De tho tl rh. oth BM) 4p NS Ee The Ge By Bins hot Ae ll -S'ih om. Angi PEA VE NO! shestette Oh j re SC en DT Pn aU MELA AL AC: Raikes Fy see ihe Pr ' N way Pe pa A OVS Lt Etta Ri WN ir} ak a : a bed) Wek SMH Passat MET gan qyg AF pia psy Haweg-e aa aes GaMOATAS BINGO Wi Ware aL my) Sera ee cee) aa EC a ee aver. Siew oi VRE EN MER AT Mee ae Get 8 He ts A Nap w eae ca bugged Hs es Y eget Ae anf bad Bo A ola Mnaaseyie Oat ea | arcane ee) Cee wy Var A mah eee Ee dey ay Ted aie iaritiinras 7 ee 4 Ao eb die nid rte eth! SATAY Ty Pre tiie RT PETE Oy tt Ve alte dite PWatribedeaedath At Uscuith gees Nake ae tana a8 ene’ Caras won set et ota Vos Negrete yd a ee wey, Baa orth Ni sy Bins ae Nav mn Wont 4} i whee 4 Yer npr ee irae Ue MMe tT ‘ esenites Ap till. oso fl VN Can gy Meath ay WA GALA HOS BU A . bi Dep a Wife fiw aid ‘ SUPP OR eC eGR eet RCM mt ee a RRL ‘ afar ayasitepnerace ho ronan) an HD see \ Ce eee arene ee i Ae ee bette abet ome AP atten Srlraraas wit anaseuanos tad ,ath wy sy hid Lehane thee TCL WO er tis Veer Merrie i So SOR Mr Ore er eat prt ec nite Bene Bart rac ee ec re Sewer jew ‘ Wel Rah B:Baty SOMA Lechawbem easy. det aarat Sete Ate i W ae Mu MATES nhs Owen WCNC TES Mun OER a een eles ve vam ais Darniiten, este re rece RN Sat Vey om arene i UE RH Se Hanae ba rene) SOM OR NOs Ser a Ley Bs laiy Vabatins wie 8 Eee . CROC ioe hia VG AG Lo ina op Nera ea eu ire rates nree ya 1/70 Bm abe a TIN | eiraree eo ane ee 1 thea. yes hha aR ahi eV ane Cie ee Lg eisin hy, tw De bswneeny Sey hgh it gels WIAA HOLA Vem Mia tmauo pe Ar tuihed We yee tetas vena re} We Oa essa a tbh Be Pets airy Wat dateane vine hp i ea onh 4) adhe oP Vb val yond d4g 99 het we Hg Anand Seren arenas Pea V5. by oy ua MUNA ome eh tte sed PoE TTR GBs ay HAST EL ate anche we eine vies EG es NF NE Neat Dake lasif ae mestmeie me ge gran Peer ener Saree sp bs (abn st a eer ae Her eaaiin yen cadertaras yawn Bae hank ately na libet fee Cate sid sinaed Naga Nel adi eS MENT RR ea nae oct tien bit pny ice oa ye ithe Weide be $4an COVES Teer a et a ih AM OANA i uty Ae DON UA Pvay PA RENE Mt mike ae opt eve amtysuge Met gh ea thet i int MVE ke as my PA sate ila fic thoes ie delta * WG seh bag iy she Sm) NaH AAAS BABS WN AG oe og date athe orn! tae hy : eerie Serene et tir) RvR Drei cciwmatanertas ts Waaeh uses “ eas Ae Seba Ge CNH Ra EDF hy Bs singe ht Py yt thy anes sty phe Ret oad Sea See oMun Metter nC tetas Aa sea UL Gooey ape Me etd a Sel SF Malas sidaati an eukesRiis Hee yy Feokoibonsibs gavtetbe gat Die Ebes Aaa e Cad bons ot ikea Horagie oct } Ye Gr Pew yt RG eatyy hi ieete aA NAHE Yes igi’ ¥ Jae gar ye Graydon kemahe de SHI a TA abe CHASM TMMU eld aeech yahll yalyhhoarhrainstsardaasiaceh A cpbeathen counties STs) Basa es ee Mak asi eyed seal ans PIT VE Oe ERs Meee hen yoy: Bs hee Ie Beer Pgh aheye Mr. Joun Barrow, Accountant. Mr. W. H. Cots, Chief Clerk. Mr. Georce ArtHur Dousuepay, Clerk of Publications. Mr. Arruur THomson, Assistant Superintendent of the Gardens. S70,6% & LIST OF CONTENTS. May 3, 1904. Page Dr. Graham Renshaw, F.Z.8S. Exhibition of a drawing of a youne ’Adrrcamebilepliamtis.sesen.ttiaees seen ces clctina clases 1 The Secretary. Exhibition of a photograph, taken by Mr. Brankgkiaess of ai@ uaa... ino eaeeneaicns as eebeaenece ees 2 Mr. F. E. Beddard, F.R.S. Exhibition of the brain of a Troupial (Quiscalus versicolor) infested by worms........- Mr. R. H. Burne, F.Z.S. Exhibition, on behalf of Prof. Stewart, of specimens of the female reproductive organs of certain Marsupials and photographs of a Leathery Riche (ReMi art en ie a meme mar Renner. Sort: Ammen AAAs a 2 1. On the Osteology and Systematic Position of the rare Malagasy Bat Myzopoda aurita. By OLDFIELD THoMAs, HORAS RE ZS. (Ce late oooa et scodes wanes an cscass sc 2 2. Contributions to the Anatomy of the Lacertilia.—(3) On some Points in the Vascular System of Chameleon and other Lizards. By Franx HE. Bepparp, M.A., F.R.S., IPTOSeChOrRuORE ney SOCIC Yarn eme ence codes ase eee se 6 3. Notes on the Gill-rakers of the Spoonbill Sturgeon, Poly- odon spathula. By A. D. Imus, B.Sc. (Lond.), Assistant Demonstrator in Zoology in the University of Birming- inenanueais(Wellayi JUL.) \ i eek Retain OHSRBA Seen taono Hae sann LaMaMOA Anat e 22. iv Page 4, On the Cranial Osteology of the Fishes of the Families Hlopide and Albulide, with Remarks on the Morphology of the Skull in the Lower Teleostean Fishes generally. By W.G. Ripewoop, D.Sc., F.L.8., Lecturer on Biology at St. Mary’s Hospital Medical School, London ......... 39 May 17, 1904. The Secretary. Report on the Additions to the Society's Menagerie im tApril i904 oo. iii socctot saniomnss is scaieesestaeeee 82 Dr. W. T. Calman, F.Z.S. Exhibition of, and remarks upon, a blind Crustacean (Munidopsis polymorpha) from the Hiclandkonwanzarokes@amarcles meveme-peeeeeieacas-eecaraceeeee 82 Mr. F. EK. Beddard, F.R.S. Exhibition of, and remarks upon, a specimen of the Lizard Chlamydosaurus kingii with HE MMOLALGVORCS cy dale hictaasnieciae wale siaisost lee tesarats no selbtawnalsbanre naletalneteirs 82 Mr. Oldfield Thomas, F.R.S. Exhibition of, on behalf of the President, a sketch by a Chinese artist of Pére David’s [Deere iigonsal |e eWiaE al: “SOBA, odonpnenonscoeéaceAdboaudeeaceceLodsec 83 1. On some Nudibranchs from East Africa and Zanzibar.— Part V. By Sir C. Extot, K.C.M.G., late H.M. Com- missioner for the East African Protectorate, F.Z.S. (Elites TEU GME), ied teehee romance sales iesieteine eseals 83 2. Description of a new Tree-Frog of the Genus Hyla, from British Guiana, carrying eggs on the back. By G. A. BOULENGER, HRI). Viena.ss abate we)! o...s.:. cece. 106 3. Notes upon the Anatomy of certain Snakes of the Family Boide. By Frank K. Bepparp, M.A., F.R.S., Prosector OPENS SOCICLY scowcesewbce'sr ae coplsies aeciaaet creo MEME sae 107 4, On Entomostraca collected in Natal by Mr. James Gibson. By G. Srewarpson Brapy, M.D., LL.D., D.S8c., F.RB.S., OMEZS: (blates Vil VIITS) Sen cateus sie cee eeemeee 121 June 7, 1904. The Secretary. Report on the Additions to the Society’s Menaeerie an) May PO04 oe stows 6.2 «brs eee RCP eeRieeets 129 Vv The Secretary. Note on two specimens of the Orang-utan he had seen in Paris Seow wees reese eee seers sere sssseroesscersesescn Dr. Giinther, F.R.S. Exhibition of, on behalf of the President, and remarks upon, a series of mounted specimens of hybrid) Pheasants ne Neha eect e ns once ice Dr. F. D. Drewitt, F.Z.8._ Exhibition of, and remarks upon, some antlers of the North-African Red Deer ............ Dr. F. D. Drewitt, F.Z.8. Exhibition of a pair of horns of Loder’s Gazelle from Southern Algeria ee cee recesses sesseceee Dr. Graham Renshaw, F.Z.8. Exhibition of photographs of, and remarks upon, a pair of Short-horned Buffaloes in the Antwerp Zoological Gardens eco ese eee soe esses esas eeesecees Mr. F. KE. Beddard, F.R.S.: Exhibition of, and remarks upon, the skull of a Cape Crowned Orane ............2..ececeeees Mr. R. E. Holding. Exhibition of photographs of, and re- marks upon, the antlers of the Wapiti in various stages OEP OLN ae SALE ASE SOLE one Mr. R. E. Holding. Exhibition of a pair of antlers of the imishmRed Deere. 4.548. i toi oer ae aaa cea ends coe teas Mr. R. I. Pocock. Exhibition of, and remarks upon, living hairless specimens of the House-Mouse and Brown Rat. Mr. R. L. Pocock. Exhibition of, and remarks upon, young examples of the Egyptian Fat-tailed Gerbille ............ 1. On some New and Little-known Butterflies, mainly from high elevations in the N.E. Himalayas. By Lt.-Col. J VEAL COLM Wawcnrn, 1 (Plate TXG)a oss asia. sets. .5- 50 2. On Seasonal Phases in Butterflies. By A. G. Burinr, 1 Bel oa BY 3)0 Sad LS pa ey Zi AIRED a asses iA Ue a ge OT 3. The Prey of the Lion. By Capt. Ricnarp Crawsuay, BZ SM capa anette vaccine aeeione sh aniinc wa duacwa rcs aa Que aa casan as 4, Note onan apparently Abnormal Position of the “Brephos” within the Body of a Skink (Chalcides lineatus). By Frank EH. Bepparp, M.A., F.R.S., Prosector to the DOGLOUY Fie imran nan eco kaminsen caacin, sr .itican ce ee Lats Page 129 130 130 130 131 131 133 133 133 vi . Contributions to the Knowledge of the Visceral Anatomy of the Pelagic Serpents Hydrus platyurus and Platyurus colubrinus. By Frank E. Bepparp, M.A., F.R.S., Prosector bo, DHE SOCLELY | 2: -en-)e/02- ele o- =e eee enact str OU 6. On the Presence of a Parasternum in the Lacertilian Genus TViliqgua, and on the Poststernal Ribs in that Genus. By Frank HE. Bepparp, M.A., F.R.S., Prosector UG, LMENSOCIEbY, vere: coe-eigect- aie «elon ariel pe eer esac 7. On the rare Rodent Dinomys branickii Peters. By Dr. Eur A. Gorupr, 0.M.Z.S., Director of the Goeldi Wiavisermin, Jee, (IERHIe: 2G) pas esaspooncosc0asdesasecasasococns 8. The Black Wild Cat of Transcaucasia. By C. SATuNIN, Gui Uintits), (CLINE Si © enone oc anonaeasesce caesuadduesnnodescgcoC eco 9. On a Buffalo-Skull from East Central Africa. By R. IUSADI SIGRID So50n scncpsnopaysdcdoee doa oe pba eaaecne so suDsHESOdouSo Od: 10. The Ichang Tufted Deer. By R. LyDEKKER............... 11. On Two New Labyrinthodont Skulls of the Genera Capitosaurus and Aphaneramma. By A. Smrra Woop- WARD. GO BARS. RVZSe (Plates x0) cs 2xcnIl) ar. cc. November 15, 1904. The Secretary. Report on the Additions to the Society’s Menagerie in June, July, August, September, and MEto ber; Mi OO4 Mee oc aesed Hamat seeoeesen eereeeee. dae ree ees Mr. R. Lydekker. Note on a sketch of a Deer from Hainan. Mr. F. E. Beddard, F.R.S. Note on the voice of a young ISBiMEAIKOO caggrasoducotuereosdnorosdunssosundanedoeldoooebesoRdsesd Mr. Frederick Gillett, F.Z.S. Exhibition of, and remarks upon, some antlers of the Altai Stag, shed in the Sociteuys Menagerie ci c.c.cearsen tess ceete ners ane ee eee Dr. P. L. Sclater, F.R.S8. Note on the specimens of the Okapi in the Congo Free State Museum .................. Mr. W. B. Tegetmeier, F.Z.S. Exhibition of a specimen of an Asiatic King-Crab picked up alive off the Isle of WaT CNG a chaantiaeagin seins lunar eros sabe baje'se tin cs te see ee eee eee Prof. J. C. Ewart, F.R.S. Exhibition of skins and de- scription of a new species of Zebra from Hast Africa ... 147 154 170 Ld 178 179 179 180 181 18] Vil 1. On Mammals from the Island of Fernando Po, collected by Mr. E. Seimund. By Ouprrexp 'THomas, F.R.S., EZ Se): (Plate Nal eae cere sou. ogden scladtonsz de tclhes bo On Hylocherus, the Forest-Pig of Central Africa. By OxupFietp THomas, F.R.S., F.Z.S. (Plates XIV. & XV.) 3. On the Species of Crowned Cranes. By P. CHALMERS Mircuet, M.A., D.Sc., Secretary to the Society ......... 4, On the Mouse-Hares of the Genus Ochotona. By J. Lewis IBONHOTE:) MESA\. RUZES i iyitad.c teeny de seieeiehs opie euboeereas On some Edible and other New Species of Harthworms from the North Island of New Zealand. By W. B. Benuam, D.Sc., M.A., F.Z.S., Professor of Biology in the University of Otago, New Zealand ...................5. oO November 29, 1904. Dr. Walter Kidd, F.Z.S. Exhibition of a drawing of, and remarks upon, the extensor surface of the hand of a C@lainmpaanee Re as ae sala Sah aise dale mie ieiae are cus Gade ctaten 1. Some Observations on the Field Natural History of the Lion, By Captain Ricnarp Crawsnay, F.Z.8.. ......... . On some Nudibranchs from Hast Africa and Zanzibar.— Part VI. By Sir C. Huror, K.C.M.G., late H.M. Com- missioner for the Kast African Protectorate, F.Z.S. (Ga Best NG VIBE SS UV AU. is 2 oe ese ety arura natin ors itnteetae a pace a bo 8. On a small Collection of Freshwater Entomostraca from South Africa. By Roserr Gurney, B.A., F.Z:S. (Plate XVIII.) ...... Se Sra ai MRts ae Rit pup a eect tates Mee DMAP TAvos 1h ol nT 4, On the Morphology and Classification of the Asellota- Group of Crustaceans, with Descriptions of the Genus Stenetrium Hasw. and its Species. By H. J. Hansen, JED, LIDS | (Uelenves OCD oO-14)) Mop sbeocsbaednedee 5. On the Lacerta depressa of Camerano. By G. A. BouLencer, F.R.S., V.P.Z.S. (Plate XXII.) ............ 6. On Old Pictures of Giraffes and Zebras. By R. LyDEKKER. 7. On Two Lorises. By R. Livprxxer. (Plate XXIII.) ... « Page 183 263 264 268 298 Vill December 13, 1904. Page The Secretary. Report on the Additions to the Society’s Menasericin November 1904 ss ..5 ee eee 347 Mr. Oldfield Thomas, F.R.S. Exhibition of specimens and description of a new Gazelle from Palestine ............... 347 1. The Characters and Synonymy of the British Species of Sponges of the Genus Leucosolenia. By EB. A. Mrncuiy, HZ.S.); University College) London)... 2.2... ae 349 2. Descriptions of Thirty-two new Species of Halticine (Phytophagous Coleoptera) from South and Central America. By Martin JAcosy, F.ELS. ..........0.0.0.cc00e 396 3. Notes on Anthropoid Apes. By the Hon. Watrer Rorusceitp, Ph.D., F.Z.8. (Plate XXIV.)............... 413 4. Descriptions of Indian and Burmese Land-Shells referred to the Genera Macrochlamys, Bensonia, Taphrospira (gen. nov.), Microcystina, Huplecta, and Polita. By We BEANFORD, WO. He. Iie BRS. VPs: (ETC XOXO ) Ode, .2): ae ee: so mee oan tgs eee nee 44] 5. On the Cranial Osteology of the Clupeoid Fishes. By W. G. Ripewoop, D.Sc., F.Z.8., Lecturer on Biology at St. Mary’s Hospital Medical School, London............... 448 ALPHABETICAL LIST OF THER CONTRIBUTORS, With References to the several Articles contributed by each. BEDDARD, FRANK E., M.A., F.R.S., F.R.S.E., Prosector to the Society. Exhibition of the brain of a Troupial (Quiscalus versi- CoLommuntested by, WOEMIS. ss... cen. see eaeeeres ees asec ester Contributions to the Anatomy of the Lacertilia.—(3) On some Points in the Vascular System of Chameleon and other) Wizards 5... heat Suh. Ge ee Vie aa eke sea paths facet . Exhibition of, and remarks upon, a specimen of the Lizard Chlamydosaurus kingit with femoral pores ......... Notes upon the Anatomy of certain Snakes of the Wamnihy. Bot lease eeeee hae ae PSone ta atatelscichdaie ato. ots otters Exhibition of, and remarks upon, the skull of a Cape Crowned i Camere sos sneer cele la arr a ahas Note on an apparently Abnormal Position of the “ Brephos” within the Body of a Skink (Chalcides lineatus). Page 131 145 5s BrepparD, Frank E., M.A. &. (Continued.) Contributions to the Knowledge of the Visceral Anatomy of the Pelagic Serpents Hydrus platyurus and Platyurus colubrinus See eseeereseser esses aesecessccesoese sere esses eeseesesseesese On the Presence of a Parasternum in the Lacertilian Genus Z7%liqua, and on the Poststernal Ribs in that Genus. Note on the voice of a young Kangaroo eee cece ern ecces eve Benuam, W. Buaxuanp, D.Sc., M.A., F.Z.S., Professor of Biology in the University of Otago, New Zealand. On some Edible and other New Species of Earthworms from the North Island of New Zealand BuanForD, WitL1AM THomas, C.I.E., LL.D., F.R.S., V.P.Z.S. Descriptions of Indian and Burmese Land-Shells re- ferred to the Genera Macrochlamys, Bensonia, Taphrospira (gen. nov.), Microcystina, Euplecta, and Polita. (Plate KOKA) wee estes asters eo esesoeesreseseseesersesossseeeeest sees esosseene Bonuotr, J. Lewis, M.A., F.Z.S. On the Mouse-Hares of the Genus Ochotona ............ BovuLencer, Grorcr Assert, F.R.S., V.P.Z.S. Description of a new Tree-Frog of the Genus Hyla from British Guiana, carrying eggs on the back. (Plate V.)... On the Lacerta depressa of Camerano. (Plate XXII). Brapvy, G. Stewarpson, M.D., LL.D., D.Sc., F.R.S., C.M.Z.S. On Entomostraca collected in Natal by My. James Ciloson., ui (Places Vall — Vela sen ec caci. as see oe eee Page 147 154 179 44] 205° X1 Burne, RicuarpD Hiaerns, F.Z.8. Exhibition, on behalf of Prof. Stewart, of specimens of the female reproductive organs of certain Marsupials and photographs of a Leathery Turtle .............:..eeseeeeee seen Butter, ARTHUR GARDINER, Ph.D., F.LS., F.Z8. On Seasonal Phases in Butterflies..................es0cese0e Catman, Witn1AM Tuomas, D.Sc., F.Z.S., of the British Museum (Natural History). Exhibition of, and remarks upon, a blind Crustacean (Munidopsis polymorpha) from the Island of Lanzarote, COU TaV TEES San REPS SHON git eR mee eae NOY a RUN rn nT Relient CrawsHay, Capt. Ricuarp, F.Z.8. Ave Preys of the giv. 26 scn nc epnemastoncceracee aceneceane ces Some Observations on the Field Natural History of the DreEwirtt, Freperic GrorcE Dawtrey, M.A., M.D., F.R.C.P., E.ZS. Exhibition of, and remarks upon, some antlers of the INorth=Atacan’ Red) Weer ii. ss eins senseoeee on saseeecenseeee Exhibition of a pair of horns of Loder’s Gazelle from Soubhermn Algeria isosceles eee cates ee sie ects nial Entor, Sir Cuarues, K.C.M.G., late H.M. Commissioner for the East African Protectorate, F.Z.S. On some Nudibranchs from East Africa and Zanzibar. Parte, (Blabesplliin ds IVE NR: bee. Riot a idee k ast On some Nudibranchs from Hast Africa and Zanzibar. arty WUsta (latesmNGV plone EVEL UR sn ee Page bo 264 130 130 xil Ewart, Prof. James Cossar, M.D., F.R.S., F.R.C.P., F.Z.8. Exhibition of skins and description of a new species ol Zebra irom bast vActrica: ssh Welle 2 tye eae eae et melee) Fawcert, Lt.-Col. J. Mancoum. On some New and Little-known Butterflies, mainly from high elevations in the N.H. Himalayas. (Plate IX.) GiILLtEert, FreperRicK WILLIAM ALFRED HeErpert, F.Z.S. Exhibition of, and remarks upon, some antlers of the Altai Stag, shed in the Society’s Menagerie.................. Gorpt, Dr. Emit Aveust, Director of the Museu Goeldi, Para, C.M.Z.S. On the rare Rodent Dinomys branickii Peters. (Plate X.) GintuER, ALBERT CHARLES Louis Gorruitr, M.A., M.D., Ph.D., F.R:S., V.P.Z.S. Exhibition of, on behalf of the President, and remarks upon, a series of mounted specimens of hybrid Pheasants. Gurney, Ropert, B.A., F.Z.S. On a small Collection of Freshwater Entomostraca from Southeainicass | (Plate XG Vil) e lee ee acne. eeenee eae eaenee Hansen, H. J., Ph.D., F.M.LS. On the Morphology and Classification of the A sellota- Group of Crustaceans, with Descriptions of the Genus Page 181 134 179 158 129 298 Stenetrium Hasw. and its Species. (Plates XIX.—XXI.) 302 Xi Page Houpine, Ricuarp E. Exhibition of photographs of, and remarks upon, the antlers of the Wapiti in various stages of growth ......... 131. Exhibition of a pair of antlers of the Irish Red Deer . 133 Imus, A. D., B.Sc. (Lond.), Assistant Demonstrator in Zoology in the University of Birmingham. Notes on the Gill-rakers of the Spoonbill Sturgeon, Halonen, Goadoudia, (Ween IU.) Senkencnosoocesoncoscoacscnas 22 Jacosy, Martin, F.E.S. Descriptions of Thirty-two new Species of Halticine (Phytophagous Coleoptera) from South and Central NTTOVSTENGEY, herd ic Coico ae OER An nO AP MIE dt GASSED BION Me em er ON 396 Kipp, Water, M.D., M.R.C.S., F.Z.8. Exhibition of a drawing of, and remarks upon, the extensor surface of the hand of a Chimpanzee ............ 263 LypEKKER, RicHArp, B.A., F.R.S., F.Z.S. On a Buffalo-Skull from East Central Africa ............ 163 AD nes Tike axes over a binnnineyel IDisteyp) Re cable seosnbabosoudsdecgoceadess ae 166 Note on a sketch of a Deer from Hainan ............... 178 On Old Pictures of Giraffes and Zebras ................5 339 Oni Two Worisess) (Plate XxX ne eee sna te seces seen ne 345 Mincutn, Prof, Epwarp Aurrep, M.A., F.Z.S. The Characters and Synonymy of the British Species of Sponges of the Genus Lewcosolenta ......ccescecscesncncens 349 X1V MircHeELL, Perer Cuatmers, M.A., D.Sc., Secretary to the Society. Exhibition of a photograph, taken by Mr. Frank Haes, Ob a Qua eaig Sih eT ile BW curl aay cate canter ee eee eee ets Report on the Additions to the Society’s Menagerie in PASTEL 9 OA iets asc tustles sepaauanicd. ee A cbyashs fk Ceara eee Report on the Additions to the Society's Menagerie in IM ctiyafi 9 OA ee RRAG SET, RENE. MOB Selon atk htatoles abet «a Mobste ame Note on two specimens of the Orang-utan he had seen sir, J Aaya vi fans af ra eres har a SOR eed ET a SOE NTS 518 Report on the Additions to the Society’s Menagerie in June, July, August, September, and October, 1904 ...... On the Species @it Orronanee! OveNES ssacaccuacooncascccnscor Report on the Additions to the Society’s Menagerie in IW@wenmnloce WOOL coopnssadccoseonacouncos onognosoduoccoatacsbooedded Pocock, Recinatp Inygs, Superintendent of the Gardens. Exhibition of, and remarks upon, living hairless speci- mens of the House-Mouse and Brown Rat .................. Exhibition of, and remarks upon, young examples of the Egyptian Fat-tailed Gerbille ...................000eseseeee Rensuaw, Granam, M.B., M.R.CS., F.Z.8. Exhibition of a drawing of a young African Elephant. Exhibition of photographs of, and remarks upon, a pair of Short-horned Buffaloes in the Antwerp Zoological Garden Se eee ote gcc BU asthe okibaiie Cai etnie cake ee rears Page 82 129 129 133 133 XV Page RipEewoop, Waurer G., D.Sc., F.L.S., F.Z.8., Lecturer on Biology at St. Mary’s Hospital Medical School, London. On the Cranial Osteology of the Fishes of the Families Hlopide and Albulide, with Remarks on the Morphology of the Skull in the Lower Teleostean Fishes generally .... 35 On the Cranial Osteology of the Clupeoid Fishes ...... 448 Roruscuinp, The Hon. Lionen Water, M.P., D.Sc., Ph.D., E.Z8. Notes on Anthropoid Apes. (Plate XXIV.)............ 413 Satunin, Constantin, C.M.Z.S. The Black Wild Cat of Transcaucasia ..........0+.0;e0000. 162 Sciater, Pain Luriuny, M.A., D.Sec., Ph.D., F.BS., WSs, EZ: Note on the specimens of the Okapi in the Congo Free Steen Meuse uae sy. aiischisd nc ced tac cule cam etencson clare anne eee 180 TEGEIMEIER, WILLIAM BuRNHARD, F.Z.S. Exhibition of a specimen of an Asiatic King-Crab picked up alive off the Isle of Wight ........................ 181 Tomas, OLDFIELD, F.R.S., F.Z.8. On the Osteology and Systematic Position of the rare Malagasy Bat Myzopoda aurita. (Plate I.) ............... 2 Exhibition of, on behalf of the President, a sketch by a Chinese artist of Pére David’s Deer from Hainan ...... 83 XV1 : Page THomas, OLDFIELD, F.R.S., F.Z.8. (Continued.) On Mammals from the Island of Fernando Po, collected by, Vir Ee Semaund).)((Plate: XeMi) eens ech eee eeeet ese 183 On Hylocherus, the Forest-Pig of Central Africa. (Plates DV XV) ne eee ouics senna celts nane eee Ree 193 Exhibition of specimens and description of a new Gazelle from Palestimes se 2.-ceeae eee eee ee eee eee 347 Woopwarp, ArtHur Smita, LL.D., F.R.S., F.Z.8. On Two New Labyrinthodont Skulls of the Genera Capitosaurus and Aphaneramma, (Plates XI. & XII.). 170 Hist OF Mina Es: 1904.—Vor. II. Plate ll, Wipes, COUGH eGo uboenoDG0o orcs cc onposocuoaedot JUL, GRIMS OE IRON OU, 5 Ra aoc coco oop eacaseangedone Il. 1. Notodoris minor. 2. Trevelyana coccinea. 8. T’. cey- | Uavaved, 4 I, CROCE weacboobrnovogr sono bcodon ooo \ IV. 1. Trevelyana bicolor. 2. Nembrotha cristata. 3. N. > affinis. 4. Martonia levis. 5. Teeth of (a) Bornedla | digitata, (b) B. excepta, and (c) B. simpler ........ a) \o JEDI COU Acaceugeucvenodoveeevnoobpoudcodncouns VI. WAG, \ 1Benserrnastneeony xo INEHENL Sob Ga oagonaoconcvccueonvan VIII. 1X. Butterflies from the N.E. Himalayas ................ GQ ID UCTONS OREITHO aco eos csc nuongeon soso onucnoDodon XI. Capitosaurus stantonensis .....-----+ ++ +++ eeee see XIl. Aphaneramma rostratum 1... cee eee cece XIII. Scotonycteris bedfordt...... 1 se eee rece ee teen es XIV. | apa : Xv. ( Hylocheeris meimert2hagent vesserssancrarsccnccrne é XVI. 1. Phyllidia nobilis. 2,3. Hervia lineata. 4,5. Facelina } lineata, 6. Stiliger varians. 7,8. Elysia marginata . | XVIL 9. Madrella ferruginosa. 10, 11. Fucelina lineata. t. 12. Stiliger irregularis. 13, 13a. Placobranchus | ocellatus. 14-17. Elysia dubia. 18. LE. marginata . | Proc. Zoou. Soc.—1904, Vou. II. b 268 Plate XVIII, XIX, XX. XXI. XXII. XXIII. XXIV. XXY. XVill Page South-African Freshwater Entomostraca ............ 298 1. Stenetrium armatum Hasw. 2. S. mediterraneum, ) NE SD) BMS CAH ACH Op TION ARON Ilr eo co bike aw ak ade 1. Stenetrium serratum, n. sp. 2. S. occidentale, n. sp, GUMS RONLVILENSE USD Wiles ss cvenaiatcta here ee eee 1. Steneirium antillense, n. sp. 2. S. stamense, n. sp. 3-6. Various Asedlota CCC aC eI Ti Ca I Yat Tt CCCI arya Tar) Lacerta depressa Camerano ote tea ke vs BieL cht oie eS eae 332 RACES Of JuOVISES:,)-soryeraterara anja wroieie cust tea te eee 345 Sumia vellerosus (Gray) (very old male) .............. 418 Macrochlamys and similar Land-Shells from India .... 441 LIST OF TEXT-FIGURES. 1904.—Vot. II. Page 1. Afferent renal system of Chameleon vulgaris ..... 0.00. s eens 8 2. Certain anterior arteries and veins of Chameleon vulgaris...... 12 3. Visceral arteries and veins of Pygopus lepidopus..........++0+ 13 4, Renal venous system of Pygopus lepidopus .........+..+200e% 15 5. Hepatic portal system of Pygopus lepidopus..... Soe HiRes 17 6. Part of hepatic portal system of Pygopus lepidopus .......... 18 7. Some arteries of Phelsuma madaygascariensts .......e++.-000: 19 Sb ORRIN CEI ADSL 6 ong MOMS tole AWOS AO DCEO Dee OdeooOns 38 Sb Joa TE COTES TENG OE OE IMMA mmds sk eomone do bone boooe 40 10. Elops saurus, hyopalatine arch, opercular bones, &c., with MORNE? ONS) 94 51 Sbig'd BB elo ie ERisecctis Wet chanel ccalee wallse irr taibey che or anemet ee 40 LC Sos: sours tic oramciiial SKE letOM list tke patel ate cle oie oe) auaregeiete 42 12 Cheyanne GEA Calla Eee UR BUMES 4 5 ols he S CoD MOOS Coad oboous 43 13. Megalops cyprinoides, right side of skull .............- 00005 46 14. Megalops cyprinoides, hyopalatine arch, opercular bones, and TRACT GUA MCE L eee odhostbcceceosdosesoer sod oat 46 Tas CieniOn io eit ook eMss GOROTCOTOUS nis dodacaconeaedeodden sean 48 16. Albula conorhynchus, right side of skull ..................-: 50 17. Albula conorhynchus, hyopalatine arch, opercular bones, and Maaichible MOtGisTOe macs. oteee eee ses Git rtcwaae ovale eigen alee (ne 50 18. Albula conorhynchus, hesgatneemacel SIKGIGIOMN 565506 886 beccud abn | OH 19, Part of intercostal arterial system of Eryx jaculus .........+.. 109 20. Part of intercostal arterial system of Eryx jaculus............ 110 21. Certam abdominal veins in Eryx conicus .......... 002 eves 114 22, Certain abdommal veins im Bry jaculus 2... eens ne 115 23. Liver and certain adjacent veins in Bryx conicus ......++.00+ 118 DAMAtntlers Gk VWapith irom phioxOoTaplse vale c miei: ool sane ses 132 25, Dissection of Chalcides lineatus..... Sc RRR Wiarton ODOR 146 Page 26. A portion of the internal surface of the lung of Platyurus COULDNINUS Hare stirs nants ee aise se OE RP 150 27. A portion of the internal surface of the lung of Hydrus PNOLYUFUS se acai ee ee Mas obs PO EE EE ie 152 28. A portion of the internal surface of the non-vascular part of the nines CHE Ne) EONS (MUM US “a escdecenesocncc¢ancasas5e 153 Soe Nodommnal xibsot Miligualscuncovdes tyes tseee eerie 156 30. Tihqua seincoides, ventral flap of musculature................ 157 31. Front view of skull of male Buflalo, Bos (Bubalus) caffer TEOCNEROBY eg oa te ahaa ROT ee Ue 6 FCO CU ec 165 32. Left lateral view of adult male skulis of Elaphodus michianus (A) andeHAchangenstsi()) warren cecil ean ee eee 167 03. Front view of adult male skulls of Llaphodus michianus (A) and EE MCCHONGENSIS (1S) 5 age enioutep es ake eukey. Were /hies aie eaten aes 168 34. Capitosaurus stantonensis, hinder view of occiput, restored .... 172 30. Ward’s Zebra, to show long ears and face-stripes ............ 181 36. Ward’s Zebra, to show “ sadheon ” and pao dorsal band .... 182 SiemeleandKor Balear.CamegulOnunune ere meinen. Hic see eee 202 TOMA MolMes QUEM CIGZOUChICE HS ene aa teens ear ie ie einen 202 BY), GLSNGL Ot VAC (ROROMIID A decbegecasgdagngseuagnccun4on 203 Ai miveadvol Baleanica cece a tery can doy is tenn A aon See 203 4). Maoridrilus mauiensis. A spermatheca ................-+++: 223 A2. Maoridrilus mauensis. A penialicheetay.. 5.6... cee doe ee 223 43. Maoridrilus mauiensis. Tip of penial cheta ................ 224 44, Maoridrilus mawiensis. 'Tip of penial cheta ................ 204 45. Octochetus michaelsent. Spermatheca............:00e+-2+0 225 46. Dinodrilus beddardi. View of part of the fore-body.......... 227 47, Dinodrilus beddardi. Spermatheca .............00+0+-+00-- 228 AS; hododralus edulis. Viembrall view tps y= ace «1 eee asisteteysieue= 232 AO hhodod rls cds e) Aymem inion recy isp eieirit ae yearn ieee 233 50. Rhododrilus edulis. The prostate of the left side ............ 233 oleiohododilus edulis. A penialichretay wage isin ieee -waeiekreiek 233 52. Rhododrilus edulis. The tip of a penial cheta .............. 233 53. Rhododrilus edulis. Diagrammatic sketch of a section through Ge PORNO DOES sa. case ny eR acs: « cay cnt acted eee ene 234 Op LULododi ius Cause, speriathecar ne see ante rae 234 55. Rhododrilus bestt. Ventral view of the clitellar segments .... 236 56. Ehododrilus bestt. An enlarged view of one of the tubercula OG ALTOS Cle SM eke oes Oh eRe SEL ES a8 0 bt Oo 236 57. Rhododrilus bestx. Portion of prostate..............++.--++- 236 Dow LodonrviusiOeste.. Au penial chester piasacni sree eer eee 237 59. Rhododrilus bestt. Tip of penial cheta ...........-..0+-0-. 237 - 60. Rhododrilus besti. Tip of penial cheta, side view............ 287 Gltwuhododnemsbestt. A ispermabtheca 20. 4. pcre ase een 237 62. Tokea esculenta. Ventral view of clitellar region ...........- 241 33, Tokea esculenta. Enlarged view of the male pore ............ 241 64. Tokea esculenta. A somewhat diagrammatic drawing of a aX dissectioniolatehwormie nn eee ee en ere 242 SOK Page 65. Tokea esculenta. An asymmetrical arrangement of the POLO SALES |v oleae een mMent te cael der ole ilchelecoce MMe Rede walt muaeh Ste eter = 243 66. Tokea esculenta. A diagram, compiled from serial sections, showing the course of the sperm-ducts .................. 244 Cie Token esculenta. . Spermat neces cisin-l ejel-yelvows acpedeske 244 68. Tokea sapida, similar view to that in text-fig.62 .......... 245 GOR Uokea sapida.., Spenmuat bie cammeryectn wetetyelleuerlets ort ater 245 70. Tokea urewere, similar view to that in text-fig. 62 .......... 247 71. Tokea urewere. View of the anterior end of the prostate .... 248 ies Lokea wreweres jp Spermarhecaly iieiss.q-ter de ae eR eae rere ot 248 73. Tokea huttont, similar view to that in text-fig.62 .......... 248 74. Tokea hutiont. View of the prostates ................505: 249 103, HokeatputionimpsSpenm atheca tyiestien sao» oc ioniielstals sey alan 249 76. Tvkea suteri, similar view to that in text-fig. 62 ............ 250 (0. Male) oie | Syren, A We sae beioo gon bobcat coop odT 250 (8. Tokea kirkt, similar view to that in text-fig. 62 ............ 251 AOS Loko kirhiy 1 SperMmirxknecay «Asked ; 1904. ] ANATOMY OF THE LACERTILIA. 21 I think it desirable to call attention as tending to emphasise the peculiarities of the Geckonide as compared with other families of Lizards. As to the apparent differences between Tarentola and Phelsuma, I am unable at present to lay much stress upon them. The arterial system shows one important agreement with that of Phelswma. There are, in fact, three small gastric arteries supplying the stomach exclusively which in’ both genera have precisely the same arrangement; that is to say, there is one anterior artery, followed after a considerable interval by the two next which are close together. But Tarentola has not a gastrosplenic artery arising after these and before the origin of the cceliac, such as that which is present in Phelsuma. I did not observe in Phelswma—but I do not assert that it does not exist—a lateral artery on either side present in Zarentola, This artery is in effect a series of anastomoses between the ex- tremities of the intercostal arteries which arise from the aorta and run along the ribs towards the ventral middle line. Whether it is to be compared to the epigastric artery of either side or not, I do not know. The hepatic portal system and the single median epigastric vein are almost exactly like the corresponding veins of Phelswma. T found, however, only a single parieto-hepatic vessel. The epigastric is connected anteriorly with the vena cava, as in Phelsuma. Suprarenal Portals.—Kach suprarenal body has, as in Phelswma, two afferent veins, The anterior vein of the left suprarenal body passes backwards to it, arising from the side of the vertebral column. This is exactly what I observed in Phelsuma. The azygos vein is as in the last genus. The lateral abdominals appear to be rather shorter than in Pheisuma, plunging at once into the thickness of the parietes, REsuMn. It may be useful to state in a few words the chief new facts contained in this communication. The most noteworthy new features in the vascular system of the Chameleon as compared with those of other Lizards are;— (1) The large number of gastric arteries situated in pairs and supplying right and left sides of the viscus. (2) The connection of the longitudinal oviducal vein with the hepatic portal system, of which it is an affluent apparently not represented in some other Lizards. (3) The entire restriction of the dorsoventral oviducal veins to the efferent renal and to its forward prolongation the anterior vertebral. In /gwana, for instance, these veins open partly into the afferent renal. (4) The absence of aggaepicnens for wanely, running lateral epigastrics. 22 MR. A. D. IMMS ON THE GILL-RAKERS [ May 3, _ (5) The short course of the lateral abdominal veins and their fusion anteriorly with the posterior vertebral. (6) The tendency towards a disappearance of a special supra- renal portal system. As to Pygopus, the following points seem to me to be deserving of special notice :— (1) The large number of gastric arteries. (2) The origin of a considerable number of the visceral arteries from the intercostals, and not directly from the aorta. (3) The large number of dorsal parieto-hepatic portals, and the existence of an equal number of portals arising dorsally on the left side and reaching the liver via the stomach. (4) The connection between the branches of the well-developed lateral abdominals and the dorsal parieto-hepatics. (5) The absence, or at most small development, of lateral epigastric veins. (6) The abundant connection (by 8 trunks) of the median epigastric with the liver and its connection anteriorly with the vena cava. (7) The presence, as in Chameleon, of a posterior vertebral vein continuous with the afferent renal. (8) The opening of a single oviducal vein into the afferent renal posteriorly, and of a single vein into the ovarian and thence into the caval vein anteriorly. Concerning ZVarentola and Phelswma, there are fewer general observations to offer. But I may direct special attention to the following :— (1) The absence of at least conspicuous lateral epigastrics. (2) The connection of the median epigastric with the vena cava anteriorly. (3) The restriction of the azygos to the right side. (4) The shortness of the lateral abdominals. 3. Notes on the Gill-rakers of the Spoonbill Sturgeon, Polyodon spathula*. By A. D. Imus, B.Sc. (Lond.), Assistant Demonstrator in Zoology in the University of Birmingham. [Received April 19, 1904.] (Plate II. +) The gill-rakers of fishes are organs which present a considerable range of variation in form and structure, but which, as yet, have been very inadequately studied. In their most familiar form they * Communicated by Prof. T. W. Barnes, F. RS., F. is { For explanation of the Plate, see p. 34. ID oe Se Oo MIP I Il. Mintern Bros ump. A.D. Imms del. M.P Parker. hth. GIULIS RVUNSEIRS Ol IP OINAOIDON. 1904. ] OF THE SPOONBILL STURGEON. — 23 occur as a single or double row of protuberances disposed along the concave inner or pharyngeal margins of the branchial arches. They may be modified, however, so as to become tooth-like, seti- form, or even filamentous. In some fishes they are completely absent. Among the Elasmobranchs these structures are found in a number of species, and, as a typical example, they are well seen in Acanthias vulgaris. In this fish they occur as lanceolate projections which are developed principally along the anterior edges of the pharyngeal margins of the arches. In the “ Basking Shark,” Cetorhinus (Selache) maximus, they are found as a series of greatly elongated, coarse, seta-like structures which are disposed along the inner margins of the branchial arches upon both the anterior and posterior edges. A somewhat similar apparatus is described as being present in the South African “Whale Shark,” Khinodon tymeus. Among the Holocephali the gill-rakers are small in size and are not conspicuous ; they are seen in Chimera as rows of small tubercles. By far the greater number of the Teleostomes possess gill-rakers, and it is in this subclass that they exhibit their widest range of variation. A contribution towards a systematic study of them in the freshwater members of the Teleostomi has been made quite recently by Zander*. He has examined a considerable number of species, and finds that in carnivorous types like Hso« and Lucioperca the gill-rakers are in the form of teeth. In numerous other genera they form a sieve- like filtering-apparatus (‘‘ Siebfortsiitze”), which is developed in some cases upon both edges of the branchial arches, as in Perca, Acerina, Lota, and the Cyprinide; or in others only upon the anterior edges, as in Clupea alosa, Coregonus fera, C’. albula, and Osmerus. ‘The relative fineness of the ‘‘Siebfortsitze” is correlated with the nature of the food of the species where they occur, and attains its extreme development in this respect amongst those fishes which subsist upon plankton. Poptat also has studied the gill-rakers in the Teleostomi, and has given brief descriptions of their structure and disposition in numerous species. As the result of his studies, he finds that they are specifically peculiar or diagnostic in all the forms which he examined, and interprets their arrangement and development in relation to the form of the mouth and the nature of the food. All three genera of the Dipnoi possess gill-rakers. They are largest and stoutest in Neoceradotus. In Protopterus and Lepidosiren they are present in the condition of minute, slender, pointed projections, and in the specimens which [ examined they are slightly more delicate in the first mentioned of the last two genera. The object of the present article is to call attention to some interesting features with regard to the structure and function of gill-rakers in the case of the “Spoonbill” or ‘“ Paddle Fish,” Polyodon spathula. Iam indebted to Prof. T. W. Bridge, F.R.S., for suggesting to me that I should examine these organs, and * Zeitsch. fir wiss. Zool. Bd. Ixxv. 1903, pp. 233-258. + Ann. Sci. Nat., Zool. t. xii. 1900, pp. 139-216. 24 MR. A. D. IMMS ON THE GILL-RAKERS [May 3, also for his kindness in placing ample material at my disposal and for valued assistance rendered to me in various ways. Anyone who has made even a cursory examination of the gills of this fish cannot fail to have been struck with the appearance of the regular comb-like organs which the rows of gill-rakers form on each face of the branchial arches. Although they are familiar to most zoologists and are characteristic of Polyodon, no one, so far as [am aware, has devoted to them more than passing notice. From among the early accounts of this fish it will serve the present purpose sufficiently if reference be made to a single source only. It is in a letter by Dr. 8. P. Hildreth to the editor of the ‘ American Journal of Science’ that the following mention is made of the gill- rakers. Heremarks: ‘“‘ The jaws are without teeth; but the fauces are lined with several tissues of the most beautiful network, evidently for the purpose of collecting its food from the water, by straining, or passing it through the ciliary membranes, in the same manner as practised by the spermaceti whale.”* Another observer, I, W, Clemens, says that Polyodon “had five pairs of gills which were double. Each of these duplicatures were thickly set with teeth, of about the diameter and consistence of best Russian bristles, and one and a fourth inches long.”+ He mentions that the particular fish he examined measured 4 ft. 8 in. in length. Very little further information is to be gleaned from the works of any of the later writers. Reference is made to their occurrence in this fish by Owen £; and Giinther§$ remarks that each branchial arch has a double series of very long setiform gill-rakers, and that the two series are separated by a broad membrane. No adequate figures of them appear to have been published by any author. A representation of a branchial arch of Polyodon, which also shows the very characteristic disposition of the gill-rakers, is given by Prof, Wiedersheim in the 2nd edition of his ‘ Vergleichenden Anatomie der Wirbelthiere,’ but has been omitted in the later edition, Before describing the gill-rakers of Polyodon, it will be necessary to refer to certain peculiar features in connection with the branchial arches. Each arch has undergone a remarkable antero- posterior compression, so that all its segments, and more especially the epibranchial and ceratobranchial pieces, assume the form of relatively thin cartilaginous plates, The plates are so obliquely disposed that the proper anterior and posterior surfaces look outward and inward respectively, while the concave inner and the convex outer margins are nearly anterior and posterior respec- tively. As usual in other fishes, the gill-rakers of the first four branchial arches form two rows in relation with each arch; and, * “Notice of the Spoonbill Sturgeon or Paddle Fish of the Ohio (Polyodon feuille of Lacépéde),”’ Silliman’s Amer. Journ. of Science, vol. xii. (1827) p. 203, + Ibid. p. 204. * ‘Comparative Anatomy,’ vol. i. p. 482. § Brit, Mus. Cat. of Fishes, yol. vili. (1870) p. 346. 1904.] OF THE SPOONBILL STURGEON. 25 again following the general rule, they are attached to the anterior and posterior surfaces, close to the concave Inner margin, in the form of an anterior and a posterior series, but from the obliquity of the surfaces of the arch the two series appear as if disposed along the outer and inner faces of an arch. They will be referred to in future, however, as the anterior or the posterior series, as the case may be. If the gill-rakers are surveyed from the first branchial arch backwards to the fourth arch, they are seen to become progres- sively shorter in length, and, furthermore, those of them that are situated along the outer or anterior aspect of an arch are somewhat longer than those carried along the inner or posterior aspect of the same arch (PI. II. fig. 2, 2.9.7. and 0.9.r.). Hence the series of the longest gill-rakers is carried on the outer aspect of the first branchial arch, while the row borne on the inner aspect of the fourth branchial arch is composed of the shortest gill-rakers. It is also worthy of note that the longest gill-rakers in either an anterior or a posterior series of a branchial arch are those situated nearest to the junction of a ceratobranchial with an epibranchial element (fig. 1, 7.), the gill-rakers gradually increasing in length from the dorsal and ventral extremities of an arch until the centre of the concavity is reached, where they attain their maximum. The fifth branchial arch carries gill-rakers along its anterior surface only, and they are slightly longer than those disposed along the posterior aspect of the preceding arch. The fifth gill-arch is itself much reduced, since it retains only its ceratobranchial element, and, in correspondence with this, gill- rakers are not developed on the opposing face of the preceding arch in relation to its epibranchial cartilage, but only with the ceratobranchial *. Owing to the extraordinary compression of the plate-like branchial arches +, combined with the attachment of the gill-rakers to their concave inner margins, the greater portion of each arch practically forms a stout cartilaginous septum, which separates the anterior from the posterior series of gill-rakers in relation with each arch, much in the same way that an inter-branchial septum would separate the double series of gill-filaments on the opposite or outer margin of a branchial arch (figs. 1 & 2). The necessity for this curious modification is by no means obvious. It would seem that, were the septum absent, such delicate and fragile organs as the gill-rakers would be very lable to get dislocated or clogged together, and perhaps damaged, through one series of gill-rakers rubbing against the other. As it is, each row is kept in a beautifully regular order, and not a single gill-raker will be noticed to be disarranged from its proper position, and all of them when not in use are closely applied to the surface of the septum. The function of this septum * A full account of the skeleton of the visceral arches of Polyodon is given by Prof. T. W. Bridge in the Phil. Trans. 1878, vol. 169; vide pp. 702-712 & pl.57. figs.8 &9. + Lacépéde (1798) speaks of the branchial arches as cartilage-plates. 26 MR. A. D, IMMS ON THE GILL-RAKERS [May 3, appears to be to prevent the gill-rakers from becoming damaged in the way suggested, and to enable them to be stowed away in a regular order, and within a small compass, so as to admit of the closing of the operculum in the acts of respiration. Tn an example of Polyodon the length of which measured 166 cm. (5 ft. 4in.) from the tip of the rostrum to the extremity of the tail, the longest gill-rakers measured 45 mm. (1? in.) in their greatest length; their average length is about 28°38 mm. (1g in.). It will be seen upon referring to fig. 3 (PI. IT.), that each gill-raker consists of two parts, viz. a slender shaft, which tapers gradually towards its free extremity, and a basal portion, which is embedded under the mucous membrane covering the branchial arch. The basal portion in the specimen figured measures 8 mm. long and 1°5 mm. in its greatest breadth; the shaft near to where it joins the base measures *75 mm. across. Each gill-raker is flattened from side to side at the base, while the shaft is nearly square in transverse section. In their natural position, the gill-rakers are disposed with their flat surfaces at right angles to the septum and are packed very closely together, the interval between any two scarcely measuring 25mm. In colour they are pale yellow-brown, and many of them are somewhat darker at their extreme points. Their surface is smooth and shining; they are extremely brittle, and when viewed with transmitted light they have a translucent appearance. When dried they are bone-coloured and perfectly opaque. The method of attachment of the gill-rakers to the branchial arch is shown in PI. IT. fig. 5, where the lower part of one is represented. It will be seen that the basal portion is inserted just under the mucous membrane (m.) covering the branchial arch, and that it lies parallel with the cartilaginous septum. It is attached to a branchial arch by numerous elastic fibres (e,f.), which are firmly inserted into the gill-raker and form a remarkably tough, strong ligament. Some additional fibres (e./f.') closely ensheath the base at its lower extremity, and others serve to unite adjacent gill- rakers to one another. Inserted on the outside of each are some muscle-fibres (m.f.) which pass downwards and inwards to be attached to the cartilage of the branchial arch. A sample consisting of 869 grm. of the shaft portions of gill- rakers, after having been washed in distilled water and then thoroughly dried, yielded upon analysis the following chemical composition * :— Organic matter ...... 32°81) °/ Mineral matter ...... 67°189 °/,. The mineral matter consists of calcium phosphate 26°80 °/,, together with carbonates, fluorides, chlorides, and sulphates of calcium, with a little magnesium and iron. On account of the small amount of calcium phosphate they * I am indebted to Mr. C. J. Thompson, of the Chemical Laboratory of the Birmingham University, for undertaking this analysis. : 1904.] OF THE SPOONBILL STURGEON, 27 contain, the gill-rakers differ very greatly in their chemical composition either from bone or dentine. I found that the gill-rakers, after being decalcified and then boiled for a few minutes in a strong solution of potassium hydroxide, retained their general form, but the only sign of structure visible in them was a coarse fibrous groundwork. After a prolonged treatment with potash this disappeared, and all that remained was a small amount of a gelatinous precipitate. If the shaft of a gill-raker be thinned by rubbing down on the surface of a fine hone, and then mounted in Canada balsam, a good deal of its minute structure can be made out. When viewed under a magnifying-power of 40 diameters, it is seen to consist of a transparent, faintly yellowish ground-substance, scattered through which are great numbers of lacune with canaliculi radiating from them. Running through the matrix or ground-substance in a longitudinal direction are yellowish-brown canals which contain blood-capillaries (P1. II. fig. 6). In thin sec- tions the ground-substance appears quite colourless, and when seen under a fairly high magnification indications of lamination are visible in it at the periphery, but they die out towards the centre. The lacune are for the most part slightly larger than those which are seen in a section of a human long-bone, and they bear no definite relation to the blood-channels, so that there are no indications of Haversian systems. The canaliculi are not nearly so numerous as those found in typical bone, but they are more frequently branched, and, moreover, they principally arise from the two opposite ends of a lacuna, and those belonging to one lacuna freely anastomose with those of several of the neighbouring ones (fig. 7). A very marked tendency is exhibited by the lacune to be disposed with their long axes parallel to the surface of the shaft with their canaliculi running in a similar direction. The channels containing the blood-capillaries, which are most numerous towards the base of the shaft (fig. 6), take the form of longitudinal canals anastomosing with one another by means of short lateral connections which are given off at frequent intervals. Traced higher up the shaft, the number of these canals becomes consider- ably less, and the anastomoses with neighbouring canals become fewer. As the extremity is reached they become reduced to two or three trunks, which eventually join with one another just under the extreme point (fig. 8). The exact method of the termination of the channels, however, is not easily to be seen in spirit material, owing to the small amount of blood that is present in the capillary vessels which are contained within them. In the basal portion of the gill-raker vascular channels are entirely absent, but running down the centre is a long, narrow, apparently empty cavity. The matrix is homogeneous, and exhibits in places faint traces of lamination. It is well supplied with lacune, and many of them differ from those found in the matrix of the shaft in having a knotted or less regular outline, and in being more profusely branched. 28 MR. A, D. IMMS ON THE GILL-RAKERS [| May 3, In a gill-raker which has been decalcified in a 10 per cent. solution of acetic acid, and afterwards stained with Kleinenberg’s hematoxylin and cut into longitudinal sections with the microtome, some additional features may be observed (Pl. II. fig. 4). The ground-substance exhibits indications of being stratified and its layers show varied capabilities for absorbing staining-reagents. Traversing it are numerous canals, which are lined internally bya definite membrane and contain one or more blood-vessels and some loose connective tissue. In the region of the shaft the lacunz have protoplasmic contents which are r eadily stained, together with one or more deeply staining bodies which are apparently nuclei. Thése lacunee which are situated nearest to the blood-channels can be distinctly seen to be in communication with them by means of their canaliculi. In the basal part of the gill-raker many of the lacunee are shrunken in their outlines and are very poor in stainable contents. The lower part of the central cavity, which is represented at c.c., appears to be nearly empty, containing only some nucleated tissue in which no definite cells are to be distinguished. The elastic fibres of the ligament, already men- tioned, penetrate deeply into the substance of the base in much the same manner as the perforating fibres of Sharpey, which are composed partly of bundles of elastic fibres, pierce the circum- ferential lamelle in bone. Im fig. 4 (ef), where these fibres are seen in section, they appear as variously shaped dots according to the angle through which they have been cut, and they are very evenly distributed through the peripheral ground-substance. The principal blood-vessel is seen to enter the gill-raker about the point of junction of the shaft with the basal part. The vessel then breaks up into several branches, which penetrate the ground- substance and reach the canals traversing it. The mucous mem- brane of the branchial arch (m. in figs. 4 & 5), accompanied by capillaries, is prolonged upwards as a complete and continuous investment to the outer surface of the gill-raker. Ina decalcified gill-raker the presence of a membrane covering it is easy to make out without cutting sections, as it can be stripped off by using a fine needle under a dissecting microscope. The only fish which possesses gill-rakers at all comparable with those of Polyodon is Cetorhinus maximus. In this species the gill-rakers are of the same general form and, in proportion to the much greater size of the animal, they are correspondingly larger andstouter. The investigations of Hannover * and Turner f have shown, on histological grounds, that there is good reason for believing them to be very greatly elongated teeth. In common with those of Polyodon, they consist at their bases of a matrix permeated with anastomosing canals containing blood-vessels; in the shaft they differ in that they contain but a single canal which runs straight up to the tip. There are no lacune, but the matrix * Kong. Danske Vidensk.-Selskabs Skrifter, 1868, p. 485. A résumé is given in French in the Ann. Sci. Nat., Zool. t. ix. 1868, p. 373: : : + Journ. Anat. & Phys. xiv. 1879, pp. 273- 286, pl. xii. 1904.] | _ OF THE SPOONBILL STURGEON. 29 contains large numbers of coarse dentine-tubes which arise from the walls of the canals and run outwards to the periphery, where they form a layer of hard dentine. Turner regards these gill- rakers as being composed of vaso-dentine, but Tomes in referring to them adds ‘ (? osteo-dentine).” * It is worthy of note that in fishes of the genus Chetodon and their allies the maxillary teeth appear to have been modified along the same lines as those by which the setiform type of gill-raker has been produced. As their name implies, the teeth of these fishes are bristle-like; they resemble the hairs of a fine brush in - being flexible and elastic, and they are composed of a yellowish, shining, semi-transparent tissue 7. I would suggest that possibly the gill-rakers of Polyodon are morphologically the much modified descendants of exoskeletal structures which have migrated along with the ectoderm on to the branchial arches. The fact that the mucous membrane covering the branchial arches is regarded as being endodermal in origin, offers considerable difficulty to any idea that such structures could have developed there independently and in sitw, unless they have arisen in the underlying mesoblast. Klaatsch, however, from a study of the placoid scales in Mustelus and some other Hlasmo- branchs, has arrived at the conclusion that their scleroblasts are ectodermal in origin and are derived from the same layer as that which gives rise to the enamel. This layer, which is at first homo- geneous, becomes divided into a portion which has been usually considered to be of mesodermal origin, while the rest remains in connection with the ectodermt. Hence he considers that a “ dermal” exoskeleton is not mesodermal in its ultimate origin. It is worthy of note that, with regard to the pharyngeal teeth of many fishes, several writers are inclined to believe that their presence is due to a migration of the ectoderm into the cavity of the pharynx. For this reason, and on account of the difficulty of reconciling them with the presence of anything except ectoderm, I would suggest the possibility that the skeletal tissue of the gill- rakers of Polyodon has arisen from portions of the epiblast forming the outer portions of the gill-clefts, which have migrated on to the inner or pharyngeal margins of the branchial arches. At all events, if any migration of epiblast has taken place, the latter route seems at least as feasible as a backward migration from the stomodeeum. In Cetorhinus the gill-rakers retain many structural features in common with the teeth of the animal, but in Polyodon they appear to have undergone a more special modification along lines of their own. The structure of the teeth in the young Polyodon has been described by Zograff§, but, after a comparison of the * “Dental Anatomy,’ p. 220. + Vide Owen, ‘Odontography,’ pp. 8 & 105, pl. i. fig. 2. ~ Morph. Jahrb. xxi. 1894, pp. 153-240. § “Ueber die Zahne der Knorpel-Ganoiden,” Biol. Centralbl., Bd. vil. 1887-88, p. 181. Ann. Sci. Nat., Zool. 8 ser. t. 1. p. 203, pl. 4. fies. 3, 4, & 6, 30 MR. A. D, IMMS ON THE GILL-RAKERS [ May 3, gill-rakers with his account, I have not been able to make out any salient points of resemblance between the two structures. With the exception of the rhombic plates and ‘“fulcra” of the tail, the scales are too degenerate to admit of a similar comparison being extended to them. The plates and “fulcra” of the tail are, however, tolerably well-developed structures. The former I have examined after having thinned them, by rubbing down on the surface of a fine hone, in the same way in which the gill- rakers were treated. The matrix of a scale is colourless in thin slices and is pervaded everywhere by lacune which are similar to those found in a gill-raker, but it does not contain any blood- channels. The substance of the plate appears to be deposited around a longitudinal core-like centre in the matrix. Adjacent plates are united to one another by means of ligamentous connec- tions the fibres of which penetrate deeply into their matrix. These fibres are comparable to what Hertwig calls the ‘ Schuppen- ligament” of the scales of Lepidosteus. In structure there is, therefore, a considerable likeness between a rhombic plate and the basal portion of a gill-raker. The matrix and its lacune are identical in both cases; the hollowed core or cavity in the base of the gill-raker might be compared to the core of one of the plates, and to this may be added the absence of blood-channels in both cases. The fibres which connect a gill-raker to the branchial arch, and also which bind adjacent ones together, are comparable with the ligaments which unite neighbouring scales. For these reasons I think that it is not improbable that the basal portion of each gill-raker is the homologue of a ganoid scale—i. ¢., of one of the rhombic plates which are found along the sides of the upper lobe of the tail. The shaft or principal part of a gill-raker may correspond to a greatly elongated spine, or to one of the evanescent spines which are found in relation with each scale in the develop- ing Lepidosteus, and which are regarded as representing the spinous portions of placoid scales. In Lepidosteus, as Nickerson has pointed out, the basal plate, which is the essential part of a ganoid scale, has come to be developed independently of the vestigial spines, instead of being a continuation of the process by which the latter are produced; and, in comparison with the basal plate of the Selachian scale, it has increased greatly in size and importance and has incorporated within itself fibres from the dermis*. Ina gill-raker, it would seem that we have a basal plate which is similarly specialised, though not to so great an extent, but that there has been no corresponding reduction in the spinous portion, which, on the contrary, has become greatly elongated. It has no trace, however, of a hard dentine layer, nor of a coat of enamel or ganoin, unless the covering of mucous membrane is to be looked upon as the representative of the latter, * Bull. Mus. Comp. Zool., Harvard, vol. xxiv. 1893, pp. 115-140, 1904. ] . OF THE SPOONBILL STURGEON. 3l The absence of such hard parts is to be correlated with their not being subjected to any use which would involve hard wear or much friction. In short, it is possible that the gill-rakers are to be looked upon as exoskeletal parts which were derived from an ancestral Selachian condition, where they exhibited little or no differentiation either in form or structure. Subsequently they became modified along lines of their own in order to fulfil particular functions, some migrating into the mouth to become teeth, while others passed on to the branchial arches and have given rise to the gill-rakers. From what I have described of their structure, the gill-rakers, at least their shaft portions, appear to be composed of a substance which bears a close resemblance to osteo-dentine, if not identical with it. Osteo-dentine was defined by Owen as that type of dentine in which the matrix is arranged around the vascular channels in the form of concentric rings, and in which lacuns similar to those of bone are found*. Tomes regards osteo-dentine as a substance which is developed by calcification proceeding through the interior of a pulp, and not by means of the calcification of a special layer of cells (odontoblasts) as is the case with other types of dentine. Consequently, inatooth or structure composed of osteo-dentine there is no single pulp, but pulp and calcified tissue are quite inextricably mixed up, the vascular channels containing masses of pulp-structure as well as blood-vessels. In vaso-dentine there is a distinct pulp-cavity from which radiate canals which contain minute blood-vessels only. He further calls attention to the fact that in some teeth neither of the characteristics defined by Owen occurs, though, if their manner of development be taken into account, they are unquestionably made of osteo-dentine?. Apart from any knowledge of their mode of development, the substance of the gill-rakers of Polyodon bears a closer likeness to osteo-dentine than to any other structure, for the following reasons. It resembles that type of dentine in the absence of a common pulp-cavity, and in the nature of the anastomosing channels which contain one or more blood-vessels and some loose con- nective tissue (pulp-remains?), The presence of bone-lacune is an additional point of resemblance, though Tomes does not look upon it as being diagnostic of osteo-dentine, since they, or spaces very similar to them, are present occasionally in other kinds of dentine. In Cetorhinus the teeth are relatively greatly reduced in size, and its food consists principally of minute surface organisms. The gill-rakers serve as a straining-apparatus which prevents the food-particles from passing into the branchial sacs with the outflowing current of water. As mentioned by Prof. Turner ¢, * Comp. Anat. vol. i. p 362. + “On the Structure and Development of Vaso-dentine,” Phil. Trans. 1878, p. 40. Also Dental Anat., 2nd edition, pp. 88-92. t Loe. cit. p. 275. 32 - -MR..A, D. IMMS ON THE GILL-RAKERS [May 3, Cornish has stated, in an account of a supposed Basking Shark, that in front of each gill a slight comb-like apparatus extended the whole length of the ray. As the mouth was opened, the comb automatically fell back to a right angle with the gill-ray, and effectually barred the egress through the gills of anything except water taken in through the mouth*. Although this apparatus has received attention from numerous zoologists, no one, so far as I am aware, has offered any suggestion as to the means by which the gill-rakers are brought to interlock with one another when they are in use. The occurrence of minute teeth in Polyodon is a well-known feature. According to Johannes Miiller, there are found im young specimens (a foot long) two rows of small teeth in the upper jaw and one row in the lower jaw. Similar teeth are found on the two anterior branchial arches where they join the floor of the mouth, and upon their opposite extremities where they join the palate. He mentions that examples over 3 feet long are. edentulous. Ina specimen in the Zoological Museum of the University of Birmingham which measures 88°4 cm. (2 ft. 10 in.) long, I find that there are unmistakable teeth arranged on the jaws, as Miiller states; those in the upper Jaw are worn down a little more than those in the lower. In another fish measuring 139°1 cm. (4 ft. 53 in.) in length, I have been unable to detect any trace of teeth. The nature of the food of Polyodon is correlated with the vestigial character of the teeth. The fish is described as stirring up with its spatulate snout the mud at the bottom of the waters of the “bayous and lowland” streams which it frequents, and feeding upon the microscopical organisms contained in it ; but the evidence which supports such a statement appears to be rarely quoted, and it leads one to believe that it is not so definite as one would wish. An early writer, already referred to, namely T. W. Clemens, remarked that the Polyodon which he dissected “had no food in its intestines—all that was observable was a small quantity of substance resembling chyle, but of the consistence of honey.” T. H. Bean §, quoting Prof. 8. A. Forbes, says that “the long snout is useful in procuring its food, which consists chiefly of entomostracans, water-worms, aquatic plants, leeches, beetles, and insect larvee.” In the hope of being able to furnish some additional obser- vations, I made a careful microscopical examination of the contents of the whole course of the alimentary canal in two specimens of Polyodon. In both cases the food appeared to have been much acted upon by the digestive secretion and very little * ‘Zoologist, 1870, p. 2253. + © Anatomie der Myxinoiden,’ p. 150. + Loe. cit. p. 204. § “Cat. of the Fishes of New York,” Bull. 60 of the New York State Museum, 1903, p. 62. 1904.] OF THE SPOONBILL STURGEON. 33 could be made out with regard to its nature. The specimens had been in spirit for a long time, which greatly increased the difficulty of identification. I was able, however, to recognise among it remains of parts of the exoskeleton of insect larvee together with portions of the spiral thickenings of the trachez, a few small Oligochete worms, fragments of plant-remains, and some earthy matter. There is every veason to believe that the gill-vakers of Polyodon ave similay in function to those of Cetorhinus. On account of their fineness, and the closeness with which they are packed together, they would form an even more effective straining- apparatus than they do im the case of the latter fish. Tn the absence of direct observations on the living fish, as to the precise method by which the eill-rakers constitute an efticient filtering-mechanism, recourse must be had to anatomical evidence. If the gillvakers act as a filter, it is clear that the anterior series of vakers of one branchial arch must be inclined forwards so as to meet the posterior series of the preceding arch, which have become inclined backwards for the purpose, so that the two series of gill-vakers interlock or interdigitate across the cleft. Under ordinary circumstances, however, the gill-rakers are closely applied to the anterior or posterior surfaces of the flattened branchial arches to which they belong, and they do not in the least incline across a cleft, or tend to meet those of an adjacent arch on the opposite side of the cleft. In all probability the necessary movements of the gill-vakers are brought about by means of the contractions of the muscle-fibres, represented in PI. I. fig. 5, which extend downwards and inwards from the outside of a gill-raker and are attached to the cartilage of the branchial areh. By the contraction of these muscle-fibres the gill-rakers would be pulled outwards so as to form an angle of about 60° with the septum. When the fibres are relaxed, the gill-rakers, on account of the pull that is exerted upon them by the stretched elastic fibres, would spring back of their own accord to close against the septum, and so take up the position they occupy when not in use. The anterior row carried by the first gill-arch is composed, as already mentioned, of the largest mdividual gill-rakers. The yeason for this appears to be that they have to bridge over the interval between that arch and the hyoid, since the Jatter is devoid of gill-rakers. In concluding these few notes it may be worth while to mention that the gillvakers of the other surviving Chondrostean Ganoids differ very greatly from those of Polyodon, although all these Fishes share the common character of having a greatly reduced dentition. In Psephurus they are comparatively short and are moderate in number. In Scaphirhynchus they are small fan- shaped structures each of which terminates in three or four points. In Acipenser the gill-rakers are small, flattened, and pointed organs which differ somewhat in relative size among the various species. Proc. Zoon. Soc.—1904, Vor. IT. No. IT. 3 D4 ON THE GILL-RAKERS OF THE SPOONBILL STURGEON. [May 3, Doubtless these variations in form ave correlated with differences in the nature of the food in the different genera. The most important features with regard to the gill-rakers of Polyodon may be summarised as follows :— 1. The gill-vakers are setiform structures, and each consists of a basal portion, which is attached by means of elastic fibres to the branchial arch, and a long, pare shaft portion which forms its prmeipal part. 2. The matrix of a gill-vaker contains numerous Jacunze which are connected with one another by means of canaliculi. In the shaft portion it contains, in addition, a series of anastomosing channels in which le blood-capillaries and some loose connective tissue. 3, Structurally, a gill-raker bears a very close resemblance to osteo-dentine. 4. The mucous membrane covering a branchial arch is pr molonged over each gill-raker in the fom of a complete inv esting coat. . The necessary movements of the gill-rakers appear to be brought about by means of the contraction of some muscle- fibres which are inserted on the outside of each gill-vaker and which pass downwards and inwards to be attached to the branchial arch. 6. The food of Polyodon consists of microscopical organisms and the gill-rakers serve as a straining-mechanism which effectually bars the entry of such particles into the gill- cavities. EXPLANATION OF PLATE TI Fig. 1. A portion of the 2nd branchial arch of Polyodon spathula showing the general arrangement of the gill-rakers. (% nat. size.) 2. A semi- diagrammatic section taken across the 2nd branchial arch (through the ceré atobranchial cartilage), to show the position of the gill-rakers in rela- tion to the septum. (# larger than nat. size.) A single gill-raker. (Nat. size.) A section taken through the lower part of a gill-raker which has been decalcified and afterwards stained with Kleinenberg’s hematoxylin. The mucous membrane of the branchial arch is prolonged over the surface of the oill-raker. The matrix of the latter shows indications of lamination and numerous lacunz are seen scattered through it. (xX 30.) 5. Figure showing the attachment of a gill- raker by means of elastic fibres to the cartilage of the branchial arch. “(Xx 5.) 6. A portion of the broadest part of the shaft of a gill-raker which has been thinned by rubbing down upon a fine hone, the preparation afterwards mounted in balsam. The matrix is seen to contain large numbers of lacunze mostly with their axes pomting in one direction. The vascular channels are seen to anastomose with one another at frequent intervals. (xX 40.) . Five typical lacune from the shaft of a gill-raker; they are seen to be in communication with one another by means of their canaliculi. From a preparation made in the same way as the one represented in fig. 6. (X 370.) . View of the extremity of the shaft of a gill-raker mounted whole in glycerine. This figure shows the termination of the vascular channels in the tip of the same. (X 20.) SIR C, ELIOT ON NUDIBRANCHS _ [May 17, generally rather sombre but gorgeous, a prevalent tint being very dark green or blue with brilliant lighter markings. The gills are few (3-5), but large and strong. The hermaphrodite gland is as usual, and not collected into globules. A very narrow labial armature is present In some species, but usually there is none. The radula is never very wide and sometimes is very narrow, consisting of a median plate with from three to twelve laterals, of which the first is large and hamate and the rest plate-like. The species are not all equally well known, all our information as to NV. morosa and cristata coming hitherto from drawings by Semper. It would appear, however, that some of my specimens should be referred to the latter species. WV. nigerrima, kubaryana, and cristata have a fairly broad radula, with about twelve laterals, and are distinguished by their dark coloration. They are evidently closely related, and may prove to be merely varieties, including WV. morosa. JN. gracilis, diaphana, gratiosa, and affinis are lighter in colour, and have a narrow radula with only three or four laterals. Nembrotha is recorded from the Indo-Pacific and West Coast of Mexico. It is fairly common onthe Hast Coast of Africa. NemBrorua cristata B. [?=V. nigerrima, var.] (Plate IV. fig. 2.) [ Bergh, 8. R. xi. p. 458, pl. xxxiii. fig. 6.] Three specimens from the East Coast of Zanzibar. The living animals are described as having a sloping back, long tail, and narrow foot, somewhat like Ceratosoma. ‘The texture was soft, and the colour a very dark but brilliant green with black spots, and also narrow stripes of brighter and lighter green. The gills were counted as five, and the rhinophore-pockets were raised. The measurements of the largest alcoholic specimen are : length 54 millimetres, breadth 15, height of body 13, height of branchiz above body 8. The texture has become hard and wrinkled, the animals having evidently been strongly contracted. The main stem of the gills is very thick, strong, and muscular, so that it almost forms a valve to protect the pinne as in otodoris. The anterior plume is distinct and separate, but the lateral pairs are almost confluent, and it is consequently hard to say where one begins and the other ends, or whether the total. number of branchiz should be reckoned as three, four, or five. The rhino- phores are not very large and are completely retracted within smooth projecting sheaths about 2°5 mm. high. The foot is narrow. ‘The relations of the external mouth-parts are much obscured and distorted by the strong contraction which has affected the whole anterior portion of the body, but it appears probable that the foot is grooved and notched with the upper lamina attached to the corners of the mouth, and that the tentacles are horizontal ridges. There is a very narrow labial armature, about half a millimetre wide and hardly visible to the naked eye. Itappears to form a complete ring, and is composed 1904. ] FROM EASY AFRICA AND ZANZIBAR. 91 of a loose mass of long, yellow, transparent rods, irregular in shape and often bent. Tn the two specimens dissected, the radula consists of 30 and 31 rows respectively, and the formula of each row is, as a rule, 10+1.1.14+10 or more rarely 114+1.1.1+11. The median tooth is squarish, not very broad, and bears, as a rule, five denticles on the anterior edge, but sometimes only four, while in one specimen there were only three denticles in the hinder rows. The first lateral tooth is large and sickle-shaped. The corner of the basal part projects over the rhachidian tooth and creates a false impression that it is an accessory denticle. The remaining teeth are generally ten, but sometimes an additional rudimentary one at the end of the row raises the number to eleven. They are little more than flat squarish plates, decreasing in size outwards. Only the first of them shows some traces of a hamate shape. Allthe internal organs are of a deep black colour, which rendered their examination difficult. The blood-gland is large. I was not able to make any satisfactory preparations of the reproductive organs, but the glans seemed to be armed with a dense mass of curved rods. IT think this form must be identified with JV. cristata B., of which, however, no specimen has been described, all that is known of it being Semper’s drawing and apparently a few notes. But it is also not improbable that it is a variety of WV. nigerrima B., from which it differs externally in little except the absence of any red coloration. The number of branchie is, as explained, uncertain, but the arrangement shown in Bergh’s plate of NV. nigerrima is certainly not that of these specimens. On the other hand, the presence of the narrow labial armature is an argument for identity. NEMBROTHA CHRULEA, sp. 0. Four specimens from Sii Island, near Vanga. No notes on the living animal, except that it was blue and had apparently no red or green mottlings. The colour of the freshly-preserved specimens was a fine bright indigo, varying in intensity in different parts. One of the specimens was much lighter than the others and also smaller. The whole of its body and the lighter parts of the other indi- viduals were marked with deep indigo spots. The largest preserved specimen is 43°) millimetres long, 18 high, and 12°5 broad. The space from the head to the branchie is 12 mm. and from the branchie to the end of the tail 22 mm., but the tail is longer in this specimen than in the others. The shape is somewhat like that of Ceratosoma without lobes, as the back rises considerably from the head to the branchie. The integuments are leathery and not at all transparent. The surface is quite smooth, and there is no indication of a mantle-edge. The rim of the rhinophore-pockets is only slightly raised. The rhino- phores themselves are large, entirely retracted, with 25-30 deep 92 SIR C. ELIOT ON NUDIBRANCHS [May 17, perfoliations. As in the last species, the gills are very thick, strong, and muscular, apparently five, but in this case, too, the lateral pairs sometimes coalesce, so that the whole number may be counted as three or four. They are bipinnate. The oral tentacles appear as large, distinct tubercles on each side of the mouth, and were doubtless fairly long in life. The foot is rather broad, with a Shallow groove in front; the upper lamina is connected with the sides of the mouth under the tentacles. The internal organs are mostly of a greyish yellow, not deep black as in the last species. Though the labial cuticle contained a few scattered yellowish rods, no connected armature is visible. The radula much resembles that of JV. cristata, and has for formula about 27 x 10+1.1.1+10 or occasionally 11+1.1.1+411, but the median plate is broader, with five distinct denticulations which do not vary in number. The first lateral has a groove near the end of the hook, and the next two or three teeth have a rudi- mentary hamate shape. The liver is large. The upper wall of the pericardium is very thick and strong. The verge resembles the figures in Bergh’s plates of Membrotha nigerrima, the glans being armed with a profuse mass of hamate teeth. Those on the top seemed rather larger and more curved than in his figures. This species is closely allied to WV. nigerrima, but appears to be sufficiently distinguished by (a) its coloration, (6) the only slightly projecting edges of the rhinophore-pockets, (c) the absence of a labial armature, (d) slight differences in the radula, (e) another form of tentacles. NEMBROTHA AFFINIS, sp. n. (Plate LV. figs. 3 a—-3 d.) [Cf. W. gratiosa Bergh, Nudibr. of ‘Blake’ Expedition, pp. 172- 175.] . One specimen caught in a trawl in Chuaka Bay on the Hast Coast of Zanzibar. Very long and narrow, being 5 centimetres in length and 1 in height. The living animal was extremely soft, dull violet-black in colour, with dull yellow stripes on the sides and somewhat brighter ones of the same colour on the back. The stems and bases of the gills were light green, and the same colour occurred between the rhinophores and round the edges of their pockets. The pinne of the gills looked black, but when seen by transmitted light were of a fine purple. The foot was very narrow, and the animal could not adhere strongly to anything. The alcoholic specimen is flabby, 28 millimetres long, 5 broad, and 10 high. As the result of this reduction in size, the yellow parts look wider and the black parts narrower, so that the animal appears to be yellow with black stripes, rather than black with yellow stripes as in Mr. Crossland’s figure. No doubt, however, the latter is correct ; it represents four lateral yellow stripes and one medio-dorsal. The stripes are interrupted in places, particu- larly on the tail, and there are some long yellow spots between them. The branchie are distinctly only three in number, smaller 1904. } FROM EAST AFRICA AND ZANZIBAR. 93 than usual in the other species, but with a very thick rhachis and bipinnate. The rhinophores are large and exserted, each bearing about 35 perfoliations. The rims of the pockets are very slightly raised. The oral tentacles are two hard black ridges, curved downwards and sideways. The foot is narrow and grooved in front. The buccal mass was extracted, but the animal was not further dissected in order to preserve the specimen. There is no labial armature. The formula of the radula is 13x3+1.1.1+3. The teeth closely resemble those of WV. gratiosa, the chief difference being that the anterior margin of the wide median tooth (fig. 3 ¢) is indistinctly bilobed, the right half being always a little higher thar. the left. The first lateral (fig. 3d) is large, rather irregular in shape, and with a double hook at the apex. This form is closely allied to JV. gratiosa, and were the latter found in the Indo-Pacific region, I should be inclined to regard them as varieties of one species. But JV. gratiosa is recorded from the West Coast of Mexico*, which lies outside of the Indo- Pacific area; and it is therefore probable that the differences pre- sented by the present animal are real and greater in living indi- viduals.. (a) It is not mentioned that JV. gratiosa is remarkably soft. (b) The present specimen shows no traces of ridges near the rhinophores or on the tail. (c) The coloration of WV. gratiosa is not dissimilar, but the pattern is spotted whereas here it is striped. (d) The tentacles do not look as if they had ever been ear-shaped. (e) The anterior margin of the median tooth is indistinctly bilobed. MARIONIA. [See especially Bergh, in Semper’s Reisen, xv. p. 737, & xvii. p- 890.] All the Tritoniade which I have collected in East Africa belong to this genus, unless the form described as Warionia sp. is regarded as sufficiently certain to constitute a new generic type. Marionia is distinguished from its near allies Z’ritonia and Candiella by the presence of a circle of horny plates or leaves in the stomach. The velum bears distinct processes, which are often ramified. The edge of the jaw has one or more rows of denticles, and the radula is moderately wide. The central tooth is broad and more or less distinctly tricuspid. The laterals are hamate, but the first one is larger and clumsier than the others. Provisionally I think it best to divide the forms here described among six species, but am by no means certain that they will all prove valid. When more material can be examined it will probably be found that the species of MJarionia exhibit many varieties in formand colour and run one into another. It is also not impossible that the denticu- * In Bergh’s ‘System der Nudibr. Gasteropoden,’ p. 1145, the locality is given as “ mare indicum, Amboina,” but this appears to be a slip. ‘The animal is described by Bergh among the molluscs of Amboina, but is expressly said to come from Mexico. 94 SIR C. ELIOT ON NUDIBRANCHS [May 17, lation of the jaw varies with age. Of the six species, IZ. pellucida seems distinct from the others, which are all nearly related to M. arborescens. WM. levis is distinguished by being quite smooth and not at all tuberculate. J. ramosa is closely allied to I, arbo- rescens, and differs chiefly in having unusually large branchie and appendages. WW. viridescens and albo-tuberculata differ from these last two forms in having branched processes on the velum, and are closely allied one to another in structure, though by no means similar in external appearance. It is noticeable that in none of these forms is the interior of the buccal cavity black, and that most of them have only one fully- developed row of denticles on the jaw. MARIONIA PELLUCIDA, Sp. 0. One specimen dredged in 10 fathoms near Wasin, East Africa. The living animal showed very httle colour but for the pink liver which shone through the transparent integuments. The back was sparsely reticulated with vermilion, turning to deep crimson near the bases of the branchiz, and also bore some opaque white raised spots. ‘The sides of the body were white and the edge of the velum sandy-coloured. The velum was not bifid, and bore 12 processes, of which 8 were 3-branched. The branchiz were 13, of moderate size, directed backwards, The finer branches very delicate and transparent. The alcoholic specimen is yellow, with small tubercles of a lighter colour on the back and sides. It is 15 millimetres long, 5 broad at most, and 4 high. The 13 branchie are rather far apart from each other; none are large, and the first pair as well as the last three are minute. The dorsal margin is not very prominent. The rhinophores are large. The velum as described above, but though the outermost processes probably represent the tentacles, they do not seem to be grooved as usual. The long narrow jaws bear three or four rows of denticles on the edge. The radula is at most 22+1.1.14+22 x 25, but many of the rows are much shorter. The central tooth is not very wide and tri- cuspid, the side cusps being as high as that in the middle. The stomach has a circular band of about 70 small yellowish plates, all of much the same size and usual triangular shape. MARIONIA LEVIS, sp. n. (Plate IV. fig. 4.) Six specimens from Chuaka, East Coast of Zanzibar, and Wasin, East Africa. Two were dissected. The living animals were high and narrow in shape, with a flat back. The sides were described as white, mottled with translucent patches. The ground-colour of the back was a light purplish brown, with stripes of the same colour but darker and others of white. The branchize and rhinophores were pink with dark red blotches. An uninjured alcoholic specimen is 26 millimetres long, 10 high, 1904. ] FROM EAST AFRICA AND ZANZIBAR, 95 and 7 broad in the widest part, but one which was dissected was about twice as large. The colour has become pale green, with a white reticulation on the sides and white stripes on the back. The skin is quite smooth, and there are no tubercles whatever. There are nine or ten pairs of branchie, of which the last three are quite small. The rhinophores have long raised sheaths with simple edges; the club is surrounded by six bipinnate plumes. The velum bears at each end a small grooved tentacle of the usual shape and six processes. The two in the middle are simple and smaller; the other four are larger and branched. The jaws are white and membranous in the smaller and probably immature specimen, yellow and corneous in the larger one. In both there are from 20-30 very large blunt denticles, and also undulations near the edge of the jaw, which in the larger specimen sometimes develop into denticles, so that in about half the length there are two rows of denticles and here and there three. The radula consists in one specimen of 47 and in the other of 45 rows, with a formula of about 804 1.1.14 80, which rises to as much as 85 marginals in one and 100 in the other for a few rows. The central tooth is broad and tricuspid; the median cusp is taller than the others, but not very pointed; all the cusps are rather irregular in shape, and have indentations here and there on the edges. The first lateral tooth is large, blunt, and very different from the rest in appearance. The othersare hamate. The stomach has a girdle of about 150 horny, yellow, triangular plates of different sizes. J do not think that this species can be identified with any of the forms the descriptions of which I have seen*. The color- ation somewhat resembles Tritonia rubra Leuckart and 7’. hawaiensis Pease, but the other details do not coincide. The species differs from the others hitherto found in East Africa in being quite smooth and having no tubercles. MARIONIA ARBORESCENS B. [Bergh, in Semper’s Reisen, xvii. pp. 890-894.] One specimen from near Wasin. The notes on the living animal suggest that it is the same species as M. ramosa, and say that it differs chiefly in that the branchie, rhinophores, and processes of the velum are much smaller. The colour appears to have been the same as in that species (7. e. cocoa and green), and it is noted that there was a greenish tinge in the branchiz. The back was warty. The alcoholic specimen does not look much like IZ. ramosa. It is rather bent, but the length appears to be about 21-5 millimetres, the breadth 11-5, and the height 9. The back and sides are covered with flat low tubercles and the epidermis comes off in flakes. The dorsal margin is unusually prominent and projects 3-2 mm. It * Tn this group as in others I have not access to the descriptions of a few ferms by the older writers, e. g. Tr. palmeri. 96 SIR C., ELIOT ON NUDIBRANCHS [May 17, bears eleven pairs of branchie, the main axis of which is bifid and the secondary axis bifid again. The first pair of branchie are set at the side of the rhinophores, which appear not to be on the dorsal margin, but this arrangement may be due to the contraction of the anterior part of the animal. The velum bears eleven simple processes of irregular length; the outermost are tentacular and grooved as usual. The jaws bear a single row of very large, bent, almost hamate denticles with slight indications of a second row. The radula con- sists of 36 rows, with a maximum formula of 27+1.1.14 27, but in most rows it is only about 154+1.1.1+15. The central tooth is broad, and, as in J. ramosa, seems to bear five cusps. The stomach is provided with the usual girdle of about 100 triangular plates, all of much the same size. This form appears referable with tolerable certainty to JM. arborescens B. MARIONIA RAMOSA, Sp. 0. One specimen dredged in 5 fathoms, north of Kokotoni, Zanzibar. The notes on the living animal are as follows :—“ Colour cocoa- like. Two rows of big branched processes which are greenish in their finer divisions. The rhinophores and processes of the velum very long. The neck part is long and the whole creature has the shape of Limax. Length about 23 inches.” The preserved specimen is of a uniform light yellowish green, much bent, but about 27 millimetres long if stretched out. The back is only 8 mm. across, but the whole animal looks much broader on account of the large branchie. These are thirteen in number, set on the somewhat projecting dorsal margin. ‘The first are a little behind the rhinophores and the last at the end of the tail. None are rudimentary, and the longest are 11 mm. long and almost ribbon-like. The largest tufts consist of three main stems, each of which is trifid again. The velum is ample and bears, in addition to two tentacles of the usual grooved shape, twelve simple digitate processes. The largest are 2 mm. long; the four in the centre are much smaller than the others. ‘The sides and back are tuberculate. The rhinophore sheaths are 5°5 mm. high, with simple but ample and spreading margins. ‘The club is surrounded by five plumes, once or twice pinnate. The jaws are not very strong, and, except that the cutting-edge is yellow, colourless. Both bear about thirty large pyramidal denticles, at the base of each of which is a small accessory denticle. In parts there are traces of a second line, which might be regarded as mere ridges on the first line of denticles, but which in seven or eight cases seem to be independent formations. The transparent radula consists of 45 rows, containing at most 29 laterals, so that the formula is 45 x 29+ 1.1.1+29 as a maximum. The central tooth is much as in Bergh’s plates of Marionia arborescens (S. R. Heft xvi. pl. Ixxxvili. fig. 34), and, as 1904. ] FROM EAST AFRICA AND ZANZIBAR. 97 there, looks as if there were five cusps, but the median cusp is in this specimen thinner and more pointed than in the plates. The liver is large and yellowish. Embedded in the front part of it is the stomach, consisting internally of a large, soft, laminated portion, and a ring of about 120 yellowish, fairly stiff, horny plates. They are not all of the same size, the largest being 2-millimetres long and 1 high, and the smallest about half as large. This species will perhaps prove to be only a variety of M. arborescens, from which it is distinguished chiefly by its long ribbon-like branchiz, which give it a remarkable appearance. The jaws also present some differences. MARIONIA ALBO-TUBERCULATA, Sp. 0. One specimen from the neighbourhood of Wasin, East Africa. Dredged. According to the notes on the living animal the sides were opaque white, with a reticulate pattern of red-brown. At the centre of each mesh was a small white projection. The back, which was dark brown at. the sides and greyish in the centre, bore a similar arrangement of reticulations and projecting spots. The sheaths of the rhinophores were tall, and the wavy edges were turned over outwards. The branchiz were much subdivided, and very large when fully extended. The main stems were of a light greenish grey, and the finer branches of a dark yellowish brown. The velum was plate-like, with five processes on each side, three of which were branched. The alcoholic specimen is 45 millimetres long, 15 high, and 13 broad. It does not taper toa point behind. The colour is dirty yellow with profuse white markings. The stems of the branchie are spotted and striped with white. There are nine pairs of branchize of which the fourth is the largest, but the left- hand plume of this pair is much larger than the other. The middle and left-hand side of the velum are injured. There remain on the right-hand side, starting from the inside, (a) a bifid process, with three branches on each bifurcation, (6) a simply quadrifid process, (¢) a simply bifid process, (¢d) a quite simple process, (¢) a tentacle grooved below. Taking into consideration the notes on the living animal, it appears that there was a similar arrangement on the left side and that the middle of the velum was smooth. There is a small oval papilla below the fourth branchia, close to the dorsal margin. The genital papilla is lower down between the second and third branchie. The jaws are yellow, horny, and large, being 9 millimetres long and 4 wide. They bear a single row of strong denticles, 10 of which are very much larger than the rest. Under five of the largest are indications of a second row. The radula is yellow, and consists of forfy rows with a maximum formula of 95+1.1.1+95. The central tooth is fairly broad and bears three cusps, of which that in the middle is pointed and those at the sides blunt. The Proc. Zoou. Soc.—1904, Vou. IT. No. VII. 7 98 SIR C, ELIOT ON NUDIBRANCHS [ May 17, first lateral is large and clumsy: the rest are of the ordinary hamate shape. The internal organs are whitish. From the large buccal mass issues a tube 14 mm. long and nearly 5 mm. wide. The interior is lined with folds, and there is a pouch-like diverticulum in the floor immediately after the buccal mass. The tube is of much the same size until it dilates into the small stomach (7-5 mm. x 5°5 mm.), which is under and partly within the liver. The stomach has a girdle of more than 100 plates, very thin and membranous, and all of about the same size, namely, 3 mm. along the base and 1:5 high. They lie in groups so as to produce a superficial impression of about 12 thick plates. The intestine is large. MARIONIA VIRIDESCENS, sp. n. [?= Tritonia hawaiensis Pease. | One specimen from near Wasin. The notes on the living animal are as follows :—“ Sides of foot light greenish brown, netted with light bright green, which becomes white near the edge of the back. There is a broad line of opaque white and green (a mixture resembling verdigris), which sends out prolongations to the bases of the branchie. Apart from this line the colour is reddish brown with a greenish network and white spots. This coloration extends into the main stems of the branchie, but the finer ramifications are white and the finest of all bright pinkish brown. The whole coloration is strikingly beautiful. The velum bears seven processes on each side; only the largest are branched. Therhinophores project but little from their pockets, which are as in MZ. albo-tuberculata. The branchize are kept moving continually, expanding and contracting. The animal is about 4 inches long.” The preserved specimen is 42 millimetres long, 21 broad, and 14 high. The shape is not tapering. The back and the sides bear small flat tubercles. The velum is large; besides the two small grooved tentacles it bears on each side seven processes, the largest of which have 2-4 short branches. The central space is wide and bears four rather indistinct tubercles not amounting to processes. The rhinophores are entirely retracted, and the club is surrounded by six bipinnate plumes. There are ten pairs of branchie, of which the fourth is the largest; they still bear traces of green colour. Thestout and strong main stem divides into four branches, each of which bifurcates, and each bifurcation is then 3—4-pinnate. The arrangement of the smaller tufts is simpler, but none are rudimentary. The foot is very narrow, being, as preserved, only 2:5 mm. wide. The mouth is large and open, showing the jaws. Tt is surrounded by a circular disk with thin free margins. All this portion of the specimen seems to have been somewhat distorted by the preserving fluid. The jaws are 9 mm. long and bear a single row of coarse denticles, of which ten are very large, the rest gradually decreasing 1904. | FROM EAST AFRICA AND ZANZIBAR. 99 in size. There are only very faint traces of a second row. The radula consists of 37 rows, with a maximum formula of about 90+1.1.1+90. The teeth are of the shape usual in the genus. The central tooth is finely striated, moderately wide and tricuspid, the central cusp being pointed, those at the sides blunt. From the buccal mass issues a long broad tube (4°5 mm. wide), which passes above and to the left of the genital organs, and then enters the liver, where it dilates into astomach bearing a girdle of plates. These are about 120 in number, horny, fairly strong, brown, triangular, and of various sizes, the largest being 4 mm. long and 1°5 high, the smallest only a quarter of the size and white. The liver is large, yellowish externally, blackish internally. In coloration this animal resembles J/. chloanthes B., but can hardly be identified with that species on account of the differences in the velum, jaws, central tooth of the radula, and stomach-plates. I think it is very probably identical with Pease’s Tritonia hawai- ensis from the Sandwich Islands, but his description is not suffi- ciently detailed to make identification certain, and the expressions “ Veil strongly digitate,” “ Tentacles [7. e. rhinophores] retractile into .. . laciniated sheaths,” hardly apply tothe present specimen. M. virescens and M, albo-tuberculata are closely allied and possibly only varieties of one species; but the present specimens exhibit some differences in the velum, median tooth, and digestive organs, as well as in coloration. MARIONTA, species. One small specimen, dredged in 10 fathoms off Wasin. Tt was dead when found and of a uniform opaque white. The velum was hardly digitate, but presented six undulations. The foot was broad. The alcoholic specimen has become deep brown and is somewhat decomposed. It is 5 millimetres long, 2 broad, and 1°5 high. The back is tuberculate, with a slightly projecting margin, which bears on each side 6 small branchiz set at a considerable distance from one another. The rhinophore-pockets are raised and simple. The velum appears simply circular. No jaws could be discerned. The radula was extremely small, and on a superficial examination appeared to be uniseriate, but on careful investigation was found to have the formula 5.1.5. The laterals are all alike, thin and hamate. They are folded over the central tooth, a narrow plate with slight indications of being tricuspid. The stomach contajned about 80 yellow plates, all of much the same size. This is perhaps an immature form in which the jaws are mem- branous ; but, if so, it is remarkable that the stomach-plates are fully developed. The extreme narrowness of the radula is also remarkable. The characters as described above are sufficient to constitute a new genus, but I hesitate to do this on the evidence of one minute specimen, V* 100 SIR C. ELIOT ON NUDIBRANCHS [May 17, BoRNELLA. _ The members of this genus are slender, elegant animals, having on either side of the back a row of cerata mostly divided into 2—4 branches and bearing gills. On either side of the mouth is a compound tentacular process consisting of a number of conical tubercles set in a sort of rosette. Over the head are a pair of large organs called in the following descriptions for brevity’s sake rhinophore-sheaths, but apparently formed by a fusion of the true rhinophore-sheaths with a pair of cerata. The pair of cerata after these organs are called the first pair. The vent is latero-dorsal between the first and second pair of cerata. The buccal mass 1s not large, but very muscular ; besides the jaws and radula, there is alsoa labial armature of scales. The radula consists of a median tooth, roughly triangular, either smooth or denticulate, and a few (9-19) smooth hamate laterals, bent somewhat forward. The innermost are generally very small, and the size increases towards the outside of the row. There are two stomachs, of which the second is armed with spines, and two accessory livers, besides the main mass. Asa rule ramifications of the liver enter the cerata, but there is some irregularity in this respect. The hermaphrodite gland lies on the liver; the preputium is smooth or armed with spines. There is considerable difficulty in dividing the genus into species. The colour presents little variety, being in all the known forms whitish yellow, with a red or yellow reticulation on the back. On the other hand, there is some variety in both the external and internal organs. The number of the cerata and their sub- divisions appears not to be specifically characteristic, but to increase with age, and is not always the same on the two sides of the body. The ramification of the liver may be present or absent in the same species (B. eacepta; see Bergh’s two descriptions), and, when present, may not extend to all the cerata. The armature of hooks on the preputium may also be present or absent in the same species (B. digitata and B. arborescens; see Bergh). Nine species are recorded, but B. hermanni Angas, caledonica Crosse, adamsii Gr., sempert Crosse, and hancockana Kel., will hardly prove valid, for even if they represent specifically distinct forms they are insufficiently characterised. Of the remaining species B. calcarata Mirch, from the West Indies, is distinguished by having appendiculate rhinophore-sheaths and smooth median teeth. The Indo-Pacific forms fall into two groups—the one represented by B. digitata, with a single process behind the rhino- phores, cerata divided into rather long erect fingers, and median teeth with faint denticulations; the other by J. excepta, with several processes behind the rhinophores, small fingers on the cerata protecting the external branchie, and much more distinctly denticulate median teeth. Whether &. digitata and B. arborescens are really distinct is discussed below. . stmplew, n. sp., 1s certainly a separate species, unless it is a monstrosity. 1904. ] FROM EAST AFRICA AND ZANZIBAR. 101 My specimens seem to be on the whole smaller than those of Bergh and to have fewer cerata. BorRNELLA DiciTaATA Adams. (Plate IV. fig. 5a.) [A. & H., Notes on a Coll. of Nud. made in India, p. 140, pl. xxxili. figs. 8,9; Bergh, S. R. vii. p. 301; id. Danish Exp. to Siam, Opisthobranchiata, p. 199.] Several specimens from Zanzibar Harbour (Bawi and Prison Island). The living animals were white, with a granulated surface. On the back was a reticulate pattern of deep orange. The cerata were tipped with opaque white, below which was a band of bright orange. The transparency of the body-walls varied in different specimens. In some the liver and its ramifications were clearly visible. The following description, when not otherwise stated, applies to the largest alcoholic specimen; the others are much like it, but the smaller ones are only half the size. Length 30 millimetres, breadth 4, height 8°5; much compressed laterally. On each side of the mouth is a large branched process with about fifteen sub- divisions; of these the four or five uppermost are larger and digitate; the remainder are round and tubercular. The back bears a pair of rhinophores with appendages, and, as a rule, four pairs of cerata behind them. The largest specimens have a fifth pair of small cerata, which in one case are fused together into a single process. The rhinophore-sheaths are raised; they bear in front three small digitate processes, and behind one long tapering process which rises 5°5 mm. above the rhinophores. The first pair of cerata are divided into two large and two small fingers ; the right-hand member cf the second pair into two approximately equal fingers, and the left into two largeand onesmall; the third into one large and one small finger; the fourth are simple; the fifth are merely small warts. In the smaller specimens the first pair of cerata are trifid only, and in the smallest bifid, with indica- tions of an incipient third digit. It appears probable that the number of digits increases with size and age. The first pair of cerata bear three branchie, the second, third and fourth two, the fifth none. The branchie are all on the inner side of the cerata and set close together. The labial armature consists of small overlapping scales, arranged in fairly regular rows. The edge of the jaws is quite smooth. The radula consists of 34 rows. The median tooth has a long central cusp, with from 8 to 10 denticulations or ridges at the base. In most rows there are 9 laterals, increasing im size from the innermost outwards, but in some the number rises to 13 and 15. The walls of the second stomach are raised into folds on which are set large brown thorns, with rather blunt tips. The ramifica- tion of the liver appears to be very irregular and to vary in different specimens. In the largest the arrangement is as follows :— A single branch runs up into the tall tapering process behind each. 102 _ SIR C. ELIOT ON NUDIBRANCHS [May 17, rhinophore; the first pair of cerata receive no branches at all ; the second and third receive on the right hand a branch which bifurcates, and on the left a simple branch which, in the third, stops at the base of the cera and does not enter it. The re- maining cerata receive no branches. I think these specimens are the B. digitata described by A. & H. and by Bergh. The best external character seems to be the tapering, finger-like shape of the cerata and of the process behind the rhinophores, to which no doubt the specific name is due. BorNELLA ARBORESCENS Pease. | Bergh, ‘Neue Nacktschnecken, No. ii.,” Journ. Mus. Godeffroy, Heft vi. 1874, p. 96; id. 8S. R. xvii. p. 886.] Several specimens from Mombasa Harbour. Note on living animals: ‘ Yellowish white, with red reticulations on back and red tips to cerata.” The alcoholic specimens are all much of the same size. All are whiter and more compressed than those of B. digitata, and the cerata are much smaller. The average dimensions are :—Length 20 millimetres, height 6, breadth 3; rhinophores and cerata about 2mm. high. The tentacular processes at the side of the mouth consist of only about six digitations. The rhinophores are as in B. digitata, but the posterior process is not solong. In most specimens there are five pairs of cerata, of which the first three are bifid and the remaining two simple. Each, from the first to the fourth, bears two gills, the fifth none. The jaws and labial armature are as in B, digitata; the formula of the radula is about 40x 9.1.9, rising sometimes to 12.1.12. The teeth are much as in Bergh’s plates (Journ. Mus. Godef. l. c. plate iv. 12), but the central cusp of the median tooth is rather longer. The median tooth is more erect than in #. digitata, and the 8-10 denticles which it bears less distinct and very hard to see. The other characters are as in B. digitata. It is not easy to say whether this form is specifically distinct from B. digitata or, if so, whether it should be called B. arbo- rescens. It represents, however, at least a well-marked variety or stage of growth in which the tentacular processes, rhinophores, and cerata are less amply developed. It could hardly be identified with B. arborescens on the strength of the original description by Pease (Amer. Journ. of Conchol. vi. 1871), but in the revised description by Bergh (Mus. Godef. 7. c.) the chief specific character seems to be “papillis anterioribus ut plurimum bipartitis.” In these specimens they are invariably bifid. With regard to the hooks on the preputium, I was unable to see the difference men- tioned by Bergh, and found only simple or bifid hooks, not trifid, in both species. BorvEwLa Excerpta B. (Plate IV. fig. 56.) [ Bergh, Challenger Reports, Nudibranchiata, p. 36; id. Danish Exp. to Siam, Opisthobranchiata, p. 202.] 1904. | FROM EAST AFRICA AND ZANZIBAR. 103 One specimen from the East Coast of Zanzibar. Notes on living animal: ‘“ Rhinophore bearers very large indeed; colour whitish, netted with orange.” The alcoholic specimen is more stoutly built than those already described. It is somewhat bent, and would be at least 30 milli- metres long if stretched out. It is 5 mm. high and 4°5 broad. The rhinophore-sheaths are 8 mm. high, the cerata 5:5. The tentacular process consists of 11 fairly long digits, all distinct and none of them merely tubercles. The large rhino- phores bear 7 digits, three in front quite separate, and five behind united at the base. Posteriorly there are traces of what may be a crest. Behind the rhinophore-sheaths are three pairs of cerata, somewhat resembling those of Doto in general appearance. They all bear three digits, above which rises the top, covered with knobs. They also all bear three stout branchiz, two of which are visible from the outer side and are protected by the digits. The jaws and labial armature are much as usual; the former have blunt indentations on the edge. The radula consists of 27 rows with a maximum formula of 16.1.16. In the median tooth the central cusp is rather longer than depicted by Bergh, and there are 10-12 denticles and ridges on each side. The second stomach and the preputium are armed with black spines as described by Bergh (Chall. Rep. 7. c.). The liver sends branches into all the cerata except the right-hand member of the first pair, but not into the rhinophore-sheaths. Tam somewhat doubtful if this is really Bergh’s B. excepta: there are differences in the arrangement of the cerata and branchiz and the rhinophore-sheaths are relatively much larger, On the other hand, the two specimens examined by Bergh did not agree in details, and the present animal possesses more or less the characters common to them. BORNELLA SIMPLEX, sp. n. (Plate IV. fig. 5c.) One specimen from Chuaka, East Coast of Zanzibar. The following are the notes on the living animal :—“ Very like B. digi- tata, but a distinct species. Anterior tentacles short and simple. Whole coloration transparent, so that the walls of the heart are distinctly visible. No opaque white or orange rings on tips of cerata, but an orange network on the back and a row of opaque white dots on the sides. Eyes not visible. Length 12 millimetres.” Superficially the alcoholic specimen looks much like B. eacepta as described above and has the same Dofo-like cerata, but it is at once distinguished by having on each side of the mouth not the usual tentacular rosette, but a single simple tubercle. The left tubercle is larger than the right. The rhinophore-sheaths bear six short digitations and a larger rounded knob behind. There are four pairs of cerata, of which the hindmost are simple warts. The others are similarly constructed, though the third pair are smaller than the first two. Hach is divided into four knob-like 104 SIR C, ELIOT ON NUDIBRANCHS [May 17, divisions, and bears a pair of trifid feathery branchie, one anterior and one posterior. The mouth-parts were taken out soon after the specimen was captured, and as preserved consist of a labial armature and radula, but no jaws. It is very likely, however, that the jaws had been lost and were not really absent. The labial armature is much as in B. digitata. Many of the scalesare heart-shaped. The formula of the radula is 21x9+4+1+9, the number of laterals being con- stant in all the rows. The median tooth has 7-8 very strong denticles on each side of the central cusp, which does not project much. The laterals are rather short and straight. The second stomach is armed with spines as in JB. excepta. 'The liver sends off diverticula into the process behind the rhinophores and all four pairs of cerata. Those which pass into the rhinophore- sheath and the fourth pair of cerata are simple, while those that pass into the other cerata are divided into four branches corres- ponding to the divisions of the cerata. The simple tentacles of this animal are a sufficient specific, if not generic character, provided they are normal. It is possible that they area monstrosity, for itis not uncommon in nudibranchs for external processes to remain undeveloped, for example, I have a specimen of Ceratosoma cornigerum in which the characteristic lobes are wanting. But apart from the tentacles, this specimen does not exactly correspond with . excepia, for imstance as regards the rhinophore-sheaths and radula. The median tooth has fewer and stronger denticulations; the laterals are fewer, shorter, and straighter. PLEUROLEURA ALBA, Sp. 0. [Cf. Pl. striata van Hass., Eliot in Nudibr. of Maldive and Laccadive Archipelagoes, p. 566-7. | Two specimens from Zanzibar. The following are the notes on the living animal :—‘“ Back white with distinct low ridges, longitudinal but not parallel to median line, each with a yellow line along its summit. The rhinophores stand vertically or point forwards and bear longitudinal perfoliations. The base is white, the main part black, the apex truncated and yellow. ‘They are not retractile into pockets, but can be withdrawn under the mantle- edge. They are not very sensitive. The large velum and the mantle are edged with bright yellow. Foot not half the width of mantle. In crawling, the underside of the mantle is applied to the substratum over which the animal moves. Length 13 milli- metres, breadth 4 mm.” The dimensions and colour of the preserved specimen have not much altered, though the yellow has become faint. The shape is elongate and tapering. The maximum breadth just behind the rhinophores is 4 millimetres, rapidly decreasing to 3 mm. and 2mm. One striation runs down the middle of the back ; on each side of it are six to eight others, not parallel to it and starting 1904. | FROM EAST AFRICA AND ZANZIBAR, 105 from various points. The external characters are those of the genus. The mouth-parts on the whole resemble those of Pl. striata as described by me (J. ¢.). The formula of the radula is 236+1.1.14+6. The rhachidian tooth has a long central cusp and about six denticles on each side. The first lateralis practically half the rhachidian tooth, having one tall cusp and about six denticulations parallel to it and rising from the base on the outside. The remaining teeth are simply hamate. The jaws are more membranous than in Pl. striata, and bear six distinct rows of denticles. This form is closely allied to Pl. striata, but differs strikingly in colour, that animal being black with yellow lines. Such variation in colour is not impossible within the limits of a species, but in this case it is accompanied by other differences :—(1) The shape is more elongate; (2) the radula is narrower; (3) the first lateral is differently shaped. These points seem sufficient to con- stitute provisional specific rank, though it is quite possible that the form may ultimately prove a mere variety of Pl. striata. EXPLANATION OF THE PLATES. Prater III. Figs. 1 a-lg. Notodoris minor, p. 84. la. Lateral view of living animal. 10. Dorsal view of living animal. 1c. Hinder part of body with the valves raised and spread. 1d. Gills with the valves removed. le. Ventral view of anterior part of body. 1f. Glans penis. 1g. Three teeth. 2a-2f. Trevelyana coccinea, p. 85. 2a-2d. First laterals of various shapes. Teeth from (2 ¢) middle and (2,f) end of row. 8a-3c. Trevelyana ceylonica, p. 86. 3a. First lateral tooth. 36. Teeth from the middle of arow. 3c. Three teeth, seen from below and behind. 4, Trevelyana crocea, p. 87. Puate LV. Figs. la-le. Prevelyana bicolor, p. 89. la. Lateral view of living animal, with the liver (the blackish tint) showing through the translucent body-wall. 16. Ventral view of crawling animal, showing the proportions of the foot and some of the internal organs through translucent body-wall. 1c. Gills as seen fully expanded. Figs. 1 q@and 16 are X 5. 2. Nembrotha cristata, p. 90. Middle of radula. 3 a-3d. Nembrotha affinis, p. 92. 3a, dorsal and (3 6) lateral views of living animal. 3c, median and (3 d) first lateral teeth. 4, Marionia levis, p. 94. 5. Median teeth of (5a) Bornetla digitata (p. 101), (66) B. execepta (p. 102), and (6c) B. simplea (p. 108). 106 ON A NEW SPECIES OF TREE-FROG. [May 17, 2. Description of a new Tree-Frog of the Genus Hyla, from British Guiana, carrying eggs on the back. By G. A. BouLencEr, F.R.S., V.P.Z.8. [Received April 30, 1904. ] (Plate V.) HYLA EYANSI. Tongue subcircular, slightly nicked behind, slightly free behind and on the sides. Vomerine teeth in a strong transverse series, narrowly interrupted in the middle, between the very large choane. Head as long as broad; snout obtusely pointed, a little longer than the diameter of the orbit; canthus rostralis strong, straight; loreal region very oblique and concave; nostril much nearer the tip of the snout than the eye; interorbital region slightly concave, nearly as broad as the upper eyelid ; tympanum very distinct, three-fifths the diameter of the eye. Fingers free; no distinct rudiment of pollex; toes three-fourths webbed, the web reaching the disks of the third and fifth, two last phalanges of fourth free; disks a little smaller than the tympanum ; no tarsal fold. Skin finely granulate above, more coarsely on the belly and under the thighs, smooth on the throat. Brown above, speckled and spotted with darker, limbs with regular dark cross-bars ; white beneath, throat, breast, and limbs speckled with brown. From snout to vent 75 millim. The unique specimen, a female, carries its eggs on the back, fitting into shallow hexagonal impressions in the skin; these eggs, 22 in number, measure 8 or 9 millim. in diameter, and contain tailed larve with rudimentary limbs and with allantois-like membranous respiratory organs. In this it agrees with Hyla goeldi Blgr.*, to which it is nearly related, and with Ceratohyla bubalus Esp. This remarkable Frog was obtained at the end of November, 1902, at Groete Creek, Essequibo, by Dr. R. Evans, who has kindly presented it to the British Museum, together with a sketch made from the fresh specimen, and showing the embryos in position ; these have unfortunately come off in transit. EXPLANATION OF PLATE V. Hyla evansi, sp.nov. Natural size. * Cf. Boulenger, P. Z. S. 1893, p. 209, pl. x. figs. 1-8. + Cf. Boulenger, P. Z. S. 1908, ii. p. 115, fig. 8. Ea SOO AG ol MPV J.Green del.et lth . Mintern Bros .imp. INA TE VAIN Syl 1904. ] ON THE ANATOMY OF CERTAIN SNAKES. 107 3. Notes upon the Anatomy of certain Snakes of the Family Boide. By Frank E. Bepparp, M.A., F.R.S., Pro- sector to the Society. [Received May 2, 1904. ] (Text-figures 19-23.) The Boide are held by most zoologists to occupy a place near the base of the Ophidian series. This view is based chiefly upon the paired lungs, the considerable rudiments of the hind limbs, and upon some other osteological points which are duly summed up by Boulenger*. The viscera also confirm this opinion; and I propose in the following pages to call attention to some new or little-known facts relating to the circulatory system which collectively support it. § Gubernaculum cordis, and Right and Left Aorte. No member of the genera Python, Boa, Hunectes, and Hryx which I have dissected possesses any trace whatsoever of a gubernaculum cordis tying down the apex of the heart to the walls of the pericardium. It is not altogether unnecessary to record the absence of a gubernaculum, though it has been stated that the Lacertilia are to be contrasted with the Ophidia by the presence in the former and the absence in the latter of this gubernaculum. A more correct statement would be arrived at if the word “ generally” were interpolated in both cases. I find, in fact, considerable vestiges of this tag in several Ophidia. It occurs for example in Coronella getula. In Colopeltis monspessulana a thin sheet of membrane runs from the ventricle some little way above the apex to the vena cava and passes down the latter to the posterior wall of the pericardium. In the Hamadryad (Ophiophagus bungarus) the covering mem- brane of the heart (=visceral layer of peritoneum) is obvious and can be stripped off. Posteriorly, this membrane forms a tubular prolongation of which one side is attached to the vena cava at its entrance to the heart and to the pericardium beyond, while the other is attached to the pericardium wall behind. The ventricle, therefore, near to the apex, but on one side, is attached not merely to the vena cava, but also to the wall of the pericardium. It isimpossible to speak of these structures but as a gubernaculum cordis. On the other hand, it is clear that they differ somewhat from the gubernaculum in the Lacertilia, which has no relation to the vena cava but attaches the actual apex of the ventricle to the pericardial wall, and is of more ligamentous consistency. _ The structure is therefore possibly a new one in the Ophidia, on which view its total absence in the Boide (so far as the material which I have examined enables me to say) may well be * Catalogue of Snakes in the British Museum (Natural History), London, 18938. 108 MR. F. E. BEDDARD ON THE [May 17, a primitive character. There is no @ priori objection to deriving the Ophidia from some Lacertilian form in which the characteristic lacertilian “tag” to the heart was absent. In the genus Varanus, for example, the gubernaculum is absent, as others as well as I myself have observed; and it may be pointed out that the Ophidia might well have been derived from some form in which, as in Varanus, the neck was long, the lungs firmly bound down to the dorsal parietes, the trachea or bronchi continued for a considerable distance through the lung, and the urinary bladder absent. The fact that in Hryx the right and left aortic arches are equal in size at their junction to form the unpaired dorsal aorta seems to me to be undoubtedly a primitive feature. J may furthermore observe that this feature is figured * by Dr. Gadow as charac- teristic of Pelophilus (Boa) madagascariensis. Apparently, how- ever, Python bivittatus has unequal right and left aorte 7. In other serpents it is common for the right aortic arch to be smaller than the left, and this is carried so far in Zamenis flagellr- formis as to give the impression that the right aorta is a mere inconspicuous forwardly running branch of the left £. §$ Iniercostal Arteries. The intercostal arteries in Hryx jaculus show some interesting features, which are partly indicated in the accompanying drawing (text-fig. 19, p. 109). In the anterior region of the body the arteries in question arise from the left aorta immediately after it has parted company with the anterior vertebral. Anteriorly to this point some arise from the left aorta, but that region of the body is supplied from the vertebral artery. The intercostal arteries which arise from the right aorta and from the first part of the conjoined aortz are single trunks given off at irregular intervals not corresponding to the individual vertebra. They join above, however, to form a continuous and slender dorsal artery which may be termed the posterior vertebral artery (P.v.); from this arise at regular intervals the paired intercostals, After the end of this vertebral artery the aorta continues to give off the dorsal intercostals which, when they reach the median dorsal line, run along it for a short distance anteriorly and posteriorly, giving off as before paired branches to the intervertebral regions. But there is no formation here of a continuous longitudinal trunk running over more than three or four vertebre. So far the arrangement is precisely such as I have lately described in another Boid, viz. Python spilotes§. But whereas in the last * Bronn’s Klassen und Ordnungen des Thier-Reichs, Bd. vi. Abth. i. pl. cxxxv. fig. 1. + Loe. cit. pl. xxxiv. fig. 2. The figure is copied from Fritsch. + Beddard, “Contributions to our Knowledge of the Vascular System in the Ophidia,” P. Z.S. 1904, vol. 1. p. 338. § “Contributions to our Knowledge of the Circulatory System in the Ophidia,” P. Z. 8. 1904, vol. i. p. 362. _ ; : 1904. | ANATOMY OF CERTAIN SNAKES. 109 named snake this arrangement of the intercostals persists to the end of the body, in Hryx jaculus another arrangement comes into force further down the body which is displayed in the drawing reproduced as text-fig. 20, p. 110. First of all, ¢. e. anteriorly, Text-fig. 19. SENSES SS (UFR aN S Ho. 7 pe SSS << Part of intercostal arterial system of Hryw jaculus. Ao., right, Ao., left aortic arch ; es., esophageal twigs; P.v., posterior vertebral artery; V., vertebral artery. the intercostal arteries begin to arise regularly from the aorta, and, when they reach the parietes, do not form short longitudinal vessels which give off paired branches to several successive 110 MR. F. E, BEDDARD ON THE [May 17, vertebre ; but each intercostal bifurcates close to the parietes'and supplies but a single intervertebral area. The arrangement in this region of the body in fact is precisely like that figured by Jacquart * for Python sebe, and which I can confirm from my own dissections of the same serpent. In Python sebe, however, this arrangement appears to persist throughout the whole body. In Eryx, on the other hand, a little way back, a third mode of arrangement of the intercostal arteries occurs. The point of bifurcation of the single intercostal advances higher up its stem (@ in text-fig. 20) until (6) a right and left intercostal is established arising separately from the aorta. A further differentiation is shown in the case of the intercostals lettered ¢ and d in the figure referred to. It will be observed that in c the left intercostal is much thicker than its fellow and than most of the neighbouring intercostals, while in d the vessel has become single owing to the complete disappearance of its fellow. These facts indicate the way in which the irregular intercostal arteries (sometimes right and Text-fig. 20. Part of intercostal arterial system of Eryx jaculus. Ao., aorta; a, 6, c, d, intercostal arteries; C.v.7., C.v.l., right and left posterior cardinals; G., gastric artery; Z., end of liver; S.mes., superior mesenteric artery. sometimes left and at unequal distances) of more modified snakes have been formed. Obliterate in the accompanying drawing (text-fig. 20) the finer intercostals and leave only those of magnified calibre, and the result would be a reproduction of the intercostal system in many Colubrine Snakes. This double series of intercostal vessels in Hryx has its counterpart in Hunectes murinus. In the Anaconda I find both intercostals arising singly which run for varying distances along the median dorsal line giving off paired branches to the parietes, and regularly paired intercostals arising separately from the dorsal vessel It is important to notice the agreement in these particulars between Eryx and Hunectes and the difference from Python, since the two former belong to the subfamily Boine, and the latter, naturally, to the subfamily Pythonine. It 1s difficult to say which of the * Ann, Sci, Nat. (4) iv. p. 321. 1904. ] ANATOMY OF CERTAIN SNAKES. 111 two arrangements of the intercostals, the Boine or the Pythonine, is the more primitive ; perhaps it is the former. In any case it is from the Boine rather than the Pythonine type that the inter- costal system of many Colubrine Serpents is derivable. Lryz conicus differs in detail from Eryx jaculus in the arrange- ment of the intercostal arteries. Anteriorly, the arteries are given off singly and regularly, bifurcating just before their entry into the thickness of the parietes. There is with great regularity one to each vertebra. The only exception that I noticed was in the case of one intercostal arising from the right aorta in front of the junction of the two aorte, and of another some way behind ; these supplied two intervertebral spaces. This apparently is the only trace left anteriorly of the arrangement characterising Eryx jaculus. Far back, much further than in £. jaculus, the paired arrangement of the intercostals is seen. Still it is evident that the two species do not differ in the type of the arrangement of the intercostals. Eryx johni again shows a fundamental likeness to, but detailed differences from, the other two species of the genus. The general agreement is that the anterior series of intercostals arise singly from the aorta and bifurcate only just before entering the parietes. They begin to be double in origin shortly behind the liver. In the region where they are double they are frequently asymmetrical in both size and in point of origin from the aorta. Thirteen intercostals arise from the right aortic arch before its union with the left. There is no trace of any formation of an azygos median vertebral artery such as occurs in Eryx jaculus. The intercostal system of this species in fact is somewhat intermediate between those of the two other species of the genus. § Some Visceral Arteries. Esophageal arteries —I have examined these arteries carefully in an injected example of Eryx jaculus. It agrees with other snakes and with the Lacertilia in the fact that the intercostals (already described on p. 109) arise from the right aortic arch only. On the other hand, branches to the esophagus, which are represented in the drawing (text-fig. 19, p. 109), arise either directly or indirectly from both arches. I have not observed this double origin of the cesophageal branches in any other Ophidian ; but, as I am not quite in a position to deny its occurrence, I cannot emphasise the fact as a characteristic of the Boide. It is, however, 1 am inclined to think, an anatomical feature not found in the Lacertilia. The cesophageal vessels, or rather vessel (for I only noticed one), arises from the left aorta ; it passes back along the cesophagus, giving off branches to that gut, and becomes continuous with the first of the csophageal arteries arising from the conjoined aorte. The right aorta does not directly give off cesophageal arteries. But from two of the intercostals of the right aortic arch such arteries arise, 112 MR. F, E. BEDDARD ON THE | May 17, In Hryxjohni the same series of cesophageal branches arising from intercostals are present. They arise from intercostals partly belonging to the right aortic arch, and partly to those arising from the common trunk. The third intercostal, after the union of the two aortee, gives off a slender vessel which runs forward and joins a vessel arising from the last but four of the intercostals belonging to the right aorta. This longitudinal trunk gives off several lateral vessels. The third, fourth, and fifth of the intercostals of the right aortic arch also give off single cesophageal vessels. I have pointed out in another paper* that the Lizard Pygopus is unusual by reason of the fact that some of the visceral arteries arise from intercostals instead of directly and independently from the aorta. In Hry« johni precisely the same mode of origin occurs not only for cesophageal arteries, but for a fat-body artery. This springs from the right-hand intercostal of the seventh pair after the posterior renal artery. I observed the exact converse of this state of affairs in Tropidonotus fasciatus. The right aorta gives off intercostal branches, but no twigs to the cesophagus that I could find. On the other hand, a single parietal vessel, accompanied closely by a vein, enters the parietes a good way to the left of the middle dorsal line and arises unmistakably from the left aortic arch, which also, of course, gives off several branches to the cesophagus. Gastric arteries.—The fact thatin Hrya there are only twot and in Python spilotes only three gastric arteries, appears to me to be an archaic point of structure in these Boid snakes. Among the Ophidia generally there is frequently a large number of gastric arteries. For example, inthe genus Coluber I have found as many as ten or eleven. The reduplication of these and other arteries, so characteristically Ophidian, seems therefore obviously to mark the more specialised members of the group. The absence of, or less, reduplication is not therefore inconsistent with the less modified, more archaic, structure. Ovarian arteries.—It is at least rare among snakes ¢ for the arteries supplying the gonads to arise from the aorta opposite to each other. Asa rule one spermatic or ovarian artery follows the other in relation to the asymmetrically placed gonads. Never- theless in a female Hryx conicus the two arteries arose side by side. They immediately follow, as is usual, the superior mesenteric. The paired condition of these arteries seems to me to be a primitive feature in the organisation of this snake. Renal arteries.—It is the general rule among the Ophidia for each kidney to be supplied with a considerable number of arteries. There are, for example, as many as eight m Coluber catenifer. Among such Boide as I have examined, the number is invariably one or two arteries only to each kidney. In Python sebe, Hunectes * Above, p. 12. + T could find only one in Hrya johni. + I have not myself observed a single instance in the Ophidia except in the case mentioned above. 1904. } ANATOMY OF CERTAIN SNAKES. i053 murinus, Hryx johni, and Hryx conicus each kidney had only one renal artery. In Hryx jaculus there was some variation in the two examples studied. In one there were two renal arteries for each kidney ; in the other the left kidney had two arteries, the right only one. Here, it may be remarked in passing, is an apparent difference between the two species of Hry«x investigated by me, which f shall refer to again later. Now, as the kidneys of Hrya are very short * and those of Coluber catenifer long, as they are in the majority of snakes, it might be held that the arterial blood-supply had merely a relation to length. That the character is one peculiar to the Boidze seems to be shown by the case of Hunectes murinus, for in this serpent there is only one renal artery to each kidney; and yet those organs, in the individual which [ dissected, measured respectively 15? and 123 inches in length. § Veins of the Posterior Abdominal Region in Kunectes. The caudal vein emerges from the thickness of the parietes some way behind the cloaca. When it reaches the level of the cloaca, two veins, asymmetrically placed with regard to one another, join it. I suppose that these are the equivalents of the ischiadic veins of Lizards. Further forward, between the cloaca and the very anteriorly situated kidneys 7, the caudal vein divides at once into three branches. The middle one of these is the right and larger anterior abdominal vein. To the left of this arises the afferent renal vein of the left kidney, and to the right of the anterior abdominal a vein which runs over the viscera to the dorsal surface of the body, where it enters the parietes to the left of the middle line, after running both forwards and backwards for a short distance. I am inclined to regard this vein as the equiva- lent of the lateral abdominal vein of its side in the Lacertilia. Its place of origin agrees with such an homology, and the shortness of its course within the body-wall is no reason against the ecom- parison, since it is of varying length among the Lacertilia. It might be held that this vein is in reality only the proximal end of the left afferent renal which has lost its connection with the remainder of that vein, were it not for the conditions observable in Hryx to which I recur later (see p. 114). There is, in fact, no afferent renal on this side of the body springing from the caudal vein and corresponding in origin to the fully developed afferent renal, whose origin on the other side of the anterior abdominal has been already referred to. It rises, I presume, from the eaudal vein further back. * In Hryx johni, which measured 26 inches from snout to vent, the kidneys were respectively = inch (right kidney) and 1 inch (left kidney) in length. In H. jaculus of 15% inch length to vent, the kidneys were proportionately larger, i. e. inch and 1 inch, and in #. conieus of 27 inches this was also the case ; they measured 14 and 1+ inch. + The kidneys are well known to occupy a very anterior position in the Boidz. Tt has not, I think, been noted that Hryx as well as Humectes agrees in this with Python and Boa. Proc. Zoou. Soc.—1904, Vou. IT. No. VIF. 8 114 MR. F, B. BEDDARD ON THE [May 17, The anterior abdominal vein is double, but the right vein is very considerably larger than the left, especially posteriorly. The two join far forward in the immediate neighbourhood of the pancreas. They are here more nearly equisized. Posteriorly there is another and the only other junction between these paired veins not far in front of the trifurcation of the caudal vein. After this point the left anterior abdominal vein receives a branch from the body-wall and ceases. It is not directly connected, as already mentioned, with the caudal vein. Just after its origin from the caudal vein the anterior abdominal gives rise to the single median epigastric vein, which runs forward at least as far as the liver, to which it gives off several branches. § Veins of the Posterior Abdonunal Region mm Hryx. It is remarkable that differences occur in these veins between the two species Lryx jaculus and EH. conicus, if, that is to say, Text-fig. 21. Certain abdominal veins in Erya conicus. a, b, ¢, parietal veins; Ant.Abd., anterior abdominal ; Ep., epigastric; L.aff-r., lett afferent renal; R.affz., right afferent renal. the single example of #., jacwluws which I have examined represents the normal state of affairs. The veins in question in Zryx conicus 1904. ] ANATOMY OF CERTAIN SNAKES, 115 (text-fig. 21, p. 114) are disposed as follows:—The caudal vein gives off first the right afferent renal; after the cloaca (¢. e. in front of that aperture) it gives off two branches to the dorsal parietes and then divides into three veins as in the Anaconda. Shortly after its origin, the anterior abdominal gives rise to the single epigastric vein. The arrangement of veins in this snake is in Text-fig. 22. Raf Loffy. 5 Certain abdominal veins in Eryx jaculus. R, branches to rectum; cl., branches to cloaca. Other letters as in text-fig. 21. fact precisely as in Humectes. In Hryx jaculus, on the other hand, there are differences which are illustrated in the accom- panying figure (text-fig. 22). The caudal vein apparently divides into two—but Iam a little uncertain—and each of these vessels g* 116 MR. FE. E, BEDDARD ON THE [May 17, gives off at least one branch to the rectum. A little further forward each vein certainly divides into two branches. These branches are, on each side, a renal afferent and a component of the anterior abdominal vein. Both afferent renals receive, as usual, branches from the dorsal parietes. The two roots of the anterior abdominal vein are unequal in size, that of the right hand being considerably the less. It is important to note this fact, since in Hryx conicus the left root of the anterior abdominal vein is the only one which persists. The epigastric vein arises from the anterior abdominal shortly after the junction of its two roots. Lryx johni agrees with H. jaculus. It is a peculiarity of Snakes as contrasted with Lizards, that the anterior abdominal is occasionally partly double, whereas in Lizards it is single after the fusion of its two roots. In the Colubrine and Viperine snakes, so far as my experience goes, the anterior abdominal is usually single except at its extreme posterior end. In Zamenis gemonensis the vein bifurcates posteriorly, and after a very short course ends in minute branches in the fat-body. In Causus rhombeatus the extent of the bifurcate region of the anterior abdominal is not much greater; for 8 inches intervene between the opening of the anterior abdominal into the portal and its bifurcation posteriorly, which is 2% inches from the vent. In Boa constrictor, on the other hand, there are 25 inches between the vent and the fusion of the two anterior abdominals anteriorly, which point is 10 inches behind the liver, and therefore less from the point of union of anterior abdominal and portal *. Tn the Anaconda the double character of the anterior abdominal vein has been already referred to. In onespecimen of Hryx conicus it was single throughout. In another it was partly double, as was also the case with two individuals of Hryx jaculus. In Python sebe (where it is figured as partly double by Jacquart 7) the fluctuation of this vein between the single and double condition was more plainly seen. Just in front of the gall-bladder the vessel communicates with the gastric portal vein ; from this point to two inches behind the gall-bladder it is single. Then fora distance of 44 inches it is formed of two tubes lying side by side ; these reunite, and finally again separate to form two tubes. In Eryx johni the vein appears to be single after the union of its two posterior roots. Jacquart figures a somewhat different state of affairs in Python. The anterior abdominal bifureates posteriorly and communicates with only one afferent renal directly as in the Anaconda. The other branch only communicates indirectly (by means of small veins) with the left renal afferent. Hochstetter { observes that in Tropidonotus natrix and Coluber csculapii there is no direct connection between the abdominal and the renal afferent veins. TI can quite confirm this by the conditions observable in Zamenis * T did not ascertain this measurement. + Ann. Sci. Nat. (4) vi. p. 321. { Morph. Jahrb, xix. p. 489. 1904.] ANATOMY OF CERTAIN SNAKES. Thy gemonensis. I take it that the posterior bifurcation of the anterior abdominal vein in the last-mentioned Colubrine snake (and possibly in Cawsus and other forms) is a reminiscence of its former origin by two roots from the renal afferent veins as in the less modified Lacertilia, and, as I have already shown, in the especially “ Saurian” Hryx jaculus and Eryx johni. It is clear from the foregoing, that the somewhat divergent opinions of previous authors are partly due to actual differences in the abdominal and afferent renal veins of different Ophidia, to our knowledge of which I have been able to add something. It is furthermore clear that the Boidee contrast with other Ophidia, so far as observation has gone, in thei greater approximation to the Saurian type of organisation ; they are, in fact, more primitive than other Ophidia. In these Ophidia alone is the anterior abdominal vein connected with one or both of the afferent renals, and in them there is generally doubling of the anterior abdominal vein in front of the junction of its constituent veins. Of all the Boidee whose anatomy is known, Lry«x jaculus (not EH. conicus) comes nearest to the Saurian type in that its anterior abdominal vein arises from two distinct roots, one from each of the renal afferent veins. § Remains of Cardinal Veins. In comparing the venous system of the adult Tropidonotus with that of the adult Lacerta, Dr. Hochstetter* arrived at the conclusion “dass das Venensystem der Lacerta. auf einer etwas niedereren Entwicklungsstufe strehen geblieben ist als das von Tropidonotus. Bei Lacerta ist das System der Vertebralvenen noch erhalten, wahrend es bei Tropidonotus nahezu vollig gesch- wunden ist.” Hochstetter’s statements are also accurate when applied to other Colubrine snakes. Ihave examined Lioheterodon madagascariensis with some care from this point of view. In that snake there is hardly any development of vertebral veins, such as I shall describe immediately in the Boide. About half- way down the liver, however, is a longitudinal vessel running for a short distance, from which arises a tributary to the hepatic portal system. Another Colubrine snake, viz. Zamenis gemonensis, showed an interesting persistence of a portion of the posterior cardinal vein precisely comparable to what is to be found in the Chameleon t and in Pygopust. The afferent renal artery in front of the left kidney, instead of ending, as is usual, towards the anterior end of that gland, passes beyond it and imbeds itself in the body-wall to the left of the median dorsal line. I could find nothing to correspond on the right side. If really absent— and Jam convinced that there is at least nothing really con- spicuous—this is another example of the asymmetry of Ophidian * Morph. Jahrb. xix. p. 493. + Hochstetter, loc. cit. p. 462, and Beddard, loc. cit. infra. ft Beddard, “Contributions to the Anatomy of the Lacertilia, No. 8,’ P. Z.S. 1904, vol. 11. p. 12. 118 MR. F, E, BEDDARD ON THE [May 17, structure. The discovery of this prerenal portion of the posterior cardinal in an Ophidian removes another point of dissimilarity between the Ophidia and the Lacertilia. Text-fig. 23. Liver and certain adjacent veins in Eryx conicus. Ao., aorta; cs., cesophageal branches of longitudinally running left posterior cardinal; Int., intercostal branches of the same; DL, liver; P.v., portal vein ; St., branches to stomach. The Azygos vein is admittedly a vestige of the posterior cardinal 1904. ] ANATOMY OF CERTAIN SNAKES. 119 of its side. This vessel appears to be well developed in some Boid serpents, and that fact appears to me to indicate the retention of a primitive feature*. In Hryx jaculus, for example, the vein, which is on the right side, extends back over no less than twelve intercostal spaces and nearly reaches the junction of the two aorte. This contrasts with the same vein in Coronella getula, which only extends over four of these spaces. In a specimen of Hryx conicus the azygos vein extended only over ten vertebra. In both specimens of Hryx conicus the azygos, after a break, reappeared in the region of the liver, where its course is shown in the accompanying drawing (text-fig. 23, p. 118). It will there be seen that the vertebral vein with one gap near the anterior end of the liver runs continuously to a point some little distance behind the liver. It gives off branches on the one hand to. the dorsal parietes, and on the other to the portal vein. Anteriorly to the liver the branches go to the ceesophagus. In Hrya jaculus (text-fig. 20, p. 110, C.v.1., C.v.r.) there were conspicuous traces of both posterior cardinals behind the liver. As will be seen in the drawing referred to, the left posterior cardinal is continued anteriorly beyond its junction by a conspicuous branch with the portal trunk in the immediate neighbourhood of the liver ; posteriorly it ends near to the commencement of the right- hand vein, the two being therefore supplementary to each other. The left is considerably the longer. The right extends back a little way beyond the origin of the superior mesenteric artery from the aorta. In Python sebe a corresponding vein occurs in the region of the liver, but it extends both further forward and backward than I have observed in Hryx. Anteriorly it extends beyond the junction of the two aorte, and posteriorly it reaches very nearly to the gall-bladder. I do not for the present suggest that these longitudinal vessels are more developed in the Boide. I simply call attention to their arrangement. § On the Specific Differences between Eryx jaculus, E. johni, and Ki. conicus. These species can be readily separated by external characters, as Boulenger plainly sets forth in the British Museum ‘ Catalogue of Snakes.’ They also, however, differ in other points besides the obtuse or pointed tail, the absent or present mental groove, and the form of the rostral scale, &c. There are differences, in the first place, in the form of the liver. In “ryx jaculus and H. johni the lobes are unequal, and the right lobe extends further down the body than the left, to the extent of about half an inch. In Eryx conicus they are as near as possible exactly equal. As these results depend upon the examination of two examples of each of the species /. jaculus and * As I have pointed out in a preliminary account of some of the facts detailed in the present communication (Ann. & Mag. Nat. Hist. (7) xii. p. 233). 120 ON THE ANATOMY OF CERTAIN SNAKES. [May 17, Lf. conicus, the character may, I take it, be regarded as a genuine, if small, difference. Secondly, the right lobe of the lung extends further back in Hryx jaculus and H. johni. In those species it reaches the gall- bladder ; in Hryx conicus it falls short of the gall-bladder by at least an mch. The superior mesenteric artery arises in Hryx jaculus and LH. johni distinctly in front of the gall-bladder. In Eryx eonicus its point of origin is as distinctly behind the gall-bladder. This character is perhaps less likely than some others to prove of value as an absolute mark of distinction. More important are other differences in the vascular system. In Fryx jaculus, though it is the smaller species, there are appa- rently, as a rule, two arteries supplying each kidney; in Lryax conicus there is but one to each kidney: and here apparently Lf. johni agrees with H. conicus. The intercostal arteries have a different arrangement m the two species. In #. jaculus there is a well-developed posterior vertebral artery formed by the junction of irregularly arising intercostals; in /. conicus this does not exist (except as the merest rudiment). The paired intercostals com- mence further forward in /. jaculus than in Z. eonicus. H. johni m these particulars is somewhat intermediate. In 4. jaculus there is a double connection of the anterior abdominal vein with the afferent renals posteriorly; in #. conicus this occurs only on one side. HH. johni agrees with H.jaculus. In relation to these anatomical differences, which appear to me to be fully as great as those which distinguish either species from Hunectes, I would draw attention to the restricted range of #. conicus and to the wider distribution of H. jaculus and EL. johnt. Résumé of principal facts. It may be convenient to deduce from the foregoing pages the main facts in the vascular system of the Boide examined ; such may be divided into two heads, 7. ¢., those which appear to argue a basal position among the Ophidia, and those which are of sys- tematic importance in the group. As to the former it may be noted that (1) The heart is always without the least trace of a guber- naculum cordis. (2) The two aorte are usually equisized at their point of union to form the dorsal aorta. (8) The renal and gastric arteries are much fewer in number than in other Ophidia, the former consisting generally of only one artery to each kidney. This distribution has no relation to the size of the kidney. (4) The intercostal arteries are always symmetrical and for the most part regular in their arrangement, frequently in regular pairs. 2) Te P.Z.S.1904:,vol.1.PLVI. Mintern Bros.imp. G.S5.B.del.G.West kth. ENTOMOSTRACA FROM NATAL. PZ 19047 ol ll Pavel mcrae ne ere Mintern Bros.imp. G.S.B.del. G. West lth. ENTOMOSTRACA FROM NATAL. PZ.3 100470! ll Pie He NEE ELA A OPT RAY renee fp EGER \E ST, Mintern Bros imp. G.S:B.del. G. West lth. JEN NO) MOS MMRV NGA. TINO IN AMIAIL, 1904.] ON ENTOMOSTRACA FROM NATAL. © 121 (5) The anterior abdominal vein always arises from one afferent renal, rarely from both as in the Saurians. Frequently, also, it is double. (6) The existence of veins continuing the azygos posteriorly is usual, Facts which are of importance for the systematic arrangement of genera and species within the family Boidee :— (1) The Boinz (Hunectes and Hryzx) differ from the Pythoninz (Python) in that the intercostals are posteriorly paired arteries, while in the Pythoninz a single median artery divides into two close to the median dorsal line throughout the series. (2) The three species Hry« conicus, E. johni, and EF. jaculus differ from each other ina large number of anatomical features. Besides these points several other anatomical features are of interest as new or rare among Ophidia. Such are (1) The continuation of the afferent renal of the left kidney in Zamenis gemonensis beyond the kidney into the parietes, as in Chameleon and Pygopus. (2) The origin of cesophageal arteries not only from the left aortic arch but from some of the intercostals of the right half arch. (3) The fusion of some of the anterior intercostals in Hry« jaculus (and Python spilotes) to form a continuous longi- tudinal trunk lying dorsally of the aorta. (4) Representatives of the lateral abdominal vein of Lizards appear to exist in certain snakes (e. g. Hunectes and Hryz). (5) Origin of a fat-body artery in Hryx johni from an intercostal, 4, On Entomostraca collected in Natal by Mr. James Gibson. By G. Stewarpson Brapy, M.D., LL.D., D.Se., F.R.S., C.M.Z.S8. [Received March 25, 1904.} (Plates VI.-VIII.*) For the opportunity of examining and describing these species I am indebted to the kindness of Mr. James Gibson, Resident Magistrate at Greytown, by whom they were collected in the summer of 1902. All are freshwater species, and were found in pools in the neighbourhood of Greytown, Natal. The identity of some of them with European forms is a point of considerable interest, and indeed the general aspect of the gatherings is quite * For explanation of the Plates, see p. 127. 122 * DR. G. STEWARDSON BRADY ON [May 17, similar to what one would expect in ordinary Northern collections. With the exception of one species, for which I have thought it best to propose a new generic name—ALctocyclops—all ave refer- able to familiar European genera. CYCLOPS LEUCKARTI Claus. This appears to be less plentiful in the Natal gatherings than any of the following species, though more conspicuous owing to its greater size. The form described by me years ago under the specific name scourfieldi* has been identified by other authors (Lilljeborg, Herrick, Schmeil) with C. leuckarti Claus. 1am doubtful as to the correctness of this identification. Both Schmeily and Herrick+ figure, with differences, peculiar pellucid marginal lamine on the last two joints of the larger antenne. I have been unable to detect any such structure in my British specimens of C. scourfieldi, neither does it exist in the Natal specimens nor in others from Ceylon which I refer to the same species. And it may be noted that Lilljeborg, while accepting Schmeil’s description and figures as belonging to C. leuckarti, does not himself refer in his definition to these antennal lamin. Herrick, on the other hand, figures and describes them, and expresses a doubt as to the identity of the species with C. scourfieldi. I do not myself possess authentic specimens of C’. lewckarti, and must be content for the present to leave the question im suspense. CYCLOPS PUSILLUS, n. sp. (Plate VI. figs. 11-18.) Female. Body slender, tapering gradually from before back- wards (fig. 11); thoracic segments not produced at the angles, except the last which is extremely short and angulated: abdomen very slender, the first segment nearly as long as the united lengths of the following three segments ; furcal segments slender, scarcely longer than the preceding abdominal segment; principal tail-seta nearly as long as the whole abdomen. Ovisacs closely adpressed to the abdomen and containing only a few comparatively large eggs. Antennules twelve-jointed, slender, somewhat longer than the entire cephalothorax, bearing numerous long setz (fig. 12). All the branches of the first four pairs of feet three-jointed ; spines of the first pair (fig. 17) very slender and setiform ; fifth pair (fig. 18) rudimentary, almost obsolete, consisting of a minute papilla from which arise two unequal sete. Length, exclusive of tail-setee, 0°46 mm. This is the smallest species of Cyclops with which I am acquainted. But its general development and the fact of many specimens bearing ova, indicate that it is not an immature form. * Brady, G.S. “A Revision of the British Species of Freshwater Cyclopide and Calanide ” (Nat. Hist. Trans. Northumberland & Durham, vol. xi., 1891). + Schmeil, Deutschlands freilebende Stisswasser-Copepoden. 1 Teil, Cyclopide. 1892. { Herrick, Synopsis of the Entomostraca of Minnesota, 1895. 1904.] ENTOMOSTRACA FROM NATAL. 123 CyCLOPs GIBSONI, n. sp. (Plate VI. figs. 1-10.) Female. Seen dorsally the outline is slender, gradually taper- ing from before backwards, greatest width equal to about one- third of the length (fig. 1); the second segment as wide as the cephalic segment, slightly produced and angulated posteriorly, third segment narrower and scarcely at all produced at the angles, fourth segment again narrower and without produced angles; last thoracic segment about as wide as the fourth from side to side, but much narrowed from before backward, its lateral angles distinctly produced and bearing a brush of six or eight rather long sete ; abdominal segments gradually tapering back- wards, the combined lengths of the second and third equal to that of the fourth, first segment rather larger than the fourth ; caudal stylets slender (fig. 10), nearly equal in length to the abdomen, about five times as long as broad; the innermost of the two’ apical setee is the longer and is about equal in length to the entire abdomen, outer setee somewhat shorter; the subsidiary sete are three in number, short, subequal, and arise from the apices of the stylets; on the side of each stylet not far from the distal end is a single very minute seta; the larger sete are very finely plumose. The antennules are twelve-jointed (fig. 2), slender, bearing numerous sete, and reach backwards as far as to the front of the second segment: the proportionate lengths of the joints are indicated in the following formula :— I 24 Sol 4a Os OME ESes (Olas Rtn PEL Om eLpe Say ey aEe ae The mandible is of the usual form, with a rudimentary palp of three setze, two long and one short (fig. 3); the chewing portion of the maxilla (fig. 4) is of normal shape, with a laminar palp bearing several plumose sete. The rami of the first four pairs of feet are all three-jointed ; terminal joint of the outer branch of the first pair bearing two slender spines on the outer margin, three long sete on the inner margin and two shorter apical sete, the outer- most being finely bipectinate (fig. 7); the terminal joint of the inner branch has one long apical spine and on its outer margin one very short spine and one slender seta, its internal margin has three setz and one at the apex. The second, third, and fourth pairs (fig. 8) are nearly similar, but with more elongated joints and more robust spinous armature. The fifth pair (fig. 9) is uniarti- culate, with three subequal setz, the innermost of which is densely plumose. Length 1 mm. Among European species those most nearly allied to C. gibsoni seem to be C. varicans, C. bicolor G. O. Sars and C. gracilis Lilljeborg; but all of these have the rami of the swimming-feet uniformly biarticulate, whereas in C. gibson they are triarticulate. CYCLOPS FIMBRIATUS Fischer. This appears to be one of the commonest species in Mr. Gibson’s 124 DR. G. STEWARDSON BRADY ON [May 17, collection, agreeing down to the minutest details with British specimens. EcrocycLops, nov. gen. Like Cyclops except that the mandibles, maxille, and maxillipeds are less fully developed and are very indistinctly jointed, the posterior antenna bears a very stout ciliated seta on the posterior margin, and the terminal joints of the rami of the swimming-feet are scarcely at all elongated. ECTOCYCLOPS RUBESCENS, n. sp. (Plate VI. figs. 19-20; Plate VII. figs. 21-27.) Female. Body depressed, rather robust (fig. 19) ; cephalothoracie segments not produced laterally ; abdominal segments short and stout, posterior margin of the last segment pectinated, furcal segments not much longer than broad; principal tail-sete marginally aculeated, the innermost nearly equalling three-fourths of the length of the body; ovisacs large and adpressed. Anten- nules (fig. 20) ten-jointed, much shorter than the first cephalo- thoracic segment, and rather densely setiferous ; antenne (fig. 21) robust, anterior margin fringed towards the base with numerous short sete, penultimate joint with about six larger sete, terminal joint with several longer sete; on the posterior margin rising from the base of the second joint is a very long and stout ciliated appendage. Mandibles bearing a rudimentary palp con- sisting of three very small sete; maxille (fig. 22) not much different from those of Cyclops ; anterior and posterior maxillipeds (figs. 23, 24) small, stout, and indistinctly jointed, the marginal curved teeth of the anterior pair very robust; rami of the swimming-feet three-jointed (figs. 25, 26), fifth pair obsolete, represented only by three short plumose setee (fig. 27). Colour of the posterior part of the body faint reddish-brown, the anterior part nearly colourless (in spirit specimens). Length 0°85 mm. ATTHEYELLA NATALIS, n. sp. (Plate VII. figs. 28-33.) Female. In general outline like Canthocamptus (fig. 28). First thoracic segment coalescent with the head; _ posterior abdominal segments spinulose on the hinder margin ; furca very short; principal tail-setee equal to at least half the length of the body. Antennules slender, nine-jointed (fig. 29), shorter than the first cephalothoracic segment, and very sparingly setiferous ; antenne (fig. 30) bearing a minute one-jointed secondary branch. First pair of swimming-feet (fig. 31) short, both branches three- jointed, outer about equal in length to the first two joints of the inner branch ; second, third, and fourth pairs having the outer branch composed of three, the inner of one joint (fig. 32); fifth pair (fig. 33) foliaceous, two-jointed, basal joint wide and bearing four long set on its inner and one on the outer lobe, terminal joint ovate and bearing three unequal apical sete. Length 0-75 mm. 1904. | ENTOMOSTRACA FROM NATAL. 125 CYPRIA CASTANEA, n. sp. (Plate VII. figs. 40-42; Plate VIII. fig. 43.) 2 Cypria lacustris Lilljeborg. Shell, seen from the side, subovate (almost semicircular), highest in the middle, height equal to nearly two-thirds of the length (fig. 40); extremities broadly rounded, the anterior bordered by a narrow translucent fillet, dorsal margin boldly and evenly arched, ventral nearly straight: seen from above (fig. 41) the outline is narrowly ovate, widest behind the middle, width equal to more than one-third of the length, tapered to an acute point in front, rounded behind. Surface perfectly smooth and polished; colour reddish-brown, without any trace of hairs or sculpture. Length 0-65 mm. Limbs closely resembling those of C. ophthalmica, but much more slender, both as regards bulk and armature (figs. 42, 43.) This seems to me to approach most nearly to C. lacustris Lilljeborg, if indeed that be truly a distinct species, and there is not much to separate it from the common Northern species C. ophthalmica, except the absence of shell-markings and the much more slender build of the limbs, of the post-abdominal claws, and of the antennal sete. The specific name now proposed must, however, be looked upon as merely provisional. CyYPRIS INERMIS, n. sp. (Plate VIII. figs. 44-49.) Shell, seen from the side, oblong-ovate (fig. 44), rather higher in front than behind, height equal to less than one-half the length, dorsal margin gently arched, ventral shightly sinuated ; anterior extremity evenly rounded, posterior very much narrower and rounded ; extremities and ventral margin fringed with fine hairs : seen from above (fig. 45) the outline is ovate, greatest width about equal to the height and situated near the middle, lateral margins evenly arcuate, extremities evenly rounded, the anterior rather the broader of the two. In young specimens the postero-inferior angle of the right valve bears two sharp teeth (fig. 46). Colour of the shell greyish white. Length 0°95 mm. The antenna (fig. 47) bears a brush of four sete which reach as far as the extremities of the very slender terminal claws, Apex of the second foot bearing a single slender curved claw (fig. 48) and a long seta. Post-abdominal rami (fig. 49) long and slender, finely ciliated on the hinder margin; all the setz slender and crowded together apically, of these two are stouter than the rest and marginally pectinated. This species occurred abundantly in one of Mr. Gibson’s gatherings, less plentifully in another. CypRIS ARATRA, n. sp. (Plate VII. figs. 34-38.) Shell, seen from the side (fig. 34), subovate, highest in the middle, height equal to more than half the length; anterior extremity broadly and rather obliquely rounded, posterior much narrower 126 DR. G. STEWARDSON BRADY ON [May 17, and less distinctly rounded off; dorsal margin boldly arched, sloping more steeply behind than in front, ventral feebly convex, with no trace of sinuation: seen from above (fig. 35) the outline is very broadly ovate, widest in the middle, width equal to nearly three-fourths of the length, extremities obtusely pointed, the left valve larger and overlapping the right both in front and behind. Surface of the shell marked throughout with delicate and closely- set longitudinal furrows and bearing scattered hairs of variable length. Colour dusky greenish grey. Length 1:0 mm. The antenne bear fascicles of setze reaching to the extremity of . the claws, exactly as in the preceding species; second pair of feet with a slender terminal claw (figs. 36, 37), one long seta and several small claw-like processes *; post-abdominal rami (fig. 38) extremely slender, bearing two apical setze and one much smaller lateral seta situated not very far from the apex. The armature of the second pair of feet is here very different from that of a typical Cypris, but this character seems insufficient to give it separate generic rank. STENOCYPRIS PERARMATA, n. sp. (Plate VIII. figs. 50-57.) Shell, seen from the side (fig. 50), elongated, siliquose, greatest height equal to rather more than one-third of the length and situated in the middle; anterior extremity evenly rounded, posterior suddenly tapered, narrow and rounded off below ; dorsal margin arcuate, sloping evenly toward the front, more abruptly and with a slight sinuosity behind, ventral slightly sinuated near the front but having a longer, shallow sinuation in the middle: seen from above (fig. 51) the outline is extremely compressed and elongated, quite four times as long as broad, obtusely pointed in front, acuminate behind. Surface smooth, slightly setiferous at the extremities; structure thin and membranaceous t. Length 1-6 mm. Antepenultimate joint of the antenne (fig. 52) bearing a small fascicle of setze, two of which are longer than the rest, reaching beyond the bases of the terminal claws; mandibles (fig. 53) of the usual form, the biting portion armed with numerous strong teeth ; maxille (fig. 54) having four slender, elongated lobes, two of the claws of the second lobe strongly denticulate (fig. 55); terminal claw of the first foot very long and slender, second foot (fig. 56) bearing a falcate terminal claw and a very long seta; post- abdominal rami (fig. 57) slender, bearing two long, strongly pectinated terminal claws and one very slender seta, and on the dorsal margin towards the apex armed with a series of about nine or ten strong spines with smaller sete in the intervals; genital lobes very similar to those of Herpetocypris and other partheno- genetic Cypridide. * The second foot of Cypris ornata, a rare British species, is very similar to this, and is figured in Plate VII. fig. 39. + The specimens, as I received them, had been for a length of time ina formalin preservative, which may perhaps have led to the disappearance of calcareous structures. 1904.} ENTOMOSTRACA FROM NATAL, 127 This interesting species seems to occupy an intermediate position between Herpetocypris and Stenocypris so far as the antennal sete, at least, are concerned, but from the shell-characters alone one would undoubtedly assign it to Stenocypris. Professor G. O. Sars in his description of S'. chevreusxiti mentions the “ partly denticulated ” spines of the lobes of the first pair of maxille, but does not figure them, and the antennal sete of that species are certainly much more fully developed both as to length and number than in S. perarmata. These points I have been able to confirm from an examination of British specimens of S. chevreuxit, which I have been fortunate enough to find in several localities :—very sparingly in a pond at Pike’s Hill, and in the Hatchet Pond, both near Lyndhurst, and more recently in considerable quantity in ditches near the River Arun at Arundel, Sussex. Figures of the shell, antenna, and post-abdomen of S. chevreuxit are given in Plate VIII. figs. 58-62. MACROTHRIX AFFINIS, n. sp. (Plate VIII. figs. 63-65.) _ Body short, length about one-third greater than the breadth (fig. 63). Head short and obtuse, with a rounded angle to which the antennules are attached; dorsal margin gently arched, obscurely angulated at its junction with the posterior margin which is finely denticulated, ventral margin convex and fringed with rather long hairs: antennules (fig. 64) club-shaped, slightly dilated and truncated distally, crenulated on the outer margin which bears a few cilia, as also does the truncated extremity. Post-abdomen (fig. 65) spinulose along the whole of its posterior border and having a short terminal hooked claw. The antenne have the normal structure, not presenting any special characters, nor does the shell exhibit any definite ornament or sculpture. Length 0°32 mm. This is a very small species, and is not very unlike the Northern M. taticornis, though only about half its size. EXPLANATION OF THE PLATES. Prats VI. Cyclops gibsoni, 2, p. 123. Fig. 1. Seen from above. Fig. 6. Posterior footjaw. 2. Antennule. 7. Foot of first pair. 3. Mandible. 8. Foot of third pair. 4, Maxilla. 9. Foot of fifth pair. 5. Anterior footjaw. 10. Furca. Cyclops pusillus, 2, p. 122. Fig. 11. Seen from above. Fig. 15. Anterior footjaw. 12. Antennule. 16. Posterior footjaw. 13. Antenna. 17. Foot of first pair. 14, Mavxilla. 18. Foot of fifth pair. Ectocyclops rubescens, 2, p. 124. Fig. 19. Seen from above. || Fig. 20. Antennule, Fig. 28. Seen from left side. Fig. 34. Shell seen from left side. i X50 35. Shell seen from above. , ON ENTOMOSTRACA FROM NATAL. [May 17, Prate VII. Ectocyclops rubescens, 9, p. 124. Fig. 21. Antenna. Fig. 25. Foot of first pair. 22. Maxilla. 26. Foot of third pair. 23. Anterior footjaw. 27. Foot of fifth pair. 24. Posterior footjaw. Attheyella natalis, 9, p. 124. Fig. 31. Foot of first pair. 32. Foot of third pair. 33. Foot of fifth pair. 29. Antennule. 30. Antenna. Cypris aratra, p. 125. Fig. 37. The same more highly mag- nified. 38. Post-abdominal ramus. 36. Extremity of second foot. Cypris ornata, p. 126. Fig. 39. Extremity of second foot. Cypria castanea, p. 125. Fig. 40. Outline of shell seen fromright | Fig. 41. Outline of shell seen from above. side. 42. Extremity of second foot. Prats VIII. Cypria castanea, p. 125. Fig. 43. Post-abdominal ramus. Cypris inermis, p. 126. Fig. 44. Shell seen from right side. Fig. 47. Antenna. 45, Shell seen from above. 48. Extremity of second foot. 46. Posterior extremity of young 49. Post-abdominal ramus. shell. Stenocypris perarmata, p. 126. Fig. 50. Shell seen from right side. Fig. 55. Maxilla, spiniferous tooth more 51. Shell seen from above. highly magnified. 52. Antenna. 56. Extremity of second foot. 53. Mandible (chewing-end). 57. Apex of post-abdominal ramus. $4. Maxilla of first pair. . 58. Shell seen from left side. 59. Shell seen from above. 60. Antenna. Stenocypris chevreuxii, p. 127. Fig. 61. Post-abdominal ramus. 62. Group of marginal setz of the same. Macrothrix affinis, p. 127. Fig. 63. Seen from left side. | Fig. 65. Post-abdomen. 64. Antennule. 1904. | DR. A. GUNTHER ON HYBRID PHEASANTS. 129 June 7, 1904. F. Du Canr Gopmay, Esq., D.C.L., F.R.S., Vice-President, in the Chair. The Secretary read the following report on the additions that had been made to the Menagerie in May 1904 :— The number of registered additions to the Society’s Menagerie during the month of May was 276. Of these 102 were acquired by presentation and 97 by purchase, 8 were born in the Gardens, 61 were received on deposit and 8 in exchange. The number of departures during the same period, by death and removals, was 132. Amongst the additions attention may be called to the following specimens, all new to the Collection : 1. Three Andaman Banded Crakes (Rallina canningi), pre- sented by the Government of India on May 2nd. 2. A Yellow-handed Howler (Mycetes beelzebul), received on deposit on May 13th. 3. An Antilopine Kangaroo (Macropus antilopinus), received on deposit on May 14th. 4, A Grey Solitaire (/yiadectes wnicolor), received on deposit on May 30th. The Secretary reported that he had recently paid a visit to Paris to inspect two specimens of the Orang-utan (Samia satyrus) which had been offered for sale to the Society. He stated that they were very fine animals and fully adult, one of them standing over five feet in height and the other not much less, the former being the largest he had ever seen either living or stuffed. Although they were savage he was able to feed them from the hand with boiled rice, which constituted their chief food. Owing to his opinion that they were in indifferent health he did not purchase them ™*, Dr. Giinther, F.R.S., Vice-Pres.Z.8., on behalf of the President, exhibited a series of hybrid Pheasants killed at various times in the coverts at Woburn, where specimens of many distinct species had been turned out into the open. He stated that nothing of a definite nature was known as to their parentage, but proposed the following determinations tentatively :— 1. Reeves’s Pheasant (Phastanus reevesi) x Common Pheasant. 2. Reeves’s Pheasant x Elliot’s Pheasant (PA. ellioti), An old female assuming male plumage. * [I have been informed that both the Orangs died within a few days of my seeing them.—P. C. M.]| Proc. Zoou. Soc.—1904, Vor. IT. No. LX. 9 130 SHORT-HORNED BUFFALOES IN THE ANTWERP GARDENS. [June 7, 3. Common Pheasant x Ambherst’s Pheasant (Thaumalea amherstie). Male. 4, Common Pheasant xX Amherst’s Pheasant. Old female. 5. Cheer Pheasant (Ph. wallichii) x Common Pheasant (? Elliot’s Pheasant). Male. 6. Cheer Pheasant x Common Pheasant. Male. 7. Silver Pheasant (Gennceus nycthemerus) x Elliot’s Pheasant. Male. 8. Lineated Kaleege (@. lineatus) x Common Pheasant. Male. 9. Lineated Kaleege x Japanese Pheasant (Ph. versicolor). Male. 10. Cheer Pheasant x ? Himalayan Monaul (Lophophorus ampeyanus). Male. Dr. F. D. Drewitt, F.Z.S., exhibited two fine antlers of the North-African Red Deer (Cervus elaphus barbarus) and made the following remarks :— “These antlers were obtained in the high forest-land of cork, oak, and cedar, extending far inland on the borders of Tunis and Algeria. “One of the most interesting links between the fauna of North Africa and Europe is this Red Deer living among lions and panthers. “‘ Barbary-Deer antlers differ from typical Red-Deer antlers in having no second tine. This seems a constant characteristic. In other heads seen in Algeria it was absent. It is absent in a specimen from Algeria in the Cambridge Museum; also in two Menagerie specimens in the Natural History Museum (one from the Gardens and one given by the Duke of Bedford). “Few Englishmen have seen a wild Barbary Stag. Sir Harry Johnston is one of the few. He reports that twenty-four years ago it was fairly common throughout the forest. Now, though protected by the French Government, it is rare ; forest fires and poaching Arabs have almost exterminated it—but a few remain. Fortunately a wild stag among trees, facing its enemy, is some- times almost invisible at a few feet, the antlers exactly copying not only the form of a branch but also the bark on it.” Dr. Drewitt also exhibited a pair of horns (154 inches in length) of Loder’s Gazelle (@azella leptoceros) from South Algeria. Some photographs, sent by Dr. Graham Renshaw, F.Z.S., of a pair of Short-horned Buffaloes in the Antwerp Zoological Gardens, were exhibited, and the following note upon them, contributed by Dr. Renshaw, was read :— “The difficulty experienced by naturalists in separating the 1904. | ON THE GROWTH OF THE ANTLERS IN THE WAPITT. 131 geographical species or races of the Short-horned African Buffaloes may render interesting the accompanying photographs, taken by myself, of a pair of Short-horned Buffaloes which have lived for many years in the Antwerp Zoological Gardens. Seen sideways, the horns of both specimens sloped back almost in the plane of the forehead. Seen from the front, the horns of the bull were broad and flattened, without transverse markings, and with but a small development of the boss, which forms a ring-like projection where the horn springs from ‘the skull. The horns bore considerable resemblance to those of the Senegambian Buffalo (Los caffer planiceros), and perhaps the animal is referable to this subspecies. The horns of the cow, however, were more curved at the tips, and recalled those of the Lake Tchad Buffalo (Los caffer brachyceros). Both animals were in the prime of life, of a deep dark blackish- brown colour; the edges of the ears were heavily fringed with hair, but there was no hair inside the pinna, A very sparse line of hair (the individual hairs being about 4 inches long) extended from the occiput to the root of the tail. The tail was very dark brown; tail-tuft black. he dorsal line of hair was hardly noticeable in the cow. The animals were labelled ‘ Bubalus pumilus, Afr. Australe et Centrale (Buffle nain du Sénégal)’ However, they do not agree at all with Los caffer nanus (Bush- Cow of the West Coast sportsmen).” Mr. F. E. Beddard, F.R.S., exhibited and made remarks upon the skull of the Cape Crowned Crane, Dalearica chrysopelargus, showing paired lateral and single median bony bosses suggestive of those of horn-bearing vertebrates. Mr. R. E. Holding exhibited and made remarks upon a series of 12 platinotype photographs, three of which are here repro- duced (text-fig. 24, p. 132), taken by Mr, Henry Irving, of Horley, showing the growth and development of the horns in the Wapiti Stag (Cervus canadensis) bred by the Society and now living in the Gardens. The first of the series was taken on March 20th, 1903, the date on which the horns were shed, and photographs were taken at fortnightly intervals until the following September, when the horns were fully adult and almost free from the velvet covering. The photographs showed in an interesting manner the rapid growth of the antlers in the Cervide, the Pe occupied being six months. On March 20th, 1904, these horns were in the usual course shed, the weight of the pair being 21 lbs. and length of beam 453 inches. Simultaneously with the development of the horns, the photo- graphs showed the rapid seasonal change in the coat of the animal, the winter coat beginning to fall off shortly after the 132 ON THE GROWTH OF THE ANTLERS IN THE WAPITI. [June7, Text-fig. 24. < Sin? Antlers of Wapiti, from photographs. (1904. ] ON THE YOUNG OF THE EGYPTIAN FAT-TAILED GERBILLE. 133 horns were shed. In five weeks the process was complete and the animal was in its summer coat. Explanation of Text-fig. 24. A. Head of Stag, taken April 17th, 1903, three weeks and four days after the horns were shed. Length of horns about 5 inches. Extremely vascular and sensitive, the skin of the forehead merging into the “velvety ’’ covering; the burr (e) being indistinct, the bifurcation between frontal tine (a) and bez tine (6) being just apparent. B. Head of same Stag, taken May 14th, viz. three weeks and four days afterwards. Here a rapid development is seen to have taken place; the frontal tine (a), bez tine (6), and beam (ce) being well divided, a small supernumerary point (x ) making its appearance at the base of the left horn. Horns still vascular and warm, and covered with thick epidermis or “ velvet ”’—the burr (e) being more definite. C. The same head, taken June 25th, five weeks and four days afterwards. Within this period the most marked development takes place; the frontal (a), bez (6), and trez (c) tines being fully developed, and the apex of the beam (d) splitting up into the characteristic three sur-royals. ‘The velvety covermg is now shrinking, and the nutrient blood-vessels can be discerned; the burr (e) is now quite prominent. On July 24th the horns, though still covered with velvet, are quite adult. During August the velvet begins to peel off, and by the first week in September the horns are usually clear. Mr. Holding also exhibited a fine pair of antlers of the Irish Red Deer (Cervus elaphus) having 14 points, which had been sent to him by Sir Douglas Brooke, Bart., F.Z.8., Co. Fermanagh. They weighed 8 lbs. and were 37 inches in length. Mr. R. I. Pocock, the Superintendent of the Gardens, exhibited living specimens of hairless varieties of the Common House-Mouse (Mus musculus) and Brown Rat (JZ. decwumanus), and remarked that the available evidence seemed to prove that the abnormality was always correlated with a wrinkled skin. The living specimens exhibited further suggested that it was accompanied by weakness of the eyes, and in the case of the Mouse by the absence, or at all events diminution, of the smell so characteristic of that animal. Mr. Pocock also exhibited four young examples, between five and six weeks old, of the Egyptian Fat-tailed Gerbille (Pachy- uromys duprest), which were born in the Menagerie and reared by hand, owing to the death of the mother when the young were only six days old. One of the young had never opened its eyes and would probably be permanently blind. There were five young in the brood, but one had died. Mr. Pocock also showed a 154 LT.-COL. J. MALCOLM FAWCETT ON [June 7, newly-born example of this species, which was naked and blind and scarcely distinguishable from the young of the Common Rat. In this respect the young differed markedly from those of the Egyptian Spiny Mouse (Acomys), which were born with the eyes open and the skin covered with hair, the brood consisting of but two individuals only. This fact further supported the view that the condition of the young at birth could not be taken as a criterion of attinity in the case of the Rodentia. The following papers were read :— 1. On some New and Little-known Butterflies, mainly from high elevations in the N.E. Himalayas. By Lt.-Col. J. Matcomm Fawcnrtt*. [Received April 9, 1904. } (Plate [X.7) [The complete account of the new species described in this communication appears here, but as the names and preliminary diagnoses were published in the ‘ Abstract,’ such species are distinguished by the name being underlined.—E 170k. | The main object of this paper is to describe some species which were taken by the native collectors of Mr. J. C. White, British Commissioner in Sikkim, at high elevations.in the N.E. Himalayas, on the borders of Thibet. I have been unable to identify them with any species in the Collection of the British Museum, or in the Rothschild Collection at Tring, and they were subsequently sent to Mr. H. J. Elwes, who is unacquainted with them. Under these circumstances it appears to me necessary that they should receive names, although future investigation may perhaps demonstrate some of them to be geographical forms of species already described. It also appears essential that the three species of Colias which have been taken at high elevations on the Sikkim-Thibet Frontier should be figured together; and for this reason a figure and description of C. eogene var. leecha (Gr.-Gr.), of wich a figure does not appear to have been hitherto published, have been added to the figures and descriptions of the two new species. At the same time that the above-mentioned species were taken, there were also four species of the genus Parnassius captured on the Sikkim-Thibet Frontier: one is clearly P. imperator (described by Herr Friihstorfer as P. imperator augustus); a second is P. epaphus var. sikkimensis Elwes ; and the other two have been * Communicated by the SecRETARY. + For explanation of the Plate see p. 141. ees 1904 oleae £.C Knight del.et Lth. Mintern Bros imp BUTTE Re Lib S Ove Dae Nhe eiIMALAVAS. 1904. ] BUTTERFLIES FROM THE N.E. HIMALAYAS. 135 identified by Mr. Elwes as varieties of P. acco and P. delphius respectively ; the two latter were described as P. acconus $ and Q by Herr Friihstorfer from the writer’s figures. A species of Ragadia, from the Battak Mts. in Sumatra, has been added to the paper, and is here figured for the first time. Family SAryrina. 1. RaGADIA sIMPLEX Fawcett. (Plate IX. fig. 1.) Ragadia simplex Fawcett, Ann. & Mag. Nat. Hist. ser. 6, vol. xx. p. 111 (1897). Eixpanse 1 in. 6 lin. Habitat. Battak Mountains, Sumatra, June 1896. Description.— Vale. Upper side pale fuliginous brown, crossed by two darker fuscous bands running parallel to one another from the costa of the fore wing to the inner margin of the hind wing, the outer being much broadened on the hind wing. There is also a shorter band proximal to the body, which, starting parallel to the others on the fore-wing costa, ends at the origin of the sub- median on the fore wing. From the apex of the fore wing to the inner angle of the hind wing runs a submarginal row of small, indistinct, fuscous spots, one in each interspace. The wing- margins are narrowly fuscous, the colour being bounded by a narrow submarginal band looped on the fore wing and on the hind wing parallel to the margin. Under side similar in markings to the upper, but the ground-colour pale buff and the row of spots silver instead of fuscous. The antenne, head, thorax, abdomen, and legs like those of f. crisia Hiibner, from which this species differs in its generally pale ground-colour, and in the ocellate band of erisia being reduced to mere pupils of silver below and fuscous above. Family NyMPHALINA. 2. MELITHA TIBETANA. (Plate IX. fig. 2.) Melitea tibetana Fawcett, Abstr. P. Z.S. 1904, No. 9, p. 8, June 14. Expanse 1 in. 2 lin. Habitat. Khamba Jong, Thibet, 15,000 ft. elevation. Description.— Upper side: both wings pale dull ferruginous, apex of fore wing pale ochreous; costa and base of wings and anal margin of hind wing dark fuscous. Fore wing with two black marks in cell and one at its apex; base of cell inside the first mark, and between the second and third marks, pale ochreous ; two transverse series of black discal spots, the inner series highly angled outwardly above the third median nervule, enclosing between them a pale ochreous fascia; a marginal series of pale ochreous lunules inwardly defined with fuscous and 136 LT.-COL. J. MALCOLM FAWCETT ON _ [June 7, outwardly by a fuscous marginal line, and a pale ochreous lunule basally between the submedian nervure and first median nervule; cilia whitish. Hind wing with the marginal fuscous line and submarginal lunules as in fore wing; the transverse series of black discal spots and ochreous fascia as in fore wing, and two black marks in discoidal cell. Under side: fore wing very pale ferru- ginous, fading into sordid white at the apex and along the costal margin ; discal markings of upper side obsolescent except near the apex. Hind wing buff, with a subbasal band, a spot in the cell, an angulated discal band, and a lunulated submarginal band all white, defined outwardly by black atoms and fine lines, the cell and subbasal band surrounded by ferruginous spots. This species is probably a local race of the variety of W/. sindura Moore described by Mr. H. J. Elwes in Trans. Ent. Soc. Lond. 1888, pt. i. p. 336, pl. x. figs. 5 & 6, which = MW. sikkimensis Moore, but differs from it in its much paler coloration, and the presente on the upper side of the pale ochreous bands and spots above mentioned, and the more prominent fuscous markings. Five specimens from Khamba Jong in my possession present no variation from the above description. J/. sindura var. Elwes, isan uniformly bright ferruginous insect with no pale fascize or spots; and I possess six specimens of it from 17,000 ft. in Native Sikkim, as also six specimens of a larger and darker form from 10,000 ft. in Sikkim interior, in which the submarginal lunules are obsolescent. 3. ARGYNNIS CLAUDIA. (Plate IX. fig. 3.) Argynnis claudia Fawcett, Abstr. P. Z. 8. 1904, No. 9, p. 8, June 14, Expanse 1 in. 6 lin. Habitat. Khamba Jong, Thibet, 15,000 ft. elevation. Drscription.— Upper side rich fulvous (pale fulvous in some specimens), the base of fore wing narrowly and that of hind wing broadly irrorated with black atoms. Fore wing with three curved streaks in the cell, and one at its apex, thicker than the remainder ; two transverse discal series of black spots, the imner one very irregular and highly angled outwardly above the third median nervule; a series of submarginal fulvous cuneiform lunules, defined inwardiy by a thick black line and outwardly by a thick black band; cilia white. Hind wing with the submarginal lunules and transverse series of spots as in the fore wing, except that the outer series is reduced to four spots (2 subapical and 2 subanal) in some specimens, and to three in others, the remaining spots being obsolete. Under side: fore wing similar to upper side but paler, the black marks and spots smaller, and the outer transverse series of spots obsolete ; apex green; marginal series of lunules silvery and two silver subapical spots below the costa; no marginal lines. Hind wing dark green, with a slight bronzy suffusion ; two basal silver spots below the costa, a linear silver 1904. ] BUTTERFLIES FROM THE N.E. HIMALAYAS. 137 spot defining the apex of the cell and a linear one near the inner margin ; the space between the two transverse discal rows of black spots (of the upper side) occupied by a series of linear silver streaks between the nervules; a submarginal series of triangular silver spots, much produced ‘inwardly, and between these two series of silver spots and streaks a faint indication of a submarginal line of yellow spots ; no marginal lines. This species is nearly allied to A. clara Blanchard, but differs from it in its rounder wings, smaller size, and paler pallone On upper side, outer mar: gins of both wings broadly fuscous, contain- ing a submarginal series of pale spots, without the submarginal and mar ginal fine black lines beyond them as in A. clara; the inner transverse series of discal spots angled outwardly above 3rd median nervule on both wings, and the outer discal series of black spots is reduced, on the hind wing, to 4 and sometimes 3 spots, the spot on the lower discoidal interspace being always absent. Under side: the fulvous colour of fore wing is paler and redder, and the green colour of apex of fore wing and hind wing is duller and darker ; bronzy suffusion less, and extends to the margin without the mar ginal fulvous border of 4. clara. The only variation from the above deser iption in five specimens from Khamba Jong is that one specimen is more melanic, while another hasa paler gr Gandleoloue | ‘The aeny, a, var. clarina of Staudinger (Cat. Lep. Pal. p. 38) cannot be referred to this species, as he says nothing about his species being smaller than A. clara, and describes it as having the black spots smaller, whereas this species is more melanic than A. clara. Family Lycanipm. 4. Lycmna ArtanA Moore, var. ARENE. (Plate IX. fig. 4.) Lycena ariana Moore, var, arene Fawcett, Abstr. P. Z.8. 1904, No. 9, p. 8, June 14. Expanse | in. 3 lin. Habitat. Khamba Jong, Thibet, 15,000 ft. elevation, July. Description.— Male. Upper side: both wings bright cyaneous blue, margins narrowly black, extending slightly up the nervules ; cilia broadly white. Under side: both wings purple-grey ; hind wing suffused with metallic greenish grey at base and on inner margin. Fore wing: apex of cell defined by a narrow black streak outwardly surrounded by white; a discal series of five minute black spots ringed with white, the subcostal spot evanescent; a marginal series of white spots with minute black centres, defined inwardly by obsolescent traces of orange lunules. Hind wing with the marginal white spots larger, and the orange lunules defining them inwardly more distinct, that at anal angle geminate ; a subcostal black spot, and a discal series of minute black spots ringed with white, varymg in number from two to three, and entirely obsolete in some specimens (this is the form 138 LT.-COL, J. MALCOLM FAWCETT ON [June 7, figured) ; a thin black streak at apex of cell, in a large cuneiform white spot, and opposite this spot another large cuneiform white spot, with its base on the centre of the line of orange submarginal lunules, This species differs from Z. ariana Moore from the N.W. Himalayas in being smaller, and in the discal row of black spots on the under side being minute, fewer in number, and obsolete, in some specimens, in the hind wing; and in the discal white spots at the apex of cell of hind wing being larger. . Lycmna puEeretes Hiibner, var. praris. (Plate IX. figs.5 g, 5a@.) Lycena pheretes Hiibner, var. pharis Fawcett, Abstr. P.Z.S8. 1904, No. 9, p. 8, June 14. Expanse 1 in. Habitat. Khamba Jong, Thibet, 15,000 ft. elevation, July. Descriprion.— Vale. Upper side dark purple-blue; fore wing with a black marginal line; margin of hind wing more broadly black, especially at apex and costa; cilia broadly white. Under side: fore wing purple-grey, paling towards apex; a black spot ringed with white at apex of cell; a discal series of small black spots ringed with white, varying in number ; in some specimens only three, usually four. Hind wing pale brownish on the disk, fading into pale greenish-ochreous at the margin, with the follow- ing pale ochreous spots: one in and extending beyond discoidal cell, cuneiform; a discal series of six beyond it, and an indistinct one basally below the costa. Female. Upper side dark fuscous, basally irrorated with blue ; under side as in male. This form differs from the var. asiatica of Elwes in having fewer discal black spots on fore wing, and in the size and length of the pale cellular spot on hind wing. 6. Lycmna (Zrzpra) zERA. (Plate IX. fig. 6.) Lycena (Zizera) zera Fawcett, Abstr. P. Z.8. 1904, No.9, p.9, June 14. Expanse | in. 11 ln. Habitat. Tounghoo, Burma, June. Descrietion.—Male. Upper side shining violet ; fore wing with a pale spot at the apex of the discoidal cell; a broad fuseous marginal band extending slightly up the nervules, but the inner border regular and bending inwards on the costa. Hind wing with the marginal fuscous band broad at the apex, narrowing to a black marginal line, inside which are placed five fuscous lunules ; cilia broadly white. Under side pale grey: fore wing with a fuscous streak, ringed with white, defining apex of cell; a discal row of six fuscous spots ringed with white; a row of marginal spots faintly defined imwardly by fuscous Junules. Hind wing: discal row and a marginal series of fuscous spots as in fore wing ; 1904. | BUTTERFLIES FROM THE N.E. HIMALAYAS. 139 three subbasal black spots ringed with white, the centre spot inside the cell, This species is nearly allied to 7. marginata from China, but differs in the discal row of fuscous spots of the fore wing being fainter and more regulary on the under side, and the black border of the upper side being more even in its inner edge and bending inwards on the costa, 7. Lycana (Nipuanpa) MAncta, (Plate IX. fig. 7.) Lycona (Niphanda) marcia Bawceth, Abstr. P.Z.8. 1904, No. 9, p. 9, June 14. Expanse 1 in, 2 lin. Habitat. Vounghoo, Burma, June. Descriprion.— Male. Upper side shining violet, with the spots of the underside showing up indistinctly. Fore wing with a narrow fuscous band on the costa and outer margin; hind wing : costa, apex, and abdominal margin broadly fuscous, and a row of fuscous lunules on the outer margin ; cilia white between the ends of the nervules. Under side pale violaceous grey, with the follow- ing brown marks ringed with white:—fore wing with a short basal streak below the costal nevvure; two spots, one above the other, and the lower one the larger, in centre of cell, and below the lower spot a large triangular space between the first median nervule and the sabmiedian nervure; a spot defining the end of the cell; a discal fascia of seven spots, interrupted in the middle, and with the outer edge of the upper four spots defined outwardly by fuscous; a pale and indistinct submarginal and marginal lunular fascia, the lunules of which are largest and most defined between the median nervules, Hind wing with a blackish-brown spot at base of cell, two black spots below costal margin, and a similar spot below the outer one; an irregular row of pale discal spots, and a marginal row of white lunules, the first three and the sixth from the anal angle with black centres. This species is probably a local race of L. (Niphanda) tessellata Moore, from Penang, but differs from it in its much less heavy markings on the under side. There is also a specimen (at present unnamed) in the British Museum which agrees with the specimen here described, Family Prerina. 8. CoLiAs BeRYLLA. (Plate IX. fig. 8.) Colias berylla Fawcett, Abstr. P. Z.8. 1904, No. 9, p. 9, June 14, Expanse 2 in. 9 lin, Habitat. Khamba Jong, Thibet, 15,000 ft. elevation. Descriprion.—Lemale, Upper side : fore wing pale lemon- yellow, base and costal margin broadly irrorated with black atoms; a diamond-shaped black spot at apex of cell; hind or 140 LT.-COL. J. MALCOLM FAWCETT ON [June 7, outer margin broadly black, irregularly dentate on nervules on its inner edge, with the following sulphur-yellow spots, viz., five forming a curved row near the apex from costa to 3rd median nervule, the lower two being larger than the upper three; also one spot between the Ist and 2nd median nervules, and one between the Ist median nervule and the submedian nervure; cilia of both wings whitish with a pink suffusion. Hind wing black, basally thickly irrorated with sulphur atoms; an orange spot at end of cell; anal and outer margin sulphur-yellow, which forms into a line of lunular spots on outer margin. Body black, head and antenne pink. Under side: fore wing —disk pale yellow ; apex and outer margin grass-green, with a submarginal line of pale yellow spots; a black spot at apex of cell, and three black discal spots in a row between the median nervules and submedian nervure. Hind wing grass-green, paler on outer margin, and irrorated with black atoms; a silver spot surrounded by a pink area at the apex of the cell. This species belongs to the Ayale section of the genus Colias, and is nearest to C. erschoffi Alphéraky, from the Tian Shan Mts. 9. Contas NINA. (Plate IX. fig. 9.) Coltas nina Fawcett, Abstr. P. Z. 8. 1904, no. 9, p.9, June 14. Expanse 2 in. 9 lin. Habitat. Khamba Jong, Thibet, 15,000 ft. elevation. Descriprion.—lemale. Upper side: fore wing pale sulphur- yellow, the disc below the median nervure and 3rd median nervule orange; a black diamond- shaped spot at apex of cell; a broad black outer marginal band, inwardly irregularly dentate on the nervules, irrorated with sulphur- yellow scales outwar dly, and a submar. einal row of ill-defined sulphur spots between the nervules, that between the 2nd and 3rd median nervules being obsolete. Hind wing black, irrorated with pale sulphur atoms, most thickly towards the base and outer mar gin; discoidal cell pale yellow, with an orange spot at its apex; a submarginal row of pale sulphur hastate spots; antenne and cilia pinkish. Under side: fore wing—disk as on upper side; apex and outer margin pale green, irrorated with black atoms; a black spot at apex of cell, and three or four ill-defined black spots discally between the nervules. Hind wing pale green, with spots as on upper side. This species appears to be a transition between the hyale and edusa sections of Colias. 10. Conzas EocENE Felder, var. LEEcHI Gr.-Gr. (Plate IX. figs, 10 g¢, 10a 2.) C. eogene Felder, Reise Novara, p. 196, t. 27. £. 7, ¢ (1865). Var. leechi Grum- Grshimailo, Hore Soc. Ent. Ross. xxvi. p. 382 (1893). Expanse | in. 4 lin. Habitat. Native Sikkim, 19,000 ft., July. 1904. | BUTTERFLIES FROM THE N.E, HIMALAYAS. 14] Descrrprion.— Male. Upper side orange, base narrowly irrorated with black atoms; a black diamond-shaped spot at apex of cell; a fuscous marginal band, with a row of pale yellow hastate spots between the nervules. ‘Hind wing orange ; inner margin green, irrorated with black atoms ; a paler spot in cell, anda submarginal row of pale yellow hastate spots; apex and outer margin fuscous. Under side : fore wing—disk paler than upper side; apex and costa grass-green, lrrorated with black atoms; a black spot at apex of cell, and a submarginal row of black spots, ill-defined towards the apex, and beyond them a row of pale yellow spots. Hind wing grass-green ; a pale sulphur submarginal fascia, irrorated with dark green atoms; a silver spot in a reddish area at end of cell. Head, antennze, and cilia pink. female. Upper side: fore wing orange, paling on the costa to yellow; base and median nervules linaeadl with black atoms ; marginal band and spots as in male. Hind wing black; cell yellow, with an orange spot at its apex; a marginal band of sordid-sulphur lunulate spots. Head, antenne, and cilia as in male. The specimens in my possession have been identified by Mr. H. J. Elwes as var. leechi. Herr Frithstorfer erroneously described two of my figures as representing a new species, under the name of C. eogene, subsp. miranda (Insekten Borse, xx. 19. 148, May 1903). These figures are introduced into this paper because this variety does not appear to have been figured previously, and with a view to complete the series of forms of Colias from the borders of Sikkim and Thibet. Since these notes were written I have seen a paper published by Mr. H. J. Elwes in ‘Iris,’ 1904 (pp. 388 e¢ segq.), on the subject of the descriptions and figures above referred to. In this paper Mr. Elwes refers to this species of Colias as “‘stoliczkana,” although he had previously written “ leechi” below the figures I submitted to him for identification. The male of this species, however, differs from the figure of the male of C. stoliczkana in the British Museum. The var. stoliczkana has hitherto been recorded only from Ladak, Fergana, and Amdo; while the var. leecht has been recorded only from China. EXPLANATION OF PLATE IX. Fig.1. Ragadia simplex Fawcett, p. 135. 2. Melitea tibetana Fawcett, p. 135. 3. Argynnis claudia Fawcett, p. 136. 4. Lycena ariana Moore, var. arene Fawcett, p. 137. 5. Lycena pheretes Hiibner, var. pharis Paw cett, 6, p. 138. 5a. Ditto, é 6. Lycena (Zizera) zera Fawcett, p. 138. 7. Lycena (Niphanda) marcia Fawcett, p. 139. 8. Colias berylla Fawcett, p. 139. 9. Colias nina Fawcett, p. 140. 10. Colias eogene Felder, var. leechi Gr.-Gr., 3, p. 140. 10a. Ditto, 2. 142 DR. A. G. BUTLER ON SEASONAL [June 7, 2. On Seasonal Phases in Butterflies. By Ay G.Burunr, Pho Dy WS: eZ Sues [Received May 10, 1904.] The fact that Butterflies emerged from the pupa in markedly different forms at different times of the year was made evident many years since by the labours of W. H. Edwards in the United States, by Mansel Weale in South Africa, and by Doherty and de Nicéville in India; but this fact was misunderstood, and there- fore not fully accepted by many workers for years afterwards, whilst not a few are sceptical as to its truth at the present day. One of the chief reasons for this scepticism is based upon the unquestionable truth that the dry-season type of a species not unfrequently emerges in the wet season and vice versa. That I was myself rendered sceptical for years on this ground will be seen by referring to some of my earlier papers in which the question of seasonal forms had to be considered. In a paper published in 1884 (P. Z. 8. pp. 478-501) I recorded the remarkable fact that, at Aden, Limnas chrysippus, Hypolimnas misippus, and Catopsilia florella produced all their varietal phases simultaneously, and that Zeracolus “nouna” = saxeus (which actually is the 7’. evagore of Klug) occurred in March, April, and May, whilst 7. yerburit was also obtained commonly in April and May; but I did not then fully appreciate the fact that all these were instances of the simultaneous emergence of phases characteristic of seasons and climates, and that they represented the probable condition of all very variable types before seasonal or climatic changes had begun to act upon them. As with protective mimicry, the more enthusiastic exponents of which have frequently erred in supposing that because this was of use against one enemy, it must necessarily be against all; so has it been with those who desired to believe in, but failed to comprehend, seasonal variation. That I misunderstood it myself in 1886 is clear from the remarks which I made in a paper upon Lepidoptera from Western India (P. Z. 8. p. 399) respecting the broods of Belenois mesentina: in 1888 I was no wiser, as my remarks emphasising the importance of dates of capture in the case of certain species of Zeracolus clearly show (Ann. & Mag. Nat. Hist. ser. 6, vol. i. p. 201). In 1895 (P. Z. 8. p. 727) I hinted at the possibility of Zypan- artia scheneia and H. hippomene being seasonal forms of the same species, and in 1896 (P. Z.8. p. 112) I considered this probable; yet later in that year (P. Z.8. p. 285) I concluded that this was an error, because both were captured on the same mountain upon two successive days. At this date, therefore, it is quite evident that I considered it impossible for wet and dry phases of any species to occur simultaneously. Indeed, it was only after reflecting upon the probable identity of 7. yerburw with the supposed 7’. nowna (P. Z. 8. 1896, p. 247) that I began 1904. ] PHASES IN BUTTERFLIES. 143 to have some idea that because a type of pattern and coloration was characteristic of a particular season or climate, it did not necessarily exclude other types: therefore that it was not im- possible for phases characteristic both of dry and wet seasons or climates to be sometimes found flying together; that in a very dry country like Aden it was the rule rather than the exception for wet, intermediate, and dry phases of a species to occur commonly together in each brood. That this polymorphic character was probably of earlier date than the more or less defined seasonal phases, of such countries as exhibit great variations of weather at different seasons, seemed evident to me from the fact that in very moist countries the extreme dry phase of species is exceedingly rare, and probably near to extinction. In Precis sesamus, the dry phase of P. natal- ensis (=calescens) from Southern and Hastern Africa, the seasonal phases are very distinct, but about equally abundant. In the wet season, as pointed out by me (P. Z.8. 1898, p. 904), both forms may be taken flying together in Mashunaland; and on that ground I proposed that the term “seasonal form” should be rejected, and the term ‘seasonal phase” substituted *. On the West Coast P. calescens or natalensis is represented by P. octavia of Cramer and a number of intermediate phases, but no extreme dry phase was recorded until 1901, when, im my “ Revision of the Butterflies of the genus Precis” (Ann. & Mag. Nat. Hist. ser. 7, vol. vill. p. 205), I mentioned an imperfect example indis- tinguishable from typical P. sesamus as having been received from Onitsha on the Niger. The extreme rarity of this phase on the West Coast, and the probability that it has become absolutely extinct at Sierra Leone, seem to indicate that it is unsuited to the conditions of a moist climate; whilst the numerous intergrades from the dry to the wet phase on the same coast certainly indicate the transition from fixed varieties, such as obtain where seasons are well defined, towards a more or less wet type. In Southern and Hastern Africa intergrades between P. sesamus and P. natal- ensis are extremely rare, the most striking of such intergrades being figured by me in 1900 (P. Z. 8. pl. lviii. fig. 1). As it is by no means rare for individuals of the wet phase of a species to emerge from the pupa in the dry season, there is no reason why Lepidopterists should be startled when this occurs. They should bear in mind the probability of the fact that, as all the phases of some species occur as simple varieties in extremely dry countries, they also formerly existed as varieties in other species; that the latter, as they gradually extended their range, were subjected to widely different conditions; that then the summer phase (as we now understand it) was so conspicuous in the winter, and the winter phase so conspicuous in the summer, that their chance of survival at the unsuitable season was lessened ; and thus it came about in course of time that one variety of the species became the prevalent wet phase, and * See also P. Z. S. 1900, p. 916. 144 CAPT. R, CRAWSHAY ON THE PREY or THE Lion. [June7, another the prevalent dry phase. But it is absurd to assume that the evolution of these seasonal phases is already perfect and complete over the whole globe, and to speak of the occurrence of wet and dry phases at any season as “ very remarkable”; as a matter of fact, it would be more remarkable if they never did occur simultaneously. 3. The Prey of the Lion. By Capt. Ricoarp Crawsuay, F.Z.S. [Received June 3, 1904. ] Prevailing opinion is so often at variance with my own ex- perience on the question of the prey of the Lion, that I venture to offer a few remarks on those experiences covering a period of many years’ travel and residence in Central Africa. It is generally believed that the Lion is a fastidious feeder, eating only what he has himself killed in the shape of the larger mammals. I have not found the Lion by any means particular in confining himself to his own kills, or in disposing of carrion when in tolerably good condition. As to his prey being only the larger mammals, this is not in- variably the case, for Iam able to establish one instance at least of his preying on the Porcupine. Mr. Selous has it on record that a Lion has eaten the skin of a Sable Antelope preserved with arsenical soap as a natural history specimen. Nothing as re- markable as this has occurred within my own knowledge, but I have known a great many instances of Lions feeding on carrion of my own killing, such as the carcases of Elephants on several occasions. The main purpose of my note is to exhibit two Porcupine- quills taken from the left fore-paw of a Lion—the skull, skin, and claws of which I also exhibit. This Lion was shot by myself two days’ journey N.W. of Kibwezi, Hast Africa Protectorate, in March 1898. Three quills were found in the paw, one of which I regret has since been lost. The two exhibited measure each approximately 1 inch in length. The cartilage in which the quills were embedded showed no inflammation, so it can be inferred that these at first most painful inflictions must have been there for a considerable time. There is no reason for supposing this Lion killed the Poreupine under the impulse of hunger, because the country teemed with game of all kinds—from rhinoceros, zebra, and ostriches, to vast numbers of antelope, large and small. My Ahenga, when I questioned them as to whether they knew other instances of Lions preying on such small game as Poreupines, seemed amused, and replied :—‘ Oh, indeed! even field rats: they tear them out of the ground with their claws”! The larger mammals, no doubt, form the Lion’s chief prey— Buffalo more especially, when these abounded; but, as can be seen, there are times when even small rodents are in requisition. 1904. | ON THE ‘* BREPHOS” IN A SKINK,. 145 4. Note on an apparently Abnormal Position of the “ Brephos”’ within the Body of a Skink (Chaledes lineatus). By Frank H. Bepparp, M.A., F.R.S., Prosector 40 the Society. [Received May 11, 1904. | (Text-figure 25.) Among a number of examples of the Lizard Chalcides lineatus which arrived in the Society's Menagerie on May 3rd, one was found to be dead on its arrival, On opening the Lizard, which proved to be a gravid female, a possible cause of death was at once obvious. The reptile presented a case of what may be called extra-uterine pregnancy, and may have been unable to get rid of the two young ones. That is, at any rate, one way—and perhaps the most probable way—of interpreting the following facts. * The Lizard itself measured 12 inches in length, including the tail, and the young ones were of the same size, and measured within a very few millimetres of 35 inches*. I did not attempt a more exact measurement, as they were somewhat twisted. The yolk-sac was not fully absorbed, and formed a string-like appendix of about 3 inch in length, In other respects these foetuses were exactly like the parent and obviously, at the very least, ready to be born. Neither brephos was within the oviduct. Both lay, in fact, apparently in the body-cavity. The position of one of them is shown in the accompanying drawing (text-fig. 25, p. 146), which was made before the relations of surrounding organs were much disturbed. Its head les not far from the head of the parent ; the end of the tail is nearly on a level with the posterior extremity of the liver. The second brephos lay much further back, but I cannot give exact details, as 1t was liberated on cutting open the body-wall of the parent. Both young ones lay in a slight spiral curve like the popular representations of asnake. There are some other facts of interest in connection with the abnormal situation of the two young Chalcides. The ovaries contained numerous round white eggs, the largest of which did not measure more than 2 min. in diameter. I presume, therefore, that they are not nearly ready for fertilisation. The two oviducts were in a condition corre- sponding to that of the two ovaries. They showed no signs what- ever of having recently contained eges or embryos. Nor, on the other hand, was there anything in the least abnormal about them or their drawn-out peritoneal orifices. They were quite un injured by my dissections. It seems to me, therefore, to be indisputable that at least a great part of the development of the two embryos must have taken place outside of the oviduct or oviducts. And it is remarkable that this took so long in producing an injury grave enough to be fatal to the mother. Indeed, there were no signs whatever of pathological conditions * The Scincidie are, at least for the most part, ovoviviparous. Proc. Zoot. Soc.—1904, Vou. II. No. X. 10 146 ON THE ‘ BREPHOS” IN A SKINK. [June 7, in the organs neighbouring upon the two young ones ; and one 1s thereby tempted to speculate as to how far this—as would be supposed by most—abnormal state of affairs is really abnormal. Such cases of extra oviducal feetation may be in part responsible Dissection of Chalcides lineatus. A, brephos; Z., lung; Zi., liver; St, stomach; Od., oviduct. for some of the legends concerning the swallowing of their young by various reptiles for protection’s sake. In the present instance, for example, as may be seen from the drawing exhibited, an unskilled observer opening the body of the parent and seeing the 1904. | ON THE VISCERAL ANATOMY OF PELAGIC SERPENTS. 147 head of the brephos lying apparently within the gullet, and at any rate anteriorly in the body, might arrive at a conclusion opposed to the real facts which more accurate observation reveals. One rather important piece of evidence is commonly omitted in those cases of alleged swallowing of the young, which I refer to here for other reasons, and only incidentally as concerned with popular beliefs. Each brephos has not only the small white vesicle adherent to the body already referred to, but considerable vestiges of the other embryonic sacs invest and are attached to it. These are distinguishable by their grey colour, and. are com- paratively bulky. In the case of one brephos, they are attached to it; in the case of the other, the young one came away from the membranes, which were found to be attached to the mesenteries, and possibly in process of being absorbed. Any suggestion of protection by the mother within her body of actually born young would of course be negatived by these facts. 5. Contributions to the Knowledge of the Visceral Anatomy of the Pelagie Serpents Hydrus platyurus and Platyurus colubrinus. By Frank E. Bepparpb, M.A., F.R.S., Prosector to the Society. (Received May 20, 1904. | (Text-figures 26-28.) Tn the comprehensive works of Milne-Edwards* and Meckel 7 and others, there are numerous references to various points in the anatomy of the pelagic Ophidia, while a more particular account of the viscera of one species has been given by Cantor. The lungs have been particularly dealt with, though very briefly, by Cope §; while Mi. G. W. Butler || has incorporated remarks upon some of these snakes into his general papers upon the asymmetry of the Ophidian lung. Both the last-mentioned papers contain references to previous literature. I have had the opportunity of dissecting one example each of the marine snakes Hydrus platyurus and Platyurus colubrinus which have been in my possession for some time, the latter specimen belonging to me, the former to the Prosector’s Stores. This dissection enables me to add something to our knowledge of the anatomy of the Hydrophiine, and to compare two quite dis- tinct generic types. I had not expected to find them so different as dissection showed them to be. (1) Platyurus colubrinus. The specimen of this snake which I dissected measured in all nearly 17 inches,.of which a little over two belong to the tail. * “Tecons sur la Physiologie et Anatomie comparée.’ + ‘Anatomie comparée.’ French Transl. < “© Observations upon Pelagic Serpents,” Trans. Zool. Soc. 11. p. 303. § “On the Lungs of the Ophidia,” Proc. Am. Phi]. Soc. xxxiii. 1894, p. 217. || In P. Z. S. 1892, p. 477, and P. Z. 8. 1896, p. 691. 10* 148 MR. F. E. BEDDARD ON THE VISCERAL [June 7, The location of the different organs of the body and the lengths: of some of them in relation to the total length of the body differ in this serpent from the corresponding situation and measurements, of Hydrus platyurus, which are placed side by side for the pur poses of easier comparison. P. colubrinus. — H, platyurus. Length of body to vent ...... 14 inches. 23 inches. From tip of snout to base ot heart. Pelkey ee oereee 53 inches. 5 inches. From tip of snout to anterior Gil GE IMEI Lodcoscssescsco006 5% inches. 6 inches. From apex of heart to anterior GING GH WAFER soaccochaosooceso0 4 inch, 3 mech. Iheneth of liver ~...-77-72-.---- 32 inches. 3 inches. Posterior end of liver to gall- Alera Gleiskee ee er sec eee une. 12 inches. Posterior end of liver to TOL Sl¢ a 7 91 71 7 anterior end of kidneys ... 1g (R), 22 (Lb). 72 (1), 7 (®). Length of kidneys ........... » > Finch, J inch. 22, 22 Posterior end of kidneys to 5 il 2) (fj D il J1 WOTI tn Os. jase nee ose eeees Ilse (Es)), ge (LI) aire ((les), es (IE) It is plain from these measurements that the situs viscerwm shows important differences in the representatives of the two Hydrophiine genera. In Hydrus the heart is placed much further forward, i.e. the neck is shorter than in Platyurws, and the kidneys of the latter are relatively further from the posterior end of the body. In both, the anterior end of the liver hes close behind the heart, as is the case in Viperine serpents, a fact also_ noted by Cantor 1m the species dissected by him. Alimentary viscera.—The liver is long and thin, and thus characteristically Ophidian. It is very long compared with that of Hydrus, measuring as it does about one-fifth of the total body-_ length (including the tail), or nearly one-fourth if the tail be omitted. Taking the length of the body (to the vent) as 100 in all cases, the following are the proportions of the liver im a number of serpents. Zamenis flagelliformis, 193. | Hydrus platyurus, 145. i gemonensis, 205. Python spilotes, 153. Coronella getula, 23%. Eryx gaculus, 223. Coluber melanoleucus, 183. » conicus, 215. Tarbophis obtusus, 213. Heiterodon platyrhinus, 223. Naia tripudians, 214. — - Bow constrictor, 174. Ophiophagus bungarus, 242. | Causus rhombeatus, 232. Platyurus corallinus, 213. | Lroheterodon madaguseari- ensis, 201, The liver shows a peculiarity not without interest, which has been also figured and referred to by Dr. Cantor * in the Sea-Snake dissected by him. In Platywrus, however, this character—the * Toc. cit. 1904. ] ANATOMY OF PELAGIC SERPENTS. 149 division of the liver by furrows running at right angles to the longitudinal axis of the body—is more marked. 1 counted five of these transverse furrows, which vary in depth and divide the liver into a series of segments of hepatic substance. As I shall point out later (see p. 151), Hydrus platyurus shows the same “ segmentation ” of the liver. Milne-Edwards observes* of the liver of Zyphlops that it is “divisé en lobes plats,” and this lobation is figured also by Cope 7; but it does not appear to be by any means so regular asin Platywrus. One cannot but put down this marked lobation to the regular bending of the body im swimming, and it forms an example of “ segmentation” probably traceable to a definite mechanical cause. The gall-bladder gives off a single duct which soon forms a very complicated network in connection with the hepatic duct. This network is very much more complex than in Hydrus, and the rete of ducts is so long before it enters the duodenum that the gall- bladder can be dissected out and pulled much further away from the alimentary canal than is possible in Hydrus. The pancreas seems to me to be smaller proportionately (it certamly is so actually) than in Hydrus. The coiled region of the mtestine is very long. When the coils are left undisturbed within their ~ coelomic space, they measure 7 of an inch, but when unwrapped no less than 5 inches. The kidneys are approximately equal in size, each measuring about 2 inch in jiength. They are broad in proportion to their length, and almost suggest those of the Boid Eryx. The right, anterior, kidney hardly at all overlaps the left, which commences where it ends. S Lung. It has been pointed out by several zoologists, including Cope <, that Platywrus and some other genera of Sea-Serpents possess the tracheal lung found also in a few genera of terrestrial Colubrines. Cope’s statement on the matter is as follows:—* Finally the tracheal lung, as I shall call it, is distinct from the true lung in Platyurus and in Chersydrus. In the former of these genera, the trachea is not separate fromthe lumen.” Ido not think, however, that any detailed description of the lung exists. J shall endeavour therefore to supply this omission by the following description. There is no trace that I could discover of a second lung. The single lung extends to within one inch of the cloacal aperture and ends abruptly without any special diminution of calibre. It les, posteriorly at any rate, on the right-hand side and is firmly bound to the dorsal parietes. The tracheal lung is, as Cope says, distinct from the bronchial lung; the two are separated at the end of the one and the beginning of the other by one of the pulmonary vessels which passes between them, The tracheal lung begins very high up in the body, close behind the head; it ends posteriorly just in front of the origin from the heart of the right aortic arch. * “Lecons sur la Physiologie et Anatomie comparée.’ + “On the Lungs of the Ophidia,” Proc. Amer. Phil. Sec. xxxini. 1894. 46 ILM. Cit Ws Alf Me 150 MR. F, E. BEDDARD ON THE VISCERAL [June ig Until very near its posterior end the trachea forms merely a gutter along its ventral side, that is it opens freely into it through- out. Posteriorly, however, there is a distinct tendency for the lung to be connected only at intervals with the trachea, a hint— as it appears to me—of the commencing independence of the two parts of the pulmonary apparatus. The thoracic lung is very plainly a respiratory organ down to about the middle of the liver and to a less extent posteriorly. Its texture is like that of many other serpents, but not like that of Hydrus (see p. 153). It presents, as is shown in the accompanying drawing (text-fig. 26), A portion of the internal surface of the lung of Platywrus colubrinus. a honeycombed and perfectly regular structure. The depressions are naturally hexagonal through mutual pressure. I could find no difference in the structure of tracheal and bronchial lungs. The bronchus is traceable for a long way down the liver. Ata little distance behind the posterior end of the liver (about ? of an inch), the lung, which has gradually become very narrow, widens out into a thin-walled dilatation of quite twice its former width. This dilatation is 3 of an inch long and fusiform in shape. It is thin-walled, and thus contrasts with adjacent regions of the lung ; but on its inner surface a honeycomb pattern is visible, the depressions, however, being very slightly marked. Behind this dilatation the honeycomb structure was also plainly visible. The greater part of the lung in fact seems to be utilised as a respiratory organ. It is particularly noteworthy that the posterior region of the 1904. | ANATOMY OF PELAGIC SERPENTS. 151 lung in this serpent has not the excessively thin-walled character that it has in most land-snakes. In the latter, when dissected, the anangious region of the lung seems to be little more than a space between the viscera of the posterior region of the body. In Platyurus the lung has thick walls throughout. Examined in transverse sections, the posterior region of the lung, some way after the dilatation, is seen to be undoubtedly a functional lung, inasmuch as the blood-capillaries are numerous and approach very near to the inner surface, being in fact only separated from it by the pulmonary epithelium. The muscular walls are very thick, especially the inner layer of circular fibres. Evidently, therefore, the lung is capable of considerable alterations In size. Dr. Cantor gives an account of the lungs of Hydrophis schistosa* Schlegel (= Enkg ydrina valakadien of My. Boulenger’s Catalogue), somewhat different from the facts as observed by myself in Platyurus colubrinus ; these differences may be doubtless put down to the circumstance that the two serpents are of different genera. In the first place, the tracheal lung would seem to be continuous with the bronchial lung, though this is not absolutely clear from Cantor’s figure. The tracheal lung is, moreover, of much less extent in Hnhydrina. The pattern of the meshwork is quite different. Finally the dilatation along the course of the pulmonic portion of the bronchial lung, such as I have found in Platyurus, is less than in Hnhydrina, and the extreme tip of the lung in the serpent dissected by me is not tied down to the parietes by any tag. The dilatation which he does figure is apparently part of the functional lung. (2) Hydrus platyurus. Of this snake the example at my disposal was a female with the ova immature. The position of the viscera has been already dealt with in considering Platyurus and a comparison of the two snakes there instituted. Alimentary viscera.—The gall-bladder, pancreas, and spleen are not unlike those of the Sea-Snake figured by Dr. Cantor =. After receiving the hepatic duct, the bile-duct plunges into the substance of the pancreas on its way to reach the duodenum. Whether it forms a reée therein or not, t have not ascertained ; but it is clear that there is not room for a very extensive one. The pancreas is much lobulated, and extends in front of, and behind, the gall- bladder. The liver, as will be gathered from the measurements on p. 148, is actually, as well as relatively, shorter than that of Platyurus colubrinus. It is, however, of a more massive structure and is less divided by transverse furrows into “segments.” I detected only four of these. The liver also appears to me to be a little closer to the heart than itis in Platyurus. The coiled region of the intestine is relatively rather shorter than in Platyurus. Kidneys.—The kidneys of Hydrus platyuwrus differ markedly from those of Platyurus colubrinus. They are in the first place * “ Observations upon Pelagic Serpents,” Trans. Zool. Soe. ii. p. 305. + Loe. cit. pl. 57. fig. 1 m. i Loe. cit. pl. a7. figs. 1, 2. 152 MR. F. E. BEDDARD ON THE VISCERAL | June ie very long and thin, measuring, as has been mentioned, 2? inches each. While, therefore, the body of Platyurus is rather more than half the length of that of Hydrws, the kidneys of the latter are five times the length of those of the former genus. They do not, however, seem to me to be much, if any, greater in bulk, since they are extremely thin and narrow, while those of Platyurus are wider in proportion to their length. Another difference shown by the kidneys of the two genera is the extent to which they overlap. In Platywrus the two kidneys hardly overlap at all, whereas in Mydrus the right anterior kidney extends alongside of the first two inches of the left kidney. There is thus a considerable approximation here towards symmetrical kidneys. : $ Lung. The lung-tissue is recognisable from two and a quarter inches behind the tip of the snout. It begins, therefore, very early in the neck. The lung-tissue begins gradually between the dorsal Text-fig. 27. A portion of the internal surface of the lung of Hydrus platyurus. non-tused ends of the tracheal rings. This snake has therefore, as has been stated, a tracheal lung. The trachea, moreover, is open throughout into the lung, and forms in fact only a gutter along its ventral surface. The neck of this serpent is short, and therefore the tracheal lung is not of great extent; it appears to cease about half an inch in front of the heart, and therefore has not a course of more than two inches. A sudden widening at that pomt I take to be the commencement of the bronchial lung. There is no other differentiation that I can detect. The inner 1904. ] ANATOMY OF PELAGIC SERPENTS. 15s) surface of the lung is marked by conspicuous thickenings ; but these do not form a choneyeonip network as in Platyurus. The folds run across the lung—that is at right angles to its longi- tudinal axis, and are wavy in outline, which of course allows of expansion during inspiration (text-fig. 27, p. 152). The trachea (or bronchus) is continued as a gutter down the lung as far as the com- mencement of the liver, that is for a very short distance behind Text-fig. 28. A portion of the internal surface of the non-vascular part of the lung of Hydrus platyurus. the heart. = The lung itself is extraordinarily long; it is indeed co-extensive with the body-ecavity, reaching as far back as the vent. Its calibre too is very consider able, and its walls have the tough, almost shiny, appearance of a fish’s swim-bladder. The dorsal w. all is firmly fixed to the parietes. The latter part of the lung, 7. e. after the end of that region which is respiratory, has undergone some modification in relation to its undoubted function as a swim- bladder. The folds already spoken of in the vascular region of the lung persist; but their object is no longer that of merely increasing the respiratory surface, and thus the efficiency of the lung as a ane eathing organ, They exist only as bands of muscular fibre, which, since their direction is mainly circular and parallel to ond other, must act as constrictors and expel air from the swim-bladder part of the lung. On a naked-eye inspection of this portion of the lung, it looks almost as if the bands in question were the bronchial rings which had in this region taken on a new development. They are, however, serially continuous not 154 MR. F. E. BEDDARD ON A PARASTERNUM [June He, with the dwindling bronchial rings, but with the vascular folds already spoken of. Microscopical examination shows them to he thick bands of muscular fibres, which are not striated. The direction of these bands is circular, but they occasionally send off anastomosing branches as is shown in the accompanying sketch (text-fig. 28, p.155). Occasionally, too, a small band detaches itselt from one of the main hoops and ends upon the wall of the lung. Between the various bands the wall of the lung does not appear to be muscular. Towards the posterior end of the lung this arrangement is lost ; there are no longer separate hoops of muscle but the walls are covered with a single sheet of musculature. It is clear from the above description, which may be compared with that of Platywrus on p. 150, that the structure of the lung differs very considerably in the two species. In Platyurus the lung has retained to a much greater extent its pulmonary function, the ‘lung- -substance extending much further back than in Hydrus. The lung has, however, a less calibre, even proportionately, than in Hydrus, and there is no trace of so marked a conversion into a ‘“ swim-bladder ” with special muscles effecting its contraction and expansion as occurs in Hydrus. In the latter genus, moreover, the whole lung extends further back in the body than it does in Platyurus. On the Presence of a Parasternum in the Lacertilian Genus Tiliqua, and on the Poststernal Ribs in that Genus. By Frank H. Brpparp, M.A., F.R.S., Pro- sector to the Society. | Received June 3, 1904. | (Text-figures 29 & 30.) The use of the term “abdominal ribs” for the ventral and superficially placed cartilages, fibrous bands, or ossifications so distinctive of certain groups of Reptiles, is open to the objection that the term “ribs” has already a definite meaning attached to it. It implies cartilaginous, fibrous, or ossified rods which have or have had a connection with the vertebral column, whereas the so-called abdominal ribs have no relation whatever to the vertebral column, but are purely ventral structures formed between the plates of the ventral abdominal musculature. Furthermore, the term ‘abdominal ribs” is actually in use as descriptive of structures which are real ribs, and which have nothing to do with what other authors have called abdominal ribs. For example, Mr. Boulenger, in his ‘ Catalogue of Lizards,’ has referred, under the name of abdominal ribs, to the ventral region of ribs in the Geckos, Chameleons, and some other forms, lying behind the sternum, which meet, pair by pair, in the ventral median line, thus completing a series of hoops encircling the abdomen. Inasmuch, as Dr. Gadow* justly points out, that * Cambridge Natural History : Amphibia and Reptiles, p. 504. 1904. | IN LIZARDS OF THE GENUS TILIQUA. 155 these ribs bear “an extraordinary resemblance to the so-called ‘abdominal ribs’ of other reptiles,” it is particularly to be regretted that the term has been used in so authoritative a work as the Catalogue to which reference has been made. Since this confusion has been quite unnecessarily introduced, it will be as well to adopt the word “ parasternum,” already used by Fiirbringer, Gadow, and others. The Lacertilia are at least generally supposed to be without a parasternum, which is one of the points of difference used to distinguish them from the genus Sphenodon. The above quotation from Dr. Gadow implies this general view, which is more explicitly stated in the ‘Royal Natural History’*. I can find no statement in such works as that section of Bronn’s ‘ Thier-Reich’ which is devoted to the Lacertilia, as well as in other textbooks, to the effect that a parasternum is to be found in the Lacertilia; and Iam therefore free to conclude that the knowledge of its actual occurrence is at least not widely spread. In a brief preliminary note in ‘ Nature’ + I pointed out the existence of a series of abdominal ribs in the Scincoid Lacertilan Tiliqua scincoides, and I herewith submit to the Society a more detailed account of these structures (text-fig. 29, p. 156), which I have not up to the present succeeded in observing in any other lizard. The chevron bones of the abdominal-vib system are thin and not always easy to see; their slender bulk, as it appears to me, fully accounts for the fact that they have been previously over- looked. They are not nearly so stout, so numerous, or so closely adpressed as these bones are in an example of Hatteria of about the same size as the two specimens of Tiliqua scincoides which I have examined. For these reasons the bones would very readily be lost in preparing skeletons of 7iliqua. The distance separating the chevrons in 7%iliqua is 8 mm., when the abdominal muscles are gently stretched but not overstretched; the same distance in the case of the dried abdominal skeleton of Hatteria was not more than 4 or 5 millimetres. In my specimen of Hatteria there were quite twenty of the chevrons; I could not detect more than seven in one of the two specimens of Ziliqua. In the other specimen these abdominal ribs were hardly at all obvious. The fact that these chevrons are quite independent of the ribs— and their purely superficial position, lying as they do in the ventralmost sheet of muscle of the abdominal wall—is in favour of regarding them as the homologues of the abdominal ribs of Hatteria. It remains, however, to be shown that they overlap the true ribs as the abdominal ribs do in Hatteria. The ribs in Tiliqua atter the sternum do not apparently reach so far ventrally as they do in Hatteria. It might therefore conceivably be held that we had here to do merely with the ventral moieties of ribs which were defective laterally, and that the condition occurring * © Another important feature in which the Order [Squamata] differs from all the preceding ones is the absence of any system of true abdominal ribs or of their equivalent a plastron ” (vol. v. p. 107). + May 5th, 1904, p. 6. 156 MR. F. E. BEDDARD ON A PARASTERNUM [J une 7, in Tiliqua was an impertect representation of that occurring in the Chameleon and other lizards, where a number of true ribs join ventrally behind the sternum. Thus in dAcontias and Typhlo- saurus, of which the former is a near ally of Zliqua, there are present these post-sternal ribs meeting in the middle line, and thus simulating abdominal ribs *. In the Chameleon the most ventral layer of the abdominal musculature is so thin that it requires a careful dissection to show that the ventrally meeting ribs do not lie in this layer but in the deeper layer. In Tiliqua, on the other hand, the ventral musculature is thick, and the two Text-fig. 29. Abdominal ribs of Tiliqua scincoides. A points to one of these ribs. A meshwork of tendons lies between and over them. layers are easily distinguishable. When the ventralmost layer is raised, the abdominal ribs are carried with it, and are seen to end off in it with shghtly forwardly-curved lateral ends. But—more important than this—on the deeper layer of the musculature are two or three cartilaginous bars, lying obliquely (see text-fig. 30, p. 157), which ave clearly the ventral portions of true ribs such as in Acontias make complete hoops. The discovery of these entirely sets at rest any doubt as to the fact that the bars of cartilage which ule superficially are really quite comparable to a parasternum. The true ventral moieties of the ribs are not in connection with their vertebral portions, so far as I have been able to ascertain. They would thus be very readily lost in macerated skeletons, and might even be overlooked m a dissection. In his ‘ Catalogue of Lizards * Peters, ‘ Reise nach Mossambique, Amphibia, pls. xii. & xiii. 1904. | IN LIZARDS OF THE GENUS TILIQUA. 157 in the British Museum, Mr. Boulenger uses as part of his definition of the Scincide * the sentence ‘ Ossified abdominal ribs are absent”; and in defining the Anelytropide + he includes in that definition the phrase ‘“ Abdominal ribs present,” It is not absolutely clear from the wording used whether these two nearly related families are to be distinguished (inter alia) by the absence or presence of “abdominal ribs” (7. ¢., as already pointed out, the ventral moieties of post-sternal ribs) or by the mere absence or presence of ossification in such ribs. In any case, not only has Peters figured complete hoops in Acontias, but I have been able Text-fig. 30. Tiliqua scincoides. Ventral Hap of musculature (B) in which lie the abdominal ribs, turned down to show the deeper-lying muscles (C), in which are imbedded the true ribs (D). to show in the present communication that considerable vestiges of the ventral moieties of post-sternal ribs occur in Tiliqua. It is to my mind possibly a matter for further enquiry as to how far the median ventral region of the post-sternal ribs which actually meet each other behind the sternum may not be actually a parasternum fused with true ribs. For the median region of the chevrons, whether of undoubted abdominal ribs or of true vertebral ribs, appears at the surface of the abdominal musculature. This median region, therefore, at least of the post-sternal true ribs may conceivably be a vestige of a parasternum in those Lacertilia where such ribs occur. * Vol. iii. p. 130. + Vol. iii. p. 430. 158 DR. E, A. GOELDI ON THE [June 7, +. On the rare Rodent Jinomys branichii Peters. By Dr. Emit A. Gortpi, C.M.Z.S8., Director of the Goeldi Museum, Para. {Received May 16, 1904. ] (Plate X.*) The zoological world was surprised in 1873 by the novelty of the discovery of a strange, large Rodent, introduced scientifically by Prof. Peters, then Director of the Berlin Museum, under the name Dinomys branickii. It was stated that the animal had been found in the neighbourhood of a town in Peru, wandering about in an orchard. It was further stated that the natives themselves were entirely unacquainted with the creature. Prof. Peters published a somewhat extensive memo on the subject, based principally upon anatomical features, especially of the skin and the more or less complete skeleton. Up to this date I have not had the opportunity of consulting this memoir, which I know only from a few lines of quotation in recent manuals on mammalogy, such as Flower and Lydekker, ‘ An Introduction to the Study of Mammals,’ London, 1891, p. 489; and Beddard, *‘ Mammals,’ London, 1902, p. 495 seg. These citations are barely sufficient to permit a certain identification of this peculiar form of Rodent and to exclude the possibility of confusion. So far as the literature at my disposal goes, there is no indi- cation of any living specimen having been examined by a zoologist, nor does there exist any notice of any further specimen as having been met with since 1873, so that the type specimen at the Berlin Museum remains till now the only one known to be in existence. I have now the extreme good fortune to be able to make further additions to our knowledge of the subject. I consider the rediscovery of Dinomys branickii in the Amazonian region, from a general point of view, as the second most important * For explanation of the Plate, see p. 162. + Since beginning to write the present article I have received both the extensive memoir of Prof. Peters as well as the preliminary communication on the subject in the ‘ Monatsberichte der kénigl.-preuss. Akademie der Wissenschaften zu Berlin, July-August 1873, pp. 551-552. Their contents correspond so nearly to my suppo- sition that I find no necessity for changing anything m the wording of my note. The coloured lithographic figure given by Prot. Peters on plate 1. identifies the animal satisfactorily, but, of course, can make no claim to rival the photographs taken from the living specimens, which give an essentially different conception of the animal’s physiognomy. The most noteworthy difference consists in the fierce expression of countenance, resembling that of an angry rat ready to bite; while my photographs show a face which might be called the personification of perfect good humour. As I foresayy, there still exists no information as to the anatomy of the scft parts of Dinomys, which in all probability is still in reserve for me to bring to hght.— April 7, 1904. ' + On reading the memoir of Prot. Peters, I find—what was absolutely impossible to presume from the brief references in the above-cited manuals—that this type specimen does not belong to the Berlin Museum, but was only lent to its Director to be described by him, and that it forms part of the collections of the Warsaw Museum, for which it had been obtained by the Polish naturalist and explorer Constantin Jelski. 2 1S, WO4s, woll i Wl 2&, Bale, Sons and Danielsson, Ltd, IDL INVOIMEYAS: IB IRUEUIN INC CIN ‘ies ) 1904. | RODENT DINOMYS BRANICKII. 159 scientific event that has occurred during the first decennium of the existence of the Para Museum, after the refinding of the long-lost Dipnoan Lepidosiren paradoxa, as fully described in former volumes of the Zoological Society’s publications. My surprise was unbounded when, a few weeks ago, I received from a friend of the Museum, personally unknown to me, a cage containing a pal of live large Rodents entirely new to me, notwithstanding a twenty years’ acquaintance with the Brazilian fauna, both from the South and the North. These, after a quarter of an hour of eager consultation of the more modern manuals of mammalogy, I recognised as being identical with the almost legendary Dinomys braniehia, of which science had lost sight for thirty years. “It seems to me to be a most opportune occasion to publish a short description of the external features and the habits of Dinomys, as it is probable that the memoir of Prof. Peters, the only source of information on the subject up to the present, is deficient in these particulars. A ‘series of photographs of the animal, taken in different attitudes spontaneously assumed, supplement my verbal explanations. Doubtless the first impression from the general appearance, especially as regards size and external features, would ally the animal closely with the South-American “ Paca” (Agouti paca) ; and the verdict of the scientist as to its systematic location has no reason to differ from the popular conception which has given to our Rodent the local name of “ Paca-rana,” a Tupy-term meaning “‘ pseudo-paca.” In size the older and larger specimen, the mother, is almost exactly of the dimensions of an average Paca ; the young one is approximately half or two-thirds of that size. The similarity is increased by its capacious head, and the brownish colour marked with distinctly visible longitudinal rows of white spots. This similarity, however, begins to diminish as soon as one undertakes a more mimute examination. It presents first the striking difference of having a hairy tail of about 2 deci- metres in length. Still further, as regards the hair, while that of the Paca is soft to the touch and of even length, resembling somewhat the fur of the Otter, Dinomys has a rougher coat of stiffer hairs, uneven in length (these stiffer hairs being white- tipped and longer, and contributing to interrupt the uniformity of the general colour), and forming a first transition-step towards the quills of the coats of the Spiny Rats proper (Loncheride, Kehinomyide). The white spots are located only on the rear part of the trunk, beginning from the shoulders. They are separated in two groups: (1) four longitudinal rows, nearly continuous at the fore part, the median line of the back being dark; (2) two lateral areas of more or less circular white spots occupying the flanks and thighs, scattered irregularly, as it would seem at first sight, but betraying a tendency to longitudinal arrangement on closer observation. The form of the head of Dinomys may be described as sub- 160 DR. E. A. GOELDI ON THE [June 7, pyramidal, while that of the Paca is subconical, due to its promi- nent cheek-bones and bulging eyes and the more slender snout. The former thus resembles more the head of the Capybara. What gives to the head a peculiar martial aspect are the immense white whiskers, equal in length to the animal’s head, forming a thick tuft on each side. The nostrils are §-shaped, and have the same valvular arrangement for closing the aperture as the Manatee, for example, showing that the animal is adapted to a semi-aquatic life. While the prominent eyes of the Paca, like large black beads, have a glaring glossy aspect, the eyes of the Dinomys are not at all prominent, are of asoft light brown colour, the pupil being a vertical slit during the hours of day- light. They are thus of similar appearance to the eyes of certain smaller Didelphyidee (Didelphys quica, &c.) and of the large Two-clawed Sloth (Cholapus didactylus); and this similarity naturally leads to the conclusion that the animal is rather of nocturnal habits—a conclusion which is corroborated by actual observation. Both front and hind feet are fow~toed. The general build of Dinomys is thick-set and inclined to corpulency. Due to the fact of setting the whole plantigrade sole on the ground, the hind feet especially, the Dmomys has a waddling gait, Al reminds one of an immense rat well advanced in development towards a bear. Both my specimens of Dinomys are of a peaceful, phlegmatic disposition ; they are most excellent boarders in our Zoological xarden, and cause very little trouble on condition of having enough to eat. They devour prodigious heaps of provender, eating being one of thei principal occupations both day and night. As far as observed up to the present, they show a preference for bread, squashes, and fruit of different indigenous palm-trees, such as the orange-coloured * Tucuma ” (Astrocaryum tucuma). While eating, their favourite position is standing erect on the hind feet, and in this attitude there is a good opportunity of admiring the dexterous manner in which they can hold a fruit in their paws, notwithstanding the lack of the counter-pressure of a thumb. Such a fruit revolves in their grasp as though it were in the chuck of a slowly-turning lathe, and the large chips which fly fill us with respect for the immense gouging faculty of the large but comparatively shghtly prominent incisors. The erect position while eating—which is the only attitude thus far observed while taking fool—reminds one more of the “ Cutia” (Dasyprocta) among the larger South-American Rodents, and forms a note- worthy difference from the *Paca,’ which I have never seen eating in any other position than on all fours. The predominant feature of the character of Dinomys is a con- bination of leisurely movements and supreme good nature. It knows absolutely nothing of haste. Spending the greater part of the day sleeping in a corner—the mother often lyimg upon the young one, or standing over it, as if to protect and to keep it warm—opening its half-closed eyes only when it hears the ‘ 1904. | RODENT DINOMYS BRANICKIL. 161 approaching steps of the keeper, it forms the resolution to move with slow gait, expecting some food, evidently governing its movements as much by hearing and smell as by sight. It is not easily irritated, and permits one to stroke and to scratch its head and back, and only occasionally manifests its displeasure by a low euttural erowl. I have never yet observed a manifest intention to bite. When let out of the cage it makes no attempt to escape, and limits its excursions to an exploration of the immediate neighbourhood in search of something to eat. It occasionally scratches itself rapidly with its long claws, which is the only occasion on which it manifests a capacity for rapid movements when required. One thing not yet definitely verified by us is its proclivity for digging, the development of the claws at least leading to the supposition that the animal is well fitted for that purpose. The amiable relations always existing between mother and son prepossess one most favourably as to the natural disposition of the animals * This phlegmatic disposition seems to me to be a very precarious endowment for the struggle for life; and considering the evident advantages which result to the smaller domestic rodents, such as rats and mice, from their nervously active constitution, it would not be strange if the species should tend to disappear. The apparent rarity of Dinom ys may possibly find its explanation in the consequences of such a psychological endowment in a more nervous environment; but it is also possible that this rarity is because of the circumstance that the real habitat of the species has not yet been clearly ascertained. As matters now stand, it would be justifiable to suppose that the true home of Dinomys i 18 not properly in the Peruvian Andes, and that the first specimen found there was merely a stray individual, and that its actual habitat may rather be located in the almost unexplored regions of the eastern slopes and tablelands of the Bolivian and Peruvian foot-hills bordering on Brazil, including geographically the head- waters of the rivers Acre, Purts, and Jurua. I shall soon have occasion to show that a scientific exploration of the region above described will result in a multitude of great surprises both from a zoological and a paleontological point of view, of which the interesting rediscovery of the lost Dinomys branickit is only a first instalment. Para, 7th April, 1904. P.S.—Unfortunately, just before I send this note, the older Dinomys, the mother, has died owing to a difficult parturition. One fetus was born under normal conditions, while the other, apparently on account of its abnormal position, could not be * After this portrayal of the animal’s peaceful character, it will not seem strange that the account of the capture of the first individual in 1873 in the Peruvian mountains, as furnished by Prof. Peters, according to which it was deemed necessary to deal two powerful sabre-strokes to lay the terrible monster low, always amuses me. Proc. Zoot. Soc.—1904, Vou. IT. No. XI. 11 162 ON THE WILD CAT OF TRANSCAUCASIA, [June 7, delivered. This unforeseen event will now give, much sooner than I expected, an opportunity to fulfil my promise to undertake a thorough study of the anatomy of the soft parts, as well as of the two fully-developed foetuses. ith April, 1904. EXPLANATION OF PLATE X. Dinomys branickii, from photographs of two living examples (a mother and young) (about <4; nat. size), and the skin of the former (about + nat. size) in the Goeldi Museum, Para. 8. The Black Wild Cat of Transcaucasia. By C. Satunin, of Tiflis, C.M.Z.8. [ Received May 16, 1904. ] Although the existence ih Transcaucasia of a Black Wild Cat was known long ago, the animal has never been described nor scientifically named. Hochenacker speaks of this cat, so far back as 1837 *, as a Felis cato affinis; but as all his text is in Latin, this cannot be regarded as a scientific name forthe animal. I myself have men- tioned this cat as Felis sp. in my paper on the Fauna of the Caucasus, as well asin the Catalogue of the Caucasian Museum $f; but I have not had the opportunity of giving a description of it until now. That it really is a Wild Black Cat I knew well, as all the spe- cimens I have had the opportunity of examining were alike, and as it is by no means rare in its native haunts. There remained one important point to decide about this animal, namely, Was it not a melanie form of the common Wild Cat, Felis catus L. % Thanks to the material in the Zoological Museum of the St. Petersburg Academy of Sciences, where I have found two mounted specimens, three skins, and three skulls of this cat, I am now certain of its specific validity, and I name it Melis demon §, of which the following is the description :— Size of a big male domestic cat. Colour ranging from black with a slight reddish tinge to reddish dark brown. ‘This colouring is somewhat lighter on the under side of the body, on the mner surface of the extremities, and on the distal under surface of tail. Very long white hairs are scantily dispersed all over the body. In a certain light, dark-black transversal stripes are visible on the sides of the fore part of the body, these stripes being more conspicuous on somewhat faded skins. The whiskers, as well as the eyebrow-bristles, are brown. The tail is considerably longer than in /elis catus. Claws white, transparent, and with a mother-of-pearl lustre. * Bull. Soc. Nat. Moscou, 1837, p. 186. + Zool. Jahrb., Syst. ix. p. 289 (1897). ‘Museum Caucasicum,’ i. p. 24( 1899). i ; “Demon” being the hero of everal Caucasian legends, and also of the poem by * db Lermontoff with the same title 1904. | ON A NEW FORM OF BUFFALO, 163 The measurements, in millims., of the two mounted specimens and of the three skins are :— Mounted specimens. Skins. I — (= rae tag aaa ? No. 671. No. 2685. No.2997. No.3001. No.4945. From tip of nose to base of tail 610 560 6) Oe 3) a odo 0) Tail, hair included ............ 309 ~=8 40 340 380 370 Ear, from external base ... 51 5] — 54 46 Skull: as compared with the skull of Felis catus that I have had for comparison, the three skulls of this new eat differ only in that the frontal part is somewhat narrower. Moreover, in Felis catus the nasals reach further back than the ascending branches of the upper mandible, whereas in Melis demon the upper man- dibular bones go further back than the nasals. The measurements of two complete skulls, in millims., are :— No. 2997. No. 3001. dis : Gireaibestillemte tas. yek bya 5 20) Lee Ge, Peete eee 98 101 iBasilarWemert byes... 22. tis ea. eee Si 81 Greatest breadth across the zygomatic bones 69:5 71 Smallest breadth behind the postorbital POIROXGISSIS) AG hn Apt ne ence pan En RRARREMPL tA cfc huo,cio 31°2 34 Smallest breadth between the interorbitals. 19 20 Greatest breadth of brain-case .............0. 46 46°5 Ibrerarerialn Ont lovoragy JRVENIE) A seancoonsboncnotcencosas nr 36 33 Width between upper premolars ............ 38°5 39°D Condylar lengchyot maxilla) 22-2 .a..ceeeneeeen 61 64 Felis demon is not scarce in the woods of the southern slopes of the chief range of the Caucasus (Nookhinsky district of Gouv. Elizabethpol) and the Zakatalsky Province. It used to be found, according to Hochenacker, also in the mountains of the Small Caucasus; but nothing is heard of it there at the present time. According to Radde, this cat inhabits the Kopet-Dagh Range in Transcaspia; and M. 8, Alphéraky tells me that it is by no means rare in the woods of Borshom, where it is often trapped, together with /elis catus and other vermin. Habits unknown. 9. On a Buffalo-Skull from East Central Africa. By R. LypexKer. [Received May 17, 1904. ] (Text-figure 31.) [The complete account of the new form described in this communication appears here; but since the name and preliminary diagnosis were published in the ‘ Abstract,’ the species is distinguished by the name being underlined.—Ep1r7or, | Through the kind offices of Rowland Ward, Limited, the British Museum has recently been presented with the skull and tail-tip of a bull Buffalo, killed by the donor, Mr. Arnold Mathews, Lie 164 MR. R. LYDEKKER ON A [June 7, of Ballynahinch Castle, Co. Galway, in the Mfumbiro district, on a volcanic mountain west of Kiva, on the borderland between the Congo Free State and German Hast Africa, in forest at an elevation of between 7000 and 8000 feet. These specimens I cannot identify with any described form, and therefore propose to regard them as representing a new race of African Buffalo. Although I cannot find Mfumbiro in any atlas, I take the district in question to be the area lying between the Albert Nyanza and Tanganyika, that is to say, approximately, the Mapi country. The following note on the Buffalo to which the specimens belonged has been communicated by Mr. Mathews :— “The hide of this animal is not bare like that of the South- African Buffalo, but covered with a dense crop of black hair all over. The height of the buffalo is considerably less than that of the big South and Hast African races. The animal seems to me to be a race midway between the Congo and the Hast-African Buffalo. It shows, however, no tendency to red. In addition to its small size, peculiar shape of horns, and density of pelt, the only peculiarity I noticed was that the tail had a white tip. I shot another bull out of the same herd, exactly like the one of which I sent you the skull, only younger and smaller.” The skull and horns are evidently those of a fully-adult bull, and the latter present the following measurements :— Along outer curve ............ 21°5 inches. Baisalioirtla e.Witee acl eeecnac cee Ser peta Gureatestispaniiwas- en. seeaanne 2599)", Tip to tip interval ............ ISH) Compared with the various local forms included in my work on ‘Wild Oxen, Sheep, and Goats,’ under the specific title of Bos (Bubalus) caffer, the skull and horns of the present form come nearest to those of 5. caffer nanus of the Congo region. They are, however, considerably larger than the latter ; and the horns (text-fig. 31, p. 165) are thicker, less sharply incurved, and with a much more sinuous front surface, bending sharply backwards immediately behind the basal frontal expansion, and then curving somewhat forwards with the commencement of the inward inclination. An interval of about an inch and a half separates the two horns in the middle of theforehead. Speaking generally, these appendages are intermediate between those of the Cape and those of the Congo Buffalo, although on the whole nearer to the latter. Here it may be mentioned that there is an almost complete gradation, as regards the extent of the frontal sinuosity, from the horns of the Cape Buffalo through those of the present and the Congo forms, to the Senegambian race (B. caffer planiceros), in which this curvature is practically obsolete. From the typical red Congo Buffalo (B. c. nanus) the present form differs by its deep-black coat, whereby it agrees with the Cape animal, from which, however, it is broadly distinguished by the dirty-white tail-tuft. In reference to Mr. Mathews’s note, 1904. ] NEW FORM OF BUFFALO. 165 it may be mentioned that two Cape Buffaloes in the British Museum have a well-developed coat of black hair. As this Buffalo has obviously nothing to do with the one from East Africa recently described by Mr. Thomas*, I propose to regard it as a new race, under the title of Bos caffer mathewsi (cf. Abstr. P. Z.S. 1904, no. 9, p. 10, June 14), the Museum spe- cimens being of course the types. It may be shortly defined as a race of the approximate size of B.c. nanus, but with the abundant hair black instead of red, and the tail-tuft white ; the horns being at the same time larger, more curved outwardly, and then inclined backwards near the lateral bend. Text-fig. 31. Front view of skull of male Buffalo, Bos (Bubalus) caffer mathewsi, from the Mfumbiro district of East Central Africa. + nat. size. I always feel I owe an apology to naturalists when adding to the list of subspecies; but the present form has certainly claim to recognition, on account of its tending to connect the Cape with the Congo Buffalo—two forms which at one time I had some compunction in regarding as specifically the same. * P.Z.S. 1904, vol. i. p. 371. 166 MR. R, LYDEKKER ON A [June 7, 10. The Ichang Tufted Deer. By R. Lyprxxmr. [Received June 8, 1904. | (Text-figures 32 & 33.) [The complete account of the new species described in this communication appears here, but since the name and preliminary diagnosis were published in the ‘ Abstract,’ the species is distinguished by the name being underlined.—Eprror. | The genus Hlaphodus has been hitherto known by two species— the typical ZL. cephalophus and the perfectly distinct H. michianus, from the Ningpo district, province of Chekiang, on the east coast of China. J am now able to adda third. A few days ago Mr. A. E. Leatham called at the Natural History Museum, bringing with him for determination the skull and skin of a young male Tufted Deer (Hlaphodus), shot by himself last January in the mountains near Ichang, province of Hupei, Central China. Ichang, it may be mentioned, is fully a thousand miles from Ningpo; and the deer killed by Mr. Leatham was shot high up in the mountains far away from water, whereas 1. michianus 1s reported to inhabit the reed-brakes on the Ningpo rivers. On looking through the specimens in the British (Natural History) Museum, I found an adult male skin and skull of an Hlaphodus from Ichang, collected by Mr. F. W. Styan in 1901 (B.M. INO eS 2c): Externally, Mr. Leatham’s specimen of the Ichang Tufted Deer differs from “1. michianus by its decidedly darker and more uniform colour, which is blackish brown, passing almost into black on the limbs, while there is more white on the tail, of which only the basal two-thirds of the upper surface is dark, so that the whole of the tip is white. The skin of the adult male sent by Mr. Styan is browner, except down the middle of the back, but exhibits the same uniformity in general colour. How different these skins really are in colour from that of H. michianus, it is not easy to deter- mine, seeing that the specimens of the latter in the Museum are more or less faded by exposure. On comparing the skull of Mr. Leatham’s specimen, which is immature and retains the milk-molars, with a skull of a male EH. michianus of nearly the same age, I find that the nasal bones are absolutely and proportionately shorter in the former, their length being 17 inch against 24 inches. Moreover, their length is contained 31 times in the basicranial length, instead of less than 3 times. Another distinction is to be found in the form of the anterior upper milk-molar, which is much shorter (antero- posteriorly) in the Ichang than in the Ningpo skull; and there also appear to be slight differences in the form of the upper molars. Both skulls, it may be observed, show no trace of the pedicles of the antlers, which must accordingly be very late in development. Equally noteworthy differences are presented when the adult 1904. | NEW SPECIES OF TUFTED DEDR. 167 skull (text-figs. 32 & 33, A) obtained by Mr. Styan at Ichang is compared with an adult male skull of 4. michianus from Ningpo (No. 78.11.14.3) (text-figs. 32 & 33, B), the latter bemg somewhat older than the former. The Ichang skull shows the shortness of the nasals characteristic of the immature specimen, the length of these bones being 23 inches, against 33 inches in #. michianus. Text-fig. 32. AN), NOY Np Oe Eo AM bts ris: ; ye A Left lateral view of adult male skulls of Hlaphodus michianus (A), and H. ichangensis (B). pr.mex., premaxilla; pr.v., preorbital cavity. The antler-pedicles are much more developed in the Ningpo skull, and have a much more outwardly-directed ridge connecting them with the orbit (as shown in text-figure 32, A); but this feature may be due, to some extent, to difference in age. The flat surface of the jugal below the orbit is quite narrow in the Ningpo, but 168 ‘MR. R. LYDEKKER ON A [June 7, broad in the Ichang skull. Very noticeable is the difference in the form of the preorbital fossa, which is rounder, deeper, and less oblique in the Ichang skull. More striking still is the difference in the canines of the two forms, that of the Ichang skull being nearly twice the length of that of the older Ningpo specimen, as well as differing somewhat in shape. Such differences cannot possibly, I think, be accounted for by wear, as I cannot conceive Text-fig. 33. Front view of adult male skulls of Hlaphodus michianus (A) and H. ichangensis (B). fr-, frontal ; na., nasal. how such tusks could become worn, while those of the Ningpo skull show no signs of wear. I may add that the figured skulls differ in the form and relations of the premaxille, but 1 find this to be a very variable feature. Compared with a female skull of the typical Tibetan 2. cepha- lophus in the British Museum (No. 92.7.13.1), which is the only specimen of the latter species available, the adult Ichang skull 1904. | NEW SPECIES OF TUFTED DEER. 169 makes a much nearer approach than it does to that of H. michi- anus, showing a similar conformation of the bones below the orbit, and apparently also a very similarly shaped preorbital cavity. The nasal bones are, moreover, of much the same relative length, their long diameter being one third of the basi- cranial length. On the other hand, #. cephalophus is at once distinguished by its greatly superior size, as is shown by the following table of the basicranial lengths of the four skulls available for comparison :— II, TOC OKTIDUIS, WHE Aon aneeoenodsjo6eor 63 inches : Female. issu doen ces heen Gy Nchanmesadulitymale yess. at i- Oe iss Li. cephalophus, female ............... (ace From this it is obvious that the Tibetan skull cannot possibly be the female of the form indicated by the adult Ichang specimen. If a male skull of 1. cephalophus were available for comparison, I have little doubt that important differences between it and the Ichang specimen might be detected. As it is, the former species appears to have a much longer type of skull than the Ichang animal, from which it is also distinguished (so far as can be determined) by the much more solid structure of the walls of the preorbital cavity. In any case, the Ichang Deer is broadly distinguished from the Tibetan Hlaphodus cephalophus by its greatly inferior size. The foregoing differences seem to leave no doubt as to the specific distinctness of the Ichang Tufted Deer, which may be named Hlaphodus ichangensis (cf. Abstr. P.Z.S. 1904, No. 9, p. 10, June 14). Itis characterised, as compared with L. michianus, by its darker and more uniform colour, whiter tail, smaller antlers, larger tusks, shorter nasals, and smaller, deeper, and more evenly circular preorbital fossa; while it is smaller than /. cephalophus. The type will be the aforesaid adult male skull, collected by Mr. Styan. In conclusion, I may take the opportunity of referring to the skull of a female Tufted Deer (B.M. No. 98.3.7.18) obtained, together with the skin, by Mr. C. B. Rickett at Fing-ling, Fokien, lying considerably to the south of Ningpo. This skull is of the same general type as that of the female of Z. michianus, but differs by its superior size, the basicranial length in the two specimens being respectively 63 and 6Linches. The Fokien skull is further distinguished by the less marked elevation of the hinder half of the frontal region, which does not develop a median crest, and by the expansion and flattening of the platform formed by the base of the preorbital cavity and the anterior zygomatic root. ‘There are also differences in the form of the paroccipital process in the two skulls, which, with other details, and their marked difference in size, appear to justify the reference of the Fokien Tufted Deer to a separate race, under the name of Llaphodus michianus fociensis. 170 DR. A. SMITH WOODWARD ON [June 7, 11. On Two New Labyrinthodont Skulls of the Genera Cupitosaurus and Aphaneramma. By A. Surra Woop- WARD, DED FuktSs) KeZs: [Received June 3, 1904. | (Plates XI. & XII.*, and Text-figure 34.) Among the remains of Labyrinthodonts acquired by the British Museum during recent years, there are two skulls of unusual interest. One was obtained from the Triassic sandstone of Staffordshire, while the other was found in a formation, appa- rently of the same geological age, in Spitzbergen. The first specimen is of importance as displaying the occipital region of the skull more clearly than any Labyrinthodont hitherto described; and the second specimen adds facts concerning the shape and relations of the quadrate bone. Each fossil represents a new species, while the second is the only skull of a Labyrintho- dont as yet described from the Arctic Regions. I. CAPITOSAURUS STANTONENSIS, Sp. nov. (Plate XI.) The discovery of the Staffordshire specimen in the Lower Keuper sandstone of Stanton, near Uttoxeter, was briefly noticed by Mr. John Ward four years agot. The block of sandstone containing the skull was split along the plane of the cranial roof, leaving most of the roof-bones adherent to one slab, while the impression of these, with the rest of the skull, remained in the counterpart-slab. When Mr. Ward examined the fossil it was still in the condition. in which it had been exposed by the acci- dental fracture; so that his description could only be of a general and preliminary nature. Since its acquisition by the British Museum, both parts of the specimen have been skilfully prepared by Mr. C. Barlow; and the principal characters of the skull, as now observable, are illustrated in Plate XJ. The skull does not appear to have been much distorted by crushing, and its shape closely resembles that of the skull of a broad-nosed crocodile. All the external bones are similarly ornamented with a coarse network of sharp ridges. The occipital border is slightly excavated, and the deep notches for the auditory meatus (aw.) are nearly, if not completely, surrounded by bone. The specimen is a little fractured in this region. The orbits (orb.) are set far back, only of moderate size and regularly oval in shape, with the long axis directly antero-posterior, not oblique. The pineal foramen (pin.) is a rather large circular vacuity. * For explanation of the Plates, see p. 176. + J. Ward, “On the Occurrence of Labyrinthodont remains in the Keuper Sand- stone of Stanton,” Trans. N. Statis. Field Club, vol. xxxiv. (1900), pp. 108-112, pls. iv., v. PZ. 5 1904 voll Pod. ' 1 1 M L. J.Green del.et Lith. CAPITOSAURUS STANTONENSIS. Mintern Bros.imp. Bate aun WOALVELSOU VANVEENVHdV Rarer ee uectney TXTd 114 vO61 SZ d 1904. ] LABYRINTHODONT SKULLS. 171 The sutures between the roof-bones in the hinder half of the skull are clear on both sides of the fossil; but the elements in the rostral region cannot be distinguished. There are the usual two pairs of small bones forming the occipital margin between the auditory notches, the middle pair being larger than the lateral pur. The parietal bones (pa.) are a symmetrical pair, twice as wide in their posterior half as in their anterior half, and trun- cated in front where they meet the frontals (/r.). Their maximum width is less than their total length, and the pineal foramen is situated at their middle point. The frontals just enter the rim of the orbit at its antero-internal portion; but the parietals are completely excluded from this rim by the antero-posteriorly elongated postfrontals (pif). The squamosal element (sq.) is longer than broad, articulating in front chiefly with the post- orbital, but also meeting the postfrontal. The postorbital (péio.) forms the posterior and half the externo-lateral rim of the orbit ; the jugal (j.) enters the outer portion of this rim; while the prefrontal (prf.) completes it in front. The large prosquamosal (psq.) is about two-thirds as deep as long, tapering to a point forwards. The quadrato-jugal (qj.) is about two-and-a-half times as long as deep, also tapering to a point forwards, and extending shightly downwards below the level of the tooth-bearing edge of the upper jaw. The jugal (j.), as usual, is irregularly triangular in shape, with the radiating ridges of its ornamentation pre- dominant. At the rostral end of the fossil represented in Plate XI., the eranial roof and underlying matrix are removed to expose the palate from above. The antero-posteriorly elongated posterior nares (pin.) are thus well seen; and the hinder edge of the large premaxillary vacuity (pv.) is also preserved. The posterior limit of the premaxilla (pmz.) is distinct on the left side. The exten- sive vomerine plates (v.), divided by a median suture, are also well preserved. The premaxillary teeth, shown in transverse section, are small, uniform in size, and arranged in close series. The maxillary teeth, exposed in side view, are similar. Hach tooth appears to have been a hollow cone with plicated walls. An excavation in the sandstone reveals hollows left by the decay of the usual large teeth, which occur on the palate in front of and behind the posterior narial openings. As already mentioned, the most interesting part of the skull is the occipital region, which is especially well preserved (fig. 2). It does not project sufficiently far backwards to admit of the occipital condyles being seen when the cranium is viewed directly from above. ‘The position of the foramen magnum (/m.) is clear; while a vacant space (socc.) between this and the roof-bones of the occipital border suggests that a supraoccipital element was present but unossified. Below the middle of the foramen magnum there is a narrow tongue of bone (bocc.) separated by a little matrix from the prominent occipital condyles. This is probably 172 DR. A, SMITH WOODWARD ON [June 7, to be interpreted as a partially ossified basioccipital element forming only an insignificant part of the occiput. The condyles themselves (¢.) are convex and seem to have been covered with unossified cartilage, while they constitute the hindermost end of a pair of bones which enter extensively into the base of the cranium and also rise upwards to bound the foramen magnum on each side. These elements (ex.) therefore exhibit the same disposition as the exoccipitals of a frog, and must be homologous with the latter. ‘They are pierced by the foramen for the vagus nerve, which is well seen on the left side. The posterior face of each exoccipital immediately above the condyle is impressed with a triangular fossa, so that the upper end of the bone, partly obscured by matrix, has the appearance of bifurcating to unite in jagged sutures with the two bones which descend from the occipital row of roofing plates. The latter plates are evidently superficial in origin—either dermal or splint ossifications—but there is no clear line of demarcation between them and the iunmediately underlying bones just mentioned. The inner half of Text-fig. 34. wt. op PP. soce. SE. Capitosaurus stantonensis; hinder view of occiput, restored, two-thirds nat. size. boce., supposed basioccipital; ¢., occipital condyle on exoccipital; ep., epiotic; ex., exoccipital; fir., foramen magnum; op., opisthotic; psq., prosquamosal ; pt., pterygoid; ptv., postero-temporal vacuity ; ¢7., quadrato-jugal; gw., quadrate ; socc., space for supraoccipital ; s¢., supratemporals. each plate of the median pair would directly overlap the carti- laginous supraoccipital ; while its outer or lateral half alone is underlaid by the bone (ep.) which occupies the place of a piscine or amphibian epiotic. The inner two-thirds of each plate of the outer pair is underlaid by the second bone (op.) of triangular shape, which exhibits the usual relationships of an opisthotic. As shown on the right side of the fossil, the jagged suture between these two otic bones coincides exactly with that dividing the two overlying plates. The pterygoids unite in a deeply jagged suture with the broad basioccipital region, and then expand behind on each side into a vertical lamina (pi.) which articulates with 1904. ] LABYRINTHODONT SKULLS. 173 the inner edge of the quadrate (qu.). The quadratojugal is in contact with the quadrate only at its lower end, thus leaving a rather large postero-temporal vacuity (ptv.) which is especially well seen on the left side. The complete occipital region, as here interpreted, is shown in the accompanying restored sketch (text-fig. 34, p. 172). It obviously conforms much more closely to the Amphibian type than to the arrangement in any known reptile. It therefore agrees with the evidence derived from the palate and the supratemporal plates, showing that the Labyrinthodonts are Amphibia. The characters of the cranial roof, the anterior end of the palate, and the teeth, prove that the skull now described belongs to the genus Capitosaurus. It cannot, however, be referred to any of the known species of this genus. It differs from the skull of the type-species, C. arenaceus, from the Keuper of Franconia”*, in the narrower and more tapering form of its rostral region, the more elongated shape of the orbits, and its more nearly circular pineal foramen. It is distinguished from the skulls of C. nasutus T and C. robustus =, among other characters by the slighter excava- tion of its occipital border and by the shape of the outer pair of its supratemporal plates. It is also distinguished from the im- perfectly known skull of C. fronto$ by the different shape of its auditory notch and the coarseness of its external orna- mentation. The new specimen therefore represents a hitherto unrecognised species, which may be named C. stantonensis. II. APHANERAMMA ROSTRATUM, gen. et sp. nov. (Plate XII.) The imperfect Labyrinthodont skull from Spitzbergen was found by Profs. J. W. Gregory and HE. J. Garwood in the Trias of Sticky Keep, associated with a few short and deep, biconcave vertebral centra, which probably belonged to the same animal. It is broken into five pieces, which show little beyond impressions of the bones. Three fragments are internal moulds, while the other two pieces bear marks of the exterior. The right half of the postorbital region and the middle portion of the rostrum are missing. Enough, however, is preserved to indicate that the skull must have been of a much elongated shape, with a slender snout. It is very little depressed behind, and the occipital plane slopes backwards. Its external surface was ornamented with large reticulating rounded ridges, which were sometimes inter- rupted, as shown by the impression of part of the left cheek * Graf zu Minster, Neues Jahrb. fiir Min. 1836, p. 580; H. von Meyer, ‘ Die Saurier des Muschelkalkes ’ (1847-55), p. 152, pl. lix. figs. 5-7. + H. von Meyer, Palwontogr. vol. vi. (1858), p. 222, pls. xxiv.-xxvi. t+ H. von Meyer, ‘Die Saurier des Muschelkalkes’ (1847-55), p. 146, pl. lix. figs. 1-4, pl. Ixi. fig. 10. Cyclotosawrus robustus, Lh. Fraas, Paleoutogr. vol. xxxyi. (1889), p. 121, pls. ix., x., pl. x1. figs. 1-4. § H. von Meyer, Paleontogr. vol. vi. (1858), p. 229, pl. xxviii. fig. 2. 174 DR. A. SMITH WOODWARD ON [June 7, (fig. 4). The grooves for the slime-canals must also have been deep and conspicuous, as shown by the same fragment. The most interesting feature in the skull, its quadrate bone (figs. 1-3, qu.), is well preserved on the left side and exposed from behind. This bone is long and narrow, antero-posteriorly com- pressed, and inclined somewhat backwards. Its inner face is sheathed by a great vertical plate of the pterygoid, while its outer or lateral border is in complete contact with the bones of the cheek. It is quite clear that there was no postero-temporal cleft or vacuity; and the quadrate bears no impression of the external auditory meatus. There are no vacuities in the bony covering of the postorbital region of the cheek, and the limits of some of its constituent elements are well shown by the natural internal mould. The long and narrow ovoid squamosal (figs. 1, 2, sq.) is distinct ; while the outline of the equally elongated and still narrower postorbital (pto.) can be traced without difficulty. There can also be little doubt that the postfrontal (pif) and prefrontal (prf.) exclude the parietals and frontals from the margin of the orbit. The parieto- frontal region, however, cannot be satisfactorily interpreted. It seems probable that the parietals (pa.), which must have been relatively large, taper rapidly in front, where they extend nearly to the anterior border of the orkit. They must also have been either depressed or thickened in the median portion, where their anterior end articulates with the very narrow pair of frontals (/r.)— suggesting an arrangement nearly like that represented by Cope in the skull of Cricotus*. The prosquamosal (psq.) is an iregularly triangular plate, somewhat longer than deep, with the apex forwards. The quadrato-jugal (qj.) is nearly twice as long as its maximum depth. The jugal (j.) is especially elon- gated and forms the greater part of the infero-external border of the orbit (orb.), which is oval in shape and directed both laterally and upwards. There are also some traces of the edge of the postfrontal (pif), prefrontal (prf.), and lachrymal (/a.) bones, but the middle part of the rostrum is lost. An internal mould of nearly the terminal portion of the snout (figs. 6, 7) demon- strates the slenderness of this region and the elongated shape of the sasal bones (7.). The palate is of the typically Labyrinthodont pattern, with a complete parasphenoid (fig. 6, pas.), which is laterally compressed in its middle portion. The large posterior lamina of the ptery- goid, in a vertical plane, abutting on the quadrate, has already been mentioned. In front of this expansion the pterygoid of each side curves outwards to meet the jugal (and probably also the hinder end of the maxilla) in a long suture. Its extent anteriorly is uncertain owing to the absence of the middle of the rostrum. Further forwards the palatines (fig. 6, pl.) are rather * HW. D. Cope, Proc. Amer. Phil. Soc. vol. xvii. (1878), p. 529. 1904. | LABYRINTHODONT SKULLS. 175 broad plates; while the vomers (vo.) bound the antero-posteriorly elongated posterior nares (pén.) both behind and within. The only teeth observable are indicated by an impression of the middle part of the pterygoid region (fig. 5), where they occur chiefly in two close parallel rows, one on the outer edge of the pterygoid (pt.), the other on the maxilla (mx.). The teeth of each series are nearly uniform in size, but those of the pterygoid are some- what larger than those of the maxilla, They are all cones with the usual thickened and folded walls. There are also traces of an irregular patch of minute teeth or tubercles still further back on the pterygoid. ‘Three vertebral centra found in the same formation and locality as the skull are deeply-biconcave discs, not pierced by any foramen for the passage of a remnant of the notochord. One obliquely crushed specimen is shown of the natural size in Pl. XII. fig. 8, and another imperfect specimen is similarly represented in fig. 9. The concavity at the end of the centrum is very slight near the outer rim and suddenly deepens towards the centre. It is difficult to determine the precise affinities of so frag- mentary a Labyrinthodont skull, but if the vertebral centra are rightly ascribed to the same animal, it evidently represents one of the higher members of the Order. Though suggestive in some respects of Cricotus, as already mentioned, the new skull differs from that of Cope’s genus in the thickening or depression of the middle part of the parietals, and in the very strong external sculpture. The associated vertebral centra also differ from those of Cricotus in being completed discs. Among other Labyrintho- donts, the specimen from Spitzbergen seems to approach most closely the skull of 7’rematosaurus*, with which it agrees in its general shape, external sculpture, dentition (so far as seen), and the relations of the quadrate bone. It is distinguished, however, by the peculiar disposition of the parietal bones. The new fossil thus represents a hitherto unknown genus, which may be named Aphaneramma and defined as follows :—Skull elongate-triangular, with the orbits widely separated and situated in its hinder half ; external bones strongly sculptured, and grooves for slime-canal deep. Parietal bones extending forwards between the orbits; frontals very long and narrow: both these elements excluded by the postfrontals and prefrontals from the orbital border. A single regular row of small teeth on the pterygoid parallel with the equally uniform row of teeth on the maxilla; clustered small teeth or tubercles further back on the pterygoid. Vertebral centra complete biconcave discs, not perforated. The type- specimen of the type-species, A. rostratum, is the imperfect skull now described. * H. Burmeister, ‘Die Labyrinthodonten aus dem bunten Sandstein von Bern- burg,’ pt. i. (1849). ‘The minute posterior pterygoid teeth are seen in a specimen of Trematosaurus in the British Museum (no. R. 1733) and in one in Mr. W. HE. Balston’s collection. 176 ON LABYRINTHODONT SKULLS. [June 7, It may be added that vertebra, ribs, and other fragmentary fossils from the Trias of Spitzbergen have already been referred to Labyrinthodonts * ; but the specimens described are scarcely sufficient for exact determination, EXPLANATION OF THE PLATES. Prats XI. Capitosaurus stantonensis, sp. nov. (p. 170); from Lower Keuper, quarries of Messrs. Peter Ford and Sons, Stanton, near Uttoxeter. [Brit. Mus. no. R. 3174.] Fig. 1. Upper view of skull, two-thirds nat. size. 2. Occiput of the same, two-thirds nat. size. Prater XII. Aphanerammea rostratum, gen. et sp. nov. (p. 173) ; from Trias, Sticky Keep, Spitzbergen. [ Brit. Mus. nos. R. 8180-82. | Fig. 1. Upper view of imperfect hinder half of skull, internal mould, one-half nat. size. 2. Left side view of the same, one-half nat. size. 3. Hinder view of quadrate region of the same, one-half nat. size. 4, Ornament of part of cheek of the same, showing slime-canal, one-half nat. size. 5. Bases of teeth on maxilla and pterygoid of the same, nat. size. 6. Upper view of end of snout, internal mould, one-half nat. size. 7. Palatal view of the same, one-half nat. size. 8, 9. Imperfect vertebral centra, one-half nat. size. LETTERING :—aw., opening for auditory meatus; boce., supposed basioccipital ; ¢., occipital condyle; ep., epiotic; ea., exoccipital; fin., foramen magnum ; fr. frontal; j., jugal; da., lachrymal; ma., maxilla; ma., nasal; op., opisthotic; orb., orbit ; pa., parietal; pas., parasphenoid; pin., pineal foramen; pl., palatine; pme., premaxilla; prf, prefrontal; psq., prosquamosal; pé., pterygoid; ptf, post- frontal; pén., posterior nares; pto., postorbital; ptv., postero-temporal vacuity ; pe., premaxillary vacuity; g7., quadrato-jugal; qz., quadrate; soce., space for supraoccipital; sqg., squamosal; sé¢., supratemporal ; v., vomer. * N. Yakowlew, “Neue Funde von Trias-Sauriern auf Spitzbergen,” Verhandl. russ.-k. min. Ges. vol, xl. (1902), p. 180, pl. iii. ; “ Nachtrag,” doc. cié. vol. xli. (1904), pp. 165-169. 1904.| THE SECRETARY ON ADDITIONS 10 THE MENAGERIE. 177 November 15, 1904. Dr. W. T. Buanrorp, C.LE., F.B.S., Vice-President, in the Chair. The Secretary read the following reports on the additions made to the Society's Menagerie during the months of June, July, August, September, and October, 1904 :— The number of registered additions to the Society’s Menagerie during the month of June was 149, of which 44 were acquired by presentation, 20 by birth, 15 by purchase, 69 were received on deposit and 1 in exchange. The number of departures during the same period, by death and removals, was 121. Among the additions special attention may be called to :— 1. A male Buffon’s Kob (Kobus kob), new to the Collection, received on deposit on June 2nd. 2. A male Goral Antelope (Nemorhedus goral), presented by Major Rodon, F.Z.S., on June 3rd. 3. A male Chimpanzee (Anthropopitheeus troglodytes), from the Congo, purchased on June 5th. 4. A male Speke’s Antelope (Vragelaphus spekii), new to the Collection, received on deposit on June 27th. 5. Two Ural Owls (Syrnivm wralense), veceived on deposit on June 29th. The number of registered additions to the Society’s Menagerie during the month of July was 181, of which 74 were acquired by presentation and 22 by purchase, 63 were received on deposit, and 22 were bred in the Menagerie. The number of departures during the same period, by death and removals, was 162. Among the additions special attention may be called to :— 1. A Riippell’s Colobus (Colobus abyssinicus), deposited on July 2nd. 2. Two Japanese Bears (Ursus japonicus), one presented by Miss Violet M. Lakin on July 11th, and the other deposited on July 28th. ‘3. A male Sing Sing Waterbuck (Kobus wnctuosus) ; 4. A male Gambian Ourebi (Ourebia wigricaudata); 5. A male Ostrich (Struthio camelus) ; and 6. Three Levaillant’s Parrots (Pocephalus robustus): presented by Capt. Sir George Denton, K.C.M.G., F.Z.8., on July 14th. 7. Two young male Greater Koodoos (Strepsiceros ludw), deposited on Suly 15th. 8. A Yellow-crowned Penguin (Ludyptes antipodum), and 9. Two Rock-hopper Penguins (Ludyptes chrysocome), purchased on July 23rd. The number of registered additions to the Society’s Menagerie during the month of August was 270. Of these 70 were acquired Proc: Zoou. Soc.—1904, Vou. II. No. XII. 12 178 ON PERE DAVID’S DEER IN HAINAN. [ Nov. 15, by presentation and 28 by purchase, 87 were born in the Gardens, and 85 were received on deposit. The number of departures during the same period, by death and removals, was 182. Among the additions special attention may be called to :— 1. A female Lar Gibbon (Hylobates lar), presented by Mr. C. R. Stokoe on August 9th. 2. A Capybara (Hydrocherus capybara), presented by Messrs. the Liebig Extract of Meat Co. on August 18th. 3. Two young female Gorillas (Anthropopithecus gorilla), obtained by purchase on August 19th. The number of registered additions to the Society’s Menagerie during the month of September was 156, of which 69 were acquired by presentation, 10 by purchase, and 77 were received on deposit. The number of departures during the same period, by death and removals, was 142. Among the additions special attention may be called to :— 1. A fine specimen of Wolf's Monkey (Cercopithecus wolfi), deposited on Sept. 5th. This beautiful species is figured in P.Z.S,. 1894, pl. vii. 2. An Aru-Islands’ Kangaroo (Jacropus brunii), deposited Sept. 9th. 3. A pair of Kelp Geese (Chloéphaga antarctica), from the Falkland Islands, presented by Mr. Vere Packe, Sept. 14th. The registered additions to the Society’s Menagerie during the month of October were 128 in number. Of these 76 were acquired by presentation and 12 by purchase, 3 were born in the Gardens, 36 were received on deposit and 1 in exchange. The total number of departures during the same period, by death and removals, was 163. Among the additions special attention may be called to :— 1. Three Beatrix Antelopes (Oryx leucoryx), one from Aden, a female, presented by Mr. G. W. Bury on Oct. 5th, and a pair from the Persian Gulf, presented by Major P. Z. Cox, I.8.C., F.Z.8., on Oct. 22nd. 2. Two King Birds of Paradise (Cicinnurus regius), from New Guinea, purchased on Oct. 15th. New to the Collection. 3. A new species of Moustache Monkey (Cercopithecus, sp. n.), from the Gaboon, deposited on Oct. 15th. 4, A DuChaillu’s Monkey (Cercopitheeus nigripes), from the Gaboon, deposited on Oct. 15th. 5. Two Schmidt’s Monkeys (Cercopithecus schnuidti), from the Congo, deposited on Oct. 17th. 6. Two Blue-tailed Fruit-Pigeons (Carpophaga concinna), from the Aru Islands, presented by Mrs, Johnstone on Oct. 17th, Mr. R. Lydekker communicated the result of recent corre- spondence connected with the sketch by a Chinese artist of a Deer from Hainan purporting to be Pére David’s Deer (Zlaphurus 1904. | ON THE ANTLERS OF THE ALTAL STAG. 179 davidianus), recently exhibited to the Society (Proc. Zool. Soc. 1904, ii. p. 83). When this sketch was exhibited it was understood to have been drawn from a specimen in the possession of Mr. E. T. C. Werner, then British Consul at Hainan, and was thus taken (in spite of the addition by the artist of a pair of Peking-Deer antlers) as evidence of the existence of Hlaphurus davidianus in Hainan. In a letter dated June 26, Mx. Werner states, however, that the artist drew the sketch from memory after the death of the animal, and also that there was considerable doubt as to whether the latter really was Pére David's Deer (Tst-pu-hsiang) at all.. Nevertheless the writer expressed his belief that the species did exist in Hainan. On the other hand, in answer to further enquiries, Mr. Hughes, now Consul at Hainan, wrote that, so far as he could ascertain, Pére David’s Deer was unknown in Hainan, and that the drawing in question must be regarded as a fancy sketch. Under these circumstances, despite the fact that the sketch undoubtedly portrayed that species, no credence could at present be given to the alleged occurrence of Pére David’s Deer in Hainan. Mr. F. E. Beddard, F.R.S., called the attention of the meeting to a fact in the life-history of Kangaroos concerning which but little appeared to be known. An example of J/acropus dorsalis, which died on the 4th Nov. last, was found to contain a young one in the pouch which had survived the death of the parent. The young kangaroo was 6 inches in length (to the root of the tail) and still perfectly naked. On being removed from the pouch, it moved its limbs vigorously and emitted a sound which was rather more voice-like than a hiss. It was difficult to describe the nature of the sound accurately. It was uttered at continuous intervals. The production of any sound in so imperfectly formed an animal was remarkable. Mr. Frederick Gillett, F.Z.S., exhibited some antlers of the Altai Stag (Cervus eustephanus), and made the following remarks on their growth based on his own observations in the Society’s Gardens :— “An Altai Stag, purchased by the Society on Aug. 10th, 1897, although in poor condition and not expected to live, in the following year produced three sets of antlers and served two hinds, becoming the parent of a stag and a hind. The young stag shed his first pairs of antlers in a twelvemonth, and I am able to show these antlers to-night. This young stag grew very large, and when two years old bore antlers with twelve points. This stag was then sold to Hagenbeck. “The stag now in the Gardens shed a set of antlers on May 28th, 1902, having produced two sets in the year before ; and the three sets which I exhibit to-night were shed by it on the following dates :—Jan. 6th, 1903; June 10th, 1903; April 23rd, 1904.” 12 3k “ 180 DR. P. L. SCLATER ON THE OKAPI. [ Nov. 15, Dr. P. L. Sclater, F.R.S., stated that in July last he had visited Brussels in order to examine the specimens of the Okapi (Okapia johnstoni) in the Museum of the Congo Free State at Tervueren near that city. This he had been enabled to do by the kind permission of M. Emile Coart, Conservateur du Musée du Congo. The mounted series of the Okapi in that Museum consisted of a fine adult pair, of which the male carried short giraffe-like horns, as shown in a lithographic plate which was exhibited, while the female had none, and of a pair of skeletons in which the male had likewise horns but the female was horn- less. There were also two other mounted specimens of immature animals. Besides these specimens, Dr. Sclater was informed that others had been sent from Tervueren by order of King Leopold to the Museums of Tring, Paris, Stockholm, Madrid, Antwerp, and Rome. All the specimens, as Dr. Sclater understood, had been received from the Station of the Congo Free State on the Ituri, which was practically in the same forest-district as Fort Mbeni, where Sir Harry Johnston’s specimens had been obtained, although the Ituri belonged to the water-basin of the Congo, and not to that of the Nile. Dr. Sclater also called attention to an article “ dus dem dunkel- sten Africa,” published in the ‘ Basler Nachrichten’ for May 22nd last, and subsequently abstracted in ‘Globus’ of July the 21st last (vol. Ixxxvi. p.. 61), whereby it appeared that the writer, Dr. T. T. David, a Swiss naturalist resident at Beni on the Semliki, claimed to be the first European who has observed and obtained an Okapi in its native wilds. Dr. David had sent one of his specimens to Prof. R. Burckhardt, C.M.Z.8. (whose former pupil he had been), for the Zoological Museum at Basel, but Prof. Burckhardt had informed Dr. Sclater that it was unfortu- nately received in a bad condition. The following was an abstract of Dr. David’s principal remarks in the ‘ Basler Nachrichten ’ :— ; “The extremely elongated skull of the Okapi presents small rudiments of horns on the frontal bones. The animal in life has the general bearing of a Tapir; it is certainly a Ruminant, but its whole appearance, its actions in the swamps in which it lives, its compressed body and the way in which it carries its head, remind one of a Tapir and not at all of an Antelope, so that the stuffed examples of this animal in London and Brussels are quite erroneously set up. The striping of the limbs is much brighter than that of the Zebras. The back is red, especially so in the male ; the ears are enormously large, and are furnished with great tufts of hairs standing up. Small horns are present in some specimens, and, moreover, in both sexes, but are absent in others, which induces me to believe in the possibility of the existence of two species of Okapi. The underskin is as thick as in the Pachyderms, which makes it a very difficult animal to prepare.” Dr. Sclater concluded by saying that, notwithstanding what Dr. David had stated and the views of Prof. Lankester and Dr. Forsyth Major, he was quite unable to believe in the existence of more than one species of Okapi in the same limited district, though it seemed that the individual specimens presented some unusual modifications. 1904. | PROF. J. C. EWART ON EAST-AFRICAN ZEBRAS. 181 Mr. W. B. Tegetmeier, F.Z.S., exhibited a specimen of an Asiatic King-Crab (Carcinoscorpius rotundicauda) which had been picked up alive off the Isle of Wight. Prof. J. C. Ewart, F.R.S., exhibited some skins and a series of lantern-slides of the Zebras of East Africa, and read the following note on a form generally resembling in conformation and markings the Mountain Zebra of South Africa :— Some years ago Mr. Rowland Ward presented me with a stuffed Zebra which, though originally “traded out of Somaliland,” has a general resemblance to the Mountain Zebra of South Africa. This Zebra (now in the Royal Scottish Museum, Edinburgh) is in several respects so unlike the other forms hitherto described that, without waiting for its exact habitat*, it may be worth while pointing out how it agrees with, and differs from, the true Mountain Zebra, Text-fig. 35 Photo by G. A. Ewart. Ward’s Zebra, to show long ears and face-stripes. This Zebra (which may be known as Ward’s Zebra) very closely resembles the Mountain Zebra in height, in the form and size of the head, ears, and muzzle (text-fig. 35), in the mane, * It probably inhabits part of the area between the upper reaches of the Tana River and Lake Rudolf. : 182 PROF, J. C. EWART ON EAST-AFRICAN ZEBRAS. [ Nov. 15 tail, and hoofs, and in having the stripes over the rump arranged to form the so-called gridiron pattern (text-fig. 36). It differs from the Mountain Zebra in having a broad dorsal band (31 inches wide as it crosses the croup), all the hairs of which are directed backwards—in the Mountain Zebra the dorsal band midway between the withers and the croup is represented by a mere line, while from the croup to the mane the hair is. directed forwards, i.e. the whorl usually at the end of the mane in the Equide is on a level with the croup. The Mountain Zebra differs also from Ward’s Zebra (1) in having a dew-lap, (2) in having decidedly larger front chestnuts, (3) in having a larger number of stripes running at right angles to the dorsal band, (4) in having the legs more intensely striped, and (5) in the ground-colour being nearly white: in Ward’s Zebra, with the exception of the inner surface of the limbs and under surface of the body, where white prevails, the ground-colour is of a rich cream tint. Text-fig. 56. Photo by G. A. Ewart. Ward’s Zebra, to show “gridiron” and broad dorsal band. In Ward’s Zebra the stripes, except in the vicinity of the muzzle, are of a dark brown colour, the muzzle and the nostril- patches are darker than in the Mountain Zebra, and the stripes above the nostril-patches are of a pale brown hue. The face is PZ. S VIOG wollte lee ate H.Gronvold del.et lth. Mintern Bros .imp. SCOTONYVCREIRIS IWEADIFOIRD I, 1904. ] ON MAMMALS FROM FERNANDO PO. 183 decorated with four pairs of nearly symmetrically arranged stripes, widest apart on a level with the eyes, and with four pairs of stripes which meet in the centre of the forehead at or near the point where the mane terminates four inches below the occipital crest. In the above-mentioned stripes, as in those on the sides of the head and on the neck, there is close agreement between the two Zebras under consideration, but, as already stated, there are fewer stripes in Ward’s Zebra in connection with the dorsal band. If the “gridiron” in the two forms is compared it will be noticed that in Ward’s Zebra the bars running across the rump are coarser than in the Mountain Zebra, apparently owing to the obliteration of several of the intervening light spaces. In text-figure 35 the colour and great length of the ears in Ward’s Zebra are well brought out—the ears are longer than in any of the Mountain Zebras I have had the opportunity of measuring, and instead of presenting a white tip and a narrow white band midway between base and apex as in the Mountain Zebra, the apex is dark, while the proximal part is only faintly and irregularly pigmented. If one may judge by the ears, hoofs, and coloration, Ward’s Zebra is adapted for a habitat similar to that of the Mountain Zebra; moreover, like the Mountain Zebra, it has the reputation of being stubborn and intractable. The following papers were read :— 1. On Mammals from the Island of Fernando Po, collected by Mr. H. Seimund. By Oxpriztp Tuomas, F.R.S., Ags | Received July 13, 1904. | (Plate XIIT.*) (‘The complete account of the new genera and subspecies described in this communi- cation appears here ; but since the names and preliminary diagnoses were published in the ‘ Abstract,’ the genera and subspecies are distinguished by the names being underlined.—EDrror. | One of the chief desiderata of the British Museum collection of Mammals has long been a proper series representing the fauna of the Island of Fernando Po. For from this island there came in the early days of the study of zoology by British workers quite a number of specimens, and these were described in the ‘ Proceedings’ of this Society by Mr. G. R. Waterhouse and others. But owing to age and exposure to light at a time when the exhibition of types was not thought criminal, the original specimens, on which all our comparisons depended, have become so faded that but little use can now be made of them. * For explanation of the Plate, see p. 187. 184 MR. OLDFIELD THOMAS ON | Nov. 15, Attention having been again called to this island by the remarkable ornithological discoveries made there by Capt. Boyd Alexander, a special collecting-trip in the interests of the National Museum was rendered possible by the generosity of our President (the Duke of Bedford), of Mrs. Percy Sladen, and the Hon. Walter Rothschild. A free passage to the island and back was also given to the collector by Messrs. Elder, Dempster & Co., through the kind offices of Sir Alfred Jones. The collector, Mr. E. Seimund, started in November 1903, arrived in the island on December 4, and left again in April 1904, so that he had in all just over four months in which to collect. The series he obtained is exceedingly valuable for the reasons above mentioned, as he got good sets of nearly all the species described so long ago by our predecessors in mammalogical research ; and these cannot fail to be of constant service to all workers on the subject. Of novelties I have only had occasion to describe two—WScoto- nycteris bedfordi, a Fruit-Bat, and Galago demidoffi poensis, a local race of the little West-African ‘Galago; but Mr. Seimund has found on the island several other mammals which had not previously been recorded from there. Our knowledge of the Mammals of Fernando Po rests chiefly on the following literature :— WarerHouse, G. R.—Descriptions of new Mammals from the Island of Fernando Po, based on specimens presented by George Knapp, Esq. P. Z. 8. 1838, p. 57. Colobus, Cercopithecus, Genetia, Lutra, and Cephalophus. WarerHouse, G. R.—Descriptions of new Mammals from Fernando Po, obtained by Mr. L. Fraser during the Niger Expedition. P.Z.8. 1842, p. 124. Anomalurus and Squirrels. ALLEN, W., and Tomson, T. H. R.—Narrative of the Expe- dition to the River Niger. Appendix, vol. ii. pp. 472 et seqq. 1848. Most of the Mammals obtained on the Expedition were collected at Fernando Po by L. Fraser. Bocace, J. V. Barsoza pu.—Subsidios para a Fauna da Ilha de Fernao do Pé: Mammiteros. Jorn. Sci. Lisboa, (2) iv. p. 1, 1895. 16 species (none new) collected by Mr. P. Newton. Bocace, J. V. BArBoza pu.—Faune des Quatre Isles du Golfe de Guinée: Mammiféres. Jorn. Sci. Lisboa, (2) vii. p. 25, 1903. Full list of species. Descriptions of isolated species have also been published by Gray, Bennett, Ogilby, A. Smith, and others. The second of Prof. Bocage’s two papers gives a full list of the 1904. ] MAMMALS FROM FERNANDO PO. 185 Mammals of the island, and I have now, by intercalating the additional species obtained by Mr. Seimund and modifying one or two doubtful determinations, drawn up a list of the indigenous species complete to date, with the names of the collectors on whose specimens the species have been determined. This contains 36 species, as follows :— 1. Colobus pennanti Waterh. 2. A satanas Waterh. Bs polycomus Schr.* | 3. Cercopithecus erythrotis Waterh. 4.. o preussi Matsch. 5. Ss martini Waterh. 6. campbelli Waterh. (6urnetti Gray). a pogonias Benn. 8. Galago elegantulus Leconte. 9. 3 alleni Waterh. HOS aS demidoffi poensis Thos. 11. Hypsignathus monstrosus Allen. 12. Rousettus stramineus Geoff. 13. Scotonycteris bedfordi Thos. 14. Rhinolophus landeri Mart. 15. Hipposiderus fuliginosus 'Temm. 16. Nycteris hispida Schr. 17. Mimetillus moloneyi Thos. [| Glauconycteris poensis Gray F. | 18. Nyctinomus brachypterus Peters f. 19. Crocidura poensis Fraser. 20. Sylvisorex johnstoni Dobs. 21. Genetta poensis Waterh. 22. Poiana richardsoni Thos. 23. Lutra capensis poensis Waterh. 24, Anomalurus fraseri Waterh. 25. Sciurus stangeri Waterh. 265) 55 rufobrachiatus Waterh. 27. 9 punctatus Temm. 28. Funisciurus erythrogenys Waterh. 29. 55 poensis Smith. 30. Mus tullbergi Thos. 31. .,, alleni Waterh. 32. Cricetomys gambianus Waterh. 33. Procavia dorsalis Fraser. 34. Cephalophus ogilbyi Waterh. 35), $5 melanorheus Gray. 36. Manis tricuspis Raf. 1. Cotospus sATANAS Waterh. Knapp, Thomson. ; Knapp, Thomson, Newton, Seimund. Knapp, Thomson, Fraser, Burton, Seimund. Seimund. Knapp. Thomson. Knapp, Thomson, Fraser. Burton, Newton. Allen, Thomson, Burton. Seimund. Newton. Newton, Seimund. Seimund. Thomson. Fraser, Newton, Seimund. Fraser, Capt. E. Downes, Seimund. Seimund. Downes. Fraser, Seimund. Seimund. Knapp. Thomson, Seimund. Knapp. Fraser, Thomson, Alexander, Newton, Seimund. Fraser, Thomson, Newton, Seimund. Fraser, Thomson, Seimund. Newton. Fraser, Thomson, Seimund. Smith, Thomson, Fraser, Alexander, Seimund. Seimund. Allen, Seimund. Smith, Newton, Seimund. Fraser, Alexander, Newton, Seimund. Fraser, Thomas, Alexander, Seimund. Thomson, Fraser, Thrupp, Alexander, Seimund. Fraser, Newton. Native skin. Bubi Town, Bantabiri, 500 m. * There would appear to be some error in the inclusion of Colobus polycomus in the Fernando Po fauna. Prof. Bocage puts it in on the authority of Gray, who mentions two of Knapp’s skins as belonging to it. But the list of Knapp’s speci- mens given by Waterhouse in 1838 does not include it, and until some confirmation is obtained of its occurrence in the island I think it should be deleted from the list. + It is stated by Allen and Thomson (J. c. p. 480) that the specimen described by Gray as Kerivoula poensis was not obtained in Fernando Po, but at Abo on the Niger. It has therefore to be deleted from the island list. { N. pumilus Dobs. Of. de Winton, Ann. Mag. N. H. (7) vii. p. 38 (1901). 186 MR. OLDFIELD THOMAS ON [ Nov. 15, 2. CERCOPITHECUS ERYTHROTIS Waterh. Cercopithecus erythrotis Waterh. P. Z.8. 1838, p. 59. dg. 78,81. 9. 74. Bubi Town, Bantabiri, 500 m. go. 147. @. 160. Bantabiri, 1800 m. Well-known as it is by menagerie specimens, few Museums. possess any wild-killed examples of this handsome monkey, and the present specimens are therefore most acceptable. The original types of this monkey, of C. martini, and of Colobus satanas were native-made skins, presented to the Zoological Society’s Museum by Mr. George Knapp in 1838, being the earliest zoological specimens known to have come from Fernando Po. 3. CERCOPITHECUS PREUSSI Matschie. 2 (young). 165. N. Bantabiri, 1800 m. Two native skins. Not previously recorded from the island or represented in the British Museum. Dr. Matschie, during a visit to London, has examined these specimens, and considers them to be the same as the species described by him under the above name. His types came from Victoria, Cameroons, and he tells me that this southern part of the Cameroons has a fauna very like that of Fernando Po, such species as Colobus satanas, Cercopithecus erythrotis, and others. occurring there without modification. The presence of C. preussi in Fernando Po is therefore not surprising. 4. GALAGO DEMIDOFFI POENSIS Thos. Galago denudoffi poensis Thos. Abstr. P. Z.8. 1904, No. 10, p. 12, Nov. 22. 3. 152, 153, 162. 9. 167. Bantabiri, 1800 m. “Shot in tree by night.”—E. 8. Similar in all essential respects to the true @. demidoffi of continental West Africa, but the under surface is paler, whitish instead of buffy. General colour above, of specimens in full pelage, pale russet or cinnamon-brown. Central light line of face white, contrasting more markedly with the general colour than the more. or less buffy one of true demidofi. Hairs of under surface slaty for two-thirds their length, then either white or pale buffy yellowish. Outer side of limbs like body, a line along the inner sides pure white; the hairs white to their bases, and forming prominently white patches below the elbows and thighs. The corresponding regions in true demidoffi are buffy or yellowish, never pure white. Upper surface of hands and feet dull whitish. Tail dark brown, darkening slightly terminally. Ears apparently rather larger, such measurements as are available running from 27 to 30 mm., as against 24 to 27 mm. in demidoff. Skull much as in demidofi, though inconspicuously larger. Dimensions of the type, measured in the flesh :— Head and body 130 mm.; tail 195; hind foot 46; ear 28. 1904. | MAMMALS FROM FERNANDO PO, 187 Skull—greatest length 38:2; greatest breadth 25-5; interorbital breadth 5:4; breadth of brain-case 19°6 ; front of canine to back of m® 12:5. Type. Adult male, no. 152. B.M. No. 4.7.1.8. Killed 6 March, 1904, at an altitude of 1800 metres. Compared with 15 well-preserved specimens of the Continental form from localities ranging from the Gold Coast to Uganda, the four skins obtained by Mr. Seimund differ so uniformly by the lightness of their under surfaces, and the pure white of the inner aspect of their limbs, that I think they should have a special sub- specific name. Galago demidoffi has not been previously recorded as occurring in Fernando Po, though a skeleton obtained there was received from Sir Richard Burton in 1862, just after the publication of Gerrard’s ‘ Catalogue of Bones of Mammalia.’ 5. RovusETrus STRAMINEUS Geoff. SLOSS TNOn MASS: 117, VTS s Mosel S enor sll yi 113, 116, 120, 121, 122,123. Bantabiri, 10 m. 3g. 178,181. ©. 87, 105, 106, 179, 184. Bantabiri, 10 m. “Very common.’—KE.8. Prof. Bocage also records Hypsignathus monstrosus as having been discovered in the island by Mr. Newton. This fine series of specimens shows a peculiar and very unusual sexual difference in colour which does not seem to have been previously noticed. The males, without exception, are more or less brown, the bright yellowish shoulder-patches contrasting strongly with the general dark colour. The females on the other hand are, both above and below, of the rich yellowish straw-colour so often described as occurring in this species. For the female of any animal to be more richly coloured than the male is an unusual phenomenon. 6. ScoronycrEeRIs BEDFORDI Thos. (Plate XIII.) Scotonycteris bedfordi Thos. P. Z.S. 1904, vol. i. p. 372. @. 31. Fish Town, 10 m. “Shot during the daytime, hanging on a tree.” —E. 8. This most interesting Bat is the only new species obtained on Mr. Seimund’s expedition, and I have thought it worthy of a figure. It is the first member of the genus to be received by the Museum, and is therefore a most welcome accession. The previously known species, S. zenkeri Matsch., was described from the Cameroons. The external characters of S. bedfordi have been already described, but the following measurements of its skull may be of service :—Greatest length 25-4 mm.; basal length 22:6; zygomatic breadth 16°5; interorbital breadth 4°8; breadth of brain-case 11-1; palate length 14; front of canine to back of molar 87 ; front of lower canine to back of m, 10. The specific distinction of S. bedfordi rests mainly on the 188 MR. OLDFIELD THOMAS ON [Nov. 15, conspicuously smaller size of the ears, these being only 11 mm. in the island species and 17 in S. zenkeri. The skulls, judging only by Dr. Matschie’s description, seem closely similar. The cheek-tooth formule of Scotonycteris and Epomophorus are considered by Dr. Matschie to be P. 3, M. 2, and that of Cynopterus to be P. 3, M. 2; but the study of a young specimen of the last- named genus shows that Dobson was perfectly right in giving its formula as P. 3, M.3*. This young specimen has milk-premolars present above the second and third cheek-teeth in each jaw, thus showing them both to be premolars; the minute anterior tooth has, as usual, no predecessor. It would follow from this that the formula in Scotonycteris is P. 3, M. 3, the same as that rightly determined for Hpomophorus by Dobson. But further, while correctly determining the teeth of Cynopterus, Dobson does not seem to have realized that the same formula, P. 3, M. 3, would certainly be applicable to Vyctymene t (“ Har- pyra”), to which he assigns P. 2, M. 2. In Dobsonia, on the other hand, with the same total number of four cheek-teeth in the upper jaw, the tooth lost has obviously been the anterior premolar instead of the last molar, so that the formula should be, as Dobson puts it, P. 2, M. 2. Matschie erroneously gives it as P. 2, M. 2. 7. HirpposipERrvs FULIGINOSUS Temm. Go Oly OW, De Ti, Gs, MOO, leaiounalonires, 10) som. OPV SS shishY Towne ltOim,: 3. 92. Sepopo, 10 m. 6. 04. Taka, 10 m. “Very common.” —E. 8. One of these specimens, a female, is bright orange, the others are of the usual dark sooty brown. 8. NycreRIs HISPIDA Schr. (2). 3. 72. Bubi Town, Bantabiri, 500 m. Not satisfactorily determinable in the dried condition. 9. MIMETILLUS MoLONEYI Thos. Bo US Oe Os, OO, WAR rateAloria, IO wi, “Shot on the wing. Flight swift and with many rapid turns and twists.” Mimetinyuus Thos. Mimetillus Thos. Abstr. P. Z.S. 1904, No. 10, p. 12, Nov. 22. Type. Vesperugo (Vesperus) moloneyi Thos. Ann. Mag. N. H. (6) vil. p. 528 (1891). * In the account of the genus Cat. Chir. B. M. p. 80; but in the synopsis of the genera on p. 3 the formula is given by an oversight as P.2, M.2. ? + Cf. P. Biol. Soc. Wash. xv. p. 198 (1902), where, however, the name is accidentally misprinted Nyctimene. 1904. | MAMMALS FROM FERNANDO PO, 189 Distinguished from Vespertilio (=Vesperus) by the abnormal reduction in the size of the wings, which look insufficient to support so large and heavy a body, and by the remarkable breadth and flatness of the skull, which resembles in these respects that of Tylonycteris. Further study convinces me that this curious Bat, which I described from a specimen sent home from Lagos by Sir A. Moloney, should be separated generically from Vespertilio. Its proportions are quite different from those of any other Bat. as is shown by the fact that its forearm is barely half the length of the head and body, its fifth finger barely longer than even this short forearm, and its third finger is only as much longer than the forearm as the fifth usually is. The result is that the develop- ment of the wings recalls that in fetal specimens. The hind limbs are also abnormally short. In the original description the wing-membranes were said to be uniformly brown, but this is a mistake due to the bad condition of the type. Inward of the fifth finger they are brown, but those between the third and fourth and fourth and fifth digits are a transparent whitish, with a few brown spots terminally. The penis is remarkable in that it has no reversible prepuce, the uncovered glans being long, conical, and covered with minute reversed sete. The skull, although larger, recalls that of Tylonycteris pach YPUs by its broad and peculiarly flattened shape. It is not quite so flat, but its anterior portion is even broader in proportion, the anteorbital projections being unusually developed. Sagittal crest practically absent, lambdoid crests strong. No distinct occipital “helmet.” Median palatal spine longer. Owing to its short velvety-brown fur and peculiar proportions, this Bat has a strong superficial resemblance to a Vyctinomus or Molossus rather than to a member of the Vespertilionide. Hence the generic name suggested for it. 10. Crocrpura (Croc.) PoENsIS Fraser. Crocidura poensis Fraser, P. Z.S. 1842, p. 200; Allen & Thomson, Expedition to River Niger, il. p. 481 (1848). ¢. 36. 2.92. Bantabiri, 10 m. 3. 26. Fish Town, 10 m. Q. 140. Bilelipi, 500 m. This Shrew seems to be the same as that afterwards described from Old Calabar by A. Murray, under the name of Rhinomus soricoides, apparently in the belief that it was a rodent. His type specimen, much discoloured, is still in the British Museum. 11. SytyIsoREX JOHNSTONI Dobs. g. 94. 9. 68,103. Bantabiri, 10 m. A separate skull (67). 190 MR. OLDFIELD THOMAS ON [ Nov. 15, SyivrsorEx Thos. Sylvisorex Thos. Abstr. P.Z.S. 1904, No. 10, p. 12, Nov, 22 Type. Crocidura morio Gray. African Shrews with white teeth, four upper unicuspids, normal mandibular dentition, and a short-haired tail without the long bristle-hairs characteristic of Crocidura. In 1887* Dr.G. E. Dobson, when describing from the Cameroons the pigmy Shrew now found by Mr. Seimund in Fernando Po, included it together with Gray’s Crocidura morio in the genus Myosorex, » genus founded for the South-African Sorex varius Smuts. The latter animal, however, is remarkable for the possession of a minute extra tooth in the lower jaw, as discovered and described by Dobson; and this character I think of such importance as to necessitate the species which do not possess it, but are in other respects allied to J/yosorex, having a special generic name. Thisnew genus would include the species S. morio Gray (type), S. gohnstont Dobs., S. sorella Thos., and S. muricauda Mill The four Fernando Po skulls of S. gohnstoni differ considerably in the relative proportions of the upper unicuspids, the second being much smaller than the third in some cases, as it is in the type, while in others it is nearly as large. I am inclined to believe that in the Soricidz generally the systematic importance of the relative sizes of these teeth has been considerably over- estimated. 12. PoIANA RICHARDSONI Gray. Native skin. Bantabiri, 500 m. 13. ANOMALURUS FRASERI Waterh. 3. 58. @. 59,185. Bantabiri, 10 m. 6. 159. N. Bantabiri, 1800 m. Although several Fernando Po specimens of 4, fraseri, including the type, are in the British Museum, all are very much faded by exposure to light, and these fresh topotypes are therefore of much value. Among the specimens assigned to this species from the mainland of Africa are two from the Lower Niger, which a comparison with Mr. Seimund’s examples shows to be subspecifically separable, as follows :— ANOMALURUS FRASERI NIGRENSIS. Anomalurus fraseri nigrensis Thos. Abstr. P. Z.S. 1904, No. 10, p. 12, Nov. 22. Closely similar to the true fraser? in all respects, but the general colour paler and greyer—body broccoli-brown, membranes smoke- grey,—and the size, as shown by the skull and teeth, decidedly smaller, The tail also less bushy. *S JP a Se USS, Ds BB. 1904. | MAMMALS FROM FERNANDO PO, 191 Coloration of head, under surface, and limbs as in true fraseri. Skull smaller and with rather a shorter narrower muzzle than in frasert, the length of the tooth-row decidedly less. Approximate dimensions of the type, measured in skin :— Head and body 330 mm. ; tail 235; hind foot (s. u.) 57. Skull—tip of nasals to back of parietals 53 mm.; zygomatic breadth 38; nasals, length 14:7, greatest breadth anteriorly 7-4 ; interorbital breadth 16; breadth of brain-case 26; palate length from henselion 22°5; diastema 12°5; palatal foramina 6; length of upper cheek-tooth series 11°9; lower jaw, incisor-tip to condyle 37; length of lower tooth-row 13°8. Hab. Abutschi, Lower Niger. Type. B.M. No. 2.11.10.5. Collected February 1902 by Mr. A. Braham. Two specimens, adult and immature. Du Chailluw’s Anomalurus beldeni, from the Gaboon *, con- sidered by Gray and Alston to be a synonym of 4. fraseri, appears to me to be referable rather to the red-backed species commonly known as A. erythronotus M.-Edw. Considering how widely different in colour the two species are, it seems curious that there should have been any doubt on the subject; but Du Chaillu’s description is extremely vague, and it is only from his statement that ‘on the back the hair is tipped with bright rufous, which gives a rufous tinge from behind the ears to the lower third of the body on the median portion to the commencement of the membranes,” that [ am able to express an opimion on the matter. This sentence, however, exactly expresses the dorsal coloration of A. erythronotus, and the locality is approximately the same, while no examples of A. f/raseri have been since recorded from the district. Tf I am right in this identification, the name A. beldeni will have to stand for the red-backed species, as it antedates A, erythronotus by many years. 14. Scrurus srancErt Waterh. Gn Wd, Qe Bs, Lrsw Mopars NO) sei: 3. 73, 80. Bubi Town, Bantabiri, 500 m. 9. 95. Bantabiri, 10 m. Se LOGs IGS Beli SeNliol aiiOM epee aN ebamtalourr, 1800 m. Ge lZGaesilelipi 10m: 15. Scrurus RUFOPRACHIAtTUS Waterh. Ow kon 2e 2h. 34, 37,47. Seeley 22223, 2, +095 » Wish Town, 10 m. 3. 3,4, 6,10, 39. 2. 7,11, 40, 41,42. Santa Isabel, 10 m. CO Ma S2AOiG NAD, VAL NGS me GF ih. 4 2. C9789) 269) Bantabiri, 500 m. 3. 54,139. @. 134, 137. Clarence Mountain, 1800 m, * P. Bost. Soc. N. H. vii. p. 303 (1861), 192 ON MAMMALS FROM FERNANDO PO. { Nov. 15, 16. FUNISCIURUS ERYTHROGENYS Waterh. 36. 29. ©. 14, 16, 20, 38, 43,50. Fish Town, 10 m. dg. 91. @. 52. Lepopo Beach, 10m. Q. 142. Bilelipi, 500 m. 6. 146. N. Bantabiri, 1800 m. The last three species of Squirrel were all described, together with ab : dd=nearly 2 aa. Tn the preclitellar region the gap aw becomes greater and ab rather smaller. Clitellum: this is complete (i. e. girdle-shaped) and well developed over the four segments 14-17, where the interseg- mental furrows are obliterated. The dorsal surfaces of 13 and 18 are also glandular, but the clitellar colouring is not so definite, and, moreover, the grooves 13/14 and 17/18 are deep. Genital pores, kc. (text-fig. 62).—There is a pair of male pores in the 18th segment; each is a small pit,;in a small, oval, pale spot, and from this small pit a little papilla projects up to the level of the body-wall (text-fig, 63). This oval poriferous area is in line with a. Tubercula pubertatis, in the form of small paired glands, are * “Toke” is the Maori for Earthworm. 1904. | EARTHWORMS FROM NEW ZEALAND. 241 present on the hinder margin of segments 16 and 17, lmediad of the line a. The two oviducal pores are close to the median line, near the anterior margin of the 14th segment. Text-fig. 62. ae A oo ——————— saa 2 a =e im | ih fa) Text-fig. 62. Tokea esculenta.—Ventral view of clitellar region, &c. (4), showing male pores and arrangement of tubercula pubertatis and disposition of chate. This and the views in text-figs. 68, 70, 73, 76, 78, and 80 are somewhat diagram- matic in that they are represented as flat projections, but the relative spacing of the cheete, &c., is correct. Text-fig. 63. Tokea esculenta.—Enlarged view of the male pore, the actual aperture is situated on a small papilla (C) which projects from the bottom of a pit (B), the skin round which is paler (A) than the surroundings. The cheta 6 is shown on the left side. There are three pairs of spermathecal pores on the hinder region of segments 6, 7, and 8; these are not intersegmental, but are situated about midway between cheta @ and the margin, in each case. I could not detect dorsal pores. Internal Anatomy. There are 8 stout septa, behind segments 7 to 14; but, in com- parison with the thickness of the body-wall, their thickness is not so great as one would expect. The dorsal vessel is single; the last of the four pairs of hearts is in segment 13. In the heart-segments there is a supra-enteric vessel, with which the hearts are in part connected. The worm is micronephric*, and these organs commence in segment 3. In the last 20 segments of the body there is, in addition to micronephridia, a pair of compact groups of tubules * | think this term, which was first employed by Vejdovsky, is preferable to my term “ plectonephric,” for it is by no means certain that a “network ” of tubules always (if ever) exists. Proc. Zoot. Soc.—1904, Vou. I]. No. XVI. 16 242 DR. W. B, BENHAM ON [ Nov. 15, constituting a meganephridium. In these the nepridial funnel is present and can be traced into communication with the nephridium, which opens to the exterior, probably in line with a, as it is in another species which was more carefully studied in this respect. I did not trace it in the present species. Not only is there this meganephric nephridiostome in these hinder segments, but a similar funnel is present throughout the worm. In the micronephric segments it is unconnected with the nephridium and has no external opening. Text-fig. 64. Tokea esculenta.—A somewhat diagrammatic drawing of a dissection of the worm, showing the tongue-shaped form and sub-enteric position of the prostates characteristic of the genus; the intestine is cut away exposing the glands; the ventral vessel and nerve-cord remaining éz sitw. INT., mtestine; D.V., dorsal blood-vessel; V.V., ventral blood-vessel; N.C., nerve-cord. The presence of a funnel, independent of the micronephridia, throughout the worm is of very considerable interest. Such an arrangement has not hitherto been recorded. The funnel is of the same size and structure as that of the meganephridia, and both in its large size and peculiar form differs from any funnel hitherto figured. In each of the species of Yokea the same general arrangement occurs, though the details as to character and arrangement and extent of the micronephridial tufts differ in each case. I am preparing a detailed account of the excretory 1904. | EARTHWORMS FROM NEW ZEALAND. 243 apparatus of these worms, so that I will not further describe it here. Alimentary system.—The gizzard is small, rounded, and hidden by the extrinsic muscles of the pharynx; it lies in segment 5. There are no definite cesophageal glands, but in segment 15 the tube is dilated, and its dark vascular wall contrasts with the paler wall of the neighbouring region, and here its lining is thrown into a series of horizontal lamelle. The intestine commences in segment 16; there is no typhlosole. Reproductive system.—The testes and funnels are free and, like the ovaries, lie in the usual segments. The two pairs of sperm-sacs lie in segments 9 and 12: their wall is smooth. Further, on the anterior wall of segment 13 is a minute curved sac, close to the gut, whose curvature it follows; it is of about the same size as a similar sac in the 14th segment, which J take to be the ovisac. Text-fig. 65. Tokea esculenta.—An asymmetrical arrangement of the prostates, such as occa- sionally occurs in some species of the genus; V.V., ventral blood-vessel. Spencer has noted in some of the Cryptodrilids studied by him such an extra sac in the 13th segment, which he regards asa sperm- sac. I could not discover any developing sperms by teasing this sac on the slide, and I have not yet studied it in sections. I have met with similar structures in some other species. 16* 244 DR. W. B. BENHAM ON [ Nov. 15, The prostates have a form and disposition on the body which appear to be peculiar and characteristic of the genus. Hach prostate is a long “‘tongue-shaped” gland (text-fig. 64, p. 242), i. é., though it is tubular in that its axis is hollow, it is not cylin- drical but somewhat flattened, extending through several segments, and its apex is usually recurved. The pair of glands, typically, he side by side, pressed against the body-wall below the gut, and indeed below the ventral blood-vessel; but in individual cases the right and left glands are asymmetrically disposed. For instance, in one case (text-fig. 65, p. 243), the right gland passes backwards from its duct to the 24th segment, and then bends for- wards on the left side, and its apex lies in the 21st segment; whereas the left gland passes forwards into the 16th segment, then curves round to the right side, passes at first backwards, and then obliquely to the left side, so that its apex lies in segment 20. Text-fig. 66. Text-fig. 67. Text-fig. 66. Tokea esculenta.—A diagram, compiled from serial sections, showing the course of the sperm-ducts, their union, and junction with the canal of the prostate, well within the region of the glandular tissue. PR., prostate; P.R.L., its lumen; PR.D., its duct ; S, septum; V.D., vas deferens. Text-fig. 67. Tokea esculenta.—Spermatheca (X12. Camera outline of mounted specimen). The surface of the gland is smooth; it is constricted by the successive septa, and at its anterior end gives rise to a muscular duct which passes to the exterior in segment 18. Sections 1904. | EARTHWORMS FROM NEW ZEALAND 245 through this region of the body show that the two sperm-ducts run back separately along the body-wall as far as the 18th segment, when, meeting the prostate duct, they pass on to its dorsal surface into the 19th segment, and after uniting with one another open into the canal of the prostate gland some distance from the commencement of the muscular duct (text-fig. 66). The prostate is traversed by a canal lined with columnar cells ; at intervals this canal receives small canalicules, around which the gland-cells are grouped and into which they open. The arrangement is similar to that described by Miss Sweet for Plutellus intermedius (loc. cit. fig. 17). There are no penial cheetze. There are three pairs of spermathece in segments 7, 8, and 9; each (text-fig. 67) is an ovoid sac passing gradually into a short wide duct, which receives the diverticulum close to the body-wall. The diverticulum is a short ovoid body, with duct; it is about 3 the length of the main sac. The diverticulum of this and the following species is placed mediad of the sac. 2. TOKEA SAPIDA, Sp. n. (Text-figs. 68 & 69.) This species is founded on a single specimen of about twice the size of the former species, but included with it as “‘ Kurekure ” by Mr. Best. Text-fig. 69. Text-fig. 68. Text-fig. 68. Tokea sapida, similar view to that in text-fig. 62. (x 6.) Text-fig. 69. Tokea sapida.—Spermatheca (x 12. Camera outline of mounted specimen). In colour it is purple-red, with an imperfectly developed clitellum of a deeper red than the rest of the body. Dimensions. 200 x 8 mm., with 190 segments. The pre- clitellar segments are biannulate. 246 DR. W. B. BENHAM ON [ Nov. 15, Cheice. ‘These are smaller than in 7”. esculenta, and have a different arrangement; for, when viewed from above, d is more laterally placed and ¢ is not visible. In other words, the dorsal gap (dd) is much greater. The formula is:—ab=ed; be>ab; aa=14 ab; dd=5 ab. I cannot detect any cheete in front of the 10th segment on examination by a dissecting-lens. Clitellum: not yet fully developed, as intersegmental grooves are still unobliterated, though the difference in colour is well marked. It is “complete,” and covers the five segments 14-18. Genital pores, dc. (text-fig. 68, p. 245).—On segment 18 is a pale, tumid, transversely disposed ridge, somewhat enlarged at each end : it extends from 6-6, and cheta «a appears to be absent. The actual male pore is uncinate, and on the mesial side of 6, but close to it. No tubercula pubertatis ave present. The three pairs of sperma- thecal pores are at 6/7, 7/8, 8/9, in line with the gap ab. Internal Anatony. There are eight stout septa, behind segments 6-13, but the first and last are less stout than the others. The last heart is in segment 13. The gizzard, in segment 5, is of large size; the cesophagus presents no dilatation. The nephridia have the same general arrangement as in T. esculenta. The genital organs agiee on the whole with the latter species ; but i the single individual in my possession the prostates are asymmetrically arranged, viz., the left gland extends back to segment 23, and then bends forwards and ends in the 23rd segment. But the right gland is bent in an $-shaped manner, and lies wholly im segments 18, 19, and 20; but, as in the preceding species, both are pressed against the body-wall. The spermathece, in segments 7, 8, 9, differ in shape from those of 7’. esculenta; the diverticulum (text- fig. 69, p. 245) being globular, with a short duct, opening into the spermathecal duct close to the body-wall. Loc. Ruatahuua, North Island, New Zealand. 3. TOKEA UREWER&, sp.n. (Text-figs. 70-72.) This worm is known to the natives as “ Pokotea,” and is described by Mr. Best as “a short white worm.” TI received three specimens, which in formol are quite a pale pinkish brown, paler still behind the clitellum, which is orange or orange-brown, with a dusky anterior margin. Dimensions. 65 to 80 mm. in length, with a diameter of 7 or 8 mm.; there are 78 segments in the larger individual. The segments are not TRTEIEUO The prostomium is 4 epilobic ; without a transverse furrow. The chete have the same general arrangement as in the 1904. | EARTHWORMS FROM NEW ZBALAND, 247 preceding species; when viewed from above d is dorsally and ¢ laterally placed, but in the tail the line ¢ approaches d. aa=ab=cd; be>ab; dd>be. But one differences, measured in millimetres, are very slight: thus aa=2°5; be=3; dd=4. The cliéellum is well marked, complete, and covers the four segments 14-17; but the ventral surface of the last segment 1s less glandular than the rest. Banta pores, de. (text-fig. 70).—On segment 18 is a pair of subcircular depressions, with a distinct, slightly raised margin ; this is in line a. Text-fig. 70. Tokea wrewere, sunilar view to that im text-fig. 62. (x 6.) There is asingle, median tuberculum pubertatis, in the form of an oval, glandular, depressed area, extending from a—a, between segments 19/20. The three pairs of spermathecal pores are in line a. Internal Anatomy. The seven septa behind segments 6 to 12 are only slightly thicker than the following. The last heart is in segment 12. The micronephridia are more delicate than in the preceding species, and do not form so dense a covering to the body-wall. The meganephridia are confined to the ee 12 to 15 segments. The gizzard (in segment 5) is longer than in 7. agen a, though not so wide. The cesophagus is more or less dilated in segments 11-15; and in the first two of these the thick vascular wall is provided with Jamelle internally. The prostates extend back to the 26th segment; the muscular duct is long, narrow, and more or less undulating, becoming thicker where it dips into the body-wall. 248 DR. W. B. BENHAM ON [ Nov. 15, The three pairs of spermathece ave in segments 7, 8, 9; each sac is ovoid (text-fig. 72), and the duct is much narrower than the sac; the diverticulum is small, and also has a narrow duct. Loc. Ruatahuua, North Island, New Zealand. Text-fig. 71. N.C. Text-fig. 71. Tokea urewere.—View of the anterior end of the prostate, showing the long, narrow and undulating duct. N.C., nerve-cord. Text-fig. 72. Tokea urewere.—Spermatheca (Xx 12. Camera outline of mounted specimen). 4, ToKEA HUTTONI, sp. n. (Text-figs. 73-75.) A single individual collected by Capt. Hutton; it is much bleached, but was apparently purplish. Text-fig. 73. Age Bs wile Tokea huttoni, similar view to that in text-fig. 62. (x 9.) Dimensions. 80X5 mm., with 63 segments; but possibly imperfect *. Prostomium 3 epilobic; segments of body not annulated. * On opening the tail, I find no meganephridia ; perhaps the true hinder end had been severed and the wound healed. ‘The lining of the intestine was continuous with body-wall round anus, but this region was thin. “ 1904. ] EARTHWORMS FROM NEW ZEALAND. 249 Chete : when viewed from above, both ¢ and d are on the dorsal surface, and ¢ is well above the lateral margin. aa=be=cd; abbe. The clitellum is fully developed, complete, and includes seg- ments 14 to 17; but both the 13th and 18th segments are glandular on the dorsal surface, the furrows 13/14 and 17/18 are, however, quite deep, whereas the intervening ones are obliterated. Genital pores, tc. (text-fig. 73)—In segment 18 is a slight circular dépression with raised margin, in line with the gap ab; the cheeta @ appears to be absent. There is a single median, oval tuberculum pubertatis on the hinder margin of the 18th segment, extending from a—a. The three pairs of spermathecal pores are on the hinder margins of segments 6, 7, 8, in the gap ab, though nearer to @ than to b. Internal Anatomy. There are seven thickish septa behind segments 7-13. The last heart is in segment 12. Text-fig. 74. Text-fig. 75. Text-fig. 74. Tokea huttoni—View of the prostates, which, though symmetrical, have assumed a reversed position (? abnormal), with the anterior end directed forwards. N.C., nerve-cord. ? Text-fig. 75. Tokea huttoni—Spermatheca (X12. Camera outline of mounted specimen). The gizzard is long, narrow, and the wall rather thin. The micronephridia form a pretty close felt over the body-wall, and commence (as in other species) in segment 3. The prostates (text-fig. 74) are shorter than in either of the preceding species, and both are directed forwards, but lie side by 250 DR. W. B. BENVIAM ON [Nov. 15, side below the gut. They only extend through the three segments Toran 16: The duct is very short and thick. s The spermatheca (text-fig. 75, p. 249) likewise has a short duct, and the diverticulum is similar to that of the foregoing species. Loc. Whangarei, North Island, New Zealand. Remarks.—Vhis species is evidently nearly allied to 7’. urewere, but differs from it in the relative position of the male and sperma- thecal pores, in the position of the tubercula pubertatis, m the cheetal formula, and especially in the relation of ab to be, as well as in the less extent of the prostate and its duct. 5. TOKBA SUTERI, sp. n. (‘Text-figs. 76, 77.) A single individual collected by Mr. H. Suter in 1899. It differs in general appearance from any of the preceding species, not only in size, but in colour. It is dark purplish brown throughout the entire dorsal surface ; the clitellum is reddish purple. Dimensions. 50X5 mm.; with 86 segments. Prostomium + epilobic; without transverse groove. The chete are practically equidistant, 1.e. aa=ab=be=dd = 2 mm., while cd is only 1°75 mm. Text-fig. 76. Text-fig. 77. A. B. Text-fig. 76. Tokea suteri, similar view to that im text-fig. 62. (x 9.) Text-fig. 77. Tokea sutert.—Spermatheca (X 12. Camera outline of mounted specimen). On the clitellum the gap aa is a little less than posteriorly, and in the preclitellar region still less; the gap ab is less in the _ eclitellar region, but not anteriorly. But these differences are not perceptible till measured with compasses. Clitellum in segments 13-17 (=5 segments), girdle-like. Genital pores, ke. (text-fig. 76).—A pair of small papille on segment 18, in the line of a. There is a pair of small twbhereula pubertatis—subarcular and pitted—between 18/29 in line a, and extending further mediad thereof than do the porophores. é 1904. | EARTHWORMS FROM NEW ZEALAND. 2 Ril 1 The three pairs of spermathecal pores are in line a, at 6/7, 7/8, 8/9. Internal Anatomy. There are five slightly thickened septa, behind segments 8 to 12. The last heart is in segment 12. The gizzard is feebly developed; its wall is not thicker, though tougher, than that of the cesophagus. The prostates are quite typical, and extend back to segment 25, with tips recurved ; the duct is short and narrow. The spermatheca (text-fig. 77) has a long duct, with a long narrow diverticulum opening into it at the body-wall, Locality. Auckland, New Zealand. Remorks.—Anatomically there is a rather close resemblance between this and the two preceding species; but in the coloration and dimensions it is remarkably distinct, while the practical equidistance of all the 8 chetie, the paired tubercula pubertatis, and the details as to the prostate and septa, mark it off as distinct. 6. ToKEA KIRKI, sp. n. (Text-figs. 78, 79.) Five specimens collected by Professor H. B. Kirk, in 1902, are about the same size and colour as 7’. esculenta, 7. e. reddish purple, but rather paler than the latter; the pigment extends further round to the ventral surface than is usually the case, and ceases about the level of 4. The clitellum is brownish red. Text-fig. 78. Text-fig. 79. pie war iG Text-fig. 78. Tokea kirki, similar view to that in text-fig. 62. (x 7.) Text-fig. 79. Tokea kirki.—Spermatheca (X 12. Camera outline of mounted specimen). This has no muscular duct distinct from the sac. Dimensions. The length varies from 80 to 100 mm., with a diameter of 6 mm. for 110 segments in the larger individuals. The chete are practically equidistant ; but on measurement it is found that ab=be=cd (=1°75 mm.), while aa=dd=2 mm. 252 DR. W. B. BENHAM ON [ Noy. 15 When viewed from above, d is dorsally placed and ¢ rather above the lateral margin. The elitellum is not fully developed on any of the specimens, as the intersegmental furrows are distinct ; but the characteristic coloration extends over the five segments 13 to 17. Genital pores, dc. (text-fig. 78, p. 251)—The male pores are situated in what appears to be a pair of tubercula pubertatis on segment 18; there is a very large, broad, subcircular papilla on each side, extending from nearly @ to c; this carries a somewhat quadrangular depression surrounded by a distinct rim or margin ; this depression is glandular and has all the characters of a tuber- culum pubertatis (as seen in section) ; and the male pore is quite small and situated in the outer edge of this gland, just within the margin, 2. e. nearly in line 6. The two oviducal pores are situated close together, in a pale area, near the anterior margin of segment 14. There are only two pairs of spermathecal pores, situated at 7/8, 8/9. Internal Anatomy. There are no specially thick septa. The last heart is in segment 12. The gizzard is quite small; the esophagus is very much dilated in segment 14, but there is no constriction separating it from the coma < ; its lining, however, is thrown into a series of horizontal, lamelliform folds. The micronephridia, instead of being spread over a great part of the body-wall, in each segment, form a very distinct and narrow row close to the septa. The prostates extend from the 18th to 23rd segments, and each is provided with a narrow duct. There are two pairs of spermathece in segments 8, 9; each (text-fig. 79) is a somewhat pyriform sac, without a distinctly marked duct; the diverticulum is of the same shape, and also without definite duct. Locality. Ohaeawai, North Island, New Zealand. Remarks.—A very distinct species, characterised both by the remarkable coincidence of male pore and copulatory tubercle, and by the possession of only two pairs of spermathece. 7. ToKEA MAoRICA, sp. n. (Text-figs. 80-82.) I have eight specimens of this species, collected at different. times by Mr. H. Suter, at different places in the neighbourhood of Auckland. The colour is a dark purplish brown, more brown than purple in these alcohol specimens, and resembles that of T. suteri. The clitellum is paler brown ; the chetz are inserted in white spots, as in Plagiocheta sylvestris, to which also they bear a resemblance both in colour and their short, stout form ; indeed, from a hasty first glance, I had placed them with the latter genus, awaiting examination. 1904. | EARTHWORMS FROM NEW ZEALAND, 253 Dimensions. The largest specimen measures 70x 4 mm., and consists of 75 segments; the shortest mature individual measures 25x 2mm., and likewise consists of 75 segments; and the same number occurs in a worm of intermediate size. We have here an instance—unique, so far as I am aware—of an earthworm increasing in length without adding new segments posteriorly. It is true that we do not know much on this subject ; but the accepted view is that the number of segments is continu- ally added to during growth. With regard to this, Beddard (95, p. 2) states :—‘ There are at present no exact data as to the constancy of the number of seoements among Earthworms. In all probability the number is not absolutely fixed, but there appears to be a mean for each species round which there is a certain amount of variation.” Text-fig. 80. ives B. Tokea maorica, similay view to that in text-fig. 62. (X 11.) Prostomiwm i epilobic; without a transverse groove. Chetal formula :—ab