in hirtipes, the inner margins of 3rd—5th joints undulate and crenulate rather than definitely notched (cf. Hansen, loc. cit., pl. i, fig. 2, d), and with more numerous and longer spine-setae; 2nd joint eS Se | a Fic. 7.—Cirolana luciae n. sp. a, dorsal view of front of head. 6, frontal lamina. c, telson. Cirolana virilis n. sp. d, pleopod 2, g. Cuirolana theleceps n. sp. e, dorsal view of head and Ist peraeon segment. ff, frontal lamina. g, telson and uropod, with cross-section of telson, and apex further enlarged. of peraeopods 5-7 more broadly oval, in peraeopod 7 its greatest width in the middle of the joint (not in distal third), but with fringe of long plumose setae as in hirtipes. Up to 13 mm., but no adult gg or ovigerous 99. Locality.—Illovo, Natal (H. W. Bell-Marley, 1934, ““An enemy of fishes, and anglers’ béte noire’’). Remarks.—The shape of the eyes, the stouter and more spinose hinder peraeopods, and the hind angle of the 4th side-plate dis- tinguish this species from hirtipes; the telson also is more pointed. The shape of the 2nd joint of peraeopod 7 seems to distinguish this species from other species belonging to this group (borealis, japonensis, etc.), in all of which the greatest width is towards the distal end. Contributions to the Crustacean Fauna of South Africa. 395 Cirolana undulata Brurd. 1914. Barnard, loc. cit., p. 353 a, pl. xxx, fig. A. Additional Locality.—Port Nolloth (Professor Stephenson, 1938). Cirolana fluviatilis Stebb. 1902. Stebbing, 8. Afr. Crust., pt. 2, p. 52. 1920. Barnard, loc. cit., p. 346, pl. xv, fig. 19 (frontal lamina). 1924. Chilton, Mem. Ind. Mus., vol. v, p. 882, pl. Ix, fig. 2 (pleonastica, non Stebbing). 1926. Id., Rec. Ind. Mus., vol. xxviii, p. 180, fig. 2 (pleonastica, non Stebb.). 1935. Barnard, zbid., vol. xxxvii, p. 310, fig. 19. The vasa deferentia open by pores flush with the ventral surface. The apex of the telson is rather broadly rounded in the young and half-grown, but assumes a narrow-rounded shape in the adult, slightly more narrow in ¢ than in @ (fig. 19 in Barnard, 1935). Additional Localities. —St. Lucia Bay, Zululand (H. W. Bell-Marley, 1919, 1 3); Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, 3S, 92, from submerged rotting logs and timber in the estuary); Knysna River (K. H. B., Nov. 1938). Distribution.—Chilka Lake, India; Talé Sap, Siam. Cirolana theleceps n. sp. (Figs. 6, c, 7, e-g.) Body smooth, coarsely but sparsely punctate, convex, sublinear in outline. Head without rostral point, slightly produced over the contiguous bases of Ist antennae, dorsal surface in ¢ with 4 rounded tubercles, 2 near the anterior margin feeble, 2 near the hind margin much stronger; in 2 quite smooth. Frontal lamina a trifle more than twice as long as wide, elongate pentagonal. Peraeon segment 1 in ¢ with 2 small tubercles approximately in the middle of dorsum (easily overlooked unless specimen is examined dry). Side-plates shallow, 4-7 with oblique ridges. Lateral margins of pleon segment 5 overlapped by segment 4. Telson triangular, longer than basal width, apex rather broadly rounded, with 4-5 strong teeth on either side of median line, 3-5 setae arising from each notch, except the proximal ones where there is only 1 seta, dorsal surface not very convex, with broad shallow median longitudinal groove of nearly even width from base to apex, and with minute scattered setules. 396 Annals of the South African Museum. Antenna 1 short, reaching to hind margin of third peraeon seg- ment, flagellum 5-6-jointed. Antenna 2 reaching to middle or end of peraeon segment 4, peduncle 5-jointed, 4th and 5th joints subequal, flagellum about 23-jointed, with brush-like hemi-whorls of setae, stronger in ¢ than in Q, on the proximal joints. Mouth-parts normal. Second joint of palp of maxilliped very broad, subcircular. Peraeopod 1 stout, 3rd joint with a very stout blunt spine on inner apex, 4th not strongly produced on outer apex, with 5 similar stout blunt spines on inner margin, 5th very short, underriding 6th, latter with 1 stout conical spine on inner apex, 7 with very short blunt spine at base of unguis. Peraeopod 7, 2nd joint oval, half as long again as broad. Penial processes on 7th sternite short and stout. Stylet on pleopod 2, 3, slender, straight, arising from base of, and extending slightly beyond, inner ramus. ‘Uropod, inner ramus extending beyond telsonic apex, apex rounded with about 8 teeth, the intervening notches with 2-5 setae rather longer than those on telson, outer ramus slightly shorter, apex rounded with 4-5 teeth, intervening notches with setae. 7-8 x 2-5 mm. Pale yellowish, with transverse series of dendritic or stellate dots, 3-4 rows across each peraeon segment, a single row across each pleon segment. Locality.—Durban and Illovo (H. W. Bell-Marley, June 1930 and July 1934, gd, 29). Remarks.—The strong teeth on the margins of the telson and the uropods are a feature of this species, though paralleled by hansens Bonnier, 1896 (see also Hansen, 1905, J. Linn. Soc. London, xxix); the uropods of Neocirolana obesa Hale, 1925, also are somewhat similar. The presence of tubercles on the head in the g is reminiscent of the genera Corallana and Lanocira, but the mouth-parts of the present species are typical of Cirolana. The presence of a secondary unguis on some or all of the peraeopods (Neocirolana Hale, 1925) can scarcely be considered of generic importance; it occurs in varying degrees of prominence in quite a number of species of Cirolana. Crrolana luciae n. sp. (Fig. 7, a—c.) Body smooth, convex, oval in general outline. Head without rostral point, bases of Ist antennae contiguous. Frontal lamina Contributions to the Crustacean Fauna of South Africa. 397 oblong, about twice as long as middle width, widening slightly in front to the gently rounded anterior margin, which is free and visible beyond the Ist antennae in dorsal view. Side-plates shallow, 4-7 with oblique ridge. Pleon segment 5 laterally overlapped by segment 4. Telson triangular, length a little less than basal width, smooth, apex rounded, sides convex, margin with closely set short plumose setae, but no spines. Antenna 2 extending to end of peraeon segment 3, 5th peduncular joint very slightly longer than 4th. Peraeopods 1-4, 4th joint not produced. Peraeopods 5-7, 2nd joint nearly linear, not expanded, not strongly setose. Uropod, inner ramus extending beyond telsonic apex, apex subacute, distal margin with plumose setae and 9 widely spaced spines, outer rami lost. 9x35 mm. ; Locality.—St. Lucia Bay, Zululand (H. W. Bell-Marley, Oct. 1919, 1): | Remarks.—In association with fluviatilis, with which species it closely agrees in the shape of the frontal lamina. The frontal lamina appears to be similar also to that of arcuata Hale, 1925, but there it does not project so as to be visible in dorsal view as in the present species. The same remark applies to zndica Nierst, 1931, which species has an even shorter telson. Crrolana rugicauda Heller. (Fig. 8.) 1868. Heller, Reise der “‘Novara”’ Crust., p. 142, pl. 12, fig. 18. 1890. Hansen, Vidensk. Selsk. Skr., ser. 6, Bd. v, p. 358. 1914. Vanhéffen, loc. cit., p. 503, fig. 40. The present specimens agree with Heller’s description as supple- mented by Vanhéffen’s account of specimens obtained at the type locality. The explanation of Heller’s figure, which shows (apparently) two knobs on the telson, seems to be that the artist mistook a pale dot in the otherwise dark pigmentation for a knob; such a pale dot is present in nearly all the South African examples between the real knob and the base of the telson. The knob is not present in a young specimen 3 mm. in length, but is present in one of 4-5 mm. The whole integument is coarsely and rather closely pitted, a feature mentioned by Heller for the telson (“‘grob puntiert’’). Van- hoffen says the telson is “gek6drnelt” and that the whole surface is covered with hexagonal markings, from each of which arises a bristle: ——- eeeeeEE—Ee “asian i \ Seeelees Gooec: ee 398 Annals of the South African Museum. which is likewise correct for the South African specimens, the bristles being mostly on the pleon and telson. The anterior margin of the clypeus is free and prominent. The bases of the lst antennae are contiguous. The anterior margin of the head is costate, and there is a very small rostral point. Vanhéffen mentions that the 3rd pleon segment overlaps laterally the 4th segment, but in his figure this overlapping is not so marked as in the present specimens, where the point of segment 3 overlaps as Fic. 8.—Cirolana rugicauda Heller. a, pleon, telson, and uropod. 6, c, d, ventral, frontal, and dorsal views of head front of, showing clypeus, frontal lamina, bases of antennae, and rostral point. far as the basal lateral swelling of the telson and is almost contiguous with the peduncle of the uropod. This feature is very distinctive and might be utilised to subdivide (subgenerically) the genus Cirolana. Penial processes on 7th sternite well developed (see figure in Van- hoffen). Up to 14mm. Colour as described by Vanh6ffen. Locality.—Port Nolloth (several specimens ex Natal Museum). Type Locality.—St. Paul Island (southern Indian Ocean). Remarks.—The discovery of this species on the west coast of South Africa, within the influence of the cold west-drift (Benguella) current, is very interesting. Heller noted (loc. cit., p. 269) two Crustacea common to St. Paul and the Cape: the crayfish Palinurus lalandei and the Isopod ‘‘Sphaeroma”’ (= Parisocladus) perforata. He might have added Neptunus sanguinolentus, as appears from his table of species on p. 256. Further species common to the two regions are Tanais gracilis Heller and Dynamenella huttona (syn. brunnea Vanhoffen and kraussi Brnrd.). This species was not found by Professor Stephenson at either Port Nolloth or Lambert’s Bay, but has recently (February 1939) been collected by him at Steenbergs Cove, St. Helena Bay. Contributions to the Crustacean Fauna of South Africa. 399 Cirolana incisicauda, n. sp. (Bie. 9. .¢,.0.) The remarkable species here figured agrees in all essentials with venusticauda-simplex except as regards the telson and the prominently acute side-plates on segments 4-7. C d Fic. 9.—Cirolana incisicauda, n. sp. a, frontal lamina, profile and ventral view. ' 6, telson. Cirolana bovina n. sp. c, frontal lamina and bases of Ist antennae. d, pleon, telson, and uropod (setae on telson and uropods plumose). The frontal lamina agrees with that of the species just mentioned. The hinder peraeon segments are crimped (with short longitudinal ridges and furrows), and the pleon segments are rugulose as in simplex. The telson is thin, slightly upturned at the apex which is shallowly notched, with 4 spines on each side in the notch. The dorsal surface is finely rugulose (best seen when dried), with minute and short, scattered setules. The inner ramus of uropod has 7 strong spines on inner margin and 5 small spinules on outer margin; the outer ramus is considerably shorter than the inner, and carries 7 strong spines on inner margin and 7 spinules on outer margin.* Length 17 mm. Creamy-white, with faint brown dendritic mottling and dots. _ * Since this was in print I have seen 4 more specimens, which show that this is not a casual aberration but a valid species. 400 Annals of the South African Museum. Localities. —Port Elizabeth and Port Alfred (Prof. T. A. Stephenson, 1936 and 1939). Inhabits the tubes of the Polychaet Gunnarea; the tail-fan forms an effective “operculum,” as in the case of Exanthura macrura Brnrd. Cirolana bovina n. sp. (Hig..9, ic\hds) Anterior margin of head evenly convex, without rostral point, with submarginal impressed line. Bases of lst antennae in contact, first two joints of peduncle not clearly distinct, flagellum 10-12-jointed. Second antennae reaching to about end of 4th peraeon segment. Frontal lamina pentagonal, half as long again as broad, sides straight, distal margins meeting in an obtuse angle, apex not freely projecting. Peraeon segments smooth, sparsely punctate, no transverse im- pressed lines, hind margins of posterior segments (6 and 7, or 5-7) feebly crenulate or denticulate. An oblique ridge on side-plates 4—7; the hinder side-plates with a few long setae. No fringes of long setae on 2nd joints of peraeopods. | Pleon segments 2—5 with denticulate hind margins; hind margin of segment 4 arcuate, with median excision, and a slightly enlarged submedian denticle on either side; segment 5 with 2 large submedian conical tubercles. The lateral margin of segment 4 is posteriorly angular (not rounded as in fluviatilis); an oblique ridge on segment 4, a much feebler one on segment 3. : Telson triangular, sides slightly concave, apex narrowly rounded, with 3 spines (sometimes 4 on one side) on each side hidden amongst the dense fringe of plumose setae; dorsal surface with a conical denticle overhanging the insertion of uropod, and a pair of strong submedian, backwardly projecting conical or spiniform tubercles near base, rest of surface smooth with rather long scattered setae. Uropods, inner ramus broadly rounded, with 7 (6-8) spines on inner, and 2 on outer, distal margin; outer ramus narrow ovate, with 4 spines on inner distal margin and about 6 on outer margin; all spines hidden in the dense fringe of plumose setae. Ovig. 9 9x4 mm., juv. but no adult g present. Creamy-white with greyish mottling, eyes black. Locality.—Kast London, “‘shelly beach” (Prof. T. A. Stephenson, July 1937). Remarks.—Nearest to pleonastica Stebb. (see Barnard, Rec. Ind. Mus., vol. xxxvii, p. 309, fig. 18, a, 1935), but distinguished by the Contributions to the Crustacean Fauna of South Africa. 401 broader frontal lamina, absence of transverse impressed lines on peraeon segments, and sculpture of telson. The specific name in allusion to the Buffalo River, Kast London, and the two prongs on the telson. Fam. AEKGIDAE. Aega semicarinata Miers. 1875. Miers, Ann. Mag. Nat. Hist., vol. xvi, p. 115. 1879. Id., Trans. Roy. Soc. London, vol. clxviu, p. 201, pl. 11, fig. 1. 1911. Bouvier, Ann. Inst. Ocean., ser. 1, vol. 111, p. 39, pl. 2, figs. 3-5. 1914. Barnard, loc. cit., p. 367, pl. 32, fig. A (wrotoma). 1916. Jd., Ann. 8. Afr. Mus., vol. xv, p. 106 (corrigendum). 1919. Stebbing, Proc. Zool. Soc. London, p. 334. Remarks.—Bouvier gives photographic reproductions of specimens from the type locality Kerguelen, but the figure of the ventral surface does not show the frontal lamina. Stebbing records the species from the Falkland Islands. The species is common off Cape Point and N.W. of Table Bay in 130-200 fathoms (stock-fish grounds). When alive the animal is pale salmon-coloured, with ruby-red eyes. A. truncata Rich. (1910, Bur. Fish. Wash., Doc. No. 736, p. 14, fig. 13) from the Philippine Islands, and A. bicavata Nordenstam (1930, Nat. Hist. Juan Fernandez and Haster Is., vol. iii, p. 547, pl. 20, fig. 11, and text-fig. 11) from Juan Fernandez, are both very closely allied to one another and to semicarinata, and a direct comparison of specimens of all three forms might prove them to be conspecific, Aega momlis Brnrd. 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 365, pl. xxxi, fig. C. Several other specimens have been obtained from off Cape Point and Table Bay, as far north as Saldanha Bay. Some of the specimens were found in Leuconia-like sponges.. The transverse rows of granules are sometimes very obscure except on the pleon and posterior peraeon segments. Gen. Syscenus Harger. 1923. Stebbing, Fish. Mar. Surv. Spec. Rep., 3, p. 9. Differs from Rocinela in the total absence of eyes, the pleon distinctly narrower than the peraeon, the absence of a linguiform process on the mandible, the 6th joint of the 3 anterior pairs of 402 Annals of the South African Museum. peraeopods not expanded, with abruptly curved dactylus, and the longer 6th joint of the posterior 4 pairs of peraeopods. Besides infelox, two other species of the genus have been described: latus Richardson (1909, Proc. U.S. Nat. Mus., vol. xxxvi, p. 85, fig. 11) from the N.W. Pacific, and intermedius Richardson (1910, Bur. Fish., Doc. 736, p. 17, fig. 16) from the Philippine Islands. Syscenus infex Harger. 1897. Sars, Crust. Norw., vol. 1, p. 67, pl. xxvii, and 1899, izbid., Appendix, p. 247, Suppl., pl. 1. 1905. Richardson, Bull. U.S. Nat. Mus., No. 54, p. 212, figs. 216, 217. 1910. Id., Bur iish Doe: Tob... wie 1923. Stebbing, loc. cit., p. 9. | There is a typ. err. in Stebbing (p. 10); “5th” and “4th” peraeopods should read 7th and 6th respectively. Stebbing gives no other details, except the length 17 mm. The flagellum of the 1st antenna in Sars’ figures of enfeliz is 5 (juv.)—7 (adult)-jointed; in zntermedius 10-jointed, and in latus 13-jointed. In these two latter species the apex of the telson is rounded, not pointed as in infeliz. Locality.—Natal coast, in a coral. Distribution.—Atlantic coast of N. America; Norwegian, Danish, and British coasts; Japan; Philippine Islands. Fam. CYMOTHOIDAE. Gen. Nerocila Teach. 1924. Monod, Parasit. Maurit. Bull. Com. Hist. Sci. Afr. Occid. frang., pp. 75 sqq. 1931. Id., Rev. Zool. Bot. Afric., vol. xxi, pp. 5 sqq. 1936. Barnard, Rec. Ind. Mus., vol. xxxvii, p. 163. In the second paper Monod admits that the forms which in 1924 he called rhabdota are not the rhabdota of Koelbel, thus recognizing the validity of my observations (1925, Ann. 8. Afr. Mus., vol. xx, pp. 390, 391) that “‘rhabdota” as figured by him in 1924 and cephalotes were connected by transitional forms and were conspecific. Very little diagnostic value can be attached to the extent of the prolongation of the hind corners of peraeon segments 6 and 7, and it should certainly not be compared relatively to the pleon segments (Monod, 1924, pp. 83, 84), because the pleon segments are subject to a variable amount of contraction according to the method of preservation. Contributions to the Crustacean Fauna of South Africa. 403 The shape of the side-plates on segments 2 and 3 seems to be a better criterion, and judged by it all the South African Museum specimens, except two, clearly fall into orbignyi. One of these two, a 29-mm. 2 “rhabdota”’ form, has side-plates 2 and 3 slightly out- standing and can scarcely be counted as an exception, but is interesting as leading on to the second one. This specimen is a 98-mm. @ “‘rhabdota” from Algoa Bay with side-plates 2 and 3 acute, outstanding and upturned, but only the left-side corner of peraeon segment 3 is slightly produced. It might well be claimed as a transition to armata Dana; it is very similar to Monod’s fig. 4 (1931). Nerocila orbignyi (Gueér. Mén.). 1829-32. Guérin-Méneville, Iconogr. Regne Anim., pl. xxix, figs. 3, a—e (Ichthyophilus orbignyt). 1832. Id.. Crust. Exp. sci. Morée, p. 47 (Ichthyophilus o.). 1881. Schioedte and Meinert., Naturh. Tidsskr., ser. 3, vol. xiul, p- 60, pl. iv, figs. 16-18 (cephalotes). 1902. Stebbing, Mar. Invest. 8. Afr., vol. 1, p. 55 (cephalotes). 1914. Barnard, loc. cit., p. 371 (rhabdota, non Koelbel). 1920. van Name, Bull. Amer. Mus. Nat. Hist., vol. xhii, p. 53, figs. 6-9 (cephalotes). 1921. Stebbing, Ann. Durban Mus., vol. iii, p. 23 (armata, non Dana). 1923. Id., Fish. Mar. Biol. Surv. Spec. Rep., 3, p. 10, pl. xv (Rosca rogans). ; 1924. Monod, loc. cit., pp. 436 (75) sqq. figs. (cephalotes and rhabdota, non Koelbel). 1925. Barnard, loc. cit., p. 390 (armata, non Dana). 1926. Hale, Trans. Roy. Soc. 8. Austr., vol. 1, p. 206, figs. 4, 5, (macleayiv). 1931. Monod, loc. cit., p. 10, figs. 5-11 (references and synonymy). 1936. Barnard, loc. cit., p. 165, footnote. 1937. Schuurmans, Stekhoven. Mem. Mus. Roy. Belg. (2), fase. 9, p. 25, figs. 19-22. Although Stebbing, in describing Rosca rogans in 1923, refers to the character of the incised pleurae of the pleon segments as distinguishing Nerocila from Rosca, the figure of his specimen corresponds so exactly with Nerocila orbignyr that there can be little doubt that he over- looked the incisions, which are often small and inconspicuous, on the first two pleurae (he says the “dissection of the mouth-organs in December fogs was unsuccessful’’). 404 Annals of the South African Museum. There is in the South African Museum an exactly similar specimen, 25 mm. in length, with the last three peraeon segments broader than the anterior ones, but without any brood-plates (see Barnard, loc. cit., 1936). | | | There are also two gd, 25 and 26 mm. in length, with the inner ramus of uropod triangular, widening to the truncate distal margin. When the change to the ? stage occurs, the inner distal corner of the ramus shifts proximally and forms the tooth on the inner margin of the now nearly parallel-sided ramus. To the list of hosts (Barnard, loc. cit., 1925, p. 391) may be added the Snoek (Thyrsites atun). Gen. Anilocra Leach. 1936. Barnard, Rec. Ind. Mus., vol. xxxviii, p. 165, fig. 7. Certain differences between capensis and leptosoma are noted and figured. | | Gen. Codonophilus Hasw. 1881. Haswell, Proc. Linn. Soc. N.S.W., vol. v, p. 471. 1883. Schioedte and Meinert, Naturh. Tidsskr., ser. 3, vol. xiii, p- 322 (Ceratothoa, non Dana, 1853). 1893. Stebbing, History of Crustacea, London, p. 354 (Meznertia). 1900. Id., Mar. Invest. 8. Afr., vol. i, p. 57 (Mewnertia). 1910. Id., Gen. Cat. 8. Afr. Crust., p. 424 (Mezinertia). 1926. Hale, Trans. Roy. Soc. 8. Austr., vol. 1, p. 223. 1931. Nierstrasz, loc. cit., p. 131. Hale’s examination of the type of Haswell’s Codonophilus argus has shown that it is a young form of Meinertia imbricata, and as Haswell’s name antedates Stebbing’s the generic name must be changed. C. imbricata remains the only species recorded from South African waters. Fam. SPHAEROMIDAE. HEMIBRANCHIATAE. Gen. Sphaeroma Bosc. Sphaeroma terebrans Bate. 1879. Hilgendorf, MB. Ak. Wiss. Berlin, 1878, p. 846, pl. iv, fig. 13 (tuberculato-crimtum). 1908. Budde-Lund, Voeltzkow’s Reise., vol. u, p. 304, pl. xvii, fig. 55 (begranulatum). Contributions to the Crustacean Fauna of South Africa. 405 1920. Barnard, loc. cit., p. 358. 1921. Calman, Proc. Zool. Soc. London, p. 217. 1926. Baker, Trans. Roy. Soc. 8. Austr., vol. 1, p. 247, pl. xxxviii, figs. 11-13. 1931. Monod, Mem. Soc. Sci. nat. Maroc., vol. xxix, figs. 22 B, 42 A, 43 J, K, L. Additional Localities —Mouth of Mtunzini River, Zululand (H. W. Bell-Marley, 1926); Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931. 33, 99, juv. in water-logged timber at Whisky Creek, 5 miles from mouth, limit of tidal water); Knysna River, at tidal limit (K. H. B., Nov. 1938. 46,-99, juv.). Distribution.—Mozambique (Hilgendorf), Zanzibar (Budde-Lund), Madras, Ceylon, Queensland, Florida, Brazil. Remarks.—There can be little doubt that Hilgendorf’s and Budde- Lund’s species are synonymous. S. retrolaeve Rich., 1904, and peruvianum Rich., 1910, are very closely allied to one another, but appear to differ from terebrans in the broadly rounded or rounded- truncate telsonic apex. Sphaeroma walkeri Stebb. 1920. Barnard, loc. cit., p. 360. 1928. Baker, Trans. Roy. Soc. 8. Austr., vol. lu, p. 49. 1931. Monod, loc. cit., p. 36, figs. 5, 23 A, 43 A, B. ian barnard, Rec. Ind. Mus., vol. xxxvm, p. 178, fig. 13, bd (epistome). 1937. Monod, Mem. Inst. d’ Egypte, vol. xxxiv, p. 13. Distribution.—Ceylon, Suez, New South Wales. Sphaeroma annandaler Stebb. fit Stebbing, Rec. Ind..Mus., vol. vi, p. 181, pl. x. 1956. Barnard, Rec. Ind. Mus., vol. xxxvi, p. 178, fig. 13, ¢ (epistome). Stebbing’s description and the distinctness of this species from walkert are fully confirmed. Front margin of head with distinct raised rim. A short transverse ridge in the middle of peraeon segment 4, and a longer one on each of segments 5, 6, and 7, on the 6th and 7th segments the ridge tends to break up into a series of transversely elongate tubercles (see Stebbing’s figure). Similarly the tubercles on the anterior portion of the pleon are transversely elongate. The apex of the telson is VOU. KNIT, PART ). 29 406 Annals of the South African Museum. broadly rounded, with an upturned rim, but not so strongly marked as in walkert. The arrangement of the tubercles on the telson is as Stebbing describes. The epistome is wider across the arms (embracing the upper lip) than in walkeri, and the raised margins narrower (see Barnard, 1936, figs.). The inner lobe of maxilla 2 broad and subquadrangular as in the Indian examples. The inner plate of maxilliped is certainly narrower and more pointed than in walker, but the actual shape depends a lot on the view-point and whether the appendage is flattened under a cover-slip; Stebbing’s figures of the maxillipeds of the two species are not, I think, strictly comparable. Up to 1 imaimar Locality.—Mouth of Mtunzini River, Zululand (H. W. Bell-Marley, 1926). Distribution.—Port Canning, and Ganges Delta, India. Gen. Hxosphaeroma Stebb. Ezosphaeroma hylecoetes n. sp. (Fig. 10.) 1902. Stebbing, Mar. Invest. 8. Afr., vol. 1, p. 69 (gigas, non Leach). 1910. Id., Gen. Cat. 8. Afr. Crust., p. 428 (gegas part, non Leach, No. 81, A). 1914. Barnard, loc. cit., p. 375 (gigas part, non Leach). Integument matt, with minute scattered setules. Head with slightly thickened rim on sinuous front margin. Telson a little broader than long, more so in g than Q, sides in ¢ subtending an angle of about 60°, rather strongly convex, smooth, sides gently concave near the somewhat truncate apex, which is narrow in ¢ but considerably broader in 9; ventral surface with low semicircular ridge, and in g a shallow transverse groove. Hpistome triangular, bluntly pointed anteriorly, surface evenly convex. Maxilliped with 2nd—4th joints of palp lobed on imner margins. Anterior peraeopods not differentiated, without natatory setae. Uropod, rami lamellate, more so in § than in 2, inner ramus not quite reaching telsonic apex, outer reaching very slightly beyond, outer margin crenulate proximally, gradually becoming serrate distally, apex acute, a few minute serrulations on inner distal margin, outer margin of outer ramus and apices of both rami setulose, inner margin of outer ramus densely setose (furry); the Cape specimens Contributions to the Crustacean Fauna of South Africa. 407 have the outer ramus apically blunt, and only 3-4 crenulations distally on outer margin. Up to 6-5 mm. (Keurbooms R.), 11-5 mm. (Cape Town). Brownish or slaty-grey, with paler mottling, the most constant being patches in the mid-dorsal line on the pleon and base of telson. Fic. 10.—Hxosphaeroma hylecoetes n.sp. a, telson and uropod, J, dorsal view on right, ventral on left, with diagrammatic sagittal section. 0, telson, 2, with Se section. c¢,epistome. d, outer ramus of uropod of Keurbooms River specimen ’ (left) and Cape specimen (right). e, palp of maxilliped (setae omitted). Localities.—Salt River, near Cape Town (W. F. Purcell, Jan. 1902, 3d, 29, juv. 14 miles from mouth, §.A.M. Reg. No. 9869); Keurbooms River estuary, Plettenberg Bay (K. H. B., Jan. 1931, gd, 29, juv. in water-logged timber in the tidal area, S.A.M. Reg. No. A7848); Buffalo River, East London (Stebbing; also coll. R. M. Lightfoot, 1914). Remarks.—The telsonic apex in the larger Cape specimens is un- usually thickened, more so in ¢ thanin 9. The palp of maxilliped and the peraeopods are of the Exosphaeroma type, but as Monod (loc. cit., 1931, pp. 11 sqq.) points out, there is a series of species ranging from typical Sphaeroma to typical Exosphaeroma which makes a sharp differentiation of these two genera wellnigh impossible. Like other estuarine and log-loving species, the integument of this species is usually more or less concealed under a coating of fine particles of foreign matter. Evidently Stebbing did not clean the specimens he examined, and consequently missed the characteristic serration on the outer rami of the uropods. 408 Annals of the South African Museum. Ezosphaeroma laeviusculum (Heller). (Hig. 11, a-c.) 1843. Krauss, Siidafrik. Crust., p. 65 (juriniz, non Audouin). 1868. Heller, Reise Novara, ii, Crust., p. 138, pl. xii, fig. 7. 1905. Hansen, Q.J. Microsc. Sci., vol. xlix, p. 117 (no opinion on generic status). 1914. Barnard, loc. cit., p. 375 (gigas part, non Leach). Integument nitidulous, with sparse and shallow punctae. Front (3 Fic. 11.—Hxosphaeroma laeviusculum (Heller). a, dorsal view of telson and uropod, g, with section. 6, epistome. c, ventral view of apex of telson, 9. Hxosphaeroma truncatitelson n. sp. d, telson and uropod, 3g, with section. é, epistome. margin of head with very feebly raised rim. Epistome with straight sides, apex triangular, surface evenly convex. In ¢g a series of very low and inconspicuous granules on hind margin of each of peraeon segments, 2 submedian ones on hind margin of first division of pleon; 2 submedian on telson and a series of 5—6 in an arc on either side; in 9 these granules obsolete. Telson broader than long, in ¢ triangular, nearly straight sides subtending an angle of about 80°, apex acute; in @ sides convex and apex rather broadly rounded, in juvenile still more so; ventral surface with well-marked semicircular ridge. Distance between tips Contributions to the Crustacean Fauna of South Africa. 409 of inner rami of uropods about 34 times in 3, 2 in Q, in length of telson. } Uropod in ¢ lamellate, apically rather broadly rounded, in 9 narrower, outer ramus extending beyond telsonic apex. Penial processes relatively short and stout. Spines on outer margins of all peraeopods as long as width of the joints. Upto 10 mm. Lighter or darker slaty-grey or bluish-black (according to habitat), move or less mottled. ! et: Localities —Port Nolloth (Prof. Stephenson, 1935); Lambert’s Bay (Prof. Stephenson, 1938); Table Bay, Mouille Point, and Hout Bay, littoral (R. M. Lightfoot, 1896; J. Drury, 1914; K. H. B., 1914); Dyer’s Island (J. Drury, 1915); Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, coast near “Cathedral Rocks’’). Nistribution.—Java; but the “Novara” locality has never been confirmed by later records. Remarks.—Through the kindness of the then (1914) Director of the Vienna Museum, I was enabled to examine the types of Heller’s species (5 gg, 19,2 juv.). The very minute granules on the peraeon, pleon, and telson, which are only to be seen with certainty in a dried or semi-dried specimen, correspond exactly in the types and the Cape specimens. The largest Cape 3¢ have the apex of the telson slightly more acute, and the rami of the uropods rather more broadlylamellate, but otherwise a side-by-side comparison reveals no differences. The Lambert’s Bay specimens are particularly smooth, only the submedian pair of granules on the telson being perceptible. Exosphaeroma varicolor Brurd. HOI Barnard, loc. cit., p. 379, pl. xxxu, C. 1926. Baker, Trans. Roy. Soc. 8. Austr., vol. 1, p. 259. 1929. Hale, Crust. 8. Austr., pt. 2, p. 276. Distribution.—Baker says: “‘They [Cymodoce unguiculata] were taken . . . in 5 fathoms at Beachport, South Australia, accompanied by a species so close to Exosphaeroma varicolor that I hesitate to separate it.” The South African and South Australian forms should, in my opinion, be directly compared before finally accepting their identity. Exosphaeroma planum Brnrd. 1914. Barnard, loc. cit., p. 380, pl. xxxu, F. Locality.—Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, 410 Annals of the South African Museum. 1 juv. from debris washed up on ocean beach); Port Nolloth (Prof. Stephenson, 1938). Remarks.—In the original description it should perhaps have been stated that in juveniles the uropods do not reach to the telsonic apex. Even juveniles, however, are easily recognized by the depressed shape, the granulations and rugae, and the epistome. The smaller Californian species amplicauda (Stimpson) appears to be very much of the same general shape, but has tubercles on the peraeon segments; probably several other differences would appear if the two species were compared side by side. Exosphaeroma truncatitelson n. sp. (Fig. 11, d, e.) Integument nitidulous, with sparse and shallow punctae. Rim of front margin of head very slightly thickened and raised only near the rostral point. Hpistome with very slightly concave sides and bluntly poimted apex, surface evenly convex. No trace of any granules on peraeon, pleon, or telson. | Telson broader than long, apex broadly truncate, apical margin very slightly convex in 3, nearly straight in 9 and juv.; ventral surface with semicircular ridge. Distance between tips of inner rami of uropods 1-75 times in length of telson. Uropod lanceolate, outer rami extending slightly beyond telsonic apex. Penial processes relatively short and stout. Spines on the peraeopods as long as width of the joints. Up to8 mm. Biscuit colour, more or less speckled with greyish. Localityn—Kleinmond, near Hermanus, Cape Province (K. H. B., Heb lo2iei5s 2 jan.) Remarks.—Both sexes of this form bear a close resemblance to the 2 of laeviusculum, but are quite distinct from the ¢ of the latter species. Up to the present it has not been collected in any other locality. Exosphaeroma pallidum n. sp. (Big. 12, a—c.) Integument nitidulous, with rather numerous punctae. Front margin of head with raised sinuous rim. LEpistome pentagonal, sides concave, surface flat or slightly concave, anterior margin projecting Contributions to the Crustacean Fauna of South Africa. 411 freely, with a deeper-lying bluntly triangular extension meeting the rostral point. Peraeon, pleon, and telson quite smooth. Telson broader than long, triangular, sides subtending an angle of about 70°, apex rounded-truncate, dorsal surface evenly convex; ventral surface with semicircular ridge. Distance between tips of inner rami of uropods about 2-6 times in length of telson. d Fie. 12.—Hxosphaeroma paliidum n.sp. a, telson and uropod, 3, with section. 6, frontal view of rostral point and epistome. c, epistome, with sagittal section. Exosphaeroma antikraussi n.sp. d, telson and uropod, g. e, frontal view of rostral point and epistome. ff, epistome, with sagittal section. Uropods ovate, apices subacute, inner ramus not reaching level of telsonic apex, outer ramus extending to or very slightly beyond it. Penial processes relatively short and stout. Up to 13 mm. Colour (as preserved) uniform pale pinkish, eyes reddish. Locality.— Woodstock Beach, Table Bay (R. M. Lightfoot, Apr. 1908, 33). Remarks.—The epistome has the somewhat shovel-like shape, with freely projecting front margin (or, in other words, a subapical trans- verse ridge) as is found in antvkrauss: and kraussi. The late Mr. R. M. Lightfoot told me that the specimens were ij 412 Annals of the South African Museum. collected amongst seaweed and debris washed up on the beach (sandy) after a storm, and that when alive they were uniformly pale in colour. I have never collected it myself on any of the rocky portions of the foreshore around Table Bay. Exosphaeroma kraussiw Tattersall. 1910. Stebbing, Ann. 8. Afr. Mus., vol. vi (Cat. 8. Afr. Crust.), p- 428 (lanceolatum, non White). 1914. Barnard, loc. cit., p. 375, pl. xxxui, D (synonymy). Remarks.—Since my 1914 paper some specimens from St. Sebastian Bay, collected by the s.s. “Pieter Faure,’ have been found in the South African Museum, which are, without reasonable doubt, part of the lot (No. 132) sent to Stebbing by Dr. Gilchrist and recorded in 1910. I have always suspected that Stebbing’s specimens were kraussw, and this is now confirmed. ‘The St. Sebastian Bay specimens have a more strongly granulate integument than the majority of examples, the micro-granulate texture being particularly noticeable on the posterior margins of the peraeon segments, and on the pleon and telson; but I have seen similar examples from Table Bay and Dyer’s Island. The shape of the epistome is constant, but in my description I omitted to mention an important feature of it: the front margin projects freely, and there is a deeper-lying rounded triangular portion which connects with the rostral point, as in antikraussi and pallidum. Additional Localities.—(West coast) Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938), Dassen Island; (south coast) Dyer’s Island, Port Elizabeth, Port Alfred. The species therefore ranges, as at present known, from Lambert’s Bay to Hast London. Ezosphaeroma antikraussi 0. sp. (Fig. 12, d-.) 1914. Barnard, loc. cit., p. 375 (gugas-lanceolatum, part). Integument granulate, more strongly so than in the micro-granulate form of kraussi (supra). Rim of front margin of head slightly thickened. Epistome quadrate, sides straight, surface flat, anterior margin truncate, projecting freely and visible beyond the antennae in dorsal view, with a deeper-lying rounded extension connecting (or almost so) with the rostral point. Peraeon and pleon without any traces of tubercles. Telson broader than long, triangular, sides subtending an angle of about 70° in g, 75° in 9, apex acute; dorsal Contributions to the Crustacean Fauna of South Africa. 4138 surface with a pair of short and feebly raised ridges proximally, and an indistinct median keel distally (less strongly marked than in normal kraussi); ventral surface with semicircular ridge. Distance between tips (outer angles) of inner rami of uropods 14-14 in length of telson. Uropod, inner ramus parallel-sided, «nner apical angle rounded, outer apical angle rectangular or slightly acute, reaching to level of telsonic apex, outer ramus ovate-lanceolate, proportionately broader in g than in 9, apex acute and uncinately curved outwards, extending slightly beyond level of telsonic apex. Penial processes relatively short and stout. Up to 8 mm. Pale buff or whitish, more or less speckled with grey, often the head and anterior four peraeon segments grey, the rest pale. - Localities.—Mouille Point, Cape Town (K. H. B., 1913, 3d, 29); Oudekraal, near Camps Bay, west coast Cape Peninsula (Prof. T. A. Stephenson, 1934, 1 3); Reef Bay, Port Elizabeth (Prof. T. A. Stephenson, 1936, 33, 99). Remarks.—This is the form mentioned in 1914 as having been considered by Dr. Tattersall as lanceolatum (White), but on which I preferred to suspend my own opinion pending confirmatory evidence. At a hurried glance it might be confused with krausswi. The epistome, however, is quite constant and distinct, and together with the uropods forms a ready means of distinguishing the two species. Ezxosphaeroma gigas (Leach). (Fig. 13, a-—f.) 1900. Stebbing, Proc. Zool. Soc. London, p. 553, pl. xxxix (references and synonymy). 1909. Chilton, Subantarctic Islands, N.Z., vol. ii, p. 652 (part). 1918. Nierstrasz, Zool. Med., vol. iv, p. 123 (remarks under calcareum). 1925. Memoria Annual, 1924, Mus. Nac. Buenos Aires, pl. xxxv, fig. 1 (fig. of lanceolatum ex Giambiagi, Ann. Soc. Cient. Argent., 1925). 1931. Monod, Mem. Soc. Sci. nat. Maroc., vol. xxix, p. 69, figs. 23 L, 25°, 304,37 .C; I. In 1902 Stebbing took the view that it was better to keep gigas (Leach) and lanceolata (White) specifically separate, “although it may not be absolutely certain which of the forms Leach had before him.”’ 414 Annals of the South African Museum. It seems impossible to settle the matter without a direct appeal to the types (if extant), but I have endeavoured to place the South African forms, which hitherto have been regarded as gigas-lanceolatum, on a definite taxonomic basis. A comparison with a few specimens from New Zealand, given me by the late Dr. Chilton, and also some Tasmanian examples from the South Australian Museum, has been of considerable help. Although all the New Zealand examples were regarded by Chilton as gigas, I am of opinion that two distinct species should be accepted, and that the Tasmanian examples represent a third species. The distinguishing characters are to be found (g¢ in all cases) in the presence or absence of a frontal ridge on the head (mentioned by Stebbing, 1900, loc. cit., p. 536), the epistome, the ventral surface of the telsonic apex, and the surface texture of the integument. The shape of the telson and the uropods, previously employed in descrip- tions, are of course also of diagnostic value. The specimens from Auckland Is., New Zealand, a g and a 2 (with embryos) 21 mm., and a juvenile 14 mm. in length, are in agreement with Stebbing’s Falkland Is. specimens (1900) in having a nearly straight frontal ridge on the head. Chilton also noticed this in his larger specimens. Stebbing’s figures of the telson and epistome seem to have been drawn somewhat foreshortened; as both structures are convex a slight difference in the angle from which they are viewed may account for the small discrepancies between Stebbing’s figures and mine. The ratio of the length to breadth of telson is: 1: 1-41 (g), 1: 1:57 (2), 1: 1-68 (Gjuv.). The ratio of the distance between the tips of inner rami of uropods to length of telson is 1: 2-4 (3), 1: 1-86 (9), 1: 1-6 Gjuv.). Thus the telson is relatively longer in the adult than in the young, and the apex is more broadly rounded in the 9 and juv. than in the 3. On the ventral surface of the telson there is a feeble semicircular ridge separating the proximal portion of telson, in which the pleopods lie, from the distal portion. As regards the epistome Stebbing draws a second line within the margin. In view of the Tasmanian form mentioned below, where there is a definite raised margin, this method of draftsmanship is ambiguous. In the present case, however, one may assume that in the Falkland Is. specimens the surface of the epistome was evenly convex, not sunken and concave in the middle. Stebbing’s description of the penial processes (on 7th sternite) Contributions to the Crustacean Fauna of South Africa. 415 VY Fig. 13.—Hzosphaeroma gigas (Leach). Auckland Is., New Zealand. a, dorsal view of head. 6, portion of integument (lst peraeon segment). c, telson and uropod, ¢ (21 mm.), with section. d, ventral view of apex of telson, juv. (14 mm.). e, ventral view of apex of telson, 3. ff, epistome, with sagittal and transverse sections. Exosphaeroma sp. Carnley Harbour, Auckland Is., and Dunedin, New Zealand. g, dorsal view of head. h, portion of integument. 7, telson and uropod, 3, with section. j, ventral view of apex of telson. k, epistome, with sagittal and transverse sections. Exosphaeroma sp. Tasmania. /, telson and uropod, ¢ (26 mm.), dorsal view on right, ventral on left, with section. m, apex of telson, juv. (15 mm.). n, portion of integument. 0, epistome, with sagittal and transverse sections. 416 Annals of the South African Museum. as “about 4 times” as long as broad is not quite in accordance with his figure unless one measures the breadth very near the tip. The surface of the integument is smooth and nitidulous, with shallow and moderately closely set punctae. The following may serve as a diagnosis of this form: integument nitidulous, shallowly punctate; head with nearly straight transverse frontal ridge; epistome anteriorly more or less pointed, surface evenly convex; telson broader than long, evenly convex dorsally, apex rounded; the distance between the tips of inner rami of uropods slightly over twice in 3, slightly less than twice in 9, in length of telson; ventral surface of telson subapically with a low semicircular ridge; uropods apically subacute or narrowly rounded, the outer ramus extending beyond telsonic apex; penial processes relatively short and stout; outer margins of 2nd and 3rd joints and apex only of inner margin of 3rd joint of peraeopod 1 furry (inner margins of 4th-—6th joints in all peraeopods furry); spines on outer margins of all peraeopods few and stout (less than width of joints). Distribution. — Falkland Is. (Stebbing); Auckland Is., New Zealand (Chilton); Kerguelen. : No South African specimens have been collected which conform to the above diagnosis or resemble the above Auckland specimens. I have seen specimens from the Falkland Is. and Kerguelen (ex. Brit. Mus.) which agree entirely with the Auckland Is. specimens. Although I leave the naming of the following two forms to some other carcinologist, I propose to outline briefly some of the characters distinguishing them from the above form which is regarded as gigas. Carnley Harbour, Auckland Is., New Zealand, and Dunedin Harbour, New Zealand (2 33 from each locality, don: Dr. Chilton). (Fig. 13, g-k.) 13-145 mm. Integument nitidulous, with closer and stronger punctae; front margin of head with a slightly raised rim, sinuous, and not forming a straight transverse ridge; epistome anteriorly in ventral view apparently truncate, but with a triangular point which curves strongly dorsally to meet the rostral point, the longitudinal profile therefore strongly convex; telson broader than long, evenly convex dorsally, apex more narrowly rounded, the distance between tips of inner rami of uropods is 2:5-2-7 times in length of telson; ventral surface of telson with the semicircular ridge so well developed medianly as to form a definite transverse groove on its distal side; Contributions to the Crustacean Fauna of South Africa. 417 uropods apically subtruncate, the outer projecting only very slightly beyond telsonic apex; penial processes more slender and longer relatively to length of animal. Tasmania (2 gg and 1 juv., don: South Australian Museum). (Fig. 13, l-o.) 26-27 mm., juv. 15mm. Integument matt, with very shallow and irregularly vermiculate impressions; frontal margin of head as in the Carnley Harbour specimens; epistome bluntly pointed anteriorly, the central portion slightly sunken and surrounded by a gently raised rim; telson more depressed, broader than long, triangular, sides straight, subtending an angle of approx. 75°-78°, apex acute; distance between tips of inner rami of uropods about 12 times in length of telson; ventral surface of telson with transverse ridge and groove; uropods apically somewhat bluntly rounded, the outer extending beyond telsonic apex; penial processes relatively short and stout (as in gigas). Exosphaeroma porrectum Brurd. 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 382, pl. xxxiu, E. The @ is similar to the g in sculpturing, but the telson does not project so far beyond the inner rami of uropods, the distal portion is not narrowed, though the apex is acutely pointed, and the outer rami of uropods are not so broadly lanceolate. Additional Localities.—Port Elizabeth (Prof. T. A. Stephenson, 1936, 1 9); Lambert’s Bay (Prof. T. A. Stephenson, 1938, 1 3). Gen. Pseudosphaeroma Chilton. 1909. Chilton, Subantarc. Is., New Zealand, vol. i, p. 653. 1931. Monod, Mem. Soc. Sci. nat. Maroc., vol. xxix, p. 73. Pseudosphaeroma barnardi Monod. 1931. Monod, loc. cit., p. 78; figs. 62-66, 67 A—D, 68-71. Localities —Hout Bay, in stream (Monod); Keurbooms River, Plettenberg Bay (K. H.B., Jan. 1931, 33, 99, Juv. in water-logged timber at Whisky Creek, limit of tidal flow); estuary of Klip Drifts Fontein stream, Potteberg, Bredasdorp district (A. C. Harrison, May 1936); Palmiet River lagoon, Kleinmond (K. H. B., April 1937, under stones, swims on surface at night). 418 Annals of the South African Museum. Zuzara furcifer Brurd. 1920. Barnard, loc. cit., p. 361; pl. xv, figs. 26, 27. Juveniles of 4 mm. in length have a short blunt apical telsonic _ projection, and a blunt (truncate) process on 7th peraeon segment extending to end of the composite pleon segment. Additional Localities—Kleinmond, near Hermanus (K. H. B., February 1927, 1 3); Keurbooms River, Plettenberg Bay (K. H. B.., January 1931, 2 $3, 3 juv. from debris washed up on ocean beach). Cymodoce valida (Stebb.). 1902. Stebbing, Mar. Invest. 8. Afr., vol. i, p. 67; pl. xu, A. 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 388; pl. xxxiu, C. A young specimen, 9 mm. in length, is creamy-white in colour, with black spots arranged very nearly in the same pattern as in Stebbing’s original figure, except that the largest and most conspicuous spot is not on the anterior part of the pleon, but on the base of the telson; a small ovate spot on inner ramus of uropod. Cymodoce unguiculata Brord. 1914. Barnard, loc. cit., p. 394, pl. xoav, /B: 1926. Baker, Trans. Roy. Soc. 8. Austr., vol. 1, p. 259. 1929. Hale, Crust. 8S. Austr., part 2, p. 285. Distribution.—Beachport, South Australia (Baker). Parisocladus perforatus (M. Edw.). 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 402, plizceqmyaee Additional Localities—(West coast) Port Nolloth and Lambert’s Bay (Prof. T. A. Stephenson); (south coast) Dyer’s Island (J. Drury), Port Alfred (Albany Mus.), Hast London (Prof. T. A. Stephenson, 1937). Remarks.—In the 99 from Lambert’s Bay the telson, instead of having 2 submedian ridges or humps each with 1-2 minute points (Barnard, loc. cit.), has 2 pairs of submedian very distinct tubercles, and the 2 tubercles on the 4th pleon segment are also much more distinct than usual. EKUBRANCHIATAR. Dynamenella dioxus Brurd. 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 419; pl. xxxiv, fig. K. Additional Localities.—Port Nolloth and Lambert’s Bay (Prof. T. A. Stephenson, 1938). Contributions to the Crustacean Fauna of South Africa. 419 Dynamenella scabricula (Heller). 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 411; pl. xxxv, fig. A. Additional Localities.—(West coast), Port Nolloth (Natal Mus.), Lambert’s Bay (Prof. T. A. Stephenson, 1938), to Table Bay and west coast of Cape Peninsula; (south coast), St James, False Bay and Keurbooms River, Plettenberg Bay (K. H. B.). Dynamenella huttoni (Thomson). 1879. Thomson, Trans. N.Z. Inst., vol. xi, p. 204; pl. x, a, fig. 6. 1909. Chilton, Subantarc. Is., N.Z., vol. ii, p. 657 (comparison with eatont). 1914. Vanh6ffen, loc. cit., p. 516, fig. 49 (brunnea). 1914. Barnard, loc. cit., p. 415; pl. xxxv. B (kraussv). 1916. Id., loc. cit., p. 106 (corrigendum = huttonz). 1917. Nierstrasz, Zool. Med., vol. iii, p. 109. O18. Id., abod., vol. iv, p. 122; pl. ix; figs. 14, 15 (kraussi). 1931. Monod, Senckenbergiana, vol. xiii, p. 25. Chilton’s distinctions (based on Calman) between this species and eatont are confirmed except the last one; if by the “sinuous transverse groove” is meant the suture between the Ist and 2nd pleon segments, I find this present in both species. The raised rim of the front margin of head in eatonz, and the shape of the animal (both juv. and adult) are clearly marked distinctions. In huttont the body is nearly parallel-sided and very convex, the side-plates being nearly vertical, the height of the body half the width (cross-section semicircular). The slot at end of telson is proportionately narrower in huttoni than in eatont. I have seen South African specimens up to 18 mm. in length. Monod remarks on the curious fact that eatont occurs at Kerguelen and South America, whereas huttonz occurs in South Africa and New Zealand. If brunnea be proved (by comparison of Vanhéffen’s types) to be the same as huttonz, the distribution becomes even more curious. Localities.—As shown by additional records the species occurs along the whole South African coast from Port Nolloth, Lambert’s Bay, and Table Bay (Melkbos Strand) to Natal. Distribution.—New Zealand, Chatham Is., Kermadec Is.; St Paul Is. (Indian Ocean) (brunnea). Dynamenella macrocephala (Krauss). 1914. Barnard, loc. cat., p. 418, pl. xxxv, C. In adult 3 the 7th peraeon segment is noticeably longer in middle 420 Annals of the South African Museum. line than the preceding segments, with convex hind margin, and is slightly gibbous, overhanging base of pleon. This species is further distinguished from huttona by having a shghtly raised rim on the front margin of head. Additional Localities.—Port Elizabeth (Prof. T. A. Stephenson, 1936); Hast London (Prof. T. A. Stephenson, 1937). Dynamenella ovalis Brnrd. 1914. Barnard, loc. qt., p. 418, pl. xxxv, D, This species has a very slightly raised rim on front margin of head. The integument is a little more strongly shagreened or micro- granulate than in eaton, but in spite of this and the feeble rim on the head, there is a possibility that ovalis should be regarded as merely a dwarf form of eatoni. Additional Localities.—Mouille Point and Oudekraal (west coast of Cape Peninsula); Still Bay, East London. Dynamenella australis Rich. 1914. Barnard, loc. cit., p. 414, pl. xxxv, E. This species seems to be rare, as since 1914 I have only seen 2 99 from Sea Point (K. H. B.), and 8 specimens from Lambert’s Bay (Prof. Stephenson, 1938). As Miss Richardson pointed out, the 2 does not differ in the ornamentation of the pleon from the 4g, the granules and the prominent tubercle in front of the apical notch being equally well developed. Dynamenella australoides n. sp. (Fig. 14.) Body nearly parallel-sided, convex, glabrous; head, peraeon and anterior part of pleon with very feeble granulation and rugae (best seen when the specimen is partially dried). Posteriorly the granules become slightly more distinct. Telson distinctly granulate-tuber- culate, but the arrangement differing from that in australis, no prominent median tubercle immediately in front of the apical notch. Epistome wider than in australis, and with evenly sinuous margins, anteriorly more pointedly produced than in huttone. Uropod, inner ramus sinuous and slightly narrowing apically as in australis, but with less acute apex; outer ramus oblong-oval, with bluntly rounded apex (not ovate with subacute apex as in australis). 9x5 to 10:5x5-5 mm. Reddish or greyish brown, more or less mottled, the granules on telson and uropods pale. . Contributions to the Crustacean Fauna of South Africa. 421 Locality. St. James, False Bay, littoral (Prof. T. A. Stephenson, August 1936, 29). Remarks.—This species on a cursory glance looks like hutton, but is easily distinguished (especially if dried) by the granules on the Fie. 14.—Dynamenella australoides n. sp. 7th peraeon segment, and pleo-telson; epistome. telson; it resembles australis in the shape of the inner ramus of uropod, but differs in the shape of the outer ramus; the sculpturing of the telson is different, and the body is not setose. Dynamenella taurus ni. sp. | (Fig. 15.) Head strongly bigibbous, front descending steeply to the rim-like anterior margin, rostral point blunt, meeting the epistome. Peraeon broad, nearly parallel-sided, markedly convex, but the height not half the width (24 times in the width), segments smooth and unsculptured, segment 7 rather prominently convex in centre; VOL. XXXII, PART 5. 30 422 Annals of the South African Museum. epimera very broad, measured ventrally to insertion of peraepods about ¢ total width of segment (distance between insertions of the two peraeopods on a segment half the total width of the segment). Telson triangular, apex narrowly truncate, ventrally notched, but notch not showing in dorsal view, surface strongly convex, with two rounded ridges. é C Fic. 15.—Dynamenella taurus n. sp. a, lateral view of head, lst peraeon segment, and Ist antenna. 06, ventral view of head and epistome (flagellum of antenna 2 not completely drawn in). c, dorsal view of 7th peraeon segment, pleo- telson, and uropod. d, posterior view of telson. e, ventral view of apex of telson. First antenna, Ist and 2nd peduncular joints robust, flagellum short, 6-7-jointed. Second antenna, flagellum 12—14-jointed. Uropod, inner and outer rami broadly oval, apices rounded, extending to, or the outer ramus slightly beyond, telsonic apex. 6x4 mm. Reddish or pinkish, with a more or less conspicuous dark-margined, pale, hourglass-shaped dorsal patch on peraeon, the projecting anterior corners of Ist and 2nd peduncular joints of first antennae, the lateral margins of the Ist peraeon segment (both dorsally and ventrally), and the peduncle and outer ramus of uropod chalky white; legs pale, with a brownish mark at base of 2nd joints near insertions; apices of mandibles also white. Localities —East London (Prof. T. A. Stephenson, July 1937, 3 Contributions to the Crustacean Fauna of South Africa. 423 immature specimens); Port Nolloth (Prof. Stephenson, Aug. 1938, 1 immature specimen). Remarks.—A very distinct species, with an even more gibbous and bull-like head than macrocephala; hence the specific name. Distin- guished by the oblong-oval shape from navicula. The coloration also appears to be distinctive. Dynamenella navicula n. sp. (Fig. 16.) Body boat-shaped, in dorsal view nearly symmetrically lenticular, in lateral view strongly convex; the legs can be completely with- drawn within the margins. Surface smooth and glabrous. Head (2) strongly gibbous dorsally, less strongly so in the juveniles. Epimera on segments 2-7 faintly demarcated. Pleon segment 1 not concealed, though only faintly indicated medianly. No ornamentation on any of the pleon segments. Telson strongly convex, triangular with truncate apex, the apical margin very feebly concave, with scarcely any groove ventrally. First antennae, Ist and 2nd joints stout, with a few short plumose setae on the anterior surface, 3rd joint short and narrower; flagellum 6§-jointed, the last 3 or 4 joints with sensory setae. Second antennae, flagellum 9-jointed. EKpistome with short quadrangular projection meeting the apically truncate rostral point which separates the bases of the Ist antennae. Marsupial lamellae overlapping in middle line, brood developed in internal pouches. Third pleopod with unjointed outer ramus. Fourth and fifth pleopods with strong pleats on both rami. The uropods can lie horizontal, or can be folded vertically so that when the animal is grasping a stem of seaweed the outer edges of the uropods are applied to the seaweed, leaving a terminal gap through which water can enter to the pleopods: a groove on the ventral surface of the telsonic apex is thus unnecessary. When horizontal, the inner apex of inner ramus almost reaches the telsonic apex, outer ramus much smaller, oval. 5x2-25 mm. As preserved (after about one month in formalin) dull crimson (probably brighter in life), with faint pale patches dorsally on the peraeon segments, uropods pale, legs pale with grey dendritic dots, eyes black. - Locality.—Port Elizabeth, amongst a miscellaneous lot of Amphi- 424 Annals of the South African Museum. pods and Isopods collected from seaweed (Prof. T. A. Stephenson, 1936, 2 ovig. 99, 2 juv.); Hast London (Prof. T. A. Stephenson, 1937, 1 ovig. 9). Cc d Fic. 16.—Dynamenella navicula n. sp. a, b, dorsal and lateral views of whole animal. c, epistome and Ist antenna. d, ventral view of apex of telson and uropod. Remarks.—It is rather unfortunate that no ¢ of this noteworthy little Sphaeromid is present. It would not be surprising if the head of the $ were found to be even more prominent thaninthe?. A swollen Contributions to the Crustacean Fauna of South Africa. 425 head, though less developed and without any marked sexual di- morphism, is found in two other South African Sphaeromids: Dyna- menella macrocephala (Krss.) and Cymodoce amplifrons (Stebb.): and a comparison is invited by Baker’s figure of an unnamed and undescribed Sphaeromid from the Gt. Barrier Reef (1926, Tr. Roy. Soc. 8. Austr., vol. 1, pl. 47, figs. 10, 11, explanation, p. 279). The smaller juvenile, 2-3 mm. in length, although scarcely differing in size (7 inch) from Stebbing’s specimen of Cymodocella algoense (cf. Barnard, Ann. 8. Afr. Mus., vol. x, p. 421, 1914), shows no resemblance, the relative lengths of the telson and uropods being as in the adult. Although not found in situ, there can scarcely be any doubt that this little boat-shaped Isopod lives on seaweed, and by closely examining handfuls of weed future collectors might be rewarded by the discovery of the exact habitat, and also the male. PLATYBRANCHIATAE. Artopoles natalis Brurd. (Fig. 17, a-c.) 1920. Barnard, loc. cit., p. 377. This opportunity is taken to give a figure of this Sphaeromid. The rostral point is small and blunt; ventrally it is pinched in to form a slight median ridge which connects with the spiniform anterior process of the epistome. Stathmos n. g. Body oblong, moderately depressed, the side-plates broad and neatly horizontal, margins not ciliate. Head laterally enclosed in Ist peraeon segment. Peraeon segment 7 not quite as wide as 6th, but forming part of the lateral margins. Telson at base much narrower than anterior portion of pleon, triangular. Epistome projecting in front, visible in dorsal view. First 2 joints of antenna 1 expanded. Fourth to 6th joints of maxilliped inwardly produced. Peraeopods normal, the hinder ones becoming slightly more slender; the anterior 3 pairs with rather long spine-setae (not natatory plumose setae). Inner ramus of pleopod 1 almost twice as long as (basally) wide. Both rami of pleopod 3 with plumose setae, the outer ramus 2-jointed. Both rami of pleopods 4 and 5 without transverse folds, the outer rami 2-jointed, feebly setose. Uropods not reaching telsonic apex, but both rami well developed. 426 Annals of the South African Museum. Stathmos coronatus n. sp. (Fig. 17, d-k.) Female.—Surface minutely granulate. Head with small rostral point, slightly raised front margin, 2 low rounded submedian tubercles, and a low rounded transverse ridge between the eyes. Hach of the : h i bh J Fic. 17.—Artopoles natalis Brnrd. a, whole animal. 6, c, ventral and lateral views of epistome and rostral point. Stathmos coronatus n. g.,n. sp. d, whole animal. e, f, g, pleopods 3, 4, and 5 respectively. h, epistome. 7, palp of maxilliped (setae omitted). Jj, peraeopod 1. k, spine from inner apex of 5th joint of peraeopods 5-7. . Contributions to the Crustacean Fauna of South Africa. 427 peraeon segments with a low median longitudinal ridge, ending on the hinder segments in a slightly projecting tubercle. Anterior portion of pleon with 2 slight submedian tubercles. Telson tri- angular, apex acute, margin slightly concave where the inner rami of uropods abut against it, 2 faint submedian keels proximally, passing into a low median keel distally. Kpistome projecting conically in front, apex truncate, ventrally smooth, dorsally grooved between 2 ridges which converge towards the rostral point on head. Antenna 1, first 2 joints triangularly expanded, the antero- distal angles sharp, an oblique ridge on dorsal surface of Ist joint, 3rd joint small, flagellum 7-jointed. Antenna 2, flagellum 10-11-jointed, extending to end of lst peraeon segment. Maxilliped, 4th—6th joints inwardly expanded, 4th rather broadly, 5th and 6th narrowly. Peraeopods normal, the hinder ones, especially the 7th, more slender than the anterior ones; the 3 anterior ones with rather long simple spine-setae on lower margins of 4th—6th joints. Pleopod 1, inner ramus subtriangular, nearly twice as long as wide; 3 coupling-spines on peduncle of both pleopods 1 and 2. Pleopod 3, both rami with plumose setae, outer ramus 2-jointed, 2 coupling spines on peduncle. Pleopod 4, outer ramus 2-jointed with a few short setae on outer margin, and some short plumose setae around apex; inner ramus non-setose, but with a single rather short and stout plumose seta on apex. Pleopod 5, outer ramus 2-jointed, the suture somewhat indistinct in the middle, outer margin with 2 short setae; inner ramus apically truncate, with a few short setae on outer margin distally. Uropods, not nearly reaching telsonic apex, inner ramus sub- quadrangular, apex truncate, slightly emarginate, outer ramus ovate- lanceolate, apex turned outwards. 6 (not incl. epistome) x3-5 mm. White, eyes black. Locality.—Oudekraal, near Camps Bay, west coast of Cape Peninsula (Prof. T. A. Stephenson, 1934, littoral, 1 non-ovigerous Q). otaduos =a cattle-pen (kraal). Fam. ASTACILLIDAE. Gen. Arcturella Sars. 1925. Monod, Bull. Soc. Sci. nat. Maroc., vol. v, p. 76 (validity of genus). 428 Annals of the South African Museum. Arcturella brevipes Brnrd. (Fig. 19.) 1920. Barnard, loc. cit., p. 396, pl. 16, fig. 27. Further specimens from a specimen of Gorgonia albicans from Table Bay, 2 up to 11 mm. in length, 3-75 mm. in width. One immature 3, 6 mm. in length, has the 4th peraeon segment 1-3 mm. in length and just under 1 mm. in greatest width. In dorsal view this segment is somewhat bottle-shaped, the anterior quarter being narrower and forming the neck; the width across posterior margin is much less than that of segment 5. There are 2 small rounded tubercles in the mid-dorsal line. Fam. IDOTEIDAE. Cleantis natalensis Brnrd. - 1925. Barnard, loc. cit., p. 394. 1936. Id., Rec. Ind. Mus., vol. xxxvin, p: 186, fest First described from juveniles which probably reached the Natal coast in drift-weed by the agency of the Mozambique current. The adult, and the tubular case it inhabits, are described in the second paper quoted above. Distribution.—Bay of Bengal. Synidotea variegata Clige. 1917. Collinge, Rec. Ind. Mus., vol. xiii, p. 2, pl. 1. 1924. Chilton, Mem. Ind. Mus., vol. v, p. 891, fig. 10 and pl. lx, fig. 6. 1927. Omer-Cooper, Tr. Zool. Soc. Lond., vol. xxu, p. 205 (? hor- tupes). 1935. Barnard, Rec. Ind. Mus., vol. xxxvu, p. 313. 1936. ld. cbed-. vol. scomyims yp. 185, nes lb: A young specimen has been compared with Indian specimens, and appears to agree in all respects. Locality.—Port Elizabeth, amongst seaweed (Prof. T. A. Stephen- son, 1936, 1 juv.). Distribution.—Coasts of India and Ceylon; Suez Canal. Paridotea ungulata (Pallas). 1910. Stebbing, Gen. Cat. 8. Afr. Crust., p. 433. 1914. Barnard, Ann. S. Afr. Mus., vol. x, p. 430. Contributions to the Crustacean Fauna of South Africa. 429 Additional Localities.—(West coast) Walfisch Bay to Table Bay; (south coast) False Bay to East London. Paridotea rubra Brnrd. 1914. Barnard, loc. cit., p. 426, pl. xxxvu, fig. A. Additional Localities.—Port Nolloth and East London (Prof. T. A. Stephenson). Glyptidotea lichtensteini (Krss.). 1910. Stebbing, Gen. Cat. S. Afr. Crust., p. 434. Additional Localities.—Port Nolloth and East London (Prof. T. A. Stephenson). Gen. Engidotea Brnrd. 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 203. The definition of the genus requires a slight modification: the side-plates on segments 2 and 3 in the Q are not as long as their segments, thus resembling Paridotea. In fact these two side-plates are not as long as the length of their segments in the 3 if the length of the segment be taken in the mid-dorsal line; it is only the peculiar intersegmental notching which develops in the J, and bevels off the postero-lateral corners of the segments, which permits the original statement to be regarded as correct. Juveniles and females are typical Paridotea, as regards the side- plates, and the two genera are certainly very close. On account of the sexual dimorphism I am inclined to let Engidotea stand. Engidotea lobata (Miers). (Fig. 18.) 1881. Miers, J. Linn. Soc. Lond., vol. xvi, p. 57, pl. u1, figs. 8, 9 (3). 1914. Barnard, loc. cit., p. 204, pl. 17, C (3). Figures of both sexes are given here to show the different shapes. The telson is not quite so widely notched in females and juveniles. The coloration is remarkable. The mid-dorsal pale line, expanding into an oval spot on segments 1, 4, and 7, is characteristi® of all specimens, from 4 mm. upwards, which I have seen, though it fades in alcohol. The margins of the peraeon segments are pale, especially around the postero-lateral corners. In the ¢ these corners become bevelled off as growth proceeds, but the pale margin remains narrow. In the 2, however, no structural alteration takes place, but the pale margin becomes a broad, more or less triangular, pale patch at the 430 Annals of the South African Museum. postero-lateral corners. Consequently when a fresh female is placed on a pale background, the coloration produces the effect of incised lateral margins simulating the actually incised margins in the 3. Length of ovig. 9 17 mm., greatest width (across 3rd peraeon segment) 5mm. The ground colour may be deep maroon, greenish, Fic. 18.—Hngidotea lobata (Miers), showing how the morphological alteration in the male (left) follows the line of demarcation between the pigmented and unpigmented areas in the female (right). The pigmentation on head and telson not shown. brown, or buff according to the colour of the seaweed and other surroundings amongst which the animals live. Additional Localities.—St. James and Buffels Bay (False Bay) (K. H. B., 1914, 1915); Sea Point and Melkbos Strand (Table Bay) (K. H. B., 1913 and 1927); Port Elizabeth (Prof. T. A. Stephenson, 1936, 1 ovig. 2); Lambert’s Bay (Prof. T. A. Stephenson, 1938, Gig A PUN) Fam. STENETRIIDAE. Gen. Stenetruum Hasw. 1920. Barnard, loc. cit., p. 398 (references). 1925. Monod, loc. cit., p. 238. Contributions to the Crustacean Fauna of South Africa. 431 Stenetrium bartholomei n. sp. (Fig. 19.) Very like diaz: Brnrd. at first sight, but with the following differ- ences. Rostrum triangular, as long as basal width, apex subacute. Ventral keel with forwardly directed point on each of segments 1-3, and backwardly directed point on segments 6 and 7, obsolete on segments 4 and 5. Antenna 1, 2nd joint shortest, $ length of Ist, es \ { ed / Z SES Fie. 19.—Arcturella brevipes Brnrd. (left). Dorsal view of 3rd—5th peraeon _ segments of immature ¢. Stenetrium bartholomei n. sp. (right). Hand of peraeopod 1. drd $ length of Ist. Peraeopod 1, 3g, elongate, 2nd joint longer than 3rd—5th together, slender at base and without tooth, 6th joint 2 length of 2nd, broadly subtriangular, palm transverse, defining tooth strong, palm with one rather long slender tooth and 3 smaller teeth, finger robust, overlapping and closing down on inner side of palm, apex subacute; in 2 as in dzaz. .Pleopod 1 as in diaz but outer margins of rami less angular. 3 5-5 mm., 2? 6-5x3 mm., length of peraeopod 1 (base of 2nd joint to palm) 5 mm. Creamy-white, with greyish-brown mottling, chiefly on lateral corners of head, a small medio-dorsal spot on peraeon segments 2-7, and the anterior half or two-thirds of the pleon, the latter with pale apex, and pale lateral and medio-dorsal spots, eyes dark. Locality—Still Bay and Port Elizabeth, littoral (Prof. T. A. Stephenson, 1932, 1 3, 1 ovig. 2, 9 juv., and 1936, 1 $, 1 9 respectively). 432 Annals of the South African Museum. Stenetrium syzygus nN. sp. (Fig. 20.) Body nearly parallel-sided, dorsally smooth, but the hinder peraeon segments feebly grooved transversely, surface finely hirsute. Head with antero-lateral angles acute, but not strongly produced; rostrum prominent, acute. Hyes reduced to 3-4 separate ocelli. Antero-lateral angles of peraeon segment 1 not prominent, nor acute. Ventral keel not prominent on any of the segments, distinct on segments 1 and 2, but not raised into processes or denticles. Pleon slightly longer than broad, the anterior 2 segments distinct. Antenna 1 short, scarcely as long as length of head, and not ex- tending beyond 4th peduncular joint of antenna 2, flagellum apparently composed of only one joint. Antenna 2, 1st joint not produced on outer apex, scale on 3rd joint short and broad, 6th joint slightly longer than 5th, flagellum longer than peduncle, multi- articulate. Mouth-parts normal; 2nd joint of maxilliped not so elongate as in crassimanus Brurd. (1914, pl. xx). Peraeopod 1 not elongate, outer apex of 3rd joint blunt, of 4th acute, 6th short and broad, outer margin strongly convex, inner margin straight, palm transverse, with stout defining spine at angle, outer margin with dense fringe of long setae curving over on to inner surface. Peraeopods 2—7 biunguiculate. Penial processes on 7th peraeon segment curving towards one another, apices nearly touching. Pleopod 1 fused, peduncles very short, rami elongate ovate. Pleopod 2, 3, normal, outer ramus 2-jointed, 2nd joint short, with one strong subapical seta, inner ramus with apical setae, and a short subapical appendage. Pleopod 3, outer ramus operculiform, 2-jointed, inner ramus shorter, narrow, with a few apical setae. Pleopod 4, outer ramus narrow, 2-jointed, apically incurved, inner ramus oval, shorter and broader. Pleopod 5 with a single ovate ramus. Uro- pods, peduncle short, outer ramus shorter than inner, both with simple setae. 6x1-5mm. Creamy-white, ocelli faintly reddish. Locality.—Still Bay (Prof. T. A. Stephenson, 1932, ovigerous and non-ovigerous specimens). Remarks.—The reduction of the eyes to 3 or 4 feebly pigmented ocelli gives this Isopod the appearance of a deep-sea form; it was, however, collected under rocks and among worm-tubes, ascidians, etc.,in the littoral zone. Except for the eyes it is a normal Contributions to the Crustacean Fauna of South Africa. 433 Stenetrium, though the peduncles (fused) of the Ist pleopods are unusually short. There is one most remarkable feature of these specimens. In all Fig. 20.—Stenetrium syzygusn.sp. a, whole animal. 6, antennal. c, antenna 2. d, hand of peraeopod 1. e, uropod. f, penial processes of 7th sternite, and lst pleopods. g, pleopod 2. h, pleopod 4. of them the pleopods are of the same structure characteristic of the 3 of Stenetrrwm, but some of the specimens have well-developed brood-plates and carry ova or embryos. This extraordinary fact is confirmed by Dr. Th. Monod (in litt. 23/10/35), to whom I had forwarded specimens. 434 Annals of the South African Museum. Fam. JAERIDARE. Gen. Austrofilius Hodgson. 1910. Hodgson, Nat. Antarct. Exp., vol. v, p. 51. 1914. Vanhoffen, loc. cat., p. 554. Austrofilius serrata (Brnrd.). 21914. Vanhoffen, loc. cat., p. 554, fig. 81 (Austrofilius furcatus, non Hodgson). 1914. Barnard, loc. cit., p. 433, pl. xxxvili, A (Jaera s.). I have a strong suspicion that Vanhéffen’s Simonstown material and my Jaera serrata are the same species; possibly the Kerguelen specimens are also the same, but I doubt whether the South African specimens should be assigned to Hodgson’s antarctic species. In any case serrata would seem to be more happily placed in Austrofius than in Jaera. These minute Asellota are very difficult to examine and usually the material is very scanty. There is a great likeness between Vanhoffen’s figure of pleopod 1, 3g, and mine. Jaeropsis curvicorms (Nicolet). 1914. Barnard, Ann. 8. Afr. Mus., vol. x, p. 224, pl. xx, fig. C (¢). Additional Localities —Sea Point, Table Bay (K. H. B., 1914), Lambert’s Bay (Prof. T. A. Stephenson, 1938). Fam. PHREATOICIDAE. Phreatoicus capensis Brurd. 1914. Barnard, loc. cit., p. 233, pls. xxi, xxiv. 1927. Id., Trans. Roy. Soc. 8. Afr., vol. xiv, pp. 141 sqq., pls. viax, and text-figs. 1-6 (general biology, and vars. abbreviatus and depressus) (January). 1927. Sheppard, Proc. Zool. Soc. London, p. 109 (April). In my 1927 paper the following correction should be made in the legend to fig. 6: for “Kogelberg” read “Steenbras”; figs. 6 and d represent the var. depressus. The following additional character of var. abbreviatus may be noted: uropods with numerous setae as well as the typical spines. var. penicillatus n. Side-plates and lateral portions of head strongly setose. Hand of gnathopod (peraeopod 1) in the largest gg as large as the head in Contributions to the Crustacean Fauna of South Africa. 435 lateral view, robust, palm scarcely longer than hind margin; in 2 and juv., however, palm distinctly longer than hind margin (cf. Barnard, 1927, fig. 4, b). Antenna 2, lower surfaces of 2nd—5th peduncular joints strongly setose, flagellum with dense whorls of setae. Telson strongly setose, setae on lateral margin long, 2 apical spines and often one lateral on each side, sometimes a subapical pair on dorsal surface (as in var. depressus); all the spines more slender than in the typical form. Uropods with spines as in typical form, but outer ramus with 3 apical spines and inner ramus with 3 or 4; peduncle and rami in addition strongly setose, most of the setae, especially the distal ones, at least twice as long as the spines; 9 not so strongly setose as 3. Locality. Hermanus (K. H. B., Nov. 1935). Remarks on the Locality—The specimens were found in a stream issuing from the base of a cliff in an amphitheatre immediately to the west of the Riviera Hotel, which stands on the 60-70 ft. sea-cut terrace. This amphitheatre or basin was formerly a lagoon opening to the sea, but the mouth is now completely closed by a sand-bar, and it has become merely a marsh overgrown with rushes, palmiet, etc. Although Phreatoicus has not been collected in the coastal mountains behind Hermanus, its occurrence there is to be expected. Evidently the animals have been carried down to their present habitat by flooding of the mountain streams, or possibly by more or less subterranean channels. As the habitat can only have been formed after the major and minor uplifts (Haughton) * the establish- ment of this colony is of quite recent geological age. The occurrence of Phreatoicus at an altitude very little above sea-level has therefore no particular significance, and the details of the habitat are given merely because the habitat was somewhat unexpected. Remarks on the Riversdale Locality.—This locality (recorded without details in 1927, loc. cit., p. 148) was also rather unexpected. It lies on the northern dip slope of the Langeberg Range, a short distance west of the north end of Garcia’s Pass (through which the Riversdale- Ladismith road runs). A small stream rises on the Table Mt. Sandstone of the Langeberg and loses itself on the Bokkeveld Beds, which are banked up against the T.M.S. and form the nearly level plain stretching northwards towards Ladismith. The Crustaceans were living amongst clumps of the liverwort Aneura fastiqiata (K. H. B., Oct. 1926). The stream apparently is not perennial, and * Geol. Cape Town. Explan. sheet 247, p.41. Geol. Surv. Union S. Afr., 1933. 436 Annals of the South African Museum. it would be interesting to visit the locality at the end of the summer to see whether the stream really is perennial, and if not what happens to the animals. Fam. BOPYRIDAE. Gen. Scyracepon Tattersall. 1905. Tattersall, Fish. Ireland Sci. Invest., 1904, vol. i, p. 35. Female broadly oval. Peraeon segments, except the first one, each with a medio-dorsal boss, increasing in size posteriorly. The first five pleon segments with the pleurae produced in long digitate processes, decreasing in length posteriorly; sixth pleon segment with very short lateral processes. All seven pairs of peraeopods present on both sides, ending in a short claw. Pleopods biramous, the outer rami elongate and digitate like the pleural processes, the inner rami short and ovate. Uropods uniramous, elongate, digitate. _ Male with the first 3 pleon segments more or less distinctly marked by lateral indents. A ventral median boss on each of the peraeon segments and on the first 2 pleon segments. Pleopods and uropods absent. Remarks.—One of the reasons for differentiating this genus, viz.: the ventral bosses in the male, does not hold good, as they are found in other allied genera (e.g. Grapsicepon). And as regards the dorsal processes in the female, there are species in which they are more or less distinct on all the segments, and others in which they are re- stricted to the posterior segments. The specimens described below show a distinct median carina on all the segments in the young female, but only three distinct bosses on the hinder segments in the ovigerous female. This character, therefore, is of little generic value (cf. Stebbing, 1904, F. and Geogr. Mald. Laccad. Archip., vol. ii, p. 716, and 1910, Tr. Linn. Soc. Lond., vol. xiv, pp. 112-115). Morphologically it is not at all easy to separate many of the genera, and one of the reasons for multiplying genera and species in the past has been the theory that each different host must have a different parasite; a theory which is not free from criticism. Scyracepon levis n. sp. (Fig. 21.) Female—Young 92 symmetrical, ovigerous Q asymmetrical. In the former there is a medio-dorsal keel on each peraeon segment, gradually increasing in height posteriorly; in the ovigerous 92 the Contributions to the Crustacean Fauna of South Africa. 437 keel is obsolete on the 1st-4th segments and is present as an elevated tubercle on 5th—7th segments, that on 7th segment being the highest. Anterior rim of head smooth, postero-lateral angles rounded. Ovarian bosses moderately well developed. All the plates forming the brood- pouch, including the last pair, smooth, not tuberculose, the last two pairs fimbriate on their posterior margins. Pleural lamellae and outer a b 3 Fie. 21.—Scyracepon levisn.sp. a, dorsal view of ovigerous 2. 4, profile of dorsum of young 9. c, dorsal view of ¢. rami of pleopods crenulate on one edge and digitate on the other edge, or digitate on both edges, with more numerous digitations than in S. tuberculosa. Male—Medio-ventral bosses on first 2 pleon segments obscure. Lateral portions of pleon segments ventrally gibbous, but without actual projecting lobes (rudimentary pleopods). Leneth, 29, 63-5 mm.; breadth, 9 7-5, ¢ 1:25 mm. Locality. Off Table Bay and Cape Point, in the branchial cavity of Scyramathia hertwigi Doflein. Remarks.—Scyramathia hertwigi is very closely allied, perhaps only a subspecies, of the Northern Atlantic S. carpenteri, and the parasites of the respective crabs are also very much alike. The northern MOM, XXXII, PART 5. 31 438 Annals of the South African Museum. parasite has the last pair of marsupial plates tuberculose, whereas the southern form has them smooth; and the latter seems to have more digitations on the pleural lamellae and outer rami of the pleopods in the &. ) Fam. TYLIDAE. Tylos granulatus Krss. 1932. Barnard, Ann. 8. Afr. Mus., vol. xxx, p. 217, fig. 11, a and 6. In this paper it was stated (p. 216) that no actually ovigerous female had been examined: In February 1937 two ovigerous females were dug up at the mouth of the Schusters River, west coast of Cape Peninsula. | In the position of the brood-pouch they resemble those woodlice which completely roll themselves up into a ball (loc. cot., pp. 225, 226, fig. 13, 6, c), that is: the 5 pairs of oostegites (peraeon segments 1-5) lie perfectly flat, simulating the sterna, while the developing ova and brood press the true sterna inwards against the dorsal body- wall. Inthe present case this squeezing of the internal organs between the sternal and dorsal plates is carried to such an extreme that these organs appear to be quite degenerate. The chitinous stomach in the head segment remains, but behind this no intestine can be traced, and the hepatic glands are gone also. In fact it would seem im- probable that the mother could recover after the escape of the brood. Each oostegite has two supporting rib-like thickenings and the margins are non-setose. The anterior margin of one overlaps the hind margin of the one in front of it. Cotyledons are developed. Kggs, and young nearly ready to escape, were found together in the brood-pouch of both these females. AMPHIPODA. Further records of distribution and notes on some South African species, in addition to those quoted below, will be found in my report on the Amphipods of the John Murray Expedition (vol. iv, 1937). Fam. LYSIANASSIDAE. Stomacontion capense Brnrd. 1916. Barnard, Ann. 8. Afr. Mus., vol. xv, p. 109, pl. xxvii, HOS 2M iaOr 1937. Id., John Murray Exp. Rep., vol. iv, p. 140, fig. 1. Contributions to the Crustacean Fauna of South Africa. 439 Distribution.—South Arabian coast. Remarks.—Monod (1937, Mem. Inst. d’Egypte, vol. xxxiv, pp. 6, 10, figs. 1-3, 4, a—c, 5, 6, a) describes a species prionoplax from the Suez Canal, very close to pepina (Stebb.). The John Murray Expedition obtained a specimen in the Red Sea which I identified as pepinir (loc. cit., swpra, p. 140). Lysivanassa ceratina (Wlkr.). 1900. Chevreux, Res. Sci. Camp. Monaco, vol. xvi, p. 16, pl. v, fig. 1. 1912. Chilton, Trans. Roy. Soc. Hdin., vol. xlviii, p. 464, pl. i, fig. 5 (cubensis, non Stebb.). 1916. Barnard, Ann. 8. Afr. Mus., vol. xv, p. 120 (cubensis, non Stebb.). 1925. Chevreux and Fage, Faune de France, Amphip., p. 42, fig. 23. 1925. Schellenberg, Beitr. Kenntn. Meeresf. Westafr., vol. iii, pe 13, fie?) 1. 1926. Id., Deutsch Siidpol. Exp., vol. xviii (zool. x), p. 250. 1928. Id., Trans. Zool. Soc. Lond., vol. xxii, p. 633 (part references). 1938. Ruffo, Ann. Mus. Civ. Genoa, vol. lx, p. 154, fig. 1. Additional Localitves.—Luderitzbucht and Simon’s Bay (Schellen- berg); Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938); Port Elizabeth (Prof. T. A. Stephenson, 1936); Hast London (Prof. T. A. Stephenson, 1937). ' Remarks.—I am not prepared to admit cinghalensis Stebb., 1897; into the synonymy of this species, as Walker and Schellenberg have done, in view of conghalensis having an enlarged Ist joint in the lst antenna. The Ist antennae admittedly tend to be stouter in the 3 than in the 9, but in a 9 collected by the John Murray Expedition (1937, Barnard, J. M. Exp. Rep., vol. iv, p. 142) I found the Ist joint of antenna 1 was enlarged as in Stebbing’s figure of the $ conghalensis. Moreover, in the South African examples now referred to ceratina, the 2nd and 3rd uropods are stouter, the 2nd with the inner ramus more strongly constricted, the 3rd with stronger (3) or much stronger (2) keel on the peduncle, than in cinghalensis. I also do not think Walker was correct in making uwrodus a synonym of cinghalensis and thus of ceratuna; but probably he and Schellenberg are right in not distinguishing bispynosa as a separate species. Distribution.—Atlantic to Senegal; Canary Is.; Mediterranean, Red Sea, and East Africa. 440 Annals of the South African Museum. Orchomenella plicata Schell. 1925. Schellenberg, loc. cit., p. 119, fig. 3 (chilensis). 1926. Id., p. 292, fig. 28 (chilensis forma plicata). Localities.—Luderitzbucht, Simon’s Bay (Schellenberg). Port Nolloth (Prof. Stephenson, 1938); Oudekraal, west coast of Cape Peninsula (Prof. T. A. Stephenson, 1934). Cyphocaris anonyx Boeck. 1926. Schellenberg, p. 244 (references). 1926a. Id., Deutsch. Tiefsee Exp., vol. xxun, p. 210, pl yee, and text-figs. 2, b, 5, a, b. Locality.—_S.W. of Cape Agulhas. Cyphocaris challenger Stebb. 1926. Schellenberg, loc. cit., p. 243 (references). 1926a. Id., loc. cit., p. 212, pl. v, fig. 3 and text-figs. 2, d, 6-10. - Locality.—S.W. of Agulhas. Gen. Hurythenes 8. I. Smith. 1891. Sars, Crust. Norw., vol. i, p. 85 (Huryporeia). 1905. Chevreux, Bull. Inst. Océan. Monaco, No. 35, p. i (Katius). 1927. Schellenberg, Nord. Plankton, Lf. 20, Amphip., p. 678 and p. 681 (Katius). 3932. Barnard, Discovery Rep., vol. v, p. 55 (Kateus) and p. 58 (discussion of identity of the two genera). 1933. Stephensen, Medd. Gronland Komm. Vidensk. Unders. Gronl., vol. lxxix, no. 7. Eurythenes gryllus (Licht.) Mandt. 1926a. Schellenberg, loc. cit., p. 217, fig. 26, d (Katius obesus). 1932. Barnard, loc. cit., p. 56, fig. 21, and pl. i, fig. 1 (coloured) (Katius obesus). 1933. Stephensen, loc. cit., p. 12, figs. 4-7. 1937. Barnard, John Murray Exp. Rep., vol. iv, p. 144. Since my 1932 exposition of the likenesses between the two genera, Stephensen from an examination of abundant material in the Copen- hagen Museum, has furnished the proof that not only are the two (monotypic) genera identical, but also the two “species” are really one: Katius obesus being the $ and juvenile form, and Eurythenes gryllus the &. Locality._S.W. of Cape Agulhas (Schellenberg), a record over- looked by me in 1982. Contributions to the Crustacean Fauna of South Africa. 441 Gen. Bathyamaryllis Pirlot. 1933. Pirlot, Siboga Exp. monogr., vol. xxxiii c, p. 123. Distinguished from Amaryllis by the pronounced rostrum, and differences in the relative lengths of peduncular joints of Ist antenna. Amaryllis conocephala Brnrd. (1925, Ann. 8. Afr. Mus., vol. xx, p. 324) is transferred to Bathyamaryllis. Gen. Chironesimus G. O. Sars. 1891. Sars, Crust. Norw., vol. i, p. 108. 1908. Holmes, Proc. U.S. Nat. Mus., vol. xxxv, p. 498 (Lakota). 1926a. Schellenberg, loc. cit., p. 219. Lakota rotundatus Brnrd., 1925, is also to be transferred to this genus. Chironesimus adversicola (Brnrd.). 1925. Barnard, Ann. 8. Afr. Mus., vol. xx, p. 327 (Lakota a.). 1926a. Schellenberg, loc. cit., p. 219, fig. 13. Locality.—S.W. of Cape Agulhas (Schellenberg). Microlysias zenokeras Stebb. 1918. Stebbing, Ann. Durban Mus., vol. ii, p. 64, pl. x. Locality—Keurbooms River, Plettenberg Bay, in “red-bait” (ascidian) washed ashore on ocean beach (K. H. B., Jan. 1931, several 99 and juv.). Remarks.—No adult 3 is present. The branchial lamellae are pleated on both sides. The 5th joint of gnathopod 1 is not so long as represented by Stebbing, the lower apex forming a narrow pro- jecting lobe (cf. Orchomenopsis nodimanus Wlkr.), and the 6th joint has a slight notch with 2-3 setae in middle of lower margin. Epistome in profile slightly concave above, its lower half together with the upper lip forming an even convex curve. Both lobes of maxilla 2 narrow. Remarks on genus, and description of a second species (from Arabian coast): 1937, Barnard, John Murray Exp. Rep., vol. iv, p. 144. Fam. AMPELISCIDAE. Gen. Ampelisca Kroyer. 1925. Schellenberg, loc. cit., p. 120 (key to West African species). 4.42 Annals of the South African Museum. Ampelisca spinrmana Chevr. 1900. Chevreux, Res. Camp. Sci. Monaco, vol. xvi, p. 39, pl. vi, fig. 2. 1925. Schellenberg, loc. cit., p. 127 (forma aspinosa). Locality.—Luderitzbucht. Ampelisca palmata Brnrd. 1925. Schellenberg, loc. cit., p. 127. 1932. Barnard, loc. cit., p. 85. Localities.—Luderitzbucht, Walfisch Bay, Angola, extending to French Congo and Senegal (Schellenberg). Ampelisca brevicornis Costa. 1925. Schellenberg, loc. cit., p. 130 and p. 133 (forma platypus). 1928. Id., Trans. Zool. SOE. Lond., vol. xxii, p. 634 (size and distribution). 1932. Barnard, loc. cit., p. 84. Localitves.—Luderitzbucht, extending to Belgian Congo, Loango, Cameroon, and Dahomey (Schellenberg). Fam. PHOXOCHEPHALIDAE. Gen. Pontharpinia Stebb. 1899. Stebbing, Ann. Mag. Nat. Hist. (7), vol.iv, p. 207 (Parharpinia). 1922. Tattersall, J. Linn. Soc. Lond., vol. xxxv, p. 4. 1930. Barnard, Terra Nova Exp. Amphip., p. 335 (Protophozus). 1932. d=. loexc7i. ps LO: 1932. Pirlot, Siboga Exp. monogr., vol. xxxii b, p. 59 (diagnosis of genus and synopsis of species). Pontharpima villosa (Hasw.). 1922. Tattersall, loc. cit., p. 4, pl. 1, figs. 7-14. 1926. Schellenberg, loc. cit., p. 300. 1931. Id., Swed. Antarct. Exp., vol. 11, p. 75. 1932. Pirlot, loc. cit., p. 60 (villosa auctorum, non Haswell). Localhity.—Simon’s Bay (1 juv.) (Schellenberg). Remarks.—According to Schellenberg (1931) and Pirlot, villosa auctorum, non Haswell is an insufficiently known species. Under this Contributions to the Crustacean Fauna of South Africa. 443 heading Schellenberg includes his own 1926 record and Tattersall’s record. The Simon’s Bay example, being young, may for the time being be referred to stimpson: Stebb., already recorded from South Africa. Gen. Cyproidea Hasw. 1904. Walker, Herdman, Ceylon Pearl Fish. Suppl. Rep., vol. xvii, p- 256 (Gallea). 1906. Stebbing, Das Tierreich, vol. xxi, pp. 157, 723. 1906. Id., zbid., p. 723 (Gallea). 1924. Spandl, Zool. Anz., vol. lxi, p. 243 (Gallea). 1925. Barnard, Ann. 8. Afr. Mus., vol. xx, p. 341. When my 1925 paper was written I was unaware of Spandl’s paper. Spandl contents himself with the opinion that Gallea is more nearly allied to the Amphilochidae than to the Leucothoidae, but suggests that it may possibly be regarded as representing a separate family. On the contrary, it seems that Gallea cannot be separated from Cyproidea. Both Walker and Spandl made the same mistake of transposing gnathopods 1 and 2. Cyproidea ornata (Hasw.). % 1904. Walker, loc. cit., p. 256, pls. ii and vii, fig. 16 (G. tecticauda). 1924. Spandl, loc. cit., p. 243, fig. 2 (G. crinita) (fig. 2, G. tecticauda for comparison). - 1925. Barnard, loc. cit., p. 341. 1927. Hale, Tr. Proc. Roy. Soc. 8. Austr., vol. li, p. 314, fig. 3 (as Stenothoe valida. Corrected Hale 1929, see next reference). 1937. Sheard, Tr. Proc. Roy. Soc. 8. Austr., vol. lxi, p. 20. 1938. Schellenberg, K. Sv. Vet. Ak. Handb., vol. xvi, p. 18. Spandl separated his crinita from tecticauda mainly on the different shape of the 5th joint of gnathopod 2 (gn. 1 in Walker and Spand)), and its armature; he also says the 3rd and 4th side-plates are fused in tecticauda, a statement apparently based on Walker’s figure, because Walker himself does not actually say they are fused. Whether enough reliance can be placed on the difference in shape of the 5th joint of gnathopod 2 is doubtful, and it is possible that Walker overlooked the fringe of setae on its lower margin in his specimens. There seems little doubt, however, that Spandl’s crimta is synonymous with Haswell’s ornata: the South African specimens agree with both. Schellenberg has no doubts about the synonymy. 444 Annals of the South African Museum. Additional Localities.—Port Alfred (Spandl); East London (Prof. T. A. Stephenson, 1937). Distribution.—S. Australia, Ceylon, Suez, Bismarck Archipelago. Fam. METOPIDAKE. Proboloides rotunda (Stebb.). 1917. Stebbing, Ann. 8. Afr. Mus., vol. xvii, p. 39, pl. vu, B (S. Afr. Crust., pl. xcvi, B). Examination of the type slide shows that Stebbing was mistaken in regarding the palp of maxilla 1 as being 1-jointed (as in Metopa); actually it is 2-jointed, and the species should therefore be transferred to Proboloides. There is no accessory flagellum. Fam. AMPHILOCHIDAE. Gen. Hoplopeon Brurd. 1932. Barnard, loc. cit., p. 105. The genotype is Peltocoxa australis Brnrd., 1916. Hoplopleon medusarum Brurd. 1932. Barnard, loc. cit., p. 105, fig. 54. Locality. Saldanha Bay, in medusae (R.R.S. “ Discovery”’). Gitanopsis pusilla Brnrd. 1925. Schellenberg, loc. cit., p. 140. Additional Localitves.—Luderitzbucht, Swakopmund (Schellenberg) ; Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938); Still Bay, (Prof. T. A. Stephenson, 1935). Remarks.—Chilton (1923, Rec. Austral. Mus., vol. xiv, pp. 82 sqq.) united this species with Amphilochus neapolitanus Della Valle, a synonymy with which neither Schellenberg nor myself agree. Fam. PHLIANTIDAE. 1936. Sheard, Rec. 8S. Austr. Mus., vol. v, p. 456. Gen. Palinnotus Stebb. 1906. Stebbing, Das Tierreich, vol. xxi, p. 202. Contributions to the Crustacean Fauna of South Africa. 445 Palinnotus natalensis n. sp. (Fig. 22.) Resembling P. thomson Stebb. so closely that scarcely any char- acters, which might be considered of specific value, can be found. (See Stebbing, 1899, Tr. Linn. Soc. Lond., vol. vii, p. 417, pl. xxxv, A). Fie. 22.—Palinnotus natalensis n.sp. a, dorsal view, 3rd and following pleon segments bent underneath, with margin of anterior side-plates further enlarged. b, margin of side-plate of young. c, telson and uropods, with apex of ramus (the same for both uropods | and 2) further enlarged. d, antennal. e, maxilli- ped. f, peraeopod 5. g, pleopod 3, setae omitted. In fact the greater expansion of the 2nd joint of peraeopod 5 in the largest specimen is the only apparent difference; in the juveniles the expansion is no greater than in Stebbing’s species. The margins of the first four side-plates are very feebly crenulate in the 9, with minute setules; in the juveniles, however, there are numerous longer setae. Nevertheless I am reluctant to identify these Natal specimens with a species which has only been found in New South Wales. If 446 Annals of the South African Museum. the specimens had been found actually in Durban harbour one might perhaps call in the agency of dispersal by ships (cf. Chilton, 1911, Tr. N.Z. Inst., vol. xliii, p. 131) as an explanation of their occurrence in South Africa, but the locality is some 20 miles south of Durban. Barring accidental capture, these small Amphipods are only likely to be collected in the course of intensive investigations such as Prof. Stephenson has been carrying out, and of which there is great need in many parts of the world. Length (incl. pleo-telson) 4 mm., greatest width 2 mm. Creamy, the gut showing through the integument as a deep red streak, as if the animal had been sucking blood, eyes black. Locahity.—Isipingo, Natal (Prof. T. A. Stephenson, 1936, 1 2 and 2 juv. from algae, littoral). Temnophlias capensis Brnrd. 1916. Barnard, Ann. 8. Afr. Mus., vol. xv, p. 158, pl. xxvi, figs. 25-35. Additional Localities.—(South coast) Still Bay (Prof. T. A. Stephen- son, 1932); (west coast) Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938). Fam. OCHLESIDAE. 1910. Stebbing, Mem. Austral. Mus., vol. iv, p. 581. Gen. Ochlesis Stebb. 1910: Stebbing, loc. cxt., p. Sei. 1932. Pirlot, Siboga Exp. monogr., vol. xxxi 6, p. 105. 1932. Id., Ann. Inst. océan., vol. xii, pp. 24, 26, 29, 30, fig. 17. 1936. Id., Siboga Exp. monogr., vol. xxxili e, p. 298. General appearance like that of Odius or Iphimedia. Integument indurated. Head strongly rostrate. Hyes present. Peraeon and first three pleon segments carinate. Pleon segment 4 elongate, 5th short but distinct from 6th. Side-plates 1-3 much deeper than long. Telson entire. Antennae short but with the normal number of peduncular joints; flagella reduced to 1-3 joints, no accessory flagellum. Upper lip elongate. Lower lip with narrow pointed lobes, no inner lobes, mandibular processes short, acute. Mandibles broad at base, tapering to an acute apex, secondary cutting-plate, spine-row and molar absent, palp set far back, slender, 1-3 spinules at apex. Maxilla 1 with small unarmed inner plate, outer plate Contributions to the Crustacean Fauna of South Africa. 447 narrow, with 5 apical spines (or 4 and a subacute apical process), palp absent or a minute rudiment. Maxilla 2, both plates narrow, with a few apical spines. Maxillipeds, inner plates very narrow, separate, with a few spinules on inner margin and apex, outer plates ovate, closely adjacent, palp absent. Gnathopod 1 simple, from the rather elongate 3rd joint onwards slender. Gnathopod 2, complexly subchelate, the 5th joint being produced in an acute process below the 6th (in the Australian species). Peraeopods 1-5 stout, 7th joints strong, uncinate. Pleopods not reduced. Uropods 1-3 biramous. Remarks.—The above diagnosis is drawn up from the Australian species (hitherto the only known species) and the South African one _ described below. The two species are obviously congeneric. In the latter the 2nd gnathopod is not known. Stebbing was unable to distinguish the 5th from the 6th pleon segment; they are here clearly separate, though the 5th is dorsally very short and more or less telescoped into the 4th segment. The remarkable feature of this genus is the absence of the palp of the maxilliped, a feature unknown in any other member of the Gammaridea at that time, but found in certain Caprellidea (Cyamidea) in the adult, and throughout the Hyperidea (Stebbing, p. 582). The nearest approach was the rudimentary 2-jointed palp in Laphystius. Since then, however, the genera Thoriella Steph., Chevreuaiella Steph., and Danaella Steph. have been discovered. These three genera are aberrant bathypelagic Lysianassids in which the outer plates of the maxillipeds are operculiform (as in Laphystius and Ochlesis) with a remnant of the palp in the first two named genera, but with no trace of it in Danaella. Pirlot (Ann. Inst. océan.) has given a comparative study of the degeneration of the palp of the maxilliped in various Gammaridea, showing a complete transition to that of the Hyperiwdea, amongst which certain forms approximate to the former. Pirlot believes that this degeneration of the palp and the expansion of the outer plates to form an operculum closing the mouth below is an adaptation to an inquiline or parasitic mode of life. Ochlesis lenticulosus n. sp. (Fig. 23.) Integument indurated, surface of peraeon and pleon closely and finely pitted. Rostrum strong, moderately deflexed, apically acute. Antero-lateral angles of head rounded-quadrate. Eyes small, circular. 448 Annals of the South African Museum. Peraeon and pleon segments 1-3 dorsally carinate; peraeon segment 7 and pleon segments 1 and 2 ending posteriorly in a blunt medio- dorsal projection; pleon segment 3 with an upstanding triangular process about in middle of its length; pleon segment 4 dorsally Fia. 23.—Ochlesis lenticulosus n.sp. a, whole animal. 6, c, d, side-plates 1-3 (anterior margin to right). e, dorsal and lateral views of telson. f, upper lip. g, mandible. h, maxilla 1, with apex of outer lobe further enlarged. 72, maxilla 2. j, maxillipeds. kk, lower lip. /, gnathopod 1, with apex further enlarged. m, 2nd—4th joints of peraeopod 1. n, 2nd—4th joints of peraeopod 3. 0, inner view of 2nd—4th joints of peraeopod 5. rounded; segment 5 dorsally very short, segment 6 with a dorso- lateral keel projecting almost horizontally on either side. Postero- inferior angles of pleon segment 3 sharply upturned. Side-plates 1-3 increasing in depth, anterior margin strongly concave, antero- basal angle slightly projecting in side-plate 1, more so in 2, still more Contributions to the Crustacean Fauna of South Africa. 449 so in 3, extending almost to anterior margin of 2; 4 shallower than 3, its basal anterior corner overlapping 3, anterior margin concave; 5-7 moderately shallow, 5 with anterior projecting point, 6 quadrate, hinder lobes of 5 and 6 deeper than anterior lobes, 7 oblong. Telson rather elongate, oval, apically narrowly rounded, entire, equal in length to peduncle of 3rd uropod, with strong medio-ventral keel. Antennae short and stout; 1st antenna with spinous projection on distal lower margin of Ist and 2nd joints, flagellum very short, l-jointed (with an obscure minute terminal joint); antenna 2, flagellum half length of 5th peduncular joint, obscurely 3-jointed, apical joint minute. Upper lip elongate, apically acute. Mandibles with apices acutely pointed, palp elongate, slender, 3rd joint almost as long as 2nd, armed with only a single apical spinule. Maxilla 1 acute, apex with a few minute adnate hooked spinules, inner lobes small, palp absent. Maxilla 2, both lobes narrow, with a few apical spinules. Maxilliped, outer and inner plates well developed, the latter separate from base, palp absent. | Gnathopod 1 simple, slender, 6th joint longer than 5th. Gnathopod 2, distal joints missing on both sides. Peraeopod 1, 2nd joint stout, strongly expanded distally, 4th expanded, twice as broad as 3rd, the upper proximal corner somewhat projecting, heel-like. Peraeopod 2 stout, but 4th joint not strongly expanded, upper proximal corner rounded. Peraeopods 3-5, 2nd joint stout, expanded (in peraeopod 4 more elongate in proportion to width than in the figure of peraeopod 3), anterior margin of 6th joint with a series of spinules, and a pair of slightly larger ones at apex. Seventh joints in peraeopods 1-5 stout and strong. Peraeopods unarmed except for the spinules on 6th joints. Uropod 1 longest, rami shorter than peduncle, outer slightly shorter than inner. Uropod 2, rami subequal to peduncle, outer ramus slightly shorter than inner. Uropod 3 a little shorter than 2nd, outer ramus distinctly shorter than inner. Length.—About 8 mm. Colour.—(As preserved after 2 days in formalin) bright orange, paler laterally, where there is a transverse red or crimson stripe on each peraeon segment, continued below on to the side-plate, on each of pleon segments 1-3 three such stripes, all the stripes ending dorsally in a lateral crimson stripe, rather sharply demarcated on its lower edge, but passing gradually into the orange colour above: ord peduncular joint and flagellum of antenna 1, flagellum of antenna 2, and the 7th joints of peraeopods 1-5 white; eyes red; cheek red 450 Annals of the South African Museum. with corner or antero-lateral angle white. The red or crimson markings were said to be violet or mauve before preservation. Locality. Simonstown, littoral (coll. Dr. and Mrs. T. A. Stephen- son, 11/6/32, 1 specimen). Remarks.—Closely allied to O. innocens Stebb., 1910, but distin- guished by the Ist antennae, the flagellum of which is shorter than the 3rd peduncular joint (instead of vice versa) and the 1st peraeopod. The telson in innocens is described as having a “process or stout spine’? on ventral surface; in the present species it is keeled from base to apex, the keel expanding in basal third into a triangular projection. The known distribution of innocens is now (1936, Pirlot, loc. cit.) extended to the Aru Islands, Kast Indies. 29 Fam. ACANTHONOTOZOMATIDAE. Iphimedia capicola Brurd. 1932. Barnard, loc. cit., p. 118, fig. 66. Locality.—Off Saldanha Bay, 4 fathoms (R.R.S. “ Discovery”). Panoploea excisa Brurd. 1932; Barnard, loc. cit., p. 129, fis. 73: Locality. Off Saldanha Bay, 4 fathoms (R.R.S. “ Discovery”). Fam. LILJEBORGIIDAE. Inleborgia proxuma Chevr. 1916. Barnard, loc. cot., p. 167. 1938. Schellenberg, K. Sv. Vet. Ak. Handb., vol. xvi, p. 31, fig. 15. Liljeborgia kinahani Bate var. capensis Brnrd. 1894. Sars, Crust. Norw., vol. i, p. 532, pl. clxxxvii, fig. 1. 1932. Barnard, loc. cit., p. 142, fig. 81, a. Locality.—Simon’s Bay (R.R.S. “Discovery”’). LIiljeborgia epistomata Brnrd. 1932. Barnard, loc. cit., p. 144, fig. 83. Locality.—_Saldanha Bay (R.R.S. “Discovery”’). Contributions to the Crustacean Fauna of South Africa. 451 Fam. CALLIOPIIDAE. Gen. Calliopiella Schell. 1925. Schellenberg, loc. cit., p. 147. Calliopiella michaelsend Schell. (Fig. 24.) 1925. Schellenberg, loc. cit., p. 147. Localities—Swakopmund (Schellenberg). Table Bay (K. H. B., 5/10/14). ZA (Z 5 Fic. 24.—Calliopiella michaelseni Schell. a, hand of gnathopod 2. 6, uropod 2. c, telson and uropod 3, with apex of inner ramus of latter further enlarged. Remarks.—A single 3 specimen, 8 mm. in length, was collected at Mouille Point, Cape Town. As it was a singleton and would not fit in with Stebbing’s key to the rather heterogeneous family Callio- pudae, it was reserved for future study. No more specimens have 452 Annals of the South African Museum. come to hand, but the 1914 specimen is clearly referable to Schellen- berg’s species. The accessory flagellum is absent. The telson has a slight apical indent. Fam. PONTOGENEIIDAE. Paramoera capensis (Dana). 1916. Barnard, loc. cit., p. 183 (references, except magellanica). 1918. Stebbing, Ann. Durban Mus., vol.u, p. 66, pl. ix, C — 1925. Schellenberg, loc. cit., p. 149. 1926. Id., loc. cit., p. 363 (forma capensis). 1926. Id., Zool. Anz., vol. Ixxxv, p. 280 (fissicauda Dana var. capensis). 1931. Id., loc. cit., pp. 194, 197 ( fissicauda). 1932. Barnard, loc. cit., p. 209, figs. 118, », 128. Additional Localities. Siest coast) Luderitzbucht, Redford Bay, Possession and Pomona Islands, Swakopmund (Schellenberg); Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1935 and 1938); Saldanha Bay (K. H. B., 1912); (south coast) Still Bay (Prof. Stephen- son); Port Shepstone, Natal (H. C. Burnup). Paramoera bidentata Brurd. 1932. Barnard, loc. cot., p. 210, figs. 118, m, 129: Localities.—Kalk Bay (8.A. Mus.). Oudekraal, west coast of Cape Peninsula (Prof. T. A. Stephenson, July 1934). Still Bay (Prof. T. A. Stephenson, 1932). : Remarks.—The length of the type specimen should have been given as 13 mm., not 15 mm. A second specimen, also a ¢, measuring 12 mm., was found at Still Bay; as preserved it is pure white (probably translucent when alive), with maroon or crimson patches as follows: a stripe on postero- lateral margin of head (next to the lower part of peraeon segment | and upper part of its side-plate), and on the postantennal angle, along the lower margins of Ist and 2nd joints of antenna 1, junctions of 3rd and 4th, and 4th and 5th peduncular joints of antenna 2, and junction of 5th joint and flagellum, apices of mandibular palps, maxillae and maxillipeds, 4th joints and palm of 6th joints of gnatho- pods 1 and 2; eyes dark reddish-brown. The Oudekraal specimen is a very fine ovigerous 2 measuring HO Misamae Contributions to the Crustacean Fauna of South Africa. 453 Fam. GAMMARIDAKE. Gen. Megaluropus Hoek. 1906. Stebbing, Das Tierreich, vol. xxi, p. 420. 1925. Schellenberg, loc. cit., p. 151. 1932. Barnard, loc. cit., p. 145 (Phylluropus). Megaluropus agilis Hoek. tes, Della, Valle, fF. Fl. Neapel, vol. xx, p. 695, pl. m, fig. 9; pl. xxxiv, figs. 1-17. 1904. Walker in Herdman’s Ceylon Pearl Fish. Rep., vol. xvii, p.278. 1925. Chevreux and Fage, loc. cit., p. 226, figs. 236, 237. 1925. Chevreux, Bull. Soc. zool. France, vol. 1, p. 304. 1928. Schellenberg, loc. cit., p. 644. 1932. Barnard, loc. cit., p. 146, figs. 84, 85 (P. capensis). Locality False Bay (R.R.S. “ Discovery’’). Distribution.—North Sea, Mediterranean, Port Said, Ceylon, Canary Islands. Remarks.—As Dr. Schellenberg has pointed out to me, the “Discovery” specimen should probably be identified with this European species. A second species, longimerus (fig. 14 is labelled “Tongumanus’’), was described by Schellenberg in 1925 from Lagos. Gen. Hriopisella Chevr. 1920. Chevreux, Bull. Soc. zool. France, vol. xlv, p: OE: 1925. Chevreux and Fage, loc. cit., p. 220. 1933. Schellenberg, Mitt. Zool. Mus. Berlin, vol. xix, pp. 408, 409. 1935. Barnard, Rec. Ind. Mus., vol. xxxvil, p. 284. This genus is separated from Hriopisa by the acute antero-inferior angle of side-plate 1, the very slender mandibular palp, inner plate of maxilla 1 with 2—3 setae at apex only, inner plate maxilla 2 narrow, without setae on inner margin, and the short almost styliform 2nd joint of outer ramus of uropod 3. It includes seychellensis (Chevr.) 1901, pusilla Chevr. 1920, and capensis (Brurd.) 19t6. Melita subchelata Schell. 1925. Schellenberg, loc. cit., p. 153 (fresnelit var. subchelata). 1932. Barnard, loc. cit., p. 211, fig. 130. Localities.—Luderitzbucht (Schellenberg), Walfisch Bay (R.R.S. “Discovery ’’). WOi. MXXIt, PART 5. 32 aed 454 _ Annals of the South African Museum. Melita orgasmos n. sp. 1916. Barnard, Ann.S. Afr. Mus., vol. xv, p. 191 (part inaequistylis, non Dana). Surface nitidulous, with scattered but rather deep pits. Peraeon segments and pleon segments 1-3 smooth, non-dentate; pleon segment 4 produced in a slender median tooth; segment 5 with 2 submedian denticles on each side, with a seta between each pair. Side-plate 1 triangular, strongly expanded forwards, the lower margin as long as the hind margin, with only 2-3 setules, and scarcely any indent near the hind corner. Side-plate 6 in 2 hooked (as in palmata and zey- lanica). Postero-inferior angle of pleon segment 3 rather strongly produced (cf. aculeata or obtusata, and also festiva Chilton, loc. cit., infra), the lower margin with feeble indents. Gnathopod 1 in § more like that of coronini than that of palmata (see Chevreux and Fage, Faune de France, Amphip., figs. 240 and 241, 1925), with upper apex produced in a lobe overhanging base of dactylus which impinges against an oblique setose surface. Gnathopod 2 in ¢ also like that of coroninu, 6th joint broadly oval and in fully-grown specimens somewhat wider distally, but not so strongly expanded as in the figures of palmata given by Chevreux and Fage (loc. cit.) or Sars (1894, Crust. Norw., vol. 1, pl. clxxix). Gnathopods | and 2 in 9 as in palmata. Peraeopods 3-5, 2nd joints spinulose but not serrate on front margin, with feeble indents on hind margin; 4th joints not markedly wider than 5th joint. Length and coloration as in Barnard, 1916. Localities—Sea Point, Table Bay; St. James, False Bay; Klein- mond, near Hermanus (K, H. B., Feb. 1927); Dyer’s Island (J. Drury, 1915); Still Bay (Prof. Stephenson); Port Elizabeth (1 3); Port Nolloth (Prof. Stephenson, 1938); Lambert’s Bay (Prof. Stephenson, 1938). Remarks.—My views on the identity of the South African specimens formerly assigned to Dana’s species have changed. In the first place my statement that side-plate 5 in 9 is not hooked is perfectly correct, but without specific import; side-plate 6 in 2 7s hooked, as in palmata and zeylanica. Closer examination has shown that two distinct forms were con- fused; one with, and one without, a toothed carina on pleon segment 4; with further differences in the hind angle of pleon segment 3, Contributions to the Crustacean Fauna of South Africa. 455 and the first gnathopod in ¢ together with its side-plate. The non- carinate form is recorded below as zeylanica. The carinate form now appears to me to be quite distinct from palmata, and is certainly not the tenwicornis of Walker (1904). In spite of Chilton’s (1909) remarks, I think that wnaequistylis remains in the category in which Stebbing (1906) left it: a species inquirenda. At any rate South Africa should be excluded from the distribution as given by Chilton (1921, Mem. Ind. Mus., vol. v, p. 535) and Schellen- berg (1931, Res. Swed. Ant. Exp., vol. 11, p. 203, as M. gayi Nicolet). The South African form is close to the Australasian M. festova (Chilton) (1916, Tr. N.Z. Inst., vol. xlviui, p. 359, figs. 1, 2) but differs in the ¢ gnathopods. Side-plate 1 is even more markedly triangular in the South African form than in festiva. Also Chilton says the 5th pleon segment in festiva is dorsally produced into two small teeth, whereas in the South African form there are two teeth on each side of the median line (four in all). The specific name refers primarily to the expanded Ist side-plate. The Lambert’s Bay examples (4) agree in all respects except in having no tooth on pleon segment 4. Melita zeylanica Stebb. 21904. Walker in Herdman’s Ceylon Pearl Fish. Suppl. Rep., vol. xvii, p. 273, pl. v, fig. 33 (tenwcornis Dana). 1904. Stebbing, Spolia Zeylanica, vol. 11, p. 22, pl. v. 1916. Barnard, Ann. 8. Afr. Mus., vol. xv, p. 191 (part onaequistylis, non Dana). 21921. Chilton, Mem. Ind. Mus., vol. v, p. 535 (enaequistylis). Peraeon and pleon segments dorsally non-dentate; 5th pleon segment dorsally with a few spinules. Side-plate 1 broadly rounded below, slightly wider than at base, but not markedly produced for- wards, lower margin setose; side-plate 6 in 9 hooked (as in palmata). Postero-inferior angle of pleon segment 3 quadrate, with a short point (cf. palmata). Gnathopod 1, 3, 6th joint oblong, without an apical lobe overhanging base of finger, which is terminal and closes against a setose lobe. Hand of gnathopod 2 in full-grown g widening distally, but not strongly expanded asin palmata. Front margin of 2nd joints in peraeopods 3-5 with widely spaced spinules, but not serrate as represented in Stebbing’s figures; 4th joints also not so broad. Up to 13 mm. Greenish-brown, somewhat mottled, the hind margin of most of the segments somewhat darker. 456 Annals of the South African Museum. Localities.—Little Brak River (Mossel Bay) (K. H. B., Jan. 1931, 33, ovig. 92); Keurbooms River estuary (K. H. B., Jan. 1931); Wilderness lagoon, George District (K. H. B., Jan. 1931); Port Elizabeth; East London (R. M. Lightfoot); Klaasjagers Lagoon, West Coast of Cape Peninsula (K. H. B., March 1938). Distribution.—Lake Negombo, Ceylon (Stebbing). Remarks.—I have little doubt that this is the same as Stebbing’s species. Those I have seen alive agree in coloration. Two small points of difference in the hinder peraeopods are noted above. The Port Elizabeth and East London specimens were probably also collected in an estuarine or brackish-water habitat, as in the other localities. Although Chilton (1909, Subant. Is., New Zealand, vol. u, p. 630) expressed the view that the Ceylonese specimens described by Walker and Stebbing were identical with Dana’s New Zealand species (inaequistylis), I feel that further study would be welcome. Walker’s description of the 4th pleon segment as having a double carina ending in two teeth is curious, and quite possibly may not be a chance variation, but a constant specific character (in spite of Chilton’s and ’ Norman’s remarks). At present I am not prepared to go farther than to identify the South African specimens with Stebbing’s species. Gen. Ceradocus Costa. 1906. Stebbing, Das Tierreich, vol. xxi, p. 430. Ceradocus aviceps n. sp. (Fig. 25.) Integument strongly pitted. None of the peraeon or pleon seg- ments dorsally dentate. Hyes dumb-bell shaped. Side-plates 1-3 very feebly notched at lower hind corner, side-plate 4 deeper than 5, hind margin slightly excavate, a small denticle at lower hind corner, lower hind margin of 7 feebly serrate. Postero-lateral corner of pleon segments 2 and 3 quadrate with a small point, the hind margin feebly crenulate, with a few outstanding setae. Telson cleft to base, lobes divergent, apically notched, with a spine in the notch. Antenna | extending back to middle of pleon, slender, accessory flagellum 11-12 jointed. Antenna 2 slightly shorter than antenna 1, not quite so slender, but not stout, 4th and 5th peduncular joints subequal, flagellum half as long again as 5th joint. Contributions to the Crustacean Fauna of South Africa. 457 Mandibular palp slender, 3rd joint very slightly longer than 3rd, tipped with 2-3 setae. Inner plates of maxillae 1 and 2 strongly setose on inner margins. Gnathopod 1, 5th joint slender, about 34 times as long (measured along upper margin) as wide, 6th joint narrow, about 23 times as long Fic. 25.—Ceradocus aviceps n. sp. a, 4th and 5th side-plates. 6, postero-inferior angle of 3rd pleon segment. c, gnathopod 1. d, gnathopod 2, 3, inner view, with finger further enlarged. e, gnathopod2,°. ff, 2nd joint of peraeopod 5. - g, telson. h, uropod 3. as wide, widening slightly to the transverse palm, finger matching palm. Gnathopod 2, 3, equal on both sides, 5th joint broader than long, 6th robust, as long as the other joints together, oblong-oval, lower distal surface minutely scabrous, with a short blunt ridge apically; finger not hinged dorso-ventrally, but folding inwards on to inner surface of 6th joint, short, stout, strongly hooked, with subacute rudimentary unguis. In Q 5th and 6th joints of normal shape, 6th with oblique palm almost as long as hind margin, with 3 spines at junction but without well-marked defining angle, finger matching palm. Peraeopods 3-5, 2nd joint broader proximally than distally, lower hind corner quadrate, not lobed, hind margin weakly serrate, in peraeopod 3 evenly convex, but in 4 and 5 slightly concave in distal half. 458 Annals of the South African Museum. Uropod 38, peduncle extending not quite to apices of uropods 1 and 2, rami extending far beyond these, both rami elongate oval, with rounded apices, upper margin with spaced spinules, apices and lower margin (especially that of lower or inner ramus) setose. 14-15 mm. Uniform greyish brown. Locality. Palmiet River lagoon, near Kleinmond, under stones at junction of river and tidal areas (K. H. B., March 1937, 3, 9). Remarks.—Agrees with Ceradocus as regards 1st and 2nd maxillae, and mandibular palp, but 4th side-plate better developed than in C. rubromaculatus and the other species, and 1st gnathopod unusually slender. The 2nd gnathopod of ¢ is remarkable for the torsion of the articula- tion between 6th joint and the finger, whereby the latter closes against the inner surface of the former, instead of against the apical margin usually termed the palm. A partial overlapping of the finger on to the inner surface is seen in C. semiserratus (Chevreux and Fage, Faune de France, Amphip., fig. 247, 1925). In the actual shape of 6th joint and finger, the nearest approach to the present form is that of Melita festiva (Chilton) (Chilton, Tr. N.Z. Inst., vol. xlviu, ‘p. 359, fig. 1, 1916). The specific name in allusion to the bird’s-head appearance of the finger of gnathopod 2, 3. Ceradocus rubromaculatus (Stimpson). 1922. Tattersall, J. Linn. Soc. Lond., vol. xxxv, p. 6, pl. 1, figs. 15, 16. 1925. Schellenberg, loc. cit., p. 154. 1936. Pirlot, Siboga Exp. monogr., vol. xxxiii e, p. 305. 1937. Barnard, John Murray Exp. Rep., vol. iv, p. 160, fig. 9. 1937. Monod, Mem. Inst. d’Egypte, vol. xxxiv, p. 10, fig. 6, b-g. 1938. Schellenberg, K. Sv. Vet. Ak. Handl., vol. xvi, p. 63. Additional Locahities.—Luderitzbucht, Swakopmund (Schellenberg) ; Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1935 and 1938). Gen. Maera Leach. 1938. Schellenberg, K. Sv. Vet. Ak. Handl., vol. xvi, pp. 37, 39 (discussion of characters). Schellenberg has shown that there is but one decisive character by which to separate the several species which have been placed in the genera Maera and Elasmopus, to wit: the mandibular palp. Species with a slender palp, the 3rd joint of which is straight and tipped only with a few setae, are assigned to Maera. Contributions to the Crustacean Fauna of South Africa. 459 The following South African species, formerly placed in Elasmopus, are consequently transferred to Maera: subcarinatus (Hasw.), boecki (Hasw.), and levis Brnrd. The last mentioned species while in the genus Hlasmopus was liable to confusion with laevis Holmes, 1905, though it might be claimed that levis is not actually preoccupied by laevis.* On transference to Maera, however, it conflicts with Maera levis 8. I. Smith, 1874, and the name is herewith changed to Maera vagans nom. nov. M. subcarinata is easily distinguished from all the other South African species by the double keel on pleon segment 4. Maera grossimanus (Mont.). 1906. Stebbing, Das Tierreich, vol. xxi, p. 435. 1925. Chevreux and Fage, Faune de France, Amphip., p. 239, figs. 248, 250. 1925. Schellenberg, loc. cit., p. 155. Locality.—Swakopmund. Also Senegal. Remarks.—Schellenberg is inclined to unite hirondellei with grossi- manus, and suspects that my Saldanha Bay specimens (referred to hirondeller, 1916, loc. cit., p. 194) are identical with his material. Chevreux and Fage employ a character to separate inaequipes and hirondelle: on the one hand from grossimanus on the other, which is not considered by Schellenberg, viz. the projecting lower hind corner of 2nd joint of peraeopods 4 and 5. As my specimens have this projecting lobe, I am inclined to retain them as hirondeller. For figure of gnathopod 2 of hirondeller see Monod, 1937, loc. cit., p. 10, ngs. 1, €, 10, c. Maera vagans nom. nov. 1910. Stebbing, Ann. S. Afr. Mus., vol. vi, p. 457 (bruzelii, non Stebbing, 1888). 1912. Chilton, Tr. Roy. Soc. Edin., vol. xlviu, p. 510 (mastersi1, non Haswell, 1879). 1916. Barnard, loc. cit., p. 200, pl. xxvu, fig. 15 (Hlasmopus levis). Additional Locality.—Lambert’s Bay (Prof. Stephenson, 1938). * The Zoological Record gives the spelling of Holmes’s species as laevis. I have not seen the original paper and can check neither the spelling nor the actual date of publication. 460 Annals of the South African Museum. Maera boeckw (Hasw.). 1916. Barnard, loc. cit., p. 199, pl. xxyn, figs. 13, 14 Further specimens (gf and juv. Still Bay, Prof. Stephenson, 1932) confirm the description of the form identified as Haswell’s species. The convexity of the hind margin of 2nd joint of peraeopod 4 is intermediate between that of peraeopods 3 and 5. The mandible was compared with Walker’s figure of that of sokotrae, but it should have been stated that only the trunk is thus short and stout, the palp being considerably longer than in Walker’s figure, more like that figured by Stebbing for bruzelii (Challenger Rep., pl. xcvii). The 3 gnathopod 2 is quite like that of the 9, only slightly heavier. Maera subcarinata (Hasw.). 1910. Stebbing, Ann. 8. Afr. Mus., vol. vi, p. 457 (references). - 1922. Tattersall, loc. cit., p. 9. 1931. Stephensen, Res. on Voy. liad Or. Neerland, vol. i, p. 11. 1936. Pirlot, loc. cit., p. 317, figs. 136-145. { 2 Fic. 26.—Maera subcarinata (Hasw.), mandibular palp and hand of gnathopod 2, 3 ae omitted). Two specimens ($ 2) from Natal. The mandibular palp is very slender, and, contrary to Walker’s statement and Stebbing’s figure (Challenger Rep., pl. xcevill, persetosus), the 3rd joint is distinctly shorter than the 2nd, with only 2 apical setae (cf. Pirlot’s figures 137, 143). The hand of gnathopod 2, 3, is in general agreement with Walker’s figure. Gen. Llasmopus Costa. 1932. Stephensen, Annot. Zool. Jap., vol. xiii, p. 487 (synopsis of species. LH. latebrachium is listed under wrong heading). 1936. Schellenberg, Zool. Anz., vol. cxvi, p. 153. 1938: id. loe. cts pp. ot, o2. Contributions to the Crustacean Fauna of South Africa. 461 Mandibular palp robust, its 3rd joint more or less faleate, and with a comb-like row of spine-setae. After the transference of subcarinata, boecki and levis (v. supra = vagans) there are only two South African species to be assigned to this genus. Elasmopus pectenicrus (Bate). O06. Barnard, loc.‘cdt., p. 197, pl. xxviii, fig. 33. 1917. Stebbing, Ann. Durban Mus., vol. i, p. 446 (brasiliensis, non Dana). 1928. Schellenberg, loc. cit., p. 647. 1936. Pirlot, loc. cit., p. 312. 1937. Barnard, John Murray Exp. Rep., vol. iv, p. 161. There is often a small denticle at base of dactylus in 2nd gnathopod, $, fitting in between the hinge and the tooth on palm. Colour: peraeon and pleon purplish-grey with white speckling, a dark medio- dorsal spot on hind margin of peraeon and pleon segments (sometimes only on peraeon, or only on peraeon segments 1-4), a dark lateral spot anteriorly on pleon segments 2-4, dark spots on distal joints of peraeopods 3-5, the 2nd joints pale, eyes black with a yellowish mark behind them. Additional Localities.—Still Bay and East London (Prof. Stephenson, 1935 and 1937). Distribution.—Indian Seas, East Indies. Also West Indies. Elasmopus japonicus Steph. 1925. Barnard, loc. cit., p. 358 (spinimanus, non Walker). 1932. Stephensen, loc. cit., p. 490, figs. 1, 2. The fact that the South African specimens have a dorsal keel (albeit rather low) on pleon segment 4 excludes them at once from being Walker’s spinimanus. Comparison with Stephensen’s figures leaves no doubt as to their correct identity. Additional Localities.—Isipingo, Natal; Port Elizabeth (Prof. Stephenson). | Distribution.—Japan. Parelasmopus suluensis (Dana). 1888. Stebbing, Challenger Rep., vol. xxix, p. 1029, pl. c. 1904. Walker in Herdman’s Ceylon Pearl Fish. Rep., vol. xvii, p. 278, pl. vi, fig. 38. 462 Annals of the South African Museum. 1935. Barnard, Rec. Ind. Mus., vol. xxxvii, p. 286, fig. 6. 1936. Pirlot, Siboga Exp. monogr., vol. xxxiii e, p. 311. 1938. Schellenberg, K. Sv. Vet. Ak. Handb., vol. xvi, p. 62, Two 36 from Still Bay (Prof. T. A. Stephenson, 1932, littoral) are white with violet markings and speckling on head, peraeon and pleon, spots and bands on antennae, legs and uropods, the posterior half of the 2nd joint of peraeopods 3-5 also violet; eyes black. Distribution.—Sulu Sea, Marshall and Solomon Is., Indian Seas to N.W. Australia, Red Sea, British East Africa. Stebbing (1922) records a young specimen from Angola. Gen. Gammarus Fabr. 1916. Barnard, Ann. 8. Afr. Mus., vol. xv, p. 202. 1926. Schellenberg, D. Sudpol. Exp., vol. xviii, p. 367 (Para- melita). | 1927. Barnard, Trans. Roy. Soc. 8. Afr., vol. xiv, pp. 167 sqq. 1937. Schellenberg, Zool. Jahrb. Abt. Syst., vol. lxix, pp. 469 sqq., esp. pp. 472, 480, 481. Schellenberg in his latest paper, comprising very valuable critical remarks on the fresh-water Gammarus species of the world, maintains Paramelita as a full genus. He considers that it has affinities with Melita rather than with Gammarus, and that it may in fact be derived from Melita-like marine ancestors, thus avoiding the supposition of a very ancient cosmopolitan distribution of Gammarus-like forms. The argument is considerably weakened by the presence of admittedly typical Gammarus forms in parts of Australia. The taxonomy and phylogeny of the genus Gammarus are beset with so many difficulties that, pace the high authority of Dr. Schellen- berg, I am unwilling as yet to admit Paramelita to more than sub- generic rank. The particular features of Paramelita appear to be inner margins of inner lobes of lst and 2nd maxillae with long setae or bristles only distally (proximally there are usually shorter and very fine setae which pass gradually into the longer distal setae; in Gammarus sensu stricto the whole inner margin of these lobes is fringed with long setae or bristles); palm of lst gnathopod, 3, with only a few spines; presence of accessory branchiae; dactylus (7th joint) of 1st—5th peraeopods with several (usually) spines on inner margin (reduced to 1 or 2, or absent altogether, in auricularius); inner ramus of 3rd uropod considerably shorter than outer ramus. Contributions to the Crustacean Fauna of South Africa. 463 Individually these characters, or some of them, are not decisive, but taken in combination they may serve to indicate that the South African representatives are not quite typical Gammarus. Fam. DEXAMINIDAKE. Polycherva atolli Wlkr. 1905. Walker, Fauna Geogr. Mald. Laccad. Arch., vol. ii, p. 926, pl. Ixxxvini, figs. 1-5. 1916. Barnard, loc. cit., p. 211 (antarctica, non Stebbing). 1925. Schellenberg, loc. cit., p. 157. 1930. Barnard, Terra Nova Exp. Zool., vol. viii, p. 390. Additional Localities.—Luderitzbucht (Schellenberg); Lambert’s Bay (Prof. Stephenson, 1938); Still Bay (Prof. T. A. Stephenson, 1932). Distribution.—Maldives, Seychelles, and British East Africa. Fam. TALITRIDAE. Gen. Talitrus Latr. 1906. Stebbing, Das Tierreich, vol. xxi, p. 524 and p. 527 (Tali- trovdes). 1925. Hunt, J. Mar. Biol. Assoc. Plymouth, vol. xii, p. 854. 1934. Schellenberg, Zool. Anz., vol. cv, p. 159. 1934. Burt, Spolia Zeylanica, vol. xviu, p. 181 (subgen. Talitropsis). Hunt discussed the various characters which have been used to differentiate the terrestrial species from the typical saltator, and came to the conclusion that two groups could not be differentiated. Schellenberg is in favour of retaining the terrestrial species in a genus separate from the typical littoral species, and shows that Talitriator Meth. falls into the synonymy of Talitroides (Bonnier) Stebb. It would seem that Burt’s subgenus should also become a synonym. Apparently the decision as to which of these two courses should be adopted rests on the value to be attached to the shape of the joints of the maxilliped palp and its armature, a character to which Hunt drew attention (loc. cit., p. 857, fig. 2). TT. saltator seems to be unique in having the joints (including the outer plate) broad and densely set with stout spines; in the other species they are narrower and bear only a few slender spine-setae near the apices. No transitional state 464 Annals of the South African Museum. has been recorded (except perhaps in 7. gulliveri. See Barnard, 1936, Ann. Natal Mus., vol. vii, p. 12). On the other hand transitions or variations have been observed in all the other characters, especially in the pleopods. I suggest therefore that the maxilliped palp and the 2nd joint of peraeopod 3 be regarded as the diagnostic features of Talitrus (s.s.) and that, following Schellenberg, all the other species be grouped together either generically or subgenerically. Gen. Talitroides (Bonnier) Stebb. 1898. Bonnier in Willem. Ann. Soc. Entom. Belge, vol. xlii, p. 208. 1906. Stebbing, loc. cit., p. 527. 1913. Methuen, Proc. Zool. Soc. Lond., p. 109 (Taltriator). 1916. Barnard, loc. cit., p. 222 (Talitriator). 1917. Stebbing, Ann. Mag. Nat. Hist. (8), vol. xix, p. 330 (April). 1917. Chilton, Trans. New. Zeal. Inst., vol. xlix, p. 294 (Aug.). 1934. Schellenberg, loc. cit., p. 159. Genotype: T. alluaudi Chevr., 1896, Seychelles. After my 1916 identification of specimens from numerous localities as belonging to Methuen’s species, Stebbing (1917) on the basis of some Natal specimens came to the conclusion that eastwoodae was identical with Bate’s africanus. A revision of my former material, together with a lot of new material, shows unfortunately that the matter is not quite so simple. The expansion of the 5th joint of gnathopod 1 was noted by Methuen, and incorporated in my diagnosis and Hunt’s table (loc. cit., p. 860). Bate’s figure, however, shows a linear (or cylindrical) 5th jot. One might pass over this detail in the figure of the whole animal, but Bate has also given an enlarged figure of this joint, in which the linear shape is unmistakable. Amongst my material there are specimens in which this joint is linear, but not one of these specimens comes from a locality east of 26° H.long. All the specimens from Natal, including some from Stella Bush, Durban (Port Natal was the old name for Durban), Zululand and the Hastern Cape Province have the expanded 5th joint as found in eastwoodae. On the other hand they all agree with Bate’s description and figure as regards the length of the lst antennae, and the pleopods are just long enough to show their tips below the pleurae as in Bate’s figure; the 2nd joint of peraeopod 3 in his figure, however, is too broad for our specimens. Contributions to the Crustacean Fauna of South Africa. 465 Although I admit that there is much justification for Stebbing’s view, I prefer to wait until specimens from the type locality showing a linear 5th joint in gnathopod 1 are available, before deciding definitely on the position of Bate’s species; and in the meantime to retain Methuen’s specific name. Whether all the South African forms should bear this name is a question which has been considerably complicated by the study of all the material now at hand. The eastern form is clearly demarcated by the characters given below, and cannot be separated from Talitrus except on the two characters given above (mxp. palp and 2nd joint peraeopod 3); the pleopods are neither degraded nor even reduced in size, and the length of the Ist antennae is connected by intergrading forms with the very short antennae of 7. saltator. This form I have called eastwoodae forma typica. All the specimens from the western Cape Province exhibit shorter lst antennae and stouter pleopods, the latter with a varying degree of degradation, at least of the inner rami. As regards the 5th joint of gnathopod 1 it is linear in specimens from some localities, and expanded in those from other localities. I have separated off four forms; no doubt with still more detailed study of further material other forms, races or local varieties, could be distinguished, but for the present I think discrimination has gone far enough. Reference must be made to the very curious form with enlarged ungues in the peraeopods, which up to now has only been found in one locality in the-S.W. Cape mountains. Talitroides eastwoodae (Meth.). (Fig. 27.) #303. Methuen, loc. cit., p..110, pls. x, xi. 1916. Barnard, loc. cit., p. 223. 1917. Stebbing, loc. cit., p. 330. Forma typica. (Fig. 27, a-d.) The 5th joint of gnathopod 1 expanded, 1.e. the inner margin more or less prominently rounded, the length (along upper margin) not more than twice the greatest width. 6th joint ovate, with indistinct palm (eastwoodae), but in larger and mature specimens more parallel- sided, with very short but distinct palm. 466 Annals of the South African Museum. Antenna 1 extending at least to half length (cotypes of eastwoodae, and samples from Pinetown, Natal) of 5th peduncular joint of Fic. 27.—Talitroides eastwoodae (Meth.). Forma typica. a, 5th—7th joints of enathopod 1, immature. 6, the same, adult. c, pleopod 1. d, pleopod 3. Forma cylindripes. e,gnathopod 1. f,9,h, pleopods 1-3. Forma macronyz. 2, gnathopod 1. Jj, 6th and 7th joints of peraeopod 4 (peraeopod 5 similar). antenna 2, usually to 2 or #, or even to end of peduncle (Durban). Pleopods slender, elongate, outer margin of peduncle even (nearly Contributions to the Crustacean Fauna of South Africa. 467 straight) with numerous close-set plumose setae along whole length (cotypes eastwoodae), or along distal half, or only a few at distal 4 (Durban); the rami subequal (the outer slightly the longer), more or less distinctly jointed. Localities —Transvaal: Woodbush (eastwoodae). Natal: Pietermaritzburg, Howick, Pinetown, Kark- loof, Durban. Zululand: M’fongosi. Eastern Cape Province: Port St. Johns, Doornnek (Alexandria Div.). Gt. Winterberg, 7000 ft. (Fort Beaufort Div.). All these localities are east of about 26° E. long. Forma cylindripes. (Fig. 27, e-h.) The 5th joint of gnathopod 1 more or less parallel-sided, length (along upper margin) at least 24 times, usually 3 times, the width, inner margin nearly straight or only very slightly convex. 6th joint elongate, ovate, tapering, with very feeble, scarcely developed, palm. Antenna 1 extending } to 4 length of 5th peduncular joint of antenna 2, usually 4-4. Pleopods stout or moderately stout, peduncle distally more or less swollen, outer margin sinuous, with plumose setae on distal half or third; rami stout, more or less distinctly unequal (the outer the longer), not jointed. Localities —S.W. Cape Province: Cape Peninsula, Sir Lowry Pass, Houw Hoek, Mossel River (Hermanus), Gt. Winterhoek Mts., 3500 ft. (Tulbagh). Forma setosa. The 5th joint of gnathopod 1 expanded (as in eastwoodae). Antenna 1 extending +-4 (usually 4) along 5th peduncular joint of antenna 2. Pleopods stout, peduncle with close-set, plumose setae along whole of sinuous outer margin, rami subequal or distinctly (though not greatly) unequal. Localities—S.W. Cape Province: Cape Peninsula, Stellenbosch, Tulbagh and Gt. Winterhoek Mts., Somerset West, Ceres, River Zonder End Mts., Onderberg Vlei (Clanwilliam). A variety of this form occurs in Orange Kloof, Table Mt., in which the peduncle of the pleopods is setose only on the distal third. 468 Annals of the South African Museum. Forma calva. As in forma setosa, but the peduncle of pleopods non-setose, the rami unjointed, but the incisions marking the limits of the fused joints very deep, so that the ramus becomes a series of subglobose segments. Localities.—Western Cape Province: Cape Peninsula, Wellington, Swellendam Mts., northern slopes of Matroosberg (Hex River Mts.), George, Knysna, Zwartberg Range at Meiring’s Poort (Oudtshoorn Distr.). Forma macronye. (Rig sie e.4.) The 5th joint of gnathopod | ovate, inner margin subparallel with outer margin, almost straight or only slightly convex; 6th oblong, parallel-sided, palm distinct though short (equal to basal width of 7th joint), transverse, with rounded angle bearing 2 very strong spines and a smaller one; 7th joint bearing an unusually strong spinule on inner apex; unguis elongate, twice as long as 7th joint. Ungues of peraeopods 1-5 longer than their 7th joints. Antenna 1 extending half-way along 5th peduncular joint of an- tenna 2. Pleopods stout, peduncle setose along whole of outer margin, rami stout, subequal (outer slightly the longer), not jointed. Locality.—S.W. Cape Province: Hottentots Holland Mts., 4000 ft. (Somerset West) (Jan. 1916, K. H. B.). Gen. Talorchestia Dana. 1916. Barnard, loc. cit., p. 215. 1917. Chilton, Trans. New Zeal. Inst., vol. xlix, p. 293 (N. Zealand species). 1922. Stebbing, Goteb. K. Vetensk. Vitt. Handb., vol. xxv, no. 2, pao: 1922. Tattersall, Mem. Asiat. Soc. Bengal, vol. vi, p. 454. Tattersall has given a key to some of the species, including two of the South African ones. He has, however, transposed ancheidos and australis, and assumed that in the latter (“‘ancheidos” in his key) ‘‘side-plates 2-4” are without a well-marked lobe on hind margin, whereas my description referred to side-plate 2 only (“side-plate . . 2 without a strongly produced lobe”). As a matter of fact this character should not be used; the lobe in question is present in australis but not nearly so prominent as in the other species. Contributions to the Crustacean Fauna of South Africa. 469 A synopsis of the four South African species is given below (p. 536) which permits the identification not only of $3, but also of 99. A very interesting feature has been noted in ancheidos, which was Fic. 28.—Talorchestia capensis (Dana). a,gnathopod1,g. 6, hand of gnathopod 2, 5, immature. c, the same, adult (inner view). d, cross-section of 2nd joint of gnathopod 2, 3 (outer side to right). e, distal joints of gnathopod 2, 9. f. 9, 2nd joints of gnathopod | and 2 respectively, 2 (anterior margin below). h, 2nd joint of peraeopod 5. 7, pleon segments 1 and 2, g. j, pleon segment 3, 0: entirely overlooked in my original description. viz.: the minute corrugations or ridges on the lower margins.of pleon segments 2 and 3 (in both sexes), causing an appearance of crimping. It occurs also in Orchestia floresiana Weber. See: Stephensen, 1935, B. P. Bishop Mus. Bull., vol. exlii, p. 24, and Barnard, 1935, Rec. Ind. Mus., vol. xxxvii, p. 288, fig. 7; but see also Schellenberg’s remarks on VOL. XXXII, PART 5. 33 470 Annals of the South African Museum. Stephensen’s and my identifications (Zool. Anz., vol. cxvi, pp. 155, 156; 1936). T. tricornuta Shoemaker, 1920, from both sides of the Congo River mouth, is near quadrispinosa, but has only 3 dorsal spines, and differently shaped gn. 2 and prp. 4 in ¢. T. skoogi Stebb., 1922, from Port Alexandra, has a distinctive 2nd gnathopod in g. T. landanae Schell, 1925, from Portuguese West Africa, is near ancheidos, but differs in the palm of gn. 2, 3, and the hind margin of 2nd joint of prp. 5. Talorchestia capensis (Dana). (Fig. 28.) 1916. Barnard, loc. cit., p. 216. Locality.—Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, 33, 22 on ocean beach); Port Nolloth (Prof. Stephenson, 1935, gd, ovig. 92, juv.). Talorchestia quadrispinosa Brurd. (Fig. 29.) 1916. Barnard, loc. cit., p. 217, pl. xxvul, figs. 29-32. 1925. Schellenberg, loc. cit., p. 159. Localities. —Luderitzbucht, Prince of Wales Bay, Walfisch Bay (Schellenberg); Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938). Talorchestia australis Brurd. (Fig. 30.) 1916. Barnard, loc. cit., p. 220, pl. xxvu, figs. 33, 34. Locality. Kleinmond, mouth of Bot River (Caledon Division) (Kee Bevan: Lote so eo): Talorchestia ancheidos Brurd. (Fig. 31.) 1916. Barnard, loc. cit., p. 221, pl. xxvu, figs. 35, 36. Localities.—Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, 33, 22 under logs, dead Zostera and other debris on margins of estuary); Masiene (near Chai Chai), Portuguese Hast Africa (R. F. Lawrence, 1924, gd, 2). Contributions to the Crustacean Fauna of South Africa. 471 c Fic. 29.—Talorchestia quadrispinosa Brnrd. a, gnathopod 1, g. 6, gnathopod 2, 3 (inner view). ¢, cross-section of 2nd joint of gnathopod 2, g. d, gnathopod 2,2. e, 2nd joint of peraeopod 5. ff, pleon segment 3. Fic. 30.—Talorchestia australis Brnrd. a, gnathopod 1, g. 6, gnathopod 2, ¢ (inner view). c, cross-section of 2nd joint of gnathopod 2, g. d, gnathopod 2,2. e, 2nd joint of gnathopod 2, 2 (anterior margin below). ff, 2nd joint of peraeopod 5, with hind margin further enlarged. g, pleon segment 3. 472 Annals of the South African Museum. Remarks.—The Masiene specimens have a slightly more robust 6th joint on gnathopod 2, 3, and a slightly broader 6th joint on gnatho- = Cc e Fie. 31.—Talorchestia ancheidos Brnrd. a, gnathopod 1, g. 6, gnathopod 2, g. c, 2nd joint of gnathopod 2, 2 (anterior margin to right). d, 2nd joint of peraeopod 5, with hind margin further enlarged. e, pleon segment 3. f, enathopod 1,¢. g, gnathopod 2, g. h, cross-section of 2nd joint of gnathopod 2,4. (f, 9, h, specimens from Masiene, Portuguese Hast Africa.) pod 1, 3, so that the finger extends only to the base, not the apex, of the apical lobe. Gen. Parhyale Stebb. 1906. Stebbing, Das Tierreich, vol. xxi, p. 556. 1934. Iwasa, J. Fac. Sci. Hokkaido, Zool., vol. ii, p. 1. Parhyale inyacka (Brurd.). 1916. Barnard, loc. cit., p. 233, pl. xxvii, fig. 4 (Hyale 7.). 1925. Chevreux, Bull: Soc. zool. France, vol. J, p. 370, fig.-17 (Hyale 1.). Contributions to the Crustacean Fauna of South Africa. 473 1933. Stephensen, Zool. Jahrb. Abt. Syst., vol. lxiv, p. 441, figs. 3, 4 (Hyale 7.). Specimens from Senegal (Chevreux) and from Bonaire Is., Dutch West Indies (Stephensen), have been identified with this species. Stephensen showed the presence of a rudimentary inner ramus on uropod 3, the same has been confirmed in Chevreux’s specimens by Fage and Monod (1936, Arch. Zool. Exper. Gen., vol. lxxviii, p. 105, figs. 3-7); and it is present also in the types of inyacka although entirely overlooked by me in 1916. All these specimens therefore are correctly transferred to the genus Parhyale. Fage and Monod, however, refer them all to the West Indian fasciger Stebb., 1897. As P. fasciger has been recorded from Cameroons (Schellenberg, 1925), there is perhaps some justification for suspecting that all the Atlantic records of wnyacka are referable to Stebbing’s species. Stebbing’s figure of the 2nd gnathopod, 2, however, is quite different from that of the type of inyacka (and from Stephensen’s 1933 figure). Until a long series of all stages of both sexes from Delagoa Bay and Atlantic localities has been compared, I prefer to keep wnyacka as a separate species, especially as the hind margin of 2nd joint of peraeopod 5 is more strongly serrate than in Stephensen’s figure, and much more strongly so than in the figures of Stebbing and of Fage and Monod. Gen. Parorchestia Stebb. 1906. Stebbing, Das Teirreich, vol. xxi, pp. 557, 735. 1916. Barnard, loc. cit., p. 226. 1922. Stebbing, Goteb. K. Vet. Vitt. Handl., vol. xxv, no. 2, p. 9. The finding of some specimens at a locality intermediate between Cape Town and Hast London (the extreme west and east localities recorded by me in 1916) has led to the re-examination of all the material, and I have come to the conclusion that the species formerly identified as the New Zealand tenwis should really be regarded as a separate species. Parorchestia rectipalma n. sp. (Fig. 32.) 1916. Barnard, loc. cit., p. 226 (tenuis, non Dana). It is unfortunate that the adult ¢ of the New Zealand tenuis has, apparently, never been figured. Thomson’s Allorchesies recens is regarded as synonymous with tenuis, and the hand of the gnathopod 2 of the immature South African $ corresponds fairly closely with 474 Annals of the South African Museum. Thomson’s figure (Trans. N. Zeal. Inst., vol. xvi, pl. 13, figs. 2-5, 1884). But if Thomson figured an adult 3, then there is considerable difference between his species and the South African one. Chilton (1909) described four new species from the New Zealand region alone, and a fortiort it seems very unlikely that one species should inhabit both New Zealand and South Africa. I therefore Fic. 32.—Parorchestia rectipalma n. sp. a, gnathopod 1, g. 6, gnathopod 2, 3 (adult). c, gnathopod 1, 9. d, gnathopod 2, 9. propose a new specific name for the South African specimens, and give figures of the lst and 2nd gnathopods in both sexes. Locality.—Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, 3d, ovig. 99 and juv., beneath logs and debris washed up on the banks of the estuary); estuary of Klip Drifts Fontein stream, Potteberg, Bredasdorp District (A. C. Harrison, May 1936). Parorchestia dassenensis Brnrd. 1916. Barnard) loc: citi p 220 uplexxvains ties. lege 1922. Stebbing, loc. cit., p. 10, pl. iv (tenuis, non Dana). Stebbing’s statement that gnathopod 2, 3, “makes a near agreement with G. M. Thomson’s figure” is incompatible with his own figure of this imb. On the contrary the palm with its two notches is very similar to that of dassenensis, except that the proximal notch is deeper in Stebbing’s figure. I have no doubt that the two forms are conspecific. Locality.—Dyer’s Island, near Danger Point (Stebbing). Contributions to the Crustacean Fauna of South Africa. 475 Hyale saldanha Chilton. (Fig. 33.) 1916. Barnard, loc. cit., p. 229, pl. xxvii, fig. 37. 1925. Schellenberg, loc. cit., p. 162. Localities. — Luderitzbucht (Schellenberg); Port Nolloth and Lambert’s Bay (Prof. Stephenson); Dyer’s Island (S.A. Museum). Fic. 33.—Hyale saldanha Chilton. Palp of maxilliped, ? on left, g on right. Remarks.—Neither Chilton, nor myself, nor Schellenberg have referred to the notable sexual difference in the palp of the maxilliped (fig. 33). The example figured is an extreme case; usually the palp is not quite so much larger in the ¢ than in the 9, but nevertheless the difference is always noticeable. So far as I am aware such a feature has not been recorded for any other species of the genus, though in camptonyx the apical seta is stated by Stebbing (Das Tierreich, vol. xxi, pp. 561 (key), 570) to be very long in 3, implying that it is longer in 3 than in 9. Gen. Allorchestes Dana. 1906. Stebbing, loc. cit., pp. 581, 736. 1926. Chilton, Trans. New Zeal. Inst., vol. lvi, p. 515. 476 Annals of the South African Museum. The character distinguishing this genus from Hyale is the small lobe of the 5th joint of gnathopod 2 in ¢ (fig. 34). The 99 of the two genera appear to be indistinguishable. A recent examination of the examples previously referred to Hyale grandicornis (1916, Barnard, loc. cit., p. 230), together with further material, has shown that in some samples the gg had the lobed 5th joint characteristic of Allorchestes. The lobe was found to be present in the youngest specimens recognizable as gg, and was as a Fie. 34.—Hyale grandicornis (Kroyer). a, hand of gnathopod 2. Allorchestes inquirendus n. sp. 6, hand of gnathopod 2. c¢, process of 5th joint of gnathopod 2. well developed as in 99, extending to the lower apex of the 4th joint. In later stages the lobe tended to become smaller, extending only about half-way along the distal margin of 4th joint. But even in the largest and fully adult gg it was present as a thin plate set trans- versely between the 4th and 6th joints, with an apical fringe of plumose setae. , In other lots, however, there was only a very small conical knob forming the outer apex of 5th joint, as is found in other species of Hyale; even in the youngest $¢ only this knob and no trace of a setigerous lobe was found. It seems, therefore, that two species, assigned to different genera, must be recognized, until perhaps breeding experiments can show whether the character has any diagnostic value. In other respects the two forms are indistinguishable, though there is a tendency for the 2nd antennae to be slightly longer in Hyale grandicornis, and for spines to be absent from the outer ramus of Ist uropod in the Allorchestes specimens. Contributions to the Crustacean Fauna of South Africa. 477 On the other hand, my remarks (loc. cit.) on the variability of characters holds good. All the Allorchestes come from localities in the Cape Peninsula, east and west sides, except one lot from Port Elizabeth; Hyale grandicornis occurs as far east as Natal, and has not yet been found on the west side of the Cape Peninsula, as the following list of localities will show :— Allorchestes. | H. grandicornis. Table Bay . x West side of Cape | Oudekraal . % Peninsula Hout Bay . x Near Cape Good Hope. x Kast side of Cape { : Peninsula Kalk Bay and St. James x x (False Bay) Bufiels Bay 3 x Kleinmond (Caledon Dives) ae x Stull Bay . aa x South and south-}‘Keurbooms R. , Pletten- east coast berg Bay ; ce x Port EKlizabeth . x Kast London Ga x Port Shepstone, Natal. ya x Allorchestes inquirendus n. sp. (Fig. 34, 6, c.) To the features given for Hyale grandicornis (1916, loc. cit., p. 231) may be added: Gnathopod 1, ¢, similar to Stebbing’s figure of that of A. humilis (ieee. trans: Linn. Soc. Lond., vol.°vii, pl. xxxni, D) the 2nd joint strongly expanding from a narrow base, the hind margin convex. | Gnathopod 2, 4, fig. 34, 6. The lobe of 5th joint relatively larger in younger specimens. In the 2 gnathopods 1 and 2 as in Stebbing’s figures (loc. cit., pl. xxxii, A, B), the hind margin of 2nd joint in gnathopod 1 being convex asin the 2. Peraeopods 3-5. Cf. Chilton’s figure of peraeopod 3 in A. now- 478 Annals of the South African Museum. zealandiae (loc. cit., fig. 3, e), but the spines on hind margin of 4th joint usually set in a tuft of setules, and 6th joint relatively longer. Palp of maxilla 1 reaching to bases of the apical spines of outer lobe. Telson cleft to base, apices of lobes subacute. Fam. AORIDAE. - Aora typica Kroyer. 1916. Barnard, loc. cot., p. 236 (references). 1926. Schellenberg, loc. cit., p. 372, fig. 59 (forma anomala). 1932. Barnard, loc. cit., p. 220, fig. 135 (forma gibbula). Localities.—False Bay (Schellenberg; R.R.S. “Discovery”’); Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938). Lembos leptocheirus Wlkr. 1909. Walker, Trans. Linn. Soc. Lond., vol. xu, p. 338, pil. x, Ae | 1926. Schellenberg, loc. cit., p. 373. 1928. Id., loc. cit., p. 662. Locality.—Simon’s Bay (Schellenberg). Distribution.—Port Said; Suez; British East Africa. Lembos hypacanthus Brurd. 1925. Schellenberg, loc. cit., p. 166. Locality.—Swakopmund. Lembos hirsutipes Stebb. 1895. Stebbing, Ann. Mag. Nat. Hist. (6), vol. xvi, p. 207, pls. viii, ix. | 1925. Chevreux, Bull. Soc. zool. France, vol. 1, p. 373. Distribution.—Dakar, Senegal. Gen. Grandidierella Coutiére. 1904. Coutiére, Bull. Soc. Philom., ser. 9, vol. vi, pp. 166, 173. 1908. Stebbing, Rec. Ind. Mus., vol. 11, p. 120. 1921. Chilton, Mem. Ind. Mus., vol. v, p. 548. 1922. Tattersall, Mem. Asiat. Soc. Bengal, vol. vi, p. 455. 1925. Chevreux, Bull. Soc. zool. France, vol. 1, p. 392. 1925. Schellenberg, loc. cit., p. 166. Contributions to the Crustacean Fauna of South Africa. 479 1933. Stephensen, Zool. Jahrb. Abt. Syst., vol. lxiv, pp. 434, 446. 1935. Shoemaker, J. Wash. Ac. Sci., vol. xxv, p. 65. 1935. Barnard, Rec. Ind. Mus., vol. xxxvii, p. 295. 1936. Schellenberg, Zool. Anz., vol. cxvi, p. 154. 1938. Id., K. Sv. Vet. Ak. Handl., vol. xvi, p. 90 (transferred to Corophidae). Grandidierella lignorum Brurd. 1935. Barnard, loc. cit., p. 300, fig. 14. Locality.—Keurbooms River, Plettenberg Bay (K. H. B., Jan. 1931, 3S, 22 under water-logged tree-trunks and driftwood in estuary); Zoetendal Vlei, Bredasdorp Division (A. C. Harrison, July 1937). Fam. PHOTIDAE. Photis uncinata Brurd. 1916. Barnard, loc. cit., p. 243, pl. xxviii, fig. 26 (longicaudata, non Bate and Westw.). 1932. Id., loc. cit., p. 223, fig. 138. Photis longimanus Wlkr. 1925. Schellenberg, loc. cit., p. 175. Locality.—Luderitzbucht. Eurystheus palmoides Brnrd. 1932. Barnard, loc. cit., p. 231, fig. 144. Locality.—Simon’s Bay (R.R.S. “ Discovery’’). Eurystheus immunens Brurd. 1916. Barnard, loc. cit., p. 250, pl. xxvii, fig. 12. 1937. Id., John Murray Exp. Rep., vol. iv, p. 165, fig. 11. Distribution.—Red Sea. Eurystheus afer (Stebb.). 1916. Barnard, loc. cit., p. 249, pl. xxviu, fig. 11. 1928. Schellenberg, Tr. Zool. Soc. Lond., vol. xxii, p. 662. 1937. Barnard, loc. cit., p. 165, fig. 12. Distribution.—Gulf of Suez and Zanzibar area. Remarks.—Cf. Monod, 1937, Mem. Inst. d’Egypte, vol. xxxiv, p. 10, fig. 10, 6. 480 Annals of the South African Museum. Cheiriphotis megacheles (Giles). 1937. Barnard, loc. cit., p. 167, fig. 14 (references and synonymy). Distribution.—Bay of Bengal, Ceylon, Hast Indies. Fam. AMPITHOIDAE. Ampithoé ramondi (Audouin). 1826. Audouin, Expl. Pl. Crust. Descr. Hgypte, vol. i, p. 93. 1916. Barnard, loc. cit., p. 253 (vacllantic) (references). 1928. Schellenberg, loc. cit., p. 665. 1937. Barnard, John Murray Exp. Rep., vol. iv, p. 170. 1938. Pirlot, Siboga Exp. monogr., vol. xxxili f, p. 346. Additional Localities.—Port Nolloth and Lambert’s Bay (Prof. Stephenson, 1938). Distribution.—Atlantic to Azores, Mediterranean; Indian Ocean, Southern Pacific. Ampithoé falsa Brnrd. 1916. Barnard, loc. cit., p. 255, pl. xxvii, fig. 34 (brevipes, non Dana). 1932. Id., loc. cit., p. 240. IOSt ld loc. sert.. p. V0; ae. 6, Localhity.—Still Bay (Prof. T. A. Stephenson, 1935). Disiribution.—Gulf of Aden, and Central Arabian Sea, on drift- weed. Gen. Cymadusa Sav. 1816. Savigny, Mem. An. sans Vert., vol. i, p. 109. 1868. Czerniavski, Syezda Russ. Est. Syezda I. Zool., p. 103 (Grubia). 1916. Barnard, loc. cit., p. 257 (references) (Grubia). 1938. Pirlot, Siboga Exp. monogr., vol. xxxiu f, p. 348. Cymadusa australis (Brurd.). 1916. Barnard, loc. cit., p. 258. Locality— Knysna Lagoon (C. W. Thorne, 1938). Macropisthopous stebbing: Brunrd. 1916. Barnard, loc. cit., p. 260, pl. xxvii, figs. 15-17. Localities—Sea Point, Table Bay (K. H. B., March 1928); Still Bay (Prof. T. A. Stephenson, 1935). Contributions to the Crustacean Fauna of South Africa. 481 Fam. JASSIDAE. Ischyrocerus gorgoniae n. sp. (Fig. 35.) 3.—Body strongly triquetral in cross-section anteriorly; in dorsal view resembling a member of the family Phliantidae. Head not Fig. 35.—Ischyrocerus gorgoniae n. sp. a, animal, 3, with gnathopod 2. b, peraeopod 1. c, peraeopod 5. carinate. Peraeon smooth, medio-dorsally carinate on segments 1-6; the keel on segment 1 moderate, those on segments 2-5 strong, and that on segment 6 feeble. Peraeon segment 7 and pleon not carinate. Side-plate 1 small, deeper than long, subtriangular; 2 very large, much deeper than long, ovoid; 3 and 4 deeper than long, oblong, antero-inferior corner rounded, postero-inferior corner quadrate; 5 with deep incision between the lobes, the hind lobe much the smaller; 6 similar to 5 but smaller; 7 small, ovoid. Postero- inferior angle of pleon segment 3 rounded. Telson broader than long, apically obtuse, a spinule on each lateral margin. Antenna | with 2-jointed accessory flagellum, the 2nd joint minute, flagellum missing. Antenna 2 missing. Mouth-parts typical. Gnathopod 1 asin megalops Sars (Crust. Norw., vol. i, pl. 210, fig. 2): Gnathopod 2, 2nd joint very long, curved, margins entire, antero- distal corner moderately lobed, 3rd with anterior margin lobed, 6th elongate ovate, upper margin convex, lower margin straight, with a narrow subacute tooth near hinge, finger a thin cultrate plate, 482 Annals of the South African Museum. lanceolate, unguis indistinguishable (the finger may be abnormal in shape, the hand of the other gnathopod is missing). Peraeopods 1 and 2, 2nd joint flask-shaped, antero-distal margin expanded. Peraeopods 3-5 slender, 2nd joint rather broadly oval, 6th joint relatively longer than in any of Sars (loc. cit.) figures. Uropods typical. Length.—4:5 mm. Colour.—(As preserved) chalky-white, eyes darker. Locality.—Somerset Strand, False Bay. 1 3, Sept. 1926, found on a piece of Red Fan-coral, Gorgonia flammea. Remarks.—This species is remarkable for its Phliantid appearance in dorsal view, and the dorsal carination, which is much stronger than in carinatus Brurd. Fam. COROPHIIDAE. Gen. Corophium Latr. 1934. Shoemaker, Proc. Biol. Soc. Wash., vol. xlvii, pp. 23 sqq. Corophium acherusicum Costa. 1916. Barnard, loc. cot., p. 272. 1917. Stebbing, Ann. Durban Mus., vol. i, p. 448. 1928. Schellenberg, loc. cit., p. 672. 1931. Id., Swed. Antarct. Exp., vol. u, p. 259, footnote. 1934. Shoemaker, loc. cit., p. 24. Schellenberg identifies the European acherusicum with the earlier cylindricum Say, but Shoemaker does not accept this on the ground that Say’s species is not recognisable with certainty. Corophium triaenonyx Stebb. 1904. Stebbing, Spolia Zeylanica, vol. u, p. 25, pl. vi, A. 1921. Chilton, Mem. Ind. Mus., vol. v, p. 555. 1935. Barnard, Rec. Ind. Mus., vol. xxxvi, p. 305. Locality—Keurbooms River, Plettenberg Bay, in water-logged tree-trunks and drift wood, in the estuary and as far up as the river is tidal (Whisky Creek) (K. H. B., Jan. 1931, 3, ovig. 29, and juv.); Knysna River (K. H. B., Nov. 1938, 3g, ovig. 99); estuary of Klip Drifts Fontein stream, Potteberg, Bredasdorp district (A. C. Harrison, May 1936), Distribution.—Ceylon; Chilka Lake; Cochin and Travancore coasts. Contributions to the Crustacean Fauna of South Africa. 483 Cerapus abditus Templ. ite. Barnard, loc. cit., p. 271. 1938. Pirlot, Siboga Exp. monogr., vol. xxxiii f, p. 349, figs. 157, 158 (detailed figures). Fam. PODOCERIDAE. Gen. Podocerus Leach. 1937. Barnard, John Murray Exp. Rep., vol. iv, p. 174. (Key to species of the Indian region.) ot. Podocerus cristatus (Thomson). 1916. Barnard, loc. cit., p. 276. 1925. Schellenberg, loc. cit:, p. 188. 1926. Chilton, Trans. New Zeal. Inst., vol. Ivi, p. 513, fig. 2. Locality.—Swakopmund (Schellenberg). Podocerus inconspicuus (Stebb.). 1888. Stebbing, Challenger Rep., vol. xxix, p. 1194, pl. cxxxi. fein Das Tierreich, vol. xxi, p. 702. 1916. Barnard, loc. cit., p. 277, pl. xxviii, fig. 23 (palinurt). 1937. Id.,loc.cit., p. 175, fig. 18 (growth changesin gn. 2, 3) (palinurt). 1938. Pirlot, Siboga Exp. monogr., vol. xxxii f, p. 356, fig. 160. Locality.—Port Nolloth (Prof. Stephenson, 1938, littoral, under stones). Distribution.—South Arabian coast, Bay of Bengal, East Indies, Port Jackson. Podocerus africanus Brurd. 1916. Barnard, loc. cit., p. 278, pl. xxviii, figs. 24, 25. P25. 1d, loc. cit., p. 367. 1937. Id., loc. cit., p. 176, fig. 19 (urop. 1, and growth changes in on. 2, 3). Distribution.—South Arabian coast. Fam. VIBILITDAE. Vibilia chuni Behn. and Wolt. 1912. Behning and Woltereck, Zool. Anz., vol. xli, p. 8. 1913. Stewart, Ann. Mag. Nat. Hist. (8), vol. xii, p. 251, pl. vi (hodgsont). 484 Annals of the South African Museum. 1925. Behning, Deutsche Tiefsee Exp., vol. xix, p. 496, figs. 68-79. 1925. Barnard, loc. cit., p. 376 (hodgsont). 1925. Stephensen, Rep. Dana Exp., vol. u, D 5, p. 246 (2 chun). 1927. Behning, Deutsche Siidpol. Exp., vol. xix, p. 120. 1932. Barnard, loc. cit., p. 262 (hodgsoni, sp. dubia). Although Behning does not refer either in 1925 or 1927 to Miss Stewart’s species, I think Stephensen is correct in identifying it with chuni; as Behning says, the 7th peraeopod is very distinctive. Distribution.—EKquatorial and Southern Atlantic. Fam. PHRONIMIDAE. Phromma atlantica Guérin. 1901. Vosseler, Hyperiidae Plankton Exp., vol. ii, p. 21, pl. u, figs. 1-10. 1932. Barnard, loc. cit., p. 285. Locality.—Several 99 and 2 $¢ washed ashore at Durban, 9/7/32 (H. W. Bell-Marley). Distribution.—Cosmopolitan. Phronima colletia Bov. 1887. Bovallius, Bih. K. Sv. Vet. Ak. Handl., vol. xi, p. 25. 1887. Giles, J. Asiat. Soc. Bengal, vol. lvi, p. 215, pl. iu, figs. 1, 2 (bucephala). 1901. Vosseler, loc. cit., p. 32, pl. 11, figs. 8-10, pl. iv, figs. 1-3. 1932. Barnard, loc. cit., p. 286. Locality.—One 3 along with the previous species, Durban, 9/7/32 (H. W. Bell-Marley). Remarks.—Bovallius (1889, K. Sv. Vet. Ak. Handl., vol. xxu, p. 378) gives the date of Giles’ species as 1888, though on pp. 346 and 351 he gives it correctly as 1887. Bovallius’ paper was received at the British Museum (Nat. Hist.) library in October, whereas Giles’ paper was not published until November 2nd, so there is no question as to the priority of Bovallius’ name. Distribution.—Mediterranean, Atlantic 41° N. -35° S., Indo-Pacific. Phronimella elongata Claus. 1887. Giles, loc. cit., p. 217, pl. i, fig. 3 (heppocephala). 1888. Stebbing, Challenger Rep., vol. xxix, p. 1362, pl. clxu1. 1889. Bovallius, K. Sv. Vet. Ak. Handl., vol. xxi, p. 389, pl. xvi, figs. 51-67. Contributions to the Crustacean Fauna of South Africa. 485 1932. Barnard, loc. cit., p. 286. Locality One 3, 6 92 washed ashore at Durban, 9/7/32 (H. W. Bell-Marley). Distribution.—Mediterranean, Atlantic 43° N.-37° S., Indo-Pacific, Antarctic. Fam. OXYCEPHALIDAE. Oxycephalus clausi Bov. 1887. Bovallius, Bih. K. Sv. Vet. Ak. Handl., vol. xi, p. 35. 1890. Id., The Oxycephalids, p. 60, pl. i, figs. 19-24, pl. u1, fig. 1. 1923. Stebbing, Fish. Mar. Surv. Spec. Rep., 3, p. 11. 1932. Barnard, loc. cit., p. 294. Localities.—One 2 Algoa Bay (s.s. “Pieter Faure’); one 2 washed ashore at Hout Bay, Cape Peninsula, 1917; several gg and 922° washed ashore at Durban, 9/7/32 (H. W. Bell-Marley). Oxycephalus latirostris Claus. 1879. Claus, Gatt. Platysceliden, p. 193. 1887. Id., Die Platysceliden, p. 71, pl. xxiv, fig. 1. 1890. Bovallius, loc. cit., p. 66, pl. 11, figs. 7-12, and text-fig. 84. The rostrum is more acute than in Bovallius’ figure (pl. 11, fig. 7). The telson extends to the level of the apex of outer ramus of uropod 3, and almost to that of the inner ramus. Locality—One 2 washed ashore at Durban, 9/7/32 (H. W. Bell- Marley). Distribution.—Subtropical Atlantic and Indian Oceans. Glossocephalus milne-edwardsi Bov. 1887. Bovallius, loc. cit., p. 35. 1888. Giles, J. Asiat. Soc. Bengal, vol. lv, p. 250, pl. vi (Elsia indica). 1890. Bovallius, loc. cit., p. 106, pl. v, fig. 5, and text-figs. 6, 6a, 71. 1931. Barnard, Gt. Barrier Reef Exp., vol. iv, p. 131. The lower (hind) margin of 6th joint of peraeopods (gnathopods) 1 and 2 is smooth, as in Walker’s Ceylon specimens, and those from the Barrier Reef. Locality—One § washed up at Durban, 9/7/32 (H. W. Bell- Marley). Distribution.—Mediterranean, tropical Atlantic, Indian Ocean.° VOL. XXXII, PART 5. 34 486 1887. 1888. 1925. 1932. 1938. Annals of the South African Museum. Streetsia pronoides (Bov.). Bovallus, Bih. K. Sv. Vet. Ak. Handl., vol. xi, no. 16, p. 36. Stebbing, loc. cit., p. 1603, pl. cevii (challengeri). Stephensen, Dan. ocean. Exp., vol. ii, D 5, p. 194, fig. 75 (growth-changes and synonymy) (challengeri). Barnard, loc. cit., p. 295 (challengeri). Pirlot, Siboga Exp. monogr., vol. xxxiii f, p. 369 (synonymy). Locality.—One immature 2 washed up at Durban, 9/7/32 (H. W. Bell-Marley). Distribution.—Mediterranean, Atlantic, 41° N.-19° S., Indo-Pacific. Rhabdosoma whitei Bate. 1927. Spandl, Deutsch. Siidpol. Exp., vol. xix (zool. xi), p. 208, fig. 31, af (typo. error, withe?). 1932. Barnard, loc. cit., p. 296. 1933. Schellenberg, Zool. Anz., vol. ciii, p. 154, figs. 1-4 (brood- lamellae). Locality Many specimens washed ashore at Durban, 9/7/32 (H. W. Bell-Marley). Distribution.—Atlantic, Indo-Pacific. Fam. CAPRELLIDAE. Gen. Pseudaeginella Mayer. 1890. Mayer, F. u. FI., Golf, Naples, vol. xvu, p. 37. Pseudaeginella tristanensis (Stebb.). 1932. Barnard, Discovery Rep., vol. v, p. 300, fig. 166. Locality.—Kast London, littoral (Prof. T. A. Stephenson, July 1937, 3 3d, 3 99). Distribution.—Tristan d’Acunha. Remarks.—The largest § (6 mm.) is larger than any specimens yet recorded, but the hand of gnathopod 2 does not differ from my 1932 figure, except that there is a distinct tooth on the distal anterior margin of the 2nd joint. This tooth is much less distinct in the 9 (ovig. 2 3-5 mm.). This one difference, without more material from both localities, scarcely justifies specific separation. () BI) KEYS TO THE IDENTIFICATION OF SOUTH AFRICAN TANAIDACEA, ISOPODA (Marine anp FresHwatTerR), AND AMPHIPODA. NoTES ON THE Keys. In the family keys where a family contains more than one (South African) genus and/or species the page number is given on which will be found the key to the genera and/or species; and similarly in the case of genera containing more than one species. Where a family, or genus, contains only one genus, or species respectively, the student is referred to the list of species. | In the Isopoda the only important family not represented in South African waters is the Serolidae. Some minor families allied to the Eurydicidae and Corallanidae are not included in the key, although it is quite possible that eventually they will be found in the warmer waters of Natal and Portuguese Hast Africa. In the Gammaridea the two families Pleustidae and Sebidae are included [in square brackets], although not yet definitely recorded from South Africa. Certain other families of minor importance are not included, mainly for geographical reasons; for these families, see Stebbing, Das Tierreich, xxi, 1906. ' All the families of Phronimidea (Hyperudea), except a few aberrant ones like the Mimonectidae and Chuneolidae, are included, because many of these pelagic forms are likely to be found, sooner or later, within the South African region. References to such species not yet recorded from South African waters and to the relevant literature will be found in Spandl, 1927 (German South Polar Expedition reports), Barnard, 1930 (“Terra Nova’ reports), and Barnard, 1932 (“Discovery ”’ reports). In the Gammaridea and Cyamidea it is customary to reckon the two anterior pairs of legs as gnathopods, so that the seven pairs of legs comprise: gnathopods | and 2, peraeopods 1-5. In the Phronimidea (Hyperiidea), however, owing probably to the fact that the anterior legs are so frequently simple and in no way prehensile (subchelate or chelate), it is customary to speak of peraeopods 1-7. As the Amphipoda are perhaps the more difficult group to identify, a few explanatory figures are given to illustrate particular differential characters utilised in the keys. 488 Annals of the South African Museum. j A distinct carapace embracing the true head and first two peraeon segments, and overhanging laterally to form a branchial chamber. Maxilliped with an epipodial process lying in the branchial chamber. First pair of the 7 pairs of peraeopods chelate. Uropods slender and terminal . : : . Tanaidacea ( =Chelifera). No distinct carapace, the apparent “‘head’’ composed of the true head plus the Ist (rarely also the 2nd) peraeon segment coalesced. Maxilliped usually without epipodial process; the | latter when present not in a branchial chamber . ; 2. Dorso-ventrally depressed (in Phreatoicidae the pleon is jateneler compressed). Pleopods, when present, lamellar, branchial in function (heart in the pleon) . : : . : . Isopoda. Laterally compressed (except the Cyamidae, whale-lice). Branchiae attached to bases of some of the peraeopods (heart in the peraeon). Pleopods, when fully developed, divided into two sets, the first 3 pairs natatory, the last 2 pairs similar to the uropods, with unsegmented rami : i : ; . Amphipoda. Inst of South African Tanaidacea (=Chelifera). Tanaidae. — manais philetaerus Stebb. », annectens Brnrd. » (Anatanais) gracilis Heller (syn.: spongicola Brnrd.). Paratanais euelpis Brnrd. Leptochelia savignyt (Kroyer) (syn.: dubia Kroyer). Haplocope oculatus Stebb. Agathotanais ingolfi Hansen. Heterotanais (?) capensis Vanh6ffen. Apseudidae. 4 pseudes grossimanus N. and 8. agulhensis Brnrd. avicularia Brunrd. ee austro-africana Brurd. (=australis Brnrd., non Haswell). Trichapseudes tridens Brnrd. Sphyrapus malleolus N. and 8. 99 29 Key to the families of Tanaidacea (after Richardson, Bull. U.S. Nat. Mus., No. 54, 1905). Body scarcely attenuated behind. First antenna with only one flagellum. Mandible without palp. First maxilla with only a single lobe and a 1-jointed palp; second maxilla quite rudimentary. Epipod of maxilliped narrow, falciform . , : ; é ; . Tanaidae, p. 489. Body narrow, elongate. First antenna with 2 flagella. Mandible with 3-jointed palp. First maxilla with 2 lobes and a 2-jointed palp; second maxilla well developed and setose. Epipod of maxilliped large, laminar. Second peraeopod with large, broad, flat hand . : ; 5 : : ; : . Apseudidae, p. 489. Contributions to the Crustacean Fauna of South Africa. Key to the genera of Tanaidae. Three pairs of pleopods. Uropods uniramous. Brood pouch formed of only one pair of plates, arising from 1 bases of 5th peraeopods. Eyes present : : Five pairs of pleopods. Brood pouch normal. (In Agatho- tanais pleopods absent and brood pouch sani Uropods biramous . 2; Uropods rudimentary. ens 2 radimentary yes absent Rami of uropod cabal: beth 2- jointed. First neraoopud similar in both sexes , : Outer ramus of uropod very small, inner ramus with Saeral joints. First peraeopod dissimilar in ¢ and 2 Pleopods biramous | Pleopods uniramous (@) . : : Peraeopod 1 ¢ slender, elongate, chelate, 6th ‘gin with fOr: [ wardly produced “‘thumb”’ ‘ ae 1 robust, imperfectly chelate, 6th jin with back: wardly directed process : : J : Key to the species of Tanais. Six pleon segments (Anatanais). Uropod 5-6-jointed. 6th 1 joint of maxilliped narrow : : : Five pleon segments (T'anais s.s.). Uropod 4-jointed f Carapace broader than long. 6th joint of maxilliped ovate \ Carapace longer than broad Key to the genera of Apseudidae. Six free peraeon segments. Antenna 2 with scale at end of 1 2nd joint : ‘ ‘ ‘ Five free peraeon segments. Antenna 2 without scale Neither the small mandibular palp nor the maxilliped fringed with plumose setae. Five (normally) pairs of 2 pleopods Both the large Sanabria pe ae the See | fea \ with plumose setae. Three pairs of pleopods Key to the species of Apseudes. Rostrum tridentate i. simple if Rostrum with acute apex and sinuous, entire margins. | Rostrum with straight, serrate margins : : { Pleon segments 1-5 laterally with spiniform processes Pleon segments laterally obtuse é : Telson apically obtuse 4~ Telson apically acutely minduceds 6th leon oumen with 2 acute tubercles 489 Tanais, p. 489. 2. 3. Agathotanais. Paratanais. 4, 5. Haplocope. Leptochelia. Heterotanais. gracilis. 2. annectens. philetaerus. 2. Sphyrapus. Apseudes, p. 489. Trichapseudes. grossimanus. 2. 3. deltoides. austro-africana. 4, agulhensis. avicularia. 490 Annals of the South African Museum. List of South African Marine and Freshwater Isopoda. (For Terrestrial Isopoda see Barnard, 1932, Ann. S. Afr. Mus., xxx.) Flabellifera or Cymothoidea. Gnathiidae. Gnathia africana Brnrd. » spongicola Brnrd. bi . var. minor Brnrd. » disjuncta Brnrd. » eryptopais Brnrd. » aureola Stebb. (juv. =Praniza form). Anthuridae. Exanthura macrura Brnrd. 8 filformis (Lucas). Haliophasma tricarinata Brnrd. be coronicauda Brnrd. Be hermani Brnrd. Les foveolata Brnrd. Aes pseudocarinata Brnrd. Malacanthura linguicauda (Brurd.). Anthelura remipes Brnrd. Cyathura carinata (Kroyer) (syn. estuarius Brnrd.). Apanthura sandalensis Stebb. (syn. dubia Brnrd.). 56 africana Brnrd. Panathura serricauda Brnrd. Mesanthura catenula (Stimpson). Leptanthura laevigata (Stimpson) (syn. fawret Brnrd.). Paranthura punctata (Stimpson). Pseudanthura lateralis Richardson. Eurydicidae (Cirolanidae auct.). EHurydice longicornis (Studer). Excirolana natalensis (Vanhoffen). Pontogeloides latipes Brnrd. Gnatholana mandibularis Brnrd. Cirolana hirtipes M. Edw. a virilis Brnrd. x5 natalensis Brnrd. se sulcata Hansen. Pe parva Hansen. 45 cranchii Leach (syn. vicina Brnrd. and ? Hurydice latistylis Vanhoffen, non Dana). ne meinerti Brnrd. ss undulata Brnrd. a venusticauda Stebb. ”» 59 var. simplex Brnrd. aS littoralis Brurd. $9 fluviatilis Stebb. Contributions to the Crustacean Fauna of South Africa. 491 Cirolana theleceps Brnrd. a rugicauda Heller. ae cingulata Brnrd. eS palifrons Brurd. sn luciae Brnrd. ee bovina Brnrd. as incisicauda Brurd. Conilorpheus scutifrons Stebb. Corallanidae. Corallana africana Brnrd. Lanocira gardinert Stebb. (syn. capensis Brnrd.). Bathynomidae. Parabathynomus natalensis Brnrd. (1924, Fish. Mar. Biol. Surv. Union S. Afr. Rep., 4, p. 2, figs.). Aegidae. Aega webbii (Guérin). » semicarinata Miers (syn. urotoma Brnrd.). », monophthalma Johnston. 5, gracilipes Hansen. », monilis Brnrd. » antillensis Sch. and Mein. Rocinela dumerilii (Lucas). a orientalis Sch. and Mein. - granulosa Brnrd. Cymothoidae. Nerocila orbignyi (Guér. Mén.). is serra Sch. and Mein. 3 trichiura (Miers). » phaeopleura Blkr. Anilocra capensis Leach. ae leptosoma Blkr. Codonophilus imbricata (Fabr.). Cteatessa retusa Sch. and Mein. Cymothoa borbonica Sch. and Mein. Irona melanosticta Sch. and Mein. Cinusa tetrodontis Sch. and Mein. Livoneca raynaudu M. Edw. Limnoriidae. Limnoria lignorum (Rathke). Sphaeromidae. Sphaeroma terebrans Bate. c walkeri Stebb. ~ annandalei Stebb. Exosphaeroma hylecoetes Brnrd. _ planum Brnrd. a brevitelson Brnrd. Be varicolor Brnrd. 492 Annals of the South African Museum. Exosphaeroma laeviusculum (Heller). = truncatitelson Brnrd. be antikraussi Brnrd. = kraussii Tattersall. - pallidum Brnrd. ae porrectum Brnrd. Pseudosphaeroma barnardi Monod. [Isocladus tristensis (Leach). Tristan d’Acunha.] Zuzara furcifer Brnrd. Parisocladus stimpsoni (Heller). - perforatus (M. Edw.). Sphaeramene polytylotos Brnrd. Dynoides serratisinus Brnrd. Paracilicaea mossambicus Brurd. Cilicaea latreillei Leach. Cymodoce tuberculosa Stebb. var. tripartita Rich. Sy falcata Brnrd. Pe uncinata Stebb. Pe unguiculata Brnrd. = umbonata Brurd. AS amplifrons Stebb. se acanthiger Brnrd. - africana Brnrd. 7 valida (Stebb.). Bs comans Brnrd. 5 setulosa (Stebb.). os cryptodoma Brnrd. = tetrathele Brnrd. 5 japonica Rich. var. natalensis Brnrd. es cavicola Brnrd. - excavans Brnrd. Dynamenella dioxus Brnrd. BD scabricula (Heller). af bicolor Brnrd. ee australis Richardson. 2 australoides Brnrd. 5 macrocephala (Krauss). a taurus Brnrd. % huttoni (Thomson). as ovalis Brnrd. By navicula Brnrd. Cymodocella sublevis Brnrd. be pustulata Brnrd. ses cancellata Brnrd. 35 algoensis (Stebb.) (species dubia, juv.). Cassidias africana Brnrd. Parasphaeroma prominens Stebb. Artopoles natalis Brnrd. Stathmos coronatus Brnrd. Contributions to the Crustacean Fauna of South Africa. Idoteidae Idoteoidea or Valvifera. Idotea metallica Bose. » mdica M. Edw. Euidotea peronii (M. Edw.). Paridotea ungulata (Pallas). bs reticulata Brnrd. » rubra Brnrd. “a fucicola Brnrd. Engidotea lobata (Miers). Synidotea hirtipes (M. Edw.). ae setifer Brnrd. ba variegata Clige. Glyptidotea lichtensteinii (Krauss). Cleantis natalensis Brnrd. Pseudidoteidae. Holidotea unicornvis Brnrd. Astacillidae. Astacilla mediterranea Koehler. os bacillus Brnrd. 493 Arciurella corniger (Stebb.) (syn. hirsuta Brnrd. and Astacilla setosa Vanhoffen). » =} var. subglaber Brnrd. - lineata (Stebb.). e pustulata Brnrd. 5 longipes Brnrd. - brevipes Brnrd. a lobulata Brnrd. Arcturina hexagonalis Brnrd. Antarcturus kladophoros Stebb. $s similis Brnrd. Neoarcturus oudops Brnrd. Pleuroprion chuni (zur Strassen). Idarcturus platysoma Brurd. Asellota or Aselloidea. Stenetriidae. Jaeridae. Stenetrium crassimanus Brnrd. x diazi Brnrd. ee bartholomei Brnrd. a . dagama Brnrd. a dalmeida Brnrd. wi saldanha Brnrd. s3 syzygus Brnrd. Jaera pusilla Brnrd. Austrofilius serrata (Brnrd.)(? syn. A. furcatus Vanh6ffen, non Hodgson). 494 Annals of the South African Museum. Jaeropsis curvicornis (Nicolet). Janira capensis Brurd. » angusta Brnrd. » exstans Brard. Laniropsis palpalis Brnrd. Iais pubescens (Dana). Haploniscus dimeroceras Brnrd. Pseudojanira stenetrioides Brnrd. Protojanira prenticei Brnrd. (freshwater). Munnidae. Paramunna laevifrons Stebb. os concavifrons Brurd. os capensis Vanhéften. Kuphomunna rostrata Brnrd. Antias uncinatus Vanhéffen. Munnopsidae. Munnopsurus mimus Brurd. Pseudomunnopsis beddardi (Tattersall). Eurycope sulcifrons Brnrd. Ss quadrata Brnrd. fusiformis Brnrd. Ilyarachna affinis Brnrd. sR crassiceps Brnrd. Desmosomidae. Eugerda sp. Ilychthonos capensis Brunrd. Macrostylis spiniceps Brnrd. Rhabdomesus bacillopsis Brnrd. Phreatoicidea. Phreatoicidae. Phreatoicus capensis Brnurd. a z var. abbreviatus Brnrd. ae zs var. depressus Brnrd. $5 “a var. penicillatus Brnrd. Bopyroidea or Epicaridea. Bopyridae. Scyracepon levis Brnrd. On Scyramathia. Pseudione munidae ee On Galatheiden. = crenulata Sars. Ss sp. On Upogebia africana. Palaegyge plesionikae Brnrd. On prawns. Bathygyge grandis Hansen On Glyphocrangon. Paragigantione papiliosa Brnrd. On Galatheidea. Hemiarthrus nematocarcini Stebb. On prawns. Epipenaeon japonicum Thielemann. On prawns. Contributions to the Crustacean Fauna of South Africa. 495 Dayjidae. Zonophryxus quinquedens Brnrd. On prawns. Cryptoniscidae. Cyproniscus crossophori Stebb. On Ostracoda. Clypeoniscus stenetrit Brnrd. On Isopoda. Aegoniscus gigas Brnrd. On Isopoda. Inriopsis sp. (Brady, 1914, Ann. Durban Mus.,i). On parasitic Cirripedes on hermit-crabs. Incertae sedis. Microniscus ornatus Vanh6offen (a free-swimming larval stage, adult ?). Gen. et sp. incert. (Barnard, 1920, fig.). On Isopoda. Key to the suborders of Isopoda. (Modified after G. O. Sars, Crustacea Norway, ii, 1896.) 1 { Uropods lateral : ‘ : : : Soe Uropods terminal or eetecitial : : 3. Uropods forming together with the last pleon Comer harans a caudal fan ; : : ; : : : . Cymothoidea or Flabellifera, p. 495. : Uropods valve-like, ae dd ventrally and arching over the pleopods . ; : : : : : , . Idoteoidea or Valvifera, p. 506. i Pleopods modified for air-breathing. Terrestrial, inland and littoral, forms . : : : . Oniscoidea. Pleopods not modified for air- Hehe Marine and fresh- water forms : 4, Pleopods usually covered by a thin Speedie pine (ie moaited Ist pair of pleopods) . : : : : : . Aselloidea or Asellota, p. 509. Pleopods never covered by an opercular plate : : aes? Pleon laterally compressed. Free-living, fresh-waterforms . Phreatoicidea. Pleon depressed. Parasitic forms, with free-swimming larval stages, marine . ‘ : : : : : . Bopyroidea or Epicaridea, p. 513. a For descriptive account, with keys, of the Oniscoidea (Woodlice) see Barnard, 1932, Ann. S. Afr. Mus., xxx; additional species, Ann. Mata) Wus., vill, p. 155, 1937. Key to the families of Cymothoidea (flabellifera) (after Richardson, 1905, Bull. U.S. Nat. Mus., No. 54). Adult with 5 free peraeon segments and only 5 pairs (2nd—6th) of normal peraeopods; peraeopod | modi- fied, peraeopod 7 wanting. Juvenile stage uae on fishes. é . Gnathidae, p. 496. Adult with 7 free peraeon eriaclite wail are 7 pairs of peraeopods : : : ‘ 3 : sf Bt 496 Annals of the South African Museum. Body narrow, cylindrical. Uropod with outer ramus arching dorsally over the telson. (Pseudanthura has rudimentary outer ramus, but the body-form throughout the family is distinctive) . d . Anthuridae, p. 497. [eas normal, depressed. Uropod normal, outer ramus lateral 3. { Pleon usually composed of 6 segments. 4, \Pleon composed of 2-4 segments. 9. Uropod with both rami well developed, sige lamellar 5. 4+ Uropod with outer ramus rudimentary, more or less ungui- form. Boring in piles and floating timber . . Limnoriidae. Palp of maxilliped free, margins of last 2 joints more or less setose, never armed with hooks . 6. eiapalp of maxilliped embracing the cone formed by ‘ie mouth-parts; apex armed with hooks, never setose. 8. No plumose branchial tufts on bases of pleopods. Eyes (when present) dorsal or lateral. : : 2 as Plumose branchial tufts on bases of pleopods. Eyes con- fined to the ventral surface . : : ; . Bathynomidae. 6 Mandible with distal half stout and conspicuous, molar 7 prominent . ; ‘ . Lurydicidae, p. 498. Mandible with distal half narrow, more or ts eoanearea | by the upper and lower lips, molar much reduced . Corallanidae, p. 500. Body symmetrical. Both antennae with well-defined peduncles and flagella. Pleopods setose. Rami of uropod large, more or less foliaceous . : . Aegidae, p. 500. 8‘ Body often distorted. Both antennae reduced, without clear distinction between peduncle and flagellum. Pleopods not setose. Rami of uropod long or short, but always narrow. : : : ' . Cymothoidae, p. 501. Pleon of 2 segments. Uropod with inner ramus 9: immovable : ‘ ‘ ‘ : : . Sphaeromidae, p. 501. | Pleon of 4 segments. Both rami of uropod movable . [Serolidae.] For an account of the Serolidae, a family confined to the Southern Hemisphere (with one exception), but which has not yet been found in South African waters, see Sheppard, 1933, “Discovery”’ Reports, vil. Key to the species of Gnathia. Frontal margin semicircularly produced forwards and downwards between bases of mandibles. 5th 1 (4th free) peraeon segment not constricted medio- dorsally . : 3 ‘ : : : . cryptopais. Frontal margin transverse . : : 5 : oN 2h Contributions to the Crustacean Fauna of South Africa. 497 Fifth (4th free) peraeon segment in § not divided medio- % dorsally. Anterior margin of head in 2 not notched africana. Z Fifth peraeon segment in ¢ divided medio-dorsally. Anterior margin of head in 2 notched . ; suinebe Mandibles in g denticulate on the straight inner margin . spongicola and 3 var. minor. Mandibles with entire, convex inner margin. : . disjuncta, G. aureola is founded on the juvenile or ‘‘ Praniza”’ form. Key to the genera of Anthuridae (see Barnard, J. Linn. Soc. London, xxxvi, 1925). {Mouth-parts normal (lower lip with rounded lobes, mandible with dentate cutting-edge, and maxilla 1 with conspicuous spine-teeth). Statocysts paired, but sometimes absent. : ee : Mouth-parts modified (lower lip with acute ‘Tones mandibles apically acute, maxilla | slender, lancet- like). Sometimes a a statocyst, but usually none : oy ko: 9 { Fifth joint of Biter ial 47 not hidacemane 6th aren: Fifth joint of peraeopods 4—7 underriding 6th . ele { Maxilliped 4-jointed * 4 : : ; : se 4. \Maxilliped 5-jointed : ; se { Pleon segments distinct . : : : : . LExanthura, p. 497. \ Pleon segments fused t : d 5 : . Haliophasma, p. 498. { Eyes present . : : : ; ; : 7 ae |Eyes absent . : ; : ; : . Anthelura. [ Unguis of peraeopod 1 hie ¢ : ; : . Haliophasma, p. 498. \ Unguis of peraeopod 1 long : ; : : . Malacanthura. { Maxilliped 4-jointed 5 ; : ; ; . Cyathura. \ Maxilliped 5-6-jointed : : : ; se Pleon segments distinct . ; z é . sO. | Pleon segments fused : : : : ; . Mesanthura. Maxilliped 5-jointed : : ; : : . Apanthura, p. 498. | Maxilliped 6-jointed : . : . . Panathura. 10 fine: ramus of uropod rudimentary . ‘ : . Pseudanthura. Outer ramus of uropod well developed é : Lu, (With statocyst. 5th joint of peraeopods 4—7 ange! riding 6th 3 : : F 5 : . Leptanthura. No statocyst. 5th joint of peraeopods 4—7 not under- riding 6th : : : 3 ; : . Paranthura. Key to the species of Hxanthura. Telson widening distally, truncate. : : . macrura. ice ovate, with a median longitudinal keel _. . filiformis. * The basal joint, which is ankylosed to the head, is included. 498 Annals of the South African Museum. Key to the species of Haliophasma. f Telson tricarinate " : ‘ ; ‘ ‘ , “ee { Telson not tricarinate me : ‘ ‘ ; 2 iee 9 Integument not pitted : : : ‘ : : . tricarinata. Integument pitted . : : ‘ : : ; . foveolata. ( Telson obovate, apically upturned, with median keel . . hermani. 3, Telson ovate, dorsal surface wholly convex, with raised central | portion like a flat hat, more or less surrounded byarim . 4. Telson distinct from 6th pleon segment : : : . coronicauda. Telson fused with 6th pleon segment . : : : . pseudocarinata. Key to the species of Apanthura. With eyes. Outer ramus of uropod apically indented . . sandalensis. Eyes absent. Outer ramus of uropod not indented : . africana. Key to the genera of Hurydicidae. Fifth pleon segment with free lateral margins . 2. Lateral margins of 5th pleon segment overlapped by shoe of the 4th segment. Inner apex of peduncle of uropod pro- duced. Peduncle of antenna 2—5-jointed ; : 2, ae rest of antenna at right angles to it. Mandibular palp 3-jointed. Inner plate of maxilliped without coupling- hooks. Pleopods without accessory lobes. Inner apex of peduncle of uropod not produced . : : . LHurydice. First joint of peduncle of antenna | not at right angles ie rest of antenna. Inner plate of maxilliped with coupling- First joint of peduncle of antenna 1 extended straight in front, | hooks. Pleopods with accessory lobes. Inner apex of ‘ peduncle of uropod produced 5 ; 2 : = ios 3 Mandibular palp 3-jointed . : : é : A . LHxcirolana. Mandibular palp 2-jointed . : , é 5 ; . Pontogeloides. Mandibles very stout, projecting forwards. Frontal lamina not distinct from median process of head. ; : . Gnatholana. (iviemnaattes not prominent. Frontal lamina distinct 5 ~ NOE Antenna 2 distinctly longer than antenna 1. Frontal lamina not projecting prominently in front. Inner ramus of pleopod 1 broad. In ¢ stylet on pleopod 2 attached near | base . : : s . Cirolana, p. 499. Both antennae anoee sea Hot atid ieee than arate Frontal lamina projecting prominently in front. Inner ramus of pleopod 1 narrow. In ¢ stylet on pleopod 2 attached far from base é ; : : ; . Conilorpheus. Contributions to the Crustacean Fauna of South Key to the species of Cirolana. Pleon segment 3 laterally overlapping both 4th and 5th 1 segments . Pleon segment 3 not Pe cre ing 4th, bee 4th seeelanpiee sth Anterior margin of frontal lamina not free or produced Anterior margin of frontal lamina free, somewhat produced, but not in the shape of a prominent horn or knob. 2nd joint of peraeopods 6 and 7 without long plumose setae . Frontal lamina very narrow, linear. 2nd joints of peraeopods 6 and 7 with long plumose setae . : Frontal lamina broad, pentagonal or scneeaereL: 2nd joints of peraeopods 6 and 7 without long plumose setae fa plate 4 with postero-inferior angle quadrate . fa plate 4 with postero-inferior angle rounded . — line on hind margin of eyes continuous across the 4 head. In ¢ stylet on pleopod 2 slender, straight . Impressed line on head not continuous. In ¢ stylet on pleopod 2 robust and strongly curved . Telson with median longitudinal furrow 6. Telson dorsally smooth Telson with 2 submedian basal eperies Ries Gotan Telsonic furrow strong. Apices of telson and uropods pointed fe furrow shallow : J Apices of telson and uropods Rinee” | Apices of telson and uropods rounded, with notches Telsonic apex broadly rounded, with 8 spines ; Telsonic apex narrowly rounded or subacute, with 13- 4 spines Telsonic apex wakinly rounded, Hie 7-8 spines; 2 setae on dorsal surface i@eontal lamina subquadrate, as broads as, OF broudes ie lene Frontal lamina linguiform, rounded in front and slightly wider than at base : : : Frontal lamina smooth. Melo with eurmdets median longitudinal keel : : Frontal lamina with transverse ridge near ious apex at a lower level and meeting rostral point of head Pleon and telson unsculptured | 7 9 Z ere of Pleon segments with transverse series of iemibles: Telson with median series of denticles and several others basally and laterally : : ‘ ‘ . : Pleon as in venusticauda. Telson with a single median denticle near base Telson apically incised 2 : : Head and peraeon with transverse grooves A impressed lines of punctae : 134 Head smooth. Hinder peraeon Conye wath dente or rugae Head, peraeon and bison smooth Africa. 499 rugicauda. vise 3. 10. 6. 5. natalensis. hirtipes. virilis. Tie ue bovina. sulcata. 8. palifrons. theleceps. parva. cranchit. meinerti. 11. 13. undulata. 12. littoralis. venusticauda. venusticauda var. semplex. incisicauda. congulata. fluviatilis. luciae. 500 Annals of the South African Museum. Key to the genera of Corallanidae. Bases of the antennae, at least the Ist joint, visible from above. 2nd joint of maxilliped much longer than broad. Corallana. Bases of both antennae hidden by rostral projection of head. 2nd joint of maxilliped very little longer than broad . 2. Maxilla 1 strongly faleate : ; : : : . Lanocira. ~|Maxilla 1 unguiform, moderate . ; ; ‘ : . [Argathona}. Argathona is included owing to the likeness of Hurydice latistylis Vanh6ffen, 1914 (non Dana), to a species of this genus. Key to the genera of Aegidae. Head with median point wholly or partly separating bases of Ist antennae. Frontallamina usually large. Maxilliped i] not less than 6-jointed ‘ : ' . Aega, p. 500. |x more or less covering bases of Ist Be anaes Frontal lamina small.. Maxilliped not more than 4-jointed os per Eyes present. Pleon not much narrower than peraeon . Rocinela, p. 500. ee absent. Pleon abruptly narrower than peraeon . . Syscenus. Key to the species of Aega. (First 2 joints of peduncle of antenna 1 much dilated, flattened; 2nd joint apically as Frontal lamina flat or concave . 2. First 2 joints of podanele of bien I eylindeae wi | dilated; 2nd joint not apically produced. Frontal lamina convex or compressed and elevated . : es ae of Ist antennae partly separated by rostral point Sup ce ae of lst antennae completely separated by rostral point monophthalma. Inner apex of 6th joint of peraeopods 2 and 3 with a linguiform 2 process. webbit. Inner apex of 6th joitit of Pecacnnde 2 en 3 aoe ee % process . : : : . : : : Sle All Telson truncate , ; 3 ; ; ; : . semicarinata. *\"Telson pointed. detente ‘ : ‘ : : . antillensis. Peraeopods very slender. Frontal lamina forming a trans- verse projecting plate. Side-plates acutely produced 5 posteriorly : ‘ : : i ‘ : . gracilipes. Peraeopods not very slender. Frontal lamina forming a rounded knob. Side-plates not acutely produced. - moniilis, Key to the species of Rocinela. | Front of head with prominently produced subtriangular process over bases of antennae . : : : . dumerilii. ere of head not strongly produced . ; ; 2. Flagellum of antenna 2 14-16-jointed. Outer ramus of uropod apically rounded . : . orventalis. Flagellum of antenna 2 10—12-jointed. SOnter ramus of uropod lanceolate, apically acute : : - . granulosa. Contributions to the Crustacean Fauna of South Africa. 501 Key to the genera of Cymothoidae. Head not immersed in Ist peraeon segment. Anterior margin | of latter trisinuate. Body symmetrical : ‘ eae : Head immersed in Ist peraeon segment. Anterior margin of latter not trisinuate. Body more or less distorted Bese oJ Head broadly rounded in front . F : F : -. Nerocila, p. 501. eiiicad distinctly narrowed in front Z : : F . Anilocra, p.501. 3) Bases of lst antennae contiguous F : : . 4, * | Birst antennae not dilated, their bases eraraiel ‘ 5. Antenna 1 much dilated. Ungues of peraeopods ations, | stout. Inferior margin of side-plates 6 and 7 concave . Codonophilus. t Antenna 1 compressed. Ungues long, strongly curved. Inferior margin of side-plates 6 and 7 convex : . Cteatessa. f Pleon manifestly distinct from peraeon : ; ; EEG: Pleon and peraeon continuous . Pate : aes Bases of Ist antennae widely senekated ‘ ‘ . . Cymothoa. Bases of Ist antennae narrowly separated . : , . Cinusa. Pleon very little immersed in peraeon : : : . ILrivoneca. ros deeply immersed in peraeon . : : : « Lrona. Key to the species of Nerocila. Inner ramus of uropod terete, slender, acute; outer ramus elongate . : , : 2. Inner ramus of uropod Teondox. margins ie a oath on inner apex; outer ramus not much longer thaninner. 3. Outer ramus of uropod very long. Side-plates rounded behind; segments rounded postero-laterally . : trichiura. Outer ramus of uropod elongate. Side-plates and deitaayts acute behind . : 2 5 : : ; . phaeopleura. Postero-lateral angles of all peraeon segments acutely pro- | duced. Outer margin of inner ramus of uropod serrate . serra. 3 Postero-lateral angles of only the 4 posterior segments more | or less acutely produced. Outer margin of inner ramus of uropod smooth . E . sie : . orbignyi. Key to the species of Anilocra. Antenna 1 geniculate, the flagellum distinct from peduncle owing to enlargement of 3rd joint of latter . : . leptosoma. Antenna | undifferentiated : : ‘ ; é . capensis. Key to the groups of Sphaeromidae. Pleopods 4 and 5 with transverse pleats on the inner rami only, the outer rami thin and membranous . Hemibranchiatae, p. 502. Pleopods 4 and 5 with pleats on both rami ‘ . Eubranchiatae, p. 505. Pleopods 4 and 5 with both rami membranous . . Platybranchiatae, p. 506. VOL. XXXII, PART 5. 35 502 Annals of the South African Museum. Key to the genera of Hemibranchiatae. Telsonic apex in 9 without a notch, rounded or some- what produced, more or less.acute; in ¢ generally as in 2, but sometimes much produced in the shape of a median process narrowed at its base. Mouth- parts in adult 2 not modified . : 2. Telsonic apex in both sexes with a notch, ee stronger in ¢, frequently divided by a median lobe. 1 Mouth-parts in adult 2 modified.* Maxilliped with 4th-6th joints lobed. Outer ramus of pleopod 3 2-jointed. Marsupialplates overlapping in middle line . ‘ 6. Telsonic apex entire in Q, in 3 San a aiedinn slit, frequently widening anteriorly to a foramen. Mouth-parts in 2 not modified. Maxilliped with 4th-6th joints lobed : : ‘ ofc edeSe ‘ Maxilliped with 4th—6th joints not lobed. Peraeopods 1-3 slender, with long natatory setae. Outer ramus of pleopod 3 l-jointed. Outer margin of outer ramus of uropod distinctly serrate. Mar- supial plates overlapping in middle line. . Sphaeroma, p. 503. Maxilliped with 4th—6th joints lobed. Peraeopods 1-3 without long setae. Outer ramus of pleopod 3 2-jointed. Outer margin of outer ramus of uropod not or only indistinctly serrate. Marsupial plates not reaching middle line Li 2 Peraeon segment 7 unarmed in both sexes 34 Peraeon segment 7 with a process in § (sometimes also in 9) s : 5 : : ; : - Oe Outer ramus of uropod well developed, subequal to 4 menaere : . Exosphaeroma, p. 503. Outer ramus ot stoped mie smallen chara inner . Pseudosphaeroma. Telsonic apex similar in both sexes, pecrne! produced, with a ventral groove : : . Isocladus. Telsonic apex somewhat produced in 9, ateonaly pro- duced in § with process narrowed at base, scarcely any ventral groove . 5 : : : . Zuzara. ( Both rami of uropod well developed. Pleon segment 4 without median process. ‘ : : . Cymodoce, p. 504. Inner ramus of uropod very short . : 2 5 es Pleon segment 4 with median process in 3 : . Cilicaea. Pleon segment 4 without process. ‘ . . Paracilicaea. * Mandibles lose their dark, strongly chitinised apices, other appendages bluntened and non-setose. See Hansen, Q.J. Microsc. Sci., n.s., No. 193, vol. 49, pt. 1, p. 82, 1905. Contributions to the Crustacean Fauna of South Africa. 503 Outer ramus of 3rd pleopod 2-jointed. Telsonic slit widening toaforamening. dg appendages on 7th segment separate . 2: Outer ramus of 3rd pleopod \. aainteds Melanie slit not widening to aforamen. J appendages on 7th segment fused basally : : : ; . Dynoides. Peraeon segment 7 in § with, in 2 with or without, a median process : : : . Parisocladus, p. 505. Peraeon segment 7 unarmed in vba sexes - . Sphaeramene. Key to the species of Sphaeroma. ae triangular, apex pointed. Epistome flat . : ae. spoon-shaped, apex broadly rounded. Epistome concave — on peraeon circular. On telson the 2 submedian series 1 of tubercles diverge slightly and continue nearly to apex Tubercles on peraeon transversely elongate. On telson 2 pairs of submedian tubercles, followed by a single median one Key to the.species of Hxosphaeroma. - Nocona minutely setulose. Outer ramus of uropod fringed with setae, especially on inner margin “| including uropods glabrous. : Epistome somewhat like a shovel, its anterior margin peek pro- jecting. Telsonic apex acute or narrowly pointed Epistome without freely projecting anterior margin, its apex curving over to meet rostral point . 2 er smooth, evenly convex. Inner ramus of open narrow- ing to a pointed apex 34 Telson with a pair of low (often snateeRnet) slges seoheasiy wa a median keel distally. Inner ramus of aes more or less truncate apically . : : Inner ramus of uropod with outer acer. corner soundea! oatee ramus serrulate on outer margin distally. Inner ramus of uropod with outer distal corner sharply anand: outer ramus with apex uncinately curved outwards . Body very depressed. Uropods very large, lamellate. Epi- stome with a concavity ‘ : : : piety moderately depressed. Uropods romeeiate. Epistome not concave . - : aedson considerably broader ea long, in ey not extending ie apices of uropods . Telson only slightly, if at all, pidadee fia ieee or sage ite broad, extending at least to level of apices of uropods Inner ramus of uropod quadrangular (parallel-sided). Telson longer than broad, narrowing to a very acute apex . Inner ramus of uropod ovate-lanceolate 8 ae apically broadly rounded-truncate ee apically narrowly rounded or subacute terebrans. 2. walkeri. annandalet. hylecoetes. va pallidum. 4, kraussit. antikrausst. planum. 6. brevitelson. porrectum. 8. truncatitelson. 9. = See a i ha + ee 504 Annals of the South African Museum. Telson extending beyond apices of inner rami of uropods. varicolor. Telson not extending beyond apices of inner rami of uropods; in g with 2 minute submedian granules and 10-12 other eranules in an arc submarginally Key to the species of Cymodoce. ($d only. to their proper species.) First peduncular joint of antenna 1 denticulate on front margin First peduncular joint of antenna 1 not denticulate Median lobe of telsonic apex ¢ dorsally with either a re- 2 curved hook or a round button-like knob Median lobe of telsonic apex without either hook or knob . fTelsonic apex with a knob Telsonic apex with a hook Pleon segment 4 without tubercles Pleon segment 4 with 2 submedian tubercles ‘Hook very long. Tubercles on pleon stronger than shot on telson. Hind margin of pleon segment + es 5 strongly setose Hook short. Tubercles on foleen aiionner nine (hore on pleon Head strongly imed ape not inflated Both rami of uropod ee rs acute oe segment 4 with submedian processes or tubercles Rami of uropod not tapering to acute points; or if one of them is acute (lanceolate) then nears segment 4 without \ tubercles : he segment 4 with 2 icone caline dan processes "pleae Ws Pleon segment 4 with 2 tubercles -| Body glabrous. Peraeon without tubercles. + blue or black spots . | Body more or less pilose or setulose . . Outer ramus of uropod not distinctly alter fen inner Often with 10, ramus : ‘ ‘ Lone: ramus of uropod ranch amailles feat inner ramus Anterior peraeon segments glabrous. uropod subacute. lobe subacute . 114 Whole body thickly setose. acute. Outer ramus of Telsonic apex deeply trifid, median Outer ramus of uropod apically Telsonic apex shallowly trifid, median lobe broad, truncate 12 Uropods not extending beyond telsonic apex Uropods extending beyond telsonic apex . laeviusculum. It is often impossible to assign isolated 9° or juveniles tuberculata var. tripartita. 7). 3. 6. umbonata. 4, unguiculata. Be falcata. uncinata. amplifrons. i 9. acanthiger. africana. valida. 10. ie 12: comans. japonica var. natalensis. 13. 14. Contributions to the Crustacean Fauna of South Africa. 505 ‘Pleon segment 4 without tubercles or bosses. Median and lateral lobes of telsonic apex acute, the lateral ones 13 curved outwards : é : : : : . cavicola. Pleon segment with 2 submedian tubercles. Lobes of telsonic apex blunt. j setulosa. (Telson with 2 submedian bosses followed bis 2 wiaicdl tubercles . : tetrathele. 14) Telson with 2 submedian hides or léneiendacelly elenmaned bosses é : : : . eryptodoma. \Telson with 2 Se nmediai eta iabebeles : 5 : . excavans. Key to the species of Parisocladus. short. Telsonic slit in g widening very little anteriorly . stimpsoni. Process long, apically entire in J; 2 without process. Telsonic slit in § widening to a circular or transverse foramen . perforatus. Key to the genera of Hubranchiatae. Peraeon. usually without processes. Telsonic apex with either a notch, or a narrow slit widening anteriorly, with or without a median lobe ‘ ‘ Zz Peraeon without processes. Telson with the ee margins bent downwards and inwards to form a tube Cymodocella, p. 506. aS 3S on peraeon segment 7 in 4 short, apically bifid; in 9 very 3 | Uropods normal . : 2 : : : ‘ . Dynamenella, p. 505. 2, Inner ramus of uropod in § rudimentary, outer ramus very | long . é : : ‘ : : : . Cassidias. Key to the species of Dynamenella. {Peraeon segment 7 in g with 2 strong submedian processes. ; Telsonic slit widening anteriorly, with a median lobe . dioxus. Whole peraeon without processes . ; oe i Telsonic apex with a slit widening xeuaeeiy es a foramen, without medianlobe. One or more of the peraeon segments tuberculate . ; : : = 3 Telsonic apex with a Phiek not adenine aikeui ely: or apex truncate. Peraeon segments not tuberculate . : 6. Mouth-parts elongate. Telsonic slit widening to a cette transverse foramen. 2 submedian longitudinal ridges on middle of telson, with a shorter pair between them basally. Inner ramus of uropod oblong, apically rounded. Peraeon with tubercles on all segments (more or less distinct) . scabricula. Mouth-parts not elongate. Telsonic slit not widening to a foramen, with a tubercle just in front of the slit : «: Ae Inner ramus of uropod oblong, apically broadly rounded. | Peraeon with tubercles on all segments . : bicolor. 4) Inner ramus of uropod sinuous, narrowly apically. Peenend | with tubercles only on segment 7, either distinct or very obscure : : : : : : Cu 506 Annals of the South African Museum. Tubercles on peraeon segment 7 distinct. A prominent median tubercle immediately in front of apical notch . australis. ae very obscure. No median tubercle inmedinitelyeie in front of apical notch australoides. | Telsonic apex with a notch . he (eae apex truncate. Body boat- shaped Head dinoueae inflated , ‘ ; . navicula. a strongly inflated. Relsna aah 2 ieee Only base of 2nd joint of peraeopods dark : ‘ . taurus. Head rather inflated. Telson with 2 low Paberctes: Whole of 2nd joint of peraeopods dark . _macrocephala. rss notinflated. Telson quitesmooth. 2nd re ofp peraeo- pods not dark : 8. { Body convex, nearly japallelended huttont. 8 ; Body depressed, oval . : ovalis. Key to the species of Cymodocella. | Peraeon without tubercles or ridges ‘ : a OR Peraeon with tubercles and ridges giving a Sneaate appear- ‘| ance. Telson with 2 large submedian processes . cancellata. Telson and pleon segment 4 both with 2 obscure submedian | tubercles : : : : : : . sublevis. Telson with 6 tubercles. Peraeon segment 7 posteriorly bilobed, at least in 3 pustulata. Key to the genera of Platybranchiatae. Eyes lateral. Head not sunk in Ist peraeon segment. Uropods biramous Parasphaeroma. Basal joints of Head sunk in Body vaulted. Basal joints of antenna 1 not expanded. Body depressed, more or less éilipsical: antenna 1 expanded. Eyes dorsal. Ist peraeon segment . : : 2. Seventh peraeon segment excluded fate the lateval margin. Seventh peraeon segment forming part of the lateral margin. Outer ramus of uropod well developed Outer ramus of uropod apparently obsolete Artopoles. Stathmos. Key to the families of Idoteoidea (Valvifera). ‘Ramus of uropod small. Pleopod | often modified in ¢. A single penial process . : 2. 1 Ramus of uropod large. Pleopod | never riptiifead in in Usually a pair of penial processes. Body depressed. Peraeopods more or less prehensile, the anterior 4 pairs often subchelate Body narrow, cylindrical, rarely se caba: Aesiee | 4 pairs of peraeopods directed forwards, not pre- hensile, slender, with long setae [Boss broad, depressed. Anterior 4 pairs of paracpods normal, prehensile, not setose 25 Idoteidae, p. 507. Astacillidae, p. 508. Pseudidoteidae. Contributions to the Crustacean Fauna of South Africa. Key to the genera of Idoteidae. | Eyes dorsal. Maxilliped 7-jointed. Pleon with 2 com- plete sutures and one incomplete | Eyes lateral Side-plates distinct. 2, Side-plates not distinct. \ Maxilliped 5-jointed 3 Pes irvca 6-jointed Maxilliped 7-jointed : { Pleon with one incomplete and 2 enters sitet: \Pleon with 3 incomplete sutures . ; Pleon with one complete and 2 incomplete sutures. 5. Pleon completely fused, with 3 incomplete sutures. | with low medio-dorsal longitudinal keel . { Side-plates contiguous in both sexes \ Side-plates widely separated in 3. A pair of penial processes A single penial process. Body Key to the species of Sunes Side-plates on segments 2 and 3 rounded. Telson with postero-lateral angles rounded. Peraeopods strongly setose. Uropod with 2 oblique ridges on peduncle Side-plates on segments 2 and 3 angular. Telson with | postero-lateral angles acute. Peraeopods not strongly ‘ setose. Uropod without ridges on peduncle 2 Peraeopods strongly setose. Ne with 2 oblique ridges on peduncle Peraeopods not strongly setose. Uabped aioe abique ridges on peduncle Key to the species of Idotea. Colour blue. Pelagic Rhlaen apically truncate. Colour variable : Telson apically emarginate. Key to the species of Paridotea. Body very narrow and elongate. Ramus of uropod longer 2 than broad . erty not verynarrow. Ramus of uropod a Tonge then broad . ‘ Posterior margins of sterna straight or convex. Pedusisle of uropod without longitudinal keel. Antenna 2 with 3rd—5th joints not produced. Lateral margin of Q¢ telson thin, not grooved : : : Posterior margins of sterna concave. Peduncle of uropod with longitudinal keel. Antenna 2 with inner apex of 3rd—5th joints produced. Lateral margin of telson thick, grooved . : Ramus of uropod distinctly ie dudes Shin ee Bastend: 3 lateral angles of telson subacute : Ramus of uropod very little broader than long. Passer lateral angles of telson acute Cleantis. 2. Bu Synidotea, p. 507. 4, 5. Idotea, p. 507. Euidotea. 6. Glyptidotea. Paridotea, p. 507. Engidotea. setifer. hirtipes. variegata. metallica. indica. fucicola. 2. reticulata. rubra. ungulata. 508 Annals of the South African Museum. Key to the genera of Astacillidae. r Body cylindrical or only slightly depressed Body strongly depressed, not bent. Peraeon segment 4 not elongate. 3 pairs of marsupial plates. Side- plates distinct. All pleon segments and telson com- pletely fused ‘ : A ‘ : ie pairs of marsupial plates. Peraeon segment 4 in J, and to a less extent in 9, elongate. Body bent be- tween segments 4 and 5. Side-plates distinct . Three pairs of marsupial plates Peraeon segments in ¢ without any veel processes. Segment 4 in 2 without lateral wing-like expansions. Pleopod 1 in 3 modified "| Peraeon segment 4 in 9 laterally expabdee Bikoped lin 6 not modified Peraeon segment 4in both sexes, but more in ¢ than in 9, 4 elongate. Body bent between segments 4 and 5 Peraeon segment 4 not elongate. Body not bent . | Peraeopod 1 stout, patent. Peraeopods 2-4 slender. \ Peraeon segment 4 often very elongate in § Peraeopods 1-4 short and stout, peraeopod 1 latent, con- cealed in a buccal chamber formed by downward pro- | jections of lateral margins of head and peraeon seg- ment 1. Peraeon segment 4 not so very longing . Side-plates distinct. Pleon with 3 segments in front of telson. Body not spinose Plsideentee not distinct. Pleon with one een in n fone of telson. Body spinose Idarcturus. Antarcturus, p. 508. Arcturella, p. 508. 5. 6. Astacilla, p. 509. Arcturina. Neoarcturus. Pleuroprion. Key to the species of Antarcturus. Peraeopods 2—7 serrate or spinose. Body granulate in J, spinose in 2. A pair of long ue on head, knobbed in g, spinulose in 2 : Peraeopods smooth. Head and peraeon ceeicnts each | with 2 submedian tubercles forming 2 longitudinal ridges down the body Key to the species of Arcturella. much less. Body in both sexes subcylindrical. Flagellum 2nd antenna 2-3-jointed Width of segment 4 2 greater than length, considezable less than lengthin §. Body in both sexes depressed. \ Flagellum 2nd antenna 1-jointed : | Width of peraeon segment 4 9 less than length, in 3 very kladophoros. similis. ee First pair of pleopods in 3 counied with the oid: formas a large operculum; peduncles of Ist pair elongate, the rami often obscure. 2nd pair large. 3rd pair not forming an operculum in either sex. Ist pair in Q large fused, opercular : Posterior 3 peraeopods not natatory. Byes vsually 9 present : Posterior 3 peraeopods eihise acto or Sede Eyes absent * But see Stenetrium syzygus. Contributions to the Crustacean Fauna of South Africa. 509 A small ventral process on 3rd segment ¢. Body 92 9 normally hirsute and strongly tuberculate corniger. No ventral process. i, Q glabrous and feably | tuberculate . ‘ . lineata. Antero-lateral angles Ist segment 9 not anmeatee Lateral margin 4th segment ° straight or nearly so 4. Antero-lateral angles Ist segment distinct, acute. Lateral margin 4th segment lobulate (only 2 known) . lobulata. ra Segment 4 tuberculate (only 2 known) . pustulata. *| Segment 4 not tuberculate . : : ee (Outer margin 2nd joint antenna 2 entire. Peraeopod 5 9 2nd joint longer than all others together. 4th segment g with median hooked tubercle near hind 5 margin longipes. Outer margin 2nd ‘aint siesta 2 watohed: petseapad 5 2nd joint shorter than all others together. Segment 4 $ with 2 tubercles in middle line. brevipes. Key to the species of Astacilla. Smooth, glabrous, very slender in §. Segment 4 in Qa little less, in $a little more than half the total length, smooth in both sexes. Amongst the spines on _Echinoids : ‘ : : : : . bacillus. ° granulate. Head spinose. Segment 4 much less than half total length, with a large median tubercle in front, and a smaller one behind mediterranea. Key to the families of Aselloidea (Asellota). (Marine species. If the specimen is freshwater see Protojanira under Jaeridae). First pair of pleopods in § not coupled with the 2nd pair, their peduncles short and fused, rami distinct. 2nd pair small. 3rd pair in both sexes forming an oper- culum. Ist pair in 9 small, fused, forming a small operculum * : : . Stenetriidae, p. 510. ee ee a a a 510 Annals of the South African Museum. Peraeon segments 5-7 not marked off from the rest, nor | smaller. Eyes (when present) dorsal or dorso-lateral, but not pedunculate. Bases of Ist antennae not 3 widely separate . : \Peraeon segments 5-7 sharply ea off from, be | smaller than, the rest. Eyes (when present) on lateral pedunculate processes. Bases of lst antennae widely separate . ‘ Posterior 3 peraeopods slender, nomorinies with long | spines, but none of the joints enlarged . Posterior 3 peraeopods with enlarged and flattened joints bearing long plumeose setae . Key to the species of Stenetrium. 1 Eyes well developed and pigmented Eyes reduced to 3-4 feebly pigmented ocelli Outer apex of Ist joint of antenna 2 produced into an acute 9 process : ‘ : : ae apex of Ist caine of Siena acute ae not produces or dentiform . Finger of peraeopod 1 3 itewaer. ena Vonea eal 3 present on all peraeon segments Finger of peraeopod | ¢ robust, not tapering. Voutenl bart | absent on segments 4 and 5. ‘Rostrum longer than broad. 6th joint of Hote ie li in Ee as broad as long, in 9 scarcely widening distally; 2nd joint 45 in ¢ without basal tooth : : : A Rostrum broader than long. 6th joint of peraeopod 1 in 9, \ widening distally | Second joint of peraeopod 1 in 3 aah bel tooth 6th joint longer than wide. Ventral keel on all segments 5° Second joint of peraeopod | in ¢ without basal tooth; 6th iain | as broad as long. Ventral keel present only on segments 3,4and7. Rostrum quadrate, heahaats than rays "6th aaa of pareeoned lin g as broad as long F ; Rostrum triangular, longer than ee an 6th joint of peraeopod | in g longer than broad Key to the genera of Jaerdae. Freshwater. First pleopods in ¢ abnormal for the family, consisting of 2 foliaceous rami arising from a short 1 basal piece, and overlapping the 2nd yoke. (not | coupled with them to form an operculum) ; Marine. First pleopods normal . . Jaeridae, p. 510. . Munnidae, p. 511. . Desmosomidae, p. 512. . Munnopsidae, p. 512. 2. SYZYGUS. 3. 4, bartholomei. CrTASSUMANUS. diaz. dalmeida. dagama. saldanha. Protojanira. a ————eE a. a a ee ee a Contributions to the Crustacean Fauna of South Africa. 511 multiarticular. Mandible with prominent molar. Uropods with rami : : Antenna 2 short, peduncle alilateas Aeolian saditientaiyy Mandible with molar obsolete. Uropods minute, \ nodular, rami obsolete . a Eyes distinct *) Byes absent Antenna 1 very ara: iplettores, 2-3- sanaiea: lueetiods short . : ... 1 well deve loged Uikopaus malt degelagad No rostral process. Peraeopod 1 not subchelate A rostral process. Peraeopod 1 subchelate gf Third joint of antenna 2 without scale (spine) Third joint of antenna 2 with a scale | Maxilliped with 4th and 5th joints very mah idahetod: | Antenna 2 long, peduncular joints not dilated, flagellum < Peduncles and rami of pleopod J in § fused and dilated Maxilliped with none of the joints much enlarged. Peduncles and rami of pleopod 1 in § usually not fused or dilated Apex of mandible in ¢ not printed. Hiya eaéh stirerai ocelli. Peraeopod 1 usually more or less modified for Fi 8 prehensile purposes 3 Apex of mandible’in § much roland. Each eye con- sisting of only 2 lenses. All peraeopods alike . Key to the species of Janira. Third joint of antenna 2 with scale. Maxilliped with 3rd— 5th joints broad. Anterior margin of head not strongly produced : : 2 : ; 1) Third joint of antenna 2 without scale. Maxilliped with all joints narrow. Anterior margin of head strongly produced. 3rd joint of peraeopod 1 in § with trans- verse rugae . : : 5 ; : : Body not very narrow. 5th joint of peraeopod 1 in 3 narrow, more than twice as long as broad Body very narrow. 5th joint of peraeopod 1 in ¢ oval, not twice as long as broad . Key to the genera of Munnidae. l es minute. Mandible with small palp . Uropods well developed Mandible with palp. Head not Sepa ina ated in 3S. Ist peraeon segment in $ not gibbous L Mandible without palp. Headin grostrate. Ist peraeon segment in § gibbous : , Jaeropsis. 4, Haploniscus. 5. te 6. Pseudojanira. Jaera. Austrofilius. Laniropsis. Janira, p. 511. Tais. exstans. capensis. angusta. Paramunna, p. 512. De Antias. Kuphomunna. 512 Annals of the South African Museum. Key to the species of Paramunna. Lateral margin of pleon serrulate . . : Be . capensis. Lateral margin of pleon entire |. : . i os Be Front of head evenly convex ‘ : : . laevifrons. Front of head quadrately produced : ‘ ‘ . concavifrons. Key to the genera of Desmosomidae. Pleon consisting of one segment. Body moderately elon- gate. Peraeon segment 1 not laterally produced . 2. 1< Pleon consisting of 2 segments. Body very elongate. Peraeon segment | laterally ee ii forwards, seg- \ ments 4 and 5 very long : : : . Rhabdomesus. Uropod biramous. Mandible with nein ; : . Hugerda. Uropod uniramous : : ‘ . oe Mandible with a feeble pabbie Weoped hoki : , . ILlychthonos. Mandible without palp. Uropodlong . : : . Macrostylis. Key to the genera of uae f Antenna 2 with scale on 3rd joint ; oy ee . Munnopsurus. | Antenna 2 without scale on 3rd joint . : x oan Head produced between bases of Ist antennae, =o are , widely separated . : 3. Head truncate in front, bases of 1st Selena aloes iectne: Ilyarachna, p. 512. Posterior part of body abruptly narrower than the anterior part. Uropoduniramous . : : . Pseudomunnopsis. Posterior part of body not narrower she anterior part. Uropod biramous . : ‘ 5 ; : . Hurycope, p. 512. Key to the species of Ilyarachna. Posterior margin of peraeon segment 6 nearly straight, segment 6 medio-dorsally longer than 7 . ‘ . afinis. Posterior margin of peraeon segment 6 concave, segment 7 medio-dorsally longer than 6 : f ‘ . crassiceps. Key to the species of Hurycope. Head produced between bases of Ist antennae ? oy ae fae concave in front, not produced . : fusiformis. Head very shortly produced. Antero-lateral wneiea of peraeon segments 5-7 and of pleon acute; segment 7 longer than either 5 or 6 : ‘ . quadrata. Head strongly produced. Antero-lateral sees hin peraeon segments 5—7 and of pleon rounded; segment 6 longer than either 5 or7. : : : : : . sulcifrons. eS ee Contributions to the Crustacean Fauna of South Africa. 518 Key to the families of Bopyroidea (Epicaridea). Q with true peraeopods and some or all of the other appen- dages. g differing from the last larval stage of 9 . Q asimple sac, without peraeopodsand with most or all ape the other appendages absent. 4 resembling the last larval stage ofthe 9. Oralcone without disc. Basal joint of antenna 1 expanded, usually with spiniform teeth. Side-plates (coxal-plates) pectinate (except Iiriopsis). Peraeopods | and 2 shorter and thicker than the others, the former prehensile, the latter slender, -with setiform dactyli. Outer ramus of uropod shorter than inner. Parasitic on various other Crustacea . ‘ , : : F . Cryptoniscidae, p.514. Q distinctly segmented, more or less asymmetrical. 7 pairs of peraeopods, all except 1st sometimes obsolete on one side. Pleopods usually present. 4 with 7 distinct peraeon segments. Last larval stage with oral cone simple, basal joint of antenna 1 entire, flagellum of antenna 2 4-jointed, peraeopods all alike, and inner ramus of uropod shorter than outer. Parasitic on Decapod Crustacea . : : . Bopyridae, p. 513 2< 2 symmetrical, segmentation (if present) visible only on the dorsal side. 5 pairs of peraeopods. Pleopods rudimentary or absent. 4 with 6 free peraeon seg- ments (the Ist segment fused with head). Last larval stage with oral cone ending in a sucking disc, basal joint of antenna 1 with spiniform process, flagellum of antenna 2 5-jointed, peraeopod | shorter and thicker than the others, and rami of uropod sub- equal. Parasitic on Decapods and Schizopods . Dajidae. Key to the genera of Bopyridae. Q asymmetrical, but not greatly swollen on one side. 7 peraeopods on both sides. Uropods present . ee Q greatly swollen on one side. Only lst peraeopod present on the swollen side. Uropods absent. 5 | pleon segments. 4 with pleon segments completely fused ‘ : i : E : : . Hemiarthrus. a! Uropods of Q uniramous . : : : : seioe a iieopads of Qbiramous . : oe | Pleurae and outer rami of weoaeds of 2 Bioneatel digitate. | 6 with medio-ventral boss on all peraeon segments; and pleon segments partly distinct : é . Scyracepon. Pleurae and pleopods of 2 not elongate or digitate. ¢ without medio-ventral bosses, and with all pleon { segments distinct : : : é ‘ . 4. | 4 ae of 2 tuberculate or warty . : : . Pseudione, p. 514. Pleopods of 2 smooth ; ‘ ‘ ‘ : . Palaegyge. 514 Annals of the South African Museum. Epimera of 2 not enlarged. Pleon segments in ¢ distinct . Paragigantione. 5, Epimera of 2 considerably or greatly enlarged, and separate. | Pleon segments in g fused . : P : ee Lateral parts of pleon segments in ? not Heveloped : . Bathygyge. 64 Lateral parts of pleon segments in 2 well developed, but | only 5 segments visible ‘ ; P , ‘ . Hpipenaeon. Key to the species of Pseudione. { Epimera of 2 rounded, pleura acute . . ‘ i . crenulata. \Epimera of 2 acute, pleura rounded. : : 5 . munidae. Key to the genera of Cryptoniscidae. Body of ¢ consisting of 2 sacs connected by a narrow neck, without attachment cord. dg with eyes, and epimera and basal joint of antenna 1 entire. Parasitic on rhizo- 1 ot ae cephalous Cirripedes on Hermit-crabs . : Liriopsis. Body of @ consisting of a single sac. 4 without eyes, ua epimera and basal joint of antenna 1 pectinate : Di First larval stage without a ventral plate. Parasitic on Ostracods . 4 Cyproniscus. Qnotattached. Ventral sae with a slit Jocea be wale like flaps. First larval stage with a ventral plate. Parasitic on Isopods. ; 3. Body slightly lobulate Teteralles nanone distin: seomontatem Clypeoniscus. Body distinctly lobulate, the segmentation extending nearly from mid-dorsal line to mid-ventral line. ¢ and larval stages unknown . é : ‘ : é F . Aegoniscus. | Q attached to host by a cord. Ventral surface without slit. \ 3 List of South African Amphipoda. Gammaridea. Lysianassidae. Trischizostoma remipes Stebb. “5 paucispinosum Brnrd. ss serratum Brnrd. Stomacontion capense Brnrd. Acidostoma obesum (Bate). Phoxostoma algoense Brnrd. Paravalettia chelata Brnrd. Euonyx biscayensis Chevr. Amaryllis macrophthalma Hasw. Bathyamaryllis conocephala (Brnrd.). Cyphocaris richardi Chevr. ee anonyx Boeck. 5 challengeri Stebb. ae faurei Brnrd. Lysianassa variegata (Stimps.). Ps ceratina (Wlkr.) (syn. cubensis, non Stebb.). | a ee ee ee ee ee Se Se _ Contributions to the Crustacean Fauna of South Africa. 515 Aristias symbiotica Brnrd. Hippomedon longimanus (Stebb.). Microlysias xenoceras Stebb. Ichnopus taurus Costa. (syn. macrobetomma Stebb.). Socarnopsis crenulata Chevr. Orchomenella plicata Schell. Uristes natalensis Brurd. » tnduratus Brnrd. Cheirimedon pectenipalma Brnrd. Tryphosa onconotus Stebb. Chironesimus adversicola (Brurd.). oe rotundatus (Brnrd.). Eurythenes gryllus (Licht.). Stegocephalidae. Stegocephaloides australis Brnrd. a attungens Brnrd. Parandania boecki (Stebb.). Ampeliscidae. Ampelisca brevicornis (Costa). a anomala Sars. a chiltoni Stebb. (? =eschrichtii Kréyer). “ fusca Stebb. Bs diadema (Costa). ss miops Brnrd. 2 natalensis Brnrd. a palmata Brnrd. = excavata Brnrd. 5 spinimanus Chevr. se byblisoides Brnrd. Bypblis anisuropus Stebb. >» gaimardi (Kroyer). Triodos insignis Brnrd. Haustoriidae. Urothoé pulchella (Costa). Platyischnopus capensis Brurda. Phoxocephalidae. Harpinia excavata Chevr. Pontharpinia stimpsoni Stebb. (? syn. villosa juv. Schell.). Amphilochidae. Gitanopsis pusilla Brnrd. Hoplopleon australis (Brnrd.). Le medusarum Brnrd. Cyproidea ornata (Hasw.) (syn. Gallea crinita Spandl.) a = eee x est 516 Annals of the South African Museum. Leucothoidae. Leucothoé spinicarpa (Abildg.). a richiardii Less. es dolichoceras Brnrd. se ctenochir Brnrd. Metopidae. Proboloides rotundus (Stebb.). Stenothoidae. Stenothoé adhaerens Stebb. (only 2 known). Za gallensis Wlkr. x dolichopous Brnrd. Be assimilis Chevr. Phliantidae. Plioplateia triqueter Brnrd. Temnophlias capensis Brurd. Palinnotus natalensis Brnrd. Colomastigidae. Colomastix pusilla Grube. Ochlesidae. Ochlesis lenticulosus Brnrd. Pardaliscidae. Nicippe tumida Bruz. Halice anacantha Brnrd. Acanthonotozomatidae. Iphimedia capicola Brnrd. Panoploea excisa Brnrd. Inljeborguidae. Liljeborgia dubia (Hasw.). a consanguinea Stebb. Fs proxima Chevr. ES kinahani Bate var. capensis Brnrd. 2 epistomata Brnrd. Oedicerotidae. Halicreion (?) ovalitelson Brunrd. Bathymedon palpalis Brnrd. Oediceroides cinderella Stebb. zs plumicornis Brnrd. Aceroides limicola Brnrd. Perioculodes longimanus (B. and W.). Synchelidium (? tenuamanus Norm.). —————— ss -— Contributions to the Crustacean Fauna of South Africa. 517 Tironidae. Tiron australis Stebb. Bruzelia diodon Brnrd. Syrrhoites tenellus Brnrd. Austrosyrrhoé crassipes Brurd. Calliopiidae. Calliopiella michaelseni Schell. Paramphithoidae. Epimeria cornigera (Fabr.). = semiarmata Brnrd. 5 longispinosa Brurd. Lepechinellidae. Lepechinella chrysotheras Stebb. Atylidae. Nototropis homochir (Hasw.). 5 granulosus (Wlkr.). Eusiridae. ; Eusiroides monoculodes (Hasw.). ELusirus minutus Sars. Cleonardopsis carinata Brnrd. Rhachotropis palporum Stebb. 5 kerguelent Stebb. 3 grimaldii Chevr. by paeneglaber Brnrd. = anomala Brnrd. ’ Pontogenevidae. Paramoera capensis (Dana) (? syn. schizurus Stebb.). ed bidentata Brnrd. Gammaridae. Megaluropus agilis Hoek (syn. Phylluropus capensis Prnrd.). Melita fresneli (Aud.). ,, subchelata Schell. » zeylanica Stebb. >> orgasmos Brnrd. Maera inaequipes (Costa). >» mastersii (Hasw.). » hamigera (Hasw.). >» hirondellei Chevr. ss grossimanus (Mont.). > oOruzelit Stebb. 1888. » vagans Brnrd. (syn. bruzeliz Stebb. 1910, non Stebb. 1888. levis Brnrd., non S. I. Smith). » subcarinaia (Hasw.). » boeckii (Hasw.). VOL, XXxil, PART 5. 36 518 Annals of the South African Museum. Ceradocus rubromaculatus (Stmpsn.). . aviceps Brnrd. Elasmopus pectenicrus (Bate) (syn. brasiliensis Stebb. non Dana). 5 japonicus Steph. Parelasmopus suluensis (Dana). Elasmopoiaes chevreuxi Stebb. Hriopisella capensis (Brnrd.). Eucrangonyx robertsi Meth.* Gammarus (Paramelita) capensis Brurd. (syn. P. ctenodactyla Schell.). es ae nigroculus Brnrd. and var. persetosus Brnrd. Ee 2 crassicornis Brnrd. AB A auricularius Brnrd. a Ag tulbaghensis Brnrd. _ x seticornis Brnrd. 5 ye kogelensis Brnrd. of Me aurantius Brnrd. = un spinicornis Brnrd. f me granulicornis Brnrd. Dexaminidae. Polycheria atolli Wlkr. Guernea laevis Chevr. Talitridae. Talitroides eastwoodae (Meth.) (?=T. africana Bate). formae cylindripes, setosa, calva, macronyx. 99 99 Talorchestia capensis (Dana). ss quadrispinosa Brnrd. a australis Brnrd. zs ancheidos Brnrd. BS (?) africanus Bate (species dubia). Neobule reynaudi (M. Edw.) (species dubia). Parhyale inyacka (Brnrd.). Chiltonia capensis Brnrd. Parorchestia rectipalma Brnrd. 53 dassenensis Brnrd. Hyale saldanha Chilton. », hirtipalma (Dana). », macrodactylus Stebb. » grandicornis Kroyer. », maroubrae Stebb. », diastoma Brnrd. Allorchestes inquirendus Brnrd. Parhyalella natalensis (Stebb.). Orchestia excavata Chevr. (Upper Zambezi.) - Pes 2 platensis Kroyer. Tristan d’Acunha. | * Retained in Crangonyx by Schellenberg, Zool. Jahrb. Abt. Syst., xix, p. 482, 1937. Contributions to the Crustacean Fauna of South Africa. 519 Aoridae. Aora typica Kroyer. Le » torma anomala Schell. ws » gibbula Brnrd. Lembos hirsutipes Stebb. » leptocheirus Wlkr. >» hypacanthus Brurd. Lemboides afer Stebb. fs acanthiger Brnrd. - crenatipalma Brurd. Grandidierella lignorum Brnrd. Photidae. Photus dolichommata Stebb. » longimanus Wlkr. , uncinata Brnrd. Cheiriphotis megacheles (Giles) (syn. walkeri Stebb., durbanensis ‘ Brnrd.) Eurystheus afer (Stebb.). 2 atlanticus (Stebb.). 53 holmesi Stebb. & imminens Brnrd. ae semidentatus Brnrd. Es palmoides Brnrd. am (?) scissimanus Brnrd. Chevalia aviculae Wlkr. Ampithoidae. Ampithoé ramondi (Aud.). us falsa Brnrd. Ps africana Brurd. Exampithoé natalensis Brnrd. Cymadusa australis (Brnrd.). Macropisthopous stebbingi Brnrd. Jassidae. Jassa falcata (Mont.). Ischyrocerus anguipes Kroyer. S carinatus Brnrd. a gorgoniae Brnrd. Isaeopis tenax Brnrd. Corophiidae. Camacho bathyplous Stebb. Ericthonius brasiliensis (Dana). Siphonoecetes orientalis Wlkr. x dellavalleit Stebb. Corophium acherusicum Costa. 5 triaenonyx Stebb. Cerapus abditus Templ. 520 Annals of the South African Museum. Cheluridae. Chelura terebrans Phil. Sebidae. Seba saundersit Stebb. (Doubtfully S. African.) Podoceridae. Laetmatophilus purus Stebb. R- tridens Brnrd. 7 durbanensis Brnrd. Podocerus cristatus (Thomson). a africanus Brnrd. - multispinis Brnrd. “a aA ae var. levis Brnrd. 55 brasiliensis (Dana). Bs inconspicuus (Stebb.) (syn. palinurt Brnrd.). Phronimidea (Hyperiudea). Lanceolidae. Scypholanceola vanhoeffeni Wolt. Vibiliidae. Vibilia armata Bov. (syn. gracilenta Bov.). ,5, Chuni Behn. and Wolt. Cystisomatidae. Cystisoma africanum Brnrd. Hyperiidae. Hyperia galba Mont. (syn. gaudichaudit). » promontorit Stebb. (syn. schizogeneios). Hyperoche cryptodactylus Stebb. Parathemisto (Euthemisio) gaudichaudii (Guer.). Phronimidae. Phronima sedentaria (Forsk.). Se atlantica Guer. Be collettti Bov. Phronimella elongata Claus. Phrosinidae. Phrosina semilunata Risso. Primno macropa Guer. Pronoidae. Parapronoé crustulum Claus (syn. Parapronoé clausi Stebb. and. Amphipronoé cuspidata Bate). Brachyscelidae. Brachyscelus rapax Claus. Contributions to the Crustacean Fauna of South Africa. 521 Oxycephalidae. Oxycephalus tuberculatus Bate. ea claust Bov. a latirostris Claus. Glossocephalus milne-edwardsi Bov. Streetsia pronoides (Bov.). Rhabdosoma whitei Bate. Calamorhynchus rigidus Stebb. Platyscelidae. Hemityphis tenuimanus Claus. Paratyphis maculatus Claus. 4 promontorit Stebb. Tetrathyrus forcipatus Claus. Cyamidea. Caprellidae. Caprella penantis Leach (syn. acutifrons Latr.). > » » var. porcellio Mayer. >» >> », var. natalensis Mayer. ied equilibra Say. = scaura 'Templ. an solitaria Stimpson (species dubia, ? =scaura). is cicur Mayer. » falsa Mayer. 3 laevipes Mayer. f triodous Stebb. - danilevskii Czern. Caprellina longicollis (Nicol.). ud spiniger Brnrd. Phtisica marina Slabber. Metaprotella makrodaktylos Stebb. Metaproto novaehollandiae (Hasw.). Paradeutella serrata Mayer. Orthoprotella mayeri Brnrd. Pseudaeginella tristanensis (Stebb.). Cyamidae. Cyamus ovalis R. de V. Paracyamus erraticus R. de V. sé gracilis R. de V. 38 boépis Liitken. Synopsis of the three sub-orders of Amphipoda. (After Stebbing. Das Tierreich, xxi, 1906.) 1. Normal Amphipoda, usually not pelagic, usually opaque and more or less pigmented. Head not fused with Ist peraeon segment. Palp of maxilliped 2—4-jointed (absent only in Ochlesis). Peraeon with 7 pairs of legs, 522 Annals of the South African Museum. Srd joint, a normal gnathopod simple 4 7) gnathopod subchelate . minutely chelate gnathopod chelate gnathopod 2 J WY entire emarginate notched cleft lower = 9 Le es inner lobes obsolete principal lobes notched inner lobes present neon = Caprellidae ciara oo peraeopod 1 SSeS reduced ~__ Phronimidea and joint of peraeopod 5 branchia Oust cuter orn Phronimidea peraeopod 7 rudimentary S Amphipoda. Figures (schematic) to illustrate some of the differential characters utilized in the Keys. Contributions to the Crustacean Fauna of South Africa. 523 the Ist joints of which form well developed side- plates. Pleon usually consisting of 7 free segments, carrying 3 pairs of pleopods and usually 3 pairs, at least 1 pair, of uropods; uropod 1 always biramous . Gammaridea, p. 523. 2. Pelagic Amphipoda, usually semitransparent or feebly pigmented. Head not fused with lst peraeon seg- ment. Palp of maxilliped absent. Peraeon with 7 pairs of legs, the Ist joints (side-plates) small or want- ing. Pleon usually consisting of 7 free segments, carrying 3 pairs of pleopods and 3 pairs of uropods; rami of the latter often evanescent. Eyes usually large . ’ : ; : : : ; . Phronimidea (Hyperiidea), p. 526. 3. Slender, elongate, cylindrical Amphipoda, modified for living among weeds (Skeleton-shrimps); or short and depressed and ectoparasitic on whales (Whale lice). Head fused with Ist peraeon segment. Palp of maxilliped 1-4-jointed. Peraeon often with less than 7 pairs of legs, the lst joints (side-plates) absent. Pleon and its appendages rudimentary. Eyes small Cyamidea (Caprellidea), p. 527. A fourth sub-order, Ingolfiellidea, is accepted (see Hansen, 1903, J. Linn. Soc. London, xxix, and Calman, 1909, in Lankester’s Treatise on Zoology, pt. 7); but no representatives of it have yet been found in South Africa. . Key to the families of Gammaridea. (Adapted from Stebbing. Das Teirreich, xxi, 1906. Gnathopods 1 and 2, and peraeopods 1-5 are reckoned.) Ant. 1 Ist joint stout,* accessory flagellum present; mandible with cutting-edge almost smooth,t with palp; gn. 2 with 3rd joint elongate ‘ ‘ : : Lysianassidae, p. 527. (* Except Amaryllis. + Except Paravalettia.) ‘These characters not combined : : 2. Body plump; ant. 1 with accessory flagellum; a (1) > mandible without molar and without palp Stegocephalidae, p. 529. | These characters not combined ‘ : seals Ss Head tapering, truncate; eyes, when present, simple, usually 4; ant. 1 without accessory flagellum; telson more or less cleft . . Ampeliscidae, p. 529. These characters not combined : : be ete 3 (2) 524 Annals of the South African Museum. Ant. 1 with accessory flagellum; mandible with 4 (3) palp normal; prps. 3-5 with joints ex- panded for burrowing 5. These characters not combined 6. 5 (4) { Prp. 4 not greatly longer than prp. 5 Haustoridae, p. 530. \ Prp. 4 greatly longer than prp. 5 j . Phoxocephalidae, p. 530. | Upper lip incised; mxp. normal; urop. 3 6 (4) < biramous; telson elongate and entire Amphilochidae, p. 530. These characters not combined a eae Ant. 1 without accessory flagellum; mxp. more 7 (6) or less abnormal; telson entire 8. | These characters not combined les 8 (7) ( Gnathopod 1 chelate Leucothoidae, p. 531. \ Gnathopod 1 not chelate y. 9 (8) { Uropod 3 biramous 10. \ Uropod 3 uniramous ne 10 (9) { Mxp. with palp Colomastigidae. Mxp. without palp Ochlesidae. 11 (9) { Mandible with palp Metopidae. Mandible without palp . 12. 12 (11) { Mxp. outer plate obsolete Stenothoidae, p. 531. ) Mxp. with both plates developed : ; Mandible with molar weak or wanting; telson Bh ((7)) | more or less divided : : These characters not combined Mxp. inner plate well developed 14 (13 ince inner plate small . 5 (1 Ae 1 and 2 simple Gnathopods 1 and 2 strongly BaDaReTaeee Eyes, when present, dorsally contiguous or con- fluent Eyes, when present, feral 4 Ant. 1 without accessory flagellum; ard ‘oint mandibular palp large; prp. 5 much longer than prp. 4; telson entire : Ant. 1 with accessory flagellum; 3rd ‘cin man- dibular palp small; prp. 5 not much longer than prp. 4; telson cleft (except in Bruzelia of the Tironidae) . Prps. 1 and 2, 4th and sth ine aninted 18 (17) (pelagic) ‘ Prps. 1 and 2, 4th and sth one Hot dilated Side-plate 4 usually excavate behind; prp. 1 and 2 not glandular; telson variable; animal usually not domicolous Side-plate 4 usually not excavate behind; prp. i animal and 2 glandular; telson entire; usually domicolous. Phliantidae, p. 531. 14. 16. Acanthonotozo- matidae, p. 531. 15. Pardaliscidae, p. 531. Liljeborgudae, p. 531. 17. USS Oedicerotidae, p. 532. 18. [Synopidae]. Tironidae, p. 532. 20. 29. Contributions to the Crustacean Fauna of South Africa. 525 Mandible with palp Mandible without palp . Telson variable, usually cleft; ant. 1 usually with accessory flagellum . 20(19){ ey Telson entire Telson cleft . Rostrum weak eae) ee well marked ‘ : ‘ Side-plates 1-4 rounded; ant. 1 longer than ant. 2 : ‘ ; , : Side-plates 1-4 angular; ant. 1 shorter than ant. 2 { Pleon segments 5 and 6 coalesced . | Pleon segments 5 and 6 not coalesced ae (24) 23 (22) 24 (21). Lower lip with inner lobes obsolete. Lower lip with well developed inner lobes { Urop. 3 not greatly elongate . \ Urop. 3 greatly elongate 27 a 26 (25) Gnathopods | and 2, hands eee Gnathopods 1 and 2, hands not powerful ie co 3, both rami well developed . op. 3, one ramus very small or wanting |Uropods 2 and 3 well developed ‘ 29 (19), Uropods 2 and 3, one or other wanting or rudi- mentary 30 (29 Grandidierella) , (29) Pleon usually depressed; urop. 3 uniramous (except Chelura) 31 (30) Uropod 3 not uncinate . Uropod 3 outer ramous uncinate pieaethopod 1 larger than gnathopod 2 Br enathopod 1 not larger than gnathopod 2 i compressed; urop. 3 biramous (except 33 (31) f Lower lip with principal lobes notched | Lower lip with principal lobes not notched Ant. 2 with flagellum not spatulate; urop. 3 uniramous, terete (or inner ramus very minute) Ant. 2 with spatulate dagelinsie urop. 3 inner ramus minute, outer ramus foliaceous \ (wood-boring) | Both gnathopods chelate; gn. 2 with 3rd hae rather long : 35 (34), Gnathopods not chelate (gn. 9 sorotns com- plexly subchelate, 5th joint prolonged); gn. 2 with 3rd joint short : 34 (30) yall 28. Gammaridae, p. 533. 22. 24. Calliopiidae. 23. [ Pleustidae]. Paramphithoidae, p. 532. 25. 27. Atylidae, p. 533. 26. Lepechinellidae. [Melphidippidae]. Eusiridae, p. 533. Pontogenevidae, p. 5338. Dexaminidae, p. 535. Talitridae, p. 535. 30. Podoceridae, p. 539. 31. 34. 32. 33. Aoridae, p. 537. Photidae, p. 537. Ampithoidae, p. 538. Jassidae, p. 539. 35. Cheluridae. [Sebcdae]. Corophiidae, p. 539. - ———————— ee ee eee ee 526 Annals of the South African Museum. Key to the families of Phronimidea (Hyperidea). (Adapted from Bovallius, K. Sv. Vet. Ak. Handl.,-xxi, No. 5, 1887.) (Peraeopods 1-7 are here reckoned.) Ant. 1 straight, 1st joint of flagellum large, the rest inserted terminally; ant. 2 straight in both sexes 1 Ant. 1 curved, Ist com of aseeiitan Lane ‘te rest inserted subterminally; ant. 2 in ¢ angularly folded (in Lycaeopsidae rudi- mentary in 4, absent in 9) | Ant. 1 flagellum few jointed . 2(1) ) Ant. 1 flagellum in ¢ many jointed, filiform, in | 9 very small, rudimentary or absent 3 eee not swollen . Head large, swollen; roe ethore anf Ant. 1 flagellum elongate, styliform; mandible without palp; inner rami of urops. fused with peduncles, outer rami rudimentary in urops. 1 and 2, small in urop. 3 Ant. 1 flagellum short, terete or laminar; mandible with palp; all rami of urops. well developed, the inner rami not fused with 4 (3) peduncles . Ant. 1 terete; last prp. rena dactyl ees af unguiform Ant. 1 flagellum ee aga or janiceolae (terete in the Antarctic Cyllopus); last prp. at more or less reduced in length, dactylus blunt, digitiform f Ant. 1 flagellum elongate, soliton Gnee forms reaching 70-80 mm.) Sx3) Ant. 1 flagellum oval or lanceolate (small reente up to 15 mm.) Urops. normal, with rami; last prp. normal 7 (2) 4 Urops. foliaceous, without rami; last prp. | reduced sin Mandible without palp . Mandible with palp All peraeopods simple “ 5 with enlarged prehensile hand a Prp. 6, 2nd joint not operculiform, the following joints articulated terminally Prp. 6, 2nd joint operculiform, the following joints articulated subterminally ‘ Ant. 1 attached to anterior surface of head; prp. 11 ( 10)) 5 very slender; prp. 6 robust . Cer 1 attached to inferior surface of head 10. ts [Scinidae]. 5. Lanceolidae, p. 540. Vibiliidae, p. 540. Cystisomatidae. [ Paraphronimidae]. 8. Phrosinidae, p. 541. 9. Hyperudae, p. 540. [ Dairellidae}. Phronimidae, p. 540. ile 14. [Lycacopsidae]. 12. Contributions to the Crustacean Fauna of South Africa. 527 Head not (or very slightly) beanies in a rostral point Head more or less ee aceed: in a Cssteal point Prps. 1 and 2 simple ; 13 (12), Prps. 1 and 2 prehensile, the sepaaded sth one | forming a chela with the dactylus f Prp. 5, 2nd joint normal | Prp. 5, 2nd joint operculiform | Prp. 7 normal 2 (11); 14 (10) [Pep. 7 rudimentary 13 Oxycephalidae, p. 541. [Lycaeidae]. Brachyscelidae. Pronoidae. 15. [Thyropidae ( = Parascelidae) |. Platyscelidae, p. 541. Key to the families of Cyamdea. Body elongate, cylindrical; both pairs of antennae well developed (free-living, among weed, coral, etc.) ; : ‘ Body short, depressed; 2nd antennae rudi- \ -mentary (ecto-parasitic on whales) Caprellidae, p. 542. Cyamidae, p. 548. Key to the genera of Lysianassidae. Mouth-parts greatly projecting below, more or 1 { less styliform. : , Mouth-parts not styliform ; , Hyes and gn. 1 strongly developed. Telkon 2 (1) { entire : A Eyes and gn. 1 not eonaly: deteloped 3 (2) Uropod 3 biramous “ \Uropod 3 uniramous 4 (3) { Maxilla 1 palp rudimentary Maxilla 1 palp 2-jointed 5 (1) ee 1 chelate Gnathopod | not chelate ‘ ; : Gn. 2 chelate. Telson entire. Uropod 3 uni- ramous . biramous Maxilla 1 without eit : Maxilla 1 with 2-jointed palp. Rostrum scarcely developed . Rostrum well developed Side-plates 1 and 2 very small Side-plates 1 and 2 not very small . 0 (9) Telson entire. Gnathopod 1 simple | Telson cleft . Side-plate 1 almost corupletoliy cwatealed be ll ‘ek : ; : ; 9 (7) side-plate 2 Side-plate 1 not concealed on ot subchelate. Telson cleft. Urop. 3 2. 5. Trischizostoma, p. 528. 3. 4. Stomacontion. Acidostoma. Phoxostoma. 6. hs Paravalettia. Luonyx. 8. 9. Amaryllis. Bathyamaryllis. Cyphocaris, p. 528. 10. Lysianassa, p. 529. LL. Aristias. 12: : : | 528 Annals of the South African Museum. al Branchial lamellae pleated on both sides. Branchial lamellae pleated on one side only mt eo lamellae simple, not pleated 13 (12) oe 1 subchelate Gnathopod | simple Prp. 5 longer than prp. 4. Urop. i inner ramus strongly constricted distally Neh 5 not longer than prp. 4. Urop. 2 inner ramus not constricted Mandible, palp attached behind iotomtmalis to) molar Mandible, palp aftached not behind molar Gnathopod 1 imperfectly subchelate Gnathopod | distinctly subchelate . Gnathopod 1, 6th joint distally widened . Gnathopod 1, 6th joint not distally widened Maxilla 2, inner plate not much shorter than outer plate . Maxilla 2, inner plate much anOrter fea outer plate 16 (15) i Gn. 2, 6th joint widening distally. iiveer lip ae " a with forwardly projecting lobe n. 2, 6th joint not widening, slender. tome and upper lip rounded in front, neither of them strongly projecting. 9 (18)) G Epis- 13. Hippomedon. 15. Microlysias. 14. Ichnopus. Socarnopsis. Orchomenella. 16. Uristes, p. 529. 17; Cheirimedon. 18. Tryphosa. 19; Chironesimus, p. 529. ELurythenes. Key to the species of Trischizostoma. Palm of gnathopod 1 with numerous minute spinules or i denticles remipes. Palm of gnathopod 1 aneeth - 2 Gn. 1 inner margin of finger denticulate; pains inne pro- duced : . serratum. les 1 inner margin of hagas noes panned anole not pro- duced pauci-spinosum. Key to the species of Cyphocaris. Prp. 3, 2nd joint, hind ete serrate, not han into a sharp point or spur Prp. 3, 2nd joint, lower hind corner arodiieedl ina shary wenn or spur : Prp. 3, 2nd joint sroducedl in a shows point the margin labore 2 and below serrate ‘ ; ‘ : Prp. 3, 2nd joint produced in a oe spur . f Spur very long, the margin above with a few serrations. | Spur moderately long, the margin above smooth. richardi. 2: anonyx. 3: challengert. faurei. wee Ave Contributions to the Crustacean Fauna of South Africa. 529 Key to the species of Lysianassa. its upper corner, lower corner rounded. Telson apically notched. Urop. 3, peduncle feebly keeled . ‘ . variegata. P segment 3, postero-lateral lobe rounded at both corners. Pleon segment 3, postero-lateral lobe with a minute tooth on Telson apically truncate, entire. Urop. 3, peduncle strongly keeled on upper margin. : 5 : . ceratina. Key to the species of Uristes. Body not indurated. Telson oblong, lobes divergent. Side- plate 1 not very small, widening distally, oblong . . natalensis. Body indurated. Telson lanceolate, lobes contiguous. Side- plate 1 very small, semicircular . ‘ : : . imduratus. Key to the species of Chironesimus. f Pleon segment 3, postero-lateral angle acutely produced . adversicola. | Pleon segment 3, postero-lateral angle rounded . ; . rotundatus. Key to the genera of Stegocephalidae. Telson entire. Mandible not denticulate . Z . Parandania. eee cleft. Mandible denticulate : ; . MStegocephaloides, p. 529. Key to the species of Stegocephaloides. Prp. 5, 2nd joint with postero-inferior apex rounded, hind margin feebly serrate . : . australis. Prp. 5, 2nd joint with postero-inferior apex karate: hind margin distinctly serrate : : : : . attingens. Key to the genera of Ampeliscidae. 1{ Pleon without tufts of setae . : : : : se gu ee jena with tufts of setae ‘ : ' .. Uriedos. 2| Prp. 5, 6th joint foliaceous, 7th dnnceokite : : . Ampelisca, p. 529. ae 5, 6th joint narrow, 7th spiniform . : : . Byblis, p. 529. Key to the species of Byblis. Uropod 3 not extending beyond uropods 1 and 2. Telson not deeply cleft . . gaimardt. Uropod 3 extending much iepenid reap i and 2. Telson deeply cleft. ‘ : : P ‘ . anisuropus. Key to the species of Ampelisca. 1 f Prp. 5, 3rd joint shorter than 4th joint . : sel) ioe | Prp. 5, 3rd joint longer than 4th joint. ; iG: 2 (1) es 5, 7th joint spiniform asin Byblis . : . byblisoides. Prp. 5, 7th joint lanceolate . : . : Snitas 530 Annals of the South African Museum. 2) Prp. 5, 2nd joint postero-distally excavate excavata. Prp. 5, 2nd joint not excavate 4, Pleon segment 3, postero-lateral margin Danate brevicornis. Pleon segment 3, postero-lateral margin not bisinuate 5. Pleon segment 3, postero-lateral angle quadrate chiltoni. Pleon segment 3, postero-lateral angle rounded . fusca. Peraeon segments 5—7 with ventral hooks . diadema. Peraeon segments without ventral hooks . Tee Antenna 2 not much longer than antenna | (the slain 4 lengths may vary in the two sexes) 8. Antenna 2 much longer than antenna | 10. Gn. 1, palm with spines. Prp. 5, 4th joint with fsa apex produced. Corneal lenses small 9: Gn. 1, palm without spines. Prp. 5, 4th joint not pro- duced. Corneal lenses large : : . anomala. Gn. 1, palm with numerous spines. Prp. 5, 4th joint 9 (8) : strongly produced : : : ; . palmata. Gn. 1, palm with about 6 spines. Prp. 5, 4th ay slightly produced spinimmanus. Corneal lenses 4. Pleon segment 3 sHontlsy produced natalensis. 10.7 Comeat lenses 2. Pleon segment 3 strongly produced mops. Key to the genera of Haustoridae. Head elongate. Prp. 4 and 5 with 4th and 5th joints strongly expanded Platyischnopus. Head slightly produced. Prp. 4 nel 5 eas 4th anal 5th joints not greatly expanded Urothoe. Key to the genera of Phoxocephalidae. Prp. 3, 2nd joint linear Eyes absent. Prp. 3, 2nd joint expanded Eyes present. Harpinia. Pontharpinia. Key to the genera of Amphilochidae. 1 Opposing margins of side-plates 3 and 4 not exactly fitting. Gutanopsis. Opposing margins of side-plates 3 and 4 exactly fitting 2. f Mandible, molar absent, palp present Cyproidea. | Mandible, molar present, palp absent Hoplopleon, p.530. Key to the species of Hoplopleon. Prp. 4 (and 5), 2nd joint with hind margin straight. Finger of gn. land 2 strongly ctenate : i . australis. Prp. 4 (and 5), 2nd joint with hind margin convex. Finger of gn. 1 and 2 not ctenate medusarum. Contributions to the Crustacean Fauna of South Africa. Key to the species of Leucothoé. Pleon segment 3, postero-lateral angle produced, with sinus above the point f Palm of gn. 2 minutely serrate | Palm of gn. 2 with a comb-like row of aes Palm of gn. 2 feebly denticulate. Antenna 1 not Bening beyond peraeon : Palm of gn. 2 strongly aenvivilbite: pleon segment 3 { Pleon segment 3, postero-lateral angle quadrate 1 nbusies 1 reaching to Key to the species of Stenothoé. Uropod 3 geniculate. crenulate Uropod 3 not Raunt! joint entire Gn. 2 g, palm with 2 teeth idfore phen one at the ines hinge. * Gn. | greatly elongate Gn. 2 3, palm with only one tooth at the fuser: eines 4 Gn. | not greatly elongate S. adhaerens known only from Q. Gn. 2 g, lower margin of 4th joint Gn. 2 3, lower te of 4th 2 Key to the genera of Phliantidae. if Body not depressed. Telson transverse . | Body depressed. Telson pyriform. f Side-plates small, separated. Head not sunk in oe peraeon | segment - Side-plates 1-4 large, Dae wonk peraeon segment : ese sunk in Ist Key to the genera of Pardaliscidae. Cutting edge of mandible and 2nd joint of palp of maxilla 1 | greatly expanded. Maxilliped, inner piates small, palp very large Cutting edge of jaendible aa 2nd some of male of reesei i | not greatly expanded. Maxilliped, plates obsolete, palp not very large inner 531 a: spinicarpa. ctenochir. richiardii. dolichoceras. gallensis. 2. dolichopous. assimilis. Plioplateia. 2. Temnophlias. Palinnotus. Nicippe. Halice. Key to the genera of Acanthonotozomatidae. Palp of maxilla 1 reaching beyond apex of outer plate Palp of maxilla 1 not reaching apex of outer plate Key to the species of Liljeborgia. Pleon segment 1 without any dorsal teeth. Epistome strongly projecting in an acute point ‘ ee segment | with at least one dorsal tooth. Epistome not very strongly projecting Iphimedia. Panoploea. epistomata. 2. 532 Annals of the South African Museum. 9 ee segments 1, 2, 4 each with 1 medio-dorsal tooth 5. ese Pleon segments 1 and 2 with more than one tooth . Aaa Side-plates 1-3, postero-inferior angles not notched . . proxima. 2 Sides 1-4, postero-inferior angles notched. Side- plate 4 with 4 teeth on hind margin. ; : . consanguinea. Pleon segments 1 and 2 quinquedentate . ; ‘ . dubia. 4 Pleon segments | and 2 tridentate . é 5 : . kinahani var. capensis. Key to the genera of Oedicerotidae. Gn. 1 and 2, process of 5th joint very long and slender . Perioculodes. Kyes completely confluent without any line of demarcation, Ditneer not completely confluent. Gn. 1 and 2, process of 5th 4 joint not very long : ; : : : enya 9 { Gn. 2, hand elongate, chelate : ; : : . Synchelidium. \Gn. 2, hand not elongate, not chelate . : : . Bs Uropod 3 much longer than uropods 1 and 2 . : . Halicreion. ee 3 not much longer than uropods 1 and2 . - Eyes wanting or not placed on the rostral projection. - OF ee or ocular pigment placed on rostral projection . . Ocediceroides, p. 532. ey Peraeopods 1 and 2 with widened joints . : : . Aceroides. °) Peraeopods 1 and 2 without widened joints . : . Bathymedon. Key to the species of Oediceroides. ‘Antenna 1, 2nd joint not plumose. Peraeopod 5, 2nd joint oval . : ‘ : : ‘ . cinderella. Antenna 1, 2nd joint ‘iuoees Peraeopod 5, 2nd joint pyriform : : : : 5 : 4 . plumicornis. Key to the genera of Tirondae. f Telson entire : : : : : : : . Bruzelia. | Telson cleft . , : : : : we oS Eyes 4. Telson deeiety spinose . ‘ ; : » Lorene Eyes 2, coalesced, separate, or none. Telson not spinose. 3. Side-plate 3 scarcely widened below. Gn. | and 2 feeble, 3 imperfectly subchelate . : , . Syrrhoites. Side-plate 3 widened below. Gn. 1] and 9 simples I stouter than 2 . i , : é ‘ ; ; . Austrosyrrhoe. Key to the species of Epimeria (Paramphthoidae). Side-plate 5 acutely produced é : ‘ : Sao ee plate 5 not acutely produced . : : . semiarmata. Peraeon segments 5-7, as well as pleon, ceneae : . cornigera. \ Only the pleon carinate and dentate : : ‘ . longispinosa. Contributions to the Crustacean Fauna of South Africa. 533 Key to the species of Nototropis (Atylidae). Branchial lamellae simple ‘ : ; : . homochir. Branchial lamellae pleated . : ‘ : : . granulosa. Key to the genera of Husiridae. Gn. | and 2, 6th joint attached to the produced apex of 5th i jemi. : ; . Lusirus. Gn. 1 and 2, 6th joint an canal attaclineas ‘ mas 9 Gn. 1 and 2, 5th joint small, cup-shaped. : ‘ ae Gn. 1 and 2, 5th joint large, triangular . : : . Cleonardopsis. 3 . stout . ‘ ; 3 : : ‘ . Lusiroides. Peraeopods slender é : ; : d : . Rhachotropis, p.533. Key to the species of Rhachotropis. Prp. 5, Znd joint with strong process on hind margin. Maxilliped with elongate palp . , . palporum. s 5, 2nd joint without process on hind margin. Balp of maxilliped not very long d ; ; P eas Pleon segment 3, postero-lateral margin not serrate . . anomala. Pleon segment 3, postero-lateral margin serrate : no oJ Prp. 5, 2nd joint acute at lower hind corner . ‘ . kerguelent. herp. 5, 2nd joint not acute at lower hind corner . . 4. al Pleon segment 3 with the dorsal keels ending in teeth . grimaldu. *) Pleon segment 3 with the keels not ending in teeth . . paeneglaber. Key to the species of Paramoera (Pontogeneiidae). Pleon dorsally without teeth. Postantennal angle of head rounded-quadrate . : ; . capensis. Pleon segments |] and 2 each with a flat devil acute Post- antennal angle of head acute . , ; : . bidentata. Key to the genera of Gammaridae. l f Telson emarginate. (Fresh-water. eel i : . Eucrangonyz. Telson cleft . 3 i es: 9 Urop. 3, rami large, tiecoue! en, 1 Sane : : . Megaluropus. Urop. 3, rami not foliaceous. Gen.1subchelate . si Be 3 Urop. 3, outer ramus elongate, its 2nd joint unusually long Hriopisella. Urop. 3, 2nd joint of outer ramus very short . : . A, 4 Mandible, 2nd joint of palp shorter than Ist . : . Parelasmopus. Mandible, 2nd joint of palp longer than Ist . : 2 Ds . with dorsal groups of spinules on pleon. (Fresh- water. Cape Province) . : : . Gammarus.* a ieae without groups of spinules on lea (Marine) ; 16% * For key to the species of Gammarus see Barnard, Trans. Roy. Soc. S. Afr., vol. xiv, p. 167, 1927. WOb. XXkih PART: 5, 37 534 Annals of the South African Museum. Urop. 3, rami very unequal . ; 3 ' ; . Melita, p. 534. Urop. 3, rami not very unequal . ‘ : : a ihe - { Maxilla 1 and 2, inner plates very setose. , : ye sMaxilla 1 and 2, inner plates not very setose . ‘ ctu Urop. 3, rami greatly developed, elongate. Mandible, 3rd | joint of palp as short as lst. : : . Ceradocus, p. 534. Urop. 3, rami short. Mandible, 3rd joint of palp rae [ longer than Ist. : : : : : . Elasmopovdes. 9 eerie palp robust 3 : ‘ . : . Elasmopus, p. 534. Mandibular palp slender : : ‘ : ‘ . Maera, p. 534. Key to the species of Melita. n. 2 in 4, one of the pair very robust, chelate. Prp.5, 2nd joint widest proximally . ; j : : . fresnelir. Gn. 2in g subchelate . : 5 : » 2a Gn. 2 in 4, one of the pair very Fanta Pep: 5, 2nd joint af widest distally : : . subchelata, Gn. 2in § moderate. Prp. 5, 2nd fie dee proximally 3. Pleon segments non-dentate dorsally. Side-plate 1 oblong . : : . zeylanica. Pleon segment 3, poker: stateriok ainclle oe produced. Pleon segment 4 with one dorsal tooth, segment 5 with 2 pairs of denticles. Side-plate 1 triangular, produced \ forwards , : : . : : 5 . Orgasmos. : | Pleon segment 38, postero-inferior angle shortly acute. Key to the species of Ceradocus. | Pleon segments dorsally dentate. 6th joint gn. 1 broadly | . ovate . ; ‘ : é : . rubromaculatus. Pleon segments not serene 6th joint gn. | narrow, . . cylindrical . : : : : A : . aviceps. Key to the species of Hlasmopus. Pleon segments without dorsal! keel : : : . pectenicrus. Pleon segment 4 with a single keel. : : : . japonicus. Key to the species of Maera. Pleon segments without dorsal keels : : : 2 | De ae segment 4 with a pair of keels : : . subcarinata. Urop. 3 scarcely or not at all extending beyond urop.1 . 3. 5 Urop. 3 extending much beyond urop. 1. : maT Prp. 4 and 5, 2nd joint not prolonged at lower hind corner . grossimanus. 3 Prp. 4 and 5, 2nd joint lobed at lower hind corner . ie Side-plates 1 and 2 serrate on lower margin . . . bruzelit. 4 Side-plates 1 and 2 not (or scarcely) serrate . : e258 3 Contributions to the Crustacean Fauna of South Africa. 535 Gn. 2, palm defined by one tooth . : 6. 54 Go 2 2, palm defined by a pair of teeth, 2 palm ath a motoh between two teeth . : . hironieller. Gn. 2, palm ¢ transverse, with neESk, 9 slightly abhgne: | regularly crenulate. Postero-inferior angle pleon seg- ment 3 with short point and notch above it (hind margin sinuous) . : : 3 : . maequipes. 6< Gn. 2, palm § oblique, an Postero-inferior angle pleon segment 3 acute (hind margin evenly concave) . vagans. Gn. 2, palm ¢@ oblique, with 4 strong teeth (incl. one at defining angle). Postero-inferior angle pleon segment 3 quadrate with very small point (hind margin straight) boeckit. “s if Body extremely slender : : : ; : . hamigera. \ Body not extremely slender . ‘ : : 5 . mastersit. Key to the genera of Dexaminidae. imperfect chela with the finger ‘ : . Polycheria. oe joint of all peraeopods distally expanded, forming an Sixth joint of peraeopods not expanded, slender ; . Guernea. Key to the genera of Taltridae. iF Maxilliped, 4th joint of palp wanting or pega size | Maxilliped, 4th joint of palp distinct. or Gay) ‘simple in ¢2. Gn. 2 feebly ener in Of. (Terrestrial. Inland) . : : . Talitroides. Gn. 1 subchelate in f9. Gn. 2 ceetee ehohelate in 6, feebly chelate in 2. (Littoral) s . Orchestia. Gn. 1 subchelate in 3, simple in 9. Gn. 2 feouniy sub- Ke chelate in g, feebly chelate in 9. (Marine. Littoral) Talorchestia, p. 536. alt Urop. 3 consisting of a single joint . : ‘ . Chiltonia. *) Urop. 3 consisting of a peduncle and at least oneramus . 4. Urop. 3 with two rami (the inner one, or both rami, very | small) . : : ‘ ‘ : , : ao Urop. 3 with one ramus ; ‘ ‘ : ‘ 26, fTelson divided. : : ‘ : : . Parhyale. | Teton entire : : : : . Neobule. ME entire (or raaanely mateo) : E : sovile i divided . : 8. Maxilliped, 4th joint of nailer not auernifontt: Yen 1 ap: chelate. Gn. 2 strongly subchelate in g, feebly chelate in 2. Gn. 2 3g, 5th joint not lobed : . Parorchestia, p. 536. Maxilliped, 4th joint unguiform. Gn. 1 and 2 subchelate, stronger in fg thanin 2. Gn. 2, 5th joint lobed below. Parhyalella. Gn. 2, 5th joint not lobed or produced between the 4th and 6th joints in g 3 . Hyale, p. 537. Gn. 2, 5th joint produced heeween 4th ane 6th joints in both sexes. ‘ é : : : : . Allorchestes. 536 Annals of the South African Museum. Key to the species of Yalorchestia. Gnathopod 1 3g, 4th joint without (scabrous) lobe on inner margin in all the species, including the West African species. A. Postero-inferior corner pleon segment 3 rounded. Gn.1 J, 5th joint without lobe on inner apex. Gn. 1 2 2nd joint broad, hind margin convex. Gn. 2 3, 6th joint with large excision on palm in adult. Gn. 2 9, 2nd joint anterior margin evenly convex ; ; : ‘ 5 . capensis. B. Postero-inferior corner pleon segment 3 quadrate, some- times with more or less acute, shortly produced point. Gn. 1 3, 5th joint with lobe on inner apex. 1. Adult $ with pairs of dorsal spines on pleon seg- ments land 2. Gn. 2 g, 6th joint with excision on palm. Gn. 1 9, 2nd joint broad, hind margin convex. Gn. 2 9, 2nd joint anterior margin evenly convex, 5th joint widest at base. Peraeo- pod 4 3g, long, 2nd joint expanded . : . quadrispinosa. 2. Adult g without dorsal spines. Gn. 2 3, 6th joint with evenly convex palm. Gn. 1 9, 2nd joint nearlylinear. Gn. 2 9, 5th joint with lower mar- gin expanded into a broadly rounded lobe. None of the peraeopods unusually long, or with ex- panded joints in g. a. Lower margin pleon segments 2 and 3 not sculptured. Peraeopod 5 2nd joint hind margin with comparatively few serrations. Gn. 2 3, 6th joint widening to palm which is only slightly convex. Gn. 2 9, 2nd joint not strongly expanded on anterior margin, anterior and posterior margins subparallel . : : b. Lower margin pleon segments 2 and 3 witha series of minute ridges appearing as if crimped. Peraeopod 5 2nd joint hind margin with very numerous crenulations. Gn. 2 g, 6th joint oval usually widest in basal third, palm distinctly convex. Gn. 2 9, 2nd joint strongly expanded on proximal half of anterior margin . . ancherdos. australis. Key to the species of Parorchestia. Gn. 1 J, 6th joint subtriangular, widening distally. Gn.2, palm oblique, straight, well-defined . ; é . rectipalma. 1 Gn. 1 3, 6th joint cylindrical. Gn. 2 g, palm not very oblique, rounded with 2 shallow notches, ill-defined. \ Qunknown . 5 : : ; : u . dassenensis. Contributions to the Crustacean Fauna of South Africa. 587 Key to the species of Hyale. Antenna 1, Ist joint with projection on lower apex, g2 . saldanha. a 1, Ist joint without projection . : : eee n. 2 g, hind margin * of hand extremely short : aoe cn 2 3, hind margin of hand not extremely short . . 4, . 1 g, hand as wide as long, palm transverse . : . maroubrae. n. i 6, hand oval, palm oblique . : é . macrodactylus. n. 2 J, finger strongly constricted at ase ; : . diastoma. n. 2 g, finger not constricted at base. : ; eo eS 2, flagellum strongly setose ‘ : : . hirtipalma. Antenna 2, flagellum not strongly setose . ; : oes Prp. 5, hind margin of 2nd joint strongly serrate : . inyacka. ie 5, hind margin of 2nd joint entire (or only feebly crenu- late) F : : 3 ; A : : . grandicornis. Key to the genera of Aorzdae. J Urop. 3 biramous . s : 3 : : g sban ‘\Grop. 3 uniramous. iy 7 : ‘ . Grandidierella. .1 3, 4th joint immensely pugiteed ; : : esha. . 1 g, 5th joint not produced 4 a ; 5% n. 1 g, 5th and 6th joints subequal in width : : . Lembos, p. 537. . 1 g, 5th joint much wider than 6th . : 4 . Lemboides, p. 537. Key to the species of Lembos. No ventral spines. Gn. 2, 2nd joint not hooked 4 oie Ventral spines on peraeon segments 3-7 in g. Gn. 2, 2nd joint hooked at front apex : : . hypacanthus. Prp. 2, 4th joint setose. Gn. 2, hand ee more fiers twice as long as broad . : . hirsutipes. Prp. 2, 4th joint not balcese Gn. 2, had very sicnddes, nearly 4 times as long as broad . : : . leptocheirus. | 2 Key to the species of Lemboides. No ventral spines . : : : : eae 1 : Ventral spines on peraeon re 3-7 ; , : . acanthiger. Gn. 1, palm oblique . : é 4 : : Ofer: 2 : Gn. 1 J, palm transverse : : ; : ; . crenatipalma. Key to the genera of Photidae. 1 Urop. 3 with inner ramus rudimentary . ; A . Cheiriphotis. Urop. 3 withtworami . : ; 3 : ; Qe 9 Urop. 3, rami unequal, the inner ramus minute. . . Photis, p. 538. Urop. 3, rami subequal, the inner not minute . : os 3 Gn. 2 stronger in g than in 9, with 5th joint relatively small Eurystheus, p. 538. Gn. 2 alike in both sexes, with large 5th joint . : . Chevalia. * Exclusive of the palmar portion against which the finger closes. 538 Annals of the South African Museum. Key to the species of Photis. Ocular lobes very long . : ‘ ; : : . dolichommata. oat lobes not very long. : ; seats Gn. 2 3, defining tooth of hand and the lobe of 3rd joint [ pointing inwards. : . longimanus. (e 2 g, defining tooth of — Seta dotnet aS 3rd joint not lobed ‘ : ; ; : . uncinata. Key to the species of Hurystheus. Prp. 3-5, hind margin of 2nd joint strongly dentate . . holmesi. ve 3-5, hind margin of 2nd joint not strongly dentate . 2. 9 ie lageniform or oblong ‘ : : : epee Eyes horizontally or obliquely anal : : : ee f Eyes lageniform . ; : : : : : . atlaniicus. | Eyes oblong . : . : 5 ; : ‘ oi He Gn. Z 3, hand ovoid, palm with several smalllobes . ~ afer: 6 2 3, hand ovoid, palm with 2 large lobes. é . imminens. Gn. 2 3, hand oblong, palm transverse with deep incision . scissimanus. None of the segments dorsally dentate. Gn. 2, finger shorter than palm, closing on to inner surface of hand . palmoides. Pleon segment 4 with 3, segment 5 with 2 dorsal teeth, Gn. 2, finger matching palm and closing on toitsedge . . semideniatus. Key to the genera of Am~pithoidae. f Antenna 1 without accessory flagellum . : : 1 ee \ Antenna 1 with small, 1-jointed accessory flagellum . . Cymadusa. Peraeopod 5 enormously enlarged . : 3 : . Macropisthopous. ee 5 not enlarged : 5 : : Set Side-plates deep. Mandibular palp stout. Prp. 3-5, 6th joint not expanded apically. 5 : . Ampithoé, p. 538. Side-plates shallow. Mandibular palp slendee Prp. 3-5 6th joint apically expanded to form a distinct palm . Hxampithoé. Key to the species of Amprthoé. Gn. 2 3, palm deeply excavated, defined by a conspicuous tooth . ; ‘ . ramondi. Gn. 2 3, palm not deeply Seine not defined ee a con- spicuous tooth , ; : 5 2. Antenna 2 not densely setose. Prp. 1 and 2, 2nd eee strongly expanded. Prp.3 and4very stout. . falsa. Antenna 2 densely setose. Prp. 1 and 2, 2nd joint not strongly expanded. Prp. 3 and 4 not very stout . africana. Contributions to the Crustacean Fauna of South Africa. 539 Key to the genera of Jassidae. l Prp. 1-5, 6th joint not prehensile. : : aoe Prp. 1-5, 6th joint prehensile . : A . Isaeopsis. Gn. 2 3, hind margin of 6th joint ‘rodubedil in a fatale 9 Gn. 2 Q much larger than gn. 1 . ' . Jassa. Gn. 2 3, hind margin of 6th joint not armed. aes 2° not much larger than gn. 1 ; : ‘ . Ischyrocerus, p. 539. Key to the species of Ischyrocerus. None of the peraeon segments keeled ‘ : . angurpes. 1) Some of the segments dorsally keeled in g . A nee a Peraeon segments 1, 2, 6 and 7 keeled 5 : . carinatus. ~) Peraeon segments 1-6 keeled . ; ; : . gorgoniae. Key to the genera of Corophidae. 1 Mandibular palp 3-jointed : : : : side Mandibular palp not 3-jointed . : : E . 4, if Urop. 3, inner ramus distinct but very minute . . Camacho. Urop. 3, inner ramus wanting, or not articulated bi eb Urop. 2 biramous . ; 5 ‘ . : . Ericthonius. Urop. 2 uniramous . ; . Cerapus. Mandibular palp 1- jointed. nitions 2 ae very stout Siphonoecetes, p. 539. Mandibular palp 2-jointed. Antenna 2 very stout . Corophium, p. 539. Key to the species of Siphonoecetes. fAntenna 1, flagellum 7-jointed . . ? ; . dellavallei. |Antenna 1, flagellum 10-14-jointed . : : . orventalis. Key to the species of Corophium. Pleon segments 4-6 coalesced. Gn. 2, dactylus with feeble decumbent denticles. ; : : . acherusicum. Pleon segments 4-6 distinct. Gn. 2, dactylus with strong outstanding denticles : é P . triaenonye. Key to the genera of Podoceridae. Pleon with only 5 distinct segments in front of the telson. Two pairs of uropods . : 3 . Laetmatophilus, p. 539. Ree with 6 distinct segments in front of telson. Three pairs of uropods : ; : : ‘ . Podocerus, p. 540. Key to the species of Laetmatophilus. 1 Gn. 2 3, palm with 3 teeth 3 : ‘ : . tridens. Gn. 3 g, palm with 2 teeth : : : : «Be 9 Gn. 1, 6th joint widened . , : : : . purus. Gn. 1, 6th joint not widened . : ; ; . durbanensis. 540 Annals of the South African Museum. Key to the species of Podocerus. Body not carinate . ; , : : : Plates XVIII, XIX.) “ae LAWRENCE, Ph.D., Natal Museum, Pietermaritzburg. (With 47 Text-figures.) F.L.S., Assistant Director. (With 19 Text-figures.) _G. A. Mavromovustaxis, Limassol, Cyprus. (With 1 Text-figure.) _ : Title Page and Index to Volume XXXII. Bees. : ey : / Ry ie : : ---—«ISSUED AUGUST 1940. PRICE 10s hes : : Eh aye PRINTED FOR THE 3 | S81 BES OF THE SOUTH AFRICAN MUSEUM, CAPE TOWN. Re ets BY NEILL AND CO., LTD., Ee es 212 CAUSEWAYSIDE, EDINBURGH 19. Ne otes on the Early Stages of Phasis felthami Trim., a Lycaenid he 20. The Genus Selenops (Araneae) in South Africa. By R. F. aL, Additional Records, and Descriptions of New Species, of South ; _ African Alder-flies (Megaloptera), May-flies (Ephemerop- ie tera), Caddis-flies (Trichoptera), Stone-flies (Perlaria), and — Dragon-flies (Odonata).—By K. B. Barnarp, D.8c., ae 2. Deseriptions of New African Anthidunae (Apoidea).—By " ¢ > * | é i ° ; } . 8g pho LA huh / j ; ES oh ne 2 oo eae J 4 + i: 3 j , f % % : 2) 4 y, ue ¥ 4 ( 545 ) 19. Notes on the Early Stages of Phasis felthami Trim., a Lycaenid Butterfly from the Cape Peninsula, and a List of some recently determined Food-plants of some other South African Butterflies. —By C. G. C. Dickson. (With Plates XVIII, XIX.) THE genus Phasis constitutes one of the largest groups of the South African Lycaenidae, being represented in the Union by about thirty known species, a number of which are peculiar to the country. Despite the fact that many of these insects are abundant (though generally local), particularly in the Western Districts of the Cape Province, the life-histories of the great majority of them are still unknown. The species whose early stages are outlined in these notes was described by Trimen in the Transactions of the Entomological Society of London, 1904, p. 233, and recently redescribed and figured by the Rev. Desmond P. Murray in “South African Butterflies: A Monograph of the Family Lycaenidae,” p. 108, and coloured illustration, No. 55. I am very much indebted to my friend, Mr. Gowan C. Clark, of Port Elizabeth, to whom I forwarded examples of the egg and larva of the insect in March 1937, for the loan of a set of beautifully executed enlarged paintings, some of which are reproduced in half-tone on Plate XIX, and for furnishing me with an interesting description of the action and function of the retractile tubercles of the larva. Near Milnerton on 20th April 1936 a female of Phasis felthami was kept under observation for the purpose of procuring eggs of the species and ascertaining the food-plant. The butterfly was followed for some time until it eventually selected a suitable spot for ovipositing, this proving to be a collection of dead leaves and other debris, which had accumulated under a low succulent-leaved shrub (Zygophyllum sessilifolium L., Plate XVIII), which grows commonly in the sandy ground bordering the coast-line. The insect alighted under the bush _and immediately commenced ovipositing, crawling about for short distances and laying eggs singly at fairly frequent intervals on the withered leaves, small twigs, and among particles of loose sand. VOL. XXXII, PART 6 38 546 Annals of the South African Museum. Hight or more eggs were laid before it moved off again into open ground. The eggs were deposited at about 3.30 p.m. The egg (Plate XIX) is about -8 mm. in diameter, subglobular in shape, flattened at the base, and creamy-white in colour. The crown is somewhat compressed and the micropyle sunken. The surface is sculptured with a fine network of ridges, the interstices being extremely small round the centre of egg and attaining a maximum size on the sides. The egg-stage lasted 20 days—cold weather prevailing during this period—and a single larva hatched out on 10th May. Prior to emergence a portion of the crown of the egg is eaten away, and before long a few hairs of the larva appear through the opening thus formed. The larva on emergence (Plate XIX) is about 1 mm. in length, proportionately broad and considerably flattened. Ground colour dull yellow. Upperside sparsely speckled with black and marked longitudinally with broken reddish streaks. Head of moderate size; black. A broad area on first segment and a disc on anal segment hardened; dark brown or blackish. Surrounding body is a ridge, set with prominent, mostly curved, bristles. On the 10th segment dorsally is a pair of long branched spines, normally prostrate but at times directed upwards. Setae, forming a double row on dorsal surface, appressed. Pair of retractile tubercles on 11th segment. The larva feeds on the leaves of the food-plant, devouring the fleshy interior substance and the overlying cuticle. Feeding occurs between long intervals of rest, when the larva remains motionless in some chosen resting place. The rate of growth in captivity was remarkably slow, though this may have been caused by the change to winter conditions usual in the month of May. Unfortunately the single larva obtained was lost, due to falling off a sprig of the food-plant, a month after emergence. It was then still in the first instar. Mr. Clark succeeded in rearing a larva from the egg to the end of the third instar and has supplied me with the following record, which may perhaps be taken as typical of the normal rate of development of the larvae :— Duration of first instar . 10 days, 7 ,, Second ,, Bele ask u ,, third e . 12. ,, (larva died while moulting). After the first moult the form of the larva closely approximates to that ultimately assumed. The pair of branched spines are lost, the Notes on the Early Stages of Phasis felthami Trim. 547 lateral bristles are increased in number, most of the more prominent longitudinal lines (which are dull light-red at this stage) and some other principal markings of the mature larva are recognisably distinct, and, as Mr. Clark has pointed out to me, the honey gland now appears on the 10th segment. The ground-colour of the larva is white at this period, but in time changes to a stone-coloured tint. On 6th June a small colony of 15 larvae, ranging in size from 5 mm. to 13 mm., was found in the original locality at the base of one of the food-plants. The larvae were associated with ants (a species of Cremastogaster) and were afforded cover and concealment among a mass of debris consisting of withered leaf and other vegetable fragments mixed with numerous grains of sand, small pebbles and a few small snails’ shells. This accumulation was loosely bound to- gether with silken web, being partially attached to the bottom of the stem, and at one point extending below the surface of the ground. A few larvae were found below ground-level in a little cavity at the side of the stem, and two others were seen some distance up the stem within reach of the leaves. It was obvious that the larvae were engaging the attention of the ants, the source of attraction being the honey-gland, a characteristic feature of many of the larvae of the Lycaenidae. The actual act of “milking” on the part of the ants was not observed on this particular occasion, though the retractile tubercles in close proximity to the gland came into action whenever the larvae were disturbed. Similar ants were previously seen under the Zygophyllum bush when the butterfly was ovipositing. The larvae were collected and later placed on a young food-plant growing in a pot, a light framework of sticks being secured over the top and covered with mosquito netting. A portion of the original “nest”? was attached to the bottom of the stem. In captivity the larvae existed independently of ants and the majority ultimately pupated, but their development was very slow, and it was not until 2nd August that pupation of the largest specimen took place. The larvae were very sluggish in their movements, and when not engaged in feeding continued to assemble together on the side of the stem and among the debris. When the larva is inactive the head is retracted into the first segment. At a later date further larvae in smaller groups, and at times occurring singly, were dis- covered below the food-plants, and occasionally in fissures that had formed between the stems. Zygophyllum flecuosum EH. & Z. (Plate XVIII) and Z. morgsana L. were found to be additional plants used by the species. 548 Annals of the South African Museum. The full grown larva (Plate XVIII, larva about 80 per cent. full grown; Plate XIX, lateral view of 7th segment) measures 16-17 mm., is broad, and ventrally flattened. Dorsum slightly arched; ridge surrounding body crenate on sides and upswept on first segment, curving round in front of well chitinised dorsal portion; last segment much flattened and sloping to anal extremity, with a circular dorsal area chitinised like upper surface of first segment. Numerous small stud-shaped spines (Plate XIX), perceptible as such only under a strong lens, are distributed over upper surface of body and over ridge (these contributing towards producing pattern of upperside); spines variously coloured: black, white, or more or less of the ground-colour of body. Bristles on sides of body and round extremities brownish grey, the majority darkening appreciably towards base, of various lengths and considerably more numerous and relatively shorter than on newly emerged larva. Ground colour of body above, stone-coloured; often tinged in parts, especially on dorsal surface, with salmon-pinkish. Ventral surface (including prolegs) pale green to greenish-grey, tinged extensively with pale vinous. Head black, tinged on clypeus with dull reddish; division between lobes whitish; basal section of antennae red. Greater portion of dorsal area of first segment blackish. Longitudinal lines reddish brown to blackish; all except a barely separated very dark medio-dorsal pair, decidedly sinuous or irregular. Narrow space between sub-dorsal line and a series of very dark markings, creamish; line above spiracles, separated from the latter series by a broad strip, darker (often considerably darker) than general back- ground; the lowest line of upperside between spiracles and margin of lateral ridge. Large areas of last three segments, particularly 11th and terminal ones, blackish. Underside of body bordered below ridge with deep maroon. Thoracic legs brown, for most part very dark; fleshy portions of the ground colour of ventral surface. In regard to the honey-gland (situated between a greenish false gland and a dull red bulge) and retractile tubercles (Plate XIX), Mr. Clark in a letter to me states :— “The tubercles of the Lycaenidae, at one time an unexplained organ, can now I think be described as a sentinel of the honey gland. “In the case of Myrina ficedula Trim. it is merely a small warning projection. In the smaller species of the genus Cupzdo it is more developed and is a larger projection crowned with barbed and pointed spines, the larvae of telicanus Lang. sometimes using the partially extended tubercle as a brush. The larvae of Phasis osbecki Aur. and Notes on the Early Stages of Phasis felthami Trim. 549 felthami Trim. definitely use the tubercle as a brush or whip and sweep the vicinity of the honey-gland with quick, determined lashes. “Dealing with Phasis felthami, the tubercle is present in the newly hatched larva and can be made to function by touching the larva with a hair of a paint brush. This organ is level with the body and difficult to detect until it functions. The honey-gland is not apparent in the first stage. “ After the first moult the tubercle is encased in a prominent black cylindrical mole and is shot out on being disturbed and remains stationary for some seconds. It resembles a white club with seven spines, three on each side and one on top, the seven being in one row. “After the second moult the cylindrical mole containing the tubercle develops a semicircle of spokes round the rim and these act as a protection to the protruding hairs; the remaining portion of the circle is open to allow the tubercle to function. “The tubercle from this stage on consists of a piston fitted with three hairs on top and two more on each side near the top, working in and out of the cylindrical case. “As the piston is protruded the three top hairs strike out and with the bending of the piston strike the vicinity of the honey-gland. The two hairs on either side shoot out with the rest but as the piston is extended their position on the side makes them sweep through an arc, the one to the side, the other to the rear on each side. In this way a large field is dusted. “The beats are very rapid, vibrating at the rate of some two or three strokes per second, and are preceded by fretful tremblings of the unextended tubercle. “Any small insects attempting to approach the honey-gland meet with a whirl of beating hairs and retreat. In this way the honey is reserved for the larva’s protectors.” In captivity pupation was in most instances effected on the surface of the pile of debris, at the base of the food-plant. The pupa (Plate XVIII) is 9-10 mm. in length, being relatively full in diameter; colour very dark brown. Front of head rounded; thorax distinctly convex dorsally; wing bases protuberant; abdomen wide to about middle, then tapering, curving downwards to extremity and terminating obtusely. Cremastral hooks small, golden-brown, disposed beneath, and over end of terminal segment. Pair of small protuberant patches above eyes and first one or two small pro- tuberant spiracular patches, pale buff; remaining spiracular patches terra-cotta. Last few segments narrowly marked at edges (principally 550 Annals of the South African Museum. in dorsal and lateral regions), and the anal segment rather prominently beneath with terra-cotta. The pupa derived from the larva previously mentioned as pupating on 2nd August produced a male imago on 6th September, i.e. 35 days later. An example bred by Mr. Clark remained in the pupal state for 21 days. The majority of the butterflies emerged earlier in the season than would have been the case under natural conditions, and all were below the usual average size of the species. The zmago (Plate XVIII) does not normally appear in any numbers before October, remaining out until the middle or end of December. It reappears.in February and continues up to the beginning of May. A badly worn specimen, however, has been noted as late as 27th May. The insect is very local, but not uncommon where it occurs. It is often found frequenting plants of Munda spinosa DC. (Schilpadbessie) and on one occasion this shrub was used for ovipositing, a single egg ‘being laid, but a larva which was subsequently offered the plant as food refused to feed upon it. The species was discovered by the late Mr. H. L. L. Feltham in the Cape Peninsula, October—February, 1900-1902. Distribution.—Localities in Cape Town district: Retreat; Ronde Vlei; Strandfontein; Durbanville; near Milnerton; Melkbosch Strand; near Brakfontein (on road to Mamre). Further locality: Near Montagu. The larva of Phasis felthami bears a close superficial resemblance to that of Phasis thysbe L. (an abundant species in most of the above localities), but may be distinguished from this by the longer bristles on body and round tubercles, the better defined longitudinal lines, the absence of a pair of small white protuberances (composed of minute closely set flower-shaped spines) on first segment, and the different form of some of the markings on the three rear segments. It should be noted, however, that a colour variety of the larva of thysbe sometimes occurs in which green replaces the usual brownish- grey hue of the upper surface of the body. Notes on the Early Stages of Phasis felthami Trim. 551 A List oF RECENTLY DETERMINED FOOD-PLANTS OF A FEW SoutH AFRICAN BUTTERFLIES’ LARVAE. By C. G. C. Dickson and Gowan C. Cuarx of Port Elizabeth, who has kindly supplied the names of food-plants (with notes) of all species marked with an asterisk. SPECIES OF BUTTERFLIES. FooD-PLANTS AND NOTES. Danaida chrysippus L. Stapelia variegata L. (Asclepiadaceae). Pseudonympha hyperbius L. Ehrharta (Gramineae): Eggs laid singly on leaf-blades. P. vigilans Trim. = Restio cincinnatus Mast. (Restionaceae): Eggs laid singly on stems. Leptoneura mintha Geyer. Ficinia (Cyperaceae): Butterfly drops eggs while fluttering in middle of clump of grass. (Observed by Mr. P. R. Robertson and myself.) Acraea horta L. Asclepias curassavica L. (Garden shrub introduced from West Indies) (As- clepiadaceae): Food-plant noted by Mr. N.S. Pillans, to whom I am also indebted for the identification of several other plants. Pyrameis cardui L. Arctotis, cultivated (Compositae): Eggs laid singly on stems and leaves. eS Precis (Junonia) cebrene Trim. Barleria pungens L. (Acanthaceae): Eggs laid singly, hidden between leaves of young shoot. * Cupido lysimon Hubn. | Amaranthus deflexus L. (Amaran- thaceae). Tribulus terrestris L. (Zygophyllaceae): Eggs laid singly under leaf. C. lingeus Cram. Salvia species (Labiatae): Eggs laid singly on leaves. C. thespis L. Saxifraga species (introduced from Europe) (Saxifragaceae). Phylica imberbis Berg. (Rhamnaceae): Eggs laid singly on flower-heads and leaves. Lycaenesthes definita Butl. Rhus species (Anacardiaceae). Acacia saligna Wendl. (Port Jackson willow, introduced) (Leguminosae): Eggs laid singly on young flower- heads and leaves. * Deudorix antalus Hopff. Schotea speciosa Jacq. (Leguminosae): Eggs laid singly. Larvae remain in pod, eating the bean. 552 Annals of the South African Museum. SPECIES OF BUTTERFLIES. Phasis zeuxo L., small dark sub- species. P. felthami Trim. P. pyroeis Trim. P. thysbe L., and var. osbecki Aur. P. thysbe, variety with blue area of male much reduced (principally in forewing) in comparison with that of ordinary form (hind-margins dentate). P. thysbe, variety resembling osbecki but with blue area in forewing of male reduced (hind-margins non- dentate). P. thero L. P, wallengreni Trim. P. malagrida Wllg. * Terias brigittu Cram. FOOD-PLANTS AND NOTES. Osteospermum moniliferum L. (Com- positae): Eggs laid singly on under- side of leaves (observed on only one occasion). Zygophyllum sessilifolium L. Z, flecuosum E. & Z. Z. morgsana L. (Zygophyllaceae): Eggs laid singly under plant. Larva associated with ants (Cremastogaster species). Zygophyllum species and association of larva with ants as above. Zygophyllum sessilifolium L. Z. flecuosum EH. & Z. Z. morgsana L.: Eggs laid singly on younger stems and underside of leaves. Larva associated with ants (Cremastogaster peringuey: Emery). | Aspalathus spinosa L. (Leguminosae). Zygophyllum species as above, also association with ants. Osteospermum moniliferum L.: laid singly on stems and leaves. Eggs Melianthus major L. (Melianthaceae): Ascertained by the Rev. D. P. Murray since the publication of his work. Rhus species. In case of latter food- plant a single egg was seen being deposited in cleft between flower- bearing and leaf stems. Senecio pubigerus L. (Compositae): Eggs laid in small batches on stems. Spinous scales from butterfly’s abdo- men adhere to surface of egg. Ifloga laricifolia Less. (Compositae). Aspalathus species: Eggs laid singly or in pairs on stems. Numerous scales adhere to surface of egg. Cassia mimosoides L. (Leguminosae): Eggs laid singly near tip of young frond. Notes on the Early Stages of Phasis felthami Trim. 553 SPECIES OF BUTTERFLIES. FooD-PLANTS AND NOTES. * Preris zochalia Boisd. Maerua triphylla Thunb. (Capparideae): Eggs laid singly, generally on upper side of leaf. * P. mesentina Cram. Capparis oleoides Burch. Maerua triphylla Thunb. (Capparideae): Eggs laid in cluster, generally on underside of leaf or near terminal of young shoot. * P. gidica Godt. Capparis citrifola Lam.: Eggs laid singly on leaf of young shoot. * P. severina Cram. Maerua triphylla Thunb. Capparis oleoides Burch. C. zeyhert Turez.: Eggs laid in cluster, generally on underside of leaf. * Herpaenia eriphia Godt. Maerua triphylla Thunb.: Eggs laid in small clusters on leaf. * Teracolus eris Klug. Capparis oleoides Burch.: Eggs laid singly, generally under young leaf. * T. omphale Godt. \ Capparis citrifolia Lam. * T. antigone Boisd. | fee juncea Harv. (Capparideae): In each case eggs laid singly on young shoot. * Hronia buquetii Boisd. Azma _ tetracantha Lam. (Salvado- raceae): Eggs laid singly on young shoot. Kedestes lenis Riley. Imperata arundinacea Cyr. var. thun- berg Hack. (Gramineae): Eggs laid singly on leaf-blades. Parnara mohopaani Wlig. Food-plant and egg-laying as above. (Observed by Mr. P. R. Robertson.) A number of the above plants were kindly identified for me at the Bolus Herbarium, Kirstenbosch, and at the Herbarium of the South African Museum. EXPLANATION OF PLATES. Prate XVIII. Phasis felthami Trim. (Plate of black and white drawings by C. G. C. Dickson.) Pupa, <3. Larva, <3 (about 80 per cent. full grown). Imago, <1, male and female. Food-plants, Zygophyllum sessilifolium L. and Z. flexuosum EK. & Z. (ant associated with larva on tip of former). 554 Annals of the South African Museum. PurateE XIX. Phasis felthami Trim. (Half-tone reproductions of coloured drawings by Gowan. C. Clark.) Egg, x20, and section of surface highly magnified. Larva on emergence, x 40, cross-section of same, and outline of last two segments showing position of tubercles and pair of branched spines. Lateral view of 7th segment of full-grown larva (from right side). Stud-shaped spines. Tubercle (half extended). : Tubercle (fully extended) and position of spiracle. Dorsal view of last three segments, showite position of honey gland (indicated by arrow) and tubercles. Plate XVIII. Ann. §. Afr. Mus., Vol. XXXII. PHASIS FELTHAMI TRIM. Neill & Co., Ltd. C. G. C. Dickson, del. Ann. S. Afr. Mus., Vol. XXXII. Plate’ XPX. PHASIS FELTHAMI TRIM. G. C. Clark, pinz. Neill & Co., Lid. ( 555 ) 20. The Genus Selenops (Araneae) in South Africa. By R. F. Lawrence, Ph.D., Natal Museum, Pietermaritzburg. (With 47 Text-figures.) THE genus Selenops has for some time stood in need of revision owing to the large number of undescribed forms which occur in South Africa. An exceptionally large number of new species has been discovered in the collections of the various Museums of South Africa, and the writer is indebted to the Directors of these institutions for the loan of their material. The revision of the genus is based on specimens from the South African Museum, where the great majority of the types are housed, the Transvaal Museum, Pretoria, the Albany Museum, Grahamstown, and the Natal Museum, Pietermaritzburg. Most of the species of Selenops in South Africa have a comparatively limited distribution and each subregion differs in its fauna from the others. The species which is best represented in tropical Africa and which appears to have the widest distribution of all African forms of Selenops, viz. S. radiatus Latr., is poorly represented in the South African region. This species appears to undergo a process of breaking up into a number of closely allied subspecies as it passes over from the tropical African to the temperate South African zone. On the other hand the smaller forms of Selenops, with a larger number of tibial spines than is the case in radiatus, are abundant, both numerically and with regard to the large number of species into which they break up. All members of the genus seem to be typical eryptozoic and nocturnal forms, living in retreat under logs and stones during the day, but often attracted by bright light into houses at night, where they can be seen resting on the walls of rooms. The following abbreviations have been used throughout the paper to indicate the Institutions where the types and other specimens are to be found: — S.A.M. South African Museum, Cape Town. T.M. Transvaal Museum, Pretoria. N.M. . Natal Museum, Pietermaritzburg. A.M. Albany Museum, Grahamstown. 556 Annals of the South African Museum. The following key may be used for distinguishing the main groups of species of Selenops found in the South African region :— Group A.—Metatarsi of anterior legs with 2 pairs of inferior spines, tibiae with 3 pairs of inferior gia (Simon’s radiatus group) ; p- 556 Group B.—Metatarsi of anterior legs with 3 pairs of sateen spines, tibiae with 4—7 pairs of inferior spines (Simon’s atomarius group) . p. 564 Sub-group B,: Anterior tibiae with 4 pairs of sna spines’ . : . ‘pr spGe Sub-group B,: Anterior tibiae with 5 pairs of inferior spines . : . See oer Sub-group B;: Anterior tibiae with 6 pairs of inferior spines. : ; . pe oer Sub-group By: Anterior tibiae with 7 pairs of inferior Spines . p-. 604 GROUP A. Key to the species of Selenops of the radiatus group. oe. 1. Vulva with two lobes meeting in the middle line; tibiae of a with black bands . : : : 2 Sy eee Vulva without lobes, with a Aieapilat more or tess roused opening; tibiae of legs uniformly blackish-brown : : 5 i 5. 2. Anterior opening of vulva rounded, a little wider tian ee radiatus radiatus, radiatus peryensis. Anterior opening of vulva triangular, considerably wider than long = diseee 3. Opening of vulva much wider than long, lateral lobes not completely separated radiatus damaranus. Lateral lobes completely separated . : : . : : . oe 4. Anterior median eyes more than their diameter apart . . radiatus krugeri. Anterior median eyes less than their diameter apart . radiatus ovambicus. 5. A line touching the upper surfaces of anterior median eyes passing below the centres of the posterior medians : : ; rhodesianus. A line touching the upper surfaces of sileniGt saedian eyes passing above the centres of the posterior medians : 2 é : : : sye'¢aGs 6. Opening of vulva wider than long : : : : : : a Opening of vulva longer than wide . ; ; 3 j é ee 7. Opening of vulva subquadrate . ; : ‘ : . brachycephalus. Opening of vulva rounded . ; : ‘ : : . zuluanus. 8. Anterior median eyes less than their aiatnetis apart. : : lesnet. Anterior median eyes more than their diameter apart . ; . tenebrosus. The Genus Selenops (Araneae) in South Africa. 557 3d. 1. Pedipalp tibia with two widely separated processes é ‘ : abwin Be Pedipalp tibia with two processes close to each other . ; : Ba 'cbi3 2. Processes of pedipalp tibia almost equal in size . ‘ radiatus krugert. Processes of pedipalp tibia unequal in size . é , radiatus radiatus. 3. Tibia of pedipalp longer than patella, subequal to tarsus : é lesnet. Tibia of pedipalp shorter than patella, much shorter than tarsus . zuluanus. Selenops radiatus damaranus n. subsp. (fig. 1). Type, 1 2, Sandfontein (185 miles east of Windhoek), South West Africa (S8.A.M., B. 5637). Colour.—Carapace reddish brown with a number of short black stripes radiating from the fovea but well removed from it, cephalic area a little darker than thoracic area, its boundaries defined by darker lines; mandibles dark reddish brown, darker than carapace; abdomen mottled with dark brown spots and smaller dots on a yellow background, the whole appearance not dark; femora of legs mottled, tibia banded with distinct blackish markings, under surface of anterior femora entirely dark, metatarsi entirely dark brown. Du aR seat eae ; amaranus n. subsp. Hyes.—Medians weakly recurved, a line Q, vulva. touching the upper surfaces of the anteriors passing well above the centres of the posterior medians, anterior medians # the diameter of the posterior medians, their own diameter apart, and a little less than a radius from the posterior medians; posterior medians half the diameter of the posterior laterals; anterior medians a little more than a radius from the edge of the clypeus. Chelicerae.—Inferior margin with 2 equal-sized teeth separated by their basal width. Legs.—Anterior tibiae with 3 pairs, anterior metatarsi with 2 pairs of inferior spines. Vulva as in fig. 1. Dimensions.—Length of carapace 5-2, width of carapace 6-2, total length 14 mm. Other Speciomens.—1 9, Loangwa Valley, N. Rhodesia (S.A.M., 9494). Selenops radiatus ovambicus n. subsp. (fig. 2). Type, 1 2, Ongandjera, Ovamboland, 8.W. Africa (S.A.M., B. 6221). Colour.—Carapace reddish brown, cephalic portion a little darker 558 ~ Annals of the South African Museum. than thoracic portion, defined at its posterior apex by two short blackish stripes, a narrow incomplete blackish stripe in the middle; thoracic portion with some submarginal spots, radiations from the thoracic stria short and ill-defined, the stria itself deeply grooved and blackish; mandibles blackish brown, ocular area dark; abdomen much macerated; femora of anterior legs with 2 distinct blackish bands on their anterior surfaces, these confluent along the under sides; anterior tibiae with 2 black and 2 lighter bands; labium blackish brown, sternum narrowly margined with black. Eyes.—Median eyes slightly recurved, a line touching the upper surfaces of the anteriors passing between the centres and the upper surfaces of the posterior medians; anterior medians a little smaller than the posterior medians, a little less than their own diameter apart, and about half their radius from the posterior medians; posterior medians about 3 the diameter of the posterior laterals; anterior medians 3 their diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 moderate _ equal-sized teeth separated by their basal width Fie, 2-—Selonpe red from each other subsp. 9, vulva. Vulva as in fig. 2 differing somewhat in detail from damaranus and the typical form of radiatus. Legs.—Anterior tibiae with 3, anterior metatarsi with 2 pairs of inferior spines. Dimensions.—Length of carapace 5-2, width 5-9, total length 13-4 mm. This specimen was designated as a typical radiatus by the writer in Ann. S.A. Mus., vol. xxv, pt. 1, p. 41, 1927. Selenops radiatus krugeri n. subsp. (fig. 3). Types, 1 2, Sabie Reserve, Transvaal (S.A.M., B. 5735); 1 g, Sabie Reserve (S.A.M., B. 7164), coll. HE. L. Gull. 2 Colour.—Carapace reddish brown, cephalic portion dark reddish brown; fovea strongly, radiations faintly marked; mandibles reddish black; abdomen rubbed above, mottled with brown spots and specks; anterior femora with 3 faint dark bands, tibiae with fairly strong dark bands. Eyes.—Median eyes weakly recurved, a line touching the upper surfaces of the anteriors passing through the centres of the posterior medians; anterior medians 3 the diameter of the posterior medians, The Genus Selenops (Araneae) in South Africa. 559 more than their diameter apart, and a radius from the posterior medians; posterior medians $—2 the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 strong equal-sized teeth, less than their basal width from each other. Vulva as in fig. 3, a. Legs.—Tibia I and II with 3 pairs of inferior spines, metatarsus I and ITI with 2 pairs of inferior spines; no lateral spines. Cc Fic. 3.—Selenops radiatus krugeri n. subsp. a, 9, vulva. 6, c, tibia of pedipalp, g, from in front and from outer side. Dimensions.—Length of carapace 7, length of abdomen 9-8 mm. 3 Colour as in description of 2, except that the tibiae of legs are entirely dark except at base and apex. Eyes and chelicerae as in description of 2, the teeth on inferior margin of chelicerae their basal width apart. Legs.—Anterior tibiae with 3 pairs of inferior spines, anterior metatarsi with 2 pairs, tibia I with a lateral spine on each side near the base, tibia II without lateral spines. Pedipalp.—Femur with 1 outer, 3 superior spines near the apex, patella without spines; tibia distinctly longer than patella but shorter than tarsus; process of tibia as in fig. 3, 6, seen from in front, fig. 3, ¢, seen from outer side. Dimensions.—Length of carapace 6, total length 12-7 mm. Other Specimens.—1 9, Potgietersrust, Transvaal (T.M., 1853); 19, Kuruman (S.A.M., 13176); 19, Lake Ngami (8.A.M., 13354), 560 Annals of the South African Museum. Selenops radiatus radiatus Latr. The species is not common in the South African region and only occupies the northern fringe of it. The South African Museum has specimens from the following localities: 1 9, Salisbury, 8. Rhodesia (2045); 1 9, Windhuk, 8.W. Africa (B. 5169); 1 9, Otjituo, S.W. Africa (B. 5028); 1 9, Namakunde, S.W. Africa (B. 5028). Lessert also records it from Portuguese H. Africa (Chemba, Nova Choupanga, Sinjal) and Angola. Selenops radiatus peryensis Lessert. S. radiatus peryensis Lessert, Rev. Suisse. Zool., vol. xliu, p. 263, figs. 57-59, 61, 1936. The types, Q and 3, were described from Vila Pery, Portuguese HK. Africa. Selenops rhodesianus n. sp (fig. 4). Type, 1 9, Salisbury, Rhodesia (S.A.M., B. 3284). Colour.—Carapace reddish brown with a distinct narrow blackish margin, decorated with indistinct markings, eyes surrounded by black; mandibles ‘reddish brown; abdomen mottled blackish brown, almost black posteriorly; legs with moderate fuscous markings, these not very clearly defined. Eyes.—Medians not strongly recurved, a line touching the upper surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians $ the diameter of the posterior medians, 14 times their own diameter Hig. d= Scenes ene ae apart, and a httle more than a radius n.sp. 9, vulva. from the posterior medians; posterior medians half the diameter of the pos- terior laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal one a little larger than the apical one, from which it is separated by its basal width. Legs.—Tibiae of anterior legs with 3 pairs, metatarsi of anterior legs with 2 pairs of inferior spines. Vulva as in fig. 4. The Genus Selenops (Araneae) in South Africa. 561 Dimensions.—Length of carapace 5-4, width of carapace 6, length of abdomen 6:3 mm. Other Specumens.—2 29, Mokeetsi, E. Transvaal (T.M., 6411). Selenops brachycephalus n. sp. (fig. 5). Type, 1 9, Salisbury, Rhodesia (S.A.M., 3292). Colour.—Carapace brown with some darker markings radiating forwards and sideways from the thoracic stria, those on the cephalic portion more strongly defined; man- dibles uniform reddish brown, hardly darker than carapace; abdomen above dark mottled brown, a few white speckles among the predominating brown ones; legs brown with some vague infuscations. Eyes. — Median eyes weakly re- curved, a line touching the upper surfaces of the anteriors passing well 4, sale ayes cache. ohne above the centres of the posterior n.sp. 9, vulva. medians; anterior medians 2 the diameter of posterior medians, their own diameter or a little more apart, and a radius from the posterior medians; posterior medians 4-2 the diameter of the posterior laterals; anterior medians about a - radius from the edge of the clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal slightly larger than the apical one, separated by less than the greatest width of the basal tooth. Legs.—Tibia I and II with 3 pairs of inferior spines, metatarsus I and II with 2 pairs of inferior spines. Vulva as in fig. 5. Dimensions.—Length of carapace 5-7, width of carapace 6-7, length of abdomen 6 mm. Selenops zuluanus un. sp. (fig. 6). Types, 1 3, 2 22, Ingwavuma, Zululand (N.M., 2398). 2 Colour.—Carapace reddish brown, darker anteriorly, with a blackish marginal border, some ill-defined radiations from the thoracic stria, and an indistinct dark stripe bisecting the cephalic area; man- dibles much darker than carapace, almost black; abdomen above almost uniformly black, speckled with some minute light dots, those WOE, XxX. PART 6, 39 562 Annals of the South African Museum. near the anterior margin larger; legs almost black, the anterior surfaces of femora a little lighter (especially near the base), remaining segments uniformly black except tarsi which are dark brown. Eyes.—Median eyes weakly recurved, a line touching the upper surfaces of the anteriors passing well above the centres of the posterior medians, anterior medians 4—2 the diameter of the posterior medians, their own diameter apart, and a little less than their diameter from the posterior medians; posterior medians half the diameter of the b c Fic. 6.—Selenops zuluanus n. sp. a, 2, vulva. b,c, tibia of pedipalp, J, from in front, and from outer side. posterior laterals; anterior medians % their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal one distinctly larger than the apical one, separated from it by its basal width. Legs.—Anterior tibiae with 3, anterior metatarsi with 2 inferior pairs of spines. 7 Vulva as in fig. 6, a. Dimensions.—Length of carapace 6-2, width of carapace 7-6, length of abdomen 10 mm. & Colour.—Carapace light yellow brown, without distinct markings; mandibles reddish brown, a little darker than the carapace; abdomen above blackish at the sides and posteriorly, the anterior two-thirds of the middle portion light yellow brown, enclosing a blackish tree-like marking; legs lighter than in the 9, femora olive green below, lighter anteriorly, remaining segments dark brown. Eyes.—Medians forming a weakly curved row, a line touching the upper surfaces of the anteriors passing half-way between the centres The Genus Selenops (Araneae) in South Africa. 563 and upper surfaces of the posterior medians; anterior medians ? the diameter of the posterior medians, a little more than their own diameter apart, and a little more than a radius from the posterior medians; posterior medians half the diameter of the posterior laterals; anterior medians a little more than a diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 subequal teeth separated by 14 times their basal width. Legs.—Tibia I and II with 3 pairs of inferior spines, I with 2 lateral spines on each side and 2 superior spines, II with 2 lateral spines on each side and 1 superior spine; metatarsus I and II with 2 pairs of inferior spines, a lateral spine on each side near the base. Pedipalp.—Femur with 1 inner and 3 superior spines in distal half above, patella with O spines above; tibia distinctly shorter than patella and less than half as long as tarsus when seen from above; tibial process as in fig. 6, b, seen from in front, fig. 6, c, seen from the outer side. Dimensions.—Length of carapace 5-6, width 6-6, total length 12-3 mm. This seems to be a fairly widely distributed form. The South African Museum has examples from Insiza, Pungwe River, and Umtali, all localities in Rhodesia; and Kaapmuiden and Potgietersrust in the Transvaal. The pedipalp resembles that of lesnei Lessert (Rev. Suisse. Zool., vol. xlii, p. 266, fig. 64, 1936) in the external processes of the tibia, but differs in having this segment much shorter than the tarsus, while in lesnei it is subequal to it. The vulva, on the other hand, resembles that of annulatus Simon (see Lessert’s figure, Rev. Suisse. Zool., vol. xxxvi, p. 124, fig. 12, ¢, 1929). Selenops tenebrosus n. sp. (fig. 7). Types, 2 99, Gravelotte, N.E. Transvaal (T.M., 2228). Colour.—Carapace rich dark reddish brown with a conspicuous pattern of markings in addition to the fovea, foveal radiations, and the boundaries of the cephalic area; mandibles reddish black; abdomen almost entirely black above with 1 or 2 pairs of light spots in its anterior half, ventral surface yellow brown, the sides and posterior margin black, spinners almost encircled with black; legs uniformly black, a little lighter towards their bases; coxae, sternum, and mouth- parts reddish brown. Eyes——Median eyes weakly recurved, a line touching the upper surfaces of the anteriors would pass a little above the centres of the 564 Annals of the South African Museum. posterior medians; anterior medians ? the diameter of the posterior medians, more than their own diameter apart (14 times), and % their diameter from the posterior medians; posterior medians half the diameter of the posterior laterals; anterior medians about 2 their diameter from the edge of the clypeus. Cheloicerae.—Inferior margin with 2 large subequal teeth separated by less than their basal width. Fic. 7.—Selenops tenebrosus n. sp. 2, vulva. Legs.—Ill, II, TVs + Th considerably longer than I; tibiae I and II with 3 pairs of inferior spines, metatarsi I and II with _ 2 stout pairs of inferior spines. Vulva as in fig. 7, resembling that of S. zuluanus. Dimensions.—Length of carapace 8-3, width of carapace 9-7, length of abdomen 13-4mm. Total length of second type specimen 23 mm. Other Specomens.—1 9, Olifants River District, Transvaal (T.M., 6364); 1 9, Louis Trichardt, Transvaal (B. 7187, 8.A.M.). Selenops lesner Lessert. S. lesner Lessert, Rev. Suisse. Zool., vol. xli, p. 265, figs. 60, 62-64, 1936. The types, 2 and 3, were described from Mouvia-Sare on the Zambesi (2), and Inhafoune near Canxixe (3), both localities in Portuguese EH. Africa. GROUP B. SUB-GROUP By. Key to the species of Selenops with 4 pairs of wnferror tubsal spines. 29. 1. Anterior median eyes their diameter apart . ; ‘ : natalensis. Anterior median eyes more than their diameter apart . . : ia 2. Vulva] plate not large, circular, without lateral lobes. : lycosiformis. Vulval plate very large, with lateral lobes . : 3 ‘ . parvulus. The Genus Selenops (Araneae) in South Africa. 565. Selenops lycosiformis Lawrence. S. lycosiformis Lawrence, Ann. Natal Mus., vol. viii, pt. 2, p. 246, text-fig. 18, 1937. The type came from the Nkandhla Forest, Zululand, and was based on an adult 2 specimen. Selenops natalensis n. sp. (fig. 8). Types, 2 99, 1 3, Estcourt, Natal (N.M., 1717). 2 Colour.—Carapace blackish brown at the sides, with a median, more or less parallel, yellow marking (as wide as the darkened area on Fic. 8.—Selenops natalensis n. sp. a, 2, vulva. 6, pedipalp, 3, from in front. each side); this lighter marking with crenulated sides and constricted behind the thoracic stria, reaching from the ocular area to the posterior margin and containing a blackish median stripe with a pair of lateral branches, one just in front of the stria and one in the middle of the cephalic area; darkened areas at the sides with a submarginal broken row of lighter spots; mandibles as light coloured as the middle of the carapace, their inner margins and apices darkened, a dark dot on the outer side near the base; abdomen above dark brown, a pair of long oval lighter markings on each side of the anterior half, a pair of smaller oval light spots above the spinners; under surface of abdomen with a considerable part of the sides and posterior portion blackish brown, the spinners surrounded by the darkened area; sternum narrowly bordered with black and with a few blackish dots in the middle. Legs almost uniformly dark brown, the black bands absent 566 Annals of the South African Museum. or barely visible; inferior surfaces of anterior femora black in front and quite light behind, posterior femora spotted black above, their anterior and posterior surfaces black with a fine yellow longitudinal line. Eyes.—Medians not strongly recurved, a line touching the upper surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians 2 the diameter of the posterior medians, their own diameter apart, and 2 their diameter from the posterior medians; posterior medians 2 ithe diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 equal-sized teeth separated by their basal width. Vulva as in fig. 8, a. Legs.—Anterior tibiae with 4 pairs of inferior spines, anterior _metatarsi with 3 inferior pairs, no lateral spines, anterior tibiae unusually short and stout. Dimensions.—Length of carapace 3 3, total length 8 mm. 3 Colour as in 9, anterior femora white below, posterior femora blackish below with a fine longitudinal yellow stripe on anterior and posterior sides. Eyes and chelicerae as in 9. Legs.—Anterior tibiae with 4 pairs of inferior spines and in addition 2 lateral spines on each side and 2 superior spines in the middle of the segment; anterior metatarsi with 3 pairs of inferior spines and 2 lateral spines on each side. Pedipaly.—Femur with 1 inner, 1-2 superior, 1 outer spine in distal half, patella with 1 inner spine; tibia much shorter than tarsus and distinctly shorter than patella, seen from above; tarsus and tibial process as in fig. 8, 6, seen from in front. Dimensions.—Length of carapace 2-7, total length 5-9 mm. Selenops parvulus Pocock. S. parvulus Pocock, A.M.N.H. (7), vol. vi, p. 352, 1900. The type came from Port Elizabeth and was based on an adult 2 specimen. No figure was given of the vulva which, however, seems characteristic, being described as situated on a very large plate and with lateral lobes. The Genus Selenops (Araneae) in South Africa. SUB-GROUP B,. 22. . Teeth on inferior margin of chelicerae their basal width apart 567 Key to the species Selenops with 5 pairs of inferior tibial spines. 2. Teeth on inferior margin of chelicerae more than their basal width apart 10. . Anterior medians a little less than their diameter apart Anterior medians a little more than their diameter apart . Anterior medians half as large as posterior medians Anterior medians more than half as large as posterior medians . Tibiae of legs with 2 dark and 2 light bands Tibiae of legs with 1 dark and 1 light band . Vulva with a median tongue-like selerite Vulva without a median tongue-like selerite . Posterior medians subequal to posterior lateral eyes Posterior medians half the diameter of posterior lateral eyes . Vulva without lateral sclerites . Vulva with distinct lateral sclerites . Lateral sclerites meeting in the middle line Lateral sclerites not meeting in the middle line . Vulva as in fig. 17 Vulva as in fig. 23 . Vulva extremely small, body size an Vulva usually large, body size larger . . Posterior medians half the diameter of posterior ieee Posterior medians subequal to posterior laterals . . Vulva with two incurved arms at its base . Vulva different, much smaller; body size smaller . Vulva with an elongate quadrate opening . Vulva without elongate opening . Anterior medians a radius from posterior medians , Anterior medians less than a radius from posterior medians . . Vulval plate rounded Vulval plate not rounded . Vulva with a median septum Vulva without a median septem . Vulva narrowed anteriorly Vulva not narrowed anteriorly . dd. . Tibia of pedipalp equal or subequal to tarsus in length Tibia of pedipalp much shorter than tarsus . Tibial apophysis of pedipalp with 4 short processes Tibial apophysis of pedipalp with 3 longer processes . Processes of tibial apophysis slender, subequal minor. os 4. 5 barnardi. namaquensis. civicus. 6. caledonicus. wc karrooicus. 8. broomi. 9, hesset. . maculosus. . atomarius. LY. 13. 13. . schonlandi. hewitti. thornet. =, d4: helenae. 15. bechuanicus. 16. longipes. 17. purcelli. . lignicolus. 2: : 4, hewitti. on . atomarius. broomi. Processes of tibial co ana strong and differing wher; in length and thick- ness 568 5 Annals of the South African Museum. 4. Tibia of pedipalp longer than patella ‘ ‘ namaquensis. Tibia of pedipalp sometimes equal to, usually are re patella .. ee 5. Anterior tibiae of legs without lateral spines : ; ; : _ * Anterior tibiae of legs with lateral spines . : : : : : a 6. Body length 6-8 mm. . : : : : : . . alticolus. Body length 5-2 mm. : : 5 : : ; . . minutus. 7. Anterior tibiae with superior spines . ; A ‘ : : : 8. Anterior tibiae without superior spines : 10. 8. The processes of tibial apophysis widely Seek like fons Ae thuiab pococki. The processes of tibial apophysis close together . : d : : o: 9. Pedipalp tarsus seen from in front more or less rounded : . lignicolus. Pedipalp tarsus seen from in front oval (fig. 21, a) ‘ : lesserti. 10. Anterior metatarsi without lateral spines . : ; : . smitherst. Anterior metatarsi with lateral spines : : : é : « pad. 11. Anterior metatarsi with 2 lateral spines on each side . ; bechuanicus. Anterior metatarsi with less than 2 lateral spines on each side. Bees 12. Anterior metatarsi with a lateral spine on each side . : immaculatus. Anterior metatarsi with lateral spines on posterior surface only . ang, eke 13. Processes of tibial apophysis far apart like a forefinger and thumb minor. Process of tibial apophysis close together . ‘ : ; z BAIL 14. Tibial apophysis as in fig. 20, b,c. : ; ; : . karrooicus. Tibial apophysis as in fig. 15, b,c. : : : : : civicus. Selenops alticolus n. sp. (fig. 9). Type, 1 3, Ingwavuma, Lebombo Mts., Zululand (N.M., 2423). Colour.—Carapace light brown with some darker patches along the lateral margins, eyes surrounded by black; mandibles reddish brown, darker than carapace, the inner half of their anterior surfaces blackish brown; abdomen brown, variegated with some minute light dots and blackish markings; legs with very weak bands and spots, the strongest being a longitudinal bar near the base of the infero-anterior surface of femur I and II. Eyes.—Medians not strongly recurved, a line touching the upper surfaces of the anteriors passing through or a little below the centres of the posterior medians; anterior medians not much smaller than posterior medians (? their diameter), their own diameter apart, and a little less than a radius from the posterior medians; posterior medians 2 the diameter of the posterior laterals; anterior medians a radius from the edge of clypeus. Chelicerae.—Inferior margin with 2 moderate subequal teeth, about twice their basal width from each other. Legs.—Anterior tibiae with 5 pairs of rather weak inferior spines, The Genus Selenops (Araneae) in South Africa. 569 no lateral or superior spines; anterior metatarsi with 8 pairs of inferior spines, no lateral spines. Pedipalp.—Femur above with 1 inner (subapical) and 1-2 superior spines in distal half; patella without spines; tibia much shorter than tarsus and shorter than patella, seen from above; tarsus and process a Fic. 9.—Selenops alticolus n. sp. a, pedipalp, 3, from in front. 0, tibia of pedipalp from the side. of tibia as in fig. 9,a, seen from in front, process of tibia from outer side as in fig. 9, b. Dimensions.—Length of carapace 3, total length 6-8 mm. Selenops atomarius Simon (fig. 10). S. atomarius Simon. Bull. Soc. Zool. France, vol. xii, p. 466, 1887. 2 Colour.—The specimens much bleached, legs with fairly distinct dark bands, abdomen yellow, its dorsal surface with chevron markings, the sides with numerous minute blackish dots increasing in number and size posteriorly. | Hyes.—Medians strongly recurved, a line touching the upper surfaces of the anteriors only cutting a small part of the posterior medians; anterior medians about 3 the diameter of the posterior medians, a little more than their own diameter apart, and less than half a radius from the posterior medians; posterior medians a little more than half the diameter of the posterior laterals; anterior medians = their diameter from the edge of clypeus. 570 Annals of the South African Museum. Chelicerae with 2 moderate subequal teeth, twice their basal width from each other. Vulva as in fig. 10, a, resembling that of S. hewitte and remarkable for its extremely small size. Legs.—Anterior tibiae with 5, anterior metatarsi with 3 pairs of inferior spines. Dimensions. — Length of carapace 3-7, total length 9-5 mm. 3d Colour as in the 9, the legs, however, with- out, or with weaker markings, median eyes differing less in size than in the 9, anterior medians a b : Ay ba their own diameter apart. Fic. 10.—Selenops atomarius Simon. a, 9, vulva. ; , b, tibia of pedipalp, g¢, from outer side. Chelocerae. — Inferior margin with 2 teeth, the basal one a little smaller and 2-3 times its width from the apical tooth. Legs.—Anterior tibiae with 5 inferior pairs of spines, II with 1 posterior lateral spine in addition, anterior metatarsi with 3 inferior pairs of spines, O lateral spines. Pedipalp.cFemur with 1 inner, 1-2 superior and O outer spines in its distal half; patella with 1 inner and 1 superior (apical) spine; tibia subequal to tarsus and distinctly longer than patella, with 3 processes on the outer side, two directed forwards and one downwards and forwards, fig. 10, 6. This figure agrees substantially with that of Simon in “ Hist. Nat. des Araignees,”’ vol. 1, p. 27, text-fig. 23, 1897. Dimensions.—Length of carapace 3-7, total length 8 mm. | Specimens.—3 99, 1 3, Port Elizabeth (S.A.M., 4524). Pocock records it from Grahamstown. Selenops barnardi n. sp. (fig. 11). Type, 1 9, Pokwani, South Rhodesia (S.A.M., 14667). Colour.—Carapace fairly dark reddish brown, cephalic a little darker than thoracic portion, ocular area black, radiations from the thoracic stria not strongly marked, a number of submarginal spots subjoined to form a wavy band; mandibles a little darker than cephalic area; abdomen above blackish brown, with some ill-defined symmetrical markings; legs with well-defined and fairly strong bands. The Genus Selenops (Araneae) in South Africa, 571 Eyes.—Medians well recurved, a line touching the upper surfaces of the anteriors passing well below the centres of the posterior medians; anterior medians about half as large as posterior medians, their own diameter apart and less than a radius from the posterior medians; posterior medians % the diameter of the posterior laterals; anterior medians their diameter, or a little less, from the edge of clypeus. Chelicerae.—Inferior margin with 2 equal-sized teeth separated by their basal width. Legs.—Tibia I and II with 5, ~. metatarsus I and II with 3 pairs of inferior spines. Fic. 11.—Selenops barnardi n. sp. Vulva as in fig. 11. uaa Dimensions.—Length of carapace 4:9, width of carapace 5-6, total length 10-3 mm. Selenops bechuanicus n. sp. (fig. 12). Types, 12,1 3, Vryburg, Bechuanaland (S.A.M., 14506). 2 Colour.—Carapace yellow-brown, narrowly bordered with black and prettily variegated with symmetrical brown markings, an oval patch on each side of the thoracic stria without markings; mandibles light yellow-brown, a narrow inner and outer dark stripe on their anterior surfaces; abdomen yellow-brown with indistinct symmetrical markings; legs with distinct dark bands, those of the femora mottled and irregular, those of the tibiae clearly defined. Eyes.—Median eyes moderately recurved, a line touching the upper surfaces of the anteriors passing well below the centres of the posterior medians; anterior medians ? the diameter of the posterior medians, their own diameter apart, and half a radius from the posterior medians; posterior medians 3—3 the diameter of the posterior laterals; anterior medians their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 subequai teeth separated by 14 times their width. Vulva as in fig. 12, a. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-6, length of abdomen 6-5 mm, 572 Annals of the South African Museum. $.—Dark bands of the legs not as well defined asin the 2. Tibiae and metatarsi of anterior legs with 2 lateral spines on each side in addition to the inferior ones. Otherwise as in description of 9. Pedipalp.—Femur above with 1-4 spines in distal half; patella with 1 superior and | inner spine; tibia a little shorter than patella, less than half as long as tarsus, seen from the outer side with process as b c Fic. 12.—Selenops bechuanicus n. sp. a, 2, vulva. 6, c, tibia of pedipalp, J, from in front, and from outer side. in fig. 12, c, seen from in front fig. 12, b. Tarsus seen from in front with style describing an almost complete circle on the outer periphery. Dimensions.—Length of carapace 4:5, total length 9-5 mm. Selenops broomz Pocock (fig. 13). S. broomi Pocock. A.M.N.H. (7), vol. vi, p. 331, 1900. The type was a 9 from Garies, Little Namaqualand, no figure being given of the vulva. 2 Colour.—Carapace reddish brown, the thoracic stria black and strongly defined, continued on to the cephalic portion as a very fine black line, radiations from thoracic stria long, fine, but distinct; mandibles darker than carapace; abdomen light brown, with some short (longitudinal) black bars and spots, a wavy transverse blackish band above the spinners; legs with dark bands, those on the femora poorly defined. especially the posterior ones, those of the tibia well defined. Eyes.—Median eyes not strongly recurved, a line touching the upper surfaces of the anteriors passing between the centres and lower surfaces of the posterior medians; anterior medians 3 the diameter of the posterior medians, a little more than their diameter apart, and The Genus Selenops (Araneae) in South Africa, 573 their radius from the posterior medians; posterior medians half the diameter of the posterior laterals, which are large and prominent; anterior medians } their diameter from the edge of the clypeus. Chelacerae.—Inferior margin with 2 teeth, the basal one distinctly larger than the other and separated from it by its width. Vulva as in fig. 13, a. Legs.—Anterior tibiae with 5, anterior metatarsi with 3 pairs of inferior spines. Dimensions.—Length of carapace 6-6, length of abdomen 8-1 mm. b Fic. 13.—Selenops broomi Pocock. a, 9, vulva. 6, ¢, tibia of pedipalp, g, from outer side, and from in front. $ Colour.—Much lighter than in 9, carapace and mandibles yellow; legs yellow, without markings. Eyes.—Anterior medians half the diameter of posterior medians and their radius or a very little more from the edge of clypeus; otherwise as 1n the 9. Chelicerae.—Inferior margin with 2 equal-sized teeth, 1} times their basal width from each other. Legs.—Anterior tibiae with 5 inferior pairs of spines, 2 lateral spines on each side and 3 superior spines; anterior metatarsi with 3 inferior pairs of spines, I with 2 lateral spines on each side in basal half, II with 2 lateral spines on each side in basal half and 2 superior spines in basal half. Pedipalp.—Femur with 1 inner, 1-2 superior, and 1 outer spine in distal half, patella with 1 inner and 2 superior spines, tibia with 1 inner, 2 superior, and 1 very strong outer spine; tibia subequal to tarsus | but longer than patella, with 3 processes at its outer apex, seen from in front as in fig. 13, c, and from the outer side, fig. 13, b. Dimensions.—Length of carapace 4:8, total length 9-8 mm. Specimens, 499, 2 33, Kleinzee, Little Namaqualand (S8.A.M., 8871). 574 Annals of the South African Museum. Selenops caledonicus n. sp. (fig. 14). Types, 2 99, Caledon, Cape Province (8.A.M., 150421). Colour.—Carapace light ‘yellow-brown with a Y-shaped darker marking formed by the thoracic stria and posterior boundaries of the cephalic area, a few small dark spots half- way between the stria and lateral margins; mandibles similar in colour to the carapace; abdomen above with 3 pairs of short curved blackish bars; well-defined brown bands on femora and tibiae of all legs. Eyes.—Medians moderately recurved, a line touching the upper surfaces of the anteriors passing half-way between the Fic. 14.—Selenops caledoni- centres and lower surfaces of the posterior MISE medians; anterior medians 2 the diameter of the posterior medians, their own dia- meter apart, and a radius from the posterior medians; posterior medians ? the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelocerae.—Inferior margin with 2 strong teeth, the basal a little stouter than the apical one and its own width from it. Vulva as in fig. 14. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-3, total length 7-8 mm. Selenops civicus n. sp. (fig. 15). Type, 1 2, Burghersdorp (8.A.M., B. 76). Colour.—In general resembling that of karrooicus, the markings of the carapace forming a fairly well-defined crenulated submarginal band, sides with a narrow blackish margin and between this and the submarginal band some blackish-brown spots; foveal radiations indistinct, legs with markings as in karrooicus; abdomen above with some indistinct brown markings on the posterior half, including some indistinct A-shaped markings. Eyes.—Median eyes distinctly recurved, a line touching the upper boundaries of the anteriors passing well below the centres of the posterior medians; anterior medians a little more than their diameter apart, their radius from the posterior medians and two-thirds as large as these; posterior medians less than half the diameter of the posterior The Genus Selenops (Araneae) in South Africa. 575 laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 subequal teeth separated by a distance equal to their width. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. b Fic. 15.—Selenops civicus n. sp. a, 9, vulva. 6, c, tibia of pedipalp, J, from in front, and from outer side. Vulva as in fig. 15, a. Dimensions.—Length of carapace 4-8, width of carapace 5-8, total length 12-5 mm. Other Specimens.—2 992, 4 juveniles, and 2 g¢ from Smithfield, O.F.S. (S.A.M., B. 465). The following is a description of one of the two males from Smithfield :— 3 Colour.—Much lighter than in the 9, the markings of the legs almost absent; teeth of inferior margin of chelicerae separated by twice the basal width of the apical tooth. Legs.—Tibia I and II with 5 pairs of inferior spines; I with 1 anterior lateral, 2 posterior lateral, and 1 superior spine in addition; II with 2 anterior lateral, 2 posterior lateral, 2 superior spines in addition; metatarsus I and II with 3 pairs of inferior spines and 1 posterior lateral spine near the base. Pedipalp.—Femur with 1 inner, 1-2 superior, and 1 outer spine in distal half, the inner and outer spines much weaker than the others; tibia subequal to patella and much shorter than tarsus, as in fig. 15, b, seen from in front, fig. 15,c, seen from outer side; 3 long spines on its inner side. Dimensions. —Length of carapace 4-2, width 4-5, total length 7-5 mm. 576 Annals of the South African Museum. Selenops helenae n. sp. (fig. 16). Type, 1 9, St. Helena Bay, Cape Province (S.A.M., 11717). Colour.—Carapace light reddish brown, ornamented with minute black specks and larger blackish spots, a tuft of white hairs above and overhanging the anterior median eyes; mandibles the same colour as the carapace; abdomen above brown with some rather ill-defined sym- metrical darker markings; anterior legs with brown bands, posterior ones without. Eyes.—Median eyes moderately recurved, a line touching the upper . surfaces of the anteriors passing well Fic. 16.—Selenops helenae n. sp. below the centres of the posterior ) WELT: medians; anterior medians half, or a little more, the diameter of the posterior medians, a little more than their own diameter apart and a radius from the posterior medians; posterior medians subequal to posterior laterals; anterior medians about ? their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 strong subequal teeth separated by twice their width. Vulva asin fig. 16. Legs.—Tibiae I and II “ue 5, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-5, length of abdomen 4 mm. Other Specimens.—2 juveniles, Stompneus, St. Helena Bay (S.A.M., 11677), and 1 adult 2 from the same locality (S8.A.M., 11666). Selenops hesser n. sp. (fig. 17). Type, 1 9, Matjesfontein, Cape Province (S.A.M., 1665). Colour.—Carapace reddish brown with a fine blackish marginal border, cephalic portion darker than the thoracic portion; thoracic stria and its radiations distinct, the remaining markings somewhat vague; chelicerae dark reddish brown, abdomen rubbed but apparently with a number of blackish-brown spots near the spinners; femora of legs with ill-defined banded markings, those of the tibiae more distinct, metatarsi dark brown. Eyes.—Median eyes distinctly recurved, a line touching the upper The Genus Selenops (Araneae) in South Africa. 577 margin of the anteriors passing well below the centres of the posteriors; anterior medians a little more than half the diameter of the posterior medians, more than their own diameter apart and aboutaradius from the posterior medians; posterior medians half the dia- meter of the posterior laterals; anterior medians a little less than a diameter from the edge of the clypeus. Chelicerae. — Inferior margin . Fic. 17.—Selenops hessei n. sp. with 2 subequal teeth separated ©: valves: by the width of the basal tooth. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. Vulva as in fig. 17. Dimensions.—Length of carapace 5-5, width of carapace 6-4, length of abdomen 8-3 mm. Other Specimens.—3 22, Matjesfontein (S.A.M., 13119); 1 2, Prince | Albert (S.A.M., 3919). Selenops hewitts n. sp. (fig. 18). Types, 3 292, Grahamstown (Albany Museum). 2 Colour.—Carapace light brown with a narrow black margin; stria and some fine radiations from it black, some spots close to the lateral margins black, a pair of short anteriorly diverging black bars behind the posterior median eyes; mandibles light reddish brown; abdomen above with two transverse black recurved markings in its posterior half, area above the spinners blackish; legs with strong black bars, especially on the femur and base of tibia, these much fainter in the posterior legs. Eyes.—Anterior row strongly recurved, a line joining the upper surfaces of the anterior medians passing well below the centres of the posterior medians; anterior medians $ the size of the posterior medians, 3 their own diameter apart and about half a radius from the posteriors; posterior medians half as large as posterior laterals; anterior medians about their radius from the edge of the clypeus. Chelicerae.—Inferior margin with 2 large subequal teeth, 14 times their width from each other. Vulva as in fig. 18, a. VOL. XXXi1, PART 6, 40 578 Annals of the South African Museum. Legs.—Tibia I and II with 5, metatarsi I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-2, total length 9-4 mm. Other Specomens.—1 3, 2 99, Grahamstown (B. 8274, S8.A.M.); 1 2, Grahamstown (S8.A.M., 5755). The Transvaal Museum has 2 99 from the same locality (8280). The following is a description of an adult 3 from Grahamstown (5755). 3 Colour.—Carapace (bleached) light yellow-brown, thoracic stria long and narrow, continued on to the cephalic area as a fine median Fic. 18.—Selenops hewitti n. sp. a, 2, vulva. 0, c, tibia of pedipalp, 3, from in front, and from outer side. line, a few faint radiations from the stria; mandibles a little darker than carapace; abdomen above with some ill-defined symmetrical markings, legs apparently without spots or bands, light yellow. Eyes.—Median eyes fairly strongly recurved, a line touching the upper surfaces of the anteriors passing half-way between the lower surfaces and centres of the posterior medians; anterior medians subequal to posterior medians (more than 3 their diameter), their own diameter apart and almost touching the posterior medians (less than half a radius from them); posterior medians a little more than half the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelicerae—Inferior margin with 2 equal-sized teeth distinctly smaller than those of the superior margin, 14-2 their width from each other. Legs.—Tibia I and II with 5 pairs of inferior spines and 2 lateral spines on each side, metatarsus I and II with 3 pairs of inferior spines and without lateral spines. The Genus Selenops (Araneae) in South Africa. 579 Pedipalp.Femur above with 1 inner, 3 superior, 0 outer spines; patella with 1 inner basal, 1 apical and 1 basal superior spines; tibia long and parallel-sided, subequal to or a little longer than tarsus, with process as in fig. 18, 6, seen from in front, and fig. 18, c, from outer side. Dimensions.—Length of carapace 3-8, total length 8 mm. Selenops ammaculatus n. sp. (fig. 19). Type, 1 3, Florida, Transvaal (S.A.M., B. 4505). Colour.—Carapace yellow-brown, thoracic stria and some very faint radiations from it a little darker, eyes surrounded by a blackened area; mandibles yellow- brown; abdomen yellow with a few blackish brown spots; legs without mark- ings except for two very faint brown bands on the anterior tibiae. Eyes.—Medians moder- ately recurved, a_ line touching the upper sur- faces of the anteriors Fic. 19.—Selenops immaculatus n. sp. a, b, tibia of pedipalp, g, from in front, and from outer side. a b passing a little below the © centres of the posterior | medians; anterior medians 2 the diameter of the posterior medians, a little less than their own diameter apart, and less than a radius from the posterior medians; posterior medians 3 the diameter of the posterior laterals; anterior medians about 2 their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 subequal teeth separated by 14 times their width. Legs.—Tibia I and II with 5 pairs of inferior spines and 2 lateral spines on each side, metatarsus I and II with 3 inferior pairs and 1 lateral spine on each side. Pedipalp.—Femur with 1 outer, 3 superior, and 1 inner spine in apical half; patella with 1 outer and 1 inner spine near its base above, _and 1 superior apical spine; process of tibia as in fig. 19, a, seen from in front, and fig. 19, b, seen from the outer side. Tibia subequal to patella, less than half the length of tarsus. Dimensions.—Length of carapace 3-9, total length 8-5 mm. 580 Annals of the South African Museum. Selenops karroowcus n. sp. (fig. 20). Types, 7 99, 4 $$, Hanover, Cape Province (S.A.M., 9499). 2 Colour.—Carapace light reddish brown, sides narrowly bordered with black, ocular area blackish, thoracic portion with irregular d Fic. 20.—Selenops karrooicus n. sp. a, 2, vulva. 6, pedipalp, g. c, d, tibia of pedipalp from outer side, and from behind. stripes and spots, the radii from the thoracic stria blackish but in- complete; cephalic portion bordered by narrow blackish lines and bisected by a narrow line; mandibles reddish brown; abdomen dull white or yellow dorsally, with symmetrical light-brown markings in the middle line, a darker —“~-shaped transverse marking just above the spinners. Legs yellow, strongly banded with black except on their ventral surfaces; femora with 3 black bands, patellae with 1 basal band, tibiae with 2 bands. Eyes.—Medians well recurved, a line touching the upper surfaces of the anteriors passing well below the centres of the posterior medians; anterior medians 2 the diameter of the posterior medians, a little more than their own diameter apart and a little less than their radius from the posterior medians; posterior medians half or a little less than half the diameter of posterior laterals; anterior medians 2 their diameter from the edge of the clypeus. ‘Chelicerae.—Inferior margin with 2 equal-sized teeth separated by their basal width. | Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. The Genus Selenops (Araneae) in South Africa. 581 _ Abdomen large, subquadrate, truncate anterior and posteriorly, wider behind than in front. Vulva as in fig. 20, a. Dimensions.—Length of carapace 5-6, width 6-6, total length 17-4 mm. 3 Colour much lighter than in 9, legs without strong black bands. Inferior margin of chelicerae with 2 teeth, the apical one a little larger, 1} times its width from the basal tooth. Eyes as in 9. Legs.—Tibia I and II with 5 inferior pairs of spines and 2 lateral spines on their posterior surfaces (occasionally also on the anterior surfaces); sometimes in addition 1 or 2 superior spines; anterior metatarsi with or without 1—2.lateral spines on their posterior surfaces and 3 pairs of inferior spines. Pedipalp.—Femur with 1 inner, 1-2 superior, but no outer spine in apical half; tibia subequal to patella, much shorter than tarsus, with 2 long spines on its inner surface; apophysis of tibia bilobed as in fig. 20, b, seen from in front, fig. 20, c, seen from the outer side, and fig. 20, d, seen from behind. Dimensions.—Length of carapace 4:6, width 5-2, total length 9 mm. Other Specumens.—From the same locality, 1 2, 1 3 (S.A.M., 11945); 2 3 (S.A.M., 11875); 1 9 (S.A.M., 11853); from Naauwpoort, 3 36, 5 99 (B. 1594, 8.A.M.). Selenops lesserti n. sp. (fig. 21). Type, 1 3, Touws River, Worcester, Cape Province (S.A.M., B. 7541). Colour.—Carapace orange-yellow with the thoracic stria and its radiations a little darker; mandibles much darker than carapace, reddish brown; abdomen light yellow-brown above, with minute scattered black dots and a wavy transverse black stripe above the spinners; legs apparently without markings or bands of any kind. Eyes,—Medians not strongly recurved, a line touching the upper surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians about ? the diameter of the posterior medians, a little less than their own diameter apart, and about half a radius from the posterior medians; posterior medians half or even less the diameter of the posterior laterals; anterior medians a little more than their radius from the edge of clypeus. Chelicerae.—Inferior margin with 2 strong equal-sized teeth separ- ated by 1-14 times their basal width. Legs.—Tibia I and II with 5 pairs of extremely long inferior spines, 582 Annals of the South African Museum. 2 lateral spines on each side, and 3 superior spines in proximal half; metatarsus I and II with 3 pairs of long inferior spines and 2 lateral spines on each side in proximal half. Pedipalp.—Femur with 1 inner, 3 superior, 1 outer spines in distal half; patella without spines; tibia shorter than either patella or tarsus, with processes as in fig. 21, a, seen from in front, and fig. 21, b, seen from the outer side. Pedipalp unusu- ally large for a species of Group B. Dimensions. — Length of carapace 5-2, total length 10-2 mm. Selenops lignicolus Lawrence. S. lignicolus Lawrence. Ann. Natal Mus:, vol) vu, pt. 2, p. 239; tsa 1937. The types were described from Hluhluwe, Zululand, Fic. 21.—Selenops lesserti n. sp. a, pedipalp, and were based on 2 0° and 6, from in front. 6, tibia of pedipalp from outer side. a gd specimen. a Selenops longipes n. sp. (fig. 22). Types, 2 99, 1immature 3, Johannesburg (S.A.M., 150495). Colour.—Specimens somewhat bleached. Carapace light yellow- brown with indistinct spots and mottling, boundaries of cephalic area and stria a little darker, areas surrounding the eyes blackish, mandibles not darker than carapace; abdomen above with indistinct spots and mottling in the anterior half, two black bow-shaped trans- verse bars in posterior half; legs with indistinct blackish brown bands, those at the bases of the anterior tibiae and femora well defined. Eyes.—Medians forming a weakly recurved row, a line joining the upper surfaces of the anteriors passing through or just below the centres of the posterior medians; anterior medians 2 the diameter of the posterior medians, their own diameter apart, and a little less than a radius from the posterior medians; posterior medians about ? the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of the clypeus. The Genus Selenops (Araneae) in South Africa. 583 Chelicerae.—Inferior margin with 2 moderate subequal teeth (smaller than those of the superior margin), separated by twice their width. Vulva as in fig. 22. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 4-7, total length 9-5 mm. Other Speciomens.—1 9, Johannesburg (8.A.M., 4206). Fig. 22.—Selenops longipes n. sp. Fig. 23.—Selenops maculosus n. sp. Q, vulva. Q, vulva. Selenops maculosus, n. sp. (fig. 23). Types, 2 99, Willowmore District (S.A.M., 12931). Colour.—Carapace with well-defined radiations from the thoracic stria and crenulated submarginal bands, a well-defined but narrow blackish marginal marking; mandibles reddish brown with some darker markings; abdomen above dirty yellow, with some ill-defined brownish spots and a number of minute dots scattered over its surface; legs with fairly well-defined bands. Eyes.—Median eyes well recurved, a line touching the upper surfaces of the anteriors passing half-way between the centres and the lower surfaces of the posteriors; anterior medians # the size of the posterior medians, a little more than their own diameter apart, and less than their radius from the posterior medians; posterior medians half the diameter of posterior laterals; anterior medians about a radius from the edge of the clypeus. Chelicerae.—Inferior margin with 2 large subequal teeth separated by a little less than their basal width. 584 _ Annals of the South African Museum. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. Vulva as in fig. 23. Dimensions.—Length of carapace 4-7, width 5-3, length of abdomen 6-4 mm. Other Specimens.—4 99, Beaufort West (B. 1634, B. 1930, S.A.M.). Selenops minor un. sp. (fig. 24). Types, 12,1 3, Empangeni, Zululand (T.M., 6389). 2 Colour.—Carapace brown with a very broad blackish-brown marginal band, crenulated along its inner margin and including some a b c Fic. 24.—Selenops minor n. sp. a, 9, vulva. 6, c, tibia of pedipalp, 3, from in front, and from outer side. light brown spots; the lighter inner portion of the carapace without distinct radiations from the thoracic stria, a V-shaped blackish marking behind the eyes, the cephalic portion with a narrow blackish margin; mandibles blackish brown; abdomen blackish, variegated with a few symmetrically arranged lighter spots, a large pair just anterior to the posterior margin, some minute black spots at the sides of the under surface; femora of legs with only one complete well- defined band in the middle, the other two bands represented by blotches and spots. Eyes.—Median eyes recurved but not strongly so, a line touching the upper surfaces of the anteriors passing below the centres of the posterior medians; anterior medians 3 the posterior medians, a little less than their own diameter apart, and a radius from the posterior medians; posterior medians about ? the diameter of the posterior laterals; anterior medians about a radius from the edge of the clypeus. Chelicerae.—Inferior margin with 2 strong subequal teeth separated by their basal width. The Genus Selenops (Araneae) in South Africa. 585 Legs.—Tibia I and II with 5 pairs of inferior spines, metatarsus I and IJ with 3 pairs of inferior spines. Vulva as in fig. 24, a. Dimensions.—Length of carapace 3-1, width 3-4, length of abdomen 4-2 mm. $ Colour as in the 2 but a little lighter, legs without black bands; tibia I and II with 5 pairs of inferior spines and 2 lateral spines on their posterior surfaces; metatarsus I and II with 3 pairs of inferior spines and no lateral spines. Pedipalp.—Femur with 1 outer, 1-2 superior, and 1 inner spine in apical half; patella with 1 or 2 spines; tibia shorter than patella and about half the length of tarsus, with 4 superior spines; tibial apophysis as in fig. 24, b, seen from in front, as in fig. 24, c, seen from the outer side. Dimensions.—Total length 5-2 mm. Selenops minutus n. sp. (fig. 25). Type, 1 3, Grahamstown (8.A.M., B. 2409). Colour.—Carapace reddish brown, with some confused submarginal markings, thoracic stria continued as a V-shaped marking, cephalic area divided by a very faint narrow blackish stripe, eyes surrounded by blackened areas; mandibles reddish brown; abdomen light brown, a broad, sharply defined, procurved band above the spinners, this transverse band whitish and contrasting strongly with the remainder of abdomen; legs with some ,. 5B gee gia confused mottlings but no definite bands. n. sp. Tibia of pedipalp, Eyes.—Medians not strongly recurved, a a = i. Atk Lt line touching the upper surfaces of the anteriors passing well below the centres of the posterior medians; anterior medians 14 times their own diameter apart and a radius from the posterior medians; posterior medians 2 the diameter of the posterior laterals; anterior medians a little less than a diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 moderate equal-sized teeth, 2-3 times their width from each other. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines; no lateral spines. Pedipalp.—Femur with | inner apical and 3 superior spines, patella without spines; tibia much shorter than tarsus but subequal to the 586 Annals of the South African Museum. patella, external process seen from in front as in fig. 25, b, from outer side as in fig. 25, a. Dimensions.—Length of carapace 2:3, total length 5-2 mm. Selenops namaquensis n. sp. (fig. 26). Types, 1 9,1 3, Lekkersing, Little Namaqualand (S.A.M., B. 8493). 2 Colour.—Carapace light reddish brown, the cuneiform thoracic stria with fine long black radiations from 11, a short black bar behind Fic. 26.—Selenops namaquensis n. sp. a, 2, vulva. 5, ¢, tibia of pedipalp, 3, from outer side, and from in front. each posterior lateral eye, an indistinct row of brown submarginal dots; mandibles a little darker than carapace; abdomen above light brown, darker posteriorly, with minute scattered black dots and a few large symmetrical spots above the spinners; legs differing from all the other species of this group in having one instead of two black bands on the tibiae, occupying the basal two-thirds of the segment, femora with vague brownish blotches; sides of patellae black, the remainder brown; anterior metatarsi black, the posterior ones lighter. Eyes.—Medians weakly recurved, a line touching the upper surfaces of the anteriors passing through or a little below the centres of the posterior medians; anterior medians half the diameter of the posterior medians, a little more than their own diameter apart, and a radius from the posterior medians; posterior medians half or even a little less the diameter of the posterior laterals; anterior medians # their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 large equal-sized teeth, their basal width from each other. Vulva as in fig. 26, a. Legs.—Anterior tibiae with 5, anterior metatarsi with 3 inferior pairs of spines. The Genus Selenops (Araneae) in South Africa. 587 Dimensions.—Length of carapace 4-8, total length 11 mm. 3S Colour in general yellow; mandibles dark reddish brown, con- trasting strongly with the carapace; abdomen above with minute scattered black dots; legs yellow, anterior metatarsi distinctly darker than remaining segments. Eyes as in 9, the anterior medians a little less than their diameter apart, and less than their radius from the posterior medians; posterior medians less than half the diameter of the posterior laterals. Chelicerae.—Inferior margin with 2 equal-sized teeth, 2-23 times their basal width from each other. Legs.—Anterior tibiae with 5 pairs of inferior spines, 2 lateral spines on each side, and 3 superior spines; anterior metatarsi with 3 inferior pairs of spines, and 2 lateral spines on each side. Pedipalp.—Femur above with 1 inner, 1-2 superior, and 1 outer spine in apical half; patella with 3-4 long setae; tibia with 4 long spines, 1 outer and 3 inner; tibia longer than patella but distinctly shorter than tarsus, with the external process as in fig. 26, c, seen from in front, and fig. 26, 6, seen from outer side. Dimensions.—Length of carapace 4:8 mm., total length 9-2 mm. Selenops pococki n. sp. (fig. 27). Type, 1 3, Lydenburg, Transvaal (T.M., 2372). Colour.—Carapace dark brown with numerous darker spots and radiations, ocular area black; mandibles a little darker than carapace; abdomen yellow’ with blackish-brown symme- trical markings; legs with weak, ill-defined, and confused markings, those of the femora strongest, those of the posterior legs almost obsolete. a b Eyes.—Median eyes not Fic. 27.—Selenops pococki n. sp. Tibia of pedi- palp, g. a, from in front, and b, from outer side. strongly recurved, a line touching the upper surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians a little more than half the diameter of the posterior medians, their own diameter apart, and about half a radius from the posterior medians; posterior medians a little more than half the diameter of the posterior laterals; anterior medians 2 their diameter from the edge of the clypeus. 588 Annals of the South African Museum. Chelicerae.—Inferior margin with 2 teeth, the basal one a little larger, 14 times its width from the other tooth. Legs.—Tibia I and II with 5 pairs of inferior spines, 2 lateral spines on each side, and 3 superior spines; metatarsus I and II with 3 pairs of inferior spines and 2 lateral spines on each side. Pedipalp.—Femur with 1 inner, 1-2 superior, and 1 outer spine in distal half, patella and tibia with long spine-like setae; tibia dis- tinctly shorter than patella and much shorter than tarsus, with its external apophysis as in fig. 27, a, seen from in front, fig. 27, b, seen from outer side. Dimensions.—Length of carapace 4-2, total length 9-4 mm. This species closely resembles S. minor, from Empangeni, Zululand, in the tibial apophysis of the $ pedipalp. Selenops purcelli n. sp. (fig. 28). Type, 1 2, Montagu Baths, Cape Province (S.A.M., 12675). Colour.—Specimen probably rather faded. Carapace yellow- brown with a blackish marginal border, thoracic stria brown, a number of short blackish stripes near the lateral margins directed towards, but not connected with the stria, cephalic portion defined posteriorly by a blackish V-shaped marking; from posterior apex of cephalic portion to just behind the anterior median eyes a very fine black stripe bisecting the ae a ro carapace; mandible with 2 or 3 longi- “n.sp. 9, vulva. tudinal blackish stripes; abdomen blackish brown above, rather rubbed, but with a clearly defined broad V-shaped marking above the spinners; legs with large black blotches on the antero-inferior surfaces of the anterior femora, and wide brown bands on the tibia, these markings becoming fainter in the posterior legs. Eyes.—Medians strongly recurved, a line joining the upper surface of the anterior medians passing a little above the lower surfaces of the posterior medians; anterior medians 3 as large as posterior medians, 14 times their own diameter apart, and a little less than their radius from the posterior medians; posterior medians equal to the posterior laterals in size; anterior medians a little less than their diameter from the edge of the clypeus. The Genus Selenops (Araneae) in South Africa. 589 Chelicerae.—Inferior margin with 2 rather small subequal teeth, separated by about 14 times their width. Vulva as in fig. 28. Legs.—Tibia I and II with 5, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-1, total length 8 mm. Selenops schonlandi Pocock (fig. 29). Saeschonland: Pocock. A.M.N.H., ser. 7, vol. x, p. 22, pl. im, fig. 15, 1902. The 9 type was from Jansenville, Cape Province. © Colour.—Carapace reddish brown with a narrow black margin, cephalic portion a little darker than thoracic portion, a crenulated submarginal marking on _ thoracic portion blending anteriorly with the darker cephalic portion; thoracic stria cuneiform and dark, with fine blackish radiations, cephalic portion bisected by a fine black line, a branch from the middle of this line running to the base of each posterior median eye; mandibles as dark as the Fic. 29.—Selenops schonlandi cephalic area; abdomen above thickly acne years covered with blackish-brown symmetrical blotches and markings, the sides more dotted, a broad transverse bow-shaped light mark- ing above the spinners, separated from them by a black wavy transverse bar; legs with strong black bands, as strong on the posterior as on the anterior legs, the bands on the lower surfaces of femora tending to coalesce, forming a continuous stripe in femur IV. Eyes.—Median eyes not strongly recurved, a line touching the upper surfaces of the anterior passing a little below the centres of the posterior medians; anterior medians a little more than half the ' diameter of the posterior medians, a little less than their own dia- meter apart, and more than half a radius from the posterior medians; posterior medians half the diameter of posterior laterals which are very large and prominent; anterior medians 3 their diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 large equal-sized teeth, twice their basal width from each other. 590 Annals of the South African Museum. Vulva as in fig. 29. Pocock’s figure apparently does not represent the colouring correctly. Legs.—Anterior tibiae with 5, anterior metatarsi with 3 pairs of inferior spines. Dimensions.—Length of carapace 4:2, total length 12-8 mm.; a smaller adult specimen with total length 10-4 mm. Specomens.—3 99, Graaff Reinet, Cape Province (8.A.M., 12573). Pocock also records it from Pearston and Graaff Reinet. Selenops smithersi n. sp. (fig. 30). Type, 1 3, Morigate, Basutoland (S.A.M., 14424). Colour.—Carapace yellow, variegated with brown markings, a narrow black marginal border; mandibles a little darker than the carapace; abdomen above mottled brown, ornamented with numerous blackish blotches and dots; legs with faint irregular markings and bands. Eyes.—Median eyes not strongly recurved, a line touching the upper surfaces of the anteriors passing panes just below the centres of the pos- Fic. 30.—Selenops smithersi n. sp. : : : : Tibia of pedipalp, g. a, from in terior medians; anterior medians front, and 6, from outer side. 2 the diameter of the posterior medians, their own diameter or slightly less apart, and half a radius from the posterior medians; posterior medians half the diameter of the posterior laterals; anterior medians 2 their diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal one a little smaller, separated by 2-24 times the width of the apical tooth. Legs.—Tibia I and II with 5 pairs of inferior spines and 1 lateral outer spine near the base, metatarsus I and II with 3 pairs of inferior spines but without lateral spines. Pedipalp.—Femur above with 1 inner, 3 superior, 1 outer spines in its distal half; patella with 1 inner and 1 superior apical spine; tibia with some long spines on the inner side near its base; tibia distinctly shorter than patella or tarsus, as in fig. 30, a, seen from in front, and fig. 30, 6, seen from the outer side. Dimensions.—Length of carapace 3-2, total length 6 mm. a b = The Genus Selenops (Araneae) in South Africa. 591 Selenops thorner un. sp. (fig. 31). Type, 1 2, Cedarberg Mts., Cape Province (S.A.M., 5867). Colour.—Carapace light yellow-brown, a broad margin on each side a little darker, thoracic stria and boundaries of cephalic area clearly defined, the latter bisected by a fine dark line; mandibles infuscated near their apices on anterior surface; abdomen infuscated above, becoming darker posteriorly, a narrow median light stripe in anterior two thirds, and two small, well-defined, circular, whitish patches above the spinners, each containing a minute brown dot; legs without bands but almost the entire anterior surface of femur I (less so in remaining femora) blackish Fic. 31.—Selenops thornei brown. n.sp. Q, vulva. Eyes.—Median eyes moderately recurved, a line touching the lower surfaces of the anteriors passing below the centres of the posterior medians; anterior medians 2 the diameter of posterior medians, their own diameter apart, and less aie a radius from the posterior medians; posterior medians fully equal to the posterior laterals in size; anterior medians a diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal distinctly larger than the apical one, separated by twice the width of the basal tooth. Vulva as in fig. 31. Legs.—Tibia I and II with 5 pairs of weak inferior spines, metatarsus T and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 2-4, total length 7-3 mm. SUB-GROUP Bs. Key to the species of Selenops with 6 pairs of inferior tibial spines. 92. 1. Vulva with distinct lateral sclerites divided by a median septum . Oe Vulva without lateral sclerites, usually a large plate with rounded opening 6. 2. Vulva with a broad median septum . ; : : : - kraussi. Vulva with a narrow median septum ; é : : : a. 3. Anterior median eyes more than their diameter apes Z . braunsi. Anterior median eyes less than their diameter apart . , ; ee Ae 4. Lateral sclerites almost touching in middle line . : : - montanus. Lateral sclerites further apart .. ; : 3 : : : SB! 11 592 Annals of the South African Museum. 5. Anterior medians subequal to posterior medians . ' : - capensis, Anterior medians half the diameter of posterior medians. sexspinatus, 6. Teeth of inferior margin of chelicerae 4 times their width from each other regalis. Teeth of inferior margin of chelicerae nearer to each other . : AAC 7. Vulval plate not longer than wide and not subquadrate ‘ : ; 8. Vulval plate longer than wide, subquadrate ‘ ; A ; : 2: 8. Inferior teeth of chelicerae their width apart ‘ é : . stauntoni. Inferior teeth of chelicerae 3 times their width apart . ; . amatolae. 9. Vulval plate a little longer than wide P ; E ; : ae 10; Vulval plate considerably longer than wide ; . ; ‘ sy ML 10. Vulva with a subquadrate anterior opening ; ; ; . tuckerv. Vulva without such an opening : : : : ‘ . reservatus. 11. Opening of vulva rounded : : : : é . barbertonensis. Opening of vulva not rounded . : ‘ : : . __ transvaalicus. dd. 1. Femur of pedipalp not longer than patella+tibia, with a patch of spines on its inferior surface : ‘ ; : ; : ; . stridulans. Femur of pedipalp longer than patella tibia, without a patch of spines on its inferior surface . : : ; : , ‘ : , : 2. 2, Anterior tibiae with 2 lateral spines on each side ‘ ; : Sie merch Anterior tibiae without lateral spines on each side , : : i a Ae 3. Anterior metatarsi with lateral spines , : : : fitezsimonsi. Anterior metatarsi without lateral spines . : : : . stauntoni. 4. Tibia of pedipalp less than half length of tarsus . : : : - gullr. Tibia of pedipalp more than half length of tarsus : , . capensis. Selenops amatolae n. sp. (fig. 32). Type, 1 9, Hogsback, Amatola Mts., Cape Province (S.A.M., 2416). Colour.—Carapace finely bordered with black, dark brown with long blackish radiations from the thoracic stria, the latter appearing as a wedge-shaped dark marking pointed posteriorly; mandibles similar in colour to carapace; abdomen above rubbed; femora and tibiae of anterior legs with broad brown bands. Eyes.—Medians moderately recurved, a line touching the upper surfaces of the anteriors would cut a small portion of the posterior medians; anterior medians 2 the diameter of posterior medians, a little more than their own diameter apart, and a full radius from the posterior medians; the oval anterior laterals unusually large and conspicuous; posterior medians subequal to the posterior laterals; anterior medians less than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 moderate subequal teeth separated by almost 3 times their width. The Genus Selenops (Araneae) in South Africa. 593 Vulva as in fig, 32. Legs.—Tibia I and II with 6 pairs of strong and very long spines (the longest about a third the length of tibia I), metatarsus I and II with 3 pairs of even longer spines (the longest about half the length of metatarsus I). Fig. 32.—Selenops amatolae n. sp. Fic. 33.—Selenops barber- 9, vulva. tonensis n. sp. &, vulva. Selenops barbertonensis n. sp. (fig. 33). Type, 1 2, Barberton, Transvaal (S.A.M., 4194). Colour.—Carapace light brown with spots and dashes of brown, thoracic radiations faint, thoracic stria and sides of cephalic area defined by brown stripes, cephalic area bisected by a faint median stripe, which at half-way gives off a branch to each posterior median eye; mandibles light brown, their inner apices darkened; abdomen light brown, finely speckled with some larger indistinct bars and spots; legs in general with irregular blotches and dots, femora with an inferior fuscous longitudinal stripe, anterior tibiae with well-defined blackish bands. Eyes.—Medians fairly strongly recurved, a line touching the upper surfaces of the anterior medians passing about half-way between the centres and the lower surfaces of the posterior medians; anterior medians about half the diameter of the posterior medians, their diameter apart, and less than a radius from the posterior medians; posterior medians 2 the diameter of the posterior laterals; anterior medians a little less than a diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 equal-sized teeth separated by 13-2 their width. — Vulva as in fig. 33. VOL. XXXIt,; PART 6, 4] 594 Annals of the South African Museum. Legs.—Tibia I and II with 6, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-9, total length 8-4 mm. Other Specimens.—2 9°, between Badplaats and Barberton, Trans- vaal (N.M., 2468). Selenops braunsi n. sp. (fig. 34). Type, 1 2, Willowmore, Cape Province (8.A.M., 12911, Coll. H. Brauns). Colour.—Carapace light reddish brown, thoracic stria well defined, with some fine lines radiating from it, a line from the stria bisecting the cephalic area and bifurcating behind the anterior median eyes; mandibles a little darker than carapace; abdomen rubbed, yellow above but with some darker spots and blotches above the spinners; legs with the dark bands very faint, almost invisible. Eyes.—Medians moderately re- curved, a line touching the upper surfaces of the anteriors passing Fic. 34.—Selenops brawnsi n. sp. half-way between the centres and 2, vulva. the lower surfaces of the posterior medians; anterior medians half the diameter of the posterior medians, a little more than their own diameter apart, and a radius from the posterior medians; posterior medians § the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 subequal teeth (smaller than those of the superior margin), separated by 3 times their width. Vulva as in fig. 34. | Legs.—Tibia I and II with 6, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 5:3, total length 12 mm. Selenops capensis n. sp. (fig. 35). Types, 19,1 3, Cape Town (S.A.M., 8970). 2 Colour.—Carapace yellow-brown with darker radi from the thoracic stria, cephalic portion a little darker than thoracic portion; mandibles reddish brown; abdomen above yellow-brown with small The Genus Selenops (Araneae) in South Africa. 595 scattered blackish dots, darker towards the posterior apex, especially at the sides and just above the spinners; legs not strongly banded, tibiae with 2 light and 2 dark bands. Eyes.—Medians fairly strongly recurved, a line touching the upper surfaces of the anteriors passing through the centres of the posterior Fic. 35.—Selenops capensis n. sp. a, normal, and 6, abnormal, form of vulva, 9. c, d, tibia of pedipalp, 3, from in front, and from outer side. medians; anterior medians a little smaller than posterior medians, a little less than their own diameter apart, their radius from the posterior medians; posterior medians half the diameter of the posterior laterals; anterior medians their radius or a little more from the edge of clypeus. Chelicerae.—Inferior margin with 2 teeth separated by more than the width of the apical tooth, the basal a little smaller than the apical tooth. Legs.—Tibia I and II with 6 pairs of inferior spines, metatarsus I and II with 3 pairs of inferior spines. Vulva as in fig. 35,a. This is the usual type of structure found in this species, while fig. 35, b, represents an abnormal form from the same locality (Signal Hill, Cape Town). 596 Annals of the South African Museum. Dimensions.—Length of carapace 6, width 6-6, length of abdomen 7-8 mm. 3 Colour and eyes as in the description of the 9; chelicerae with the two teeth of the inferior margin distant 3-4 times the width of the apical tooth from each other. Tibia and metatarsus of anterior legs without lateral spines. Pedipalp.—Tibia about 3 the length of tarsus and subequal to patella, seen from in front as in fig. 35, c, seen from outer side as in fig. 35,d. Dorsal surface of femur with a spine in the middle and a transverse row of 3-4 spines between this spine and the apex of segment; tibia with 3 stout spines on its inner surface. Dimensions.—Length of carapace 6-3, width 6-8, total length 13-3 mm. Other Localitves.—2 99, Kalk Bay (S8.A.M., 1173 and 4602); 1 9, Steenbraas River, Gordons Bay (8.A.M., 3348); 3 99, Sir Lowry’s - Pass, Somerset West (S.A.M., 6072); 1 3, 2 99, Clifton, Cape Penin- sula (S.A.M., 4181); 1 9, Table Mountain (S.A.M., 4602); 1 2, Her- manus (S8.A.M., 8488); 2 92, Matroosberg Mountains (4000-6000 ft.) (S.A.M., 3454); 2 99, Kirstenbosch, Cape Town (B. 8305-8306, S.A.M.). Selenops fitzsimonsi n. sp. (fig. 36). Type, 1 3, Barberton, Transvaal (T.M., 2439). Colour.—Carapace yellow-brown, without radiations from thoracic stria, a lighter parallel-sided broad yellow area behind the eyes as a wide as the ocular row, bisected by the stria which is continued as a brown stripe on to the cephalic area; the broad median area of carapace bordered at the sides by a wavy brown line; eyes Fic. 36.—Selenops fitzsimonsi n. sp. Tibia of dado leks pedipalp, ¢. @, from in front, and 6, from SUfTounde ¥ ack; outer side. mandibles yellow, with a blackish-brown stripe of uniform width along their inner anterior margins; abdomen yellow above with a few indistinct brown mark- ings ; under surface and legs yellow, legs without black bands. Eyes.—Median eyes fairly strongly recurved, a line touching the upper surfaces of the anteriors, passing a little below the centres of the posterior medians; anterior medians about 3 the diameter of z ~— The Genus Selenops (Araneae) in South Africa. 597 posterior medians, less than their own diameter apart, and almost touching the posterior medians (less than 4 their radius from them); posterior medians ? the diameter of ere: laterals; anterior medians their radius from the edge of clypeus. Chelicerae.—Inferior margin with 2 moderate subequal teeth, separated by more than twice their basal width. Legs.—Tibia I and II with 6 pairs of inferior spines and 2 lateral spines on each side, metatarsus I and II with 3 pairs of inferior spines and 2 lateral spines on each side. Pedipalp.—Femur above with 1 spine in the middle and a trans- verse row of 4 spines between this spine and apex of segment; patella above with 2 spines, tibia with 4 spines above. Tibia with two widely separated processes, as in fig. 36, a, seen from in front, and fig. 36, 6, seen from the outer side. Palp in general resembling that of minor from Empangeni, Zululand, to which this species is allied. Dimensions. wbety of carapace 3-8, width of carapace 4-2, total length 8-6 mm. Selenops gilli n. sp. (fig. 37). Types, 2 33, Hell’s Gate, Uitenhage, Cape Province (S.A.M., 8448). Colour.—Specimens doubtlessly bleached. Carapace light reddish brown, a little darker anteriorly, thoracic stria brown, with some fine long radiating lines from it, cephalic area bisected by a light brown line which is duplicated for most of its length; man- dibles a little darker than cara- pace; abdomen above with some brown spots and wavy cross-bars over most of its surface, near its posterior extremity a transverse, procurved, fairly broad black Fic. 37.—Selenops gilli n. sp. Tibia of stripe, followed behind by a pedipalp, g. a, from in front, and slightly broader light transverse o,tnom outer side: marking; legs yellow entirely without spots or dark bands. _ Eyes.—Medians moderately recurved, a line touching the upper surfaces of the anterior medians passing a little below the centres of the posterior medians; anterior medians % the diameter of the posterior medians, a little less than their own diameter apart, and a little less than a radius from the posterior medians; posterior medians a 598 . Annals of the South African Museum. about # the diameter of a posterior lateral; anterior medians a radius from the edge of clypeus. Chelicerae.—Inferior margin with 2 moderate equal-sized teeth (distinctly smaller than those of the superior margin), separated by 3-34 times their width. Legs.—Tibia I and II with 6, metatarsus I and II with 3 pairs of inferior spines; no lateral spines. Pedipalp.—Femur with 1 inner, 2 superior, 1 outer spine near apex; patella with a superior apical spine; tibia subequal to patella and less than half the length of tarsus, with outer process seen from in front as in fig. 37, a, and from the outer side as in fig. 37, b. Dimensions.—Length of carapace 3-5, total length 6-8 mm. Selenops krausst Pocock. S. krausst Pocock. A.M.N.H. (7), vol. u, p. 350, fig. 3, 1898. The type came from “‘Cape Colony” and was based on an adult 9. It is without doubt closely allied to S. capensis, but differs from it in the general form of the vulva, which has a wider median area between the lateral lobes. Selenops monianus n. sp. (fig. 38). Type, 1 2, Clanwilliam, Cape Province (S.A.M., 5347). Colour.—Carapace light reddish brown with distinct radii from the thoracic stria, one of these passing straight forwards and terminating behind the median eyes; areas surrounding the eyes _ blackish; mandibles reddish brown, a little darker than the carapace; abdomen above with a number of minute brown dots becoming more numerous posteriorly, otherwise uniformly light yellow; legs with black bands on tibiae, more distinct on the anterior legs, femora without mark- ee ings except at the base below. Fic. 38.—Selenops montanus n. sp. Hyes.—Median eyes moderately 2, vulva. recurved, a line touching the upper surfaces of the anteriors, passing well below the centres of the posterior medians; anterior medians % the diameter of the posterior medians, a little less than their own diameter apart, and less than a radius from the posterior medians; posterior The Genus Selenops (Araneae) in South Africa. 599 medians $ the diameter of the posterior laterals; anterior medians little more than a radius from the edge of clypeus. Chelicerae.—Inferior margin with 2 moderate subequal teeth separated by twice their width. Vulva as in fig. 38. Legs.—Tibia I and II with 6 pairs of extremely long and strong inferior spines; metatarsus I and II with 3 similar pairs of spines. Dimensions.—Length of carapace 6-3, total length 14-5 mm. Other Specimens.—1 9, Gt. Winterhoek Mts., 4000-5000 ft. (S.A.M., 2731 and 2828); 2 99, Matroosberg Monrtamae (S.A.M., 3454 and 3519) (S.W. Cape Province). The mandibles much swollen and rounded in this species which is closely allied to S. capensis, differing from it, however, in the form of the vulva. Selenops regalis n. sp. (fig. 39). Type, 1 9, Knysna, Cape Province (N.M., 1910), coll. Rex. Colour.—Carapace brown, a roughly circular patch in the middle of the posterior half much lighter, a few fine radiations from the thoracic stria, eyes surrounded by a blackened area; mandibles dark reddish brown, much darker than carapace; abdomen above mottled brown; all legs with well-defined and distinct dark bands on both femora and tibiae. Eyes.—Medians weakly recurved, a line touching the upper surfaces of the anteriors passing below the centres of the posterior medians; Ree es ee ae i ia anterior medians ? the diameter of the ae medians, a little more than their own diameter apart, and about 2 their diameter from the posterior medians; posterior medians Edius to posterior laterals; anterior medians a little less. than their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 subequal teeth, far removed from each other by about 4 times their basal width, or the distance between the basal and the apical tooth of the superior margin. Vulva as in fig. 39. | Legs.—Anterior tibiae with 6, anterior metatarsi with 3 pairs of inferior spines. 600 Annals of the South African Museum. Dimensions.—Length of carapace 5:4, width 5-5, length of abdomen 6-5 mm. Selenops reservatus Lawrence. S. reservatus Lawrence. Ann. Natal Mus., vol. vi, pt. 2, p. 241, text-fig. 15, 1937. The type was described from Hluhluwe, Zululand, and was based on an adult 9. Selenops sexspinatus n. sp. (fig. 40). Type, 1 2, Concordia, Little Namaqualand (S.A.M., 5353). Colour.—Specimen doubtlessly much bleached. Carapace light yellow-brown, darker anteriorly than posteriorly; thoracic stria, the boundaries of the cephalic area, and the areas surrounding the eyes brown, darker than the remainder; _ radiations from the stria faint; * mandibles light reddish brown, ? darker than the carapace but not darker than the areas surrounding the eyes; abdomen above bleached Fic. 40.—Selenops sexspinatus n. sp. or rubbed, yellow; all legs with 2, vulva. distinct bands, especially on the tibiae. Hyes.—Medians fairly strongly recurved, a line touching the upper surface of the anterior medians cutting the posterior medians a little above their inferior surfaces; anterior medians half or less than half the posterior medians in size, their own diameter or a little less apart, and 4 their radius from the posterior medians; posterior medians 2 the diameter of the posterior laterals or a little less. Anterior medians their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 large subequal teeth, a little less than twice their width from each other. Vulva as in fig. 40. Legs.—Tibia I and II with 6, metatarsus I and II with 3 pairs of inferior spines. Dimensions.—Length of carapace 5, length of abdomen 9-3 mm. Selenops stauntont Pocock. S. stauntont Pocock. A.M.N.H. (7), vol. x, p. 330, 1902. S. elusus Lawrence. Ann. Natal Mus., vol. viii, pt. 2, p. 242, text-fig. 16, a—c, 1937. The Genus Selenops (Araneae) in South Africa. 601 The type which was based on an adult 9, came from Durban. No figure was given of the vulva. Various localities are given for the species by Lawrence in “Ann. Natal Mus.,” vol. vii, pt. 3, p. 492, 1938. The South African Museum has it from Kentani (13079), Bashee River (13103), Grahamstown (B. 5691), and East London (2441). Selenops stridulans n. sp. (fig. 41). Type, 1 3, Steinkopf, Little Namaqualand (S.A.M., B. 8852). Colour.—Carapace pale yellow, almost without markings, a few indistinct darker spots near the lateral margin, a dark bar behind each a Fic. 41.—Selenops stridulans n. sp. a, 6, tibia of pedipalp ¢, from in front, and from outer side. c, femur of pedipalp, outer and inferior surface. posterior lateral eye containing two spines; eyes surrounded by large black areas; mandibles pale yellow, the inner two-thirds of their anterior surfaces somewhat darkened; abdomen yellow, a few minute spots at the sides of the posterior extremity; legs yellow with a few indistinct blackish spots and stripes. _ Eyes.—Median eyes strongly recurved, a line touching the upper surfaces of the anteriors, passing a little above the lower surfaces of the posterior medians; anterior medians half or a little less than half the diameter of the posterior medians, a little less than their own diameter apart, and less than half a radius from the posterior medians; posterior medians about half the diameter of posterior laterals; anterior medians 4—3 their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 moderate teeth, the basal one a little larger than the apical one and separated from it by its own basal width. 602 Annals of the South African Museum. Legs.—Tibia I and II with 6 pairs of inferior spines, some of them extremely long, 2 lateral spines on each side and 4-5 superior spines; metetarsus I and II with 3 pairs of inferior spines and 2 lateral spines on each side in basal half. Pedipalp.—Femur differing from those of all other species of the genus 1n being short and globose (equal to, or a little shorter than the combined patella and tibia, while in other species it is distinctly longer and subparallel), and in having a patch of strong black spines and teeth on the outer side of the inferior surface (text-fig. 41, c). This perhaps represents a stridulating organ, as the inner surface of the trochanter of leg I also bears a fairly large number of minute chitinous teeth and spines which may represent the opposing surface of friction. Femur with 1 inner, 1-2 superior, and 1 outer spine in basal half, patella with 1 inner and 2 superior spines, the distal one the stronger, tibia with 2 unusually strong inner spines and | superior spine; tibial apophysis as in fig. 41, a, seen from in front, and fig. 41, 6, seen from the outer side. Dimensions.—Length of carapace 4, total length 7-8 mm. This species is characterised by the large number of strong spines on the legs and pedipalp, and by the peculiar patch of short spines and teeth on the pedipalp femur. Selenops transvaalicus n. sp. (fig. 42). Types, 2 99, Lydenburg, N.W. Transvaal (S.A.M., 13171). Colour.—Carapace reddish brown, cephalic portion hardly darker than the rest, with a trident-shaped marking in the middle behind the median eyes; thoracic portion with the radiations of the stria ill-defined, each bearing a blackish dot in the middle, a few brown dots near the marginal border which is not well defined; mandibles reddish brown with a narrow black stripe along their inner margin and an apical black marking; abdomen with some large ill-defined blackish-brown markings above; legs with weak and ill-defined dark bands. Eyes.—Median row not strongly recurved, a line touching the upper surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians half as large as posterior medians, their own diameter apart, and less than a radius from the posterior medians; posterior medians 2 the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of the clypeus. The Genus Selenops (Araneae) in South Africa. 603 Chelicerae.—Inferior margin with 2 equal-sized teeth, 13-2 their width from each other. Legs.—Tibia I and II with 6, metatarsus I and II with 3 pairs of inferior spines. , Vulva as in fig. 42. Dimensions.—Length of carapace 4:8, width of carapace 5-5, total length 11-4 mm. Other Specimens.—1 92, Lydenburg (S.A.M., 13555); 1 9, Komati- poort, KE. Transvaal (8.A.M., B. 4363). Fic. 42.—Selenops transvaalicus n. sp. Fic. 43.—Selenops tuckert Q, vulva. n.sp. 9, vulva. Selenops tuckeri nu. sp. (fig. 43). Type, 1 2, Junction of Crocodile and Marico Rivers, Transvaal (S420 1B. 3718), Coll., R. W. Tucker. Colour.—Carapace light reddish brown with indistinct markings, thoracic stria and sides of the cephalic area darker, a faint stripe in the middle of the cephalic area bifurcating half-way between the eyes and the stria, eyes surrounded by black areas; mandibles lighter than the carapace except on their inner anterior surfaces; abdomen light yellow with symmetrical brown markings; legs without bands except for 3 on inferior surface of femur I, and some fainter ones on inferior surface of femur II. Eyes.—Median eyes moderately procurved, a line touching the upper surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians } the diameter of posterior medians, their own diameter or slightly more apart, and a radius from the posterior medians; posterior medians subequal to the 604 Annals of the South African Museum. posterior laterals; anterior medians 2 their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal distinctly larger than the apical one, separated by 14 times the width of the basal tooth. Vulva as in fig. 438. Legs.—Tibia I and II with 6, metatarsus I and II with 3 pairs of long inferior spines. Dimensions.—Length of carapace 3-6, total length 10-3 mm. SUB-GROUP By. Key to the species of Selenops with 7 pairs of inferior tibial spines. 29. 1. Vulval plate large, black, subquadrate : ; : : - spencert. ~ -Vulval plate not large or black . : : : : 2. 2. Vulva with lateral lobes partly meeting in ‘ie nals fine : - : 3. Vulva without lateral lobes meeting in the middle line . ‘ : ge es 3. Total body length 18 mm. . ! : : : : : . basutus. Total body length 9 mm. . : ; ; 3 ; : rubicundus. 4, Plate of vulva wider than long . : : : : : . decoratus. Plate of vulva longer than wide . Avie ls ‘ : . septemspinatus. 3S. 1. Apophysis of pedipalp tibia with 2 processes 5 : 2 . decoratus. Apophysis of pedipalp tibia with 4-5 processes. Br) ate 2. Apophysis of pedipalp tibia with 5 processes, the Inter process bifimeato septemspinatus. Apophysis of pedipalp with 4 processes, the lateral process not bifurcate 3. 3. No angular prominence at base of pedipalp tarsus : , - whiteae. An angular prominence at base of pedipalp tarsus : : . marshalli. Selenops basutus Pocock. S. basutus Pocock. A.M.N.H. (7), vol. vii, p. 288, 1901. The type was based on a 2 from Teyateyaneng in Basuotland. No figure was given of the vulva. The species is unusual for its large size, 18 mm. in total length. Selenops decoratus n. sp. (fig. 44). Types, 5 99, 1 3, Ingwavuma, Zululand (N.M., 2422 and 2424). 2 Coleur.—Carapace brown, narrowly margined with black, decorated with spots and short bars of black, including a row of > is il The Genus Selenops (Araneae) in South Africa. 605 black spots along the lateral margin, and a large one on each side of the stria along the posterior margin; mandibles brown, their inner margins black, a black dot on the outer side near the base; abdomen mottled with black spots and bars, the whole effect dark brown; b Fic. 44.—Selenops decoratus n. sp. a, 6, vulva, 2. c, pedipalp, g, from in front. d, tibia of pedipalp from outer side. all legs with very clearly defined black blotches and bands, the pos- terior surfaces of femora white, metatarsi with 2 distinct black bands. Eyes.—Medians very strongly recurved, a line touching the upper surfaces of the anteriors would not cut the posterior medians; anterior medians small, less than half the diameter of the posterior medians, a little more than their own diameter apart, and a radius from the posterior medians; posterior medians $ the diameter of the posterior laterals; anterior medians a little less than their diameter from the edge of the clypeus. Chelicerae.—Inferior margin with 2 equal-sized teeth distinctly smaller than those of the superior margin, 24 times their basal width from each other. . Vulva as in fig. 44, a and 0. Legs.—Anterior tibiae with 7, anterior metatarsi with 3 pairs of inferior spines. Dimensions.—Length of carapace 3-3, total length 9 mm. $ Colour.—Carapace light brown, with ill-defined markings; eyes surrounded by deep black areas, a curved black bar behind the posterior lateral eyes, a black marginal spot on each side above the 606 Annals of the South African Museum. mandibles; mandibles almost white, a broad black band at their apices and a narrow black stripe along their inner sides; abdomen above mottled blackish brown; legs with fairly distinct blackish bands and blotches, these, however, not nearly so well defined as in the 9. Eyes and chelicerae as in description of the 9. Legs.—Anterior tibiae with 7—9 pairs of inferior spines and a labaral spine on each side near the base, anterior metatarsi with 3 pairs of inferior spines. Pedipalp.—Femur with 1 inner and 1-2 superior spines in distal half, patella with 1 inner and 2 superior spines; tibia shorter than patella and much shorter than tarsus seen from above; tarsus and tibial process as in fig. 44, c, seen from in front, tibial process seen from outer side as in fig. 44, d. Dimensions.—Length of carapace 3, total length 6-2 mm. One of the female specimens which agrees with the types in colour, eyes, chelicerae, and the number of spines on the anterior legs, differs markedly in the form of the vulva (fig. 44, 6). I presume that this figure represents the vulva at a later, perhaps more mature stage. Selenops marshalli Pocock. S. marshallt Pocock. A.M.N.H. (7), vol. x, p. 21, fig. 6, 1902. The type was a ¢ from Estcourt, Natal. Other localities: 1 2, Durban (N.M., 2183). The total length of the type was 11 mm. Pocock records a ¢ of S. spencers from Estcourt, so that it seems likely that these two species, spencerr and marshalli, may be synonymous. Selenops rubicundus n. sp. (fig. 45). Type, 1 2, Witpoort, near Belfast, Transvaal (T.M., 1155). Colour.—Carapace yellow-brown, cephalic area not darker than thoracic area, a marginal and submarginal row of ill-defined blackish spots, thoracic stria continued on to the cephalic area as a fine blackish median line with a lateral branch on each side; eyes surrounded by blackened areas; mandibles with their inner halves blackened, the remainder coloured as in the carapace; abdomen with a symmetrical pattern of black spots intermixed with red hairs; femora of legs with black irregular spots tending to merge into each other as stripes; tibiae and metatarsi with weakly defined dark bands, that at the base of each tibia strongest. Eyes.—Medians not strongly recurved, a line touching the upper The Genus Selenops (Araneae) in South Africa. 607 surfaces of the anteriors passing a little below the centres of the posterior medians; anterior medians 3 the diameter of the posterior medians, a little more than their own diameter apart, and 2 their diameter from the posterior medians; ee posterior medians #? the diameter of the posterior laterals; anterior medians 2 their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 teeth, the basal one distinctly larger than the apical one and 14-2 its basal width from it. Vulva as in fig. 45. Legs.—Tibia J and II with 7, metatarsus Fic. 45.—Selenops rubi- I and II with 3 inferior pairs of spines. ee ee Dimensions.—Length of carapace 4, total length 9 mm. Selenops septemspinatus Lawrence. S. septemspinatus Lawrence. Ann. Natal Mus., vol. viii, pt. 2, p. 243, fig. 17, a—c, 1937. The species was based on 3 $¢g and 5 99° from Kosi Bay, Zululand. Selenops spencer Pocock (fig. 46). S. spencert Pocock. A.M.N.H. (6), vol. xvu, p. 55, figs. 8, 8a, 1896. The type 2 was from Durban, Natal. Other Localities.—1 2, M’fongosi, Zululand (8.A.M., 4166); 1 9, Umtata, C. Province (8.A.M., 9256); 1 9, Manubi Forest, Transkei, C. Province (S.A.M., 14509); 1 2, Blue Chiff, Uitenhage, C. Province (8.A.M., 5710); 1 9, Grahamstown (S.A.M., B. 8283); 2 29, Port ni. Johns, C..Provinee (S.A.M., 11735): 1.9; Winkel Spruit, Natal (N.M., 1798); 1 ¢ from Estcourt, Natal (4000 ft.), has also been recorded by Pocock (A.M.N.H., ser. 7, vol. ii, p. 224, 1898). Vulva of one of the 2 specimens from Port _ St. Johns as in fig. 46. oe Glee oo The average total length (6 adult 99 measured) was 14:5 mm.; the type was 11 mm. in length. Pocock unfortunately did not describe or figure the male specimen which he recorded from Estcourt (loc. cit.), and it is possible that his S. marshalli is the male form of this species. 608 Annals of the South African Museum. Selenops whiteae Pocock (fig. 47). S. whiteae Pocock. A.M.N.H. (7), vol. x, p. 21, fig. 7, 1902: The type was a g from Grahamstown; the following is a description of the 2 based on an adult 9 and ¢ from the Pirie Forest, Kingwilliamstown, Cape Province (N.M., 1516). Colour.—Carapace light yellow in the thoracic region, with a number of ill-defined darker lines radiating from the stria; cephalic portion darker, light reddish brown, bisected by a discontinuous median blackish stripe; ocular area in general blackish brown; mandibles yellow, blackish towards their apices; abdomen brown with some darker broad chevron markings; femora with blotched markings; tibiae with ill-defined brown bands, metatarsi with strong blackish bands (a black band at apex and base, a yellow between them). Hits agi aserename Lyes.—Median eyes strongly recurved, a line whiteae Pocock. touching the upper surfaces of the anteriors Yale just cutting the lower edges of the posterior medians; anterior medians very small, less than half the diameter of the posterior medians, their own diameter or a little less apart, and less than a radius from the posterior medians; posterior medians subequal to posterior laterals; clypeus unusually high and concave so that the anterior medians project over it; anterior medians about twice their diameter from the edge of clypeus. Chelicerae.—Inferior margin with 2 large subequal teeth about twice their basal width from each other. Vulva as in fig. 47, occupying a large subquadrate plate and rather resembling that of S. spencer Pocock. Legs.—Anterior tibiae with 7, anterior metatarsi with 3 pairs of inferior spines. Dimensions.—Length of carapace 5-2, total length 13-7 mm. Other Specimens.—1 3, Port St. Johns, Cape Province (S.A.M., 13072). ( 609 ) 21. Additional Records, and Descriptions of New Species, of South African Alder- flies (Megaloptera), May-flies (Ephemeroptera), Caddis- flies (Trichoptera), Stone-flies (Perlaria), and Dragon- Jlies (Odonata).—By K. H. Barnarp, D.Sc., F.L.S., Assistant Director. (With 19 Text-figures.) Tuis is the eleventh report on the Fauna of the Mountain Ranges of the Cape Province, for the investigation of which I have received grants from the Royal Society of South Africa and the Research Grant Board.* My thanks and acknowledgments are herewith tendered to these bodies. I have also to thank Dr. Hewitt, Director of the Albany Museum, and Dr. Lawrence, Director of the Natal Museum, for submitting material; and to Mr. H. G. Wood (H. G. W.) and Mr. C. W. Thorne (C. W. T.) for their help and co-operation in the field- work. : MEGALOPTERA. With the exception of a n. sp. of Platychauliodes, and some additional localities for the other species, no great advance has been made in our knowledge of the Cape Alder-flies since my 1931 paper (Trans. Roy. Soc. 8. Afr., vol. xix, 2, 1931). A 3d and 2 Leptosialis africana have been collected, and their genitalia are described; but the larva still remains undiscovered. No larvae or pupae of the Taeniochauliodes type have been found which might be assigned to the genus Platy- chauliodes. The discovery of the common Cape genus, Taeniochauliodes, in the Drakensberg in Natal, not only greatly increases the known distribu- tion of the genus, but indicates that the mountain streams in Natal have an Alder-fly fauna hitherto unsuspected. * Previous reports: 1. ‘‘ Freshwater Crustacea,’ Trans. Roy. Soc. 8. Afr., vol. xiv, 1927. 2. ‘Colophon (Coleoptera),”’ ibid., vol. xviii, 1929. 3. “‘Alder-flies,”’ ibid., vol. xix, 1931. 4. ‘‘May-flies,” 2bzd., vol. xx, 1932. 5. “Terrestrial Isopoda (Woodlice),”’ Ann. S. Afr. Mus., vol. xxx, 1932. 6. “Further New Species of Colo- phon,” Stylops, vol. i, pt. 8, 1932. 7. ‘““A new Corduline Dragonfly,” ibzd., vol. ii, pt. 7, 1933. 8. “‘Caddis-flies,’’ Trans. Roy. Soc. 8. Afr., vol. xxi, 1934. 9. “‘Stone- flies,” Ann. S. Afr. Mus., vol. xxx, 1934. 10. “Dragon-flies,” 2bid., vol. xxxii, 1937. VOL. XXXII, PART 6, 42 610 Annals of the South African Museum. Chloromella peringueyi HK. P. 1931. Barnard, loc. cit., p. 171. Localitves.—French Hoek Pass (head-waters of River Zonder End) (13th December 1932, K. H. B. and H. G. W., 1 3); same locality (October 1933, H. G. W., larva); Oudebosch, River Zonder End Mts. (September 1933, H. G. W., larva); Gt. Winterhoek Mts., Tulbagh (November 1932, K. H. B. and H. G. W., larva); Kaaiman’s River, near George (January 1936, H. G. W., 1 9). The French Hoek $ has fewer costal cross-veins in the fore-wing, viz. 23; the 5th cross-vein is at the point of origin of Rs, and the 9th at the first cross-vein R—Rs. The @ genitalia resemble those of Taeniochauliodes, but, owing to the abdomen of the single specimen not being in good condition, the presence of mobile apical papillae on the lower appendages could not be determined with certainty. Platychauliodes capensis Brnrd. (ie aia.) 1931. Barnard, loc. cit., p. 175. Localitves—Bosch Kloof, Keeromberg, Worcester (January 1933, H. G. W., 13); Jonkershoek, Stellenbosch (26th February 1931, H.G. W.,2 3g; 4th April 1931, H. G. W., 1 ovig. 2; 14th February 1933, 082 ie Beand sei Ge Weel gala ®)e Gemtalia 9.—A figure is given which applies to all three species of this genus. The genital plate on the 8th segment is ovate and strongly chitinised, with an apical membranous projection. The subanal plate (lower appendages) is divided into two halves, separated by a groove with membranous lining; the apical lobes have no mobile papillae, such as are found in Taenochauliodes and Leptosvzalis (v. wnfra). Platychauliodes wood: Brurd. (Biss c.) LOSie barnande |oc..es) peli de Locahities.—Palmiet River (December 1932, H. G. W., 2d); Oudebosch, River Zonder End Mts. (January 1933, H. G. W., 14, 19; and January 1934, 13); Meiring’s Poort, Zwartberg Range (January 1935, H. G. W., 1g); Kaaiman’s: River,’ mearpGreonge (January 1936, H. G. W., 3 3d). Genitalia §.—An additional figure is given here to show the subanal lobe (or penis), the apex of which is subpentagonal and shield-like. Descriptions of New Species of South African Alder-flies, etc. 611 Fie. 1.—Platychauliodes capensis Brnrd. a, 6, lateral and ventral views of 8th and 9th abdominal segments, 9, in b the more strongly chitinised areas dotted. Platychauliodes woodi Brnrd. c, g genitalia viewed from behind and slightly froin below, for comparison with d. Platychauliodes thornei n.sp. d, 3 genitalia, from same viewpoint asc. e, pro- jections of the genital plate viewed from above, for comparison with fig. 8, c, in Barnard, loc. cit., 1931. Platychauliodes sp. micropterous 9. f. meso- and meta-thoracic segments with winglets, and one winglet further enlarged. 612 Annals of the South African Museum. Platychauliodes thornez n. sp. (Big. L,.wyie.} Locality.— Hottentot Holland Mts. (January 1933, K. H. B. and H&G. Wo ls ee): Similar to the other two species, but larger and distinguished by the ¢ genitalia. The latter are of the same general structure as in woodi, but, as may be seen by comparing the figures here given, show several differences. The superior appendages are more falcate and the patch of spinules on inner (ventral) surface is less strongly developed. The median projections of the genital plate are much broader, triangular when viewed from behind, but strongly cowl- shaped asin woodi. The lateral projections do not differ very much, but are feebly spatulate, only the top margin being curled over slightly. A subanal lobe, which may function as penis or intro- mittent organ, ends in a strongly chitinised triangular point. Fore-wing 6 27 mm.,92 42mm. _ Named after Mr. C. W. Thorne, of the South African Museum, who has accompanied Mr. Wood and myself on many collecting expeditions in the Cape mountains. It is possible that the large @ described and figured on p. 178 (loc. cit., 1931) belongs to this species. ? Platychauliodes sp. Micropterous form, or aberration.—In the Gt. Winterhoek Mts., January 1939, Mr. C. W. Thorne found a 9, 28 mm. in length, which is remarkable for having greatly reduced wings. All four winglets are alike; each consisting of a roughly semicircular membrane, spoon-like, 7.e. convex above and concave below, without trace of venation except a costal and (?) subcostal thickening meeting at the rounded apex. The length of each winglet is about 4 the width of the thoracic segment (fig. 1, f). The winglets are considerably smaller than the wing-pads of a normal pupa of Taeniochauliodes, and the unexpanded wings con- tained therein; moreover, they show no evidence of being normal- sized wings which have failed to expand on emergence. Judging by the shape of the head and the labrum, and the position of the ocelli, the specimen appears to belong to the genus Platy- chauliodes; the lower appendages, however, have mobile papillae at their apices as in Taeniochauliodes. Descriptions of New Species of South African Alder-flies, etc. 613 Although of frequent occurrence among the Stone-flies (see Barnard, Ann. 8. Afr. Mus., vol. xxx, p. 576, and also infra, p. 658), I am not aware of any record of a micropterous Alder-fly. Taeniochauliodes ochraceopennis H. P. fel. Barnard, loc. cit., p. 179. Localities—French Hoek Pass (River Zonder End _ system) (December 1932, K. H. B. and H. G. W.); Palmiet River (December 1932, H. G. W.); River Zonder End Mts., 1500-4000 feet (January 1933, H. G. W., pupae and imagos, and January 1934, K. H. B. and H. G. W.); Seven Weeks Poort, Ladismith, Cape (February 1932, K. H. B. and H. G. W.); George, Outeniqua Mts. (January 1931, K. H. B., larva); Robinson Pass, Outeniqua Mts. (February 1932, K. H. B. and H. G. W.); Kalk Bay Mts., Cape Peninsula (May 1933, A. C. Harrison, young larvae); Cathkin Peak, Drakensberg, Natal, 6000 feet (R. F. Lawrence, January 1938). The largest 2 yet collected has the fore-wing 38 mm. in length. The 2 genitalia resemble those of Platychauliodes, but the sub- genital plate is transverse, broader than long, without definite membranous distal projection; and the lower appendages have an apical mobile papilla as in Leptosvalis. Leptosialis africana E. P. (Fig. 2.) 1931. Barnard, loc. cit., p. 184. Localities —Gt. Winterhoek Mts., Tulbagh (3rd November 1932, K. H. B., 1 3); Hottentots Holland Mts. (23rd November 1932, C. W. Thorne, 1 2); Upper Olifants River, north of Ceres (8th October 1937, C. W. Thorne, 1 9). The wings are not held roof-like over the body when at rest, but curled round the body as in Taeniochauliodes and other local Cory- dalids. In both these specimens the junction of M and Cu, in fore-wing is more proximal, only slightly distal to the level of origin of Rs. In the hind-wing the fork of Cu, is slightly proximal to the cross-vein M344 — Cu. As the $ comes from the type locality there is no reason to doubt its identity with africana, of which the genitalia have not been de- scribed (they are missing in the type specimen). It is unlikely, but nevertheless just possible that a second species may exist in the 614 Annals of the South African Museum. Hottentots Holland Mts., although the 2 shows no differences in venation. | . Genitalia g. Superior appendages lateral in position, short, in- curved, clasper-like, hollowed on inner lower surface; supra-anal Fic. 2.—Leptosialis africana E.P. a, b, c, lateral, dorsal and ventral views of 6 genitalia. d, supra-anal lobe viewed from behind or flattened, not fore- shortened asinbandc. e, lateral view of 7th, 8th and 9th abdominal segments 9, with portion of integument further enlarged. /f, dorsal view of apical segment 9. g, ventral view 9, the more strongly chitinised areas dotted. lobe (210th tergite) triangular, with narrow apical cleft, the two apical points curved downwards and strongly chitinised; subgenital plate broadly rounded. 2. Subgenital plate on 7th segment broadly rounded, with minute median notch; superior appendages short, rounded-conical; subanal plate divided into two halves by a groove, and ending in a blunt lobe Descriptions of New Species of South African Alder-flies, etc. 615 with a mobile papilla on lower apical surface; the lining of the groove is feebly chitinised and the two halves of the lobe can be moved to and fro, and evidently serve to place the ova in position. The whole surface of the integument (in both sexes) is minutely hirsute (fig. 2, e) in addition to the larger setae. EPHEMEROPTERA. Since the publication of my 1932 paper (Trans. Roy. Soc. 8. Afr., vol. xx, pt. 3) numerous additional locality records have accumulated as the result of continued collecting. A few more dried specimens have been found in the 8.A. Museum, and small collections have been submitted for identification by Mr. H. K. Munro of the Etomological Department, Pretoria, Dr. J. Hewitt, Director of the Albany Museum, and Dr. R. F. Lawrence, Director of the Natal Museum. Three undescribed species also have been collected; and a nymph which appears to be the true nymph of Elassoneuria trimeniana has been discovered. Mr. A. C. Harrison has observed another case of longevity, and proved the ovoviviparity of the Red Border Wing (Cloeon lacunosum). A large Polymitarcid from Natal has been described by me in Ann. Natal Mus., vol. vin, 1937. Some unfortunate typographical errors occurred in the 1932 paper and are herewith corrected :— Weaation, figs. 1, 2, 4,5, 17, 19, 42. For “1R,, 1R,, IMA, 1M, 1Cu (but not 1A),” read “I (=intercalary) R,, IR, etc., re- spectively.” Pages 204, 208, and 244, line 32. For “‘westermanni”’ read “dis- locans.”’ Page 245, fig. 35b. For “‘westermanni”’ read “‘dislocans’’; fig. 35c, for “dislocans”’ read “auriculata.” Page 217, fig. 7. For “‘perkensi”’ read “‘perkinst.” Page 220. Delete last sentence at bottom of page: “‘the nymph . . . discovered.” Page 230, line 1. For “‘access”’ read “assess.” Page 246, line 12. For “long stalk of MA” read “long stalk of M.” Page 252. Delete last sentence of first paragraph: “Other similar Ha October 1951.’ Taxonomy.—Dr. Ulmer (Stettin. Entomol. Zeit., xciii, 1932, pp. 204-219) has briefly commented on the new genera proposed by me. In his opinion ELuthraulus, Austrocaenis, and Austrocloeon are 616 Annals of the South African Museum. of doubtful validity, but he declines to express an opinion as to whether nymphal characters alone are sufficient to justify generic separation (p. 216). For my part I feel that the nymphal stage has very strong claims for recognition in classification, although its full value cannot be estimated until the nymphal stages of all the known species of imagos have been correlated. As examples of the value of nymphal charac- ters one may refer to Torleya Lest., Ephemerellina, and Lzthogloea, Aprionyx and Atalophlebia, Epeorus and Iron (see Ueno. Annot. Zool. Japon, xiii, 3, 1931, p. 192), and Acentrella (v. infra). Tillyard (Proc. Linn. Soc. N.S.W., lviii, 1933, p. 2) speaks in favour of founding new genera on nymphal characters alone; and in 1934 (Pap. Proc. Roy. Soc. Tasman. for 1933) he describes the nymph of the genotype of the genus Atalophlebia as having denticulate claws, thus confirming the validity of the genus Aprionyz. _ Genotypes.—It is usually understood that when a new genus con- taining more than one species is proposed, and the genotype is not specifically designated, the first species listed or described is to be reckoned as the genotype. In view, however, of the resolution of the Budapest Congress of 1927 (see Entomol. Monthly Mag., December 1932), Austrocloeon africanum (EH. P.) and Aprionyx tabularis (Haton) are herewith designated as the genotypes of their respective genera. The other new genera proposed by me were monotypic. Hot Springs—Nymphs of Baetis bellus and Centroptilum sud- africanum have been found at the stream issuing from the Warmbaths near Citrusdal. At the spot where the nymphs occurred the water would be a few degrees lower than 108° to 110° F. (see “South African Caddis-flies,”’ Trans. Roy. Soc. 8. Afr., xxi, p. 297). Fam. PROSOPISTOMATIDAE. 1921. Lestage in Rousseau, Larves et Nymphes Aquatiques, 1, p. 177 (Prosopistomidae). 1932. Lieftinck, Tijdschr. Entomolog., Ixxv, suppl. p. 44 (references). The remarkable May-fly nymph (only one subimago has ever been discovered) known as Prosopistoma has been recorded from Europe, Madagascar, and Java. Recently (August 1939) I have seen a speci- men taken from a trout stomach in Kenya. It seems therefore worth while to mention the occurrence of Prosopistoma in Africa, in the hope that some day it will be found in Descriptions of New Species of South African Alder-flies, etc. 617 South Africa. According to Lieftinck the Javanese species lives in the crevices on the undersides of boulders in fast-flowing streams. Fam. POLYMITARCIDAE. Polymitarcys savignyi (Pict.). 1932. Barnard, loc. cit., p. 209. Ulmer (1932, loc. cit., p. 208, and in litt. 26/5/32) points out that the vein labelled in my figure 1 as Cu, is really only a branch of Cu,, and that the true Cu, is the vein labelled 1a. Gen. EXEUTHYPLOCIA Lest. 1918. Lestage, Rev. Zool. Afric., vi, p. 74. 1933. Ulmer, Peking Nat. Hist. Bull., vii, p. 197. 1937. Barnard, Ann. Natal Mus., viii, p. 275. Distinguished from Polymitarcys by the subparallel sigmoid cross- veins in the anal area of the fore-wing. Ezeuthyplocia sampsoni Brnrd. 1937. Barnard, loc. cit., p. 276, fig. Localities—Umzimkulu River (28th October 1936); Yarrow and Mooi Rivers (Mid-October 1933); Mooi River, Nottingham Road (6th November 1933). (All in Natal.) Nymph provisionally assigned to Exeuthyplocia sampsont. (Fig. 3.) On the 25th October 1939 Dr. B. Sampson found a single empty nymph-shuck floating in an eddy on the Umzimkulu River, Natal. No imagos were seen, and Dr. Sampson says there was no hatch between 22nd and 30th October. From the size of the nymph one suspects it to be the nymph of E. sampsoni, which is the largest and bulkiest May-fly in the Natal fauna, excepting perhaps Hatonica schoutedent. It resembles the EHuthyplocia nymph described by Eaton (Trans. Linn. Soc. London, 2nd ser., zool., ili, p. 37, 1883, and pl. 29, 1884), and also Palingenia, in the 3-jointed maxillary and labial palps; but differs in having a small, non-fimbriate gill on the Ist abdominal segment. In the latter character it resembles Ephemera, Hexagema, and Polymitarcys. The 2nd-7th gills, however, are not like those of Polymitarcys. 618 Annals of the South African Museum. Safe ONS) ore eee > Y Peg cy = > HY — as Tra if i) 4 Ze A s ae eee ——— Y wulalls, = . Ss ar SS yy \ a : Sea f V1 TTX SNL Ss 9 Malin POT yi oR if Adele 3 # cs Wig \' yn LES Uh ~ Ad a= wer Rena Wik / Saw, ae as) — 5 vay WY) pa SST Wh Law Wie eee: Lif > ees — mer} ~ / ’ QOS r {7 (Feed > U7, F Pte * ee) SEE LETT td SS Le 73 Y, ? Fic. 3.—Nymph provisionally assigned to EHxeuthyplocia sampsoni Brnrd. a, nymph drawn from empty shuck, showing split in middle line of thorax; left wing-cases cut away to show small Ist gill; cerci cut off. 6, ventral view of left mandible, and molar portion of right mandible. c, dorsal view of apex of left mandible, setae omitted. d, maxilla. e, labium. f, ventral view of fore-leg. g, ventral view of apex of tibia of middle leg, only a few of the long setae shown. A, Ist gill. 7, 2nd gill. Descriptions of New Species of South African Alder-flies, etc. 619 Front of clypeus to end of abdomen 25 mm., cerci 12-13 mm. (apparently complete). Clypeus projecting horizontally, setose. Mandibles triquetral, thickly setose on inner and outer margins, outer margin with short spinules, a line of spinules on upper margin, double proximally, becoming a single series distally. Tibia of fore-leg with lower surface flattened and spinules along inner and outer margins, upper surface convex, densely setose. Claw of fore-leg short, almost concealed in a thick bunch of short blackish spines and bristles on apex of tarsus. Tibia of middle leg cylindrical, spinose and densely setose on outer surface, long setae also on outer distal portion of lower surface, inner apex with a tuft of stiff bristles. Femur of hind leg broadly ovate, flattened but not specially so on lower surface, both margins and upper surface densely setose, inner (front) margin in addition with spinules, which are inserted ventrally to the long setae. Hind tibia expanding distally, densely setose on hind (outer) margin, inner margin with long setae and stiff bristles, the latter especially numerous and densely packed distally. Abdomen in cross-section trapezoidal, the dorsal surface narrower than the ventral surface, the former marked by a dorso-lateral setose ridge on segments 3-7; on segments 8 and 9 there is a fringe of setae but no very definite ridge. A single transverse fringe of setae on segments 1, 2, and 8, near hind margin on segments | and 2, anterior on 8; on segments 3-7 two transverse fringes of setae. Gills, 7 pairs; the Ist pair very small, bilamellate, non-fimbriate and non-setose; 2nd—7th pairs elongate, gently curved with very numerous (far more numerous than represented in the figure, especi- ally on anterior margin of upper lobe) long slender filaments. Owing to the dorsal transverse and dorso-lateral fringes of setae on the abdominal segments, it is doubtful if the gills are held curved over the back in life; apparently they can be folded against the sides and more or less protected between the dorso-lateral and ventro-lateral fringes of setae. Fam. EPHEMERIDAE. Eatonica schoutedeni (Navas). 1932. Barnard, loc. cit., p. 210. Locality.—Nelspruit, Transvaal (November), H. K. Munro. 620 Annals of the South African Museum. Fam. OLIGONEURIIDAE. Elassoneuria tromeniana (McLach.). (Fig. 4.) 1932. Barnard, p. 212, fig. 4 (references). Lestage (1916, Rev. Zool. Afric., iv, p. 314, figs. 1-5) described a nymph (20 mm. in length, excl. cerci) from the Belgian Congo, resembling in general that of Oligoneuria rhenana (see Lestage, Larves et Nymphes Aquatiques, i, p. 214, fig. 54, 1921). The head was not figured, but part of the description says, “bord antérieur finement cilié et offrant sur la partie médiane une saillie longitudinale caréniform bien visible légerement prolongée en avant et formant comme un petit nez procéphalique.’’ In this respect it differs from Oligoneurra (loc. cit., 1916, p. 318). In 1917 (Rev. Zool. Afric., v, p. 122, fig. 1) Lestage assigned this nymph to £. trimeniana, and gave a figure of the whole animal showing the head with its anterior point and dorsal keel. The two nymphs (13 mm. in length, excl. cerci) before me, from Cathkin Peak, 6000 ft. alt., Drakensberg, Natal (Dr. R. F. Lawrence, January 1938), differ in the shape of the head which is rounded in front, without any longitudinal keel or ridge, widest in front of eyes instead of across the postero-lateral angles, and in the shorter antennae (as in Oligoneuria). Since HL. trimeniana was originally described from Natal, there is the greatest probability that this is its true nymph. In which case, either the Congo nymph belongs to another species whose imago is as yet unknown, or the &. trimeniana of Ulmer (1916, Arch. Naturg., Ixxi, 1915, Abt. A, p. 4) is not the true tromeniana of McLachlan. The further alternative that the carinate and pointed head and the longer antennae are assumed only in nymphs larger than, say, 13 mm. does not seem at all likely. This latter statement is confirmed by some nymphs, 13-19 mm. in length, from Kenya, which I have recently examined. These have the head widest across the postero-lateral angles, without medio- dorsal keel or anterior point, antennae short as in the Natal specimens, but the median cercus at least half the length of the lateral ones. It seems probable that more than one species will eventually be recognised in Africa. Descriptions of New Species of South African Alder-flies, etc. 621 Fig. 4,.—Elassoneuria trimeniana (McLach.). a, whole animal, with seta from fore- leg, and claw of middle and hind legs further enlarged (the claw of fore-leg is without denticles). 6, ventral view of head, showing maxillary gills projecting from under the labium (setae omitted). c, labrum. d,.mandible. e, hypo- pharynx. f, maxilla, the submarginal row of bristles is on the outer (ventral) surface. g, labium, inner (dorsal) view, with cross-sectiqn of palp. h, one of the lateral gills in dorsal view. 622 Annals of the South African Museum. Fam. BAETIDAE. Cloeon lacunosum Brnrd. 1932. Barnard, loc. cit., p. 214. Longevity.—Mr. Harrison records a further case (cf. Barnard, p- 214). A Q subimago emerged at 8 p.m. on 15th July 1932, and transformed to the imago during the night 16th-17th. She lived until 8 p.m. on 7th September 1932, 7.e. for 54 days after emergence from the nymph. Although $¢ were supplied, she died unfertilised, as in the case previously recorded. Ovoviviparity.—On 12th December 1932 Mr. Harrison was able to prove that this species is ovoviviparous like C. dipterum. “A female alighted on the water with wings flat out, and as she lay there a stream of larvules left her. After placing in a tube of water a few more larvules came away. Immediately afterwards a second female alighted and was promptly placed in a tube, when the larvules streamed out and swam fast in all directions.”’ The larvules possess two cerci and the buds (anlagen) of the seven pairs of gills. Localities.—Reservoir, Platteklip stream, Cape Town (26th Novem- ber and 26th December 1932, 99, K. H. B.); Orange Kloof, Table Mt. (Ist March 1933, 99, K. H. B. and H. G. W.); St. James (21st January 1933, A. C. H., gg, “swarming”); Muizenberg Reservoir (December 1932, A. C. H., 99, with larvules); Diep River, near Philadelphia, Cape Div. (26th April 1932, K. H. B., nymphs). Cloeon chaplint Brnrd. 1932. Barnard, loc. cit., p. 216. Locality.—Jonkershoek, Stellenbosch (November 1931, F. G. Chaplin, nymphs). Cloeon aeneum Brnrd. 1932. Barnard, loc. cit., p. 216. | Localities —Palmiet River mouth (3lst July 1932 and 25th December 1932, H. G. W.); French Hoek Pass (Ist October 1933, K. H. B.); Upington (November 1920, Father Sollier). The latter mutilated specimen in the 8.A. Museum collection seems to be this species. Cloeon perkinsi Brurd. 1932. Barnard, loc. cit., p. 216. Locality.—East London (March, H. K. iunkey Descriptions of New Species of South African Alder-flies, etc. 623 Austrocloeon africanum (E. P.). 1932. Barnard, loc. cit., p. 218. Localities.—S.A. Museum grounds, Cape Town (12th April 1932, 8th May 1933, and 7th December 1932, Dr. E. L. Gill); Welgemoed Farm, east of Ceres (February 1922, K. H. B.); Swellendam Mts. (October 1925, K. H. B.); Platteklip stream, Table Mt. (26th April 1933, K. H. B.); Meirings Poort Mts. (February 1932, K. H. B. and H. G. W.); Boschluis Kloof, Zwartberg Range (February 1932, ie. and H. G. W.). Austrocloeon virgiliae Brurd. 1932. Barnard, loc. cit., p. 219. Localities.—Reservoir, Platteklip stream, Cape Town (26th December 1932, K. H. B.); Ida’s Valley, Stellenbosch (20th February 1932, A. C. H.); Boschluis Kloof, Zwartberg Range (February 1932, K. H. B. and H. G. W.); Meirings Poort, Zwartberg Range (January 1935, K. H. B. and H. G. W.); Near Cango Caves, Oudtshoorn Distr. (September 1933, K. H. B.); Willow R., north slopes of Cockscomb Peak (Uitenhage Div.) (K. H. B., October 1938). Austrocloeon nymphs. Locality.—Drieling’s Kloof, between Laingsburg and Ladismith (February 1932, K. H. B. and H. G. W.). Pseudocloeon vinosum Brurd. 1932. Barnard, loc. cit., p. 220. Localities.—Orange Kloof, Table Mt. (1st March 1933, K. H. B. and H. G. W.); French Hoek Pass (Ist October 1933, K. H. B.); Hotten- tot Holland Mts. (January 1933, K. H. B. and H. G. W.); du Toits Kloof, Rawsonville (25th March 1932, K. H. B.); Seven Weeks Poort, Ladismith, Cape (February 1932, K. H. B. and H. G. W.); Upper Olifants R., north of Ceres (October 1937, K. H. B. and Caan. I). Pseudocloeon magae Brnrd. 1952. Barnard, loc. cit., p. 221. Localities.—Orange Kloof, Table Mt. (Ist March 1933, K. H. B. and H. G. W., 33, 1 2); Silvermine Stream, Kalk Bay Mts. (12th March #955, A.C. H.). 624 Annals of the South African Museum. Baetis harrisont Brurd. (Fig. 5, a). 1932. Barnard, loc. cit., p. 229. Localities.—Groot Drakenstein (25th October 1933, K. H. B. and H. G. W.); French Hoek Pass (1st October 1933, H. G. W.); Platte- klip stream, Cape Town (26th April 1933, K. H. B.); du Toits Kloof, Cer! iJ AS Cac Song . ' 4 <= ' \ ’ A . ee or ~, shane Ven X eather: tan u L J ¥ 3 i 1 a b c Fic. 5.—Baetis harrisom. a, apex of abdomen of nymph, with subimago ready to hatch. Acentrella capensis. b, c, apex of abdomen of nymph and of imago. Rawsonville (25th March 1932, K. H. B.); Seven Weeks Poort, Ladi- smith, Cape (September 1933, K. H. B., nymphs); Near Cango Caves, Oudtshoorn Distr. (February 1932, K. H. B. and H. G. W., nymphs, and September 1933, K. H. B., imagos); Fairy Glen, Worcester (12th August 1932, K. H. B. and H. G. W.); Porterville (August 1937, A.C. H.); Upper Olifants R., north of Ceres (October 1937, K. H. B. and C. W. T.). Baetis bellus Brurd. 1932. Barnard, loc. cit., p. 222. Localities. —Great Winterhoek Mts., Tulbagh (November 1932, K. H. B. and H. G. W.); Warmbaths, Citrusdal (September 1932, K. H. B., nymphs); Platteklip stream, Cape Town (26th April 1933, K. H. B.); Blinkwater stream, above Camps Bay, Cape Peninsula (21st August 1932, K. H. B.). Baetis sp. Nymphs of what seems to be a different species of Baetis were found at Meirings Poort mountains, Zwartberg Range (February 1932, Ko BB. and Ee .G. W.): — ee a Be rc Descriptions of New Species of South African Alder-flces, etc. 625 Acentrella capensis Brurd. (Bia. 551b,6): ma2q barnard, loc. cit., p. 259, It seems impossible to separate with certainty the adults of this species from those of Baetis harrisont except by breeding. Unless examined when freshly caught the abdominal coloration is liable to be inconclusive. The short branch of the fork 1 A may be clearly disconnected, or it may be almost connected. In any case this feature is variable among the species of Baetis, as Dr. Ulmer has kindly pointed out to me (in lett. 26/5/32), and should not be used as a specific character. Figures are here given of the apex of the abdomen of the nymph and the adult. For comparison a figure is also given of the abdomen and cerci of a nymph of Baetis harrisonz, showing the subimago ready to hatch. The median cercus is fully developed in the nymph, but in the final instar the enclosed adult cercus degenerates and becomes - detached near the base, remaining behind in the empty shuck after emergence of the subimago. Localities. —Naudesberg, Worcester (August 1932, K. H. B. and H. G. W., nymphs); Gt. Winterhoek Mts., Tulbagh (November 1932, K. H. B. and H. G. W., nymphs). Centroptilum sudafricanum Lest. 1932. Barnard, loc. cit., p. 224. Localities.—Orange Kloof, Table Mt. (lst March 1933, K. H. B. and H. G. W.); Platteklp stream, Cape Town (26th April 1933, K. H. B.); Blinkwater stream, above Camps Bay (31st January, 21st March, and 21st August 1932, K. H. B.); Hottentots Holland Mts. (January 1933, K. H. B. and H. G. W.); Gt. Winterhoek Mts., Tulbagh (November 1932, K. H. B. and H. G. W.); Warmbaths, Citrusdal (September 1932, K. H. B., nymphs); Farm Ezelfontein, 7 miles east of Ceres (September 1933, K. H. B.); Prince Albert, mouth of the Zwartberg Pass (September 1933, K. H. B.); Palmiet River, Kleinmond (March 1932, H. G. W.); Jonkershoek, Stellenbosch (14th February 1933, K. H. B. and H. G. W.); Oudebosch, River Zonder End Mts. (September 1933, H. G. W.); Groot Drakenstein (25th October 1933, K. H. B. and H. G. W.); Meirings Poort, Zwart- berg Range (February 1932, K. H. B. and H.G. W.); du Toits Kloof, Huguenot (31st March 1934, H. G. W.). VOL. XXXII, PART 6, 43 626 Annals of the South African Museum. Centroptilum excisum Brnrd. 1932. Barnard, loc. cit., p. 224. Localities.—Meirings Poort, Zwartberg Range (February 1932, K. H. B. and H. G. W.); Onderbokfontein, Cold Bokkeveld, north of Ceres (September 1932, K. H. B.); M’fongosi, Zululand (April 1917, W. E. Jones). Centroptiloides bifasciatum (H.-P.). 1932. Barnard, loc. cit., p. 226. Locality.—Yarrow River, Karkloof, Natal (18th March 1934, L. A. Day). Fam. BRACHYCERCIDAE. Austrocaems capensis Brnrd. 1932. Barnard, loc. cit., p. 227. Localities.—Diep River, Philadelphia, Cape Div. (26th April 1932, K. H. B., nymphs); Palmiet River (March 1932, H. G. W., nymph); du Toits Kloof, Rawsonville (25th March 1932, K. H. B.); Seven Weeks Poort, Ladismith, Cape (February 1932, K. H. B. and H. G. W.); Boschluis Kloof, Zwartberg Range (February 1932, K. H. B. and H. G. W.); near Cango Caves, Oudtshoorn Distr. (September 1933, K. H. B., nymphs, imagos). Austrocaenis sp. A g imago caught at Groot Drakenstein (25th October 1933, H. G. W.) is smaller and much darker than the normal capensis. Length of wing 3:25 mm. Head and thorax dark Vandyke brown, abdomen strongly suffused with sepia, legs normal, but the dark streaks almost black, Sc and R brown, cerci white. Also 2 $3 from Umzimkulu R., Natal (Dr. B. F. Sampson, October 1937). Gen. TRICORYTHUS Eaton. Ulmer (1932, loc. cit., p. 215) remarks that T. discolor, with its peculiar nymph, is not typical of the genus Tricorythus. If that is so, possibly a new genus may later be required for the South African species. The characteristic nymphs have been found at the following localities: Hogsback, Amatolas, King Wilhams Town Div. (February 1933, R. F. Lawrence); Seven Weeks Poort, Ladismith, Cape (September 1933, K. H. B.); Cathkin Peak, 6000 feet, Drakensberg, ; ; } a ee ee ee ee a a Descriptions of New Species of South African Alder-flies, etc. 627 Natal (January 1938, R. F. Lawrence); Bushmans and Umzimkulu Rivers, Natal (Dr. Sampson). Tricorythus reticulatus Brnrd. 1932. Barnard, loc. cit., p. 232. Locahty.—Pretoria, at light (February 1934, H. K. Munro, 99) (one with egg-mass), wings, 6-5-7 mm. Fam. LEPTOPHLEBIIDAE. Gen. APRIONYx Brnrd. 1934. Tillyard, Pap. Proc. Roy. Soc. Tasman. for 1933, pp. 5, 6 (discussion). Aprionyx petersenr (Lest.). 1932. Barnard, loc. cit., p. 236. Localities —Du Toits Kloof, Rawsonville (25th March 1932, K. H. B.); Lemoenshoek, Langeberg Range (November 1927, K. H. B.); Palmiet River, Kleinmond (March 1932, H. G. W.); Klands Kloof, Citrusdal (5th March 1933, H. G. W.); French Hoek Pass (December 1932, K. H. B. and H. G. W.); Oudebosch, River Zonder End Mts., 1500 feet (January 1933, H. G. W., and January 1934, K. H. B.); Tulbagh Valley and Gt. Winterhoek Mts. (February 1934, K. H. B.); Meirings Poort, Zwartberg Range (January 1935, cee Band H..G. W:). Aprionyx intermedius Brunrd. 1932. Barnard, loc. cit., p. 238. Localities.—Fairy Glen, Worcester (12th August 1932, K. H. B. and H. G. W.); Groot Drakenstein (12th September 1932, A. C. H.); Ruiterbosch, Robinson Pass, Outeniqua Range (February 1932, K. H. B. and H. G. W.); Great Winterhoek Mts., Tulbagh (September io2 ak. EL. B.). Aprionyx rubicundus Brnrd. 1932. Barnard, loc. cit., p. 239. Localities.—Hottentots Holland Mts. (east side of Spitzkop) (November 1932, K. H. B. and H. G. W.); Witte River, Wellington Mts. (1st October 1933, H. G. W.); French Hoek Pass (1st October 1932 and 1933, K. H. B. and H. G. W.); Oudebosch, River Zonder - End Mts., 1500 feet (December 1931 and January 1933, H. G. W., and January 1934, K. H. B.); Fairy Glen, Worcester (12th August 628 Annals of the South African Museum. 1932, K. H. B. and H. G. W.); Gt. Winterhoek Mts., Tulbagh (November 1932, K. H. B. and H. G. W.); Upper Olifants R., north of Ceres (October 1937, K. H. B. and C. W. T.). When freshly caught most specimens are considerably darker than would appear from the description; in fact, when settling on white stones in the sunlight the flies look quite black. Wing length of 22 9-10 mm. Aprionyx pellucidulus (K. P.). (Mie G5a,) 1932. Barnard, loc. cit., p. 239. Localitves.—Hottentots Holland Mts. (January 1933, K. H. B. and H. G. W.); River Zonder End Mts., 3500-4500 feet (January 1933, H. G. W., and January 1934, K. H. B.); Gt. Winterhoek Mits:, Tulbagh (November 1932, K. H. B. and H. G. W., nymphs and adults). Nymph.—With smooth claws, and other features as in diagnosis of Aprionyx. Front tibia and tarsus as in petersent, without the fringe of long hairs found in tabularis. Mottled brown, antennae and cerci ochraceous. Femora banded. Gills greyish-white, tracheae blackish. Abdomen pale yellowish or ochraceous, segments 1-9 each with narrow transverse dark band on hind margin, segment 6 slightly suffused on anterior margin, segments 7-9 with anterior half suffused but leaving a narrow pale median line, on either side of which is a small pale spot (fig. 6, a). Females are rather more strongly suffused, the transverse bands being enlarged at both ends into a small dark spot; the whole abdomen is deeper in colour owing to the eggs developing inside. The suggestion that the Zwartberg nymph (loc. cit., fig. 29, c) might be the nymph of this species is herewith withdrawn. Aprionyx argus Ni. sp. (Fig. 6, d-f.) Imago.—8-9 strong cross-veins before bulla, 9-10 in pterostigmal area, oblique, straight or nearly so, only occasionally forming forks at costal margin or anastomosing. Hind-wing with Sc extending nearly to end of wing, 4-6 cross-veins in subcostal area. Penis 3S thin dorso-ventrally, broad, apically notched, with the openings of the vasa deferentia laterally at apex. Clasper with inner margin angularly curved proximally. Ventral plate 2 apically indented. Descriptions of New Species of South African Alder-flies, etc. 629 Body $ 10 mm., 9 12 mm.; wing 3 10-11 mm., 2 12 mm. g. Head and thorax dark Vandyke brown. Legs ochreous, femora with dark band in middle and at apex. Abdomen brown, Pipes RS Os es pits a Fic. 6.—Aprionyx pellucidulus (E.P.). a, abdomen of nymph. Aprionyx natalica (Lest.). 6, penis and clasper g. c, segment of abdomen. Aprionyx argusn.sp. d, penis and clasper, with lateral view of former. e, seg- ment of abdomen. /f, fore-wing. segments 1, 9, and 10 unicolorous, segments 2-8 each with a pair of large pale spots at base dorsally, pale spots laterally and ventrally. Neuration dark fuscous-brown, some specimens more heavily marked than in the specimen figured. Cerci whitish, annulated. Penis pale, apices (or only inner apical margin) of claspers whitish. @ similar. Locality.—Cathkin Peak, 6000 feet, Drakensberg, Natal (R. F. Lawrence, January 1938). 630 Annals of the South African Museum. Remarks.—Somewhat like natalica but more heavily veined, and reminiscent of the Cape species petersent. A figure of the penis and clasper of natalica (fig. 6, 6), from the above same locality, is given for comparison. Both the Natal species differ from all the Cape species in the shape of the penis. Adenophlebia auriculata Eaton. 1932. Barnard, loc. cit., p. 242. Localities.—Stutterheim (September, H. K. Munro); Hogsback, Amatolas, King Wiliams Town district (February 1933, R. F. Lawrence); Kaaiman’s Gat, near George (April 1933, H. G. W.); Howieson’s Poort, near Grahamstown (October 1933, J. Hewitt). Adenophlebia peringueyella Lest. ~ 1932. Barnard, loc; cit... p. 942). Localitues.—Seven Weeks Poort, Ladismith, Cape (February 1932, K. H. B. and H. G. W., also September 1933, K. H. B.); Gt. Winter- hoek Mts., Tulbagh (September 1932, K. H. B.); Schoemann’s Poort, Cango, Oudtshoorn Distr. (September 1933, K. H. B.); Huis River, between Cango and Calitzdorp (September 1933, K. H. B.). Three small gg from Seven Weeks Poort, February 1932, appear to be this species. Wing-length 8 mm.; in two of the specimens there are 8 cross-veins between R,,; and MA,, and also between MA, and M,,.,, in the third specimen 9 cross-veins. The wings are rather heavily spotted, but not so much as in the wing dissected from a Zwartberg Pass nymph (loc. cit., fig. 31). Another ¢ caught on the same occasion has wing-length 8-5 mm.; 9 cross-veins on either side of MA; veins no more enlarged or spotted than in auriculata (loc. cit., fig. 30). In the extreme western Cape peringueyella is a spring fly and does not seem to last on until the summer. It is possible that these small forms may be a later brood (the typical perongueyella with wing- length 10-12 mm. occurs in the same locality in September), but it must be remembered that the Zwartberg Range may get some heavy rainfalls in summer, so that a marked diminution in the flow of the streams and consequent lessening of the food supply cannot be in- voked to explain the small size of these later forms. If it were possible, observations throughout a whole year at Seven Weeks Poort would prove very interesting. Descriptions of New Species of South African Alder-flies, etc. 631 Adenophlebia dislocans (Wlkr.). 1932. Barnard, loc. cit., p. 244. Localities.—Platteklip stream, Cape Town (3rd January 1932 and 26th April 1933, K. H. B.); Oudebosch, River Zonder End Mts. (September 1933, H. G. W., and January 1934, K. H. B.); Fouches Hoek, Mosterts Hoek (Breede River valley) (17th April 1933, K. H. B.); du Toits Kloof, Huguenot (31st March 1934, H. G. W.). Adenophlebia nymphs. Locality.—Ezelfontein, east of Ceres (September 1933, K. H. B.). Gen. CASTANOPHLEBIA Brnrd. 1932. Barnard, loc. cit., p. 244. 1932. Ulmer, loc. cit., p. 214. This genus appears to be accepted by Ulmer. Unfortunately Ulmer received a reprint of my paper with the typographical error “long stalk of MA” instead of “long stalk of M” (see corrigenda). The beautiful May-fly described below is a typical Castanophlebia as regards the imago, but shows remarkable features in the nymph, which necessitate an emendation of the generic diagnosis. Nymph.—Maxillary palp with 3rd joint either short, its junction with 2nd joint transverse, or elongate with oblique junction. Gills seven pairs, either double or single, but always slender and narrow lanceolate. The reduction of the inner lamina of the 7th gill and its occasional complete absence in calida, seems to indicate the course of evolution of the new species. Genotype: C. calida Brurd. Castanophlebia albicauda nu. sp. (Fig. 7.) Imago.—7-8 cross-veins before bulla, 10 ($) 14-15 (9) in ptero- stigmal area, the latter oblique, straight or slightly sigmoid, with occasional Y-forking on costa; subcostal 16-17 (2); between R,.; and MA, 8 (3) 9-10 (2); between MA, and M,., 9 (g) 9-10 (9); 1-3 (4) cross-veins between IMA and MA, and MA, (not counting proximal connecting veins); IR, usually not connected with Rs. Hind-wing as in calida (not always so many cross-veins in subcostal area as represented in fig. 36, 6). Tenth sternite, penis, and forceps 632 Annals of the South African Museum. 3S as in calida, each lobe of the penis with a small knob in middle, which appears as if it could be extruded into an acute process as in calida. Ventral plate of 2 short, broader than long, with deep excision as in calida. Fore-wing: ¢ 9 mm., 29 11 mm. Castaneous, occiput, and prothorax slightly paler; a dark brown bar on lateral margin of scutum as in calida; abdominal segments Fie. 7.—Castanophlebia albicauda n. sp. Fore-wing (2), and maxilla. 1-6 3g, 1-7 @ castaneous, the junctions of the segments and the pleural junctions of tergites and sternites pale, segments 7-9 3, 8 and 92 pale fawn. Legs and cerci pale fawn, uniform. Neuration fulvous-castaneous, costal and subcostal areas in pterostigmal region semi-opaque. Neuration in $ much paler and less conspicuous than in 9, the cross-veins especially faint. Subimago similar but paler. Nymph.—As in calida, but the labrum more angular, the spines on front margins of femora acute. The maxillary palp as figured, the junction of 2nd and 3rd joints oblique, 3rd greatly elongate, with dense brush of setae along its inner margin, most of the setae apically falcate, with extremely minute setules on inner margin. Labial palp with 3rd joint not so slender relatively to the 2nd as figured for calida (even for calida fig. 87, g, shows the 3rd joint too slender). Seven pairs of gills, each gill consisting of a single narrow lamina, with unbranched trachea. Up to 10-11 mm., cerci 15-18 mm. Fulvous-castaneous, mature nymphs deep castaneous, pro- and meso-thorax slightly mottled dorsally. Abdominal segments 1-7 castaneous, segments 8-10 pure ivory-white. Ventral surface of thorax and abdominal segments 1-7 pale brownish, usually with the Descriptions of New Species of South African Alder-flies, etc. 633 3 dark spots as in calida. Femora ochraceous, tibiae and tarsi paler. Cerci white. Anterior gills greyish, with dark tracheae, the posterior ones white with inconspicuous tracheae. Maxillary palp with 2nd and 3rd joints pale. Localities —Gt. Winterhoek Mts., Tulbagh, 4000-5000 feet (end September 1932, nymphs, K. H. B., beginning of November 1932, nymphs and adults, K. H. B. and H.G. W.); Witte River, Wellington Mts. (October 1933, H. G. W., nymphs); Bosch Kloof, Keeromberg, Worcester (end of January 1933, half-grown nymphs, H. G. W.). Remarks.—A larger and apparently more local fly than the “Chestnut Dun” (C. calida). It is easily recognised by the white or pale apex of the abdomen, which is very conspicuous in the nymph (cf. similar coloration in nymphs of Lithogloea harrisoni, and Agrionid Dragon-fly nymphs from the Gt. Winterhoek Mts.). Under a hand-lens the head of the nymph with its maxillary palps projecting forwards resembles a dust-pan with a pair of hand-brushes; and the palps evidently function as brushes to sweep up particles of food-stuff into the mouth. The fly may be known as the Larger Chestnut Dun. Castanophlebia calida Brnrd. 1932. Barnard, loc. cit., p. 246, figs. 36, 37. Localities.—Hottentots Holland Mts. (November 1932, K. H. B., H.G. W., and C. W. T.); Tradouw Pass, Langeberg Range (February 1932, K. H. B., H. G. W., and C. W. T.); Meirings Poort, Zwartberg Range (Mebruary 1952, K. H. B., H. G. W., and C. W. T.); Gt. Winterhoek Mts. (September 1932, K. H. B., H. G. W., and C. W. T.); Porterville (August 1937, K. H. B. and A. C. H.); Upper Olifants R., north of Ceres (October 1937, K. H. B. and C. W. T.). Fam. EPHEMERELLIDAE. The discovery of a second species of Lithogloea necessitates another alteration in the family diagnosis, as the nymph of this n. sp. has only four pairs of gills on segments 1-4. Gen. EPHEMERELLINA Lest. 1932. Barnard, loc. cit., p. 347. 1932. Ulmer, loc. cit., p. 214. The generic diagnosis may be completed thus: fore tarsus ¢ slightly longer than tibia, which is twice as long as femur; Ist tarsal joint very 634 Annals of the South African Museum. short, 2nd longest, 3rd—5th decreasing, claws alike. Ventral plate 9 broader than long. In the subimago the fore-leg of 3 is no longer than that of the 9. Although the fly has not actually been bred, further collecting, especially on Table Mt., leaves no doubt that the nymph described and figured by me (loc. cit., p. 252, fig. 43, d-g) is correctly assigned. Both nymphs and flies have been found in the Gt. Winterhoek Mts. and on Table Mt.; and nymphs in several other localities, including the river system in which the type locality is situated. Whether there is more than one species in the Cape mountains is a moot point. The $¢ from the Gt. Winterhoek Mts. have a much more strongly bilobed penis than the single ¢ from the Wellington Mts. Ephemerellina barnardi Lest. 1932. Barnard, loc. cit., p. 251, fig. 42, e, and p. 252, fig. 48, d-g (nymph). Imago.—Some specimens collected in the Gt. Winterhoek Mts., 4000 feet (beginning November 1932, K. H. B. and H. G. W.), are Cc oe a b Fic. 8.—Ephemerellina barnardi Lest. a, ventral plate, 9. 6, penis. c, d, dorsal abdominal spinules of imago, lateral view and cross-section. very much darker than the original specimens. The whole body is dark Vandyke brown, with lighter marks laterally on the meso- thorax. Wings hyaline, slightly milky along the costal margin, longitudinal veins dark brown or sepia, the cross-veins very 1mcon- spicuous. | Penis somewhat hour-glass shaped, apically deeply bilobed. Ventral plate 2 broader than long, lateral margins feebly angular, and distal margin slightly concave. . Abdominal segments 2-6 each with a small transverse shicledaitie in the middle of the tergite, each bearing 2 small conical tubercles or spinules. Descriptions of New Species of South African Alder-flies, ete. 635 Fore-wing ¢ 9 mm., 2 10 mm. These specimens are smaller than the Gt. Winterhoek 2 (November 1916), but barring the colour are exactly similar. The ventral plate and the abdominal spinules were not mentioned in my description. The $3 differ in the penis from the Wellington Mts. ¢ figured by me. A subimago from this locality (Ist January 1934, H. G. Wood) confirms this figure, though a comparison between subimago and imago is not too reliable. The abdomen in the latter specimen is too mutilated to determine the presence of the spinules. Some 9° from Table Mt. (5th January, 2nd February, and 29th December 1933, K. H. B.) agree in the shape of the ventral plate and other features, but the coloration is distinctly castaneous or rufous. Fore-wing 10-11 mm. The original pair described by Lestage is not in the S.A. Museum (Barnard, loc. cit., footnote, p. 202), so that further comparison must wait until more material is obtained from the type locality. Nymph.—Nymphs similar to that described from the Cedarberg Mts. have been found in the Gt. Winterhoek Mts., end September 1932, K. H. B.; in the head-waters of the Palmiet River (Groenland Mts., south side of Viljoen’s Pass, Elgin, 29th October 1931, K. H. B., and Hottentots Holland Mts., November 1932 and January 1933, K. H. B. and H. G. W.); French Hoek Pass (head-waters of River Zonder End) December 1932, K. H. B. and H. G. W.; Table Mt., August-September 1933, K. H. B., half and three-quarters grown. The Elgin nymph is a large one measuring 12 mm., cerci 9 mm., 3 mm. across the wing-cases and 2:75 mm. high at the same level. The largest from the Gt. Winterhoek Mts. is 10 mm. in body length, and is evidently in a younger stage as the projecting tips of the wing-cases are not so prominent. The larger nymphs are very solidly built, plump and high; they can be picked off the rock with thumb and finger without damage. They inhabit only those portions of the mountain streams which are shallow and where the water is always rushing tumultuously over the rocks and stones. Gen. LirHociora Brnrd. 1932. Barnard, loc. cit., p. 252. The diagnosis stands, except that the penis is not always broad in contrast with that of Ephemerellina, and that gills are present in the nymph on segments 1-6 or 1-4. The confusion of two species under the name harrison: was due to 636 Annals of the South African Museum. my collecting very young nymphs from Table Mt. streams and referring them without actual breeding to the Drakenstein specimens bred by Mr. Harrison. We were led to think that there might be two species in the Drakenstein material by differences in coloration; but this was definitely proved not to be so, as stated on p. 255, lines 5-6. This has been further confirmed by breeding from nymphs of another different colour-pattern. Barring a slight difference in size, and the penis of the dg, there is no distinction between the imagos of the two species; but the nymphs are quite distinct. Genotype: L. harrison: Brnrd. Inthogloea harrison Brurd. 1932. Barnard, loc. cit., p. 253, figs. 42, a-d, 44 (harrisoni part; not the young nymph described and figured). The locality Table Mt. should be deleted, as the true harrisoni has not yet been found in the Cape Peninsula. Add the following localities: Gt. Winterhoek Mts., Tulbagh (November 1932, K. H. B. and H. G. W.); Palmiet River, Kleinmond (December 1932, H. G. W.); Hottentots Holland Mts. (November 1952; Ke Be B. and. Ba Ge W.): In the nymph the following colour varieties have been observed: From Stellenbosch Mr. Harrison has collected nymphs of a uniform very dark Vandyke brown, almost black. In the Gt. Winterhoek Mts. nearly all the nymphs have definite white spots; the ground colour in the younger nymphs is mottled as originally described (p. 255), but in the larger ones becomes very dark sepia, almost black. The following parts are pure white: 2 small dots on head behind lateral ocelli and contiguous with inner margins of eyes, a large oval median spot on anterior part of mesonotum, whole of 10th tergite and the lateral projections of the 9th segment, the distal ends of all the femora, and the distal half of the cerci except the extreme tips. In life the contrast between the blackish ground- colour and the white thoracic spot, 10th abdominal segment and “knees”? is very striking. The nymphs are very conspicuous on the stones and are not in the least protectively coloured as are the nymphs of the following species. It is interesting to note that in the same locality the nymphs of one of the Agrionid Dragon-flies has the same scheme of coloration: blackish with white terminal segment and white knees; and also the nymphs of Castanophlebia albicauda (cf. supra). Descriptions of New Species of South African Alder-flies, etc. 637 Lnthogloea penicillata n. sp. 1932. Barnard, loc. cit., p. 255, fig. 43, a-c (harrisont part, young nymph). Imago.—As in harrisoni, but slightly smaller, and penis narrower (cf. fig. 42, e, of EL. barnardi). Ventral plate 2 as in harrisoni. Fic. 9.—Lithogloea penicillata n. sp. Nymph. a, 6, left and right mandibles. c, d, e, gills on 2nd, 3rd, and 4th segments, ventral view, the external margin on the left in each case. Body ¢ 5-6 mm., 2 6 mm.; fore-wing $ 5-5-6 mm., 2 6-5 mm. Nymph.—Full-grown nymph resembling the young described and figured. Abdomen triquetral in cross-section, medio-dorsally slightly carinate, but none of the segments raised or imbricate. Only three pairs of lamellate gills, those on segments 2 and 3 being double, that on segment 4 single. Guill on segment 1 as in harrisons. Mandibles even more oblong than in the young, the apices being squarely truncate, the outer angle usually worn quite blunt. Cerci feebly setose. Up to 5-6 mm. 638 Annals of the South African Museum. Ochraceous or straw-coloured, with faint darker markings on thorax, abdomen with a series of dark spots dorso-laterally, that on segment 2 being the largest and most conspicuous. Localities.—Streams on slopes of Table Mt. (Blinkwater, Platteklip, and Skeleton Ravines, Orange Kloof) (November to March, K. H. B.); Amandel River, Hex River, Worcester (October 1931, mature nymphs from trout stomachs, A. C. H.); Gt. Winterhoek Mts., Tulbagh (September 1932, K. H. B., and November 1932, K. H. B. and H. G. W.); Porterville (August 1937, A. C. H.); Seven Weeks Poort, Ladismith, Cape (September 1933, nymphs and subimagos, K. H. B.); Prince Albert, mouth of Zwartberg Pass (September 1933, nymphs, KeUEE RS: Fam, ECDYONURIDAE. Afronurus peringuey? (EK. P.). (1932. Barnard, loc. cit., p. 255. Localities.—Yarrow River, Karkloof, Natal (18th March 1934, L. A. Day); Umzimkulu River (April 1936, nymphs, Dr. Sampson); Cathkin Peak, 6000 ft., Drakensberg, Natal (January 1938, nymphs, Dr. R. F. Lawrence). Afronurus harrisont Brurd. 1932. Barnard, loc. cat., p. 257. Localities—Du Toits Kloof, Rawsonville (25th March 1932, 9 imago, K. H. B.); Michell’s Pass, Ceres (September 1932, nymphs, K. H. B.); Seven Weeks Poort, Ladismith, Cape (September 1933, nymphs, K. H. B.); Upper Olifants River, north of Ceres (October i9o1, 2 imate; kK. EL B))- TRICHOPTERA. A considerable amount of material has accumulated since the publication of “South African Caddis-flies”” (Barnard, Trans. Roy. Soc. 8. Afr., xxi, pt. 4, 1934), comprising undescribed species and numerous new locality records due to collecting by Mr. H. G. Wood and myself. Only the most important localities are recorded, namely, those extending the already recorded distribution or linking up widely separated localities. Descriptions of New Species of South African Alder-flres, etc. 639 INAEQUIPALPIA. Fam. SERICOSTOMATIDAE. Gen. GorRopES Ulmer. 1907. Ulmer, Cat. Coll. Selys. fasc. 6, p. 37. 1927. Martynov, Ann. Mus. Zool. Ac. Sci. U.R.S.S., xxvii, p. 471 (Crunobiodes). 1939. Mosely, Ruwenzori Exp. (Brit. Mus.), 111, p. 4. This genus is characterised by a fold in the fore-wing of the 3, and a certain type of genitalia. As the South African species conform with the genus Groerodes as redefined, Mosely (loc. cot., swpra, p. 2) has transferred it, together with other African species, to Goerodes, leaving Crunoeciella with a single Madagascan species. Goerodes caffrarvae (Brnrd.). (Fig. 10.) 1934. Barnard, loc. cit., p. 302, fig. 2 (3) (Crunoeciella c.). Imago 2.—Fore-wing narrower and more oval than in g, covered with hairs only, no scales. Venation similar to that of sjoestedti Ulmer, a Fic. 10.—Goerodes caffrariae (Brnrd.). a, fore- and hind-wings 9, dotted line representing forking of R, in one of the specimens. 8, c, lateral and ventral views of 2 genitalia. but fork of Cu, midway between distal cross-vein of thyridial cell (M,,,—Cu,) and cross-vein Cu,,—Cu,. Basal cross-vein Cu, —Cu, perpendicular and situated at the fork of M and Cu, (this cross-vein 640 Annals of the South African Museum. was omitted in the figure of the g). Hind-wing as in sjoestedti; Sc and R separate throughout (as is probably the case in sjoestedti, though the figure shows them fused proximally). One specimen has R, forked in both fore-wings, as indicated by dotted lines in fig. 10, a. On the membrane the cross-vein R;—M, is surrounded by an oval clear space, and a clear space along M and M,,,, with a slight expansion at forking of M,,, and M,,,. Coloration as in g. Fore-wing 6-6-5 mm. Genitalia, terminal tergite projecting rather prominently, in dorsal (or ventral) view rounded-truncate; terminal sternite apically tri- lobed; penultimate sternite forming a flat semicircular plate. Locality.— Grahamstown (Carl’s Rust and Paradise Kloof) (J. Hewitt, 2 99). Dyschimus collyrifer Brurd. 1934. Barnard, loc. cit., p. 306, fig. 5, a-e. - Locality.—Malgas River, George (K. H. B., November 1938, 1 3). Dyschimus ensifer Brnrd. (Fig. 11.) 1934. Barnard, loc. cit., p. 306, fig. 5, fEk (3). Imago, 2.—Venation; in fore-wing Cu,, complete to margin, cross- vein from fork of Cu, nearly straight to margin, joining Cu, and 1A, neither of the latter veins continued to margin. In hind-wing distal portion of Sc running obliquely upwards and forwards to C (not recurved as in 3g), R incomplete basally and distally, discoidal cell present, fork 3 shortly stalked. Genitalia, supra-anal plate prominent, in dorsal (or ventral) view broadly subtriangular with apical notch; subanal plate small, longer than wide, apically rounded. Fore-wing 10 mm. Colour as in ¢, antennae dark umber. Locality.— Grahamstown (Carl’s Rust) (J. Hewitt, 1 9). Remarks.—In the venation of the hind-wing and the genitalia, in both sexes, this species is somewhat isolated from the other two species of the genus. | Rhoizema montanum Brurd. 1934. Barnard, loc. cit., p. 311, fig. 8. Locality.—Prince Alfred’s Pass, Avontuur, Outeniqua Range, 3500 feet (K. H. B. and C. W. T., January 1940, 1 3). Descriptions of New Species of South African Alder-flies, etc. 641 Petroplax phleophila Brnrd. 1934. Barnard, loc. cit., p. 319, fig. 14, a-g. Locality.— Willow R., north slopes of Cockscomb Peak (Uitenhage Div.) (K. H. B. and C. W. T., October 1938). Fie. 11.—Dyschimus ensifer Brnrd. a, fore- and hind-wings, 2, dotted line in fore-wing representing a fold. 6, c, lateral and ventral views ° genitalia. Petroplax anomala n. sp. (Fig. 12.) Imago, §.—Head and thorax dark umber with blackish hairs. Legs and antennae fuscous, the latter unicolorous. Wing membrane with the anastomosis and a streak along M clear white; hairs on fore-wing sepia-brown with golden-brown hairs intermixed, androconia blackish; hairs on hind-wing sepia, fringe blackish. Fore-wing 7 mm. Venation.—M prolonged basally and without apparent connection with Cu,; between these two veins a fold forming a deep groove on lower surface; on anterior side of Cu, at base on lower surface a series of androconia, extending about 1 mm. in length; Cu, and 1A obsolete, represented by a fold which does not extend more than 4 length of wing; thyridial cell resembling that of the 9 of the other species; two of the branches of M forming a reversed fork or cell which is sym- metrical in both right and left wings. Hind-wing with an indication of a cross-vein between M and Cu,,. VOL. XXXII, PART 6. 44 642 Annals of the South African Museum. Gemtaha, 9th and 10th tergites fused, the portion representing the 9th somewhat raised and medianly scabrous, the distal portion (10th) apically cleft, the lower lateral margins with fine recurved serrulations; penis with 2 stout subterminal spiniform processes, whose apices are slightly upturned, the membranous apex bilobed; clasper spatulate, hollowed on inner side, apex rounded (not bifid); 9th sternite with Fic. 12.—Petroplax anomalan.sp. a, fore- and hind-wings ¢ (androconia actually on lower surface). 6, diagrammatic cross-section of fore-wing near base, showing androconia attached to lower surface of Cu,. c, d, e, lateral, dorsal and ventral views of 3 genitalia. /, lateral view of apex of penis. stout process, the apex slightly excavate; lamina on 7th sternite transversely crescentic. Locality.—Grahamstown (Carl’s Rust) (J. Hewitt, 1 3). Remarks.—In general features this species unmistakably belongs to the genus Petroplax (Barnard, loc. cit., p. 316), but it is markedly distinct from the other three species in the modification of the fore- wing. The patch of androconia is much larger than in the other species. In consequence of the suppression of Cu, and 1A in the fore-wing, M and Cu, are pushed farther towards the hind margin, and the area between Rs and M becomes very broad. Descriptions of New Species of South African Alder-flies, etc. 643 Cheimacheramus caudalis Brurd. 1934. Barnard, loc. cit., p. 316, fig. 12. The Palmiet River specimens (H. G. W., December 1933) are even smaller than the Tradouw ones, viz.: fore-wing ¢ 4:5 mm., 2? 4:75 mm. The posterior projection of the ¢ claspers is more slender, the prongs of the fork of 10th tergite longer and less divergent, and the ventral process of 9th sternite bluntly pointed. Locality.—Upper Olifants River, north of Ceres (K. H. B., October iat, 1 3). Barbarochthon brunneum Brnurd. 1934. Barnard, loc. cit., p. 321, figs. 15 and 1, a. The specific identity of the Table Mt. form with that from the type locality has been established by breeding. In fresh specimens, especially noticeable in those collected in the Hottentots Holland Mts., there is a pale or white band along the side of the abdomen. Locality.—Upper Olifants River, north of Ceres (K. H. B. and C. W. T., October 1937, gd, 29). AEQUIPALPIA. Petrothrincus triangularis (Hagen). 1934. Barnard, loc. cit., p. 325, fig. 18, g-g. Cases were found (K. H. B., September 1933) in the Valsch Gat stream on the Ceres (north) side of Matroosberg in the Hex River Mts. Fam. LEPTOCERIDAE. Pseudoleptocerus cupreus Brnrd. 1934. Barnard, loc. cit., p. 329, fig. 19. Mosely (Ann. Mag. Nat. Hist. (10), xi, p. 541, figs. 6-10, 1933) has redescribed P. schoutedeni Navas. These two species would seem to be closely allied, but Mosely says nothing about the spines on the penis; the shape of the process on the 9th sternite in schoutedent is more like that of cwpreus var. subfuscus (fig. 19, 7). Leptocerus schoenobates Brurd. 1934. Barnard, loc. cit., p. 333, figs. 21, 22, a—I. Localities.—Upper Olifants River, north of Ceres (K. H. B. and C. W. T., October 1937, $3, 99); Clarkson, between Humansdorp and Tsitsikama (C. W. T., November 1938, 32). 644 Annals of the South African Museum. Leptocerus bibulus n. sp. (Fig. 13, a-d.) . Imago.—In fore-wing anastomosis between R, 43 and Cu, straight, or almost so, slightly oblique; stalk of fork 1 subequal to upper Fic. 13.—Leptocerus bibulus n. sp. a, 6, lateral and dorsal views of ¢ genitalia. c, penis. d, ventral view of one of the claspers. Leptocerus spatulan. sp. e,f, lateral and dorsal views of 3 genitalia. g, ventral view of penis and lower sheath. Leptocerus dieselin. sp. h, i, lateral and dorsal views of 3 genitalia. branch (R,) of fork 1; apical cell 4 g¢ with stalk almost or quite obsolete, sessile on anastomosis; stalk of cell 4 9 (M,,,) subequal to stalk of fork 1. In hind-wing apical fork 1 very short, about + length of its stalk; stalk of apical cell 4 very short, not more than 4 the cross-vein between Rs and M, 7.e. the cell almost sessile. Descriptions of New Species of South African Alder-flies, etc. 645 Genitalia $ in general similar to those of potes Brunrd. (loc. cit., p. 341), but distinguished by the large dorsal plate, different shape of the two dependent processes, spination of penis, and claspers. Fore-wing 5-6 mm. Coloration as in potes. Locality.—French Hoek Pass (K. H. B. and H. G. W., October 1933). Leptocerus spatula n. sp. (Fig. 13, e-g.) Imago.—In fore-wing anastomosis between R,,, and Cu, straight, oblique; stalk of fork 1 equal to upper branch (R,) of fork; apical cell 4 § with short stalk not quite equal to cross-vein between M and Cu,. In hind-wing apical fork 1 short, less than 4 length of its stalk, much shorter in the Clanwilliam specimens; stalk of apical cell 4 about 14 times length of cross-vein between Rs and M. Genitalia $ somewhat like those of tuckeri Brnrd. and variety, but distinguished by the extraordinarily long spinous processes and the more lamellately expanded claspers. Fore-wing 9mm. Head and thorax umber brown, abdomen paler; wings pale brown, neuration slightly darker. Locality.—French Hoek Pass (H. G. W., October 1933, 1 3; K. H. B., October 1936, 1 3g); Jan Diesel’s River, Clanwilliam (K. H. B., September 1936, 2 3g); Upper Olifants R., north of Ceres (esrb. and C. W. T., October 1937, gd, 29). Leptocerus tabularis Brurd. 1934. Barnard, loc. cit., p. 337, fig. 24. Top of Kasteel’s Poort, Table Mt. (7th May 1935, K. H. B.), in very dry places. All attempts to locate the larva of this species have been unsuccessful. Leptocerus harrisont Brnrd. 1934. Barnard, loc. cit., p. 337, figs. 25, 26. Jan. Diesel’s River, Clanwilliam (K. H. B., September 1936, 1 3). Leptocerus diesels n. sp. (Fig. 13, h, 2.) Imago.—In fore-wing anastomosis between R,,, and Cu, straight, oblique; stalk of fork 1 subequal to upper branch of fork; apical cell 4 3 with stalk 14 times length of cross-vein between M and Cu,. SS 646 Annals of the South African Museum. In hind-wing apical fork 1 absent; stalk of apical cell 4 subequal to cross-vein between Rs and M. Genitalia g—a transverse band projecting laterally in a short setose process on each side, below this a pair of strongly chitinised processes, with sinuous dorsal margin in lateral view, and a short retrorse spinous projection on lateral margin; penis very broad in dorsal view, with 4 weak spines dorsally; no titillators; claspers relatively small, twisted so that the upper margin (in lateral view) becomes the front margin, in general somewhat like those of L. cedri (Barnard, loc. cit., fig. 27, h, 2). Fore-wing 6-6-3 mm. Colour of fore-wing uniform greyish. Locality —Jan Diesel’s River, Clanwilliam (K. H. B., September 1936, 4 33). Remarks.—At first sight resembling L. cedri in the genitalia, but without the small medio-dorsal bifid process of the latter species, and apical fork 1 in hind- wing completely absent. Leptocerus cedri Brnrd. 1934. Barnard, loc. cit., p. 341, fig. 27, h, +. Upper Olifants River, north of Ceres (K. H. B. and C. W. T., October 1937, 3d, 29). Leptocerus potes Brnrd. 1934. Barnard, loc. cit., p. 341, fig. 28, a-g. Locality.—Kaimans River, near George (K. H. B., November 1938, 3d, $9). Leptocerus longistylis Brnrd. 1934. Barnard, loc. cit., p. 341, fig. 27, 7-1. Locality.—Upper Olifants River, north of Ceres (K. H. B. and CG. W. T., October 1937, $3). Leptocerus securis Brurd. 1934. Barnard, loc. cit., p. 3438, fig. 30, a, 6. Localities.—Wellington Mts. (H. G. W., October 1933); French Hoek Pass, east side (K. H. B., October 1936, 33, 99); Pakhuis Pass, Clanwilliam (K. H. B., September 1936, 33, 99); Willow R., north slopes of Cockscomb Peak (Uitenhage Div.) (K. H. B., October 1938); George and Kaimans River (K. H. B., November 1938). In the George and Kaimans River specimens (1 ¢ from each locality), the medio-dorsal process, instead of being laterally com- Descriptions of New Species of South African Alder-flies, etc. 647 pressed, forms a dorso-ventrally flattened process. When more material is available, this may prove to be of varietal or specific value, though in other respects the genitalia do not seem to show any differences from the typical form. Leptocerus stephanus Brnrd. — 1934. Barnard, loc. cit., p. 348, figs. 22, k, 31, d, e. Localities.—Tradouw Peak, Langeberg Range (K. H. B., January 1935); Meirings Poort, Spitzkop, Zwartberg Range (K. H. B. and H. G. W., January 1935); Rust en Vrede, near Cango, Zwartberg Range (H. G. W., January 1938). The Zwartberg Range specimens have the 10th tergite with out- standing lateral spines as figured in fig. 31, e. Leptocerus elaphus Brnrd. 1934. Barnard, loc. cit., p. 348, fig. 31, f, g. Localities.—River Zonder End Mts., 4000 feet (K. H. B., January 1934); du Toits Kloof, Wellington Mts. (H. G. W., March 1934). Leptecho scirpi Brurd. 1934. Barnard, loc. cit., p. 349, figs. 22, p—s, 31, 7, k. Localities —River Zonder End Mts., 4000 feet (K. H. B., January 1934); French Hoek Pass (K. H. B. and H. G. W., April 1935). Gen. LeprocEeRINA Mosely. The suggestion (Barnard, loc. cit., p. 350) that furcata Mosely was synonymous with ugandanus Ulmer is confirmed by Mosely (Ann. Mag. Nat. Hist. (10), xvu, p. 429, 1936) after examination of Ulmer’s type. Oecetis afra Mosely. 1934 (May). Mosely, Eos., ix, p. 24, figs. 14-16. It is possible that kunenensis Brnrd., described from the 9 only, will prove synonymous with the Northern Rhodesian afra. Oecetis lucipetens n. sp. (Fig. 14.) - Imago.—Tibial spurs 1, 2, 2. Fore-wing with Sc and R thickened and connected by an oblique cross-vein distally, Sc only faintly 648 Annals of the South African Museum. continued to costal margin; apical fork 1 sessile, fork 3 shortly stalked; anastomosis stepped; hind-wing narrow, fork 1 absent, M and Cu, forked. Genitalia 3.—9th (or 9th and 10th?) tergite ovoid, with 2 minute setose apical lobes, and 2 subapical groups of setae on a slightly Fic. 14.—Oecetis lucipetensn. sp. a, fore- and hind-wings. 0, c, dorsal and lateral views of $ genitalia. raised boss-like projection; penis short, curved downwards; clasper < distally abruptly narrowed, upper margin proximal to the narrow apex with a fringe of rather long setae. Fore-wing 6 mm., antenna 18 mm. Head and thorax fulvous, wings pale with sepia venation, in fore- wing a small dot formed by black hairs at base of fork 1, at base of Descriptions of New Species of South African Alder-flies, etc. 649 thyridial cell, and at junction of Cu,,, with hind margin. Legs and antennae pale. Locality.—Olifants River, Citrusdal, Clanwilliam Division, Cape (K. H. B., 2 $3, 3 99, 13th February 1939). Remarks.—Although this is rather an aberrant species with features reminiscent of Ptochoecetis, I am referring it for the time being to Oecetis on the advice of Mr. Mosely, to whom I submitted specimens. The flies were caught round the lamp on the bank of the river; the taxonomist also seeks light on many puzzling cases. Gen. Homit1a McLach. 1934. Barnard, Tr. Roy. Soc. 8. Afr., xxi, p. 354. 1936. Mosely, Ann. Mag. Nat. Hist. (10), xvu, pp. 429, 432. Mosely describes two species, one of which, H. lomza, is found at Sabie, Transvaal, and in Sierra Leone. He does not give the tibial spur formula for either species. In 1934 under Leptocerus fissus (loc. cit., p. 336), I remarked that as this species was stated to have the formula 2, 2, 2 in both sexes, it could not be included in Homilia in spite of the similarity of the $ genitalia. The two n. spp. described below have a 2, 2, 2 formula, in the case of elephas in both sexes; in other respects they are essentially repre- sentatives of the genus Homilia. Although the tibial spurs are cus- tomarily regarded as of considerable taxonomic importance, I am including both these n. spp. in Homilia, and suggest that L. fissus may eventually also be transferred to this genus. Of the two species here described, knysnaensis agrees with vetulata Brnrd. and malia Mosely in lacking the lateral appendages of the 10th tergite; and elephas is more closely allied to lomia Mosely than to vetulata Brnrd. from the Kunene River. Homuilia elephas nu. sp. (Fig. 15, d-f.) Imago.—Tibial spurs 2, 2, 2; the two on the fore-tibia are subequal and easily distinguished in both sexes when the limb is cleared. Venation as in lomia Mosely (loc. cit., fig. 5) in both sexes, stalk of apical fork 1 shorter than its upper branch, and shorter than stalk of apical cell 4. Genitalia §.—9th tergite, as in vetulata and lomia, with 2 patches of minute spinules or granules, preanal (superior) appendages elongate, slender, sparsely setose; 10th tergite (upper penis sheath) elongate, 650 Annals of the South African Museum. deeply cleft, without prongs, apices of lobes acute, near the base of the 10th tergite on each side there arises an elongate process (lateral appendage) (not found in either vetulata or lomia), sinuous, with setose apex; penis robust, in side view bent downwards at right angles, deeply grooved dorsally and posteriorly; clasper in side view widely bifid, proximal upper branch ending in a sharp, incurved, strongly chitinised apex (cf. lomia), distal branch slender, elongate, apically incurved, a slender process arising from inner side, lower margin with a tooth-like process. Fore-wing 6-7 mm. Antennae 14 mm. Head and thorax fuscous with black and white hairs. Wings umber brown, somewhat speckled apically with white hairs; in fore-wing a whitish costal mark proximal and another distal to the pterostigmal area, which in the denuded wing by transmitted light is suffused, a whitish mark at junction of Cu,4, with hind margin. Tarsi and antennae annulate. -Locality.— Olifants River, Citrusdal, Clanwilliam Division, Cape (K. H. B., 138th February 1939, 2 33, 3 99). Homilia knysnaensis n. sp. (Fig. 15, a-c.) ~ Imago, 3.—Tibial spurs 2, 2, 2. Fore-wing not scaly, with stalk of apical fork 1 subequal to its upper branch, stalk of apical cell 4 shorter; thyridial cell longer than discoidal cell; anastomosis nearly perpendicular to long axis of wing, cross-veins R,—M,,, in line, distal to cross-vein M—M,,,. Gentalia 3, preanal appendages somewhat clavate, setose; 10th tergite an elongate process, its depth basally in lateral view greater than its width in dorsal view, bifurcate for about half its length, with a pair of movable slender processes dorsally; no lateral appendages; penis stout, curved, with subterminal lateral expansions when viewed dorsally; claspers with an internal ventral bifid process at base, one lobe of which bears short spines, the other long setae, on its ventral surface a small digitiform process, with apical spinule, the longer external arm of the clasper apically trifid, the upper two processes being movable. Fore-wing 8mm. Antennae 18 mm. Head and thorax piceous-brown. Fore-wing dark sepia with 2 cuneiform white marks distally from the costa inwards, subcosta and hind-margin as far as end of Cu, pale, a pale speck also at junction Descriptions of New Species of South African Alder-flies, etc. 651 of cross-vein M—M,,, with M. Antennae pale sepia, annulated. Legs pale ochreous. Fig. 15.—Homilia knysnaensis n. sp. a, fore- and hind-wings. 0, lateral view of 6 genitalia. c, dorsal view of same, with penis drawn separately to same scale, and ventral basal process further enlarged. Homilia elephas n. sp. d, lateral view of $ genitalia.. e, dorsal view of same, the appendages of 9th tergite displaced, lateral appendages omitted on left side, inner process of clasper omitted on right side. /, posterior view of penis (on a larger scale). Locality.—Goukama River, near Knysna (K. H. B., 5th November 1938, 1 3). Remarks.—The colour pattern of this single specimen suggests its inclusion in the genus Homilia, apart from other features. 652 Annals of the South African Museum. Gen. ADIcELLA McLach. 1932. Barnard, loc. cit., p. 357. 1932. Ulmer, Peking Nat. Hist. Bull., vis. 1936. Martynov, Rec. Ind. Mus., xxxviii, p. 279. Adicella monachus n. sp. (Fig. 16.) Imago.—Venation like that of pulcherrima Ulmer, and biramosa Mart. (loc. cot., 1936); in fore-wing M, and M, arising together at upper distal corner of thyridial cell, anastomosis almost straight, R, thickened distally, stalk of fork 1 subequal to its upper branch : ve! Ny Ly}; b Fic. 16.—Adicella monachus n. sp. a, b, lateral and dorsal views of ¢ genitalia. (R,); in hind-wing fork of M nearer to level of fork of R than in biramosa. Gemtalia 3, 9th tergite transverse, preanal appendages ovate, strongly setose, 10th tergite forming a broad, cowl-like process, apically bilobed, with a short transverse median ridge, penis stout, simple, strongly curved, clasper uniramous, upturned, somewhat twisted, with a few short spines on apex. Fore-wing 5-5 mm. Head and thorax ochreous, with greyish hairs, antennae pale ochreous, faintly annulate, wings grey, anastomosis on fore-wing faintly indicated by a whitish streak. Larva.—Similar to that of filicornis (Rousseau, 1921, p. 636, fig. 219). Case composed of small pieces of vegetable matter forming a tube more or less 4-sided in cross-section; sometimes sand-grains are incorporated; one case is composed entirely of sand-grains, but is circular, not 4-sided in section. Descriptions of New Species of South African Alder-flies, etc. 653 Locality. Natal (Natal Mus.). Remarks.—As regards the ¢ genitalia this species is very like syriaca Ulmer (1907, Notes, Leyden Mus., xxix, p. 52, figs. 71, 72). In lateral view (fig. 71) there is very little difference; but the claspers are described and figured (fig. 72) as 2-jointed. One suspects that this is an error, as they appear to be only 1-jointed in fig. 71. In syriaca the fore-wings are yellow, and the species is stated to be the palest or brightest of the species then known. This coloration does not fit the present species. Although syriaca has been recorded from widely separated localities (Lestage, 1919, also Barnard, loc. cit.), it seems better to regard the Natal form as a separate species until a revised description of syriaca has been given. Potamoryza modesta Brurd. 1934. Barnard, loc. cit., p. 352, fig. 33. Larvae and pupae (imagos dissected from latter) were collected at the northern entrance to the Zwartberg Pass, Prince Albert (K. H. B., September 1933); Willow R., north slopes of Cockscomb Peak (Uitenhage Division); Patentie (Gamtoos valley); and Knysna (K. H. B., October-November 1938). Fam. HYDROPSYCHIDAE. Gen. CHEUMATOPSYCHE Wallengr. 1891. Wallengren, K. Sv. Ver. Ak. Handl., n.s., xxiv, pp. 138, 142. 1934. Barnard, loc. cit., p. 360. (Hydropsychodes.) 1936. Mosely, Ann. Mag. Nat. Hist. (10), xvu, p. 438. H. maculata (Mosely). 1934 (May). Mosely, Eos., ix, p. 22, figs. 11-13. 1934 (July). Barnard, loc. cit., p. 362, figs. 37, a-f, 38 (lateralis). It is evident that Mosely and I have described forms of the same species. This very common and somewhat variable Western Province species must therefore bear the name proposed by Mosely. In Ann. Mag. Nat. Hist. (10), xv, p. 231, 1935, Mosely gives a list of the African species of this genus. 654 Annals of the South African Museum. Leptonema natalense Mosely. 1933. Mosely, Rev. Gen. Leptonema, Brit. Mus., p. 24, figs. 43-48. 1934. Barnard, loc. cit., p. 370, fig. 41, f-2 (occidentale, non Ulmer). By an oversight the correction was not made in my paper. In this paper delete the words: ‘Distribution: Cameroons.” Macronema natalense Ulmer. 1934. Barnard, loc. cit., p. 372, fig. 42, a-e. Localities —Kromme River at Assegai Bush (C. W. T., Ist November 1938, 1 2); Goukama River, near Knysna (K. H. B., 5th November 1938, 33). Fam. POLYCENTROPODIDAE. Dipseudopsis capensis Wlkr. (Fig. 17.) 1934. Barnard, loc. cit., p. 374, fig. 43, a-t. Pupa (described from empty skins).—Mandibles bulbous at base, distally very slender, curved, terete. Presegmental dorsal plates on a b & Fic. 17.—Dipseudopsis capensis Wlkr. Pupa. a, mandible. 6, dorsal view of 5th and 6th abdominal segments, showing postsegmental plates on 5th and presegmental plates on 6th segment, with postsegmental plate in lateral view and further enlarged. c, dorsal view of 9th abdominal segment. segments 3, 4, and 6-8, sometimes (in 1 out of 13 specimens) also on segment 5; postsegmental plates dorso-laterally on segment 5. The presegmental plates bear spines, 6-8 on the plates on segments 3 and 4, 4-5 on those on segment 8, 10-12 on those on segment 7, and about 14 on those on segment 6; the postsegmental plates are oval in shape, nearly vertically upstanding, and with 6-8 strong spines on upper edge. 9th segment with a transverse dorsal series of 8 chitinised conical projections, the outermost (lateral) one on each side smaller than the others, each bearing a single rather long seta; Descriptions of New Species of South African Alder-flres, etc. 655 posterior margin arcuate, postero-lateral corners with several setae of varying length. Localities.—Dolley Reservoir, Zwartkops River, Uitenhage (K. H. B. and C. W. T., 28th October 1938, 3, 99, and empty pupal skins); Goukama River, near Knysna (K. H. B., 5th November 1938, 2 99). Fam. PSYCHOMYIDAE. Eenomus similis Mosely. 1934. Barnard, loc. cit., p. 378, fig. 45, e, f. Locality.—Grobelaars R., Schoemann’s Poort, north of Oudtshoorn (K. H. B., October 1937). One g agreeing with Mosely’s description, but the inner and distal margins of the claspers are very distinctly scalloped or crenulate. Ecnomus natalensis Ulmer. 1934. Barnard, loc. cit., p. 378, fig. 45, g—2. Localityi— Knysna (K. H. B., November 1938, 3 39). Gen. PADUNIELLA Ulmer. 1912. Ulmer, Notes, Leyden Mus., vol. xxxv, p. 80. 1934. Martynov, Tabl. Anal. Fauna U.R.S.S., vol. xi, p. 209. 1936. Mosely, Ann. Mag. Nat. Hist. (10), vol. xvii, p. 444. Tibial spurs g 2, 4, 4,2 1, 4, 4. No ocelli. Maxillary palp ¢ 6, 2 5-joited. Labial palp ¢ 4, Q 3-jointed. Wings narrow (much resembling those of Hydroptilids), especially the hind-wing, lanceolate, fore-wing with anal lobe, and forks 2, 3, 4, 5, hind-wing with forks 2 and 5. Originally instituted for a Javanese species. Padumeella capensis n. sp. (Fig. 18.) Imago.—Venation very similar to that of africana (Ulmer) (ef. Mosely, loc. cit., fig. 30), but venation very difficult to trace. Genitalia 3. 9th tergite in dorsal view triangular with truncate apex, upper appendages ovate both in dorsal and lateral views, with slightly scalloped or serrate edges, setose; penis in lateral view apically enlarged, a slender dorsal process arising near base; claspers oblong, apically bifid, with a setose lobe on middle of upper margin. 656 Annals of the South African Museum. Fore-wing § 4, 9 4-5 mm. Head, thorax, and abdomen pale brown, fore wing fawn brown, darker (sepia) in the Cu and anal Fic. 18.—Paduniella capensis n. sp. a, fore- and hind-wings. 8, c,d. lateral, dorsal and ventral views of § genitalia. e¢, penis. areas where the scaling is denser; antennae annulate, legs pale ochreous. Locahty.—Schoemann’s Poort, north of Oudtshoorn (K. H. B. and C. W. T., October 1937, 3d, 22). Remarks.—Differs from africana (Ulmer), from Cameroon and Sierra Leone, in the shape of the upper appendages and the claspers. Fam. PHILOPOTAMIDAE. Chimarrha ambulans Brnrd. 1934. Barnard, loc. cit., p. 382, figs. 47 and 1, b. Localities.—Meirings Poort, Zwartberg Range (K. H. B., January 1935); Bosch Kloof, Clanwiliam (K. H. B., September 1936); George and Kaimans R. (H. G. W., January 1938); Kaimans R. and Tsitsikama R. (K. H. B., November 1938); Goukama R., Knysna (K. H. B., November 1938). Descriptions of -New Species of South African Alder-flies, etc. 657 Thylakion urceolus Brurd. 1934. Barnard, loc. cit., p. 386, fig."49, a-. Localities.—Palmiet R., near Kleinmond (Caledon Div.) (K. H. B., December 1934); Willow R., north slopes of Cockscomb Peak (Uitenhage Div.) (K. H. B., October 1938). Fam. RHYACOPHILIDAE. Myspoleo agilis Brurd. 1934. Barnard, loc. cit., p. 388, fig. 50, a—m. Locality.—Upper Olifants River, north of Ceres (K. H. B. and C. W. T., October 1937). Myspoleo murinus Brurd. 1934. Barnard, loc. cit., p. 390, fig. 50, n—p. Localities.—Drakenstein (K. H. B., October 1933) and French Hoek Mts. (H. G. W., November 1933); Malgas River, George (K. H. B. and C. W. T., November 1938). Fam. HYDROPTILIDAE. Argyrobothrus velocipes Brnrd. 1934. Barnard, loc. cit., p. 393, fig. 52. Locality—F ound in the Kirstenbosch Botanic Gardens, Newlands, Cape Peninsula (H. G. W., August 1933). PERLARIA. To the records in Barnard (Ann. 8. Afr. Mus., vol. xxx, 1934) may be added the following :— Fam. NEMOURIDAE. Aphanicerca capensis Tillyard. 1934. Barnard, loc. cit., p. 524, figs. 7, 8, a—h. Localities.—Gydo Pass, north of Ceres (K. H. B. and C. W. T., October 1937, 3,2 (var. 8)); Kaimans River, near George (K. H. B., November 1938, 33). VOL. XXXII, PART 6. 45 658 Annals of the South African Museum. Aphamicerca bovina Brunrd. 1934. Barnard, loc. cit., p. 531, fig. 12. ‘ French Hoek Pass (east side) (H. G. W., April 1935, 2 $3). Aphancercella barnardi Tillyard. 1934. Barnard, loc. cit., p. 537, fig. 15. French Hoek Pass (east side) (H. G. W., April 1935, 3 gd, 1 9 (var. y)). Aphaniecercella scutata Brnrd. 1934. Barnard, loc. cit., p. 540, fig. 16. French Hoek Pass (east side) (H. G. W., April 1935, 1 hemipterous 3, T*micropterous gd, 7 micropterous 99, including one pair in cop); same locality (H. G. W., September 1935, 1 g, 292, and 1 hemip- terous g, | hemipterous 9). In the hemipterous ¢ the fore-wing measures 3 mm., extending to the end of the 7th abdominal segment; the hind-wing measures ‘75 mm. In the micropterous specimens the fore-wing measures 1 mm. in the $3, and 1-5-2 mm. in the 99; in both sexes the hind-wing measures -4—-5 mm. Aphanicercella brfurcata Brurd. 1934. Barnard, loc. cit., p. 542, fig. 18. Hermitage Kloof, Swellendam (H. G. W., January 1938). Aphamecercella nigra Brurd. 1934. Barnard, loc. cot., p. 544, fig. 20. Additional specimens, including gg and 99 in copula, from the type locality (K. H. B., 3rd October 1936). Desmonemoura pulchellum Tillyard. 1934. Barnard, loc. cit., p. 546, fig. 21. ) Localities—Rust en Vrede, near Cango, Oudtshoorn District (H. G. W., January 1938); George (K. H. B., November 1938). Descriptions of New Species of South African Alder-flies, etc. 659 ODONATA. Chlorocypha caligata (Selys). 1937. Barnard, Ann. 8. Afr. Mus., vol. xxxii, p. 185, fig. 4. At Patentie, in a side stream of the Gamtoos River, a young nymph was found under a stone (K. H. B., 31st October 1938). The gizzard shows 7-10 denticles in a double row (somewhat irregular) on each major fold, and 4-5 in a single row on each minor fold. Pseudagrion massaicum Sjést. var. cogmant Brurd. 1937. Barnard, loc. cit., p. 215, fig. 15. The Cogman’s Kloof River flows into the Breede River, and this Damsel-fly has now been collected (K. H. B., February 1938) higher up the main valley at Robertson. Gen. SyMPETRUM Newman. 1921. Rousseau, Larves et Nymphes Aquat., vol. i, p. 141. 1930. Lucas, Aquat. Stage Brit. Dragonfl. Ray Soc., No. 117, pp: 19-91, pls. 19-22. 1937. Barnard, loc. cit., p. 251. The diagnosis given by me in 1937 and based on that of Rousseau, must be slightly altered to include the now authenticated nymph of S. fonscolomber, viz.: medio-dorsal projections on abdomen small or absent, and lateral spines on segments 8 and 9 strong or feeble. Sympetrum fonscolomber Selys. 1929. Brain, Insect Pests and their Control in 8. Afr., p. 158, fig. 76 (nymph). 1930. Lucas, loc. cit., p. 83. 1937. Barnard, loc. cit., p. 251, fig. 29 (penis). Several nymph skins were at hand in 1937, but owing to their extraordinary resemblance to those of Crocothemis erythrae, they were not recognised as being really nymphs of S. fonscolomber. Imagos have now been bred and the nymph authenticated. The description of the nymph of C. erythrae (Barnard, loc. cit., p. 250) applies except for the following differences: length up to 15-17 mm.; antennae longer than the distance between their bases 45* 660 Annals of the South African Museum. by the length of the last two joints; divided spines on all the tibiae, “but only on the fore and hind tarsi (spines on the mid tarsi being simple). Gen. Pantata Hagen. 1921. Ris, Ann. 8. Afr. Mus., vol. xvii, pp. 388, 431. Imago.—Arculus proximal to 2nd Ang. More than 10 Anq in fore- wing, the last one incomplete, 7 in hind-wing. Sectors of arculus with common stalk. Triangle not crossed in either wing. One Ac in hind-wing, and a second cross-vein running obliquely to proximal corner of triangle. Anal field in hind-wing very broad. Ry, strongly sinuous. Irregularly one or two rows of cells in IR,—Rspl. Nymph.—Body sparsely setose. Legs slender, moderately long; divided spines on all joints of all the tarsi. Antennae inserted a . little in advance of level of anterior margin of eyes, lateral margins of head behind eyes convergent. Abdomen oval, without medio- dorsal tubercles; segments 8 and 9 with very strong lateral spines. Mask extending to middle coxae; lateral lobes with deep indents. Pantala flavescens (Fabr.). (Fig. 19.) 1904. Needham, Proc. U.S. Nat. Mus., vol. xxvu, p. 712, pl. 40, fig. 5 (nymph). 1921. Ris, loc. cit., p. 431. 1937. Barnard, Ann. 8. Afr. Mus., vol. xxxu, pp. 182, 260. Nymph.—Up to 25 mm. Antennae about 14 times the distance between their bases. Abdomen widest across segment 6. Anal pyramid as long as segments 8 and 9 together; appendix dorsalis longer than the cercoids, and subequal to the cerci. Median lobe of mask with straight margins, meeting in an angle very slightly greater than 90°; lateral lobes with 9 deep indents, each intervening tooth tipped with about 4 graduated spines; lateral setae 13-14, mental Pi setae 16-20. Gizzard with 2 denticles on each ventral tooth, and a series of 6 on each dorsal tooth. Rectal gills with 48-50 lamellae in each hemibranch. | Localities.—Olifants River, Clanwilliam (April 1937, K. H. B.); Letjesbosch, Beaufort West (L. Boonstra and C. W. Thorne, March 1937. Nymph). Remarks.—The gizzard and gills are described from the Letjesbosch specimen; the imago was bred from the Clanwilliam nymph. Descriptions of New Species of South African Alder-flies, etc. 661 Fic. 19.—Pantala flavescens (Fabr.). a, apex of abdomen of nymph. 06, median and lateral lobes of mask. c, one ventral and one dorsal tooth of gizzard. & ( 663 ) 22. Descriptions of New African Anthidiinae (Apoidea). By G. A. Mavromoustaxis, Limassol, Cyprus. (With 1 Text-fig.) Anthidium mveocinctum Gerst. I wave both sexes from Uganda: Hoima (C. C. Gowdey); the mesonotum is smooth and shining, and sparsely punctured, the scutellum is produced and broadly emarginate in middle. Anthidium karossense n. sp. Female.—Length, 11 mm. Black; clypeus a little above the apical margin transversely and somewhat elevated, apical margin very slightly emarginate in middle and angular at sides; clypeus densely punctured (the punctures finer above), ochreous, with two short longitudinal black and parallel linear stripes above, apical margin narrowly brown; lateral face marks (fillimg space between clypeus and eyes) nearly reaching level of insertion of antennae, large mark between ocelli, a rounded spot above each eye, ochreous; mandibles 6-dentate, ochreous, apex black-brown; vertex and occiput densely punctured and moderately shining; antennae black, scape with yellow-brown mark in front; third antennal joint longer than fourth or fifth; clypeus, sides of face, supraclypeal area, with white hairs; vertex and occiput with some- what dense yellow-brown hairs; cheeks with dense and short snow- white hairs. Thorax black; mesonotum and scutellum dull, finely and very densely punctured; scutellum rounded, united with axillae, apical margin very slightly emarginate in middle; a short pale yellow linear stripe at sides of mesonotum; apical margin of scutellum with narrow pale yellow stripe interrupted in middle; outer side of axillae narrowly pale yellow; tegulae finely punctured, brown in middle, and rest ochreous, with subhyaline margin (broadly in front); mesonotum and scutellum with some short and yellowish-brown hairs, apical margin of scutellum with long and dense erect hairs; sides of thorax with dense snow-white hairs, wings subhyaline, marginal cell with transverse clouded streak above; second recurrent nervure interstitial 664 Annals of the South African Museum. with second transverse cubital nervure; femora black, anterior and middle ones with a narrow and short yellow stripe beneath; tibiae black, anterior and middle ones with a narrow yellow stripe nearly reaching the apex above; tarsi black, last joint with brown apex; pulvili absent; hind spurs black; anterior and middle femora with dense and short shining white hairs beneath; tibiae and basitarsi with dense and short shining white hairs above; hind basitarsi with dense and short black-brown hairs on inner side. Abdomen shining, basal half of tergites 1 to 3 somewhat finely punctured, rest very finely and densely punctured (the punctures covering the apical margins); first tergite with basal half yellow-brown (suffused with orange), rest black (the black narrow at sides); second tergite with basal half yellow-brown (suffused with orange), and pale yellow beneath, rest black (the black narrow at sides); third tergite with base yellow-brown (suffused with orange), and with a pale yellow stripe beneath, rest black; fourth tergite black, with a transverse median dull yellow stripe; fifth tergite with basal half dull yellow, rest yellow; sixth tergite black, with two very large dull yellow rounded marks; base and sides of first tergite with white hairs; fourth and fifth tergites with some short and sparse white bristles; sixth tergite with short shining white bristles (the apical ones denser) ; first and second sternites yellowish brown, third and fourth with apical margin yellowish brown, rest of sternites black; ventral scopa shining white, the hairs on sixth sternite short and black-brown. Soutu West Arrica: Kaross, 1 9 (type), 1 9 (paratype), February 1925, in South African Museum. Anthidium karossense n. sp., differs from Anthidium niveocinctum Gerst., or Anthidium banzonis Strand, in the form of the scutellum, the puncturation of the mesonotum and abdomen, the colour of the pilosity and the integument. Anthidium tuberculiferum Brauns. Female.—Length, 10 mm. Clypeus black, lower margin with six rounded tubercles of the same size; vertex and occiput with dense yellowish-brown hairs; thorax with dense yellowish-brown hairs above, and shining white hairs at sides; wings slightly clouded, upper part of marginal cell clouded; second recurrent nervure slightly out of second transverse cubital nervure; hind tibiae black; hind tarsi with basitarsi black, rest of joints reddish brown; pulvilli absent; scopa black-brown, sides narrowly whitish. Descriptions of New African Anthidiinae (Apoidea). 665 Male.—Length, 9-5 mm. Similar to the female; mandibles yellow, apex narrowly deep reddish brown, teeth black brown; clypeus creamy, lateral face marks not reaching level of insertion of antennae, antennae black- brown above and brown beneath; sixth tergite broadly yellow at sides, and with a subapical narrow and short transverse brownish streak; seventh tergite with a broad and deep rounded incision; sixth ventral segment smooth and shining, rounded at the apex; stipes rounded at the apex. Care Province: Willowmore, 1 9, 5. xi. 1919, 1 3, 25. xu. 1903 (H. Brauns); both specimens compared with the type, by my friend Mr. G. van Son (Transvaal Museum). Both agree with the description, but the clypeus of the female has six, instead of five tubercles, as mentioned by Brauns. Anthidium tuberculiferum subsp. namaqualandicum nov. Male.—Length, 8-5 mm. Supraclypeal area with a linear transverse basal cream-coloured stripe; sixth and seventh tergites light yellowish brown; seventh tergite as in A. tuberculiferum; tibiae light yellowish brown, with a basal hight yellow stripe above, that on anterior ones nearly reaching the apex; tarsi light yellowish brown, hind basitarsi with a light yellow stripe on outer side. NAMAQUALAND: Bowesdorp, 1 ¢ (type), November 1934; Klip Vlei (Garies Distr.), 1 3 (paratype), November 1931; both in South African Museum. Similar to the typical male, but smaller. The type from Willow- more has the hind tibiae and the supraclypeal area black, and sixth tergite mostly black. Anthidium katbergense n. sp. (Fig. 1.) Male.—Length, 8 mm. Related to Anthidium tuberculiferum Br.; black; mandibles black, with small lemon-yellow mark, apex narrowly deep reddish brown, teeth black; clypeus moderately shining, strongly punctured, with a very small median impunctate and shining area above; clypeus black, with a very broad reversed T-shaped mark ochreous, the middle branch oval and not reaching supraclypeal area; the ochreous lateral face marks not reaching apex of clypeus; antennae black, and black-brown beneath; vertex and occiput strongly punctured, dull; entire face below antennae with dense and somewhat long shining 666 Annals of the South African Museum. white hairs; front with somewhat dense and long dull greyish-white hairs; cheeks with shining white hairs mostly on outer side. Thorax black; mesonotum strongly and densely punctured, dull; scutellum entire; tegulae minutely and very sparsely punctured, black-brown, narrowly bordered by reddish brown and pale in front; mesonotum and scutellum with somewhat short greyish-white hairs; thorax with dense shining white hairs at sides; eae Sere wings somewhat clouded, apical margin Fic. 1—Seventh abdominal very clouded; nervures black; second hae of Anthidium yecurrent nervure slightly out of second atbergense n. sp., 3. : transverse cubital nervure. Femora and tibiae black (apex of anterior and middle femora narrowly yellow- ish brown beneath); anterior tibiae yellowish brown on inner side, and narrowly on inner side above; middle tibiae yellowish brown beneath and narrowly on inner side above; anterior tarsi yellowish brown; middle tarsi yellowish brown, basitarsi brown; hind tarsi with basitarsi, small second and third joints black-brown, rest of joints yellowish brown; legs with shining white hairs; hind basitarsi with short and dense yellowish-brown hairs on inner side; spurs yellow; pulvilli absent. Abdomen black and shining, apical margins broadly and minutely coriaceous; first and second tergites with the base broadly and strongly punctured (the punctures somewhat dense at sides), and somewhat finely punctured in middle, subapical area some- what finely punctured; tergites 1 to 4 with a small lateral light yellow spot (that on fourth very small); tergites 3 to 5 on each side of middle with a light yellow mark; sixth tergite black, apical margin yellowish brown (broadly at sides), rounded at sides and slightly emarginate in middle, as in Anthidium tuberculiferum; seventh tergite yellowish brown, apical margin with two small lobes and between the lobes a shallow elongate incision (fig. 1); sides of abdomen with dense shining white hairs, tergites with sparse white bristles, those on the apical ones more sparse; sixth sternite yellowish brown, flat and polished; apical margin rounded; stipes pointed at the apex. Female.—Length, 8 mm. Similar to the male; head entirely black; lower margin of clypeus with five tubercles (three large ones in the middle and two smaller ones laterally, with an incision between the median and the sublateral tubercles, and without an incision between the sublateral and the lateral ones); mandibles 5-dentate (basal one stout), black, with subapical linear space reddish brown; first tergite with light yellow Descriptions of New African Anthidiinae (Apoidea). 667 lateral mark; second tergite with two lateral and two sublateral yellow spots; sixth tergite black, with somewhat shining white hairs, apical margin emarginate in middle; ventral scopa shining white, but narrowly from middle to apex the apices of hairs are tinged with light yellowish white. Care Province: Katberg, 4000 feet, 1 $ (type), 1, 15.1.1933, 1 9 (allotype), 15, 30.1.1933 (R. E. Turner), in British Museum. Differs from the female Anthidiwm tuberculiferum Br. in the colour of scopa, mandibles, pilosity, the form of the lower margin of clypeus, and from the male in the form of seventh tergite, the colour of pilosity and other details. Anthidiwm crassidens Cameron, from Katberg, is larger, and it has the mandibles 4-dentate, the scutellum slightly incised in middle, the tegulae black, the hair of the front, vertex and thorax above, fuscous, and the white ventral scopa tinged with rufous. Anthidium severint Vachal, subsp. erzksoni nov. Scape black; occiput with a yellow mark on each side; thorax black; first abdominal tergite black, yellow at sides, and below the yellow light reddish brown; second tergite with basal lateral yellow stripe, rest light reddish brown; third tergite with basal yellow stripe nearly reaching middle (the yellow suffused with hght reddish brown), rest light reddish brown; fourth and fifth tergites with broad yellowish stripe, suffused with light reddish brown, rest light reddish brown. S.W. Arrica: Kunene River, Erikson’s Drift, 1 2 (type), March 1923, in South African Museum. This subspecies is related to Anthidium severint melanaspis (CkIl.), but the colour on abdominal tergites is sufficiently different. Lan Scie | it 4 i ey eee wey ' RS ae Vol. 1.—Part 1, ‘out “of Pope Bare Salk ot Bint” his? Vol. I1.—Part 1, out of print; Part 2, 0lGs Part EY ol woe : Part 5, 1/6; Part 6, 3/-;— Si5 anes . Part 10, 7/-; Part LG 3/-3 Bla. J Cady Vol: 18f.=-Part 1, 2/6 5° Part 9, 1/6; Parke 3, 5/63, pte eee nas 5/6; Part 6, 7/-; Part G out “Part 8, 3/-; Pa Part 9 _ 'Vitle, ete., 1/6. * Vol. IV (containing ’Palaeontological ae published in con Geological Survey).— 4 Part 1, 11/-; Part 2, 7/-; Part 3, 4/6; ‘Part 4, 4/6; Part 6, 4/6; Part 7, 14/-; Part 8 (with Index, Ti Title, e : ~ Vol. V.—Part 1, 4/6; Part 2, 8/6; Part 3, 2/6; Part 4, 1/6; * Part 6, 5/-; heeds ie 3/-5 Part 8, 4/6; Part 9, 4/6; ee etc., 1/6. U\ ae Vol. VI.—Part 1, 13/6; Part 2, 4/6; Part 3, 3/6; Part 4, 30/3 Ind eX, etc., 1/6. ‘ Vol. VIL (containing Palaeontological papers published in conjunction Geological Survey).— Part 1, 3/-; Part 2, out of print; Part 3, 5-3 Part 4, 8/-; 5/6; Part 6, 1/6; Index, Title, etc., 1/6. Sie AS Vol. VIII.—Complete, out of print. Index, Title, etc., 1, /6. Vol. [X.—Part 1, 4/6; Part 2, 5/6; Part 3, 10/-; Part 4, 6/6; Part 6, ll/-—; Part 7, 9/-; Index, Title, ete.s TGs 85 Vol. X.—Part 1, 3/-; Part 2, 2/6; Part 3, 2/-; Part 4, 3/-; Part 6, 3/-; Part 7, 10/-; Part 8, 2/6; Part 9, 5/-; 2/6 Part 11, 20/-; Part 12, 7/-. “Complete. Title, ete. » ae ed i \ Part 11. » oa e, Vol. XI.—Part 1, 3/6; Part 2, 2/-; Part 3, 13/6; Part 4, 1/6; Part 5, J 7 wa . Part 6, Li/=; Part. 7, muarearas Index, oe Jets a Appee ; Plate ILL, 3/-. Vol. XII (containing Palaeontological papers published in eon julie Geological Survey).— Part 1, 15/6; Part 2, 3/6; Part 3, 4/6; Part 4, 3s Part . Part 6, 6/—; Part 7, 20/-; Part 8, 20/-; Index, Title, ete., 1 Vol. XIII.—Part re 6/-; Part 2, 2/6; Part 3, 3/-; Part 4, 8/6; Pe Part 6, 5/-; Part 7, 30/-; Part 8, Ts Index, Title, etc., , Vol. XIV.—Part 1, 8/6; Part 2, 8/-; Part. 3. 6/-;. Part 4, 17/6; P Part 6, 9/—; Index, Title, etc., 1/6. Vol. XV.—Part 1, 17/-; Part 2, 17/-; Part 3, 14/-; Part 4, 12/65 Pa: Part 6, 3/6; Index, Title, etc., 1/6. Vol. XVI.—Part I, 30/6; Part 2, 4/-; Part 3 (with Title, (aoe ete. 3 25/ ie Vol. X VII.—Part ip 12/-; Part 2, 9/6; Part 3, 3/-; Part 4, ea ‘Part eS ws Part 6, 2/6; Index, Title, etc., 1/6. Vol. XVIII.—Part 1,'20/-; Part 2, 7/6; Part 3, 30/-5 Part 4, 12/6; Index, Title, etc., 1/6. Mog TM Aa Vol. XIX.—Part 1, 22/-; Part 2, 17/6; Part 3, 11)-3 Part 4 (with, Index, mh ey etc.), 5/6. wii a aed Vol. XX.—Part 1, 8/6; Part 2, 12/6; Part 3, 4/-; Part 4, 10/-; Part 5, 4-5 a Part 6 (with Title, etc.), 4/6. a ae yeh Vol. XXI.—Part 1, 25/-; Part 2 (with Title, etc.), 30/-. , . Vol. XXIT (containing Palaeontological papers paPhshee in ‘conjunction with the a . Geoiogical Survey).— Pp Part 1, 20/-; Part 2, 10/-; Part 3 (with Title, etc.), 3/6. ; Vol. XXITI.—Part 1, 12/6; Part 2, 8/-; Part 3 (with Index, Title, ete. ), of 7 i is Vol. XXIV.—Part L, 10/-; Part 2, 2/6; Part 3, 4/6; Part 4, U/-., Part 5 om : Index, Title, etc.), 16/-. 2 e.i> Vol. XXV.—Part 1, 12/6; Part 2, 10/-s Part 3 (with Index, Title, € oto. oh 10 aE ee Vol. XX VI. —Complete, 25/-. Migs es Vol. XXVII.—Complete, 25/-. | eee eh Vol. XXVIII (containing Palaeontological papers published in ‘conjunction with ‘the Geological Survey).— ; Part 1, 10/-; Bart 2, 17/6; Part 3, 6/-; Part 4 (with Index, Title, 2 etc. )s 10/-. ; cat eae: Ting ~ Vol. XXITX.—Part 1, 20/-; Part 2 (with Index, Title, etc.), 20/-. Bs Vol. XXX.—Part I; 15/6; Part 2, 20/-; Part 3, 8/-; Part 4, Of-5 Part 5 vith Index, Title, etc.), 9/- ys Vol. XXXI (containing Palaeontological papers published in compas with the : Geological Survey).— Pees | : Part 1, 20/-; Pat 2, P3/6;. Part; 3, 17/6. ES , Vol. XX XII.—Part 1, 2/6; Part 2, 7/6; Part 3, 17/-; Part 4, Ws Part 5, 20/5 : Part 6 (with Index, Title, etc.), 10/-. . nee SENG yes 5 a} re = 4 “ * Vou eX XT. Pari, Sb/=. 0 cee ees Se Vol. XX XIV.—Complete, 40/-. : "(oor gos ida ana INDEX, Subjects and Authors, to Vols. I-XXX, 1/-. vies * ok re as The Annals of the South African Museum will be issued at argue intervals, € as matter for publication is available. . oa pees IS ROIS NASA NS a Copies may be obtained from— | aie gems Messrs. WHELDON & WESLEY, Lrp., Soe See 2, 3, and 4 EarnsHaw STREET, New OXFORD STREET, Lonnox W.C. 23. Or; _ The TTRRARTAN Saum : “by RS Ba A Pica a" AAA | 1] | LOOSE. . ~~, Ad AIA AA YD AAO A Nad halal (0 lal PI IDG 4 AAA As le AAA I AAAnAAAWAANAAN nasaananenmaga get: alae AANA) AA AAAAAAAAAN AAA AIA Annan aaasaunees pAaAnn AAAA / fe 4 a” | nat ) ‘ amt Y \ ; . ’ ? \p is fs a & / ; = am \ 4 a a I a | ‘ ) , | \ ; = A aia in| yyaAyy yy) vyyv lol main YY). ARR | j (fom | I ; nj fn fe ie | oe | a | se & oe ; | aa) Pe, = ; | | | } | , } rs 1, q " , | f r -_ | y ian ila ay Sy PR Fe PF ‘an'.. |e / y / / f f | | oy { fl | tl ost 6 : Yo Vrs | \ f / f y | | | AW AAR Aap AY AAAAAY AAA . A AAARARAARAARRARA | | | | ARR Van NIAVANE AAR navn san aay VAAAANAA > AAA ' i } y wey y 1 : ! | Y J { 1 ) 7 | \ | — AA ~~ | gil \/- t | 1) Mi AUAAIARBEA A ARN A ANA A ae Antaae Ar A AWA | a YAAAAAA ala | a Maan AAA AAAAAAA AAA PAP aAanny alana NY AAAZ ANAAAAAAAAAAA, RAAB AA AAAAAR AA Lu eeacannnansnennen ‘i AANA aa SAAAR A A P| ACAAAEAARAARaAMUAMMAAManance cet of AAAAAA Ala A RASA Wal YY AAR wasnt RERECAARAAPYL Ca AAR S fan a wean AVENE. \ a. \ NAAN |, aalallala lIAAniAR la Al ~ ! ~ f™ ie ; a VW AAAAAY AAAAAAAAAA eterna aA AAAN WAN Nandviaenn " A roorentaccaionncae aoosee aanumnnnnneecees : AWA Na A @»\ o|\|& Bee ae ae ee en a ee Nes A ceeenneeeeeee AA \AAAAAAARAA preenemsceniirnt AAARAAAABAARAAGAA A AAAAAAA NAA =| ons BN EN Les Fo / WV\VAVAAIAARR AAANA ARR ARANAAAAAAA ) WON 3 9088 012