ceestita Soe Near Se ae ewig 0 ww EE egin ve AN lh wih (tn ain 4 ea hts ty EA TO Ae KR AEE Nim Sehincites tree et Specks eo RT sik Tra Ea! et Rieti Sa Fite gt ietiatie ye t-5e when's ay = sagt Se os aber atin et — Ed etn oO ite awn tan 2 RS we AS letra crab: Reg BAGR ST ME. BE m2 Ae tr een ah 4 oat > yeet ee * a Sp SRD HOR AR he Bh ea NS artis Te Secs? oR > ae Ee RO Goat = . es eee Aer oe SS ee ee ee —— - m MRE Ae RINT AS OD RS OO a: wee S - ago Daw as Rat etn CRA Ce ae Sa ee Sa a kScwk doa be wane : - ‘ ‘ aS ert rf : : Sor eckoe ee = eterna ee ee - my eee ae 4 ¥ na. Faerie Ae ERY Oe ‘ gp ; S ese ¥- per ener ee ST OLS aide Sea PROCEEDINGS OF THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1905, vol. II (MA Y—DECEMBER.) PRINTED FOR THE SOCIETY, AND SOLD AT THEIR HOUSE IN HANOVER-SQUARE, LONDON: MESSRS. LONGMANS, GREEN, AND CO, PATERNOSTER ROW, Ow onian institu, Eeiise OF THE COUNCIL AND OFFICERS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1905. COUNCIL. His Grace Tue DuKE or Beprorp, K.G., President. Sir ALEXANDER Batrrp, Br. Grorce A. BouLencer, Hsq., | E.R.S., Vice-President. Tuomas H. Burroveues, Hsq. Freperic G. Dawtrey Drewirt, Esq., M.D. Hersert Drucs, Ksq., F.L.8., Vice-President. Cuartes Drummonp, Ksq., Treasurer. Sir Epwarp Duranp, Br., C.B. FREDERICK GiLLeTT, Esq. F. DuCanr Gopmay, D.C.L.,F.R.S., Vice-President. W. R. Octuvir-Grant, Esq. Hsa., | JOSEPH JACKSON ListER, Hsq., M.A., F.R.S. Sir Epmunp Gitrs Loprr, Br. EK. G. B. Meapr-Watpo, Esq. P. Cuatmers MircHey, Ksq., M.A., D.8c., Secretary. E. Lort Puitures, Esq. Howarp Saunpers, Ksq., Vice- President. H.S.H. Prince TECK. CuHarues 8. Tomus, Hsq., M.A., F.R.S., Vice-President. Aveustus F. Wiener, Esq. Henry Woopwarp, Esq., LL.D., E.R.S., Vice-President. FRANCIS OF PRINCIPAL OFFICERS. P. Cuaumers Mircuenr, Hsq., M.A., D.Sc., Secretary. Frank H. Bepparp, Hsq., M.A., F.R.S8., Prosector. R. I. Pococgr, Hsq., F.L.S., Superintendent of the Gardens. CHARLES GABRIEL SELIGMANN, Hsq., M.R.C.S., L.R.C.P., Pathologist. Me. F. H. Wateruouss, Librarian. Mr. Joun Barrow, Accountant. Mr. W. H. Cots, Chief Clerk. Mr. Grorce ArtHur Doupiepay, Clerk of Publications. Mr. ArtHur THomson, Assisiant Superintendent of the Gardens. Y, LIST OF CONTENTS. May 2, 1905. The Secretary. Exhibition of photographs of Hippo- [DOWATIMNEISSS) ShyamoMNATTAMIe? So Gavide acnonoaseanne sac ve odunooasraqtTee Mr. R. KE. Holding. Exhibition of, and remarks upon, a series of the first-year antlers of certain Deer ............ My. R. I. Pocock, F.Z.8. Exhibition of a specimen of the Syoeiouteln “beneryoniblless cocoeboncosendnostcopec: Nt SABE eee aS Mr. W. Bateson, F.R.S. Exhibition of, and remarks upon, specimens of Fowls illustrating peculiarities in the hereditymolewhitepplumaceyy. sp see chen eee enon acne 1. On the Sponge Leucosolemia contorta Bowerbank, A scandra contorta Haeckel, and Ascetia spinosa Lendenfeld. By HK. A. Mincnin, F.Z.8., University College, London. (CER Tetb ea) eai-van ear: veh Se charts i ceotte Mpsaldalo Ss \sniomimphign ae ees 2. Some Notes upon the Anatomy of the Ferret-Badger, Helictis personata. By Franx HE. Bepparp, M.A., TD Bide LesROEAONe AO) HOEY SOXGENY deonanadonbocososesqosonconese 3. Contributions to the Osteology of Birds.—Part VII. ELurylenude ; with Remarks on the Systematic Position of the Group. By W. P. Pycrart, F.Z.8., M.B.O.U. (Gedy IES Ne Sane PPA ae ames ee sade Rela Rees Serie crt ue Page 21 iV May 16, 1905. The Secretary. Report on the Additions to the Society’s Menagerie during the month of April 1905 ............... Mr. Oldfield Thomas, F.R.S. Description of a new Golden Mole (Amblysomus corric) from Cape Colony eeoeeteecore Mr. H. B. Fantham, F.Z.S. Exhibition of microscopic slides and description of a new Sporozoon, Lankesterella ETUC OTIS oe otis s ct SIC EE RRR PE ae Eka en oe ede 1. A Contribution to the Knowledge of the Encephalie Arterial System in Sauropsida. By Frank E. Bepparp, M.A., F.R.S., Prosector to the Society 2. On the Nomenclature of the Anthropoid Apes as pro- posed by the Hon. Walter Rothschild. By Sir H. H. JOHNSTON, G.C.M.G., K.C.B., E.Z.S. Seer ese re eee ase eeeseesve (su) . On some Bats of the Genus Rhinolophus, with Remarks on their Mutual Affinities, and Descriptions of Twenty- six new Forms. By Kyup Anpersen. (Plates ITT. & TV.) sRelleishese XVII. Figs. 7-15. Palemon (Eupalemon) lar, var. Figs. 16-19, +537 ula CLA DUCUEINOM)) pUAMe: raleread cca s caked Heron eee ) XIX. a BES OR CaHA ee eae a ae cn eee 564 — Dare ww LIST OF TEXT-FIGURES 1905.—Vot. II. lirst-year antlers of certain species of Deer...... sede blo Spicules of a specimen of Clathrina contorta from Roscoft Spicules of two specimens of Clathrina contor/a {rom Banyuls. . Spicules of the ‘ spinosa” variety of Clathrina contorta ...... Spicules of Leucosolenia, Sycon, and Clathrind 1... ccs. cece es Abnormal gigantic spicules of the class of the monaxons from a specimen of Clathrinu contorta from Banyuls. . . CC ee et Brain of Helictis personata, dorsal aspect ...............00 Brain of Helictis personata, ventral aspect, with the arterial system shown in thicker cit thinner black lines ......... Pancreas and adjacent regions in Helictis personata .......... . Pancreas and adjacent region in Galictis barbata ..........4. . Intrathoracic aorta of A. Helctis personata; B. Galictis TOURS ee 006 Breese shaeans ane: Mera Uhre diego Intrathoracic aorta of Suricata tetr adect “Ale done ence ok ie Sternum of Calyptomena, showing the Sines (animmcnted) SPUMAROX LORNA OLE. ana. crater Restrain bapa se aionrces te ut 219100 01d OND Portion of the shoulder-girdle of Ca branes Maton atin vaults : . Dissection of arm, dorsal aspect, of Hwrylemus ochromelas .... Eumeces algeriensis. Ventral aspect of brain, showing chief EL OTIOS 0015 sesamin SH OCU COLAO THOODE OKs Cb dn ae : Gerrhosauus flavigularis. ‘enim aspect of brain, showin ing Chietfarteniesma canmcicrriar a5 SRN RCE annie eda epee Tupinambis nigropunctatus. V anne aspect of brain, esiendth ing Chiehanteniesia see een Sasa ema Mss he FERRO HONS He . Python molurus. Ventral aspect of brain, owine chief arteries. . Testudo vicina. Ventral aspect of brain, clowdass chief arteries . . Testudo vicina. Lateral aspect of brain, showing chief arteries . Side views of nose-leaves of Bats, showing the principal forms of the connecting process in the Rhinolophus simplex group (a) aavel Waa Jia, Lemans eiamnyy (O, GC!) oovhdacednoodondasoons 2M XXIV Page . Wing-structure of Chionts alba ........ 0. sees seen ee eee 156 24. Diagram of intestinal pattern of Rhynchea capensis .......... 157 25, Diavram of intestinal pattern of Hydrophastanus chivurgus .... 158 . Shoulder-muscles of Gtdicnemus scolopav ....... eee veen, 159 pBicepsior (OA7a7ts) a 10c: iterate elke ffenestr 161 DS, Wine ners or Claws GL cooacecnnogcndcen00gs2055 05% 166 } Wileyniart MIGROS aA s eo aso ucokoo cud oes eweBaae OD 3 ea ee 193 . Diagram of the distribution of 250 Mexican species according to WA WRAVECTEN Cn toe cene a eR RENS eSNG tio BLO: O DIG ERO ERS PNP se 39:9 020 "0 229 . Diagrams to illustrate the contours of Mexico at different geo- ios Gell: SYS as ota, cal erecinoip.2.ckaie sean Mibtbhele ohare ReO oe Acre o's 235 2, Diagrams to illustrate the contours of Mexico at diferent geo- logical CEC ah cient tan cen as DIA ee GSS MTOM SO Octo > 236 . Liver of Gerrhosaurus flavigularis, ventral aspect ............ 259 . Liver of a second example of Gerrhosaurus flavigularis, ventral AGUOGE cooocrongooonaoracaopgnsee000b000 aianeacsasekaroteeaene 259 . Gastro-hepatic ligament of Gerrhosaurus flavigularis, showing COUMNS Ort wemsomllaye lyme, co ceanAcancaccucascasdoauover 261 36. Pancreas of Lacerta ocellata and of Gerrhosaurus flavigularis .. 262 . Aortic arches and first part of dorsal aorta of Gerrhesaurus LORUTOUNDIOS “a obavccseosecbaasonsaasasccanguvaeconacoar 268 . Abdominal region of aorta of Gerrhosaurus flavigularis ...... 264 . Lateral view of brain of Varanus exanthematicus and of Tupi- JOCHMOS TOU SUIVOKDUIS. ga got ecg egoeoosoooebensanouaene 268 . Dorsal view cf brain of Varanus evanthematicus ............ 268 , Whe ka OP AlGDMOS COUOMUB soo 00cncscedcossbccoucecadan 981 2), hs jalan Grvslenmays ooo asdpockoaheacoosnnaso ses boo 0G0- 282 3. A section of a portion of the placenta of Acomys ............ 983 . A section through the junction between trophoblast and tropho- sponpia of Acomys ......... 220 eee eee eee eee ee 284 . A diagram of a section through the centre of the placenta of ACT Gs o one 000 poganeon an sorb de soaedsassaaenoncodeoe 285 . Supposed clavicula of Diplodocus ..................... 0005 290 , Seine Ione Git IDG WOCOGNSs onc osc on ns codon cs edooecsopodbor BOO . Diagram of penis of Struthio ..... 66. eee eee eee eee 291 . Os penis of European Otter ................. 0. seen eee 292 5), Wine Sages cadhens OF Uwlhyms 0. sb sccooocanseaassneccsoccs 299 51. Early figure of Chimpanzee, from Astley’s ‘Travels’ ........ 301 52. A, lateral, and B, upper view of the skull of Vulpes ferrilatus.. 304 53. A, upper, and B, lower right molar series of Mecrotus waltoni.. 307 Rul, Wing lBlaelke Cineneves (COMO SHUICUS) sob ehoccoeenct dns ses6e 325 55. The Mantled Guereza (Colobus palliatus) ..................5. 326 56. Sharpe’s Guereza (Colobus sharpet) ..... 1.60... eee eee eee, 527 57, The White-tailed Guereza (Colobus caudatus) ........6.++.. 327 58. The White-thighed Guereza (Colobus vellerosus) ..........45 328 59, Principal arteries of Hatteria ..........----+++2++s+2 sees 463 60. Anterior abdominal and portal veins of Hatteria ............ 465 61, Heart and aortic trunks of Ophisaurus, to illustrate mode of OMIM, Oh CAMOUCL A bebgduoe cvcodcodadaceanccannecansaac 469 XXV Page 62. Intestinal arteries and portal of Ophisawrus, from left side ATL Gaueenalearteries) of Opnisarncsmam ead tyes. al eter rats) =e . 478 64, Hepatic portal system of Ophisaurus ........... 2550 ee Sede ACO 65. Renal veims! of Ophisaurus . 1.02.1 ee ee ee ee eee 47 66. Membranes uniting vena cava and left lung in Amphisbena UVR & 0 bho 0 Oke DH Denes CONMe dG woo OG ODOUUpOOCooN 481 67. Liver of Amphisbena brasiliana, ventral view ..........+++- 485 68. Kidney, testis, and intervening veins of Amphisbena brasiliana, 486 69, Origin of subclavian in Amphisbena brasiliand 1.1... 6.0. eee 486 70, Tail-vertebree and regenerated appendix of a species of Central EXSY JUNCTIONS 0B Sneed ode'9 00.0.8 Hib oo olla ole Go OcleIDIb 0.4 493 71. Tail of a species of Graphiurus from Fernando Po showing regenerated appendix ........... see eee eee esters ees : 493 72, Melia tessellata from Mauritius, holding an actinian in oucla Claw wari ers fe eenitaictieets eee rihchat ation aaetare cet afcomel ca eet wo. 495 73. a. Melia tessellata from the Maldive and Laceadive puidefipalie- goes, bearing in each claw a sea-anemone, 0. The “ hand” Joollionayer hil Byateaalons; Gh oeooconnnodoac Be al hen ROE 496 74, Melia tessellata from the Hawaiian Islands, bearing an expanded actinian in each claw ...... Sie he toa a Gnionn pions Liao anots Crete 498 75. Claw of Melia tessellata showing the two rows of spines ...... 500 76. Melia tessellata dislodging a feed actinian by means of its first mmo wa cory: Ianb ie ey telesales SRSA Eee PLU ancien 3 mrt 502 77. Téte de Morenoa orizabensis, en dessus et de profil.,.....-... , 618 78. Ventral view of Pontodrilus crosslandt......60. 11sec eens soo WSO) 79. Ventral view of Pontodrilus laccadivensis ........+-++.eeveees 560 80. Lycosa tasmanica ........ DD a Bean vcs Rnmestrtethey cme pee Gn cog 572 Sil, Jiyaose HN Re oa Ae poo yous Orage dope 5 yaa t 574 G2) Eycosa) MOlNeULD neo ee ee se eee oe 575 SB, JUNCOST COSMMNAD .whbs took one deaccacad . O77 Sul. IGNGOSE GiwkS gon odoovabardvonos daa coeeson soot oos conc Doe 579 Sa, IMGT WOOP papas encboces gates cobhenoactogo Non Donen lak 581 SO, JONCICU GUGUE Baannbneeds sees 008000 Ans > Oud oC Omu a anCD 6 582 Gils ILpICOSe GHATS Vo one bbs be ce moce ss eroor uuu ouoDCuon oan oOr 085 SSN cosa arenanis no esaaee- esses esos Sdver seas oso nso do 587 SOs eMolamedes Wabtusinnn 1 saute aul an > ys ae CY Ren CCRC Ea 588 Proc. Zoou. Soc.—1905, Vou. II. LIST OF NEW GENERIC TERMS PROPOSED IN THE PRESENT VOLUME (1905, vol. I1.). Page Dryomys (Mamm.) ............... 348 EHurynannus (Hemipt.) ......... 153 Glirulus (Mamm.) ............... 347 Page Hedonistes (Coleopt.) ............ 278 | Levenna (Hemipt.) ............... 151 | Morenoa (Ophid.) ......... 517, 518 — PROCEEDINGS OF THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1905, vol. IL. PART I. CONTAINING PAPERS READ IN MAY ann JUNE. OCTOBER 1905. PRINTED FOR THE SOCIETY, SOLD AT THEIR HOUSE IN HANOVER SQUARE. LONDON : MESSRS. LONGMANS, GREEN, AND CoO., PATERNOSTER-ROW. a [Price Twelve Shillings} _ 5S lS OF 90 ON aN ES: 1905.—Vot. IT. Parr I. May 2, 1905. The Secretary. Exhibition of photographs of SEUNG Swindining’ st... |. ana Mr. R. E. Holding. Exhibition of, and remarks upon, a series of the first-year ‘riggs of adil ID Se Sad Gas OOP RIGNe Oconto 6 Man ms SoA OG MOU b Oeics Sr bhd ae CaN oie: ~ Mr. R. I. Pocock, F.Z.S8. Exhibition of a specimen of the Spanish Tarantula Mr. W. Bateson, F.R.S. Exhibition of, and remarks upon, specimens of Fowls illustrating peculiarities in the heredity of white plumage ..+......---.. ec eee cece eee tee i On the Sponge Leucosolenia contorta Bowerbank, Ascandra contorta Haeckel, and Ascetia spinosa Lendenfeld. By E. A. Mincuiy, F.Z.S8., University mae London. (Plate L.) cee cece cee eee ee cee tere eee e ee cece ee eee etree nets ensue ee [2x] 2, Some Notes upon the Anatomy of the Ferret-Badger, Helictis personata. ae Reine K. BHDDARD, MuAT HORS, eb rosector tO tae SOCIeby, cr. )cive No a Sunita cre 'enelis Weve lelsuaiereiaiare ti 3. Contributions to the Osteology of Birds.— Part VII. Hurylemide ; with Remarks on the Systematic Position of the ne, By W. P. Pycrart, E.ZS., M.B.0.U. - 7 TSES GU, Vy a RRO ani neh ATT Marner eh OU CA beta UNSEEN A Ba May 16, 1905. The Secretary. Report'on the Additions to the Society’s Menagerie during the month of iparp PM Getic af teks al ere bbeifare ju 6 Oak cue tere Ste Ne PSUR Ce es ciet bio egate’' «ta! « s 9 Mr. Oldfield Thomas, F.R.S. Description of a new Golden Mole (Amblysomus corrie) Hyori ibl CFs 0\2. Offa Uh as Maeno naaresce ba era cis Monnet rein Stic) SA CnG eMmnyERE: -\cr0 Mr. H. B. Fantham, F.Z.8. Exhibition of microscopic slides and description of a new Sporozoon, Lankesterella, tritonts.. 2. +2. cece cece eee ete cee cet eee teens 1. A Contribution to the unwise of the alte Arterial System in Sauropsida. By Frank E. Bupparp, M.A., F_BS., Brosector tothe, Societyacace sane toes ao. 2. On the Nomenclature of the Anthropoid Apes as proposed by the Hon. Walter Rothschild. By Sir H. H. Jounsron, G.O.M.G., K.C.B., F.Z.8. .-..c0cs cece ee eeee 3. On some Bats of the Genus Rhinolophus, with Remarks on their Mutual Affinities, and Descriptions of We nis new Forms. By Kyup Anpursen. (Plates III. & SAV 2) te 59 4. On Stridulating Hemiptera of the Subfamily Halyine, with Descriptions of new © Genera and new Species. By Dr. E. Bercrowu, C.M.Z.S., Tammerfors, Finland . 146 Odiiaas continued on page 3 of Wrapper. PROCEEDINGS OF 'THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1905, Vol. II. (May to December). May 2, 1905. Dr. W. T. Buanrorp, C.I.E., F.R.S., Vice-President, in the Chair. The Seeretary exhibited three large photographs (now in the Society’s Library), presented to the Society by Mr. Howard B. Turner, of Hippopotamuses swimming in a river in their native haunts. ae Mr. R. EH. Holding exhibited and made remarks upon a series of antlers of the first year of the Roebuck, Red Deer, Fallow Deer, and Wapiti. The exhibit had special reference to a paper read by Mr. Martin A. C. Hinton at the meeting of the Society held on March 21st, on some antlers of the Red Deer (Cervus elaphus) which were obtained from the Post-Pliocene deposits in the South of England, and in which it was stated that “these antlers belonged to individuals that had suffered testicular injury at an early period of life, by which the characters of youth were retained for a longer period than usual.” Mr. Holding pointed out from the specimens exhibited (text-fig. 1, p. 2) that the long pedicle, suppression of tines, and presence of rudimentary offshoots were characteristic of the antlers of all the Cervide at the first year or “ pricket” stage, and were not therefore due to testicular injury, and that any inter- ference or injury to the generative organs, as in castration, did Proc, Zoou, Soc.—1905, Vou. I. No. I. 1 ON FIRST-YEAR ANTLERS OF CERTAIN DEER. [May 2, Text-fig. 1. First-year antlers of certain species of Deer. A, Red Deer; B, Wapiti Deer; C, Fallow Deer; and D, Roebuck—showing adven- titious points marked X not being analogous to or the predecessors of the characteristic “tines” of the adult antler. E, lower portion of a pair of antlers of an aged Fallow buck, showing reappearance at the base of the left antler at X of one of these points or characters of the immature stage. IP 409,90 Saro HL 211 ale & Damelsson [tt B CLATHRINA CONTORTA 1905. ] ON THE SPONGE CLATHRINA CONTORTA. 3 not prolong or retain youthful characters of the antlers, but, quite the contrary, caused them to grow irregularly or had the effect, of entire suppression of the antler. He stated that very frequently an aged Stag or Fallow buck would throw up supernumerary snags at the base of the antler (text-fig. 1, E) or along the side of the beam, which somewhat resembled, and were probably a reversion to, these immature characters, and that there were several records of aged or barren hinds growing the simple “ pricket” antlers of the first year. Mr. R. I. Pocock, F.Z.S., exhibited and made remarks on a specimen of the Spanish Tarantula, Lycosa hispanica, that had died in the Society’s Gardens. On behalf of Mr. R. C. Punnett, F.Z.S., and himself, Mr. W. Bateson, F.R.8., F.Z.8., exhibited specimens of Fowls illustrating peculiarities in the heredity of white plumage, and made the following remarks :— A pure white breed such as White Leghorn, crossed with a dark breed such as Brown Leghorn, gives a cross-breed substantially white, the colour being recessive. The White Rose-comb Bantam, however, crossed with a coloured breed gives coloured cross breeds, the white being recessive. But in every specimen examined carefully these recessive whites were found to have one or more minute ticks of black pigment. Though, superficially regarded, these ticked whites would be classified as white, experiment proves them to be entirely different in nature. These facts elucidate the paradoxical accounts given by Darwin and others that Black and White Bantams crossed together give both blacks and whites; for the black may fully dominate over the white in this particular Case. The following papers were read :— 1. On the Sponge Leucosolenia contorta Bowerbank, Ascandra contorta Haeckel, and Ascetta spinosa Lendenfeld. By H. A. Mincamy, F.Z.8., University College, London. [Received March 16, 1905.] (Plate I.* and Text-figures 2-6.) The Calcareous Sponges have been a very unfortunate group, from the systematic point of view. From the time when Haeckel swept away all previous generic names, in order to found his so- called natural system, up to the present day, scarcely any two * Wor explanation of the Plate, see p. 20. |* 4 PROF, E, A. MINCHIN ON THE [May 2, authors have been in agreement as to the names to be employed for the genera or as regards the grouping of the species, especially in the more primitive and interesting section of the Calcarea Homoceela. The characters, for instance, by which Breitfuss defines the genus Leucosolenia of Bowerbank (1864) are suchas would exclude from it all, or nearly all, the species which I should refer to it, including, as I have shown elsewhere, even Bowerbank’s type species of the genus, LZ. botryoides ; while Lendenfeld has always consistently declined to make any use at all of the oldest generic name amongst the Ascons. In short, with the exception, perhaps, of the malarial parasites, there is probably no other group in the animal kingdom in which the nomenclature is in so confused a state as in the Homocela. The species which forms the subject of the present memoir illustrates well the statement just made. It is a veritable comedy of errors that I have to set forth. The name Leucosolenia contorta was given by Bowerbank in 1866 [1] to certain small sponges from the Channel Islands— Guernsey, and the Guliot Caves, Sark. It is not very clear, however, what Bowerbank considered the distinctive characters of his species, since his diagnosis would apply to almost any Ascon, He states that “the form of this sponge is so distinctly different from that of Z. botryoides that ....it cannot well be mistaken for that species... . Z. contorta always appears to consist of a mass of contorted inosculating fistule.” Further, that “the external surface of Z. contorta is also sparingly furnished with recumbent acerate spicule, mostly disposed in a longitudinal direction, and I have never observed like spicule on the surface of L. botryoides.” He was a little doubtful if his sponge were not really identical with Spongia complicata Montagu (1816), but came to the conclusion that Montagu’s figure of complicata was “really a very characteristic figure of Spongia botryoides of Hillis and Solander,” and that therefore the name complicata was to be rejected. Finally, Bowerbank remarks that contorta and coriacea might be mistaken for each other in the dried condition, but that ‘“‘the total absence of defensive spicule on the cloacal cavity of L. coriacea” (meaning apparently the gastral rays of the quadri- radiates) readily distinguishes it, If we put Bowerbank’s description into more modern terms, it amounts to this—that Z. contoria was characterised (1) by form and appearance (contorted inosculating tubes), (2) by the presence of triradiate, quadriradiate, and monaxon spicules. The term “equiangular ” applied by him to the triradiate systems need not be taken into account, since he applies the same term to the sagittal spicules of botryoides. It is not necessary to point out that the characters given by Bowerbank are not sufficient to define a species of Ascon; and when it is seen that botryoides always has monaxon spicules, as [ have shown elsewhere, and that contorta may frequently lack them; that the specimen of botryoides from which Bowerbank figured spicules (Brit, Spong, 111. pl, iii, figg, 3, 4) 1905, | SPONGE CLATHRINA CONTORTA, 5 was really a specimen of variabilis, while the specimen of contorta of which the spicules were figured (U. c. figg. 8, 9, 10) was really a specimen of complicate ; and that amongst nine of Bowerbank’s specimens examined by me I have found four distinct species confused together—to wit, complicata, variabilis, coriacea, and “ Ascetta spinosa Lendenfeld”: I think it is not necessary to say more in support of the statement that Bowerbank’s species contorta was of absolutely no systematic value whatever, but represented merely an ill-defined jumble of different species. In 1872 Haeckel, in his ‘ Kalkschwiimme’ [2], used Bowerbank’s specific name contorta for a sponge which he described in detail. Haeckel pointed out quite rightly that the external characters of contorta as set forth by Bowerbank were no guide whatever to its’ identification, since a quite similar mode of growth characterises other Ascons. Haeckel therefore diagnosed contoria by details of its spiculation. The diagnosis given is incorrect in two points, namely, in stating that “the monaxons possess a lance-head at their distal extremity, and that the gastral rays of the quadri- radiates are “curved oralwards”; two statements that lead me to suspect that Haeckel’s material of contorta was, like Bowerbank’s, contaminated by admixture of Lewcosolenia complicata. Haeckel, in his description, also affirmed, in his usual manner, definite characters in the spiculation without taking into consideration the variability which is so marked a feature of the sponge. It is a puzzle to me how Haeckel arrived at the definition which he gave of Ascandra contorta, since the specimens named and identified by him which I have seen do not agree with his description, and belong, indeed, to other species—a fact which easily explains any errors of description on his part. It is even more mysterious that Haeckel should have considered his contorta identical with Bowerbank’s contorta, since, of Bowerbank’s specimens examined by me, eight in all, not one agrees with Haeckel’s diagnosis! These enigmas are not, however, of importance to the present enquiry. Taking Haeckel’s description as it stands, and allowing for a certain margin of inaccuracy, I have been able without difficulty to refer to Haeckel’s Ascandra contorta a sponge extremely abundant on the Mediterranean coasts of France, and occurring elsewhere also. As I have stated in a previous memoir, I consider that where previous writers leave us in doubt as to the characters of a species, Haeckel’s description fixes the application of the name. J will proceed now to describe the sponge which I regard as the true contorta, and then to consider the synonymy and application of the name. Ascandra contorta H. is a species which, for reasons stated elsewhere [4, &c.], I refer to the genus Clathrina Gray (1867). It has a closely reticulate mode of growth, equiangular triradiate systems, collar-cells with basal nucleus, and parenchymula larva ; all these being characters which make up my diagnosis of the genus Clathrina. 6 PROF. E. A. MINCHIN ON THE [May 2, The specimens of this sponge which I have studied nearly all came from Banyuls-sur-Mer, where this species is extremely abun- dant. By the kindness of Monsieur Topsent, however, I have seen a specimen from Roscoff, not differing in any respect from the Mediterranean specimens. ‘The sponge therefore has a wide range of distribution, and is almost certainly to be ranked as a member of the British Fauna, though it does not appear to be common on our coasts. Hanitsch has, indeed, recorded it from Liverpool : I have no reason to doubt the correctness of this record beyond the fact that my experience of specimens labelled contorta by the most eminent authorities has left me very sceptical as to the correctness of any identification of this species which I have not checked ; a scepticism heightened, in the present instance, by the fact that Hanitsch names his specimens Ascaltis contorta. I may add that the sponges named Ascandra contorta by Breitfuss in various memoirs have nothing to do with this species, and should not therefore be taken into account in considering its geographical range. At Banyuls-sur-Mer Clathrina contorta is not only one of the commonest, but also one of the largest Ascons occurring there. Colonies frequently measure 8 centimetres or more across. They consist of a massive or spreading growth of twisted anastomosing tubes, running in all planes, and forming a dense feltwork from which arise at intervals the short, straight, not very conspicuous oscular tubes, which reach two or three millimetres in height, and are of slightly larger calibre than the body-tubes, as the basal growth may be called. The body-tubes are centred round the oscular tubes more or less distinctly, and in the region of the oscular tube the basal system of tubes is usually slightly raised up to form a conulus bearing the oscular tube on its summit; but these conuli are generally very shallow, so that the upper surface of the spreading colony is nearly flat, not lobulated like that of cerebrum, nor cushion-like, as in reticulwm—two species occurring commonly with contorta, but both very easily distinguished from it at sight. Photographs will make the external characters of contorta clearer than any description (Plate I.). Of its allies, it is perhaps coriacea with which contorta might be most easily confused, on simple inspection ; the latter, however, with its greatly developed gastral rays, is not found contracted up, with closed oscula, like coriacea, and when expanded its body-wall is much thicker and less delicate. The spiculation of Clathrina contorta comprises in typical specimens all the three kinds of spicules found in calcareous sponges. The triradiate systems are equiangular, with the rays straight, tapering imperceptibly for the proximal half or two-thirds; after that tapering more rapidly to a sharp or moderately blunt point (text-fig. 2, 1a-1f). The distal extremities of the rays are often irregular in outline, sometimes markedly so. ‘The rays vary in length from 80 to 130, in different specimens, but may be said to average 90-100. The breadth at the proximal end of the 1905.] SPONGE CLATHRINA CONTORTA. 7 Text-fig. 2. 1) , ie / Spicules of a specimen of Clathrina contorta from Roscoff. Figg. 1a, triradiate; 1 6-1 e, quadriradiates in facial aspect; 1, abnormal quadri- radiate with one basal ray wanting; lg-17, quadrivadiates in side view, showing gastral rays in profile; 1j-1m, monaxons (the spicule represented by 1, being too long for the page, has been drawn in two pieces). 8 PROF, E. A. MINCHIN ON THE [May 2, ray is usually 8 or 9 u, but may reach 12; speaking generally, slender triradiate systems, with rays not exceeding 10p in breadth, can be distinguished from thick ones with rays exceeding 10 (text-fig. 3,2a-2f). In some specimens the triradiate systems are all, or nearly all, of the slender type; in others, triradiate systems of the thick type are more abundant. Some of the triradiate systems develop gastral rays, becoming quadriradiates, and others do not. Asarule the quadriradiates are more abundant than the simple triradiates. Im some specimens there is a tendency for the simple triradiates to be of rather stouter build than the quadriradiates, but in other specimens this cannot be noticed. The gastral rays of the quadriradiates are attached at the centres of the triradiate system, and are remarkable for their slenderness and usually also for their length (text-fig. 2, 1 g-1 2). Arising from a slightly expanded base, the gastral ray sometimes tapers rapidly to a point, then reaching a length equal to about one-half or one-third of that of the basal rays; but more usually the gastral ray is prolonged to a considerably greater length than the basal rays, reaching 130, 140, or even 150 » in length. The gastral ray then becomes excessively slender for the distal half or two-thirds of its length, and ends ina sharp point; it is not bent oralwards as Haeckel describes it, but it is either quite straight or irregularly curved. Haeckel’s figure of a quadri- radiate (Kalkschwamme, iii. pl. 14. fig. 6c) obviously represents a spicule of LZ. complicata (compare his fig. 1 e on pl. 15, l.c.). Quadriradiates are also to be found in which, with gastral rays of great length, are found basal rays much shorter than usual text-fig. 2, 1g; text-fig. 4, 4e); these are probably young forms in which the rapid growth of the gastral ray * has caused it to attain its full length before the basal rays have done so. In the thick quadriradiates found in many specimens, I have observed a curious point with regard to the gastral ray, when seen in the facial aspect of the spicule. When the basal system is focussed so that the bases of the rays show sharp contours, the origin of the gastral ray appears as a dark central spot roughly triangular in outline, each side of the triangle being transverse to the base of one of the rays of the triradiate system, and the angles of the triangle rounded off (text-fig. 3, 2a, 2). If now the focus is slightly raised, the base of the gastral ray appears as a sharp ring, within the triangle. The dark triangle appears to be the expanded base of the gastral ray, but it is only to be seen in the case of the thickened triradiate systems, not in the slender ones. The monaxon spicules of Clathrina contorta vary in the most singular manner, constituting the most remarkable feature of the species. The variations are best considered, first, from the point * As I have described in a former memoir (Quart. Journ. Micr. Sci., n.s. xl. pl. 42. fig. 55), the elongated gastral rays of contorta are covered by a plasmodial mass containing four nuclei, more than I have observed on the gastral rays of any other Ascon. 1905. | SPONGE CLATHRINA CONTORTA. 9 Text-fig. 3. gh Ga eh Spicules of two specimens of Olathrina contorta from Banyuls. Figg.2a@ & 6, thick quadriradiates; 2¢ & d, slender quadriradiates; 2e & f, triradiates; 2g, quadriradiate showing gastral ray in profile; 2/,a monaxon, 3a & b, quadriradiates of another specimen ; 3 c-3 h, monaxons. 10 PROF, E. A. MINCHIN ON THE [May 2, of view of substantive variations of form and size; secondly, as regards numerical variation, that is to say abundance of monaxons compared with other types of spicule. The monaxons are all of large size, being at least twice as thick as the basal rays of the triradiate systems, and not less than 300 » in length, allowing for those which are apparently not full- grown. But in some specimens the monaxons reach a size which can only be called gigantic. In a specimen from Banyuls sent me by Topsent (which I will refer to as Topsent 12¢), the monaxons, when drawn to the same scale as the other spicules figured here, come out 32 centimetres in length, corresponding to an actual length exceeding 1000p (1 mm.), with a breadth of about 50 at the thickest part. Even these proportions are exceeded by a specimen in my collection from Banyuls, in which | the monaxons when drawn to scale measure 75 centimetres in length, corresponding to an actual length of 2343 (2°3 mm.). I do not think that spicules of such size have been recorded from any Ascon. The large monaxons of Ascandra densa and A. parus figured by Haeckel (J. c. pl. 14. figg. 2c, 3,f) fall far below those that I have mentioned in dimensions. With these extraordinary variations in size, the form and characters of the monaxons are fairly constant (text-figg. 2 and 3, 1j-lm, 2h, 3c-3h). They are spindle-shaped, pointed at both ends, slightly curved, some- times distinctly so when more slender, or nearly straight when very thick. There is no lancet-head present at the distal ex- tremity, as figured by Haeckel; his figure (/.c. pl. 14. figg. 6 d, 6 ¢) almost certainly refers to complicata (compare his figg. 1 g—-1 k, on pl. 15). It is, indeed, impossible to say which is the distal end of these monaxons, as they do not project from the sponge like the true (primary) monaxons of other Ascons. Near the middle of the spicule, sometimes at about one-third of the length from one end, a slight constriction can be observed, sometimes very distinct, in others very shallow, in others again represented by an annular thickening, and sometimes not to be made out at all. This con- striction is more distinct in young spicules, and appears to become more or less obliterated with growth. In big spicules the contours are often so sinuous and irregular that the primary constriction may be masked by secondary curves. I consider this primary constriction, as I propose to call it, of great morphological importance, as indicating probably that these spicules are not primary monaxons*, comparable to those of Leucosolenia conyplicata, for example, but in reality derived from a triradiate by loss of one ray and shifting of the two others into approxi- mately the same straight line. In very young monaxons of contorta I have noticed a delicate transverse line in the region of the constriction (text-fig. 3, 3¢), and I have also found a spicule of which it would be difficult to affirm whether it is a young * A primary monaxon is derived from a single mother cell which divides into two formative cells, thus originating in exactly the same manner asa single ray of a trivadiate system. 1905.] SPONGE CLATHRINA CONTORTA. 11 monaxon or an abnormal triradiate (text-fig. 3, 3); probably it is both! My friend Mr. Alford has also found, in the slide of Topsent 12 e, four abnormal monaxons which have additional rays growing out laterally and thus become triradiates (text-fig. 6, 9a-9c). In one of these (96) the three rays are approximately equal in size and meet at the angles of an ordinary triradiate. For all these reasons I consider there is much to be said for regarding the monaxons of contorta as secondary monaxons derived from a triradiate system by suppression of one ray and hypertrophy of the two remaining, which become piaced in the same straight line, or nearly so. The numerical variation in the monaxons is not less re- markable. In some specimens scarcely any monaxons are to be found; in others they are extremely abundant. Thus in a specimen recently examined by me, I took a fairly large piece of the sponge, separated the spicules with Eau de Javelle, and mounted all I could get up with the pipette, covering three slides. After prolonged searching I found five monaxons to many thousands of triradiate systems. In another specimen in which IT could find no monaxons, Mr. Alford by careful searching found two. It is often extremely difficult to be certain if a specimen has monaxous or not. Mr. Alford has kindly undertaken for me the task of counting the numbers of each kind of spicule found in different specimens, with the following results :— Triradiates. Quadriradiates. Monaxons. | Kind of | 5°21 2 Specimen. | Actual |—Per- | Actual | Per- | Actual | Per- Moerene of number | centage | number | centage number centage servee: | Spicules. counted.'of whole.| counted. |of whole. counted. of whole. | sea “yi Pace aerial Pn =o Alea 93 | 3-278 27 | 96:128 5 Large. | 2837 (Zz 3.a-3h). 3 | 3278+] 272 23+) 17 599+ arge. 3 MIO. 311 | 8304+] 3423 |91402 | 11 | 204+) Lar 3745 (z2a-2h) +| 342 02 aes ae || arge. 3) INI@ B soosos 886 |12°512+| 2658 |86158+ 41 | 1329+ |Very large.| 3085 No. 4 o47 | 5835+/ 3965 |93668+ 21 | “496+ Very large.| 4233 y (Plate I. B). | | IN@s 6) coonee 146 5'144+| 2686 | 94°644+ 6 | ‘211+ | Gigantic. | 2838 No. 6 | Cow 267 |10°349+] 2188 |84:806+, 125 4-844 | Gigantic.) 2580 12 e). | Total for ; : Ad Spacing, } 1450 7506 | 17647 | 91°35 221 1144 19318 These results were obtained in the following way :—‘“ Hach specimen was put into Eau de Javelle to separate the spicules, and after careful washing, and being allowed to stand for some 12 PROF. E. A. MINCHIN ON THE [May 2, considerable time after each washing, the spicules were transferred to the slides by means of a pipette. “Hach slide, when ready, then had marked upon its under surface twenty circular areas, each being brought into the micro- scopic field in turn and all spicules in each area carefully counted. When all the spicules were counted the circle was erased and the next circular area dealt with. “The counting was done with the aid of a camera lucida and three differently coloured crayons, thus ensuring that all spicules were counted and counted once only. “Hach quadriradiate spicule had a number in blue marked upon it; the triradiate spicules were marked with successive red numbers and a green number noted a monaxon. At each counting a check could be made, and the counting was complete when each spicule was seen to have one number of a special colour upon it.” The spiculation of Clathrina contorta thus shows, on the one hand, comparatively slight variation in the triradiate systems, and, on the other hand, extraordinary differences in number and size of the monaxons in different specimens. The variability is so marked, and the monaxons are frequently so difficult to find, as to suggest at once a possible extreme of variation in which the monaxons would be totally absent. Were this to occur we should have a variety of the sponge characterised by a type of spiculation which would lead to its being placed, in many current systems of classification, in a genus distinct from the variety in which monaxons occur. As a matter of fact, I may state at once that the variety of contoréa in which monaxons are completely lacking is very common, and it has been described by Lendenfeld from the Adriatic under the name of Ascetéia spinosa, This is no mere surmise on my part ; I have been able to examine, in the collection of Canon Norman, a slide obtained by him from Lendenfeld, and bearing in Len- denfeld’s handwriting the label “ Ascetéa spinosa.” Text-fig. 4, 5 a—5 h, vepresents some spicules drawn by me from this slide. As will be seen, the spiculation differs in no single particular from that of the true contorta, except for the lack of monaxons. Since the preparation consists of tubes of the sponge mounted whole, it was not possible to obtain profile views of the gastral rays, except at the torn ends of the tubes, and in no case was I able to see an unbroken gastral ray in side view, but the fragments which I have drawn (5/5 h) are sufficient to prove that the gastral rays of this specimen attain the degree of length and slendernegs characteristic of the species. Lendenfeld’s specimen is, in fact, identical in character with other specimens of “spinosa” which I have from Banyuls (text-fig. 4, 6 a6 g), and these again differ in no respect from the true contorta except for the absence of monaxon spicules. If Ascetta spinosa Lend. is to be regarded, as I believe, merely as a variety of Ascandra contorta H., how is this variation to be explained? The specimens of spinosa that have come under my notice agree perfectly in external characters with contorta, but are 1905. | SPONGE CLATHRINA CONTORTA. 13 Text-fig. 4. 62 Spicules of the “ spinosa” varicty of Clathrina contorta. Figg. 4a-4f. Spicules of Bowerbank’s type of Leucosolenia contorta in the British Museum (Bowerbank Coll. 988), showing gastral rays with tendency to irregular curvature.—Figg. 5 a-5 h. Spicules of a specimen in Canon Norman’s collection labelled “ Ascetta spinosa” in Lendenfeld’s handwriting ; the elongated gastral rays (5f-5h) are broken off.—Figg. 6 a-6g. Spicules of a specimen from Banyuls. 14 PROF. E, A, MINCHIN ON THE [May 2, all of small size. The big, spreading colonies of contorta always have monaxons. It is my belief that the absence of monaxons is simply a juvenile feature, so to speak, of the sponge, and that they are only formed when the sponge has grown to a certain size. Such changes of spiculation with age are probably more frequent in sponges than is usually supposed. Fora parallel case I need only refer to Topsent’s observations on Cliona celata. A point which requires brief discussion, however, is why Lendenfeld found only the spinosa-form in the Adriatic, and not the contorta-form, if these two forms are really only age-variations in one species. Are we to suppose that in the Adriatic the sponge does not acquire monaxons? In my opinion the explanation of this point is to be sought in quite a different manner. In his ‘ Kalkschwiimme der Adria’ [3] Lendenfeld describes another species of Clathrina occurring commonly in the Adriatic, namely C. reticulum. I have also found this species very abundant at Banyuls, and I possess many specimens of it; but my experience of this species at Banyuls differs sharply in one respect from Lendenfeld’s observations upon it in the Adriatic. I find reticulum to be more constant in external form and characters than any other species of Ascon. All the specimens I have seen—and at one time I had some hundreds of specimens, collected in order to obtain the larval development—are compact, rounded, cushion- like masses of slender, closely-knit tubes, forming a dense and finely-meshed reticulum from which arise one or more oscular tubes of much larger calibre than the tubes forming the body of the sponge. I have figured such a specimen elsewhere (4, p. 6, fig. 6). In short I have never had the slightest difficulty in recognising reticulum at sight, though its spiculation often approaches that of contorta very closely. My astonishment was therefore great to find that Lendenfeld describes this sponge as occurring (at Sebenica and Lessina) in nearly all the forms generally found in Ascons. There is thus a great discrepancy between Len- denfeld’s observations and mine with regard to this species,and I am inclined to think that this is to be explained simply by Lenden- feld not having recognised the true contorta, but having confused it with reticulum. This is a supposition which I am unable to prove or test; but if correct, it would explain why Lendenfeld did not find the true contorta occurring in the Adriatic as well as spinosa, and also why he finds vedtiewlwm so variable in form when in my experience it is so extremely constant. I may add, finally, that the figures of monaxons of reticulum given by Lendenfeld (3, pl. viii. figg. 7 e-7f) are more like those of contorta than those of reticulum, though not exactly like those of either, as these sponges are known to me. I will now describe some of the historically important specimens to which I have had access, and I begin with the type-specimens of Bowerbank’s Leucosolenia contorta in the British Museum (Bowerbank Coll. 988). The “type” consists of seven dried specimens, all very small, stuck on a card, The largest specimen, 1905. | SPONGE CLATHRINA CONTORTA. 15 Text-fig. 5, Spicules of Leweosolenia, Sycon, and Clathrina. Figg. 7 a-7 1. Spicules of a specimen in Norman’s collection, received from Bower- bank with label Leuwcosolenia contorta and identified by Haeckel as Ascandra contorta ; showing spicules of Leucosolenia variabilis (7 a-7j), mixed with spicules of Sycon sp. (7k, 71).—Figg. 8a-8m. Spicules of a specimen in Norman’s collection received from Bowerbank with label Leucosolenia contorta ; showing spicules of Leucosolenia complicata (8 a-8j) mixed with spicules of Clathrina coriacea (8 k-8m), 16 PROF, E, A. MINCHIN ON THE [May 2, the original of Bowerbank’s fig. 7 on pl. ii. of Brit. Spong. vol. i11., is at the top over the middle of the card; the other six arein two vertical rows of three each to right and left. As I have stated elsewhere, J have examined six out of these seven specimens, and all of them, except the larger one at the top, are quite typical specimens of Lewcosolenia complicata ; the large specimen alone is atrue Clathrina. 1 give figures of its spicules (text-fig. 4, 4 a—4/), and it is not necessary for me to describe them in detail, for it is evident from the figures that this specimen agrees with the true contorta in all respects but one, namely, in that the monaxons are wanting. In short, Bowerbank’s type-specimen of ‘“ Leucosolenia contorta,” or, to be more accurate, the only one of his type- specimens which does not belong to a species of prior standing, is a specimen of “‘ Ascetta spinosa” Lendenfeld ! I have also examined two other specimens of Bowerbank’s*, given by him to Canon A. M. Norman, and now in the latter gentleman’s collection. The first of these was sent by Canon Norman to Haeckel, and returned by him after examination. It has the following label in Norman’s handwriting :— “ Teucosolenia contorta Bow. ““ Guernsey “(A type-specimen from Dr. Bowerbank).” Also a label in Haeckel’s handwriting :— “ Ascandra contorta H. ue (Leucosolenia contorta Bwhk.) “ Guernsey, Bowerbank.” If any specimen in the world ought to have been a specimen of contorta, surely this ought, bearing, as it does, a double testimonial to character from the two founders of the species. What, then, was my astonishment, on examining the spicules, to find it a quite typical example of Leucosolenia variabilis Haeckel! I figure its spicules in text-fig. 5, 7a—-71. The only point to notice about them is a certain admixture of Sycon spicules (7%, 71), which, as I have set forth in another place, frequently occurs in preparations of variabilis. The second specimen in Canon Norman’s collection bears a label in Bowerbank’s handwriting as follows :— “ Leucosolenia contorta, Guernsey.” According to information furnished me by Canon Norman, this particular specimen was not sent to Haeckel, but it is one of the same lot as the type sent to him, and has an equal claim to be regardedas atype. Examination of the specimen shows a mixture of Leucosolenia complicata and Clathrina coriacea (text-fig. 5, 8 a-8 m). * Bowerbank in his Monograph mentions twenty-eight specimens of contorta, but I have had access to only nine of them, Ido not know what has become of the others. 1905. | SPONGE CLATHRINA CONTORTA. 17 From the foregoing it will be seen, I think, that the name- question, in the case of the species under consideration, isa tangled problem, one, indeed, which I feel some diffidence in approaching. IT could wish, in fact, as I have said elsewhere, that there were in existence some sort of International Hague Tribunal to which these knotty points of nomenclature could be referred for arbitra- tion and authoritative settlement. In the absence, however, of any such body, I extract from the facts above set forth the following conclusions :-— (1) Bowerbank’s Leucosolenia contorta was a jumble of different species, and his description could not be used for identification of any particular species. Hence Leucosolenia contorta Bowerbank is a nomen nudum, of no systematic validity. (2) Haeckel’s Ascandra contorta, though not in all respects correctly described, can be applied to an existing species of Ascon, which can be identified by his description. This I consider the true contorta: ought the species, however, to be written contorta Bwk.or contorta H.? Pending the constitution of the International Nomenclature Tribunal, in order to settle this important point, I content myself in following Haeckel in calling it contorta Bwk. (3) Ascetta spinosa Lend. is probably the young form, without monaxons, of contorta. IT arrive therefore at the following synonymy and diagnosis :— CLATHRINA CoNTORTA (Bowerbank). ? Nardoa spongiosa Kolliker *, 1864, [cones Histologice, Abth. 1. pp. 63, 64, pl. vil. fig. 10, pl. ix. fige. 6-8. Leucosolenia contorta Bowerbank 1866, Mon. Brit. Spong. 11. pp. 29-32; 1874, op. cit. 111. pp. 7-8, pl. 11. figg. 5-10. Leucosolenia (Nardoa) contorta Gray, 1867, P. Z. 8. p. 555. Leucosolenia (Leuciria) contorta Haeckel, 1870, Jen. Zeitschr. v. p. 243. Ascandra contorta Haeckel, 1872, Kalkschwiimme, ii. pp. 91— 93, 11. pl. 14. fige. 6 a6 e. ? Ascaltis contorta Hanitsch, 1890, Tr. Biol. Soc. L’pool, iv. pp. 195 & 233. Ascetta spinosa Lendenfeld, 1891, Zeitschr. wiss. Zool. liii. pp. 203-205, pl. vii. figg. 2, 16, 21, 22. Leucosolenia contorta Topsent, 1891, Arch. Zool. Exp. (2) ix. p. 925; Bull. Soc. Zool. France, xvi. p. 128; 1892, Résult. Cam- pagnes Sci. Albert 1°, fase. 11. p. 22; 1894, Rev. Biol. Nord France, vii. pp. 7 & 22. Clathrina contorta Minchin, 1896, Ann. & Mag. Nat. Hist. (6) Xvlil. p. 399. * Nardoa spongiosa Kolliker has been put by Haeckel as a synonym of either Ascaltis cerebrum or A. gegenbawri, but the figures of the external form, no less than those of the spiculation, given by Kolliker, seem to me to indicate that the author was dealing with the spinosa-form of contorta. I have discussed this point elsewhere (Quart. Journ. Micr. Sci. n.s. xl. p. 533, footnote). Proc. Zoou. Soc.—1905, Vou. II. No. II. 2 18 PROF. E. A. MINCHIN ON THE [May 2, Clathrina spinosa Minchin, ibid. Leucosolenia spinosa Breitfuss, 1898, Arch. f. Naturges. lxiu. 1, pels. (The following references, on the other hand, probably do not relate to the true contorta.) Ascandra contorta Barrois, 1876, Ann. Sci. Nat. (6) 11. Article 11, p. 35, probably refers to Leuwcosolenia complicata. Leucosolenia contorta Carter, 1880, Midland Naturalist, 11. p- 195. The author remarks that ‘ Bowerbank’s illustration of the linear spicule is defective. There are two forms, quite different from each other and from Dr. Bowerbank’s figure.” I consider it probable from this statement that Carter was dealing with a specimen of Leucosolenia complicata. Ascandra contorta Breitfuss, 1898, Arch. f. Naturges, lx. 1, . 214, refers to a specimen of Leucosolenia complicata ; so pro- bably also the sponge described and figured by the same author in Mém. Ac. St. Pétersbourg, 1898 (vili.) vi. p. 15, pl. i. fig. 1, and cited by him in other memoirs. And finally it should be mentioned that the numerous specimens sent out from Sinel and Hornell’s Zoological Station, Jersey, are all, so far as I have seen, specimens of Lewcosolenia complicata. Diagnosis.—Trivadiate systems equiangular, with or without gastral rays; the quadriradiates generally more numerous than the simple triradiates. Rays of the triradiate systems tapering imperceptibly for the proximal half or two-thirds, then narrowing more rapidly to a sharp or moderately blunt point. Gastral rays sometimes short, more usually longer than the basal rays, very slender, sharp, and straight or irregularly curved. Monaxons at least twice as thick as the basal rays of the tri- radiate systems,—varying in different specimens from a moderate size to gigantic proportions, spindle-shaped, usually slightly curved, and usually with a distinct constriction near the middle of their length ; sometimes very few in number, sometimes absent altogether. The chief objection that can be made, it seems to me, with regard to my treatment of the species, relates to the position of spinosa. Naturalists concerned chiefly with the arrangement of specimens in bottles'on shelves will perhaps object to my “lumping” together two forms which can be separated by a definite character, although by one only. Those who reason thus will, no doubt, prefer to retain spinosa as a “species” distinct from contorta ; in that case the type of Bowerbank’s contorta belongs to the former species, a fact which raises alarming problems of nomenclature. The range of variation seen in contorta has its natural and logical termination in the form spinosa, and justifies, in my opinion, placing the latter asa synonym. Moreover it is often extremely difficult to be certain that monaxons are really absent in a specimen of “ spinosa.” They may be so scarce that they have been simply overlooked. After arriving at the above conclusions with regard to the 1905. ] SPONGE CLATHRINA CONTORTA. 19 identity of contorta and spinosa, it is hardly necessary for me to express my opinion with regard to those systems of classification which define not only species but even genera of Ascons by the presence or absence of monaxon spicules. Before such a character as the presence or absence of monaxons can be used for systematic Text-fig. 6. Abnormal gigantic spicules of the class of the monaxons from a specimen of Clathrina contorta from Banyuls (Topsent 12¢). Magnified about 150 lincar (i. e. half as much as the spicules figured in text-figg. 2-5), purposes, it is necessary to understand clearly what is meant by a monaxon spicule. In calcareous sponges a spicule of this class may be one of two perfectly distinct things. It may be, on the one hand, 2% 20 ON THE SPONGE CLATHRINA CONTORTA. [May 2, a primary monaxon spicule, derived from a single mother-cell, and developing exactly in the same way as a single ray in a tri- radiate system, with which it is strictly homologous. It may be, on the other hand, a secondary monaxon, derived by modification of an entire triradiate system by loss of one ray, perhaps in some cases two rays. Good examples of monaxons undoubtedly of secondary nature aie the elbowed monaxons in the stalk of Clathrina lacunosa Johnston (renamed Ascandra angulata by Lendenfeld). TI believe also, as stated above, that the monaxons of contorta are to be regarded as secondary. It is clear that a character which is sometimes one thing, in other cases quite another thing, cannot be usefully employed for purposes of systematic classification, not, at least, until more is known about it. If Ascetta spinosa be put as a synonym of Clathrina contorta, it is seen that the species has a wide range, extending from the Adriatic round the coasts of France into the English Channel, and probably also on to the coasts of Great Britain. Tt is my pleasant duty finally to express my thanks to friends who have assisted me in the preparation of this memoir, put together from observations for the most part of long standing, at a time when the stress of other work, caused by preparations for my departure for the Tropics, was very great. My friend Mr. G. R. Alford, who is making a special study of the variation of this sponge, has given me valuable assistance, as will be evident from the facts I have quoted from him above. Mr. Alford has also kindly undertaken to see this memoir through the press for me. My friend and pupil Mr. L. R. Crawshay has given me great help in preparing the illustrations. Finally, I have to thank Monsieur Topsent, of Caen, for his kindness in sending me specimens from Roscoff and elsewhere and for answering many queries. BIBLIOGRAPHY. (1) Bowrrsank, J. 8. A Monograph of the British Spongiade, London, Ray Society, 3 vols.: 1864-1874. (2) Harcken, E. Die Kalkschwimme. Berlin, 1872; 3 vols. (3) LeENDENFELD, R. v. Die Spongia der Adria: I. Die Kalk- schwimme. Zeitschr. wiss. Zool. li. (1891) pp. 185-321 361-433, pls. vill.—xv. (4) Mincuin. EK. A. Sponges in: Lankester, ‘A Treatise on Zoology, London, 1900. p) Other references are cited in the list of synonymy, p. 17 above. EXPLANATION OF PLATE T. Clathrina contorta from Banyuls. A from above; B from above, and C from the side, to show the oscular tubes (O). 1905. ] ON THE ANATOMY OF THE FERRET-BADGER. 21 Some Notes upon the Anatomy of the Ferret-Badger, Flelictis personata. By Frank E. BeEpparp, M.A., F.R.S., Prosector to the Society. | Received March 21, 1905. ] (Text-figures 7-12.) The dissection of a female example of //elictis personata, which was acquired by the Society on the 4th and died on the 14th November, 1904, enables me to lay before the Society some new facts in the anatomy of this genus of Carnivora. So far as Iam aware, the only zoologist who has investigated the anatomy of the soft parts of the genus Helictis is the late Prof. Garrod *, whose memoir deals with the essentials in its structure. The species examined by him was Helictis subaurantiaca. It is not therefore unnecessary to report upon the anatomy of another species, though the differences between the two are, as might be expected, but slight. I deal, moreover, with a few points upon which Prof. Garrod did not touch in his account. S Brain. The brain of Helictis swbawrantiaca has been described and figured (in dorsal and lateral view) by Prof. Garrod in his memoir already referred toy. The figure of the brain of Helictis personata submitted herewith (text-fig. 7, p. 22) shows certain differences, which I regard as worthy of record in view of the little knowledge which we possess upon the matter. The most salient difference which this brain shows from that of H. subaurantiaca is the very slight appearance upon the dorsal surface of the intercalary prolongation of the calcarine sulcus. This furrow, as will be seen in the figure (text-fig. 7), only appears dorsally for a short distance quite at the posterior end of the hemispheres, and also of course anteriorly where the two sulci join the crucial sulci. The precrucial sulcus in my specimen is not so fully developed, particularly upon the left side (text-fig. 7, Pc.S.), as in Garrod’s specimen of Helictis subaurantiaca. It does not entirely delimit the ursine lozenge in front. The Sylvian fissure on both sides of the brain joins the supra- sylvian, the gyrus anterior tothe Sylvian beingapparently depressed below the surface of the hemispheres. There is a hint of this in Garrod’s figure, but hardly in that of Dr. Eliot Smith, though it refers, I imagine, to the same brain. The remaining fissures agree absolutely with those of Helictis subaurantiaca. 1 pass on therefore * “Notes on the Anatomy of Helictis subawrantiaca,” P.Z.S. 1879, p. 305. + This brain is also figured in the Catalogue Physiol. Series Roy. Coll. Surgeons, vol. ii. (2nd ed.) p. 273, by Dr. Elliot Smith. 22 MR. F, E. BEDDARD ON THE [May 2, to the arteries of the brain, which are most satisfactorily injected in my specimen and which show all the Arctoid characters*. The rhomboidal area formed by the bifurcation of the anterior spinal and its junction with the basilar is of considerable calibre and uniform throughout, as in all Carnivora which have been examined. The vertebral arteries are, however, peculiar in their mode of joining this rhomboidal vessel. Each vertebral artery in fact divides before joining the rhomboidal, and each branch opens separately into it, as is shown in the accompanying figure (text- fig. 8). The carotids join the circle of Willis just before the middle cerebral arteries are given off. Text-fig. 7. Text-fig. 8. Text-fig. 7—Brain of Helictis personata, dorsal aspect. Cr. Crucial fissure; Za¢. Lateral fissure; Orb. Orbital fissure; Pe.§S. Precrucial fissure; S.S. Supra-Sylvian fissure. Text-fig. 8—Brain of Helictis personata, ventral aspect, with the arterial system shown in thicker and thinner black lines. The dotted lines delimit regions of the brain. b.a. Basilar artery ; Ca. Carotids; Call. Callosal arteries ; P.c. Posterior cerebellar; v.a. Vertebral arteries. Anteriorly the circle of Willis is completed by the fusion of the two callosal arteries, that of the right side being distinctly smaller than that of the left. * Beddard, P, Z.S. 1904, vol. i. p. 183. 1905. | ANATOMY OF THE FERRET-BADGER. 23 The posterior cerebellar arteries are asymmetrical in their origin from the basilar, the left being considerably in front of the right. The middle cerebellar arteries arise in front of the sixth nerve. $ Some Notes on the Muscles. The muscular anatomy of the Carnivora has been lately treated of in an exhaustive fashion by Messrs. Windle and Parsons*, As a supplement to that paper (which does not deal with Helictis) I am able to offer a few notes upon the musculature of Helictis personata. The Sterno-mastoid consists from the very beginning of two parts: the larger of these is inserted on to the mastoid next and superficial to the cleido-mastoid muscle; the smaller part crosses the cleido-mastoid and joins the cephalo-humeral. This latter portion of the muscle has been spoken of as a portion of the trapezius, with which, indeed, it is plainly confluent above. The Sterno-hyoid and Sterno-thyroid appear to arise from the sternum as one muscle. I could find no tendinous intersection. The Omohyoidis apparently completely absent. I could find no trace of it. This muscle is usually present in Mustelide. The Omotrachelian has exactly the relations described by Windle and Parsons. The Rhomboideus profundus, which arises from the supra-spinous fossa of the scapula near to the root of the spine, is a slender muscle inserted on to the atlas deep of the omotracheal. It is perfectly distinct at its origin from the Rhomboideus cervicalis. Its discovery in Helictis gives further support to Messrs. Windle and Parson’s belief that the muscle is eminently characteristic of the Mustelidee. The Rhomboideus capitis has only asingle origin in common with the Rhomboideus cervicalis, not the double origin of Jctonya (a near ally of Helictis) as figured by Windle and Parsons. The Dorso-epitrochlear is contiguous to and hardly if at all distinguishable from the extra head of the Z’riceps occurring in this as in many other Carnivora. The Dorso-epitrochlear itself is of course part of the Latissimus dorsi; in passing by the scapula it receives a mass of fibres from the lower border of that bone and thence becomes continuous with a sheet of fibres arising from the Teres and constituting, as I imagine, the “extra head” of the Z'riceps of Messrs. Windle and Parsons, which those anatomists state to be characteristic of the Mustelide. The iceps has only one head. Helictis appears to possess two distinct Palmaris longus muscles. The Sartorius is Single and fused at its insertion with the also single Gracilis. The Pectineus, often a double muscle, is single in Helictis. * P. Z.S. 1897, p. 370, & 1898, p. 152. 24 MR. F. E. BEDDARD ON THE [May 2, I found it impossible to subdivide the Adductor mass. The Semimembranosus is divided into two muscles for some way in front of its obviously double insertion on to the tibia and the femur. I could not find, however, that this muscle was divided at its origin from the ischium, The Scmutendinosus, as in some other, but not in all, Mustelidee, has a very distinct caudal head. There is no Agitator ‘eaudee. The Zenwissimus is plainly present. The Zbialis anticus is single. § Lungs. As Prof, Garrod pointed out in H. subauwrantiaca, the lungs in H. personata consist of four lobes on the right side and two on the left. Prof. Garrod, however, made no observations upon the relative sizes of the several lobes. On the right side the first lobe is rather larger than the second; the third is the biggest of all and quite twice the size of the first; the fourth or azygos lobe 1s the smallest of all. The two lobes on the left side are more nearly equal in size, but the second or lower lobe is the larger. § Lier. The liver of this species appears to be much like that of H. subaurantiaca. The enormous right central lobe is deeply * fissured and exposes the gall-bladder on the diaphragmatic side. This lobe is quite twice the size of the left lateral lobe, which is the next largest ; this lobe again is larger than the right lateral, which does not show any great difference of size from either the left central or the caudate. The Spigelian lobe is minute. § Pancreas. The pancreas of Helictis is almost exactly like that of the Tayra (Galictis), with which Arctoid I have specially compared it. It is not clear from Garrod’s description what is the precise form of the gland in the species investigated by himself. In H. personata there is a circular portion of the pancreas running right round the duodenal loop +; this ends in a straight piece running parallel with the spleen. The chief difference which Helictis shows from Galictis is in the mesenterial attachment of the straight part of the pancreas. In Galictis a transparent mesentery, apparently anangious, 1s attached to the whole length of the straight region of the pancreas, and is inserted on to the mesocolon along a line which commences in front of and ends behind the left kidney. * But not quite so deeply as in Galictis. + As in many Carnivora, cf. e.g. Owen’s Comp. Anat. vol. ili. p. 496. 1905. | ANATOMY OF THE FERRET-BADGER. 25 In Helictis, on the other hand, this membrane is of much less extent. It is only attached to about half the length of the pancreas and is inserted on to the mesocolon along a line which begins a little before the left kidney but ends at about its middle. This characteristic difference is illustrated in the figures (text- Imes, Wh, 10), Text-fig. 9. P Pancreas and adjacent regions in Helictis personata. D. Duodenum ; P. End of pancreas; K. Kidney; S¢. Stomach. $ Ovary and Broad Ligament. As is very frequently, if not constantly, the case with the Arctoidea, the ovary is completely encapsuled and thus continuous, anatomically, with the Fallopian tube. An interesting point concerns the suspension of the ovary and oviducal canal. The mesoarium is continued forwards for a short distance in front of the ovary, running attached to the parietes to the outside of the kidney. In Galictis there is the same forward prolongation of this fold, which has the same position in relation to the kidney, but it extends much further forward on 26 MR. F, E, BEDDARD ON THE [May 2, both sides, in fact nearly to the diaphragm. In Cynictis levaillanti and Arctictis binturong, which I examined for purposes of comparison, the eonditions are a little different. In the former the fold of peritoneum in question runs over the kidney instead of avoiding it, and ends on the parietes a little way in front and outside of that gland. In the Binturong the mesoarium on the right side extends nearly up to the diaphragm, passing over the Text-fig. 10. Pancreas and adjacent region in Galictis barbara. Lettering as in text-fig. 9. kidney and being naturally attached to it on its passage. On the left side, this fold of peritoneum actually reaches the diaphragm, passing also over the kidney of its side. I will not assert at present that there are here characters which serve to differentiate the Arctoid from the Ailuroid Carnivora, but they do as a matter of fact differentiate certain Aluroids from certain Arctoids. 1905. | ANATOMY OF THE FERRET-BADGER. 27 § Arterial System*. The Aortic arch gives off first an innominate artery and then the lefé subclavian ‘separately. These matters are not mentioned Text-fig 11. Intrathoracic aorta of A. Helictis personata; B. Galictis barbara. Ao. Aorta; L. Branches to lung; es. Branches to cesophagus; v. Intercostals ; zg. Azygos vein; 7’. Branch to trachea; D. Phrenic arteries. by Garrod in his account of Helictis subaurantiaca, and indeed * The arteries of the brain are dealt with under the description of that organ. 28 MR. F. E. BEDDARD ON THE [May 2 he gives no account of the vascular system at all. The innominate first gives off the left carotid, and then very shortly after divides into the right subclavian and right carotid. The aorta in the thoracic region gives off eleven pairs of intercostal arteries, the Text-fig. 12. Intrathoracic aorta of Suricata tetradactyla. Lettering as in text-fig. 11. eleventh being just in front of the diaphragm. It is important to notice that these arteries are paired throughout, each artery of the pair arising separately from the aorta: important because in some mammals (e. g. Chinchilla) the intercostals arise as single 1905 | ANATOMY OF THE FERRET-BADGER. 29 arteries and afterwards divide into right and left halves. The first pair of intercostals corresponds to the first branch of the Azygos*, The first six pairs of intercostals lie entirely to the left of the Azygos; the 7th artery on the right side and those which follow le to the right side of the Azygos. This point, I take it, is where originally the now missing right aortic arch joined the left aortic arch. In this region the aorta also gives off a number of fine slender branches to the cesophagus and to the lungs. The first of these branches arises a little way down the first right intercostal and supplies the windpipe; from or in the immediate neighbourhood of the next four right intercostals ‘arise twigs for cesophagus and lungs; then follows a gap of two intercostals, the last twigs arising thon the 8th right intercostal. From the last intercostal in front of the diaphragm arises a diaphragmatic artery on each side; another diaphragmatic artery springs directly from the aorta behind the diaphragm, and independently of an immediately following suprarenal artery. I have carefully and, I hope, exactly compared the pre- diaphragmatic arteries of Helictis with those of its ally Galictis and with those of the Afluroid Swricata. The former, as might be expected, shows greater resemblances to FHlelictis than does the latter. There are, however, also difterences. There are 10 instead of 11 pairs of intercostals in front of the diaphragm, or, to be more absolutely accurate, 10 on one side and 9 on the other; for the first intercostal has not a fellow and belongs to the left side. The fifth right intercostal is the first which passes to the outside of the Azygos vein. The pulmonary and esophageal branches arise in every case from the right intercostal vessels, and I counted four of them which have the following position: the first three arise from the first three right intercostals; the fourth springs from the fifth right intercostal. In Suricata tetradactyla there are 12 intercostals on the right side in front of the diaphragm and two additional ones on the left side. The most important pulmonary and cesophageal arteries arise separately from the aorta, though some spring “from right intercostals. The 8th right intercostal is the first which passes over the Azygos vein. * This vessel, as in most mammals, is present only on the right side. 30 MR. W. P. PYCRAFT ON THE [May 2, 3. Contributions to the Osteology of Birds.—Part VII.* Hurylemide ; with Remarks on the Systematic Position of the Group. By W. P. Pycrart, F.Z.S., M.B.0.U. [Received March 30, 1905. | (Plate I]. + and Text-figures 13-15.) CoNnvTENTS. i. Introductory Remarks, p. 30. vil. The Pelvic Girdle, p. 48. ii. The Skull of the Adult, p. 30. viii. The Pectoral Limb, p. 48. ili. The Skull of the Nestling, p. 40. ix. The Pelvic Limb, p. 49. iv. The Vertebral Column, p. 43. x. Summary, p. 50. v. The Ribs, p. 45. xi. List of Literature referred to, p. 56. vi. The Sternum and Shoulder-girdle, xii. Explanation of Plate II., p. 56. p. 45. 1. LnrropucTORY REMARKS. The present paper is intended to form the first of a series on the osteology of the Passeres, and, in order to increase its value to the systematist, characters other than osteological will be discussed where necessary. By this means it is hoped that that most difficult of ornithological problems—the classification of the Passeres—will be materially aided. The labours of Garrod, Forbes, and Firbringer have resulted in the accumulation of a considerable pile of facts concerning the soft parts of the Kurylemide, but comparatively little has been done in the way of osteology. My work, it may be. as well to state here, has been hampered by paucity of material, since several genera are entirely unrepresented in the Collection of the British Museum (Natural History). Doubtless these gaps will be filled in course of time, and the lacune, unavoidable in this contribution, can then be filled up. Skeletons of nestlings are especially wanted. 11. THE SKULL OF THE ADULT. The skull of the Eurylemide is remarkable for the extreme specialisation which it displays, though these birds are of an undoubtedly primitive type. ‘That changes so considerable as are here to be noticed should have taken place in the skull is unfortunate, since thereby valuable evidence on questions of ancestry has been lost. It is not an easy matter to express exactly what are the characteristic features of the Eurylemid skull, or, rather, it is not easy to set down diagnostic characters, since it presents considerable and often wide differences in d werent genera. Superficially it * For Part VI. see P. Z.S. 1903, vol. i. p. 258. + For explanation of the Plate, see p. 56. H.Grénvold, del. OSTEOLOGY or P.Z.S ISOS pvGli0L iL J0 Bale & Danielsson, L* coll. THe BURY AMID zs. 1905. ] OSTEOLOGY OF THE EURYL/EMID#. 31 presents an undoubted resemblance, in some respects, to the aberrant Procnias, in others to the Swallows. The following characters will, however, probably suffice :— The beak is of great size, nearly as broad as long, and joins the cranium by a more or less perfect nasal hinge; free lachrymals are wanting, save in Calyptomena; palate egithognathous; palatines short, broad, wide apart, and produced backward into prominent spurs; vomer truncated, much reduced and terminating posteriorly in a pair of slender limbs; pterygoids and palatines articulating by means of an oblique joint; maxillo-palatine processes reduced to long slender rods slightly expanding at their termination beneath the vomer; basipterygoid processes wanting; postorbital processes obsolete ; squamosal process prominent The Occipital Region. The foramen magnum is cordiform, its apex rising only slightly above the level of the superior margin of the rim of the tympanic cavity. The plane of the foramen inclines downwards rather than backwards, as in the Capitonide, but not to such an extent asin the Bucconide. The base of the foramen is not raised above the level of the basi-cranial axis. The supra-foraminal ridge is barely traceable. There is no lambdoidal ridge, such as is met with in the Capitonidee for example, but the cranium above the occipital foramen presents a fairly prominent cerebellar dome, bounded on either side by a subcircular depression (the supraoccipital fossa). Above this region the skull rises considerably and presents a gently rounded surface. The tympanic wings of the exoccipital are considerably developed to form a pair of downwardly directed plates, the processus ale exoccipitalis inferior, having a convex border and a convex surface with recurved free edge: through these plates the semicircular canals can be faintly traced. The Cranial Roof (Pl. 11.).—The cerebral rises vertically above the cerebellar dome and is of considerable width, being wider than long. In regard to the position of the cerebral with relation to the cerebellar dome, the Kurylemide agree with the typical Passeres and the Cypseli, and differ from the Capitonide, for example, wherein the cerebral les tm front of the cerebellar dome. The parietal region is marked by a moderately well-defined temporal depression, the “‘ temporal fossa,” which, however, does not extend further inwards than the outer margin of the supraoccipital fossa. This is a Passerine feature; in the Coraciiformes these fossee usually meet in the middle line, forming a more or less well- marked sagittal crest. The temporal fossee in the Eurylemide are mainly responsible for the formation of the well-marked squamosal prominences. The interorbital region is marked with a more or less distinct median groove, sometimes with alow ridge. Immediately behind 32 MR. W. P. PYCRAFT ON THE | May 2, the base of the beak it expands considerably and is supported from within by outstanding antorbital plates. Lachrymals, except in Calyptomena, are absent, and consequently take no share in the formation of the preor bital region of the skull. In this particular the Kurylemide agree with the bulk of the Passeres, in which, however, vestiges of the lachrymal are frequently present. The Re ontalsy terminate abr uptly i in front, not extending beyond the level of the anterior border of the mesethmoid. The nasals and nasal-processes of the premaxilla are also sharply truncated caudad ; thus, at their meeting with the frontals and mesethmoid a freely moving nasal hinge is formed (Pl. IT. figs. 26, 3a, 4). The incipient stages in the development of such a hinge can be studied in Chasmor hynchus—one of the Cotingide. The Base of the Skull. The basitemporal plate is shghtly hollowed in the middle line, and is continued forward for some distance on to the parasphenoidal rostrum; owing to the small size of the brain its free edge projects beyond the level of the brain-case. To appreciate this point the skull of one of the Eurylemidz should be compared with say that of Menwra or Corvus, where, it will be found, the basitemporal plate fails to conceal the brain-case when the skull is seen from below. The edge of this plate is free only at its apex. Not even vestiges of the basipterygoid processes remain. The parasphenoidal rostrum is long and slender. The occipital condyle is spherical and depends from the roof of a shallow pre-condylar fossa. The Lateral Aspect of the Cranium. (Pl. IT. fig. 2.) The tympanic cavity is small, shallow, and has little or no floor, The constriction of the skull-wall in the temporal region, to form the “temporal fosse,” gives the tympanic cavity the appearance of the aperture of a tube, the cylinder of which is formed by the ‘¢squamosal prominence” and lateral occipital wing. The roof of this cavity is formed by the under “surface of the processus 2 gomaticus squamosi. Its floor in part by the lateral occipital wing and in part by the ossification of tissue extending between this wing and the external angles of the basitemporal plate; but this region is much cut away. Within the cavity three apertures will be found in the dried skull after the removal of the tympanic membrane. The largest of these is the mouth of the recessus tympanicus anterior. Immediately outside this, and below the otic articular surface for the quadrate, is the fenestral recess : this is very small, and neither the fenestra ovale nor the rotunda can be distinguished within it, though the columella is in position above; and behind the fenestral recess a cluster of minute pneumatic apertures will be found, corresponding to a similar group commonly found in the higher 1905. ] OSTEOLOGY OF THE EURYLEMIDA, 33 Passeres. In shape and position, however, this group of foramina more nearly resembles its counterpart in ‘the Bucconidee. These foramina form a sort of cribriform plate guarding the mouth of the recessus ty ympanicus posterior, which is much reduced. The recessus tympanicus superior is of small size, and opens externally into the tympanic cavity by a small aperture lying between the squamosal and otic heads of the quadrate. The aperture is bounded externally by a short, pointed processus articularis SqUamosi. The Squamosal Prominence.—It has already been pointed out (p. 32) that the constriction of the temporal region of the cranium has given the tympanic region a sort of individuality not met with in the skulls of the higher Passeres, but common among the lower types, and among the Coraciiformes, In the Huryleemidz the free edge of this prominence projects shelf-like beyond the head of the quadrate. It is continued forwards into a hastate processus zygomaticus squamosi directed downwards and outwards. From the base of the inferior surface of this process projects a short. pointed processus articularis squamosi; between these two processes the head of the quadrate is firmly grasped. The temporal fosse are especially deep in Corydon. As in other genera, they are linguiform in shape and do not extend inwards beyond the outer border of the supraoccipital fossa. The trigeminal foramen pierces the skull-wall at about the level of the otic articular process for the squamosal, but some con- siderable distance mesiad thereof. The orbito-sphenoid does not ossify. The inéerorbital septum 1s largely fenestrated. The interorbital region of the frontals is generally very narrow so that the orbits are only very partially roofed. In front the orbit is bounded by a f--shaped antorbital plate. In Calyptomena the interorbital region is wide. The lachrymal, in Calyptomena (Pl. II. fig. 2, U.), has the form of a sigmoid rod more or less clubbed at each endl The upper end would perhaps more correctly be described as hammer-shaped, and is completely overshadowed by wide expansions of the frontal. The whole ossicle is embedded in a groove carved out of a very much swollen antorbital plate. The close resemblance between the lachrymal of Calyptomena and that of Chasmorhynchus is most remarkable. Both are embedded in the antorbital plate, and both have the same sigmoid flexure. Only in the larger size of the orbital end can the lachrymal of Chasmorhynchus be distinguished from that of the HKurylemid Calypiomena. In all the other Eurylemide, however, the lachrymal appears to have been lost; further, the antorbital plate has been reduced to a thin e-shaped plate. The Ethmoidal Region.—The mesethmoid is greatly reduced by the fenestration of the interorbital septum. The antorbital plate which bounds the orbit in front is -shaped and attached to Proc, Zoot, Soc,—1905, Vou, Il. No. ILI. 3 34 MR. W. P, PYCRAFT ON THE [May 2, the mesethmoid by a horizontal plate of bone, almost rod-like in some species. “The vertical, hamulate portion of the plate, by its upper limb, considerably adds to the width across the frontal, the lower, descending, process turns outwards to reach the quadrato-jugal bar. In the extraordinarily wide-mouthed genus Corydon, however, the quadrato-jugal bar stands far from this descending process. The olfactory chamber, owing to the extremely reduced con- dition of the maxillo-palatines, in the macerated skull is without a floor, in the majority of the genera of this group; but in two skulls, Zurylemus and Cymbirhynchus, in the British Museum Col- lection, this is more or less filled up by the ossification of a pair of turbinals, one on either side of the septum nasi, which apparently answer to the concha media. Pyriform in shape, each extends from the narial aperture backwards to the anterior horn of the vomer, where it becomes attached. Above and behind this is an oat-shaped and laterally compressed turbinal answering to the concha posterior. The nasal septum, in Calyptomena, is formed by a thin sheet of bone running along the whole length of the under surface of the nasal process of the premaxille. In Hurylemus, Cymbirhynchus, and especially in Corydon, this septum becomes greatly swollen and grooved on its under surface. The Cranial Cavity.—The mesencephalic fossa is capacious. Tts floor sweeps rapidly upwards to form astrongly marked basin- shaped cavity. This upward rising of the floor is much more conspicuous than in some other genera, e. g. Menura or Corvus. The internal auditory meatus is represented only by a shallow depression. Immediately above and somewhat in front of this lies the trigeminal foramen. ‘This, opening under a strong ridge, leads immediately into a deep groove across the floor of the mesencephalic fossa and thence through the under wall of the skull. All the branches of v leave by this foramen. There is no separate foramen for the ophthalmic (v’) (orbito-nasal); and in this respect the Hurylemide appear to agree with all the other Passeriformes. The vagus foramen lies at the bottom of a deep fossa. The cerebellar fossa is small, relatively to the cerebral, sharply defined, and has the supra-occipital region marked with prominent horizontal ridges. The floccular fossa forms a conspicuous moderately deep and more or less pyriform depression, sharply bounded caudad by the anterior semicircular canal. The mesencephalic fossa is of considerable size and, as in other Passeriformes, extremely well defined by a strong vertical ridge above, and an equally prominent ridge formed by the pro-otic below. The pituitary fossa takes the form of a narrow tube rising vertically from the floor of the skull. The dorsum selle is reduced to a knife-like edge. The pre-pituitary region is produced into a moderately well-defined optic platform, triangular in shape, 1905. ] OSTEOLOGY OF THE EURYLEMIDA. 35 The cerebral fosse are relatively of considerable size, though relatively smaller than in Corvus for example. Thus, in the Hurylemids the cerebral fossa is only distinguishable from the mesencephalic fossa by reason of the boundary-line of the tentorial ridge. In Corvus the mesencephalic fossa forms a totally distinct basin-shaped cavity, lying as it were within the cerebral fossa, which dips down to the outer side and below the level of the fossa in question in the form of a deep pocket. Menura represents a half-way stage between the Corvide and Kurylemidze. In Menura, moreover, the roof of the cerebral fossa is marked by a low ridge roughly dividing the fossa into two equal parts. There is a well-developed bony falx. The olfactory fossee are reduced toa pair of small pits. But there are strong impressions of an olfactory tube to be found in the fore part of the cerebral fossa of the Eurylemide. The Premaxilla. The premaxilia in the Eurylemide forms the major part of the upper half of the beak. Hooked at the tip, and of extreme breadth, it recalls in many respects that of many of the Coraciide, e. g. Hurystomus, on the one hand, and of some Caprimulgi, e. g. Podargus, on the other. When these several types come to be compared, however, these resemblances will be found to be but slight. More significant is the close resemblance to the Cotingide. This is well brought out in the skull of Calyptomena, which, as will be shown presently, presents many features in common with Chasmorhynchus. 'The number of other skeleton characters which these two forms possess in common suggest affinity between the two groups, rather than homoplasy. In Calyptomena, which I propose to take as the typical Kurylemid for the purpose of comparison, the body of the pre- maxilla is moderately large. The nasal process, fusing with the nasals, is sharply truncated caudad, and articulates with the frontals by a hinge. In this respect the Kurylemid skull re- sembles that of the Podargide, and not of the Coraciide. The low position of the skull of Calyptomena is indicated by a comparison of the narial aperture with that of the skulls of other Eurylemid genera. In Calyptomena the nasals are of the typical holorhinal shape with an obliquely sloping descending process. The nasal fossa, in the dried skull, is a long oval aperture showing, within the cavity, a narrow ridge of bone continued from the palatal border forwards to meet a low septum hanging from the middle line of the nasal process of the premaxilla. The septum represents the ossified remains of the septwm nasi; the small plate of bone running inwards from the level of the tip of the palatine is a portion of the alinasal cartilage which has become ossified. 3* 36 MR. W. P. PYCRAFT ON THE [May 2, Corydon, Hurylemus, and Cymbirhynchus differ conspicuously from Calyptomena in this matter of the narial aperture, as may be seen by a comparison of figs. 26, 3a, 4, Pl. IT. Tn all three genera the nasal is reduced to its smallest possible limits, little more than an arcuate bar being left. Of this, one half represents the descending process of the nasal, the other the body of the bone,—now merely a rod joined at its inner end to the nasal process of the premaxilla, and affording the means of articulation with the frontals. This, as I have remarked, takes the form of a nasal hinge. In Hwrylemus and Corydon the nasal fossa, as in Calyptomena, is open in the dried skull, the actual position of the nasal orifice in the living bird being indicated by semicircular grooves in the anterior border of the nasal fossa. The circle completing the rest of the fossa in the living bird was roofed by the alinasal wall. In Cymbirhynchus this wall almost completely ossifies, leaving an oval narial aperture, and a small semilunar space immediately in front of the nasal (fig. 3 Pl. IT.). In the Kurylemide the floor of the olfactory chamber is open behind, revealing in Calyptomena an ossified sheet-like nasal septum, which in Corydon becomes immensely swollen. In the Coracie, certain Caprimulgi and Pici the floor of the olfactory chamber is more or less ossified. In Hurystomaus and the Bucconide there is a long palatal fissure, which at first sight appears to correspond to the huge palatal cavity of Hurylemus. An examination shows, however, that this vacuity leads into a spacious cavity underlying the olfactory chamber and formed by the inflation and absorption of tissue of the nasal septum. In Podargus the palatal surface of the premaxilla is completely ossified, and the olfactory chamber is reduced to the smallest possible limits. In Corydon the nasal process of the premaxilla is immensel swollen and rises far above the level of the nasal hinge. The frontal is similarly swollen immediately above this hinge. The intermediate stages between this condition and that found in Calypiomena can be studied in Cymbirhynchus. Corydon, indeed, would appear to have reached the high-water mark of speciali- sation in the matter of the jaws, among the Eurylemide. b) The Ma«illo-jugal Arch. The maxilla, as usual, is in the adult completely fused with the premaxilla. In Corydon the maxillary region of the jaw is highly developed and forms a large semicircular plate, the convexity forming its free edge and projecting downwards far beyond the level of the quadrato-jugal bar. The mawillo-palatine processes in Calyptomena take the form of a pair of delicate rods projecting backwards at a very marked angle from the body of the maxilla, which, at this point, is perforated by small pneumatic apertures. These rods, on each 1905. | OSTEOLOGY OF THE EURYLEMIDE. 37 side of the skull, terminate immediately beneath the free end of the vomer. That these processes are degenerate there can be no doubt. They have probably been derived from a condition precisely similar to what obtains in Chasmorhynchus. In the latter, these processes are swollen and spongy in character. Arising from the maxilla at a point almost immediately below the descending process of the nasal (in Calyptomena they arise distad of this point), they extend backwards so as to run on either side of and beneath the vomer for nearly one-fourth of its length. In Corydon and Cymbirhynchus these processes are more slender than in Calyptomena. In Cymbirhynchus they are hook- shaped. Probably, as I have remarked, the maxillo-palatines of Calypiomena at an earlier stage closely resembled those of Chasmorhynchus. It seems also highly probable that these, in turn, were derived from yet more primitive and much more extensive triangular plates such as have been retained by the ‘Tyrannide. The palate of Vityra, indeed, shows how easily the Eurylemiform palate could have obtained its peculiar maxillo- palatines. The quadrato-jugal bar in Calyptomena as in Chasmorhynchus is sigmoidally curved, as much so as in some Spheniscide. In Corydon and Cymbirhynchus it is straight. There are no separate elements distinguishable in this bar. The Vomer, Palatines, and Pterygords. The vomer (Pl. 11. fig. 2a), in Calyptomena, is roughly oar- shaped in front and terminates caudad in a pair of long, slender limbs, bowed outwardly so as to enclose a space through which the parasphenoidal rostrum may be seen, and fused completely with the palatines. The free end of the blade is truncated, and has the angles produced into minute processes, thus showing that the vomer was earlier of a more pronounced Aigithognathous type. In Corydon the vomer is much reduced, being represented by a short, broad, oblong body produced caudad into a pair of widely separated and slender rods which articulate with the palatines. The free end of the vomer is squarely truncate with prominently produced angles. The dorsal aspect of the vomer is closely applied to the base of the septum nasi. Oymbirhynchus resembles Corydon in the shape of the vomer, but differs therefrom in that it is slightly constricted between the free end and the origin of the posterior cornue, which fuse com- pletely with the palatines, forcing the parasphenoidal plates thereof away from their normal relationship to the parasphenoid. In the Coraciidee the vomer is either wanting or reduced to a mere spicule, e. g. Hurystomus. In Chasmorhynchus the vomer is larger than in the Kurylemide. Aigithognathous anteriorly, it terminates posteriorly ma a pair of 38 MR. W. P. PYCRAFT ON THE [May 2, broad limbs indistinguishably welded with the palatines, agreeing in this with Calyptomena and Cymbirhynchus. The palatine (Pl. LI. fig. 2a) in Calyptomena is a long bone: anteriorly rod-shaped, it extends backwards as far as the under surface of the antorbital plate, when, after sending out- wards a prominent, rounded elbow—“ transverse bone ”-—it turns abruptly inwards, ultimately forming a roughly spatulate plate, bent upon itself so as to form a long linear surface running along the parasphenoid rostrum, and fusing mesiad with the vomer and a free downwardly hanging curtain to form a cavernous space in the roof of which is the base of the vomer. The palatine of Chasmorhynchus differs from that of Calypto- mena in the greater width of the hinder laminated portion and the more extensive development of the inferior free edge, forming the cavernous space beneath the vomer. This edge now appears rather as a shelf-like projection developed from the inner border of the shaft of the palatine. In Cymbirhynchus the palatine shaft is broader than in Calyptomena, and this increased breadth is especially noticeable at its junction with the body of the premaxilla. The latter, as has already been pointed out, is much more conspicuous than in Calyptomena and terminates in a doubly crescentic free edge synchronously with the palatines. Thus a relatively enormous oblong space is enclosed. In Corydon these features are still more exaggerated, the “‘ elbow” is also more strongly marked. In broadness and the truncated form of the shaft of the palatines, the more specialised Kurylemide recall the Podargide, wherein the body of the premaxilla is still more developed and the truncation of the palatine distally more marked. Both in the specialised Hurylemide and the Podargide the truncation of the vomer appears to have been brought about to facilitate the move- ments of the nasal hinge, which in both types les immediately above the anterior ends of the palatines, while in the more generalised Calyptomena, which lacks a nasal hinge, the palatines run far forwards. The pterygoid in Calyptomena (PI. II. fig. 2a) is a long, slender, rod-shaped bone, laterally compressed, and perforated by a pneumatic foramen at its articulation with the quadrate. At its anterior end it meets its fellow of the opposite side in the middle line; and immediately sends upwards and forwards a sub- crescentic plate which, embracing the parasphenoidal rostrum by its plane surface, affords attachment along its inferior border to the vomer. Late in life the articulation with the vomer is succeeded by anchylosis. Certain points concerning the morphology of the end of the pterygoid will be discussed in the section dealing with the nestling skull (p. 43). In Chasmorhynchus the anterior ends of the pterygoids do not meet in the middle line but impinge instead directly against the parasphenoidal rostrum, forming therewith a pedate articu- lation, which is largely augmented by “ hemipterygoid” elements 1905. ] OSTEOLOGY OF THE EURYLEMIDS. 39 corresponding to the sub-crescentic plates of Calyptomena. These hemipterygoids, in both the Eurylemid and Cotingid forms, articulate with the palatines, in the adult, by means of an oblique suture. Corydon and Cymbirhynchus difter in no essential features from Calyptomena. The quadrate is peculiar in that, in common with the Tyrannide and some other Passerine forms, it sends out a strong spur for the articulation of the quadrato-jugal bar. This spur projects like a buttress laterad of the outer condyle for the lower jaw. The squamosal and otic heads are closely approximated. The former is wedged in between a prominent processus articularis squamosr behind, and an equally well-developed processus articularis zygo- maticus in front. In Corydon these processes are expanded laterally so as to overhang the head of the quadrate, but at the same time they afford this element a greater freedom of movement than in Calyptomena. Cymbirhynchus and Eurylemus are inter- mediate in character in this respect. The Mandible. The mandible, in the Eurylemide, is much bowed outwards to a very considerable extent. In Calyptomena it has only a relatively small symphysis: is truncated posteriorly, and shows little or no trace of the separate elements of which it is composed. The internal angular process is moderately well developed, and is perforated by a small pneumatic foramen. The rami, in their general shape, are rod-like, and slightly compressed laterally. In Corydon and Cymbirhynchus, however, there is an abrupt transition between the malar region of the mandible and that portion covered by the rhamphotheca, which is most markedly thicker and broader than the hinder region. The symphysial region is very broad and spoon-shaped. ‘The internal angular process is more spine-like than in Calyptomena, and there is a feebly-developed posterior angular process. The Hyoid. The hyoid of the Eurylemide resembles that of the higher Passeres. The basihyal (os entoglossum) is made up of a pair of boomerang-shaped ossifications placed dos @ dos, so that a long free process is produced backward beyond the articulation with the basibranchial 1. Basibranchials 1-2 are fused; the latter, however, is a long cartilaginous style. The ceratobranchial and epibranchials are of moderate length; the latter are cartilaginous at the free ends. In Corvus, for example, among the higher Passeres, the basi- hyals are long and straight, and run parallel with one another, yet so as to leave a median space between them. 40 MR. W. P. PYCRAFT ON THE [ May 2, iii. THE Skuxt or THE Nustiine. (PI. IT. figs. 1 & 1a.) It is a matter for regret that the British Museum Collection of nestling skulls of Hurylemide is limited to half-grown specimens of Hurylemus ochromelas, and these have suffered somewhat severely as a consequence of having been preserved in formol. a. Cartilage-bones. The basioccipital cannot, in these skulls, be more than imper- fectly traced, having become fused with the lateral occipitals. The exoccipital, or lateral occipital, is a large, more or less linguiform plate presenting a broad convex external border, the inferior segment of which forms the tympanic cavity, while the superior arc of the curve is applied in part to the base of the squamosal and in part to the parietal. It is bounded mesiad by the supraoccipital and the occipital foramen, which excavate a considerable moiety from its internal border. Inasmuch as the exoccipital comes into contact with the parietal, it resembles that of the Cuculide. The supraoccipital is short antero-posteriorly, and is not yet ossified ; its superior margin being W-shaped and leaving a large fontanelle between itself and the parietals. Laterad it has fused with the lateral occipitals, leaving only a faint tell-tale notch to indicate the junction. The pro-, epi-, and opisthotic bones are now completely concealed when the skull is viewed externally. The basisphenoid is also concealed, being underfloored by the basitemporal plate. The alisphenoid appears as an oblong plate, having its long axis horizontal. The orbito-sphenoid is still membranous, while the presphenoid has fused with the basisphenoid. The mesethmoid has only just commenced to ossify, and is represented by a small linguiform plate supporting the yet carti- laginous antorbital plate, and bounded in front by the cranio- facial fissure. The interorbital septum formed by the backward extension of the plate is as yet only outlined in cartilage. The olfactory cavities occupy less than half of the so-called anterior narial apertures as seen in the dried skull. The actual anterior nares, in Hurylemus ochromelas for example, are small and round, and placed at the extreme anterior angle formed by the divarication of the nasal and maxillary processes of the premaxilla. The superior segment of this circle is formed by membrane, and this extends backwards and inwards as a subtubular sheet to be attached to the antorbital plate. Mesially this tube is shut in by the nasal septum, and inferiorly by membrane forming the roof of the palate.. Within the chamber thus formed lies a long, somewhat spatulate cartilaginous turbinal extending backwards, by a short stalk, to the anterior end of the vomer, ‘The free end 1905. | OSTEOLOGY OF THE EURYLEMID. 4] of the spatulate process lies on a level with, but mesiad of, the external aperture. Without this chamber is a large sinus roofed by the rhamphotheca, floored by membrane supported by the maxillo-palatine process, and closed posteriorly by the antorbital plate lying external to the nasal chamber. In the dried skull this sinus is included as part of the external narial aperture. The guadrate, though not yet completely ossified, differs in no material particular from that of the adult. b. The Membrane-bones. The parietal is roughly quadrangular in shape; its superior external angle is drawn upwards into a point, its inferior external angle forms a sweeping curve. Its mesial border is not yet ossified in the skull now described. A small portion of its inferior border, lying between the supraoccipital and squamosal, comes into actual contact with the exoccipital. The frontal along its posterior border follows the curve of the parietal: anteriorly, in the mid-orbital region, it becomes reduced to a narrow band, and finally terminates in a strap-shaped process underlying the nasals. Before leaving the cranial cavity its free edge passes downwards and inwards to join the alisphenoid inferiorly. The rim of this inturned plate is overlapped by a long tongue-shaped process of the squamosal (PI. II. fig. 1 a). The sguamosal is a somewhat remarkable bone. Roughly i-shaped, the horizontal region overlaps, mesiad, the lateral occipital and extends so as nearly to reach the supraoccipital ; laterad it overhangs the tympanic cavity and terminates in a pointed processus zygomaticus squamost. ‘The vertical shaft arising from this base is roughly sword-shaped, with a slightly decurved pointed tip. About one-third of this blade arises above the level of the parietal to overlap the frontal as already described. Immediately above the level of the superior border of the ali- sphenoid this blade develops a barely perceptible prominence, which supports a small cartilaginous nodule—the anlage of the postorbital process. Another most noteworthy feature of the squamosal in this skull is the fact that the greater part thereof appears on the inside of the skull: only, indeed, the extremities of the horizontal and vertical portions being excluded. Compare figs. 1, 1@ (PI. I1.). In the most primitive types of Avian skull, it will be re- membered, the squamosal is either entirely excluded from any participation in the formation of the brain-case, or only a very small area is admitted. Originally a quite superficial bone, it has gradually absorbed the underlying osseous tissue, till eventually it has forced itself into the very walls of the cranial cavity, and this is especially the case in the skull of Hurylemus ochromelas. I am unfortuately unable at the present time to make any extensive series of comparisons between the form of the squa- mosal in the Hurylemide and that of the Coraciiformes, or the 42 MR. W. P. PYCRAFT ON THE [May 2, Menuride and other Passeriform types, owing to lack of material. Such a comparison I believe would be valuable. So far, the peculiar squamosal of the Eurylemide resembles most nearly, among the Coraciiformes, that of the Capitonide. But the likeness is but general, and seems to point to the Capi- toniform type as being the more primitive. Herein, this element is roughly quadrangular in type, but has the antero-dorsal angle produced into a point, which, however, does not extend on to the frontal. Its mesial border is, indeed, exactly coterminous with the external lateral border of the parietal. The alisphenoid in this skull is prominent and forms a large triangular block fitting into the deeply concave anterior border of the squamosal on the one hand, and overhung by the postorbital region of the frontal on the other. The postorbital process appears to be formed in part by the alisphenoid, and in part by the frontal. The squamosal takes no part whatever in its formation. These relations can be seen in the skull of Calorhamphus. The resemblance to the squamosal of the Passeriformes is close, but is of a kind such as to leave little doubt but that this element in the Eurylemid is much the more specialised: a fact which is somewhat surprising, and is at the same time not without significance. Comparing the squamosal of Hurylemus ochromelas with that of the Rook (Corvus frugilegus), it will be found that in the latter this element is of considerable size, conical in form, and rises superiorly to overlap the frontal as in Hurylemus. The base of this cone is broad, and its postero-internal angle is produced backwards and inwards to form a wedge between the parietal and lateral occipital. It is from a squamosal of this type that the squamosal of Hury- lemus has been derived. This evolution has resulted in a much greater extension of the base mesiad, between the parietal and exoccipital, and in the lateral reduction of the body of the bone so as to transform the sometime cone into a xiphoid shaft springing from a broad base. These changes will become the more apparent by a reference to fig. 1 a, Pl. IT. The nasal varies considerably in form in this group. Unfor- tunately, I have not material at my command which will enable me to make a comparison of the early stages of growth of these several varieties. The lachrymal is not yet ossified. The premaailla apparently lacks palatine processes. What appear to be vestiges of these seem rather to be ossifications of the mem- brane forming the floor of the anterior region of the nasal chamber. This point can only be solved by a further examination of well- preserved material. The masxilla appears to be unusually large in the skull, but the decalcification caused by the formalin in which this specimen was preserved has almost obliterated the premaxillary and quadrato- jugal sutures, 1905. ] OSTEOLOGY OF THE EURYLEMID&. A3 The quadrato-jugal is long, extending to beyond the middle of the orbit. The vomer is not yet ossified. The palatines difter from those of the adult in that the trans- palatine elements (?) are as yet membranous. The pterygoid is rod-shaped, bent at its posterior extremity at an obtuse angle, so as to fit closely to the parasphenoidal rostrum. The free end of this rod is pointed, and bears a small pointed piece of cartilage. Whether this represents the hemipterygoid, or, as seems more likely, the unossified extremity of the shaft, is a point which can only be determined by the examination of somewhat older skulls. The palatine extends backwards beneath these bent limbs of the pterygoid. The apparent absence of the hemipterygoid is a point of con- siderable interest. The interpretation to be placed upon this fact is, I think, not that the pterygoid shaft retains its primitive integrity, but that the hemipterygoid element has been Jost, just as it has in many other groups of birds. My reason for this view is that the vomer, which shows various grades of reduction in the Eurylemide, is supported entirely by the palatines, as in all other cases where the hemipterygoid has been greatly reduced or is wanting. There is nothing remarkable in the absence of this element, because, as has been shown, the skull in this group is highly specialised in many ways. The elements of the mandible are as yet distinct. iv. THE VERTEBRAL CoLUMN. All the presynsacral vertebree are heteroccelous and free. The cervical vertebre are characterised by the deeply incised neural plates of the 6th—10th vertebrae, where the posterior zygapophyses are borne upon the under surface of the free ends of long beams. The atlas has the odontoid ligament perforated. The axis bears a large tooth-like neural spine and a large pair of hyperapophyses. The second and third have large quadrangular neural plates, the hinder angles of which in the third vertebra are produced upwards into strong hyperapophyses. The outer borders of these plates are pierced, on each side, bya smallforamen. The hyperapophyses of the 5th to 8th vertebre are placed about midway between the neural spine and the posterior zygapophysis. From the 5th to 11th vertebra the neural plates are deeply incised both before and behind the neural spine. The neural spines gradually decrease in size from before backwards, so that from the 9th to the 12th they are represented only by the merest tubercle. Hypapophyses are borne by the 2nd, 3rd, 4th, and 5th vertebre ; the 7th to 10th bear catapophyses, feebly developed ; hypapo- 44 MR. W. P. PYCRAFT ON THE [May 2, physes again succeed from the 11th vertebra and are continued backwards to the thoracic. The cervical vertebrae are 12 in number. There are three cervico-thoracic; that is to say, there are three vertebre bearing free cervical ribs. The Ist pair are reduced to the merest vestiges ; the 2nd pair are long, bear vestigial uncinates, but no sternal segment; the 3rd pair bear large uncinates and a long styliform sternal segment, which does not, however, reach the sternum. Thus, it is obvious that, at no distant date, these three vertebre formed part of the thoracic series and articulated with the sternum. ‘They differ, moreover, in form from the true cervicals, and resemble the thoracic series in having broad outstanding diapophyses. The cervical and cervico-thoracics of the Eurylemide differ conspicuously from those of the Menuride and of the Coracii- formes, and resemble rather those of the higher Passeres. The thoracic vertebree, six in number, have moderately developed, quadrangular, neural spines. Only the Ist thoracic bears a small hypapophysis. The centra are pierced by pneumatic foramina. The last thoracic has been incorporated with the synsacrum. Twelve vertebrz enter into the composition of the synsacrum in Calyptomenaand Cymbirhynchus; 13in Corydon. The numerical differences are as follows :-— Calyptomena. Cymbirhynchus. Corydon. Mnoraciel.....: 1 1 1 umbar ...... 2 2 3 Lumbo-sacral 3 2 2 acral ease ee 2 2 2 Caudaleeeneess. 4+ 8freecaudal 5+8free caud. 5+ 8free caud. Total ae: 12+8 uF 1248 = 13+8 on Thus Calyptomena appears to have lost 1 post-sacral and Cymbirhynchus 1 pre-sacral. Corydon would appear to express the primitive number of these segments. The 2nd lumbar in Calyptomena bears a large pair of ventri- lateral processes abutting against the pre-ilia. The Ist sacral vertebra lies immediately caudad of the hinder margin of the acetabulum. The dorsi-lateral processes of the sacral and caudal vertebre are long, and, by the ossification of the tendinous tissue overlying them, form a broad bony plate dividing the innominates. There are 8 free caudals, including the pygostyle. The diapophyses of those immediately following the synsacral series are not embraced by the innominate, owing to the fact that these are kept apart by the outstanding dorsi-lateral processes of the synsacral series. Corydon and Cymbirhynchus differ from Calyptomena chiefly in that the dorsi-lateral processes of the sacral and post-sacral 1905. | OSTEOLOGY OF THE EURYLEMID&, AD components of the synsacral vertebra are shorter, so that the diapophyses of the first precaudal are overlapped by the hinder ends of the innominate. Well-marked intercentra appear on caudals 5-7, in each of the three genera here described. v. Tur Rips. The cervical ribs extend from the 4th to the 12th vertebre. The Ist and 2nd, 10th and 12th, are represented by little more than broad pleurapophysial lamelle; in the remaining vertebre, however, these lamelle are narrow and band-like, and the ribs slender and styliform, extending the whole length of the centrum. There are 3 cervico-thoracic ribs. ‘The first is reduced, only the capitulum and tuberculum remaining connected by a common base. The second is long, but bears no uncinate ; the third bears a short sternal segment, which, however, does not reach the sternum. There are 5 thoracic ribs, of which 4 only reach the sternum. The uncinates are long and extend backwards to reach the 3rd rib from their base of attachment. In Corydon the sternal segment of the 5th rib articulates by a special facet with the sternal segment of the rib next in front. vi. THE STERNUM AND SHOULDER-GIRDLE. (Text-figs. 13, 14, pp. 46, 47.) The sternum of the Hurylemide is typically Passeriform, and presents no very close resemblances to that of any other group. In some features it recalls that of the Cuckoos, in others of the Caprimulgi, but these are not of a nature likely to cause difficulty in confounding the sterna of either of these groups with Passerine sterna. In the Kurylemide the corpus sterni is short and broad—the breadth nearly equal to the length. The posterior lateral processes are long, extending forwards to a point corresponding to a trans- verse line through the middle of the corpus sterni; the free ends of these processes are spatulate. The metasternum has its free border squarely truncate, so as to form a continuous line with the free ends of the posterior lateral processes; broken only by the notch enclosed by this process. ‘The anterior lateral processes in Calyptomena have their free ends truncated and curved slightly backwards. In Corydon these processes are long, pointed and directed forwards. Cymbirhynchus is intermediate in this respect, the process being long, truncate, and directed forwards. The articular surfaces for the sternal ribs are confined entirely to these processes, There is no spina interna. The spina externa shows only the faintest indication of the bifurcate free end which prevails among the Passeriformes (text-fig. 13, s.e.). In Cymbirhynchus and Corydon this spine is triangular in section; in Calyptomena it 46 MR. W. P. PYCRAFT ON THE [May 2, should rather be described as blade-shaped, the dorsal edge of the blade being much thickened ; in other words, the ventrally placed keel of the triangle seen in Corydon has in this genus extended downwards, The median line of the dorsal surface of the corpus sterni is deeply grooved, and pierced anteriorly by a large pneumatic foramen. In Cymbirhynchus, and to a less extent in Calyptomena, the groove is laced across by narrow, irregular bars of bone. The carina is deep, and has the free (ventral) edge produced forwards, The anterior (vertical) border bears a hollow groove for the reception of the hypocleideum. The coracoid grooves look directly forwards, and do not meet in the middle line. The dorsal lips are prominent and thickened ; the ventral lips are well-defined, but have knife-like edges con- tinued inwards on to the spina eaterna. Text-fig. 13. Sternum of Calyptomena, showing the simple (unbifurcated) spina externa, s.e. a.l.p., anterior lateral process; c., carina; p.l.p., posterior lateral process. The coracoids are long; as long as, or longer than, the corpus sterni. The procoracoid process, though reduced, is still moderately large, and forms a narrow flange of bone, arising beyond the middle of the coracoid shaft and having its free edge directed downwards, Cephalad it articulates with the scapula, and with the clavicle forms the foramen triossewm. The procoracoid of the Eurylemide is larger than in the Cotingide. In the Picide the procoracoid appears to be wanting, and in the higher Passeres it is reduced to the merest vestige, e.g. Corvus. The breadth of the base of the coracoid is increased by narrow phalanges of bone, one on either side extending forwards, for about one-fourth the length of the shaft. On the dorsal aspect of the shaft a prominent tongue of bone is sent up to abut against the dorsal lip of the coracoid groove. This at least obtains in Calyptomena; in Corydon and Cymbirhynchus it is less marked. The scapula is long, narrow, and scimitar-shaped. The furcula 1905.] OSTEOLOGY OF THE EURYLHMIDA. AT is long, slender, gently arched, and bears a large hypocleideum, roughly quadrangular in form, and articulating with the anterior edge of the carina. The free ends of the furcula are expanded to form flat plates articulating with the acrocoracoid, procoracoid, and acromion process of the scapula (text-fig. 14), thus enclosing the JSoramen triossewm. In the form of the sternum and shoulder-girdle the Eurylemide closely resemble the Cotingide, especially in the form of the spina externa, which is simple, and thereby differs from the typical Passerine form wherein it is forked. In the Eurylemide this process is more or less spike-shaped, whereas in the Cotingide it appears to be generally flabellate. In Chasmorhynchus the posterior sternal notches are not so deep as in the Eurylemide. The hypocleideum articulates nearer the antero-ventral angle of the carina; and the coracoids do not develop the internal basal flange found in the Eurylemide. Text-fig. 14, Portion of the shoulder-girdle of Calyptomena, to show the meeting-point of the scapula, coracoid, and clavicle, forming the inner wall of the foramen triosseum. se., scapula; ac., acrocoracoid ; f., furcula. In the relations of the articulations between the procoracoid, furcula, and scapula, where these unite to form the foramen triosseum, the Hurylemide are distinctly Passerine, though this arrangement also obtains among the Picide—a fact of some significance. In Calyptomena, for instance, the acromion process of the scapula extends downwards along the anterior border of the free edge of the procoracoid, and affords an articular surface cephalad, for the posterior angle of the expanded free end of the fureula. In Chasmorh ynchus this articulation for the furcula is markedly increased ; and this increase apparently reaches its maximum in the Corvide, where the acromion forms a long beam-like roof to the foramen triosseum, and a very extensive articular surface for the furcula, \ 48 MR, W. P. PYCRAFT ON THE [May 2, vil. Tae Petyic GIRDLE. Outside the Passerine series the pelvic girdle of the Kurylemide resembles most nearly that of the Capitonide ; from which, however, it may be distinguished by the fact that whereas in the Hurylemide the post-acetabular ilium is produced caudad into a spine, in the Japitonide this backward extension is broad and bifurcate. Among the Passeres it approximates most closely perhaps to that of Chasmorhynchus. Cymbirhynchus only, among the Kurylemide, appears to possess even a vestige of the pectineal process. In Calyptomena the pre-acetabular ilium is broad throughout its whole length, and sharply truncated anteriorly. The inferior bor der thereof is markedly sinuous. The pre-acetabulz of the right and left sides are widely separated one from another ; and rise so as to lie nearly level with the ridge of the neural crest of the synsacrum, thus forming a large, open, canalis dleo-lumbalis. The post- acetabular region of the ilium is expanded to form a broad dorsal plane, and passing backwards terminates in a long spine, which, in Cymbirhynchus and Corydon, is closely applied to the free ends of the transverse processes of the post-synsacral caudal vertebre. The ischiwm in Calyptomena is long, produced backwards consi- derably beyond the level of the post- acetabular region of the ilium, and terminates in adownwar dly-directed hook- -shaped process which fuses with the pubis. Cymbirhynchus differs but slightly from Calyptomena is this respect. In Corydon the ischium is shorter antero-posteriorly, and deeper, than in the two genera just described, and does not project beyond the level of the free end of the post- acetabular ilium. Further, the dorsal border of the pre-ilium is much cut away anteriorly so as to expose a great portion of the synsacral neural crest. The ischio-pubic fissure is closed im all three genera here described ; the obturator foramen is shut off therefrom by a bony bar. The pubis is long and straight, and projects beyond the level of the ischium, especially so in Calypiomena. The close approximation of the post-acetabular ilium to the transverse processes of the free caudal vertebre is due to the shortness of the transverse processes of the synsacral vertebrze already referred to. The fovea lumbalis is small; and the fovea ischiadica and pudendalis are confluent. vill. THe Pecrorat Lime. The pectoral limb of the genera here described presents no marked differences by which they can be distinguished one from another. It resembles that of the Coliide and Capitonide in that metacarpal IT. sends backwards from its proximal end a small triangular bony spur (intermetacarpal process) to abut against 1905. ] OSTEOLOGY OF THE EURYLEMID#. 49 metacarpal ITT. In the Eurylemide this spur is, however, much larger than in the Coraciiform genera referred to. The humerus only is pneumatic; and is subequal to, or shorter than, the manus. The forearm isthe longest segment of the limb. The sulcus transverswm ov covaco-humeral groove is shallow. The crista superior is triangular in form. The incisura capitis is fairly sharply defined; the fossa subtrochanterica is large. There is a small ectepicondylar process, which, it is to be noted, is not forked as in the higher Passeres ; the entepicondylar process is still smaller. Ventrad of the tuberculum ulnare is a prominent spur-like blunt-pointed tubercle directed backwards and outwards so as to interlock with the olecranon process of the ulna in the extended wing. On the palmar surface immediately above the radial condyle is a small tubercle for the attachment of the inner head of the extensor metacarpi ulnaris. The ulna has a prominent, pointed, olecranon process, and bears a row of small tubercles, for the attachment of the secondary remiges, along its postaxial border. The radius is slender and slightly bowed. The forearm is the longest segment of the wing. The manus is well developed. As in the Capitonide and the normal Passeres, the base of Me. II. sends backwards a bony plate to overlap and fuse with the base of Mc. III. In the EKurylemide this plate (intermetacarpal plate) is of considerable size, 1ts base extending down the shaft for some distance. In some Coraci, e. g. Hurystomus, there is also an intermeta- carpal plate, but feebly Abvalieea! and not fused with Mc. IIT. ix. Tae Petyvic Limes. The pelvic limb, in the Eurylemidze, has, in common with the Cotingidee, a syndactyle pes ; and in this respect these two families resemble many of the Coraciiformes. None of the bones are pneumatic ; in which respect the Eurylemide differ from the Cotingide, which have a pneumatic femur, and resemble many of the Coraciidee. The femur islongand slender. The popliteal fossa is represented only by a shallow depression. The tibio-tarsus has moderately well-developed ecto- and ento- cnemial crests and a long fibular crest. The shaft is curved first forwards, then inwards, so that the distal end thereof is markedly inflected. The extensor bridge is ossified. The intercondylar gorge is deep. The fibula extends to below ple level of the middle of the shaft of the tibio-tarsus. The tarso-metatarsus is moderately long. The hypotarsus is complex. The distal end of the shaft is flattened from before backwards, and laterally expanded to form the condyle for digits II-IV. These condyles all extend forwards to practically the same level, the middle condyle scarcely projecting beyond the level Proc. Zoou. Soc.—1905, Vor. II. No. 1V 4 50 MR. W. P. PYCRAFT ON THE [May 2, of those on either side. In section the shaft is subeylindrical. Me. I. is long, as in the Passeres. The pelvic limbs of the Eurylemide and Cotingide can be distinguished from the limbs of the syndactyle members of the Coraciiformes by the fact that, m the latter, the tarso-metatarsus is either broad and flat, or deeply grooved anteriorly, and is more or less triangular in section. Further, the cnemial crests of the tibio-tarsus are, in the Coraciiformes having this type of feet, but feebly developed. x. SUMMARY. Regarded, by common consent, as. the most lowly of the Passeriformes, the Eur ‘ylemidz are at the same time an extremely specialised group ; much more so than has been hitherto recognised. Such a condition might have been expected indeed, inasmuch as this is a common feature among primitive groups. Nowhere is this specialisation more conspicuous than in the skull. The basipterygoid processes have entirely disappeared ; the maxillo-palatines have been reduced from broad triangular plates to rod-like splints; and a singularly perfect fronto- nasal hinge has been developed. In some genera, as in Corydon, the beak has vastly increased in size, and has acquired a markedly hooked shape, as well as a great increase in breadth. Nor is this all. The vomer presents a number of gradations in the direction of reduction and degeneracy ; and this is true also of the nasals, whereby the anterior narial fossa—which, by the way, is only in fact a narial fossa in so far as its extreme anterior end is concerned— is enormously enlarged. The lachrymal has been reduced to a mere vestige embedded, though still free, in the anterior face of the antorbital plate as im Calyptomena, or it is wanting as in Corydon. The palato-pterygoid articulation is also specialised ; so too is the nature of the vomerine support, this haying been transferred from the pterygoids to that of the palatines. The hemipterygoid element appears to be wanting, but traces of this may turn up in the nestlings of Calyptomena. Evidence of yet further specialisation is obtained from a study of the nestling skull. Besides the disappearance of the hemipteryg oid just referred to, the squamosal gives unquestionable proof in this direction; yet, at the same time, having preserved the essential characters of its shape, this element, more than any other bone in the skull, affords testimony of no uncertain kind as to the truly Passerine character of the group. Roughly L-shaped, there can nevertheless be no doubt, from the general contours of the bone, that it has been derived from a larger and more conical plate resembling that which obtains in the Corvide for example. Further, as in all the Passeres, the long axis of this bone is continued upwards and forwards beyond the parietal so as, in short, to overlap the frontal. So far as I have yet been able to ascertain, such an extension does not obtain anywhere among the Coracii- 1905. | OSTEOLOGY OF THE EURYLEH MIDE, 51 formes. A further indication of specialisation is the fact that the squamosal, in all the Passeriformes and most of the Coracii- formes, has absorbed the underlying bones so that it now appears, almost in its entirety, within the cranial cavity. The remarkable variations which obtain in the Class Aves, mm the form and arrangement of the membrane-bones are of considerable interest. These changes seem to follow along certain definite lines, and are the more remarkable because, save for the first few weeks of the bird’s life (the nestling period in short), these bones, as separate entities, cease to exist, being fused to form one homogeneous tissue. Yet progressive evolution is as obvious as in, say, the sternum or pelvis; though they cannot be individually influenced by the strains and stresses incident to the struggle for existence in the same way as if they maintained their individuality throughout life, or for at least some considerable time after leaving the nest. This is a point to which I propose to return later. So far we have described only the specialised features of the skull; what of the primitive? It is difficult to speak with any degree of certainty on this point. The small size of the anterior, posterior, and superior tympanic recesses, and of the tympanic cavity, may be reckoned in this category; and so too, probably, should the long narrow vomer as seen in Calyptomena. The close approximation of the otic and squamosal heads of the quadrate is an undoubtedly primitive character. These, in the mee are barely separated ; in the Corvide, for example, they are comparatively wide apart. These few points seem to sum up all the evidence that is obtainable on this question. How far specialisation has gone in the skulls within this group may be seen at a glance by comparing the skull of Calyptomena with, say, that of Cor ydon. In the latter the antorbital plate is oveatly reduced in size and thickness, and the lachrymal is wanting. The beak is markedly wider, more hooked, and articulates with the frontal by a more pronounced nasal hinge, while the nasal septum is obliterated by the inflation of the nasal processes of the premaxilla. The simple, unforked condition of the spina externa of the sternum is undoubtedly a primitive character; and in the form of the pelvic girdle this group is less advanced than in the remaining Passeres. Before proceeding to discuss the relationship of the Hurylemid to the remaining Passeres, it would be well to saya few words as to the wider question, of the probable allies of the Hurylemide outside the Passeres. This is a matter on which it is impossible to dogmatise ; at most, one can but throw out suggestions, of a very nebulous character. It will be found, probably, that Fiirbringer (3) has come nearest to the solution of this problem. He points to a relationship between the Hurylemide and the Cypseli, and a yet closer alliance with the Pici. Affinities to the Coraciidse he regards as remote indeed. A* 52 MR. W. P. PYCRAFT ON THE [May 2, My own work most certainly tends to support Fiirbringer’s conclusions. it 1s possible that the Hurylemidz will prove to be related both to the Caprimulgi and Cypseli. As regards the connection with the Pici, it is significant to note that the squamosal, in the nestling, closely resembles that of the Passerine type, mas- much as it overlaps the frontal, an arrangement which does not appear to occur elsewhere among the Coraciiformes. Coming now to the question of the relationship of the Eury- lemide to the remaining Passeres, I would remark, at the outset, that there seems scarcely sufficient ground for separating the former so widely from the latter as has been done by many during recent years. This separation foreshadowed by Garrod, and consummated by Forbes, has been widened even further than either of these distinguished workers would have considered justified. Forbes, just twenty-five years ago (2), summarised the main features of the Eurylemide, from the systematic point of view, as follows:—“.... They are not Tracheophone; and in that they possess the sciatic instead of the femoral artery, they differ from the Pipride and Cotingide, with which they have so often been associated. From these, too, they differ, as uey do from the Tyrannide, Pittide, and Rupicola, in the details * of the syrinx as well as in the simple manubrium sterni and other points. As has already been stated, they differ from all the other Passeres in the retention of a vinculum in the deep plantars of the foot... .” In a second contribution to this subject during the same month these views were repeated. After referring again to the syrinx and syndactyle foot, he goes on to remark :—‘‘ The peculiarities of the EKurylemide, and especially their oft-spoken-of retention of the plantar vinculum, are sufficient, I think, to justify their forming a main division of Passeres by themselves, as suggested by Prof. Garrod, which may be termed Desmodactyli, in distinction from the wunens. Eleutherodactyl . . .” It seems to me open to question whether so wide a separation is justified. After all, the existence, or rather we may say the survival, of the plantar vinculum is not so very surprising, not more so than the persistence of basipterygoid processes for example—which crop up sporadically among groups which have, as a whole, long since lost them. In Calyptomena, according to Beddard, this vinculum is wanting. Some importance has “been given to the statement made by Forbes, that in Hurylemus ochromelas there is a second vinculum : the additional slip “‘being given off lower down, from the hallux tendon, which joins the tendon of the digital flexor at the point where the latter, splitting into three, receives the main vinculum.” Gadow (4), commenting on this statement, remarks that this arrangement closely agrees with what obtains in Upupa and Jrrisor, a fact which suggests the origin of the Passerine plantars from this type. * Italics mine.—W. P. P. 1905. | OSTEOLOGY OF THE EURYLAMID&. D3 Though I looked carefully for this slip, I failed to find it, yet I examined three or four specimens. Forbes showed that, in the matter of the syrinx, the Hurylemide agree most nearly with the Philepittide of the Old World; and, after that, with the Cotingide, Pipridee, and Tyrannide of the New World. This organ is of the ‘ Mesomyodian,” “ tracheo-bronchial ” type, or, to adopt Gadow’s term, the syrinx is tracheo-bronchial and ‘“* Anisomyodean.” Had the syrinx instead of the plantar tendons been adopted as the basis of classification for this group, then the Cotingide would have been regarded as the more primitive group, inasmuch as in Lipaugus cineraceus the intrinsic muscle, according to Beddard, is of great width, “which seems to foreshadow its division in the Oscines into a complex of muscles... .” The many characters which the Eurylemide and Cotingide share in common—skeletal, muscular, syringeal, pterylological, &e.—are surely proofs that these two groups are much more nearly allied than is generally supposed to-day: the lkenesses are too many and distinct to be put down to convergence or correlated variation. The fact that the spina externa of the sternum is simple is generally bracketed together with the plantar tendons, and other characters, so as to emphasise the primitive character of the Eurylemide. But this same peculiarity of the sternum occurs again in the Cotingide. The pterylosis of the Eurylemide is generally regarded as peculiar : as a matter of fact, it is hard to distinguish from that of the Cotingide. The syndactyle foot again turns up—in the Cotingide. We have already described the close resemblances which obtain in the skulls of these two groups. Turning now to the muscular system. The syringeal muscles we have already referred to. They offer no striking peculiarities of structure. Indeed, the only muscles which seem to call for comment in this summary are the brevis and longus divisions of the deltoidews. The separation of this muscle into two distinct parts is nowhere so complete as in the Passeres. In its primitive (archicentric) condition, this muscle arises, in part from the acromion and inner face of the expanded free end of the clavicle and in part from the os humero-scapulare and crista lateralis of the humerus. It is inserted by a common tendon into the base of the ectepicondyloid process; the tendon forming the terminal of a practically homogeneous muscle. I have not yet had time to study the apocentricities of this muscle, but it would appear that as specialisation proceeds it breaks up into two more or less equal and perfectly distinct muscles terminating in a common tendon: later the brevis portion becomes suppressed and the longus much shortened, each receding farther and farther up the shaft of the humerus. T have only just realised the potentialities of this muscle as a factor in systematic work, and therefore have no large series of 54 MR. W. P. PYCRAFT ON THE [May 2, data to support this interpretation. But the facts, in so far as they are relevant to the present paper, seem to show that the primitive (archicentric) condition is represented fairly well in, say, Paradisea. In Corvus corax the longus portion is degenerate and fuses with the brevis just below the middle of the shaft of the humerus, the brevis portion then running downwards, ultimately becoming tendinous and passing to its Insertion at the base of the ectepi- Text-fig. 15. Dissection of arm, dorsal aspect, of Hwrylemus ochromelas, to show the deltoideus major longus and brevis. The longus portion has been cut through the middle, and the two halves drawn in opposite directions. The brevis portion has now become very degenerate and quite functionless. d.m.b., deltoideus major brevis; d.m./., deltoideus major longus; a., anconeus ; i., humerus: ., nervus radialis. condylar process. In Séwrnws both portions are extremely well developed, and perfectly separate until the distal end is attained, where they fuse in a fleshy insertion in which may be traced two distinct incipient tendons. 1905. ] OSTEOLOGY OF THE EURYLEMID#. DD In the Eurylemidz and Cotingide—at least in so far as Rupicola is concerned—the major portion is well developed, but the brevis portion has now receded, not extending beyond the middle of the humerus, and having an entirely fleshy insertion ; the longus portion, on the other hand, is slender and terminates in a long tendon. This interpretation of the transformations of the deltoides major et minor, it will be noticed, runs directly counter to that of Dr. Chalmers Mitchell, who, in a paper “On the Anatomy of Gruiform Birds” (6), contended that apocentricity in this muscle was shown by the gradual extension down the shaft of the major portion. It would seem, rather, as if the archicentric condition were represented by the maximum downward extension, and that apocentricity is represented by the gradual reduction of muscular tissue. That this reduction and inevitable suppression of the brevis portion represents an extremely specialised condition there can be no doubt; and the fact that it is shared also by the Cotingidee seems to me, coupled with the numerous other points which these two groups share in common, to show conclusively that the EKurylemide and Cotingide must henceforth be regarded as very closely related forms. These two groups differ in some other myological characters, as might be expected. The most noticeable is the fact that the latissimus dorsi posterior in the Cotingide appears to be wanting, though it must be remarked J have only been able to examine a single specimen of Rupicola in this connection. In the Euryle- mide both muscles are present, strap-shaped in form, and widely separated ; therein differing from the Corvide, in which they are of considerable size and shghtly overlap one another. But this feature is one of many primitive characters which the Corvide have retained. The peculiar myological resemblances which these birds share do not necessarily imply relationship ; but, as I have just remarked, there are so many structures in which these two groups agree, that it is impossible to entertain any notion of convergent resemblance between the two. The poimts of likeness are so peculiar, and affect such different, independent systems, that correlated variation and convergence cannot be regarded as a satisfactory explanation of the case. When two apparently con- vergent forms come to be particularised, each new point of resemblance which is brought to light is to be regarded as an additional link in the chain of evidence, establishing the common origin of the two forms in question. Thus, then, I contend there is no evidence which will justify the present isolated position which has been almost universally assigned to this group during the last few years. It is quite possible that further investigation will show that the Eurylemide are entitled to rank no higher than a subfamily of the Cotingide. 56 ON THE OSTEOLOGY OF THE EURYLAMID&. [May 2, But this point, as well as the status of the ‘‘ Passeres Clamatores,” I propose to deal with in a further communication at no distant date. xl, List or LivERATURE REFERRED TO. (1) Bepparp, F. E.—The Structure and Classification of Birds. (2) Forses, W. A.—* Contributions to the Anatomy of Passerine Bucdsy) Parts ies Ma ae Zias, es: (3) Firprincer, M.—“ Zur vergleich. Anat. des Brustschultes- apparates.” Jenaisch. Zeitschr. f. Naturwiss. xxxvi. 1902. (4) Gapow, H.—Bronn’s Thier-Reich. Systemat. Theil, Band vi. Vogel. 1893. (5) Garrop, A. H.—‘ On some Anatomical Characters which bear upon the Major Divisions of the Passerine Birds.” Part I. IP, Ay So MSG. (6) Mrrenent, P. C.—‘ On the Anatomy of Gruiform Birds.” lin Mop LEO, WOls il, 10, O28). (7) Suarpe, R. B.—A Review of Recent Attempts to Classify Birds. 1891. xii. EXPLANATION OF PLATE II. Fig. 1. Inner aspect of skull of nestling Hurylemus showing the large area occupied by the squamosal. Ua, Outer view of same skull showing the peculiar form of the squamosal. « Side view of Calyptomena viridis showing the peculiar lachrymal, large narial aperture, and large and spongy antorbital process. 2a. Ventral view of same skull to show the vomer, maxillo-palatine processes, and wide-set palatines. 2b. Dorsal aspect of same skull to show the large size of the nasal fossa and the nasal hinge. 3. Ventral aspect of skull of Cymbirhynchus showing the slender maxillo- palatine processes, short vomer, and sharply bent palatines. Note the difference between the palatal surface of this species and that of Calyptomena. 3a. Dorsal aspect of same skull to show the closing in of the nasal fossa and the formation of pseudo-nasal apertures. 4, Dorsal aspect of the skull of Corydon to show the large size of the narial apertures and nasal hinge. Heplanation of Letters. a.0.p. = antorbital process. na. = nasal. als. = alisphenoid. n.h. = nasal hinge. a.p.v. = anterior palatine vacuity. p. = parietal. b.oc. = basi-occipital. pa. = palatine. b.s. = basisphenoid. pa. = posterior nares. ex. = exoccipital. /p.o.p. = postorbital process. Jr. = frontal. pro. = prootic. h.pt. = hemipterygoid. pt. = pterygoid. 1. = lachrymal. S.n. = septum nasi. mes. = mesethmoid. sq. = squamosal. ma.p. = maxillo-palatine. v. = vomer. 1905. ] MR. OLDFIELD THOMAS ON A NEW GOLDEN MOLE. 57 May 16, 1905. G. A. BouLencer, Esq., F.R.S., Vice-President, in the Chair. The Secretary read the following report on the additions that had been made to the Society’s Menagerie in April 1905 :— The registered additions to the Society’s Menagerie during the month of April were 205 in number. Of these 67 were acquired by presentation and 19 by purchase, 104 were received on deposit, 9 by exchange, and 6 were born in the Gardens. The total number of departures during the same period, by death and removals, was 126. Amongst the additions special attention may be directed to :— A young female Chimpanzee (Anthropopithecus troglodytes), deposited on April 8th. A young female Giraffe from Northern Nigeria, probably belonging to the race known as Giraffa camelopardalis peralta, purchased on April 7th. A young male Huanaco (Lama huanacos), from Punta Arenas, Tierra del Fuego, presented by Mr. Moritz Braun and Capt. R. Crawshay on April 10th. A pair of Concave-casqued Hornbills (Dichoceros bicornis) from India, purchased on April 4th. Mr. Oldfield Thomas, F.R.S., exhibited examples of a new - Golden Mole from Knysna, Cape ‘Colony, which had been obtained by Mr. Grant in connection with Mr. C. D. Rudd’s exploration of South Africa, and which he proposed to name in honour of Mrs. Rudd, who had taken much interest in the results of the exploration. AMBLYSOMUS CoRRIZ Thos.*, Abstr. P. Z.S. No. 20, p. 5, May 23, 1905. Rather smaller than A. hottentotius. General colour dark smoky blackish, darker than in 4. iris, with a beautiful iridescent sheen, greenish to coppery violet, over the whole upper surface. Sides and belly not or scarcely lighter, a slight brownish tone occasionally present along the centre of the abdomen. Hairs of back 8-9 mm. in length, their bases dark slaty grey, them ends lighter and more brownish grey subterminally and their tips iridescent blackish brown. Crown and forehead like back. Cheeks paler, greyish or yellowish, but not conspicuously contrasted. Limbs and upper surface of hind feet smoke-grey. Skull (Pl. XVI. + fig. 3) in its general characters like that of * (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.—Hprror. | + P.Z.S. 1905, vol. i. p. 254. . 58 MR. H. B. FANTHAM ON A NEW SPOROZOON. | May 16, A. hottentottus, but markedly narrower across the brain-case ; and the zygomata less thickened at their posterior base. As a result, the two skulls beg of about the same length, the general outline was much less broadly triangular. Muzzle and inter- orbital region narrow and delicate. Outer edge of permanent teeth narrow antero-posteriorly, and of milk-teeth broad with conspicuous cusps, as shown in the Plate. The anterior premolar triangular, not extended transversely as in A. obtusirostris and chrysillus. Dimensions of the type, measured in the flesh :—Head and body 129 mm.; hind foot 13. Skull—ereatest length 28, basal length 22:6; greatest breadth across brain-case 16°6; greatest height 12°6; imterorbital breadth 8; front of i to back of m* 10:5; palate, breadth across premolars 8-1. An adult female had a head and body length of 118 mm. ; ereatest skull length 25:7. Hab. Kuysna, 8. Cape Colony. ‘In Forest.” Type. Oldmale. Original number 1021. Collected 25 January, 1905, by C. H. B. Gheats and presented to the British Museum by Mr. ©. D. Rudd. Ten specimens examined. This handsome little species was not only a very interesting discovery in itself, but the fine series of it obtained by Mr. Grant, of both sexes and different ages, had enabled Mr. Thomas to identify with confidence the milk and permanent dentitions of the specimen figured in the plate illustrating his paper on the Zululand Mammals collected by Mr. Grant. No proper knowledge of the respective characters of the two den- titions had hitherto existed. My. H. B. Fantham, B.Sc., F.Z.8., exhibited microscopic slides of and made remarks upon Lankesterella tritonis, N. Sp., a Hemogregarine parasitic in the red blood-corpuscles of a Newt, Triton cristatus (Molge cristata). Blackboard sketches were made illustrating the life-history of the parasite so far as was known. This parasite was found some time ago by the exhibitor while working in the Zoological Laboratory, University College, London. Afterwards his observations were independently confirmed by Dr. A. C. Stevenson. Up to the present the trophozoite and schizogonous stages only had been seen, and the sporogony pr obably took place in an intermediate host. Schaudinn’s and Siegel’s recent observations on the sporogony of allied parasites in the lizard and water-tortoise were quoted in support of this view, and mention was made of the inaccuracy of Hintze’s account of the sporogony of Z. ranarum in the intestine of the frog, the cysts therein mentioned probably being Eimerian stages of a Coecidian. The trophozoites, vermiform in shape, were apparently 5 py to 6 » in length, and slightly over 1» broad. They became U-shaped 1905.] ON THE ENCEPHALIC ARTERIAL SYSTEM IN SAUROPSIDA. 59 and gave rise to “rosette-stages,” about 2°5 p to 3°5 jin diameter. A “rosette” consisted of a schizont dividing up into merozoites. This parasite was probably the smallest Hemogregarine yet described, and it occurred in large red blood-corpuscles, those of Triton cristatus being about 30, in long diameter. The research on this parasite and allied forms was being continued. The following papers were read :— 1. A Contribution to the Knowledge of the Encephalic Arterial System in Sauropsida. By Frank E. Bepparp, M.A., F.R.S., Prosector to the Society. [Received March 29, 1905. | (Text-figures 16—21.) The following pages contain some facts relating to the principal vessels of the arterial system of the brain in a number of Lizards, in a Python, and in the giant Tortoise, Zestudo vicina. Some of these have not been hitherto studied; some have been examined by Rathke and others, and references to these anatomists will be found in the proper place. Most of the brains which I describe are now in the Museum of the Royal College of Surgeons. My principal object has been, next to the recording of new facts, to ascertain how far the characters offered by the distribution of these vessels, which are undoubtedly of use in the systematic arrangement of mammals, are also of use in the remaining Vertebrata for a like purpose. § Brain of Varanus exanthematicus. Although the cerebral arterial system of Varanus griseus has been described by Corti*, I have a few notes to add to his description and comparisons to make with the other genera treated of in the present communication. The two vertebral veins are strong and mark the posterior end of the medulla, precisely as is the case with Jguana. The posterior pair of cerebellar arteries arise, as in /gwana, from the basilar artery at the middle of the medulla, and are larger than the anterior pair, which arise from the fork of the basilar in front. This fork is not quite so symmetrical as in Jguana. ‘The left side and the left carotid are rather thicker than the right, and there is thus a suggestion of the marked inequality of these arteries in Python. The branches to the corpora bigemina and to the rest of the brain are quite as in /guana; but the large size of the ophthalmic arteries is a point of likeness to Python. * De systemate vasorum Psammosauri grisei. 1853. 60 MR. F, E, BEDDARD ON THE ENCEPHALIC [May 16, § Brain of Iguana tuberculata. The plan of the cerebral arteries in this Lizard differs in a number of particulars from that which will be shortly described. The anterior spinal artery, though of considerable size, is yet of less calibre than the basilar, with which it is nevertheless in perfect continuity. The exit of the posterior pair of cerebellar arteries marks the middle of the medulla. These arteries are slightly asymmetrical, the left being a little in advance of the right. They arise behind the point of origin of the 6th pair of cranial nerves. The anterior pair of cerebellar arteries arise just after the division of the basilar artery to form the carotids on each side; they are distinctly smaller than the posterior pair. The two branches of the basilar are approximately equal in size, as are the carotids which join them very shortly after the bifur- cation of the basilar. The point of junction is just at the point of origin of the anterior cerebellar arteries: In this, it will be observed, is a slight difference from the figure of the cerebral arterial system of this Lizard given by Rathke*. The next artery arising from the circle of Willis is in front of the third nerve (to the inside of which nerve passes the forward continuation of the carotid, as in other vertebrates) and supplies chiefly the corpus bigeminum of its side ; but it also gives off a branch each to the cerebellum and to the cerebral hemisphere. A little way anterior to this is a much more slender vessel which is absolutely symmetrical on both sides of the body and which almost at once divides into two branches; one of these ends upon the in- fundibulum, the other reaches the optic nerve of its side. Beyond this again arises the posterior cerebral artery. This artery reaches the hemisphere just at the furrow which divides it from the corpus bigeminum and runs parallel to the cerebral branch of the bigeminal artery. A little further forward the carotid finally divides into two arteries. The outer and stronger branch may be termed the middle cerebral; it runs forwards, curving outwards in the middle so as to be crescent-shaped, to the long and slender olfactory bulbs, giving off numerous slender branches to the hemisphere on its way. The inner branch very soon again divides into two: the innermost of them is the ophthalmic artery; the outer runs forward along the median ventral line of the brain in close contact with its fellow of the opposite side. * “ Untersuchungen itiber die Aortenwiirzeln &c. der Saurier,” Denkschr. k. Akad. Wiss. Wien, xiii. 1857, p. 51. [Since this paper was read Mr. R. H. Burne has kindly directed my attention to a paper by Dr. Hofmann in Zeitschr. f. Morph. u. Anthr. ii. 1900, in which the arterial system of the brain is described in a number of Fishes, Amphibia, Birds, and Mammals, and in the following Reptiles, viz. Iguana, Tropidonotus natrix, Croco- dile, and Testudo greca. That of the last alone (among Reptiles) is figured. This paper has been apparently overlooked by the recorders of the Mammalia, Aves, and Reptilia in the ‘ Zoological Record’ for 1900; but it is catalogued by the recorder of “General Subjects.”’—July 6th. | 1905. ] ARTERIAL SYSTEM IN SAUROPSIDA. 61 § Brain of Tropidurus hispidus. In comparing the arterial system of the brain of this Iguanoid with those of the other species of Lacertilia with which I have dealt, I am unable to say anything about the cerebellar arteries, which were not visible in the specimen examined by me. The bifurcation of the basilar artery in front at rather an acute angle consisted of equally-sized vessels, and the carotids which joined these arteries behind the third pair of nerves were also equal. The arteries to the corpora bigemina disappear at once in the groove separating each corpus bigeminum from the hind brain. The other arteries of the brain seem to be as in other Lacertilia. $ Brain of Eumeces algeriensis. The arrangement of the arteries of the brain in this Skink, which, so far as I am aware, has not been described, shows certain differences from that of both Varanus and Jguana. These features are illustrated in the accompanying drawing (text-fig. 16, p. 62). The fusion of the vertebral arteries with the basilar marks, as 1s usual, the end of the medulla. From the basilar artery arise a number of branches of which the posterior cerebellar arteries are the most important; of these the left artery arises in advance of the right and it is shortly reinforced by another branch. The bifurcation of the basilar anteriorly begins further back than in both Varanws and Jqguana; and another difference from the conditions observable in these two genera is to be noted. In these Saurians the carotids join the circle of Willis behind the origin of the third pair of nerves; in Hwmeces these arteries join the circle of Willis well in front of the third nerves, and therefore also in front of the slender anterior cerebellar arteries, and of the artery supplying the corpus bigeminum on each side. This artery not only supplies the corpus bigeminum but also the cerebellum, and it sends a branch forward which runs parallel to the posterior cerebral artery, and like it is lost in the groove separating the fore brain from the mid brain. Between this artery and the middle cerebral or Sylvian is a slender twig hke that of Zguana which runs to the base of the optic nerves. The anterior cerebral, which gives off the ophthalmic artery, is considerably thicker than the middle cerebral artery. § Brain of Gerrhosaurus. As is the case with Hwmeces, the basilar artery in Gerrhosaurus (see text-fig. 17, p. 62) divides rather further back than it does in either Jguana or Varanus. There is, moreover, a very distinct in- equality of calibre in the two arteries ; the right is in fact consider- ably larger than the left. This inequality does not, however, extend to the two carotids, which are equal in size. These join the circle of Willis only just in front of the point of origin of the anterior 62 MR. F. E. BEDDARD ON THE ENCEPHALIC [May 16, cerebellar arteries, which latter, as in other Lizards, are smaller than the posterior pair. The place at which the carotids join the circle of Willis is only just behind the third pair of cerebral Text-fig. 16. Text-fig, 17. Text-fig. 16.—EHumeces algeriensis. Ventral aspect of brain, showing chief arteries. ca. Carotids ; op. Optic nerves; opth. Ophthalmic arteries; 3, third nerves. Text-fig. 17.—Gerrhosaurus flavigularis. Ventral aspect of brain, showing chief arteries. Lettermg as in text-fig. 16. nerves. The order in which the remaining arteries of the brain arise 1s quite similar to that of the other Lizards described here, . and there are no particular comments to be made upon them. § Brain of Tupimambis nigropunctatus. The most important branches arising on either side from the 1905. ] ARTERIAL SYSTEM IN SAUROPSIDA. 63 basilar artery are the posterior cerebellar, and these arise a little behind the middle of the medulla. The two arteries are perfectly syminetrical with each other as to their point of origin. They are, however, different in their branching. The right artery gives off, shortly after its origin from the basilar, a strong artery running backwards along the side of the spinal cord. This branch exists and pursues the same course on the left side; but on that side of the brain it arises separately from the basilar artery. Between the origin of the posterior cerebellar arteries and the bifurcation of the basilar anteriorly are three pais of small arteries supplying adjacent regions of the medulla. A slightly larger artery, which is the anterior cerebellar, arises from the fork of the basilar. This fork is U-shaped in the Teguexin (text- fig. 18, p. 65), and not V-shaped as in the other Lizards described here. The U-shape is due to the fact that the two carotids run parallel to and almost in contact with each other for some distance before they join the circle of Willis. The carotids, moreover, lie within the area bounded by the third nerves very close to and about ona level with those nerves. Theartery formed by the junction of the basilar and carotid on each side, often spoken of merely as the carotid, passes outwards and slightly backwards at first, when it is practically at right angles with the basilar. In this region the artery shows different relations on the two sides of the body. On the left side it runs in front of the third nerve; on the right side it lies behind that nerve. The first branch arising after the carotid is at the bend of the artery, where it turns forward ; this very stout artery supplies the cerebellum and optic lobe; im- mediately in front of this is the artery of the optic lobe. This state of affairs occurred on the left side of the body; on the right side the two arteries arose by a common trunk. On both sides the artery of the corpus bigeminum gives off an artery to the cerebral hemisphere which buries itself in the furrow between the hemi- sphere and the optic lobe. From the inner side of the circle of Willis, just opposite to the bigeminal artery on the left side and to the conjoined arteries just mentioned on the right side, arises an artery which runs to the optic chiasma. This artery is precisely like that of other Lacertilia. The next artery to be given off is the posterior cerebral, which plunges at once into the furrow lying between the optic lobe and the cerebral hemisphere. The middle cerebral artery, which is the largest of the cerebral arteries, runs in the usual way along the Sylvian depression, and just in front of the point of origin of this the circle of Willis practically ends in the strong ophthalmic arteries which follow the optic nerves. There are therefore no differences of importance between the arterial system of the brain of Zupinambis and of the other genera of Lizards reported upon in the present communication. § Cerebral Arteries in the Lacertilia. We may deduce from the facts just described the chief 64 MR. F. E. BEDDARD ON THE ENCEPHALIC [May 16, characters of the encephalic arterial system in the Lacertilia for purposes of comparison with those of other Vertebrates *. {1) The entrance of the vertebral arteries into the anterior spinal marks the end of the medulla oblongata. (2) The posterior cerebellar arteries are the only conspicuous arteries arising from the basilar; they arise at about the middle of the medulla oblongata and behind the 6th pair of cranial nerves; they are occasionally asymmetrical with each other. (3) The anterior bifurcation of the basilar is at a more or less acute angle according to its position; the slender anterior cerebellar arteries are invariably given off from the bifurcated basilar behind the point of origin of the third nerves; the two branches of the basilar produced by the bifurcation may be imequisized. (4) The point of entrance of the carotids is not invariably the same; it is sometimes in front of and sometimes behind the third pair of nerves. (5) The artery on each side to the corpus bigeminum sends branches to the cerebellum and to the cerebral hemispheres. It arises in front of the entrance of the carotids. (6) In front of this artery is one which runs towards the optic chiasma. (7) There are three cerebral or hemispheral arteries: the posterior reaches each hemisphere just at its junction with the corpus bigeminum ; the middle one is Sylvian in position; the anterior cerebral gives off the ophthalmic; there is no distinct completion of the circle of Willis anteriorly. (8) There is no strongly marked asymmetry in the cerebral arterial system of the Lacertilia. S$ Brain of Python molurus 7. I have been able to study two injected brains of this serpent, of which one is more completely injected than the other. The most obvious and plain difference from the brains of other Sauropsida is the marked asymmetry in the arterial system (text-fig. 19, p. 65), which agrees of course with the vascular asymmetry shown else- where among the Ophidia. This asymmetry, however, only concerns the carotids. The other arteries of the brain, so far as I have been able to study them, do not show anything of the kind, but indeed a perfect regularity quite comparable to that shown in other Sauropsida. Of the two carotids the left is very much the larger. The basilar artery is single where it runs along the ventral surface of the cord and brain, until of course it bifurcates anteriorly at the commencement of the circle of Willis. The entrance of the vertebral arteries marks the end of the medulla. These arteries, which lie exactly opposite to each other, are very much stouter than the basilar, which they combine with the anterior spinal to * See below, pp. 66, 67, and 69, for comparison with Ophidia and Testudinata. + Rathke describes but does not figure brain-arteries of Ophidia in Denkschr. Akad. Wiss. Wien, xi. 1855. 1905. ] ARTERIAL SYSTEM IN SAUROPSIDA, 65 form. In one of the two specimens at my disposal, I could not see very well the actual mode of junction of the vertebral arteries with the basilar. In the other it was plain and very complicated. The basilar artery itself divides and immediately reunites, thus forming a circle; the two vertebrals join below this circle, and from the lower surface of this transversely running trunk two Text-fig. 18. Text-fig. 19. Text-fig. 18—Tupinambis nigropunctatus. Ventral aspect of brain, showing chief arteries. Lettering as in text-fig. 16. ‘ext-fig. 19.—Python molurus. Ventral aspect of brain, showing chief arteries. v. Vertebral arteries. Other letters as in text-fig. 16. To the right of the figure is an enlarged representation of the junction of the vertebral arteries with the basilar. branches are given off, each of which joins one side of the circle already referred to. I should not like to lay undue stress upon the fact as absolutely characteristic of Pyfhon, since the arrange- ment was not obvious in one specimen through deficiency of injection. Proc. Zoot. Soc.—1905, Vou. II. No. V. 5 66 MR. F. E. BEDDARD ON THE ENCEPHALIC [May 16, A noteworthy difference exists between the two specimens in relation to the course of the large left carotid. In the one brain this artery lies outside of the dura mater for a large part of its course, and gives off at least one branch to the brain which perforated that membrane ; thus giving additional proof of the fact that the carotid itself lies outside of the dura mater. In the other brain I did not observe this state of affairs. It follows that the left carotid exhibits an aloofness from the brain which is remarkable, and that the branches therefrom do not run on the same plane with it. The arteries to the optic lobes arise from the basilar artery after its bifurcation, between this point and the entrance of the carotids, and further back still there is a smaller cerebellar artery. The posterior cerebral artery is small and arises just in front of entrance of the carotids. The next important artery is a cerebral, which arises in front of the inflow of the carotids. This artery is the middle cerebral or Sylvian of other animals, since it runs along the rudimentary Sylvian fissure. The posterior cerebral is also partly represented by several small branches of the artery to the optic lobe. In front of the middle cerebral artery is a smaller anterior cerebral artery. Anteriorly to this the circle of Willis is completed in the following way: the large left carotid bifurcates to form the two nearly equally stout’ ophthalmic arteries which of course accom- pany the optic nerves. Just before this bifurcation the slender right carotid effects a junction with the common trunk. Immediately in front of this a single trunk arises from the point of bifurcation of the left carotid, which at once divides into two. These vessels run closely side by side in the furrow which separates the two hemispheres and rejoin at the extreme anterior end of the brain, their course in fact recalling that of the callosal arteries in mammals. The arteries are by no means inconspicuous, as 1s shown in the annexed figure (text-fig. 19, p. 65). T now draw, of course quite in a preliminary and tentative way, a series of comparisons between the Ophidian and Lacertilian brain arteries, enumerating the characters of the former in the same order as already given (on p. 64) for the latter. § Cerebral Arteries in the Ophadia. (1) The entrance of the vertebral arteries into the anterior spinal marks the end of the medulla oblongata. These arteries seem to be stouter than in the Lacertilia. (2) There is no markedly large pair of cerebellar arteries arising from the basilar artery, but a number of more or less equisized arteries supplying the cerebellum and adjacent region. (3) The two branches produced by the bifurcation of the basilar are equisized. The anterior cerebellar arteries arise from the bifurcated region. (4) The point of entrance of the carotids appears to be rather 1905. | ARTERIAL SYSTEM IN SAUROPSIDA. 67 fay forward as compared with the Lacertilia; but this appearance is at least partly due to the great length of the bifurcate region of the basilar artery in Python as compared with that of any Lacertilian. (5) The artery to the corpus bigeminum on each side arises behind the entrance of the carotid instead of in front as in Lacertilia. It gives off branches to the cerebrum and also to the cerebellum. (6) In front of this artery and also in front of the carotid is an artery which runs towards the optic chiasma. (7) There is a very marked completion of the circle of Willis anteriorly. (8) There is a strongly marked asymmetry in the arterial system of the brain due to the greater size of the left carotid. § Brain of Testudo vicina. The most salient characteristic of the arterial system in this Reptile is the double basilar artery (text-fig. 20, p.68). The artery is double for the whole of its course beneath the medulla oblongata. The anterior spinal artery in fact divides into two well behind the medulla. The right-hand one of the two branches is not larger than the left; the two arteries do not run close side by side, but are separated by a considerable distance. They are joined each of them by the carotid in front of the origin of the third nerve. Behind the origin of the third nerve a large number of arteries arise from the basilar on each side; there are certainly eight or nine of them on each side, and they supply the cerebellum, the medulla, and the cranial nerves of this region of the brain. The fifth artery (on the right side at any rate), which arises from the basilar behind the third nerve, is par excellence the cerebellar artery ; it fuses with its fellow of the opposite side at the end of the cerebellum. In front of the third nerve arise two arteries rather close together, of which the anterior has several branches and is the larger artery: it partly supplies the cerebral hemi- spheres and corresponds, as I imagine, to that artery in the Lacertilia which supplies the corpus bigeminum on each side. Asinthe Lacertilia, there are two cerebral arteries on each side. The first and largest of these (text-fig. 21, p. 68) may be termed the Sylvian, as it runs along the lateral groove upon the hemisphere which has been compared to the Sylvian fissure of mammals The branches of this artery are not altogether symmetrical on tne two sides of the body; it is possible, however, to distinguish the main trunk which runs towards the top of the brain, where it divides into a forwardly running anda backwardly running branch, several branches from the main stem which pass backwards over the temporal region of the hemisphere, and a strong branch running forwards to the olfactory lobe. Moreover, there is plain on one side a branch arising immediately after the origin of the Sylvian artery, which plunges at once beneath the hemisphere. A second 5 68 MR. F, E, BEDDARD ON THE ENCEPHALIC [May 16, cerebral artery arises from the circle of Willis a very short way in front of the Sylvian artery. This vessel runs forwards parallel Text-fig. 20. Text-fig. 21. Text-fig. 20.—Testudo vicina. Ventral aspect of brain, showing chief arteries. w. Junction of two halves of the circle of Willis anteriorly. Other lettering as in text-fig. 16. Text-fig. 21.—Testudo vicina. Lateral aspect of brain, showing chief arteries. S. Sylvian. w. Junction of two halves of the circle of Willis anteriorly. Other lettering as in text-fig. 16. with and close to the olfactory branch of the Sylvian, and finally ends in an anastomosis with the main stem, from which the 1905. ] ARTERIAL SYSTEM IN SAUROPSIDA. 69 olfactory branch arises beneath the olfactory lobe at the junction of the latter with the cerebral hemisphere ; before this point of junction a branch is given off to the olfactory lobe. The circle of Willis is completed anteriorly ; it also ends in two strong branches which run along the under surface of the brain, anteriorly, giving off numerous branches at the junction of the hemispheres with the olfactory lobes. One or more of these bend downwards (as the brain is viewed from beneath) and pass through the gap between the hemispheres running to the dorsal side of the brain in a way which suggests the callosal artery of the mammals. I could not detect anything more than a very small branch arising where the ophthalmic arteries arise in the Lacertilia. I cannot think that this artery is absent, but it is clearly not so conspicuous as in the Lizards. It is evident that the encephalic arterial system differs quite as much from that of either Lizards or Snakes as do the encephalic arterial systems in the two last mentioned groups. It is thus plainly possible to distinguish between several types of distribution of the cerebral arteries among the different divisions of the Sauropsida, and there is, as is well known, another type characteristic of mammals. It will be interesting to ascertain how far these several types confirm views as to the relative positions of the groups of Sauropsida under consideration. It will not be held by anyone, I presume, that the class Aves represents a primitive Sauropsidan type; and in agreement with this presumption we find clear evidence of modification in the encephalic arteries*, in the abortion of one or other of the normal two branches of the basilar. On the other hand, the arteries in question of birds are, as I think, undoubtedly primitive in that there is no completion of the circle of Willis anteriorly. A completed circle of Willis appears to me to be a secondary modification mainly for the reason that in mammals, where it occurs universally, it is there brought about in more than one way, and is moreover associated with strong arteries in the anterior region of the brain in close communication, or rather in close apposition, and there is apt to be confluence between closely apposed spaces and vessels. If this view be correct, we can set aside the brain of the Python and that of Vestuwdo as showing primitive characters by virtue of the fact that they have a closed circle of Willis. And in addition to this, it may be pointed out that the asymmetry of the arterial system in the Snake indicated by the carotids, and the changes in the disposition of the vessels due to the prevalence of the left carotid, can be fairly regarded as being secondary. This conclusion is obviously in accord with the current views of the relations of the Ophidia to other reptiles. There only remains the Lacertilia. It is, in my opinion, probable that in this group (and in Hatteria) the whole question lies of the antiquity of existing * P.Z.S. 1905, vol. i. p. 102. 70 SIR HARRY H. JOHNSTON ON THE [May 16, reptiles. And it must be admitted at onee that the facts dealt with in the present communication do not conform with any certainty to one view or to the other. On the whole, however, they seem to point to the Lacertilian; since from that type the remaining schemes of encephalic arterial arrangement can be derived, while the extraordinary modification of the basilar artery in Z'estudo, found nowhere else, would seem for that very reason to be a divergence from the original condition. 2. On the Nomenclature of the Anthropoid Apes as proposed by the Hon. Walter Rothschild. By Sir H. H. JouNnsToN, G.C.M.G., K.C.B., F.Z.S. [Received May 5, 1905. | I should like to make a few remarks on the admirable paper written on this subject by Mr. Walter Rothschild, which has just appeared in the ‘ Proceedings’ (1904, vol. ii. p. 413). Unfortu- nately, I did not know that this paper was going to be read in December 1904, or I should have endeavoured to be present. T am disposed in a general way to agree with Mr. Rothschild’s classification of the great Apes of Africa. I have only one criticism to offer with respect to the nomenclature of the Jhimpanzees. Since Mr. Rothschild has done so much to revise, revive, and establish the nomenclature of these Apes, I should lke to see him introduce a more rational spelling into the third of his species of Chimpanzees—the Bald Chimpanzee, which he gives, following Du Chaillu, as Simia koolookamba. Du Chaillu was very inaccurate in his transcription of African words, and he used the cumbrous system of English transliteration which prevailed until the rational spelling was introduced thirty or forty years ago by various scientific societies and departments of the Govern- ment. Koolookamba is really two words, which are pronounced nkulu-nkamba. J think that this spelling might stand in preference to Koolookamba [Simia nkulunkamba). A much more serious point, however, is the generic name which Mr. Rothschild gives to the Orangs—Pongo. Mr. Rothschild is undoubtedly right in reviving Simia as the most appropriate and the earliest name for the Chimpanzee genus, to which it was applied in the first instance by Linneus. Linneus evidently thought that the differences between the Chimpanzee and the Orang, which animal was later brought to his notice, were not more than specific, so that he included the Orang in the Chimpanzee genus. Much later, in 1799, Lacépéde applied the generic name Pongo to the Orangs; and although in the same year the Orang genus was named Satyrus, Mr. Rothschild prefers Pongo to this very appropriate designation, and wishes to establish Pongo as the generic name for the Orangs. I would certainly protest against this. There is much to connect the Satyr of the Classical world and Medizval mummeries with traditions of a red-haired man-of- 1905. | NOMENCLATURE OF THE ANTHROPOID APES, 71 the-woods—the Orang—which had filtered to Europe through India and the Levant, and the Arab sea-berne trade from Sumatra ; but Pongo is an African word originally applied to the Chimpanzee, and in all probability derived from the Bantu dialects of Angola, south of the Congo. The proper spelling of this word is Miponee, and it is a root which, in varying forms, is found in a number ot Bantu dialects and languages in Western and Equatorial Africa, and used to indicate either a chimpanzee or a big baboon*. I have not got access to various old books at the time of writing, but I think I am correct in saying that English and Dutch travellers on the West Coast of Africa in the 16th, 17th, and 18th centuries referred to the Chimpanzee as ‘‘ Pongo.” I also fancy that the same allusion and the same name are made use of by Buffon. As in zoological nomenclature the preference is for the adoption of a Latin or Greek name, it is a pity to introduce into our lists a barbarous word in preference to one derived from either of the classical languages. But when in addition an African word is taken as the name of an Hast Asiatic genus, then the choice is singularly inappropriate. = Tt may be of interest to add the names for “Chimpanzee” in a number of African languages, mostly collected by myself :— LANGUAGE. DISTRICT. WorpD FoR CHIMPANZEE. ILE sdgoneiscoocn vesdodenbooa | SHS IVAOS socconcoocaspsnconseasonccs uRaanien (KG ns Gig prefix). Vai. sonosoode sadanbecooupl) NAVGRtEGIN WURLRE) ” —Sossusadedadcoesoogs | a!Dutloe Busi oo cccccccscsessese. North-western Liberia .. : Guru. JHGRMWDGO 00 ona necaconcn North of Sierra Leone, "Liberia, Iburu. and Ivory Coast. GO WeSta ree eee an COntral Palberialerrt se Ibulu. (GROIRD cooesosocebosysesescouen) tela Jems liner, VW est Central Onyi. Le Liberia. e Basauand eens Coast ol Centraluluibertares sees Ibe. | Kru and Grebo............ Southern Liberia ..................... Tuawe JZOP DIN: sonocosonboccoo scones LUMTSBOP QE IVES oho cen: ww ane Obo. Tjo languages............... Brass, Bonny, “Niger Weltay eo: Tele. LED OMe en eons sen eash (LOWE NIGER derek a ne ee Ozodimbas ED Ulcer ere ae ere eae Old Calabar .o.cccccceccseeeis Idiok. aaa ies fete vi } Lower Crosspiveriee cee rcs Enop or Enowi. INKG ow ........10...s....... Extreme Upper Cross River ...... Boki. Mbudikum ............... Sources of Cross River, N.E. of Apu or Epfu. Cameroons. Barondo and Isubu...... North Cameroons Coast ............ Ewaka. WG XORGAHD caonadvaanedanccay “ERIKOOIN nosnnogaoacscaconsveonsadocasages ANIA TERD GING! ING NEEKO)s also Nkulu. TON GO conn sse tee ees Lower Congo and Congo Coast .... Mpongi. Kimbundu .................. Angola (south of Lower Congo) ... _Mpongo. [x*, The origin of the name “ Pongo,’ of Buffon and others. | Kiwemba [or me Bemba ... South-west of Tanganyika ......... Koroe. Ki-guhea .. ee Westbancanyikaeanenressasasos elolires Kabwari ............... North-west an canyalkay ere secees Soko. Manyema .................. West of Tanganyika and extreme Soko. Upper Congo. Ruande ..................... North of Tanganyika ..............,_ Enjangwe. Kifipa veces Hast Coast, Tanganyika ........... Isike. DLu-ganda@ .................. West and North Coast of Victoria Edzike or Izike. Nyanza. 72 SIR HARRY H. JOHNSTON ON THE [May 16, T do not suppose much deference will be shown to my own suggestions ; but it seems to me that the best generic name for the Orangs would be Satyrus; or, if that is strongly objected to because it may be confused with the specific name of one or two Chimpanzees, then possibly Pithecus. IT cannot help thinking that in this case, as in many other instances, when we are settling for good and for all our biological nomenclature, we carry too far the passion for asserting the prior rights of the first invented name, which is occasionally a singularly inappropriate one. I will conclude my paper with a few remarks on the definite knowledge of the different species of Anthropoid Apes from the dawn of zoological science in Greece to the end of the 18th Century of the present era, by which time European zoologists had begun to discriminate pretty clearly between the Gibbons, the Orang, and the Chimpanzee. Knowledge of the Gorilla of course was not clearly defined till about 1848 or even later. It is possible, however, that a living specimen of the Gorilla was brought over to Holland in the latter part of the 17th Century. A figure of this creature (which was a female) is given in Dr. Tyson’s work on the Chimpanzee, published in London in 1699. _ Aristotle, writing in about 330 B.c., divided the mammals that were nearest to man into three closely allied groups: the Pithekoi or Apes, the Keboi or Monkeys, and the Kunokephaloi or dog- faced Baboons. In the Latin translations of Aristotle these designations are rendered Simia, Cebi, and Canicipes. Aristotle’s LANGUAGE. District. Worp For CHIMPANZEE. Nyoroand Hima dialects. Unyoro, Toro, Ankole, and south- Isike, Yisiki, west of Victoria Nyanza. Kchikuya, Empundu, Kitera. TROPPO cnpsenscagoarcenoonna Wiki, TREN KEMATN p00 5. sovanegeaa LWESKREED, TARUEW occ ec ese esses... Forest, north-w est of Semliki Neule. River (Mboga Country). Bibira .....0.c.......... Congo Forest between Semliki R. Kika. and Upper Congo (Aruwimi basin). Mangala and allied Upper Congo, between Aruwimi (Mu) Kumbuso languages. and confluence of Mubangi Welle, (Mu- is only the and between course of Congo and singular prefix). vicinity of Welle River. Lend y...... es: West of Albert Nyanza ............ Nerrr (2’s trilled). Bambute pyginies ee Semliki-Aruwimi Forest . Un. Momfu ....... + North-east Borderlands (Ce ongo- Tato. Nile water parting) of Congo Forest, west of Lendu. SAL atropine dara tern eatin ese North of Albert Nyanza een Bin. Madi ..........00.05..4..... (Western dialects of) West of Arugu. Mountain Nile, north-east of Congo Forest. Makarka (Nyamnyam). Southern and Western parts of Baham or Bamu; Bahr-el-Ghazal province of also Irangba and Egyptian Sudan. Nderuma. Mund... cccccecseees, North-east of Makarka.. ... Angé. JECT ~ coreoae0oor North-west of Makarka redline. oe Deddar foe sine tray selles and Philologists -¢ can fill up the gaps in this series of names 1905. | NOMENCLATURE OF THE ANTHROPOID APES. 73 general description of the Pithekoi delineates very distinctly an ‘Anthropoid Ape, and reads as though it was derived from a generalised knowledge of the Chimpanzee, a knowledge obtained no doubt from specimens which had been brought down the Nile from the Egyptian Sudan (in the southern parts of which the animal still exists) to Lower Egypt. A good summary of Aristotle’s description of the Pithekoi is given in Dr. Tyson’s celebrated book ‘“‘On the Anatomy of a “Pyemie, sive Homo sylvestris,’ which, as before stated, was published in 1699, and of which there are copies in two or three of the principal libraries of London. T think I am correct in saying that in an Egyptian fresco or papyrus which is exhibited in the Egyptian collection of the Museum at Naples, a Chimpanzee is depicted amongst other strange animals brought to Egypt from the Sudan. I believe also there is a representation of the Chimpanzee on one of the Roman mosaics recently brought to light at or near Carthage, and now preserved in one of the Museums, either at Carthage or Tunis. The Byzantine Greeks, who, after Alexandex’s conquests, ex- tended their trade to India, and the Arabs of west, south, and east Arabia, who maintained commercial relations with Sumatra, the Malay Peninsula, North-west Borneo, and the ports of the Persian Gulf and the Red Sea, may haye introduced some knowledge of the Orang utan to Constantinople, to Egypt, and to the Mediterranean world between 100 B.c. and the fall of the Byzantine Empire. Sir Walter Scott in his novel ‘Count Robert of Paris’ introduced somewhat fantastically a captive Orang utan into the story. Iam not aware what foundation he had for this incident ; and I think it somewhat improbable that an Orang utan could at that period have survived the overland journey from the Persian Gulf to the Mediterranean, or the transit through Egypt. Marco Polo, the Venetian, in 1296 or thereabouts, travelled overland from Asia Minor to China and the Malay Peninsula, and reached Sumatra and possibly Borneo, bringing back with him stories of man-like apes, some of which cer tainly referred to the Gibbons, while one or two may be attributed to the Orang utan. Odorie, a friar of the Order of St. Francis, travelled overland from Constantinople to India during the first half of the 14th Century, and from India reached Sumatra by sea. He brought back distinct accounts of both Gibbons and Orangs. Ibn Batuta, a Morocco Arab, also journeyed to those parts about the same time, and described the Orang utan in his records. Friar Giovanni dei Marignolli, a Franciscan like Odorie, also travelled overland from France to China and thence to the Maiay Archipelago during the first half of the 14th Century, and br ought back from Sumatra, or more likely North Borneo, very distinct accounts of the Orang utan. At the commencement of the 16th Century the Portuguese conquistadores reached Malacca and Sumatra in their ships, and 74 ON THE NOMENCLATURE OF THE ANTHROPOID APES. | May 16, by 1521 had placed more or less roughly on the map all the big islands of the Malay Archipelago. They were followed a few years later by Spanish, Dutch, and French adventurers. During the 17th Century many British ships visited Sumatra and Borneo, and the Malay name Orang utan was in current use in scientific Europe during the second half of the 17th Century, having been originally definitely applied to the man-like apes of Sumatra and Borneo*. But towards the close of the 15th Century the Portuguese had already become acquainted with the West Coast of Africa and the Chimpanzee. They first noticed this creature in the southern part of what is now the colony of Sierra Leone. They called it in their earlier writings “ Selvage” (savage), and later ‘“ Barri.” Later still they came to know more of the Chimpanzee in dealing with the Lower Congo and Northern Angolay. It there went under the name of Pongo, which as already explained is the Angola name Mpongo. Andrew Battel, of the 16th Century, was an Hssex fisherman. Through being shipwrecked off Brazil he got conveyed into Portuguese captivity in Angola, Escaping, he travelled into the northern part of Angola towards the Congo. He returned to England and brought back with him stories of the ‘““Pongos,” which obviously referred to the Chimpanzee. The name “Chimpanzee” does not seem to have come into vogue till the latter part of the 18th Century, or to have been much used until the 19th Century. I have no certain clue as to its origin; but I have been told that it is a Loango word of which the root would be -mpanzi or -mpangi (possibly, therefore, cognate with the Congo name for Chimpanzee, mpongi), with the well-known Bantu prefix chi (kz) added. This prefix is sometimes an augmentative, so that chimpangi or chimpanzi might merely mean a big ape. At the close of the 18th Century, Buffon, Linnzus, Lacépede, and other zoologists had finally discriminated between the Gibbons, the Orang utan, and the African Chimpanzee; and to this list was added in the period between 1847 and 1860 the definitely established genus (afterwards species, then again genus) of the Gorilla. ‘The discovery of the Gorilla was really due to the American Evangelical missionaries, who established themselves in the early part of the 19th Century in the Gaboon; but complete specimens of this Ape and a far more extended knowledge of it were brought to the civilised world by Du Chaillu. Stanley asserted the existence of the true Gorilla as far east as the forest between the Upper Congo and the Nile watershed; and this statement has seemingly been confirmed by the specimens received from that region by Dr. Matschie, and described and figured by Myr. Rothschild. * Though often misapplied to the African Chimpanzee in the 17th and 18th Centuries by English and Dutch sea-captains, who, having first made acquaintance with the Orang in the Malay Archipelago, saw Chimpanzees at the West African ports on their return voyage. + When I visited Angola in 1882 Chimpanzees were still found in forested regions inland south of the Congo and north of the Quanza River, especially in the old kingdom of Congo. ee Pin. . Eng el Term ; del. I 4, CASO Solel lel. VU. Bale & Damelsson, L&. SIAUILIGS Ol IRISUON OIL OI TEU. IP 4. SAGO Savel, WLP. OV, Taal ips tary © — A. J.Engel Tera, del Bale & Danielsson, L%@. SAUILLS Of IRUSUUNOILOPIEIL 1905. ] ON BATS OF THE GENUS RHINOLOPHUS. (i) 3. On some Bats of the Genus Rhinolophus, with Remarks on their Mutual Affinities, and Descriptions of Twenty- six new Forms. By Knup ANDERSEN *. [Received May 12, 1905. ] (Plates ITI. & IV.+ and Text-figure 22.) The present paper is, chiefly, an attempt to disentangle some of the more complicated groups of Eastern Rhinolophi, to make out the probable interrelations of the species, and to describe the many new, imperfectly known, or hitherto confused forms. I have appended some general remarks on the affinities of the Ethiopian and Western Palearctic species. The material placed at my disposal has been more extensive than that of previous writers on these Bats, namely, Prof. Peters (1871) and Dr. Dobson (1878); and I have approached the subject from a different point of view, basing the diagnoses of the primary groups, and, where possible, of the species and subspecies too, not on external and dental characters alone, but also on important differences in the skulls. This may account, partly at least, for the essentially different conclusions on many points at which I have arrived. On the other hand, the following pages afford ample proof that my material has not been complete enough to enable ine to venture an answer on all the difticult questions, taxonomic or phylogenetic, that occurred to me during my work. I shall feel satisfied if my paper is considered of some use as a basis for further investigations. T owe my sincere thanks to Mr. Oldfield Thomas for entrusting me with a revision of these Bats, for giving me unlimited access to the recently acquired, still unregistered specimens in the British Museum, especially those of the large and important ‘Tomes Collection,” and also for having favoured me with much valuable information during the progress of my work. T also have to acknowledge the kind assistance of Mr. Gerrit S. Miller, Jv., who sent me for inspection almost all the Indo- Malayan Lhinolophi preserved in the United States National Museum, including many new and interesting forms, part of which will be dealt with below. For the loan of specimens for comparison, or for information on examples preserved in Continental Museums, I am indebted to Geheimrath Prof. Dr. Ehlers, Gottingen ; Prof. Matschie, Berlin ; Prof. Dr. Kurt Lampert, Stuttgart; M. Ch. Mottaz, Geneva ; M. A. Ménégaux, Paris; and Prof. A. Cabrera Latorre, Madrid. I. THe RAINOLOPHUS SIMPLEX GROUP. Diagnosis. Basioccipital, between cochlez, not unusually nar- rowed. Posterior connecting process low and rounded off (text- fig. 22a, on p. 121). * Communicated by OrpFreLD Tuomas, F.ZS. + For explanation of the Plates, see p. 14. 76 MR. K, ANDERSEN ON BATS [May 16, Tinclude in this group 40 different forms (22 species), correspond- ing to Lh. megaphyllus, affinis, capensis, clivosus, and ferrum- equinwm in Dobson’s ‘ Catalogue of the Chiroptera in the British Museum. Only the Austro-Malayan, Oriental, and Palearctic forms will be described below, and only the first species in some detail, the description of the other forms being, as a rule, confined to the points in which they differ from the fundamental type. The Ethiopian species will be briefly mentioned in the ‘“ General Remarks” on the group (p. 117). 1. RHINOLOPHUS SIMPLEX, sp. n. (Plate III. fig. 1.) Diagnosis. Cranial character : supraorbital crests meeting at a point behind the middle of the orbit. External: sella distinctly constricted at middle. Forearm 44:2 mm. Details. Nose-leaves large, as compared with those of the other Austro-Malayan species (2h. truncatus, nanus). A supplementary leaflet distinctly visible in front of, and on the anterior part of the sides of, the horseshoe ; a character common to all the members of the present group, but becoming gradually less pronounced in the more highly developed species (affinis, ferrwm-equinum, and their allies) ; it seems to point back to the much more primitive genus ilipposiderus. Horseshoe so broad as to completely cover the upper lip; a slight indication of a tooth-like projection on either side of the median notch. Sella decidedly broader at base than at summit, and distinctly constricted at middle; summit rounded ; height of sella, from angle between vertical portion and nasal lobe, about 4°8 mm., width at base 2°3, at constriction 1:9, at summit 18 mm.; front of sella densely covered with exceedingly short white hairs (scarcely observable without a lens). Posterior con- necting process low and broadly rounded off. Lancet long, almost cuneate ; length, from posterior transverse bridge, about 4°7 mm. Three mental grooves, as in all forms of this group, except the highest-differentiated species (ferrwm-equinum and its nearest relations). Hars, compared with those of the closely allied Austro-Malayan species, rather large, almost reaching the tip of the muzzle when laid forwards. Upper part of outer margin somewhat concave ; tip blunt ; no constriction below the tip. Wing-structure very primitive: 4th and 5th metacarpals sub- equal in length (the 5th, if anything, a little shorter), and both of them but very slightly longer than 3rd; III.°* less than 14 the length of ITI.1; IV.° and, especially, V.° very short, being only a trifle longer than IV.‘ and V.’ This structure of the wing is characteristic of all the primitive members of this group (simples, megaphyllus, truncatus, nanus, celebensis, borneensis, malayanus, rouai, &e.); it is first in so highly-developed forms as affinis and its various modifications (ferrwm-equinum, &e.) that we find an important progress: prolongation of III.°; shortening of the 3rd * Hor brevity’s sake I call the proximal phalanges of the 3rd, 4th, and 5th fingers IIL}, TV.1, and V.!, the distal phalanges of the same fingers IIT.7, [V.2, and V.2 1905. | OF THE GENUS RHINOLOPHUS. ne metacarpal, as compared with the 4th and 5th; the 5th meta- carpal decidedly longer than the 4th; &e. Tail a little longer than the lower leg. Plagiopatagiuim inserted on tarsus. Colour (of a spirit-specimen, unfaded). Fur of upper side a very dark shade of “drab,” approaching “ Prout’s brown”; base of hairs rather more distinctly drab; under side somewhat darker than drab. Skull. Four anterior nasal swellings and two posterior. The four anterior arranged in a transverse row, forming the upper and lateral borders of the nasal opening. Externally these anterior swellings are separated only by extremely faint linear depressions ; internally by three bony lamelle, also easily observable through the thin, transparent outer wall of the swellings. The posterior nasal swellings, situated immediately behind the anterior ones, at the front corner of the orbital cavity, are much lower, slightly concave at summit; three very faint lines divide them, rather indistinctly, into an upper, middle, and lower swelling.—The shape and arrangement of the nasal swellings, as here described, are, roughly speaking, the same in almost all the members of the simplex-group; there is some variation in the size of the swellings in the different species; but the more noteworthy deviations from the general scheme are two only: Rh. malayanus and Rh, stheno. Postnasal depression triangular in shape, rather long; the supraorbital crests, which constitute the lateral border of this depression, meeting (and joining the sagittal crest) at a point more or less behind the middle of the orbital cavity. ‘“ Supraorbital length ” of skull (¢. e. distance between the point of junction of supraorbital crests and median anterior point of nasal swellings) greater than extreme width of nasal swellings.—The shape of this part of the skull, as here described, is characteristic of only the four most primitive members of the group (simplex, mega- phyllus, truncatus, nanus). Palatal bridge comparatively long (in antero-posterior direction); measured in the median line equal to about one-third the length of the upper tooth-row ; median anterior point opposite the front of m’, median posterior point opposite the middle of m?. Dentition. As a general guidance: in all existing species of the genus the upper p’ * is completely lost ; in all the more primitive * T write the dental formula (excel. of incisors and canines) of a Rhinolophus with the (cf. Herluf Winge, p?_-p* m! m? m3 Po P3 Py M, M, mM, “ Jordfundne oz nulevende Flagermus fra Lagoa Santa; med Udsigt over Flager- musenes indbyrdes Slegtskab” ; E Museo Lundii, vol. ii. pt. 1 (1892), p. 56). As already mentioned by Winge, we have no positive proof whether the upper premolar lost in all known species is p? or p?. For two reasons I regard the former alternative to be the more probable :—(1) In all Rhinolophi, also the most primitive forms, the lower p, 1S on the point of being reduced, in the more highly-developed species pushed definitely out to the external side of the tooth-row, in the still higher forms completely lost; it is but reasonable to suppose that the premolar quite lost in the upper jaw of al7 species corresponds to the premolar which is on the point of being lost in the lower jaw of all species, in consonance with the general rule that the teeth of the upper jaw show a more advanced stage of evolution than those most complete known dentition as follows : 78 MR. K. ANDERSEN ON BATS [May 16, species of the simplea-group also the lower p, 1s very much reduced in size and on the point of being driven out of the tooth-row, to the external side; in all the more primitive species of the group also the upper p* is reduced in size, but still, trnvariably, in the tooth-row. The following remarks apply to RA. simplea and Rh. megaphyllus, the dentition of these two species, the most primitive within the present group, being practically exactly alike :—p, very small, but decidedly less reduced than in the other species of the group. The position of this tooth, m relation to p, and p,, varies individually (in the same geographical race, and in examples from the same locality and of apparently the same age): completely in the tooth-row (one specimen), or slightly towards the external side (two), or half external (one), or almost quite external (one), or completely external (one). This “ vacillation ” in the position of p, is of some interest as being the first indication of a tendency towards driving this premolar out of the tooth-row, a tendency gradually increasing in a long series of more highly developed species, and culminating in the forms in which the tooth is quite lost, even in young individuals (4h. acrotis).—p? is comparatively large, with a well-developed, pointed cusp. From its base to its tip this cusp is directed obliquely mwards, under an angle of about 25° to 45° with the vertical line; also in those species of the present group in which the cusp is so much reduced as to be scarcely perceptible without a lens, it is invariably point- ing obliquely inwards, only to a still higher degree. The upper canine and p' always widely separated. In some individuals there is a very narrow interspace between p* and p’, on either side of the jaw, or on one side, no doubt a remnant of the place where p’, lost in all existing species, was situated (see footnote on p. 77). Measurements *, On p. 80. of the lower jaw. (2) When the lower p; is external in position, or even when it is completely lost, we still, rather often, find p, and p, separated by a narrow inter- space, reminiscent of the time when p; had its normal position in the tooth-row; if we can find, sometimes at least, a similar “atavism” in the upper jaw, our sup- position will be strengthened; and such cases are, in fact, not very rare :—in some individuals, and just those of the most primitive species of the genus (simplex, megaphyllus, borneensis, refulgens, philippinensis), I find an arrangement of the upper teeth which can be graphically expressed as follows: cp pm!m?mn®, 7,e. the anterior of the upper premolars in contact with the canine, the posterior mm contact with the first molar, but between the two “p” still a narrow interspace, apparently a remnant of the place where the lost premolar was situated; if so, however, the lost p is, of course, p%, those present p? and pt. * Only the following measurements require some explanation :—Hars, length from base of inner margin to tip. Forearm, trom posterior point of radius to front curve of carpus (wing bent), therefore somewhat greater than the length of radius measured on skeletons. Metacarpais, as far as possible the true length of the bones. 2nd phalane, always exclusive of the cartilaginous “3rd phalanx” (this restriction beg of especial importance in measurements of the 3rd finger, the terminal cartilagmous rod of which is comparatively large). Hind foot, with claws. Skull, total length, to front of canines (not to front of premaxilla). Width of brain-case, above root of zygomata. Supraorbital length, distance between point of junction of supraorbital crests with sagittal crest and median anterior point of nasal swellings. Mandible, condylus to front of incisors. Upper and lower teeth, exclusive of incisors. 1905. | OF THE GENUS RHINOLOPHUS. 79 Type. 2 ad. (in aleohol). Lombok, 2500 ft., June 1896. Col- lected by A. Hverett, Esq. Brit. Mus. no. 97.4.18.4. 2. RHINOLOPHUS MEGAPHYLLUS Gray. (Plate IIT. fig. 2 a, b,c.) Diagnosis. Allied to Lh. sinuplex, but considerably larger. Fore- arm 46-50 mm. Details. This is a large continental representative of the simplea- type. The evidences of its close connection with the Lombok species are clear enough: the general shape of the facial portion of the skull; the wide interspace between the upper canine and p'; the presence, individually at least, of an extremely narrow interspace between p* and p*; the distinctly constricted sella; the strong development of the nose-leaves; the large ears. On the other hand, it has in several respects taken its own course of development: the sella is, also proportionately, broader than in simplex, the constriction at the middle is more abrupt; the nasal swellings are, also proportionately, considerably broader ; the size of the animal is markedly increased: as regards this latter, Lh. megaphyllus bears quite the same relation to Rh. simplea as Rh. rouxi does to Rh. borneensis. Distribution*. EKastern Australia. Louisiade Archipelago. Geographical races. There are two apparently well-marked forms of Rh. megaphyllus, differing in size and in geographical habitat. 2a. RHINOLOPHUS MEGAPHYLLUS Gray, TYPICUS. Rhinolophus megaphylius J. Ki. Gray, P.Z.S. 1834, p. 52. Rhinolophus megaphyllus (partim) Peters, MB. Akad. Berlin, 1871, p. 306 7; Dobson, Cat. Chir. Brit. Mus. (1878) p. 110. Diagnosis. Larger: forearm 46°5-50 mm. Sella. In one, out of eleven specimens, the summit of the sella is completely square-cut ; in all the others (some of them from the same locality) it is broadly rounded off. Conf. with this Rh. borneensis. Colour. (1) Dark phase (two skins, one adult and one full- grown, but young): Like 2h. simplex. (2) Russet phase (one skin, full-grown individual, but young): Uniform “ russet” above and below; base of hairs of upper side “ clay.” Measurements. On p. 80. Distribution. Hastern Australia: Queensland, New South Wales. Technical name. The type of Rh. megaphyllus is in the British Museum. *« The information on the “distribution” of the species and subspecies reviewed in this paper is based eawclusively on the material examined by myself. 7+ I amunacquainted with Peters’s hypothetical Rh. keyensis, based on an example in the Leiden Museum, and characterised as “eine vielleicht nur etwas kleinere Varietit [of megaphyllus]| oder Art” (1. s.c. p. 307). No further information has been published, and nine years later Peters records “ Rh. megaphyllus” from the Key Islands without any reference to Rh. keyensis (Anu. Mus. Civ. Genova, xvi. (1880) p. 32). It is not very likely that the typical Rh. megaphyllus should occur in the Key Islands. 80 MR. K. ANDERSEN ON BATS [May 16, 26. RHINOLOPHUS MEGAPHYLLUS MONACHUS, subsp. n. Diagnosis. On an average smaller than the typical form: fore- arm 46 mm. Details. Sella a trifle broader at base than in the typical form ; summit completely square-cut ; front face a little more distinctly haired. Length of forearm almost as in the smallest individuals of the typical form, but metacarpals distinctly shorter. Tail also comparatively somewhat shor ter. Brain-case decidedly more slender. Tooth-rows somewhat shorter. In colour scarcely different from the dark phase of the typical form. Measurements. Below. _Lype. 2 ad. (in alcohol). St. Aignan’s Island (Misima), Louisiade Archipelago. Collected by Albert 8. Meek, Esq. Brit. Mus. no. 98.4.1.1. Measurements of Rh. simplex and megaphyllus. Rh. simplex. Rh. megaphyllus. | | | f . typica. monachus. fi a i 11 specimens, © ad. Till 5 skulls. Type. | Min. Max. mim. | mm. mm. mm. Ears, length .. ME Ne eI te 18 I TS TIGR 19'8 » greatest yea th yiees tie beige 13°5 || 135 15 16 Nose-leav es, total Jength ............... 14°5 Halon lO, 14'8 y breadth of horseshoe ...... 85 | 88 98 88 SHOKeAT TR ME rr aL eon RR 44-2 || 465 60 46 GixGl WNGUACAAOAN co an ccoeoo oon nas ovaoee ego00e 318 || 33°38 36 32°7 11) GES ae sree Mens at Coen onemminS Heaeen ie 13 | 13 146 13:2 Il? . Son cies ae teanene TR Tene ee 178 175 20 17°8 4th metacarpal Seana Gah Se aaa ae ey 32 | 343 368 33°5 UTNE) hisktcweon a heeteants icon Wiley, | Sane aby eatin 9:2 98 11:2 97 Ine eh On Acree chee ye eee || 115 13:3 10 Sth metacarpal deat aliens take ati 318 || 3843 3865 32°7 VRIES NEN os enh FAM el Se br 10 || 104 12°7 10°2 Ve). SUE Seth ty Se Sees Areal cemneptie neers efaeti 11:2 elle amen 11°7 Gialamas. oe A aly tn ae 245 | 29:2 268 20°5 Mo wersleas ener ctee cn e c eet ess eee 19°7 | 185 22 19 Foot ..... Senaneaaenuccea aes? 88 | 9 102 87 Skull, total length aba usaacstisnasneneeas 187 || 19°9 20 19°3 s)) mastordiwidth! (0 so .ceeee.--.-- ee 3 98 98 95 » width of brain-case ............... 78 85 86 8 >» Zygomatic width ...............-. 9°4. | .. 10 96 » supraorbital length ............... 6 | 6 68 59 width of nasal swellings ...... 52 | 58 6 al Mandible, length icra 12°8 SESS ai 13:2 Uipperatecth a ese eee eee 72 ad eit 73 IWORFEE WOEWD 05 000 cvs v0¢ c60. 008 960 codon soo 00s 78 82 87 8 3. RHINOLOPHUS TRUNCATUS Peters. Rhinolophus truncatus Peters, MB. Akad. Berlin, 1871, p. 307. Rhinolophus megaphyllus (non Gray), var. a, Dobson, Cat. Chir. Brit. Mus. (1878) p. 111. Diagnosis. Allied to Rh. simplex. Sella more slightly constricted 1905. } OF THE GENUS RHINOLOPHUS. 81 at middle. Summit of sella square-cut, or even concave. Base of fur almost blackish. Forearm 44°7-46°8 mm. Details. In this species the sella* is not of the shape charac- teristic of Kh. simplex and megaphyllus. It is narrower, not considerably broader at the base than at the summit, and the constriction at the middle is less distinct. This points decidedly away from simplex, and towards nanus, celebensis, and borneensis. The square-cut (or concave) summit of the sella seems to be a rather common feature in those forms of the present section of the group which are inhabitants of small islands (cf. Rh. megaphyllus monachus, Rh. nanus, kh. borneensis spadix). Lancet long and cuneate. Wing-structure and proportionate length of tail as in simplex, Plagiopatagium inserted on tarsus. Colour (six skins; adult individuals, but teeth quite, or almost, unworn). Very peculiar. Generalimpression: avery dark brown. Details: hairs of upper side “ broccoli-brown” at tip; below the tip, for a broad space, almost “‘ clove-brown” (more exactly: an exceedingly dark shade of “ hair-brown,” very much approaching clove-brown); the extreme base of the hairs, immediately at the skin, again somewhat lighter. Individual hairs of the under side much of the same colour, but the tips more brightly broccoli- brown, giving the under side a somewhat lighter appearance. All the specimens are exactly alike in colour. Skull. Essential characters as in Rh. simplex. Nasal swellings narrow. Dentition. p, is, f anything, a little more reduced than in simplex. In two skulls I find it placed in the tooth-row, but shghtly towards the external side; in a third, on the one side half external, on the other external; in a fourth, external on both sides, and the interspace between p, and p, therefore very narrow. p is always in the tooth-row; its cusp rather well developed, though somewhat smaller than in simplex. No interspace between p” and p’. Measurements. On p. 84. Distribution. Batchian. Technical name. One of the two typical specimens (in the Berlip. Museum) was collected on Batchian by A. R. Wallace and for- warded to Prof. Peters by Tomes. The whole series in the British Museum is from the same island and the same collector, and four of the examples belong to the recently acquired Tomes Collection ; they are therefore practically (though not technically) co-types. kemarks. The dentition of Rh. truncatus proves it to be ona slightly higher level than simplea; the interspace between the upper canine and p’ is a little narrower, p’ a little more reduced. The vacillation in the position of p, gives evidence of the same ten- dency as in simplex: towards the more advanced members of the group. In the shape of the nose-leaves it has taken a course point- ing towards borneensis. In its coloration it seems to stand alcne. * A good series of skins, but no spirit-specimens, are at my disposal. This. description is from the sesoftened nose-leaves of three examples. Proc. Zoou. Soo.—1905, Vou. If. No. VI, 6 82 MR. K. ANDERSEN ON BATS [ May 16, 4, RHINOLOPHUS NANOS, sp.n. (Plate IIT. fig. 3.) Rhinolophus megaphyllus (non Gray), var. (2 (partim), Dobson, Cat. Chir. Brit. Mus. (1878) p. 111 (Goram). Diagnosis. Essential cranial characters as in Rh. truncatus, but brain-case remarkably slender. Sella so slightly constricted as to be practically parallel-margined. Small: forearm 43-3 mm. Detwils. This species marks a further step towards the celebensis- borneensis type. Externally Rh. nanws is exceedingly like these two species, but the skull is of the stmplex type. The sella (compared with that of the foregoing three species) is considerably reduced in breadth; its width at the base is but very little greater than at the summit; the constriction at the middle is much reduced (it requires some attention not to be overlooked); and the whole of the sella therefore might very well be called almost parallel-margined; summit completely square-cut (there will probably, in a large series, be some individual variation in this respect). The horseshoe, too, is a little narrower. Lancet almost cuneate, the lateral margins being but very slightly coneave. The sizeof the ears, both length and breadth, is reduced ; the tip slightly more attenuated (less blunt than in 2h. simples). In the structure of the wings it stands exactly on the same level as the foregoing species. Colour (one skin; adult; teeth almost quite unworn).—Fur of the upper side uniform dull “ mars-brown”; base of hairs slightly lighter ; under side very much of the same colour as the upper side, but with a slight tinge of “ drab.” Skull. Postnasal depression and supraorbital crests as in Rh. simplex. Nasal swellings very narrow (4°9 mm.). Chief character (compared with the three foregoing species): the very narrow brain-case (7 mm.). Dentition. p, quite external, and cingula of p,and p, in contact (a sufficiently large series will presumably show some vacillation in the position of p,). p®? in the tooth-row ; its cusp very small. Measurements. On p. 84. Type. Ad. (skin). Goram Island. Collected by Dr. A. R. Wallace. Brit. Mus. no. 61.12.11.10. Remarks. This species is readily distinguished from Lh. celebensis and Lh. borneensis by the different shape of the facial portion of the skull. Dobson regarded the specimen here described, together with two others from N. Celebes (Menado), as a variety (“@”) of Rh. megaphyllus, characterised chiefly by having “‘the summit of the vertical process of the sella broadly rounded off, much broader than the base.” But, firstly, it should be remembered that a sella, much broader at summit than at base, would be exactly the reverse of what is found in megaphyllus; it would even be unique in the whole genus. Secondly, on resoftening the nose-leaves I found the sella, in all the three specimens, quite of the same general shape as in Ah. borneensis, i.e. practically 1905. | OF THE GENUS RHINOLOPHUS. 83 parallel-margined. It would evidently have been much more to the point if Dobson had called these Bats Rh. borneensis, not Rh, megaphyllus. But Rh. borneensis, again, was confused with Rh. minor, which, however, not only IS a distinct Species, but belongs to a different group of the genus. 5, RHINOLOPHUS CELEBENSIS, sp. n. (Plate IIT. fig. 4 a, 0.) Rhinolophus megaphyllus (non Gray), var. 3 (partim), Dobson, fat. Chir, Brit. Mus. (1878) p. 111 (Menado), Diagnosis. Supraorbital crests meeting at a point more or less in front of the middle of the orbit. Nasal swellings narrow. Nose-leaves as in Rh. nanus and Rh. borneensis. Small: forearm 43-44‘7 mm. Details. In the foregoing species (2th. simplex, megaphyllus, truncatus, nanus), all of! which are Australian or Austro- Malayan, the supraorbital crests join the sagittal crest at a point more or less behind the middle of the orbit. In 2h. celebensis, as in all the other species of the present group, which are all Oriental, ‘Paleearctic, or Ethiopian, the supraorbital crests meet at a point more or less in front of the middle of the orbit. This makes 2 comparatively shorter postnasal depression, the supraorbital crests being the lateral borders of this depression. In this point therefore Hh. celebensis agrees with the MWesterm forms of the group, differing from the Eastern. The mechanical reason for this modification is evidently the following: a slight increase in the size of the temporal muscle has pushed the sagittal crest more forwards; this involves a shortening of the supraorbital crests; this again a reduction in the length of the postnasal depression. The nasal swellings are narrow (4:8 mm.), as in the closely related Eastern forms (anus, truncatus), In the more Western Rh. borneensis they are, at least sownewlba, and as a rule con- siderably, broader. Compare figs. 4 and 5 on PI. IIT. Tt is worth noticing that the cranial characters of this species are, so to say, “fin accordance with” its geographical habitat : Celebes is, geographically, intermediate between the Austro- Malayan and Indo-Malayan subregions, and in its more im- portant cranial characters 2h. celebensis points partly westwards (shortening of supraorbital crests), partly eastwards (narrow nasal swellings). The nose-leaves, ears, wings, and the general size are as in Lh, nanus and Kh. borneensis, Colour. (1) Makassar specimen (2 ad.; in alcohol; unfaded ; teeth unworn).—General impression of upper side : brown : the true colour is a deep brown shade of “drab”; base of hairs a little lighter than drab ; under side drab with a tinge of “ broccoli- brown.” (2) Menado specimens (two skins; ad.; teeth almost un- worn).—Above uniform dull “ mars-brown,” base of hairs but 6* 84 MR. K, ANDERSEN ON BATS [May 16, slightly lighter ; colour of the fur of the under side very much as on the upper side. The Makassar specimen seems to represent the true “dark phase” ; the mars-brown tinge of the Menado skins may indicate a tendency towards a “russet phase.” Similar differences in colour are very common in this section of the group. Dentition. As in Rh. nanus. Measurements. Below. Type. 2 ad. (in alcohol). Makassar, 8, Celebes, November 1895. Collected by A. Everett, Esq. Brit. Mus. no. 97.1.3.19. Distribution. Celebes : Makassar Menado. Measurements of Rh, truncatus, nanus, and celebensis. } | Rh. truncatus.| Rh. nanus. Rh. celebensis. - | 6 specimens, | Ad. | 3 specimens, | 4 skulls. Type. 3 skulls. | Min. Max. | Min. Max. mm. mm. mm. | mm. mm. Mars leno bhiyae , width of brain-case ............... he a 7 8 » zygomatic width .................. “es Zen 9 Br >» supraorbital length Paste DIDEEON 58 48 48 width of nasal swellings Bee Sil Beal 4:9 48 48 Mandible, IGMEFAN —soovooosoocsenoeavooacce| AS USE 13 122 12:7 (Wippersteethy 26 cscs sececsetecene se ccennn set Al 98} 72 7 72 Wowerjteeth 3oicc.. shee esaensqeaSeeioese TD OY 78 TA 78 6. RHINOLOPHUS BORNEENSIS Peters. (Plate III. fig. 5 a, b,c.) Diagnosis. Similar to Rh. celebensis, but with broader nasal swellings. Small: forearm 41°2-46°3 mm. Details. Sella so slightly constricted as to be almost parallel- margined from base to summit; in some individuals the con- striction is completely obsolete; height of sella about 3 mm. ; 1905. | OF THE GENUS RHINOLOPHUS. 85 width at base, at middle, and at summit: 2, 1°8, and 1:7 mm. Lancet almost cuneate, or the lateral margins but slightly concave, never abruptly narrowed at the middle (as i in Fh. rout) ; length of lancet about 4°2mm,. Ears and wings quiteas in Rh. celebensis. Plagiopatagium inserted on tarsus, or as much as 1°5 mm. above the tarsal joint. Colowr. There is an extreme dark phase and an extreme red phase, connected by several intermediate stages. (1) Dark phase.— ?, Banguey Isl. (Brit. Mus.); two ¢, Pulo Sarutu (Un. St. Nat.-Mus.); all of them full-grown, but with unworn teeth; distal epiphyses of metacarpals in two of them ossified, in one not completely so; in alcohol, unfaded. General impression of upper side: brown. The true colour isa deep brown shade of “drab”; base of hairs next to ‘ broccoli-brown.” Under side between ‘‘ wood-brown ” and “ broccoli-brown.” The individuals are not precisely, but almost, alike in tinge. (2) Intermediate stage, nearer to “ dark phase.”— gad., Qad., Labuan (B.M.); ¢ ad., N.W. Borneo (B.M.); teeth either quite unworn, or almost wnworn,; distal epiphyses of metacarpals ossitfied ; in alcohol, unfaded. Upper side “russet,” base of hairs but slightly lighter. Under side ‘‘ wood-brown.” (3) ian emediate stage, nearer to “red phase.” — @ ad., Sirhassen (URNES Ms) co Grade 2 afl. Karimata(U. N.S, M.); teeth either quite unworn, or very slightly worn; distal epiphyses of meta- carpals ossified ; in alcohol, unfaded. Much like the foregoing, but also the under side of the body “ russet.” (4) Extreme red phase.— g ad., Sirhassen (B.M.); teeth wn- worn; epiphyses ossified; in alcohol, unfaded. Much like the extreme red phase of 2A. rowxi: not far from “ cadmium orange” above; ‘“‘ orange” beneath. As proved by the above, these differences in colour are inde- pendent of the geographical habitat and of the sex of the individuals, seemingly also of the age. So far as the present material goes, the only “ phase” in which a quite young, though full-grown, individual occurs (epiphyses not quite ossified) is the dark phase; but it may be accidental: the individual which represents the extreme red phase is, at all events, only a few months older (teeth unworn). Skull. As in Rh. ¢elebensis, but with broader nasal swellings (5:4 mm., on an averzge), Dentition. ps almost always completely external, but in one skull (out of eleven) half in row. Cingula of p, and p, in contact (six), or very slightly separated (four), or distinctly separated (one), p always in the tooth-row; cusp very small. In four individuals there is an extremely narrow interspace between p* and p* (the former place of p’). Distribution. N. Borneo; 8. Natunas; Karimata Group. Technical name. The type of Rh. borneensis, in the Berlin Museum, is from Labuan, There are two specimens from the 86 MR. K. ANDERSEN ON BATS [May 16, same island in the British Museum*, As, however, i. borneensis has for many years been completely confused not only with several more or less closely related species, but also with the widely different Rh. minor, the following remarks may not be out of place here :— The salient point in the original description of Rh. borneensis, as given by Prof. Peters (loc. infra cit.), is this: “Sattel....an dem vordern obern Hade abgerundet, die hintere, zusammenge- driickte Spitze [7. ¢. the posterior connecting process] kaum héher, abgerundet.” have emphasised the last three words, because they clearly prove that 2h. borneensis belongs to what here is called the. simplex group (connecting process low and rounded off), and has. nothing to do with Ah. minor or its allies (connecting process pro- jecting and pointed). But ten years later (MB. Akad. Berlin, 1871, p. 306), Peters himself believed Rh. borneensis to be identical with Rh. minor, described by Horsfield so long ago as 1824. The reason was, beyond all doubt, this: to identify Horsfield’s Bats without an examination of the types is, in most cases, impossible ; and Peters had not seen the type of Rh. minor (then in the Indian Museum, London, now in the British Museum), but only the bad figure in the ‘ Researches in Java’; as, further- more, the two species in many respects (size, wings, sella, ears, &e.) are, externally, puzzling alike, the mistake is easily explained. Thus, according to Peters, there were two small Indo-Malayan. Rhinolophi: the one, with a low and rounded connecting process,, he called Rh. minor, Horsf. (synonym: Rh. borneensis, Peters) ; the other, with a projecting and pointed connecting process, he identified with Temminck’s 2h. pusillus, stated to be from Java.. Under these circumstances, a quite reasonable conclusion: we had a name for either “species,” and perfectly clear diagnoses. Dobson, who examined the type of Ah. minor, states, quite correctly, that the connecting process is projecting and pointed ; when, nevertheless, he put Rh. borneensis down in the list of “synonyms” to Lh. minor, he must have overlooked the most important point in Peters’s description of borneensis, the shape of the connecting process. Dobson, therefore, called the small Indo- Malayan Rhinolophus with pointed process Rh. minor (synonym : Rh, borneensis): thus, the names were the same as employed by Peters, but the diagnosis exactly the reverse ; Temminck’s “ih. pusillus he identified with Fh. hipposiderus (sic); and as to the small Indo-Malayan Rhinolophus with rounded process (the true borneensis) he put it down under Lh. affinis, Horsf. (!), with which species he alsc united the very different Rh. rowwi, Temm., at the same time keeping a genuine Rh. rowaxi separate as Rh, petersi. This accumulation of errors and wrong identifications * On one point there is a discrepancy between Peters’s description of Rh. borneensis and the series before me: according to Peters the length of the forearm is 37 mm.; in the smallest (adult) specimen I have seen, it measures 41°2 mm. I am informed by Prof. Matschie, who kindly re-examined the type for me, that Peters’s statement must be a mispr int ora slip of the pen; the forearm of the type specimen (a rather young, but apparently full-grown individual) measures 41 mm. 1905.] | OF THE GENUS RHINOLOPHUS. 87 is the true reason of the exceedingly confused state in which this group of Bats has remained, making a safe determination of Specimens procured almost impossible. Geographical races. There seems to be two forms of Rh, bor- neensis, differing, slightly, in the size of the ears, and in geogra- phical habitat. 6 a. RHINOLOPHUS BORNEENSIS Peters, TYPICUS. Rhinolophus Borneensis Peters, MB, Akad, Berlin, June 25th, 1861, p. 709. Rhinolophus minor (partim, nec Horsf.), Peters, MB. Akad. Berlin, 1871, p. 306; Dobson, Cat. Chir. Brit. Mus. (1878) p. 114. Rhinolophus afinis (partim, nec Horsf.), Dobson. op. cit. (1878) ) pele Diagnosis. Kars slightly shorter: 16-17 mm., and narrower : 12°2-12°8 mm. Forearm 41°:2-43°7 mm. Details. In one specimen (from Banguey Isl.) the summit oi the sella is completely square-cut; in the others (Labuan, N.W. Borneo) it is broadly rounded off. This is, no doubt, an individual variation, but, it would seem, of more frequent occurrence in indi- viduals inhabiting smaller islands (cf. Rh. megaphyllus monachus, Rh. nanus, Rh. truncatus, Rh. borneensis spadia). Measurements. On p. 88. Distribution. N.W. Borneo; Labuan; Banguey. 6 6. RHINOLOPHUS BORNEENSIS SPADIX Miller, Rhinolophus affinis rouxi (non Temm.) Thomas, Nov. Zool. i. (1894) p. 656. Rhinolophus spadi« Gerrit 8. Miller, Jr., Proc. Wash, Ac. Sci. ii. (March 26th, 1901) p. 136. Diagnosis, Kars slightly longer: 17-19°5 mm., and broader : 12°5-14:2 mm, Forearm 42°5- “46-3 mm, Details. In one specimen (Sirhassen Isl.) the summit of the sella is completely square-cut ; in all the others (one of them from the same island) it is broadly rounded off. Measurements. On p. 88. Distribution. 8. Natunas (Sirhassen); Karimata Group (Kari- mata and Pulo Sarutu). Technical name. The type of “ Rh. spadix,” in the Washington Museum, is from Sirhassen. There is a specimen from the same island in the British Museum. I am indebted to Mr. Miller for the loan of a paratype, also from Sirhassen, and of the series from the Karimata Group, collected by Dr. Abbott. Remarks. 1 should not have separated these two forms (if they be so) of borneensis, if the latter of them had not, accidentally *, got aname. There is no tangible difference in the skulls, not even * When describing Rh. spadix as a new species, Mr. Miller compared it with. Rh. affinis. He could not, very well, compare it with Rh. borneensis, which was regarded as identical with Rh. minor. 88 MR. K, ANDERSEN ON BATS [May 16, (as might perhaps be expected) in the measurements of them. It may well be that the few examples from N.W. Borneo, Labuan, and Banguey (four only) happen to be rather short-eared (and short-armed), and therefore do not show the true limits of indi- vidual variation in these respects. I prefer to keep them separate, provisionally at least, to call attention to the possible existence of two very slightly differing forms of the species. 7. RHINOLOPHUS VIRGO, sp. n. Diagnosis. Similar to borneensis, but much smaller. Forearm 37°5-38°8 mm. Details. This is decidedly the smallest species of the present group. The horseshoe is markedly narrower than in any other form of the borneensis type; the sella considerably smaller than in borneensis, but of the same shape; the ears much shorter and narrower. Colour. Probably not far from being the same as in the dark phase of borneensis (the two specimens examined are evidently somewhat faded in alcohol). Measurements of Rh. borneensis aid virgo. | Rh. borneensis. || Rh. virgo. f. typica. spadix. | 4 specimens, | 6 specimens, || 2 specimens, 4 skulls. 7 skulls. | 2 skulls. Min. Max. Min. Max. || Min. Max. mm. mm. | mm. mm. |) mm. mm. Ears, length puscnsovooconecscoanasenceal| AK 2 SIL, SN lee Na 2 Sreatest breadth ..................| 122 128 | 125 142 ||. 10:7 108 Nose-leaves, total length ...............) 125 137 | 127 142 ‘|| lov 112 5 breadth of horseshoe ...| 8 83 8 9 C2 GD Forearm NBeaoovopEeooudososscosnosceel| AMIGA dehy 425 463 || 375 388 3rd metacarpal 2.0... .....| 28°7 B12 ASS) Be ||| Wee aise gt ee aeoe 121 1355 117 142 || 102 10-7 WES? socledondpedsoneoineeeeomnoueinete erate GE Theorem Iho aiarae Mera) It Peale 15°2 4th metacarpal... | 297 822 | 307 345 || 98 286 IV.) SUL GON GaE IMB Re bu Mice Ne a SHEA Nera en Un 8:8 OFT 82 98 73 82 LOE Tap Hae Ba) ohh eee CMD acme 10 118 98 12 9 9 oth metacarpal ..........................| 29'S 322 30°7 33°8 27 = (282 AG gh Ser Suet coeuSsc quell amr oeh. sere Sa 95 103 9 10°3 81 88 NE coupe SRGee MEN ee ma i a ag 10°2- 11°8 98 12:2 82 83 Tail pan psnddaduupsoacsdeasbucnddoseagonenad|| lls) 19°2 183 21°5 179 202 owerdesy Rec thc. beet aeeee vale mye@e 1972 17-2 19 14:2 152 Foot nobridedad Gsniine Haakon SebeH a we aA eae ook 88 9 85 91 . '3} Skull, totallength | 195 | 182 20 162 169 » Mastoid width ...0...0..0..00...5| 9:2 88 95 & 82 » Width of brain-case...............) ... 8 78 82 Weal manda: >» Zygomatic width .................. say 98 | 9 9:9 81 82 » Ssupraorbital length ............... Sl Hy 5 5:2 47 5 » _ width of nasal swellings 53 57 52 55 43 43 Mandible, length .................... 192) 11 Pil 122 13°7 10°8 11°5 Wipperstecthic se nn Gran nit 7 72, @ 76 61 62 LSwersheethr ee eee ete (Odes T4 8 65 68 1905. ] OF THE GENUS RHINOLOPHUS. 89 Skull. As in borneensis, but considerably smaller; the nasal swellings are, also proportionately, narrower than in the Bornean species (perhaps as a consequence of the much smaller nose- leaves). Dentition (two skulls). ps half in row (one skull), or external (the other). p, and p, in the former skull, of course, separated 5 in the latter almost in contact. p° in the tooth-row. Upper canine and p* widely separated. Type. Qad. (in alcohol), §. Camarinas, Luzon, Philippine Islands. Collected by L. M. McCormick, Esq. Un. St. Nat. Mus. no. 101966. Remarks. This species is readily distinguished from any other form of the simplex group by its small size, narrow horseshoe, and short ears. The shape of the connecting process ought to prevent a confusion with the equally small species of the minor group, to which it, in other respects, bears a very striking external resemblance. 8. RutyoLopHus MALAYANUS Bonhote. (Plate ITI. fig. 6.) Rhinolophus malayanus Bovhote, Fasc. Malayenses, Zool., 1. (Oct. 1903) p. 15. Diagnosis. Closely allied to Rh, borneensis, but median anterior nasal swellings somewhat more differentiated. Small: forearm 4]:2-42°8 mm. Details. Externally this Bat is exceedingly like kh. borneensis, but the shape of the anterior nasal swellings is somewhat different. The colour, too, seems to be constantly different. The sella is, in vertical direction, a trifle shorter, but the difference is scarcely appreciable without actual comparison with borneensis. The lateral margins of the sella are, practically, parallel from base to summit; an extremely faint constriction can be traced, at least under a lens; summit of sella rounded. Plagiopatagium inserted on tarsus, or very nearly so. Colour. (1) Biserat specimens; two 2 ad.; August and Sep- tember ; teeth slightly worn; in alcohol; unfaded.— Upper side a rather dark brown shade of “drab”; this colour is confined to the tips of the hairs; the much broader base of the hairs so light ‘“‘eeru-drab” as to approach whitish; under side whitish “ ecru- drab,” somewhat darker on the sides of the body. (2) Laos specimen; ad.; teeth slightly worn; skin.—Very much lighter. Upper side bright ‘“ cinnamon,” base of fur ‘“‘eream buff”; horseshoe patch * on back dark brown; under side buff. * A dark-coloured patch on the upper side of the body, horseshoe-shaped, or like a V, the branches starting on each shoulder, convexity (or angle-point) directed backwards. It is curiously characteristic of many species of the families Rhino- lophide and Phyllostomatide, but often (quite individually) more or less, or even completely, obliterated, especially, of course, when the fur also is dark-coloured. Being, as a rule, more common and more distinct im young or immature individuals, it is, probably, an inheritance from some remote ancestors of the two families. Rhinolophide and Phyllostomatide have, probably, had a common origin. 90 MR. K. ANDERSEN ON BATS | May 16, It looks like a dark and a light “ phase.” The dark phase differs from that of Ah. borneensis, chiefly, in having the under side of the body much lighter, in strong contrast to the colour of the upper side, and in having also the base of the hairs of the upper side much lighter. The light phase is, as will be seen from this description, totally different from the ‘“ cadmium orange ” phase of borneensis (and more approaching the light phase of. Lh. affinis himalayanus). Skull, Wssential characters as in Rh. borneensis, but the median anterior nasal swellings somewhat more distinctly marked off from the lateral anterior swellings. Dentition. p, external; p, and p, almost in contact; p* im row, cusp extremely small. Measurements. On p. 92. Distribution. Biserat (Jalor, Malay Peninsula). Laos Mts. (Siam). Technical name. The type is in the British Museum. Remarks. From the Laos Mountains, Siam, I have seen one dried skin only (Tomes Collection); it looks like a light-coloured phase of Rh. malayanus; the nasal swellings of the (fragmentary) skull have the shape characteristic of this species. But fresh material from that region is desirable. 9, RHINOLOPHUS NEREIS, sp. n. (Plate IIT. fig. 7 a, b, ¢.) “ Rhinolophus rouxii?” (non Temm.) Gerrit 5. Miller, Jr., Proc. Wash. Ac. Sci. ii. (Aug. 20th, 1900) p. 234. Diagnosis. Allied to Rh. borneensis, and of about the same size, but with much larger skull and teeth. Lower leg considerably longer: 21 mm. Tail comparatively very short: 17 mm. Fore-. arm about 45 mm. Details. In addition to the above :——The second phalanx of the third finger is more than 14 the length of IIT.’; this is the first time we have to note a decisive lengthening of III. in Lh, bor- neensis, as in all the foregoing species, III.” (always, i in this paper, measured without the terminal cartilaginous rod) is invariably less than 14 the length of III.’; compare with this Rh. stheno, thomas, affinis, ferrum- equinum. IV.! is comparatively shorter than in Rh. borneensis, only about 7 the length of the meta- carpal of the same finger ; compare with this 2h. stheno. Colour. 2 ad. (type); September; teeth almost quite unworn ; first preserved in formalin, now in alcohol; probably unfaded.— ‘“¢ Mars-brown ” above; base of hairs “ ecru-drab”; of a peculiar yellowish ‘‘ drab” beneath (?the yellow due to the influence of formalin). Skull. Of the same general shape as in Lh. borneensis, but much larger, with considerably larger teeth, and therefore longer tooth-row ; orbital constriction very narrow. The following measurements, in millimetres, will give a more precise idea of the differences (the ciphers in brackets are the measurements of eleven skulls of 2h. borneensis) :—total length, inion to front 1905. | OF THE GENUS RHINOLOPHUS. 91 of canine 21-2 [18-2-20]; length of brain-case, inion to anterior point of proencephalon 13-7 [11-3-12°5]; width of brain-case above zygomata 9°5 [7-9-8:2]; zygomatic width 10°8 [9-9-91; maxillar width, across antero-exterior corners of m* 8°5 {6°8-7-2]; inter- orbital constriction 2:2 [2°4-2°8]; palatal bridge, median length 2°6 [1°8-2°3]; maxillar tooth-row 8°7 [7-7-6]; extreme width of m! 2-2 [1°5-1°9]. Dentition. I have not seen the mandible of this Bat. p° in row; cusp almost imperceptible. Measurements*. On p. 92. Type. 2 ad. (in alcohol). Pulo Siantan, Anambas Group ; September, 1899. Collected by Dr. W. L. Abbott. Un. St. Nat. Mus. no. 101714. Remarks. As already pointed out above, the Bats of the borneensis type inhabiting the 8. Natuna and Karimata groups, rather close to the north-western and western coasts of Borneo, are so extremely like the typical borneensis as to be, perhaps, searcely separable. But farther westwards, on the much more isolated Anambas Islands, the borneensis type has developed into the present, peculiarly modified species. In the lengthening of I11.7, the shortening of IV.!, and the shortening of the tail (com- pared with the tibia), Rh. nereis has taken the same course as the still more western Lh. stheno (described below). But the shape of its skull sufficiently proves it to be an offshoot, not of that species, but of 2h. borneensis. Compare with this the “remarks ” under 2h. stheno. 10. RHINOLOPHUS STHENO, sp.n. (Plate IIT. fig. 8, a, 6.) Diagnosis. Allied to Rh. borneensis, but anterior nasal swellings much more projecting. Lower leg long: 19°8-20°8 mm. Tail extremely short: 15°5-17°8mm. Slightly larger than borneensis : forearm 45-2-48 mm. Details. This is a third modification of the borneensis type, in several respects recalling Rh. nereis, in others quite peculiar. The shape of the facial portion of the skull is unique within the present group. As in Rh. nereis, III" is lengthened, IV.’ shortened; the tail is extremely short. The general size of the animal is slightly increased. Plagiopatagium inserted 1-3 mm. above the ankle-joint. Colour. 3 ad., Penang; teeth unworn; skin.—General 1n- pression: reddish brown above; under side much lighter, con- trasting with the upper side. ‘“ Mars-brown” above; base of hairs light “drab”; under side almost “ broccoli-brown.”— Three spirit-specimens (Selangor; teeth unworn) apparently agree In colour with the skin. Skull (three individuals). Owing to the much more projecting anterior nasals wellings, the skull of Rh. stheno, in side view, is strikingly different from that of Rh. borneensis. This peculiarity * The tip of the ears and the posterior nose-leaf are damaged ; forearms broken. 92 MR. K, ANDERSEN ON BATS [May 16, in its outline is produced, not by a heightening of the anterior swellings, but by a reduction of the posterior pair; these latter, which in all the allied species form a sort of transition between the anterior swellings and the adjacent part of the supra- orbital crests and interorbital constriction, are in stheno so much reduced as to leave the anterior swellings more isolated, @. e., more abruptly projecting. Dentition. p, external; p, and p, im contact; p* in row, cusp extremely small. Measurements. Below. Type. 3 ad. (in alcohol). Selangor, Malay Peninsula. Pre- sented by H. N. Ridley, Esq. Brit. Mus. no. 98.3.13.1. Distribution. Selangor ; Penang. Remarks. Rh. stheno differs from Rh. borneensis in the series of characters pointed out above. From &h. nereis, in the shape of the facial portion of the skull, the much slenderer brain-case, and the shorter tooth-rows. From 2h. rowwi, in the shape of the facial portion of the skull; the much shorter metacarpals (al- though the forearm 1s of the same length as in smaller individuals of pour) the long III.’ (compared ‘with ILI. ); the short IV.’ Measurements of Rhinolophus malayanus, nereis, and stheno. \Rh. malayanus.|Rh. nereis.| Rh. stheno. 3 specimens, @ ad. 4, specimens, | 2 skulls. Type. 3 skulls. Min. Max. Min. Max. mm. mm. | mm. mm. mm. Kars, length .. A sae tne PETA aOR, ALOIS ae Wy es > greatest breadth eel e tl toss 8 1 13-7 18. ery | Nose- leaves, TOWAN NEM — soo cveoccesocasl| SFA) SERA a | 138 14:2 55 breadth of horseshoe... 73 8 | 9 1 8 $3 Forearm.. Saasdod tide Godan care encws socntee| meee Ate) 245 |} 45:2 48 3rd metacarpal — Bor Ate eee Li dente ape ee nea(), MESS TL | PSeea 9) Biles Sys JOU eee Bee Se oe Ree ear eter emt ft ic Bane beg 13°2 12°6 18 NEE ee he AE Geach cis Bas callie LOTS GIB: al Ql 201 217 | 4th metacarpal — LUNE oagheenecumoostecaoal ly Ora nail an a nner? 330 38a'8 er OME AL nee ate) BAP NOT Sy sm MEN MOLT FSS, | JE NWP. D La iso Spao ee bakaaiaiacriac crs sericea thee te) Go) 105 | 12°8 11 125 | 5th metacar pal SAbsruope al adomeo Ud eor ote 30. 315 =| 34 335 3842 | Veena ce oe mPen she acetal usae a machetes Oy OS | 10°8 9 104 | ane 97 10 | 10°2 105 115 | Tail PER so ZRRMEL OM EMionn Mia eee Lp gOS eM wah lg 155 17:8 | Tio werwle ge: nsenrt 6 ade terccacieccteioneiaee 168 178 | 21 19°8 203 | Foot ....... Se BHOU ROE CHORD OES eS eaten | 9°3 85 92) | Skull, total leneth tee ate eee op b anne ahi tone! 21-2 197 202 | 57 DASTONG! WAH 555m 060 280002 00008 Scar 10:2 9:3 10 > Width of brain-case ...............| 8 3: 9°5 Siommoud 5 AVEOTIENNG WACWIN so6o0c0n5c0a0cee00) 9: aa 108 Pee LOwt ;, supraorbital length .. eG Sulee O2aee| 56 5 sik | width of nasal sw ellings ‘ee | 52 56 5'8 ee 55 | Mandible, length .. Geanaeouncaal e Lieel ee ea! es [Por ls ba be tel Upper teeth Mite aM LaRue a hl aca Veal 8:7 Cs 1D | howerrbeetht tee cncce racer seneerercmniie t3 dio a Sl ee | 1905. | OF THE GENUS RHINOLOPHUS. 93 (compared with the fourth metacarpal); the excessively short tail; and the smaller hind foot. Phylogenetically, Rh. stheno is evidently more closely connected with Ah. nereis than with any other hitherto known Bat, To call the resemblance between these two species (in III.*, [V.’, the tail) “ convergence,’ would be a phrase only, not an explanation, There can scarcely be any doubt that the type of Rhinolophus to which the now existing Rh. borneensis belongs, sent off a branch west- wards; a part of this branch, isolated on the Anambas Islands, developed into Fh, nereis; another part, in the Malay Peninsula, into Fh, stheno (cf. the diagram on p. 120). 11. Rurotopnus rouxt Temm. (Plate IIT. fig. 9 a, b, ¢, d.) Diagnosis. Allied to Rh. borneensis, but larger, and with con- siderably longer metacarpals. Third metacarpal 34-38 mm. Forearm 46—51°5 mm. Details. This is a large, continental representative of the borneensis type, characterised chiefly by the much longer meta- earpals and the shape of the lancet. In general size, the continental 2h. rowxi bears the same relation to the insular Rh. borneensis as the continental Rh. megaphyllus does to the insular Lh. simplex. The sella is practically parallel-margined from base to summit ; not rarely some faint indication of a constriction at the middle can be traved ; summit broadly rounded off. In simplea and its closest allies the lancet is long and quite (or almost) cuneate ; in borneensis there is some tendency towards a slight emargination of the lateral margins of the lancet ; this tendency has been carried almost to an extreme in rowai: the lancet is hastate, i. e., abruptly narrowed in the middle, the tip well developed and slender (not abnormally shortened, as in thomas); but still, individually (though, as it seems, rather rarely), in rout, the lancet is less abruptly narrowed, as an atavism towards a passed stage, The ears are as in borneensis. Wing-structure almost on the simplex-borneensis stage, 7. e., III? almost always less than 13 the length of III.!. The rare individual exception, that IIT.* is equal to (or a mere trifle more than) 13 the length of ITI.', is of some interest as fore- shadowing the next important step to be taken in the series of evolution, viz., from rowxi to affinis, in which species III.” is always considerably more than 14 the length of IIT.’ Plagiopatagium inserted on, or 1-4 mm, above, the tarsus, 7. e., there is evidently some tendency to draw the insertion of this membrane away from the ankle-joint, a little higher up on the tibia; compare with this 2h. afinis. The proportionate length of the tail is as in borneensis. Skull. The skull of Rh. rousxi is larger than that of borneensis, but I fail to find any appreciable difference in the shape—a strong evidence of the very close relationship between the two species. The individual variation in the size of the skull, in 94 MR. K. ANDERSEN ON BATS [ May 16, vouxi, is rather considerable (as is also the variation in the ex- ternal dimensions of this Bat); but among 18 skulls of the typical form of rowxi, from localities so many and so distant inter se as to represent practically the whole area covered by this form, I do not find any so small as the largest among 11 skulls of borneensis (and 6. spadix); in so far there is no difficulty in diseriminating them. The tooth-rows, too, in rowaxi, are longer. As to the small 8. Chinese race of row«i (described below), the skull has the same length as the largest of borneensis, but the brain-case is decidedly broader, the zygomatic and maxillar width greater. Dentition (19 skulls). p,, most often, quite external (12 skulls) ; not rarely half in row, or ? in row (6 skulls); in one aged individual (teeth much worn) p, is wanting, on both sides of the mandible, and the alveoli have disappeared. Cingula of p, and p,, most often, in contact or separated by a very narrow, sometimes almost hairfine, interspace (13 skulls); in the remaining (6) individuals, distinctly separated, but the width of the interspace is not always quite the same on both sides of the mandible. The upper canine and p‘ are, with rare exceptions, distinctly separated, p* completely in the tooth-row (17 skulls, out of 19), as in all the foregoing species. The size of p* and, therefore, the width of the interspace between ¢ and p‘ vary, however, to a certain extent; but in »o instance is the width of the interspace as broad as (p° as well developed as) in stmplea: this is a thing of the past. As to the remaining two skulls (Ceylon, Nepal), the interspace is very narrow, p° half external. This 1s the first time we have to note instances of p’ not being completely in the tooth-row. As a general conclusion :—(1) In “Ah. rowxt p, has arrived so far on its way towards disappearance as to be, generally, external; but still, not rarely, the individual variation falls back to a former stage: p, partly in the tooth-row; and in some aged individuals the dentition (p, disappeared) points forwards to sub- sequent stages in the series of evolution: Rh. ferrwm-equinum (p, rather often lost) and 2h. acrotis (p, always lost). (2) As to p’ in rouwi, it is generally in the row, rarely half external; this latter, again, points forwards towards subsequent stages: thomasi, ferrum-equinum, and acrotis (p° always external, or lost). Distribution. From 8. China through the Himalayas to the Indian Peninsula and Ceylon. Technical name. As Rh. rouxt has for many years been com- pletely confused with Ah. affinis, some remarks are necessary to prove that the name rowxt belongs to the species here under consideration. The type locality of Rh. rowai is “ Calcutta” * ; the types (in the Leiden Museum) were collected by the French naturalist, M. Roux. There is in the Tomes Collection (British Museum) a skin also collected by Roux. The essential points ~ * Temminck, loc. infra cit., p. 30 ¢; Jentink, ‘Catalogue systématique des Mammiferes,’ Mus.:d’hist. nat. Pays-Bas, xii. (1888) p. 161 (under Rh. affinis). 1905. ] OF THE GENUS RHINOLOPHUS. 95 in the original description as given by Temminck are the following :— (1) In “taille, forme du corps, des oreilles et des follicules accessoires du nez” very much like Java specimens of Zh. affinis Horsf. It may be said so; the difference in the shape of the sella is not easily ascertained in dried skins. (2) “ Des proportions moins grandes,” as compared with affinis. As measurements Temminck gives :—Of rowxi: forearm “1 pouce 10 lignes” (49°5 mim.), expanse of wings “10 pouces.” Of affinis: forearm ‘1 pouce 10 lignes,’ expanse “11 & 12 pouces.” 49-5 mm. is one of the commonest measurements of the forearm in the series before me. It looks a little contradictory that Temminck, having stated that rouxi is smaller than affinis (which is quite correct), gives precisely the same measurement of their forearms, though, at the same time, a considerably larger “expanse” of the latter species. But just that is the salient point. Asa matter of fact, the two species can have the forearm of exactly the same length (very large rowai, and small affinis) ; but also in that case, the expanse of Rh. aflinis is always markedly larger than that of Rh. rouxi, for the obvious reason that in the former species the second phalanx of the third (longest) finger is always absolutely longer than in the latter. (3) A ved, a dark, and an intermediate phase of rowaxi were known to Temminck. I have the same phases before me. That similar phases occur in Ah. borneensis has no bearing on the present technical question; borneensis lives far away from “ Calcutta.” The “phases” of Rh. affinis are different. (4) “Les molaires de la machoire supérieure sont en méme nombre que dans lafinis, celles de Vinférieure en compte cing, ou une de moins, par le manque total de la petite dent dont lafinis est pourvu, et qui forme la sixiéme molaire.” Since Temminck emphasises the “‘ manque total” of p,, I suppose that he has not overlooked this small tooth, but has examined a (probably aged) individual in which it was wanting (cf. the specimen mentioned above). The word “ sixiéme” is, of course, a lapsus for “cinquiéme” (Temminck counted the “molars” from behind forwards). To sum up:—There can be no doubt that Temminck’s RA. rowni is the Bat here under consideration, being a species (1) bearing much resemblance to 2h. affinis ; (2) of almost the same size, but with a markedly smaller expanse of wings ; (3) with a red, a dark, and an intermediate phase; and (4) inhabiting the Continent of India. “Rh. peterst.” —The original description of Rh. petersi is meagre and vague ; the figures of the head and nose-leaves published four years later are badly drawn; the type specimen (in the Calcutta Museum) has no indication of locality. This may sufficiently account for the fact that no technical name in the genus has been the source of more confusion. I therefore think it of some use to give a brief sketch of its rather complicated history in literature :— (a) As to the identification of *“ Rh. petersi,” in the original 96 MR. K, ANDERSEN ON BATS [May 16, sense of the term*, there are only two alternatives: it is either Rh, rowwi or a species of the Ah. acuminatus section, I have not the slightest hesitation in referring the name as a synonym to the former species. As, however, Dobson himself later on applied the name to two Bats of the aewminatus section, it will only be necessary to give evidence, from his own description, that he was mistaken. The only important points in the description of “ Rh. petersi” as given by Dobson in 1872 and 1876, i.e. at the time when he had access to the type specimen, are the following (the italics are mine)—(1) The nose-leaves are “as in Lh, acuminatus, except the upper border of the posterior connecting process, which is much less acute.” This statement alone would be sufficient. In acwminatus the shape of the sella and lancet is very much as in rouwai, but the connecting process, both in acuminatus and in all its allies (swmatranus, calypso, wudax), 1s projecting and pointed; there is, in this respect, no difference between the species of the aewminatus section, and there is also no appreciable individual variation. When, therefore, Dobson in this decisive point (the chief character of the whole group to which acuminatus belongs) declares his Rh. petersi to be very different from acuminatus, it may safely be said that it has nothing to do with that group. Dobson had evidently before him an example of Rh. rouxi with a slightly raised connecting process (‘‘ much less acute” than in acwminatus); such individuals are by no means rare ; there are severalin the British Museum, and the peculiarity is purely individual. Dobson found, quite naturally, that this peculiarity recalled that shape of the connecting process which had been described, one year earlier, by Peters in a species called by him Rh. acuminatus Tt, and, consequently, he compared it, in his paper, with this latter species, at the same time emphasising that there was a considerable difference. (2) The figure (side view) in Dobson’s ‘ Monograph,’ however bad it is, can scarcely represent the shape of the connecting process in acwminatus. Dobson has, no doubt, called the attention of his artist to the connecting rocess of the specimen to be figured as 2h. petersi, and the artist, in due obedience, has made his best to ‘‘emphasise” that point : this may account, I think, for the process being somewhat more exaggerated than in ordinary individuals of rowxi; but it is still not the process of an acuminatus. (3) The measurements of petersi are, without any exception, perfectly like those of several unquestionable specimens of rowai measured by myself ; there is not the slightest indication of a difference. (4) The type of petersi is from “ India, precise locality unknown.” The aewminatus section is distributed over Sumatra, Engano, Java, and Lombok. When Dobson wrote his ‘Monograph,’ there was not, in the Calcutta Museum, any specimen.of any species of Rhinolophus from those islands; so that, if RA. peterst were a member of the acuminatus section, the type, without locality, would have been * Dobson, J. A. S. B, xli. pt. ii. (Dece 22, 1872) p. 337; id., Monogr. Asiat. Chir. (1876) p. 49, text-figs. a, b. + Peters, MB. Akad. Berlin, 1871, p. 302. 1905. | OF THE GENUS RHINOLOPHUS 97 the only Rhinolophus in the museum from any of those islands. This is, of course, not beyond the limits of possibility ; but it is certainly much more likely that RA. petersi, as also the vast majority of the Bats in the Caleutta Museum at Dobson’s time, came from some part of the Indian Peninsula or the Himalayas, the habitat of Rh. rowxi, and far from the home of Rh. acuminatus and its allies. To describe a new species which subsequently proves to be an old one is no rare occurrence, and, as a rule, it does no very serious harm. But the strong emphasising of a purely individual peculiarity, combined with the circumstance that the type had no “ locality,” caused in this case a series of confusions: Rh. petersi emerged, like a ghost, very unexpectedly at such different places as the Gold Coast, Sumatra, the Himalayas, and 8. India. And, curiously enough, the author of the “species” inaugurated the mistakes. When he had returned to London and was working out his ‘Catalogue,’ Dobson had no longer access to the type of Rh. peterst; he had his own short description only, and perhaps some private note. It is quite evident that, in these circumstances and occupied with the study of many other Bats, he lost the precise idea of the type specimen ; he only kept in his memory, as its most important character, its ‘‘ projecting” connecting process. So it eame that he referred a specimen labelled ‘Gold Coast” to Rh. petersi*; for it is a genuine acuminatus, beyond all doubt from Java, and Dobson himself would scarcely have been able to tell why he called it peterst instead of acuminatus. Two years later, Dobson had for determination a collection of Bats belonging to the Gottingen Museum; among these he again believed he found a Fh. petersit. I have had this example for inspection £; it is neither ‘“‘ Rh. petersi” nor Rh. acuminatus, but Rh. sumatranus. (6) In a paper on some Himalayan Bats, Capt. Hutton § records Rh. petersi from Masuri. All the Bats mentioned by Hutton were presented to the ‘‘ Indian Museum,” and are now in the British Museum. The two specimens labelled “ Rh. petersi” are Rh. monticola, a species closely allied to Rh. lepidus ||. * Dobson, Cat. Chir. Brit. Mus. (1878) p. 114. + Dobson, “On some new or rare Species of Chiroptera in the Collection of the Gottingen Museum,” P. Z.S. 1880, p. 462. ¢ I am indebted to Geheimrat, Professor Dr. Ehlers, Gottingen, for the loan of this specimen. § Hutton, “On the Bats of the North-western Himalayas; with Notes and Correc- tions in Nomenclature by Prof. W. Peters,” P. Z.S. 1872, p. 700. || As Hutton’s article is one of the very few papers which give information respecting the habits of Himalayan Bats, and therefore has been frequently quoted by subsequent writers, I think it advisable to correct the following errors in the identifications of the four species of Rhinolophus dealt with in that paper :—“‘ Rh. affinis’’ (p. 696) is Rh. pearsoni; “ Rh. rouwi” (p. 697) is Rh. affinis ; “ Rh. minor” (p. 698) is Rh. rouxi; and, as pointed out above, “ Rim petersi” (p. 700) is Rh. monticola. Hutton’s Bats were (as also stated in his paper) determined, not by himself, but by Prof. Peters in Berlin. But the mistakes are so strange that they cannot, certainly, be due to Prof. Peters; an extensive confusion of labels must have occurred (I can rather easily, from Peters’s point of view, as laid down in his papers, guess the original arrangement of the labels), but the confusion had at all events taken place before the specimens were returned to Hutton. Proc. Zoou. Soc.—1905, Vou. II. No. VII. 7 98 MR. K. ANDERSEN ON BATS [May 16, (c) In Blanford’s ‘Fauna of British India’ (loc. injra cit.) Rh. peterst is recorded from Masuri and from Nilghiri. The former statement is borrowed from Hutton’s paper. The latter is based on an example collected by W. Davison in Coonoor, Nilghiri*. This specimen is now in the British Museun. It is a Ah. rouse. In short:—(1) For reasons given above I regard Dobson’s Rh. peterst (1872 and 1876) as a synonym of Lh. rouxt; (2) Dobson’s Rh. petersi (1878) is Rh. acuminatus; (3) Dobson's Rh. peterst (1880) is Rh. swmatranus; (4) Hutton’s hh. petersi is Rh. monticola; (5) Blanford’s Rh. petersi is partly Rh. monticola (Masuri), partly Rh. rowaxi (Nilghiri). Geographical races. There are, at least, two forms of 2h. rouwi, differing in size and geographical habitat. lla. RHINOLOPHUS ROUXI SINICUS, subsp. n. Diagnosis. Skull smaller, tooth-rows shorter. Forearm 46 mm. Details. The general size is as in the very smallest examples I have seen of the typical form. Skull still a little smaller, with slenderer brain-case and shorter tooth-rows; nasal swellings, in front view, slightly lower. Colour as in the dark phase of Himalayan specimens of the typical form (see below). Measurements. On p. 100. Type. & ad. (skin). Chin Tah, Anhwei, Lower Yangtse 7. Presented by W. Styan, Esq. Brit. Mus. no. 99.3.1.6. 116. RuroLopuus Rouxt Temm., TYPICUS. Rhinolophus Rowxit Temminck, Mon. Mamm. ii. 8° monogr. (1835) p. 300. Rhinolophus rubidus, cinerascens, rammanika Kelaart, Prodr. Faune Zeylanice (1852), pp. 13, 14. Rhinolophus Rouxii (partim) Peters, MB. Akad. Berlin, 1871, . 308. ; Rhinolophus petersii Dobson, J. A. 8. B. xli. pt. ii. (1872) p-. 337 (nec Dobson, 1878, 1880); Blanford, Fauna Brit. India, Mamm. pt. ii. (1891) p. 275 (partim). Rhinolophus minor (non Horsf.) Hutton, P. Z.8. 1872, p. 698. Rhinolophus affinis (partim, nec Horsf.) Dobson, Cat. Chir. Brit. Mus. (1878) p. 113. Diagnosis. Skull larger, tooth-rows longer. Forearm 46—01'5mm. Colour.—(1) Specimens from Nepal and Darjeeling. (a) Dark phase: onead.; Nepal; teeth unworn ; skin :—Upper side “ mars- brown”; horse-shoe patch on back distinguishable, though some- what obliterated ; base of hairs light ‘‘ drab,” almost ‘ ecru-drab ” ; under side “drab,” with a tinge of ‘“russet” ; sides of body some- what darker. With this skin agree in colour another adult specimen from Nepal (teeth somewhat worn; skin) and a 9 ad. from Darjeeling (in alcohol). * Blanford, J. A.S. B. lvii. pt. ii. no. 3 (1888) p. 261. + For the exact position of this locality, see ‘ Ibis,’ 1899, p. 289. 1905. | OF THE GENUS RHINOLOPHUS. 99 (6) Light phase: one ad.; Darjeeling; teeth slightly worn ; skin :—Above inclining to “clay”; a strongly marked, deep brown horse-shoe patch; base of hairs and fur of under side almost “ cream-buff.” (2) Specimens from Ceylon and India.—{a) Dark phase: three adult individuals; Ceylon; teeth rather slightly worn; skins :—Upper side a shade of brown, darker and duller than “mars-brown ”; horse-shoe patch more or less effaced; base of hairs “drab,” with a tinge of ‘ ecru-drab”; under side ‘‘ wood- brown” or light “ drab.”—This is Kelaart’s 2h. cinerascens. A skin (ad., January, teeth unworn) from Sirzi, Kanara, comes extremely near to the last-mentioned specimen, being only a little darker. A. spirit-specimen from Nilghiri seems to be of very much the same colour. (b) Intermediate stage: 3g ad.; January; Sirzi, Kanara; teeth unworn. Upper side between “russet” and ‘ mars-brown ” ; base of hairs ‘‘ ecru-drab” ; under side almost “ clay.” —This is Kelaart’s th. rammanika. (c) Red phase: one ad.; Ceylon; teeth worn; skin :—Above light “hazel” with a tinge of ‘orange-rufous”; horse-shoe patch almost obliterated ; base of hairs and under side of body light “ orange-rufous.”—This is Kelaart’s Rh. rubidus. A skin (¢ ad., February, teeth unworn) from Jellapur, Kanara, represents the extreme of light colour: upper side next to “ tawny- ochraceous”; base of hairs and fur of under side almost ‘“ orange- ochraceous.” Conclusions :—-The dark phase in specimens from the Himalayas (Nepal, Dar ‘jeeling) 1 is of a richer brown, more tinged with russet, than in specimens from Ceylon and 8. ‘India (Kanara, Nilghiri). The light phase, in specimens from the Himalayas, seems to be more inclining to “clay”; in specimens from Ceylon and§. India more “hazel” or * tawny-ochraceous.” I do not think the series examined affords evidence conclusive enough to justify the sepa- ration of a Himalayan “race” and a soit hern (Ceylonese and S. Indian) “race.” In all the other characters (external, cranial, dental; variation in general size) there is no appreciable dif- ference. If they were *t0 be separated subspecifically, the southern form would have to stand as “ 2h. rowai rubidus Kelaart,” the Himalayan as ‘“ 2th. rouxt typicus.” Measurements. On p. 100. Distribution. Himalayas (Darjeeling, Nepal, Masuri). 8. India (Nilghiri, Kanara) and Ceylon. Remar hs. Of the two forms here recognised, 2h. rowan sitirvicus and Rh. rowxi typicus, the former, as coming nearest to Lh. borneensis, is no doubt the more primitive. The rouai-type, therefore, has spread from an eastern point of the continent westwards, through the Himalayas, down the Indian Peninsula, to Ceylon. Q/ 100 MR. K. ANDERSEN ON BATS [May 16, 12. RHINOLOPHUS THOMASI, sp. n. (Plate ITI. fig. 10.) Rhinolophus affinis rouxi? (non Temm.), Thomas, Ann. Mus. Civ. Genova (2) x. (1892) p. 15, pl. xi. fig. 3. Diagnosis. Allied to Rh. rouxi, but p* external to the tooth-row. Smaller than rowwi, with considerably shorter metacarpals, and the tip of the lancet excessively shortened. Third metacarpal 30°-4-31 mm. Forearm 44°8-45-7 mm. Details. While being similar to 2h. rowwi in the shape of the sella and the ears, and the proportionate length of the tail, Rh. thomasi differs, externally, from that species in the following particulars :— The horse-shoe is considerably narrower ; it is even narrower than in the smaller borneensis and in the much smaller malayanus. - The tip of the lancet is exceedingly short, almost rudimentary ; it is the hastate lancet of rowai carried to an extreme. The general size is smaller, as seen by the measurements of the forearm. But the metacarpals are proportionately mach shorter, as short as inthe much smaller malayanus. ITI.° is comparatively Measurements of Rhinolophus rouxi and thomasi. Rh. rouxi. | Rh. thomasi. sinicus. Eypicus. 6 ad. 30 specimens, || 2 specimens, Type. 18 skulls. | 1 skull. Min. Max. Min. Max. min. mm. mm. || mm. mm. Devas, Jes RERAN 289 s3a44s ceaeeaeae coaceanee ony bb0 see 166 19 | 168 168 » greatest breadth.....................| se 13 «86415 |} 12 122 Nose-leaves, total length ...............| Hee 135 162 || 117 118 | F breadth of horse-shoe ...| abe eens 927 = a2) ES Forearm EAR eee Stigs tcc ericnealg AO 46 515 || 448 45°7 Brdumetacatpal...es.eesehsa tesserae | ee 34 = 38 || 3804 31 WORT pre tee conenine cc sean/escmemnemaal pg) hate 137 15'8 13°) 1331 fl U UC 2 BA a a aes ean nantehcosaaRmEBRentas.. aan 20°8 185 25°5 20°2 20:2 Ath metacarpal............c0c0cc | BHT | 3845 389 313 319 PW a aubk fet Aart SEU eats esadacianence tr ikl isla 7 12 10 102 Ia IsWite, sSeeteeny Renee nm natin seater gt sccercenl (mk acer 17 145 = «|| 12:2 12-7 | 5th metacarpal..........-.........:.-.-.-..| 354 354 389 || 3823 327 others sett pada coe ponbosnpoandieegatondacdes 119 106 132 11 11:2 Veo ee ed opments ack valine ures le DZ eS 9 9:7 I ARES uioh onuedude pay aasouebesbooee sadoeacntso4b 21 21 265 || 19 19 Us Gonueese Gea ele ho ndociagsesecandenuescocenel| SA8MS) 19 235 || 18 18 DY GXE a cmences cononet uucases ane sancenene buy adel & 9. 11:2 8 88 Sheol #rotalilene-thissssreeree sire veces ele ae 20°3 23 nish teaiiinn oe = ,, mastoid width .....................] 9°5 ‘O'7 10° 9:2 » width of brain-case............... 87 87 98 8:7 » zygomatic width..................| 108 104 118 10 », supraorbital length ............... 48 48 58 4-4, » Width of nasal swellings......... 5°8 65 59 53 Mandible, length ............0..0...0..[ 185 13 «164 12°8 Wipper teeth 20sec. eesceee essen TT 82 92 Hilt TORRE THA, icone cod eau see ane0oo cbosons00080 81 85 103 ai 1905. | OF THE GENUS RHINOLOPHUS. 101 longer than in rouwi, 7.e. more than 13 the length of IIT.’ (cf. neveis and stheno). V. is extremely short. Colour. To judge from specimens preserved in alcohol, probably not far from being the same as in the dark phase of Nepal examples of Rh. roux. Skull. The essential characters are as in rouat, thus proving Rh. thomasi to be an offshoot from that type of Bat, not (as might very well be supposed, in view of the short metacarpals) from borneensis. The skull of Rh. thomasi agrees with that of rowat im the broad brain-ease ; it differs from rowxt in the much smaller size. Compared with borneensis, the skull of 2h. thomast 18 as small ag in the smallest individuals I have seen of borneensis (even as small as in malayanus), but the brain-case is markedly broader, even broader than in the largest bormeensis, and the supraorbital length is exceedingly short (¢/. measurements, p. 100). Dentition. p, external; p, and p, in contact; p* external, Upper canine and p* in contact. Both of the specimens examined are identical in dentition. Measurements. On p. 100. Type. 2 ad, (in alcohol). Karin Hills, Burma, 1888. Collected by Signor Leonardo Fea. Presented by Marquis G. Doria. Brit. Mus. no. 90.4.7.10. I venture to connect with this fine species the name of Mr. Oldfield Thomas, who already thirteen years ago (J. s. ¢.) pointed out that it could scarcely be identified with any hitherto known form, but refrained from describing it as new, owing to the general confused state of this group of Bats. 13. Ruronopuus arrints Horsf. (Plate HT. figs. 11-13.) Diagnosis. Sella pandurate. p° in the tooth-row. Forearm 50-56 mm. Details. This species marks an important progress in develop- ment as compared with 2h. rowwi. It is the base of the ferrum- equinum section. The chief modifications are four: in the shape of the sella; in the structure of the wings; in the size of the animal; in the shortening of the palatal bridge. In the borneensis-rouwi type the sella is practically parallel- margined; in afinis it is pandurate, %. ¢. the lateral margins concave, as in ferrum-equinum, though generally to a slightly less degree. In simplex and its closest relations the lancet is almost cuneate; in borneensis there is a tendency towards emargination of the lateral margins; in rows this tendency is carried to an extreme ; in afinis the lancet falls back to the former stage : it is almost cuneate. E Throughout the whole series of forms reviewed above, with the exception of the somewhat aberrant Rh. nereis, stheno, and thomas, the wings have remained at the same primitive stage : no length- ening of the second phalanx of the third finger. In affinis this phalanx has considerably increased in length, being always more 102 MR, K, ANDERSEN ON BATS: [May 16, than, and with very rare exceptions considerably more than, 14 the length of the first phalanx, a peculiarity which is preserved in the subsequent stage of evolution: ferrwm-equinum. The aberrant species just alluded to, viz. Rh. nereis, stheno, and thomasi, are, from this point of view, of especial interest, as being Bats of the rouxt type which already show the wing-structure characteristic of the more highly developed affinis. kh. affinis is larger than rouxi ; but small affinis have the same length of the forearm as very large rouxi. In such cases, Ph. afjfinis, provided the specimens examined are fresh or preserved in spirit, can, of course, easily be discriminated by the shape of the sella and the length of III’; if preserved as dried skins (in which the shape of the sella is often difficult to recognise), still the latter character remains unchanged. Colour. The many forms in which this species is differentiated seem to agree, rather closely, in colour :—- (1) Darker individuals: gad., Darjeeling (2h. a. himalayanus) ; Oct. 22nd; teeth unworn; skin :—Upper side “ mars-brown” with a rather strong hue of “drab”; no horse-shoe patch; base of hairs ‘“ ecru-drab”; under side ‘ broccoli-brown.” Still darker isa ¢ ad. from Lombok (2h. a. princeps); teeth ° somewhat worn ; in alcohol; unfaded :—‘ Prout’s brown” above, base of hairs ‘‘wood-brown”; under side almost “ tawny-olive.” (2) Light-coloured individuals: ¢ ad., Nanking (2h. a. hima- layanus); July 5th; teeth somewhat worn; skin :—Extremely light. Above light “clay,” almost “ ochraceous-buff,” hinder back somewhat darker; a rather distinct, ‘‘ mars-brown ” horse- shoe patch; base of hairs “ cream-buff”; under side very light, almost ‘ cream-buff.”—-A spirit specimen ( gf ad.) from the same locality (June 15th) is quite of the same colour. Skull. The essential characters as in rouaxi, proving that Rh. affinis originated from a Bat of that type. The skull is generally larger, and the gap in front between the maxillary bones wider. Chief character: the exceedingly short palatal bridge, as a rule only 7 the length of the maxillary tooth-row, or even less; in rowxi, with very rare exceptions, decidedly more than t, sometimes almost 3. The teeth, too, are slightly larger. Dentition. p, external and extremely small; but, as a rare exception, this premolar may still, in this comparatively highly- developed species, show some tendency towards the tooth-row (one skull, out of 19), or be halfway in row (one). p, and p, generally quite, or almost, in contact (14 skulls); in the remaining somewhat more distinctly separated. p> always in the tooth-row, extremely small, and the interspace between the canine and p* rather narrow. In no less than five skulls there isan exceedingly narrow, In most cases almost hair-fine, interspace between p* and p' (the former place of p’) Distribution. From the N.W. Himalayas to 8. China; through Indo-China, the Malay Peninsula, and N. Natunas, to Sumatra, Java, and Lombok. 1905. | OF THE GENUS RHINOLOPHUS. 103 Technical name. The type of Rh. affinis is in the British Museum. From the original description it would have been quite impossible to identify the species. Remarks. Of all the races of Rh. affinis, the Himalayan form (Rh. «a. himalayanus) is the most ordinary-looking: in_ the horse-shoe, the ears, the nasal swellings, the brain-case. There can hardly be any doubt that the affinis type originated in the Himalayas, and from there spread eastwards to 8. China, south- eastwards through Indo-China, as far as Lombok. Geographical races. There are, at least, seven forms of Rh. affinis, differing in certain cranial characters, in the size of the ears and horse-shoe, in the length of the tail and tibia, in general size, and in geographical habitat. Some of these forms may be called distinct species by other authors. 13@. RHINOLOPHUS AFFINIS HIMALAYANUS, subsp. n. (Plate IIT. fe. Ue, b.) Rhinolophus affinis (pavtim) Dobson, Cat. Chir. Brit. Mus. (1878) p. 112. Diagnosis *, External characters :—Size largest ; ears small ; horse-shoe narrow; tail short ; lower leg short. Cranial: length of skull, width of brain-case, length of tooth-rows, moderate ; nasal swellings narrow. Type. Q ad. (in alcohol). Masuri. Collected and presented by Capt. Hutton. Brit. Mus. no. 79.11.21.148. Distribution. Himalayas (Masuri, Nepal, Darjeeling) ; 8. China (Nanking). 13 6. RHINOLOPHUS AFFINIS TENER, Subsp.n. (Plate ITT. fig. 12.) Diagnosis. External characters: Size small; ears small, horse-shoe broader; tail short; lower leg rather long. Cranial : skull short; nasal swellings and brain-case narrow; tooth-rows short. Type. 3 ad. (in aleohol). Pegu. Collected and presented by W. Theobald, Esq. Brit. Mus. no. 87.5.4.11. 13c. RHTNOLOPHUS AFFINIS MACRURUS, subsp. n. Rhinolophus afinis Thomas, Ann. Mus. Civ. Genova (4) 3, (1892) p. 922. Diagnosis. External characters: Size moderate; ears larger ; horse-shoe broader; tail long; lowerleg longer. Cranial: length of skull, width of brain-case, length of tooth-rows, moderate ; width of nasal swellings moderate. Type. gad. (inalcohol). Taho, Karennee, Burma; Febr. 1888. Collected by Signor Leonardo Fea. Presented by Marquis G. Doria. Brit. Mus. no. 90.4.4.7. % As the characters of the different forms of Rh. affinis are sufficiently clearly expressed in the table of measurements, p. 105, they will not be reviewed im detail, but only rendered in general terms, in the “ diagnoses” of the subspecies. 104 MR, K. ANDERSEN ON BATS [| May 16, 13d. RHINOLOPHUS AFFINIS SUPERANS, subsp. n. Rhinolophus afinis (pavtim) Peters, MB. Akad. Berlin, 1871, p- 306 ; Dobson, 1. s. e. Diagnosis. External characters: As macrurus, but with short tail. Cranial: skull rather long; nasal swellings still broader than in macrurus; brain-case broad ; tooth-rows rather long. Type. 9 ad. (in alcohol). Pahang, Malay Peninsula. Pre- sented by the Selangor Museum. Brit. Mus. no. 0.7.3.2. Distribution. Lower Siam (Trong); Malay Peninsula (Pahang) ; Sumatra. Remarks. A specimen from Sumatra is in every respect, cranial, dental, and external, indistinguishable from those from Pahang and Trong (the latter sent for identification by the United States National Museum). 13. RHINOLOPHUS AFFINIS NESITES, subsp. n. Rhinolophus affinis Gerrit 8. Miller, Jr., Proc. Wash. Ac. Sci. iii. (1901) p. 135. Diagnosis. External characters: As superans, but smaller, and with shorter tibia. Cranial characters unknown. Type. Qad.(inalcohol). Bunguran Isl., N. Natunas, Aug. 24th, 1900. Collected by Dr. W. L. Abbott. Un. St. Nat. Mus. no. 104753. Remarks. This is evidently an offshoot of the Malacca form, Rh. a. superans, isolated on the outlying N. Natunas, and developed into a well-marked race (or species). It still shows some of the chief characters of swperans: the large ears, broad horse-shoe, and short tail; but, to judge from the metacarpals (the forearms are broken), it is decidedly smaller, it would seem still a little smaller than RA. a. tener, and the tibia is very short. The skull is so much damaged that I have only been able to examine the teeth and the lower jaw. 13. RHINOLOPHUS AFFINIS Horsf., TyPrIcus. Rhinolophus affinis Horsf., Zool. Res. Java (1824), pl. [7], figs. A, B. Rhinolophus affinis (partim) Peters, |. s. c. (1871); Dobson, 1. s. ec. (1878). T am unable to give a definite diagnosis of this, the “ typical,” form of Rh. affinis, having seen only one very old skin (the type) and a fragment of the skull, representing the facial portion and the tooth-rows. But these are sufficient to show, first of all, of course, the specific characters (pandurate sella, lengthened IIT.’, dentition, &c.); secondly, that this form is quite different from any of its next neighbours, on Sumatra and the Malay Peninsula (superans), on the N. Natunas (nesiées), or on Lombok (princeps). The horse-shoe seems, allowing for some shrinkage, to be quite as narrow as in 2h. a. himalayanus ; the nasal swellings, too, are as narrow as in himalayanus and tener. But, although the OF THE GENUS RIHINOLOPHUS. 105 1905.] L6éL TL 8.6€ 6-P1 ¢.01 ¢.6€ VG ¢.cT 9.16 6S LIL 6-81 6.9T €-16 “UU “pe P ‘sdaoutud “TUUL ‘ad AY, “PY. snoidhy 6-01 8.66 GG G-GL ¢.IL 8.96 | 8-71 GOL 9€ | FG PL 6.26 | 8-6 G-V1 6.06 “UU ad Ay, “ped “saqisau TOL 4-6 | 6.6 L.6 6 6-6 6-91 8ST L.9L 19 6.9 B.9 G9 L.S 6.¢ L6t IT 9.11 GOL 8-6 3-6 GIL 6.01 II 8-66 8.66 G6 ol TIL ait 8.96 GG V-6 GGG -€.16 €.66 SST G6L 8-F1 611 &I11 €L c6e LLe Gi 7 FL G.9T 6.0L ol G-GL 8-8E 8-96 L.0F 96 8-86 2.96 GST 9-PT OL 6-8E 9€ 66 eg 66.19 8.€¢ OL 8-6 8.6 11 8-61 ¢.9L LST 8-P1 8.ST £06 6.06 1.06 “TU “UU “UU ‘xB ULIAL “XU “s][NYS ¢ ‘stauttoads ¢@ ‘supsadns 4 10 19 10 do & toe sp gUcyWI~ HE il Ne) a ier) = “WU “ULIN SIMS & ‘suamuoeds ¢ *SMLMLLODU > © 19 19 19 G SD a DADAM ONAN =! ) 6-€1 8.11 “ULL ‘ad AY, “pe é “wag ‘xe BOL 4.6 ¥.6 6 @.9T 8ST a) Om qo ag 8-IL TIT OL 6 6.0L GOT es L.2e eae Bes 8.66 e818 Lp LT Sel ScD Veh 6.68 a = Ol BIL V-OL GIP 8.88 ae Oe oe Ol OF 4-8 Gch ace 8.8 8 TL 6.1 GFL 8.81 GSI BLT *ULUL “TUL “UAL ‘STINYS L ‘suamads 9 ‘snuphpppury ~" Yyoay TIMOTT Uneeth woddy Tiere) ‘OIqIPUB TN © SSUITIOMS jeseu Jo [AprM “yySuey peqqaoradus = YIP orywuosdz = ** aseo-ulntq JO Yyprm Ee ployseur : sue] yeqoy TnyAS 4007 Sa] IOMOTT TL ZA | vA _[edavonjou qg¢ oe AT cAI Te eee “ah? edawomaut pag ) ULIBO.10 il SOYs-as10r] JO TZ pPRa.tq a aoe “* YVSUAT [RJO} ‘SOAual- 9SO NT mu eeand qsayewais =“ YQouey ‘sven | ‘savaadsqns Pup stuyys snydopouryy fo SPMAIUWAMNS DA T|" 106 MR. K. ANDERSEN ON BATS [May 16, specimen is slightly smaller than the smallest example of hima- layanus I have seen, the tibia is fully as long as (if anything, a trifle longer) than in the very largest of these latter. On the whole, I have but very little doubt that Ah. a. typicus will prove to be much more closely related to the Burmese and Himalayan forms than to any of the others. This would be an additional evidence of the closer connection between the fauna of Java and that of Indo-China and the Himalayas—closer than between Java and the geographically nearer Sumatra, Malacca, and Borneo. Distribution. Java. 13g. RHINOLOPHUS AFFINIS PRINCEPS, subsp. n. (Plate III. fig. 13.) Diagnosis. External characters: General size moderate; tail short; but largest in the size of the horse-shoe and ears, and the length of the tibia. Skull, nasal swellings, tooth-rows: the extreme. Type. 3 ad. (in alcohol). Lombok, July 1896. Collected by A. Everett, Esq. Brit. Mus. no. 97.4.18.13. Remarks. Placed side by side with Rh. a. himalayanus, this form is strikingly different; the horse-shoe is no less than } broader than the broadest in himalayanus, and the skull is distinguishable at a glance by its excessive width and the very broad nasal swellings. But it must be remembered that sawperans leads, not up to, but decidedly in the direction of, princeps, and we do not yet know the extreme limits of individual variation, either in superans or In princeps. When considering the geographical races * of 22h. affiiis from a more general point of view—and excluding “ typicus,” owing to the peculiar geological history of Java, as well as nesites, owing to its having, probably, been influenced by somewhat exceptional con- ditions, far away on the small isolated N. Natunas,—the following rule will be observed: the more southern or south-eastern the habitat, the longer the ears, the broader the horse-shoe, the longer the tibia, the larger the skull, the broader the nasal swellings, and the longer the tooth-rows. 5 14. RetNoLopHUS FERRUM-EQUINUM Schreb. (Plate IV. figs. 14, 15.) Diagnosis. Sella pandurate. p* completely external or wanting. Hars more than 20 mm. Width of horse-shoe less than 10 mm. Forearm 52°8-65 mm.?f Details. The ferrwm-equinum type originated from a Bat in all * T am unacquainted with Dobson’s Rh. andamanensis (J. A.S. B. xli. pt. ii. (1872) p. 337). The only specimen known is in the Calcutta Museum. It seems to be a local representative of the affinis type. + The first and second characters, combined, are sufficient to distinguish fermein- equinum from all Oriental species of this group. The others are added to prevent confusion with those Ethiopian species of the present group which also have the sella pandurate and p? external or wanting (clivosus, darlingi, acrotis; augur and deckent). 1905.] OF THE GENUS RHINOLOPHUS. 107 essential points similar to 2A. afinis. It agrees with the now existing affinis in the pandurate sella and the prolongation of Iii’. But it is considerably higher-developed, chiefly in the following respects: (1) the dentition; (2) the wing-structure ; (3) the length of the tail; (4) the beginning, or complete, reduction of the lateral mental grooves; (5) the general size. The peculiar prolongation of the second phalanx of the third finger, described above under 2h. affinis, is preserved in Rh. ferruim- equinum: III.’ is more than (or, extremely rarely, at least equal to) 14 the length of III... Also IV.’ is lengthened, i.e. more than 14 of IV.’; it is an interesting fact that, in this particular point, Rh, ferrum-equinwm (all races) agrees with Lh. affinis himalayanus, but not with any of the other races of affinis. Besides these two characters, which are simply inherited from an affinis-like ancestor, there is an important modification in another part of the wing, to which we have no parallel in any of the foregoing forms*, viz. «@ change in the proportionate length of the third, fourth, and fifth metacarpals, as shown in the subjoined table :— 3rd meta- 4th meta- 5th meta- . Forearm. carpal. carpal. carpal. All the foregoime species (94 examples) ..................... 1000 715 739 740 Rh. fervrum-equinum (all races; 121 examples) ...... 1000 644. 724) 743, This table shows:—(1) In all the foregoing 21 forms of this eroup the fourth metacarpal is but very little longer than the third (24 mm., for a supposed length of forearm of 1000 mm.), and the fifth metacarpal is practically of the same length as the fourth f. (2) In ferrum-equinum «a considerable shortening of the third metacarpal has taken place ; at the same time a much smaller reduction of the fourth metacarpal has occurred, so as to make the fifth metacarpal, slightly but decidedly, the longest of all. The tail is proportionately longer than in the foregoing species, being, on an average, in the eastern races of ferrwm-equinwm (nippon, tragaltus, regulus) exactly 14, in the typical form 13, the length of the lower leg, whereas proximas, in this point (as well as geographically), is intermediate between the eastern and western races 7. In all the foregoing forms, without exception, there are three * But there is an exact parallel in an Ethiopian species, of the affinis type, viz. Rh. darlingi (see the “General Remarks,” below, p. 118). + It would only have made the table more complicated if I had given separate ciphers for all the foregoing species. The only difference (and an exceedingly small one) is that in simplex, megaphyllus, truncatus, nanus, celebensis, borneensis, virgo, and malayanus the fourth metacarpal is, almost always, a mere trifle Jonger than the fifth; in xereis, stheno, rouxi, thomasi, and affinis a mere trifle shorter than the fifth. However small this difference is, it is evidently the first faint trace of the modification definitely carried out in ferrwm-equinum : the fourth metacarpal always shorter than the fifth. £ It is hardly necessary to say that a short tail cannot be a primitive character in the order Chiroptera, taken as a whole. But, for some reason or other, we find in the most primitive species of the genus Rhinolophus a very short tail; in the higher forms of the present group we see, again, a lengthening of the tail. 108 MR. K, ANDERSEN ON BATS [May 16, vertical grooves on the front of the lower lip. In the eastern races of ferrum-equinum (nippon, tragatus, regulus) sometimes exactly the same, but very often the lateral grooves are more or less reduced ; in the western races (proximus, typicus, obscurus) they have, as arule, almost or quite disappeared *. As to the general size, the eastern races are, as it seems, always larger than any form of affinis; proximus and typicus at least on an average so; while obscurus is nearly of the same size as affinis himalayanus. The remaining external characters need only a brief record :-— The supplementary leaflet is slightly more reduced than in affinis, and more closely united to the upper lip; this latter it™is (more than the reduction) which makes it less distinctly visible. The posterior connecting process is more lengthened in antero- posterior direction, also a little more projecting, but quite rounded offat the summit. But, curiously enough, in one specimen (from Transcaspia) I find the process quite as in @finis (in all other specimens from W. Asia it is normal). The lancet has a marked tendency towards assuming a hastate shape, rather than a cuneate, the extreme tip being, generally, long and slender; but sometimes, and both in the eastern and western races (though more often in the former), individuals are found in which the lancet is almost cuneate, as in affinis.—These two individual variations are worth noticing, as, both of them, pointing back to affinis. The ears are somewhat modified: more attenuated below the tip, and more pointed. The plagiopatagium is inserted on the tarsus, on the base of the metatarsus, or about 1 mm. above the ankle-joint. But in one individual (from Cyprus) it is mserted no less than 6 mm. in front of the ankle-joint. It, again, recalls 2h. affinis. Colour. A small series of skins from Tessin, Switzerland, affords some information as to the difference in colour dependent on the age of the individuals; all the specimens are of the same sex, from the same locality, and the same month :— (1) Two full-grown, but younger individuals (females, December); distal epiphyses of metacarpals ossified, but teeth unworn; they are probably about six months old :—Upper side * According to Blanford (J. A. S. B. lvii. pt. 11. no. 3 (1888) p. 263), Rh. tragatus Hodgs., regarded by him asa distinct species, and corresponding to what is here called the eastern races of ferrwm-equinum, has three mental grooves, ferrum-equinum one only. If this were so, I should have no objection to separating Rh. tragatus ‘specifically from ferrum-equinum. But there is, in this as in other respects, a complete intergradation. The details are these:—(1) “Rh. tragatus” (10 spirit- Specimens) : in three individuals (Kashmir, Almora, Darjeeling) the three grooves are perfectly distinct; in three (Masuri, Nepal) the lateral grooves are less distinct than the central one; in two (Nepal) they are so far on the way towards obliteration that it requires close examination to discover them; in the two remaining (Shanghai) they are still more reduced. (2) Rh. ferrum-equinum (s. stv.): rather often traces of the lateral grooves are easily observable; a number of individuals before me, from various places in Europe and W. Asia, have either a slight depression or a short linear groove on either side of the central one; ina specimen from Ttibingen (one instance only, among several) they are at least not moze obliterated than in two “tragatus” trom Nepal and two “ nippon” from Shanghai. 1905. ] OF THE GENUS RHINOLOPHUS. 109 greyish “ drab,” lighter on the head and neck; base of hairs ‘ ecru; drab”; astrongly marked, dark brown horse-shoe patch ; under side almost “ ecru-drab” on throat and breast, very Lght “drab” on belly. (2) One (female, December); teeth almost unworn ; must be very nearly of the same age as (1):—Intermediate in colour between (1) and (3), but nearer to (3). (3) Three aged individuals (females, December); teeth worn ; two of them are at least 14 years old, the third (teeth very camel worn) still older :—Upper side, a shade of brown which might be described as “ mars-brown” with a pronounced tinge of ‘ ‘drab’; base of hairs light “ecru-drab”; scarcely any indication of a horse-shoe patch ; under side light ‘ wood-brown” with a tinge of * ecru-drab.” In a series from the Hautes-Pyrénées (January) I find the same differences in colour, but have not been able to verify the comparative age of the individuals by means of the skulls. Three skins from Minorca (spring) are like the aged Swiss individuals or, if anything, a trifle lighter. The teeth are worn, showing the animals to be, probably, at least about two years old. Skins of aged individuals from England are indistinguishable from Swiss specimens of a like age. A very young (not full- grown) example from Somerset is quite like the younger (greyish- drab) individuals from Switzerland. As a general conclusion: young individuals are, broadly speaking, ‘dark grey, old individuals brown; the Colon of the young animal is retained, at least in some individuals, till December, beyond the time when the epiphyses of the metacarpals have become ossified. For those who have an opportunity to watch these Bats in the caves during the winter, it would be an object of some interest to ascertain how the colour-change is effected, by a moult or by a recolouring of the hairs. Shull. The essential characters as in th. affinis, the general shape hardly different, but as a rule, of course, the skull is larger. The four anterior swellings are slightly more differentiated ; ‘the median ones almost circular in outline, the lateral ones oblong. Chief character: the much longer palatal bridge: very nearly 3 the length of the maxillar tooth-row, a little more or less, but never so short as $ the tooth-row (as in affinis). Dentition. p, external and exceedingly small, or, very often, lost, also in younger individuals. p, and p,in contact. p* completely external, extremely small, not rarely y lost, also in younger individuals. Upper canine and p’ not only in contact, but their cingula, as a rule, considerably overlapping each other (the cingula of p* being external to that of the canine). Measurements. On p. 115. Distribution. From 8. Chinaand Japan, through the Himalayas, the Mediterranean Subregion (exclusive of Egypt), and Central Europe to 8. England. Geographical races. There are, at least, six forms of Rh. ferrum- 110 MR. K. ANDERSEN ON BATS [May 16, equinum, three eastern (iippon, tragatus, regulus), and three western (proximus, the typical form, and obscurus). They are sufficiently differentiated to need technical names, but in no respect—in the external characters, in the skull, in the dentition— is there a sharp “hard-and-fast” line between them :— Tn the extreme east (S. China and Japan) we find a Bat (mippor) of moderate size and with rather small teeth; the dentition, too, has remained on a rather primitive stage of development; but the horse-shoe and nasal swellings are very broad. Some of these peculiarities, viz. the broad horse-shoe and nasal swellings, are preserved in the Central Himalayan tragatws, but the general size of the animal is increased, the skull and teeth very large, the dentition more highly developed. This latte character reaches a climax in the next form, regulus, from the N.W. Himalayas, but at the same time the horse-shoe and nasal swellings are markedly narrower; in this respect regulus evidently shows tendencies towards the western races, as also might be expected from its habitat.—These three Bats constitute what I call the ‘“ eastern” races of ferrwm-equinum. The geographical line separating them from the western races must be drawn somewhere between Masuri and Gilgit, at the border between the Oriental and Palearctic Regions. East of that line the individuals are generally larger, with broader horse-shoe; the lateral mental grooves not rarely fully developed; the tail on an average only 13 the length of the lower leg. Passing from Masuri (still regulus) to Gilgit, on the extreme north-western, ‘ Palearctic” side of the Himalayas, we find a form (proximus) with small and slender skull, narrower horse- shoe and nasal-swellings ; which give it a decidedly ‘“ western ” aspect, and contrast it with its eastern neighbour, vegulus ; but it has retained the somewhat shorter tail characteristic of the eastern races. The typical form has got rid also of this reminiscence, but, as a matter of fact, also in this race now and then, though rarely, individuals occur which ‘fall back” to the shorter-tailed eastern stage. The typical form leads to the generally smaller, extreme south-western race (obscwrus : Spain, Algeria). A closer study of these races, as compared with the Ethiopian Rh. augur and Rh. deckeni, will throw some ‘light on the past history of the ferrwm-equinum type (see the “ General Remarks” on the simplex group, below, p. 118). 14a. RHINOLOPHUS FERRUM-EQUINUM NIPPON Temi. Rhinolophus nippon Temminck, Mon. Mamm. 1. 8° monogr. (1835) p. 30a; Temminck & Schlegel, Fauna Japonica (1842), p. 14, pl. iii. figs. 1, 2; Peters, MB. Akad. Berlin, 1871, p. 312. Rhinolophus ferrwm-equinum (partim) Dobson, Cat. Chir. Brit. Mus. (1878) p. 119. Diagnosis. Size moderate, horse-shoe very broad. Skull smail, but with rather broad nasal swellings; tooth-rows very short. 1905. ] OF THE GENUS RHINOLOPHUS. 111 Details—(1) Compared with tragatus: On an average (as a rule also absolutely) markedly smaller: forearm 57:2-59-3 mm. (tragatus: 59-63); but the horse-shoe is, nevertheless, of the same excessive breadth: 9-9°5 mm. (éragatus: 8°8-9°7), Skull considerably smaller and narrower, but {(in conformance with the broad horse-shoe) with rather broad nasal swellings: comparatively as broad as in tragatus, but, owing to the smaller size of the skull, not absolutely so. Teeth markedly smaller, the tooth-rows shorter. (2) Compared with regulus: Of approximately the same size (or nippon rather smaller), but horse-shoe considerably broader : 9-95 mm. (regulus: 8:2-8°8). Skull generally smaller and narrower, but nasal swellings, nevertheless, quite as broad as in regulus (comparatively, therefore, decidedly broader). Tooth- rows markedly shorter. (3) Compared with the western races: The broad horse-shoe prevents it from being confused with any of the western forms. Colour. AS im adult individuals of Serrum-equinum from Europe*. No quite young specimens examined. Dentition (5 skulls). In two skulls p, is present on both sides; in two (teeth unworn) on one side only; in one (teeth very slightly worn) lost, but the alveoli not quite obliterated. p? is present in all skulls examined. The cingula of the upper canine and p* not only less completely overlap than is generally the case in the other races, but in one skull the two teeth are very slightly, in one quite distinctly, separated. his dentition is decidedly more primitive than in the western neighbours of this race, tragatus and regulus. Distribution. S. China (Shanghai). Pt. Hamilton. Japan. Remarks. 1 find the examples from Shanghai and Pt. Hamilton (S. of Korea) indistinguishable from those from Japan. 146. RArNOLOPHUS FERRUM-EQUINUM TRAGATUS Hodgs. (Plate IVe fie. 14a, 0, c,d.) Rhinolophus tragatus Hodgson, J. A. 8. B. iv. no. 48 (Dec. 1835) p. 699; Peters, MB. Akad. Berlin (1871), p. 312. Rhinolophus ferrum-equinum (partim) Dobson, 1. s. ec. Diagnosis. Size largest, horse-shoe very broad. Skull and tooth-rows: the extreme. Details —({1) Compared with nippon : see this form, supra. 2) Compared with regulus: Onan average larger, with markedly broader horse-shoe (but no sharp line of separation, the maxima * According to Temminck the fur of nippon is “plus long, plus abondamment feutré, plus soyeux et moins lustré” than in ferrwm-equinwm from Europe, and the colours “ différent également.” In the length and abundance of the fur I am unable to find any tangible difference between nippon, tragatus, and ferrum-equinum. As to the colours (two well-preserved skins: Fuji and Nikko), it is quite the same as in darker individuals of ¢ragatus, and this again as im-fully adult individuals of the typical ferrum-equinum ; laid side by side these Bats are indistinguishable in colour, ple MR. K. ANDERSEN ON BATS » [May 16, of regulus-being equal to minima of tragatus). Skull generally larger, and with broader nasal swellings. (3) Compared with the western races: The large size, broad horse-shoe, shorter tail, large skull, broader nasal swellings, and longer tooth-rows prevent it, in most cases, from being confused with any of the western forms. Dentition. In one only, out of six pairs of mandibles, p, is present on both sides; in two (teeth unworn, or very slightly worn) on one side (alveolus disappeared on the other side); in no less than three completely wanting, although the teeth are either quite or almost unworn. A similar high development of the upper teeth (eight skulls): p* present in five; completely wanting, and alveoli disappeared, in three (teeth unworn or slightly worn). Cingula of the upper canine and p* always over- lapping. ‘This is unquestionably a higher stage than in nippon. Distribution. Darjeeling. Nepal. Technical name. Hodgson’s cotypes of Rh. tragatus (three examples; Nepal) are in the British Museum. 14¢. RHINOLOPHUS FERRUM-EQUINUM REGULUS, subsp. n. Rhinolophus ferrum-equinum Hutton, P. Z.8. 1872, p. 698. Diagnosis. Size rather large, but width of horse-shoe moderate only. Skull large and broad, with long tooth-rows, but narrow nasal swellings. Details. Compared with the western races: The large size, combined with the short tail, will, in most cases, make it readily distinguishable. The skull is, almost invariably, larger, the tooth- rows longer. Dentition (4 skulls). In none of the skulls examined could If find any trace of the lower p,, although they all have the teeth unworn. Tn two skulls p* is present, in two completely wanting. Cingula of the upper canine and p* always overlapping. This is the highest stage of dentition in any race of ferrum-equinum (in the present group it is surpassed only by 4. acrotis, but this species is an Ethiopian modification not of the ferrwm-equinwm type, but of the affinis type). Type. 3 ad. (in alcohol). Masuri. Collected and presented by Capt. Hutton. Brit. Mus. no. 79.11.21.153. Distribution. Almora. Masuri. 14d. RHINOLOPHUS FERRUM-EQUINUM PROXIMUS, Subsp. n. (Plate IV. fig. 15.) Diagnosis. Size moderate, horse-shoe very narrow, tail short, Skull small and slender, with very narrow nasal swellings and short tooth-rows. Details.—(1) Compared with the typical form: Although being of the same size as the larger and medium-sized individuals of the typical form, proximus has a very short tail; in so far, it might, very properly, be characterised as a “typical” ferrum-equinum 1905. ] OF THE GENUS RHINOLOPHUS. 113 which has preserved the tail of the eastern races (¢f. also its geographical habitat); the horse-shoe is remarkably narrow. The skull very small and slender; the nasal swellings narrow. (2) Compared with obscurus: Larger, but proportionately with narrower horse-shoe. The skull is even smaller and more slender than in any individual of obscwrus I have seen. (3) Compared with the eastern races: The small size, combined with the very small horse-shoe, distinguishes it sufficiently. The skull is smaller and, especially, more slender, the nasal swellings narrower, than in any of the eastern forms. Dentition (one skull). p, and p” present. Cingula of the upper canine and p‘ overlapping. ‘This dentition is more in accordance with that of the typical ferruwm-equinwm than that of regulus, showing the “western” character of proximus (notwithstanding the short tail), a conclusion borne out by the general external aspect of this Bat, and the size of the skull and the tooth-rows. Type. 9 ad. (in alcohol). Gilgit. Presented by Dr. J. Scully. Brit. Mus. no. 81.3.1.10. 14 ¢. RHINOLOPHUS FERRUM-EQUINUM Schreb., TYPICUS. Le fer-a-cheval Daubenton, Mém. Acad. Rov: Sci. Belg. 1759, pp. 377, 382, pl. 15. fig. 4. Vespertilio Ferrum equinwm (partim) Schreber, Siiugthiere, 1. (1775) pp. 174, 188, pl. 62 (the two upper figures). Vespertilio equinus (partim) P. L. 8S. Miiller, Natursyst., Suppl. (1776) p. 20. oe Ungula (partim) Boddaert, Elenchus animalium, i. GU7S5) ape Gl Vespertilio Ferrum equinum, a. major Gmelin, Linn. Syst. Nat. . (1788) p. 50. Vespertilio Hippocrepis (paxtim) Schrank, Fauna Boica, i. (1798) . 64. Lthinolophus uni-hastatus Geoffroy Saint-Hilaire, Descr. de VEgypte, ii. (1812) p. 132; id., Ann. Mus. d’Hist. Nat. xx, (1813) . 257, pl. 5. Rhinolophus JSerrum-equinum var. germanicus et var. téalicus Koch, Jahrb. Ver. Naturk. Nassau, 1862-63, pp. 522, 523 * Phinolopha s ferrum-equinum (partim) Peter s, MB. Akad. Berlin, 1871, p. 810; Dobson, Cat. Chir. Brit. Mus. (1878) p. 119. Rhinolophus libanoticus, conchifer, et rufescens ‘ Khrbg, et Lichtst. Mspt.” Peters, loc. cit. (1871) (nomina nuda). Diagnosis. Size moderate, horse-shoe rather narrow, tail long. Skull rather small and slender, with narrow nasal swellings and short tooth-rows. ** Koch’s two “varieties” of ferrum-equinum must have been based on too small a material, or there must be some mistake in his statements. That individuals from S. Europe, é. e., Europe 8. of the Alps (his “ var. italicus”’), should, generally speaking, be larger than those from Europe N. of the Alps (his “var. germanicus”’), is at all events not correct. The statement that var. germanicus is “iiber den tticken mehr braungrau oder aschgrau gefarbt,” whereas var. italicus “stets in das Rothliche neigt,”’ raises the suspicion whether Koch has not compared immature individuals from Germany with fully adults from Italy. Proc. Zoou. Soc.—1905, Vou. IT. No. VIII. 8 114 MR. K, ANDERSEN ON BATS [May 16, Details.—(1) Compared with obsewrus : the subjoined particulars will make the difference evident :— 59 specimens of the typical form have been examined from the following localities :—Transcaspia (1); Euphrates Valley (3) ; Syria (2); Galilee (2); Cyprus (2); N. Bulgaria (1); Trans- sylvania (31); Hungary (1); Moravia (2); Dalmatia (2); Turin (1); Genoa (1); Sicily (2); Switzerland (Tessin and Geneva * 7) ; Tiibingen (1). Forearm, in these specimens, on an average 57-5 mm. In no less than 44, 7. e. 75 per cent., the forearm measures 5/7 mum. or more (up to 60°3 mm.) ; in the remaining (and quite independent of the locality) Jess than 57 mm. (down to 53-5 mm.). Of obscurus 31 specimens have been examined from :—Troubate, Hautes- Pyrénées (8); Cintra, Portugal (1); Madrid (3); Valencia (12); Minorca (5); Algeria (2). Forearm, in these specimens, on an average 55:5 mm. In no less than 25, 7. e. 81 per cent., the forearm measures less than 57 mm. (down to 52°8 mm.); in the remaining between 57 and 58mm. Although the series is smaller than that of the typical form, the facts here pointed out cannot be due to mere chance ; the contrast is too well marked. As a conclusion: in the typical form the forearm measures generally 57 mm. or more; in obscwrus almost always less than 57 mm.; maximum of obscurus is but a trifle larger than the average size of the typical form. (2) Compared with the eastern races: the proportionately longer tail prevents, in almost all cases, its confusion with any of these races. The skull is rather easily discriminated from that of tragatus and regulus (cf. measurements, p. 115), but I fail to find any point by which to distinguish it from the Japanese nippon. British specimens. 13 specimens have been examined. Yorearm on an average 55°4 mm., i.e., British specimens of ferrum-equinum are on an average of the same size as the extreme south-western (Spanish) race, Rh. f. obscurust. Of the 13 specimens, 2 only have the forearm 57 mm. long or more (up to 58 mm., quite as in obscurus); all the others between 53°8 and 56:2 mm. These indi- cations require, of course, verification by a much larger series §. Dentition (11 skulls). In seven skulls p, is present on both sides (teeth in very different stages of wear); in one, on one side only (teeth worn); in three (teeth almost unworn, or much worn) completely wanting (no alveoli). p* is present in all the skulls examined, two of which are of very aged individuals. Cingula of the upper canine and p' generally more or less overlapping, but in two skulls separated by an extremely small interspace, This dentition is almost exactly as in xippon. * For the loan of some Bats from the neighbourhood of Geneva I am indebted to M. Ch. Mottaz. + A very elaborate table of measurements of fourteen Spanish specimens was kindly sent to me by Prof. A. Cabrera Latorre, Madrid. These are the only examples, dealt with in this paper, not examined by myself. + Compare with this Rh. hipposiderus minutus, below, p. 142. § To keep the typical form uninfluenced by the smaller British individuals, I exclude these latter from the table of measurements on p. 115. OF THE GENUS RHINOLOPHUS. 115 1905.] 9-6 6-6 18 9.8 6ST &-ST ¢.9 6-9 gg g GGL &.-IT 6-6 6 9-01 OL GEG -& 86 &-16 &@ LE LG AT &8T GPL GET Gh SBE 06 9T L11 OL S-1V LE GE GL 61 ST 8& TE 8G 8.69 £.8 LL GST 4-81 ST &1 GG 81 “UU “TUT XV ULTL SIMS P ‘susuttoods Te *SNANISGO OL 66 6 ¢.8 LL 9ST. 1.9 6-9 9 8-P L600 1-61 6.6 ¢.6 It &.OT 8-86 4.66 G&L Il G16 8-6 Iv VE AT &-&T CT $61 LUV ¥-68 606 LT $<6L €.0L L-8V 8E GCE GLE 6:06 81 60F 8-7E €.09 ¢.€¢ 8:8 8. 8:91 FT 8-91 ST 8.66 1G “UU “UU ‘XR | ULTT “S]DYS TT ‘suoutloeds @g ‘snardhg 2.6 8:8 9 6. G11 8-8 G6 GIL 01 6-6 6.96 ve g.1& ZT 8.ST LEt L&T Ty 8.07 $-6T 9.81 L10 4.01 $07 8-66 $66 $.86 0G I-61 LE G96 8g 8.99 8 LL S-7L ia GST ST 6-16 8-06 “UU = “HU “XBI ULIAL TMS T ‘suoutoeds Z “snmixord L01 6.0L 1.6 6-6 GAT 4.91 49 19 9 1g &I L-6L GOL 4.6 GIT 4.01 L¥G L.&% GPL 8-61 8-96 9.66 ge. 18 AT 8.91 401 GET Fi i LCL 6.06 4-81 GEL 6.11 SEP 8.11 GE G0 16 &.6T G8E ELE 09 «4.89 8:8 6.8 Z-S1 ST TL4L 6ST 8.86 GG “WUL = “WU Xe “ULIN LCE ‘suouloeds ip “sinha Z0L 6.01 6-6 6-6 8I 8.47 GL 1.9 LG 6G G&L 8-61 IL 6.01 g.1T IL L6 6.VG €-71 €1 9-46 9-96 LE 8-VE 61 81 8-ST ial 6Lbp 8.87 GB G06 8-6 9-IT LSP GV SVE G& 6 9.06 600 GLE €9 6g 4.6 8:8 8-41 G.9T 8-81 8-91 $76 8.16 “UUL = UU Xe UTI “ST[NAS 8 ‘suauttoeds g ‘snp hpi, 8-6 6-6 6 €-8 8-91 8.ST 8-9 €.9 9 6.G GGL 4.11 OL 9.6 6.0L GOT L:86 6-66 GPL L.61 €96 G.PG 8.9€ Té GLI 4.ST SVL 4-81 €Sp 1-87 S.6L G8T GL IL TY §8-1P GE 1G 16 §-6T 8.88 GLE 6.69 G.Lg ¢.6 6 8ST 61 ZL GST 8-66 P.66 “UU = “TU “XBYL ULTAL “SOAS G ‘sueuttoods y ‘uoddvu "4904 TeMorT DOD Ob OC COMOGUOO UOC ORCLOOO GOD.CO GCCOOo = Gye jedd page sie BaUIT Te Ma [eseu jo WPL eee ees “Yysue] peyiqiovidns =“ rr TIPIA OL{BUIOSAZ ce Myce S OSvo-UIBIG JO YypIm we aes proyseur yg suey Te904 TRNAS — q00KF Foes Caen een cdnbA taddeecenodanioenacnoss TL “oA tA “qedworjace qe Soere oe ee .. SAI _ledivoejout WP s III vIil _ [edavovjour pre TWIG WT “s0Us-9s10y 2@ YI pvaiq & : Y}SUET [2404 Boneeye O80 NT doo sotsodo0bnDe opp O0O _Wapread qsoqvois 66 p\etabafatalstelatw(ulelsyeratsyalelsy>tsier " YP. UOT ‘SLB ‘sawoadsqns pun wmutnbe-un«zr1e; snydojouryyy fo spvawaimsnary 116 MR. K, ANDERSEN ON BATS [May 16, Distribution. From Transcaspia and the Euphrates Valley through Southern and Central Europe, exclusive of the Spanish Peninsula. 14/. RHINOLOPHUS FERRUM-EQUINUM OBSCURUS Cabrera. Rhinolophus ferrum-equinum obscurus Cabrera Latorre, Mem. Soe. Espafi. Hist. Nat. ii. (1904) p. 257. Diagnosis. Smaller than the typical form. Detwils.—(1) Compared with the typical form: see above, p. 114. (2) Compared with the Eastern races : the small size, combined with the narrow horse-shoe, make it readily distinguishable. The skull is apparently slightly smaller than in nppon. Dentition (4 skulls). As in the typical form. Distribution. Spanish Peninsula, with the Balearic Islands. Algeria *. General Remarks on the Rhinolophus simplex Group. The place of origin.—Of all the existing forms, the Australian Rh. megaphyllus is one of the most primitive in dentition. But it is very unlikely that the Australian Continent has been the place of origin of the group. Rh. megaphyllus is the only Australian species of the whole genus; this might suggest the assumption that it is an immigrant into the country, vather than an ancient inhabitant: secondly, Australia is the extreme eastern border for the group (as well as for the genus), no species being known from the islands to the east of the Continent; it would probably not be so, if Australia had been a centre of dispersal for the group: thirdly, megaphyllus has at least two characters which certainly are not primitive—the large nose-leaves, and (probably as a consequence of that) the rather broad nasal swellings: fourthly, megaphyllus looks extremely like an enlarged, continental representative of the Lombok species, Rh. simplex (just as Rh. rouwi is the larger, continental repre- sentative of Ah. borneensis). These arguments seem to support the conjecture that, not the Australian Continent, but the “ Indo- Australian Transitional Tract,” now broken up into numerous larger and smaller islands, and still inhabited by such. very primi- tive forms as simplex, truncatus, nanus, celebensis, and borneensis, has been the centre from which the group spread eastwards and westwards. Differentiation t.—The ancestral species seems to have divided into two branches, an eastern and a western. In the eastern, more primitive branch the sagittal crest does not reach quite so far forwards as a point corresponding to the middle of the orbit ; in the western the temporal fossa is comparatively a little wider, and the sagittal crest produced forwards more or less beyond that * The type of Rh. f. cbscwrus, in the Madrid Museum, is from Valencia, Spain. As will be seen, I take the name:in a wider sense. Valencia specimens were separated by Prof. Cabrera, as a distinct subspecies, mainly on account of a difference in the ratio between the length and breadth of the horse-shoe. In a large series of ferrum-equinum from Europe and W. Asia there is, however, no small, and quite ndividual, variation in this respect. + Compare the diagram on p. 120. 1905.] OF THE GENUS RHINOLOPHUS. ITE point. The geographical line separating the two branches coin- cides with the line separating the ‘‘ Austro-Malayan” from the “Tndo-Malayan” subregion (Celebes being a part of the latter). The eastern branch is, as yet, represented by four known species Rh. simplex, megaphyllus, truncatus, and nanus. The western by all the others. The further evolution, from borneensis to ferrum-equinum, has been discussed above, and is summed up, in the briefest possible form, in the subjoined diagram (p. 120). But the sketch of this group would be deprived of some of its most instructive features if the Ethiopian species were left quite out of consideration. They belong to three closely related types :— (1) Ethiopian species of the borneensis-stheno-rouxi type.— Far south in Africa, in Bechuanaland and Mashonaland, we find two small species, Rh. denti and simulator, described quite recently *. They are the Ethiopian representatives of the borneen- sis type: the same general shape of the skull; essentially the same dentition; the same parallel-margined sella, with a faint or almost imperceptible constriction at the middle; the same style of connecting process; the same proportionate length of the fourth and fifth metacarpals ; even the same length of the tail, &c. But there are, in these species, three characters of especial in- terest, because they enable us to determine still more precisely their phylogenetic place: the nasal swellings (side view) are more projecting than in borneensis, but less than in stheno; III. is lengthened, and IV.’ somewhat shortened, as in this species,— proving that they have originated from a Bat which had already traversed a part of the distance separating borneensis and stheno. The dentition is on a slightly higher level than in borneensis and stheno, the only difference being that p”, although still in the tooth-row (as in the Oriental species), shows a distinct. tendency fowards the external side.: In the extreme south of Africa (Cape Colony) we find a species, Rh. capensis, which, quite superficially, looks like an enlarged Rh. simulator. It is an African representative of Rh. rowxi: the skull is to such a degree that of roux? that it would be hard to find any tangible difference, even the measurements being practically the same (on an average smaller than in rowwi); the nose-leaves (sella, process, lancet) are the same; proportionate length of fourth and fifth metacarpals, of tail and tibia, the same. But the dentition is somewhat more advanced: p is generally ex- ternal, but still, very often, a quite distinct interspace between the canine and p‘ indicates its former place; IIT.° is somewhat lengthened. In short: Rh. capensis isa “Rh. rowxt ” which in the wing-structure has taken a course towards, in the dentition very slightly beyond, the affinis-stage. (2) Ethiopian species of the aftinis-type.—On the coasts of the Red Sea we find a species, Rh. clivosus, first made known by Cretzschmar from Mohila in Arabia; I have seen examples from * Thomas, Ann. & Mag. Nat. Hist. (7) xiii. (1904) p. 386; Andersen, op. cit. (7) xiv. (1904) p. 384. 118 MR. K, ANDERSEN ON BATS [May 16, the African coast of the Gulf of Aden. Jé is the closest existing relative of the Himalayan Rh. affinis: the same shape of the skull; the same shape of the sella, of the connecting process, of the ears; the same structure of the wings (also the same lengthening of III.*); the same proportionate length of the tail. But it is more advanced in dentition : p, is not only external (as in affinis), but very often lost; p*, which in affinis is still in the tooth-row, is in clivosws external and very small. In short: Rh. clivosus is a“ Rh. affinis” with ferrwm-equinum dentition. The clivosus type has found its way very far into the Ethio- pian Region. kh. darlingi*, from Mazoe to Angola, is a modification of this type (as proved by the skull), differing from clivosus in the more pronouncedly pandurate sella, the much broader horse-shoe, the much smaller ears, and, by far the most interesting, i the shortening of the third metacarpal. This last peculiarity is the same as that pointed out above, under th. ferrum equinum: in the wing-structure Rh. darlingi differs from “h. clivosus quite in the same way as Lh. ferrwm-equinum from Kh. afinis. It is a suggestive fact to find this peculiarity so exactly copied by the South-African species. Kh. acrotist, from Egypt and Erythrea, is, externally, very similar to Ph. clivosus; also the wing-structure is the same. But the tendency, in elivosus, towards an obliteration of p, and p* has been further developed by acrotis: it has completely lost both of these teeth, thus being, in this particular respect, the highest member of the whole group. h. acrotis is a “ Rh. affinis” with a dentition still more advanced than in ferrwin-equinwm regulus. (3) Hthiopian species of the ferrum-equinum type.—Rh. augur £ is widely distributed, in several geographical races, over the southern part of the Ethiopian Region: the Orange River tract, Natal, the Lower Zambesi. Jt is the closest existing relative of Eh. ferrum-equinum ; the skull, the nose-leaves, the wing-structure are the same; but the dentition is a trifle less advanced, and the ears are smaller. We find the ferrwm-equinum type also further northwards in Tropical Africa (Mombasa): Rh. deckent; the skull and dentition, and all external characters of any importance, are as in augur; but the horse-shoe is broader. The area occupied by these two Ethiopian representatives of the ferrum-equinum type extends, broadly speaking, from the Orange River to Mombasa. It is completely cut off from any other region inhabited by that type of Bat; it forms a large enclave bordered to the north and west by vast tracts where no representative of ferrwm-equinum occurs; we must go so far away from South and Equatorial Africa as the Euphrates Valley, Syria, and Algeria before meeting with the closest relatives of those Ethiopian species. Thus the question suggests itself, by which way the ferrwm-equinum type reached Tropical Africa, and why its range there is now so peculiarly insulate. When * Andersen, Ann. & Mag. Nat. Hist. (7) xv. (1905) p. 70. ~ + Andersen, op. cit. (7) xiv. (1904) p. 454; (7) xv. (1905) p. 73. {£ Andersen, op. cit. (7) xiv. (1904) p. 380. _ 1905. ] OF THE GENUS RHINOLOPHUS. 119 trying to answer this question, the following facts must be borne in mind :—Firstly, that all paleontological evidence is wanting, which detracts from what we know about the affinities and distribution of the now existing representatives of these Bats. Secondly, that the ferrwm-equinwm type is unknown in Egypt, as well as in the whole region of the continent north of British East Africa, and that we have no reason, of any kind, to believe that it ever existed there. Thirdly, that we have to account not only for the distribution of Rh. augur and deckeni as compared with the other members of the same section of the genus, but also for the presence in Tropical Africa of representatives of the borneensis and rouxi types, and, be it noticed, representatives which, without exception, are more highly differentiated than their Oriental allies. These facts, so far as they go, seem to allow of no other satisfactory explanation than this: the im- migration of these Bats, as of so many other Oriental types in the Kthiopian fauna, has taken place by way of the broad tract of land which, as commonly supposed, in a geologically late period connected Southern Asia with the African continent. In the case of the ferrum-equinum type this explanation would make it evident, why it, though vastly distributed in South and Equatorial Africa, is absent from the whole north of the con- tinent with the exception of the extreme north-western (Medi- terranean) coast-region, which it, no doubt, has reached from South-western Europe, since the Algerian race is subspecitically indistinguishable from the Spanish form (A. f. obscurus). In the case of the borneensis and roux types it would account for the fact that they are common to the Oriental and Ethiopian Regions, but absent from the whole of the Palearctic Region. And it would also account for the presence of the genus Rhino- lophus in the Ethiopian Region, for, as I shall have to show later on in this paper, all the Ethiopian representatives of the genus are undoubtedly of Oriental origin. Such being the case, | am able to draw up the following rough sketch of the history of Rh. augur, deckeni, and their Oriental and Palearctic relatives :— The ferrum-equinwm type has originated somewhere in South Asia; we find there the long series of more primitive forms which lead up to that type, whereas in the whole of the Ethiopian Region there is not any species with which it can be brought in genetic connection. The ancestral ‘“ferruwm-equinwm” broke up into three branches: a south-western, a western, and an eastern. The south-western branch, which had spread directly from South Asia into the Ethiopian Region, was cut off from the main stem by the submergence of the connecting tract of land, and is now differentiated into two species—the southern Rh. augur and the northern Rh. deckent. Both of them have retained at least two “ancient” characters: a slightly more primitive dentition (the upper canine and p* often more or less separated; p* sometimes half in row*) and a short tail. To the external difference * 35 skulls of Rh. augur (all races) have been examined :—In 17 the upper canine and pt are more or less separated, in 7 in contact, in 11 more or less overlapping 120 MR. K, ANDERSEN ON BATS [May 16, between these two Ethiopian species, viz. a broad horse-shoe in deckent and a narrow one in augur, we have a parallel in ferru- eguinum: a broad horse-shoe in nippon and tragatus, a narvew one in the other races. The western branch spread over South and Central Europe: the dentition slightly more advanced, the tail lengthened. The third branch is now represented by what I have called the Eastern races of ferrwm-equinum; all of them have retained the short tail; mippon (which, so far as the dentition is concerned, has remained on a relatively less advanced stage) leads through fragatus to regulus, in which the dentition has reached the highest stage of development found in any race of ferrum-equinum. According to this the mutual affinities of the species of the simplex group might be expressed as follows? (the Ethiopian species are marked with an asterisk) :— *deckeni. — SUUgur. SS \ eee nen. *acrotis. *darlingt. / / xv. = / *clivosus. thomasi. “capensis. wy stheno SS al | , *denti. LOUKT ve / | nereis. *simulator. malayanus. | | | virgo. ~borneensis. celebensis. NENUS. | | | truncatus. | | megaphyllus. Y SS eae ol | | O (Jepidus-group_)< — each other at base; in 4 pis half in row. To this latter I find no parallel in any specimen of ferrwm-equinum (all races) I have seen, and in 4 skulls only, out of 33, there is a more or less distinct remnant of the interspace between the canine and p‘. Of Rh. deckeni I have seen one skull only; the dentition is as in many specimens of Rh. augur: c and p! separated, p2 external. t I give the diagram the form of a genealogical tree, only because it is convenient to 1905. ] OF THE GENUS RHINOLOPHUS. 121 Il. Tat RA#NOLOPAUS LEPIDUS GROUP. Diagnosis. Basioccipital, between cochleex, not unusually narrowed, Posterior connecting process projecting and pointed. I include in this group :—(1) All the forms with projecting connecting process comprised by Dobson under the technical name “Kh. minor” ; their close relationship is unquestionable; their differences will be pointed out below ; (2) Rh. acuminatus and its allies, which are scarcely more than giant forms of the lepidus- type; (3) the Rh. blasti and (4) Rh. ewryale sections, peculiarly modified Ethiopian and W. Palearctic representatives of the subbadius-type. ‘The two former sections only will be reviewed below ; the two latter will be briefly mentioned in the “ General Remarks” on the group (p. 135). Text-fig. 22. c d Side views of nose-leaves, showing the principal forms of the connecting process in the Rh. simplex group (a) and the Rh. lepidus group (8, ¢, d). a. Rh. borneensis typicus; b. Rh. cornutus pumilus ; c. Rh. monoceros ; d. Rh. empusa. As this is a first attempt to disentangle the many different forms hitherto confounded with MHorsfield’s Rh. minor, the following preliminary remarks are necessary, as a general guidance :— The first of the above-named sections (the “‘ lepidus-section ”), viz., all the small Oriental and EK. Palearctic Rhinolophi which have the connecting process projecting and pointed, fall into three show, at a glance, the probable interrelations of the species. As sufficiently emphasised in the foregoing pages, I am far from being of opinion that ferrwm-equinum is derived from the now-existing affinis (or capensis from rouxi, or stheno from borneensis, &c.). But ferrum-equinum has originated from a Bat which had the more essentiui characters of affinis (besides several others, unknown to us). The technical names in the diagram are, in other words, to be taken, not in their strict specific sense, but as names of the sections (“ types,” “ branches”) of which the species, as we now see them, are the surviving representatives. 122 MR. K. ANDERSEN ON BATS [May 16, natural groups (sub-sections): the lepidus-type, the minor-type, and the subbadius-type. , I propose to characterise these types at once. It will enable me to confine the diagnoses of the various species to the points in which they differ from the subjoined general characteristic. (i) The lepidus-type—Chief characters : skull larger, width of braim-case about 7-7-7-8 mm.; connecting process (in side view) projecting as a small, erect triangle (not curved forwards as a sharply pointed “ horn ”). Description, based on Rh. lepidus (Wynaad, Mysore, Indian Peninsula).—Supplementary leaflet as in simplex and its allies. Horse-shoe not completely covering the upper lip; a small tooth- like projection on either side of the median notch; front border sometimes, not always, slightly crenulate (individual variation). Sella decidedly broader at base than at summit, slightly, but quite distinctly, constricted at middle, narrow at summit: there is a tendency towards producing an almost subacute summit to the sella (compare with this the borneensis-type : sella broadly rounded off, or even truncated, at summit); height of sella 3-2 mm.; width at base, at constriction, and at summit: 2, 1:8, and 1:2 mm. Connecting process projecting as an acute, sometimes only sub- acute, triangle beyond the summit of the sella. Lancet strongly hastate, about 3 mm. long. Three mental grooves. ; Ears much as in the celebensis-borneensis type, but somewhat more blunt-tipped. . Wing-structure quite primitive, 7. e. no lengthening of III.’, this phalanx being always less, and very often much less, than 14 the length of III.'; no shortening of the third metacarpal ; fourth metacarpal slightly the longest (individually it may fall short of the fifth by a fraction of a millimetre). This wing-structure is perfectly like that of Rh. simplex and its allies. Tail slightly longer than (individually equal to, or a trifle shorter than) the lower leg. Plagiopatagium inserted on the ankle, slightly above or below. Skull. General shape: the simplea-borneensis type, but consider- ably smaller, with smaller teeth, and shorter tooth-rows. The orbital cavities (the confluent orbital and temporal fosse) are shorter and narrower than in borneensis, the zygomatic arches, therefore, less projecting laterally, making the zygomatic width of the skull, as a rule, only equal to, or even a trifle smaller than, the mastoid width. These peculiarities combined make, as a rule, the skulls of the species of the lepidws-type rather easily distin- guishable from those of the borneensis-type-—Arrangement of the nasal swellings, essentially, as in borneensis. Palatal bridge, on an average, somewhat less than 4, but more than + the length of the maxillar tooth-row. Denxtition. Position of p, (in, or external to, the tooth-row) “ vacillating.” _p? invariably in the tooth-row. ‘This dentition is precisely as in simplex-borneensis. Species. Rh. lepidus, monticola, refulgens. 1905. ] OF THE GENUS RHINOLOPHUS. 123 (2) The minor-type—Chief characters: skull, also propor- tionately, very small; width of brain-case about 6°8-7°2 mm. ; connecting process of the lepidus-type (text-fig. 22, b, p. 121). Deser iption, based on 2h, cornutus pumilus (Loo- choo Islands).— Nose-leaves as in the lepidus-type, but: sella narrower; height about 2°8mm.; width at base, at constriction, and at Semen - 1:7,1°5, and 11mm. Connecting process slightly higher, slightly more acute, but of the same general shape. The other external characters as in the lepidus-type. Skull. Considerably smaller; nasal swellings narrower. Teeth smaller, Dentition. As in lepidus. Species. Rh. minor, cornutus, “ minutus” (Miller, nec Montagu), gracilis. (3) The subbadius-type.—Chief character: connecting process long, slender, very sharply pointed, curved forwards, projecting like a small, curved “ horn” (text-fig. 22, ¢, p. 121). Nose-leaves, and other external characters, much as in nwn07, but connecting process as described above; lancet more or less approaching the shape of an equilateral triangle; length of sella about 2°4 mm.; width at base, at constriction, and at summit: 1:7, 1:3, and 0:9 mm. Skull. To judge from fragments, and the skull of a quite young individual, much of the minor-type. Dentition. As in lepidus and minor. Species. Rh. subbadius, monoceros. 15. RHINOLOPHUS LEPIDUS Blyth. Rhinolophus lepidus Blyth, J. A. S. B. xiii. pt. i. (June 1844) p. 486. Rhinolophus minor (partim, nec Horsf.) Dobson, Cat. Chir. Brit. Mus. (1878) p. 114. Diagnosis. Skulland external characters: lepidus- type. Larger: forearm 41°8-42 mm. Details. This species difters from Rh. monticola in its broader nasal swellings, larger size, and considerably longer metacarpals. Colour. Ad., skin: Ganges Valley ; teeth almost unworn; two ¢ ad., in alcohol: Wynaad ; teeth unworn. General colour above between “ wood-brown” and “ cinnamon,” lighter on the anterior part of the back; base of hairs very light “ ecru-drab”; under side “‘ wood-brown ” or tending to “ ecru- drab. a Dentition (three skulls). p, external. p, and p, separated, or almost or quite in contact. p* in the tooth-row, with a well- developed cusp, pointing inwards. Measurements. On p. 125. Distribution. Indian Peninsula: Wynaad (Mysore); Ganges Valley. Technical name. I identify this Bat with Blyth’s Rh. lepidus (to which I find no reference in Dobson’s ‘ Catalogue’), for the following reasons :—(1) lepidus belongs to this group of the genus, 124 MR. K. ANDERSEN ON BATS | May 16, as proved by Blyth’s description of the connecting process, “ still more developed [than in his Rh. subbadius| and obtusely angulated behind m3 the words “still more developed” mean, evidently, ‘““bigeer,” not extremely slender as in suwbbadius. (2) The types were “ probably obtained in the vicinity of Calcutta”; one of the specimens in the British Museum is from the Ganges Valley, therefore in all probability from the very same locality as the types. (3) The colour, as described by Blyth, agrees very well with that of the specimens before me. (4) The forearm was stated to be “13 inches” (41°5 mm.); the longest finger “21 inches” (57°2 mm.); the tibia “above 3 inch” (above 16 mm.) ; these measurements are as in the British Museum examples: forearm 41:8-42 mm.; third finger 58°3-59'1 mm.; lower leg 16-17 mm. These facts leave no room for doubt as to the identification of 2h. lepidus. 16. RHINOLOPHUS MONTICOLA, sp. n. Rhinolophus petersi (errore*) Hutton, P. Z.8. 1872, p. 700. LRhinolophus minor (partim, nec Horsf.) Dobson, wt supra. Lhinolophus subbadius (non Hodgs., nec Blyth) Scully, J. A.S. B. lvi. pt. i. (1887) p. 244. Diagnosis. Skull and external characters: lepidus-type. Smaller: forearm about 37°5 mm. Details. This species differs from /h. lepidus in its narrower nasal swellings, somewhat smaller size, and considerably shorter metacarpals. The horse-shoe seems to be narrower. Colour. Unknown (faded in alcohol). Skull. As in Rh. lepidus, but somewhat smaller, and with narrower nasal swellings. Dentition (two skulls, one belonging to a quite young individual). p, im row (skull of an adult), or external (young). p, and p, well separated, or almost in contact. p* in row; a distinct cusp, pointing inwards. Measurements. On p. 125. Type. & ad. (inalcohol). Masuri. Collected and presented by Capt. Hutton. Brit. Mus. no. 79.11.21.151. 17. RHINOLOPHUS REFULGENS, sp. n. (Plate IV. fig. 16a, 5, c.) Diagnosis. Skull and external characters, essentially of the lepidus-type. But brain-case somewhat higher in front, making the anterior slope of the sagittal crest, towards the postnasal depression, somewhat more abrupt. Forearm 40°6-41:5 mm. Details. Very nearly of the same size as 7h. lepidus, but meta- carpals, also proportionately, somewhat shorter; tibia shorter. The horse-shoe is, if anything, shghtly broader. * There is no doubt that this is anaccidental error. Prof. Peters (who determined Hutton’s Bats) cannot, possibly, have identified the specimen here under considera- tion (forearm 37°5 mm.) with “Rh. petersi” (forearm of type 51 mm.). As already pointed out above (p. 97, footnote), the labels must have been confused; the name “ Rh. petersi” was, probably, intended for Hutton’s examples of Rh. noua’. 1905. ] OF THE GENUS RHINOLOPHUS. 125 Colour. Q ad., skin; Perak; March; teeth almost unworn. Very different from Rh. lepidws. General effect of the colour of the upper side: a dark shade of “‘ Prout’s brown” with a tinge of ‘“‘hair-brown.” On closer examination the fur of the upper side proves to be composed of two kinds of hair : longer, thinner, straight hairs, quite black; and somewhat shorter, crinkled hairs of a ‘“‘hair-brown” colour; the mixture of the colours of these two kinds of hair produces the general effect. Base of hairs of wpper side not lighter coloured. ‘The fur of the upper side has a silvery reflection (iridescence). Under side between ‘“ broccoli-brown ” and ‘“hair-brown.” _ width of nasal swellings......... 4 42, Ba Ay. 4. AD Mandiblewcntcs ice ener, Ot ame Ose mee OMe 105 112 Wippentechlin 24s eit ee ke See ee 5:9 6 57 57 6 63 Wowersieebliges GAGs ae Ae rere 6:2 63 6 61 6:2 6°8 1905.] OF THE GENUS RHINOLOPHUS. 129 (three); in none, completely in contact. p® in row; a well- developed cusp, pointing inwards. Upper canine and p' widely separated ; in one skull there is a small interspace between p° and p’ (the former place of p’). Distribution. Japan proper. Remarks. In general size, as well as in the skull and dentition, the T'su-sima Bat agrees with the typical form; but the colour is that of Rh. c. pumilus*. 20. RHINOLOPHUS GRACILIS, sp.n. (Plate IV. fig. 18 a, }, c.) Rhinolophus minor (partim, nec Horsf.) Dobson, ué supra. Diagnosis. Skull: the minor-type. Sella parallel-margined ; tail extremely short. Very small: forearm 36°2 mm. Details. This is an aberrant species of the minor-type. The connecting process is quite of the same shape as in the foregoing species (very different from that of subbadius). But the sella is parallel-margined, as broad at the summit as at the base; by means of a lens (probably not without) an exceedingly faint trace of a constriction can be observed; the summit of the sella is broadly rounded off, as in borneensis, not witha tendency towards a subacute shape, as in the foregoing forms of this group; length of sella 2°8 mm.; width at base 1:8 mm., at summit 1-7 mm. The lancet is, considering the small size of the Bat, remarkably fo) long (4mm.), with the lateral margins almost straightly converging towards the tip; it recalls the lancet of Rh. midas and hippo- siderus (with which species &h. gracilis has no very close affinity). The tail is extremely short (13°5 mm.), shorter than the lower leg. Plagiopatagium inserted a trifle above the ankle. The colour (a little faded in alcohol) has probably been rather like that of 2h. lepidus. Skull. Quite of the minor-type. Dentition (one skull). p, external. p,and p, distinctly separated. p in row; cusp extremely minute (unworn). Measurements. On p. 132. Type. 9 ad. (in alcohol). Malabar Coast. Purchased. Brit. Mus. no. 73.4.16.2. 21. RHINOLOPHUS SUBBADIUS Blyth. Rhinolophus subbadius Blyth, J. A. 8. B. xi. pt. i. no. 150 (June 1844) p. 486. Rhinolophus garoénsis Dobson, J. A. 8. B. xli. pt. i. no, 4 (Dec. 22, 1872) p. 337; id., Mon. Asiat. Chir. (1876) p. 48, text- figs. a-c; id., Cat. Chir. Brit. Mus. (1878) p. 115. * T have examined a paratype of Gerrit S. Miller’s Rh. minutus (Proc. Wash. Acad. Sci. 1900, p. 235), the type of which is from the Anambas Islands. It is an offshoot of the minor-type, but undoubtedly a distinct species, differme from Rh. minor (trom Darjeeling) in having the brain-case decidedly higher in front, giving the skull, in side view, a very characteristic outline. The name “minutus” is, however, preoccupied by Montagu’s “ Vespertilio minutus,’ which is the British form of Rh. hipposiderus. Mr. Miller will rename the Anambas species. Proc, Zoou, Soc,—1905, Vou, I. No, LX, 9 130 MR. K. ANDERSEN ON BATS | May 16, Diagnosis. Subbadius-type (cf. p. 123). The smallest species in the genus: forearm 34°2 mm. Details. The very characteristic shape of the connecting process, formed as a long, sharply pointed, slightly curved ‘“ horn,” pre- vents the confusion of this (and the next-following) species with any of the foregoing forms. Also the shape of the lancet is peculiar : : short, ine bank almost as an equilateral triangle ; but I doubt that this character, in a large series, will prove to be quite as safea guide for the discrimination of the species as the shape of the connecting process ; there is, in all species of Rhinolophus, a little more individual variation in the lancet than in other parts of the nose-leaves. The sella is, essentially, of the minor-type (not as in gracilis), much broader at base than at summit; below the constriction the margins are almost parallel, above the con- striction slightly converging ; the summit somewhat more subacute * than in any of the foregoing species; tip of sella bent forwards. Plagiopatagium inserted a trifle above the ankle. The colour (a little faded) is probably not very different from that of Rh. lepidus. Skull. Unknown. I have seena small fragment only; it seems to be of the minor-type. Dentition (one example). p, external. p, and p, in contact. p* in row; cusp small, but distinct. Measurements. On p. 132. Distribution. Nepal (type locality). Garo Hillst. (The only example of this species in the British Museum is without exact indication of locality.) Technical name. Hodgson’s “ Vespertilio subbadia” (J. A. 8. B. x. pt. 11. (Nov. 1841) p. 908), from the ‘Central Region of the Himalayas,” is a nomen nudum (no word of description). The head of this Bat is figured in his unpublished drawings (pl. 8. fig. 3); it is not a Rhinolophus, but a Hipposiderus, probably H. bicolor or an allied form. * T emphasise this peculiarity (and, on the whole, enter intoa detailed description of the sella), because it is this “pattern” of sella which has been carried to an extreme insome of the Ethiopian and W. Palearctic representatives of the swbbadius- type (Rh. empusa and blasii; cf. the “General Remarks,” pp. 136-37). + In Dobson’s ‘ Monograph’ and ‘ Catalogue’ (1. s. c.) Rh. garoénsis (= subbadius) is recorded from Masuri. The species is very likely to occur there, only it must be said that till now there is no proof. Its alleged occurrence in Masuri can be traced back to two examples in the British Museum (Capt. Hutton) identified by Dobson with Rh. garoénsis. They are, however, Rh. monticola, differing in all im- portant points (process, lancet, size) from his own original description of garoénsis. Quite as in the case of Rh. petersi: as Dobson had no longer access to the type, he lost the precise idea of it. Still later (Rep. Brit. Assoc. 1880, pp. 175-76) he gave up the separation of Rh. garcénsis as a distinct species, and then we arrive at the stage when all small Indian and H. Palearctic Rhinolophi with a projecting process were called Rh. minor, irrespective of differences in the skull, the process, the sella, lancet, general size, and geographical habitat. What led Dobson to this conclusion was the fact that the position of the lower pz varies in individuals from the same locality (which, however, also is the case in all the more primitive species of the simplex group, as high up in the series as Rh. affinis), and he was quite right in arguing that, from an exclusively taxonomic point of view, this character had no value; but he overlooked the other and more important characters by which the members of his composite species differ from each other, 1905. ] OF THE GENUS RHINOLOPHUS. 131 Blyth’s kh. swbbadius (1844) from Nepal, erroneously believed by himself to be the same as Hodgson’s V. subbadia, is a genuine Rhinolophus. The following analysis of the original description will make it evident that it is the species here under consideration : (1) The connecting process is stated to be “ conspicuously deve- loped, and pointed”; one of the chief characters of subbadius. (2) The lancet is but “slightly emarginated towards the point” ; also one of its principal characters; for the salient point in the sentence is the word “ slightly,” as proved by a comparison with the immediately subsequent description of lepidus, in which the lancet is called ‘‘ considerably emarginated towards the tip.” (3) Forearm “12 inches” (34°8 mm.); third finger “ 14 inches” (47°6 mm.); these measurements, as being smaller than in any other species, and like those of the individual before me (forearm 34:2, third finger 46-4 mm.), settle the identification beyond all doubt. Rh. garoénsis.—Dobson’s Rh. garoénsis (1872) is evidently the same species as Blyth’s Ah. subbadius* (to which there is no reference in Dobson’s ‘Monograph’ or ‘Catalogue’). The two authors emphasise the same points :—(1) The connecting process is described by Dobson as “ forming an acutely pointed elevation.” (2) The lancet is a “ broad, triangular, pointed process,” or, as he says in his ‘ Monograph,’ “almost an equilateral triangle” ; both of these features are the same as already pointed out by Blyth. (3) The Bat is said to be ‘probably the smallest known species of the genus,” the forearm measuring only 1°3 in. (33mm.). (4) Width of horse-shoe 0:2 in. (5'1 mm.); a very narrow horse-shoe is also characteristic of the species (5°5 mm., as measured by myself). In the type of garoénsis p, is, according to Dobson, in the tooth- row; this is of no importance for the identification ; the position of this tooth is ‘“‘ vacillating ” in the whole lepidus section. 22. RHINOLOPHUS MONOCEROS, sp. Nn. Diagnosis. Subbadius-type. Larger: forearm, in a not full- grown example, 38°2 mm. Details. Connecting process (text-fig. 22 ¢, on p. 121) and lancet as in subbadius. Horse-shoe markedly broader. General size considerably larger. Tail proportionately longer. The type, and only specimen known to me, is not full-grown (supraorbital crests still separated posteriorly ; no saggital crest ; metacarpals far from having acquired their full length). In the table p. 132 I give only those measurements which may be of some use for comparison with Rh. subbadius. Dentition. p, external. p, and p, in contact, p* in row; cusp very minute. Type. © jyav. (in alcohol). Baksa, Formosa; June 5th, 1893. Collected by Mr. P. A. Holst. Presented by Henry Seebohm, Esq. Brit. Mus. no, 94.2.4.1. * This view was held by the late Dr. Blanford, who, however, put the names down as synonyms of Rh. minor (J. A. S. B. lvii. pt. ii. no. 3 (1888) p. 262; Fauna Brit. Ind., Mamm. pt. ii. (1891)'p. 277). . 9 132 MR. K, ANDERSEN ON BATS [May 16, Measurements of Rh. gracilis, subbadius, and monoceros. Rh. gracilis. | Rh. subbadius. | Rh. monoceros. @ ad. gad. ® juv. Type. Type. | mm. mm. mm. Ears, length . se eoadomoae 15°7 145 - createst ‘breadth... rep ineeievaee 11 112 Nose- leaves, total length. ons 112 10 a “i breadth of horse-shoe _... 6:2 55 65 Rorearm.. EE Maen eee een 36:2 34:2 382 3rd metacarpal .. Be ems 25 24:8 she GB ifs 8 ame A toe eee A ee a NE 9-7 9°8 11:2 1 Gane nee Reet tantn to sacecaanicn 12 118 4th metacarpal... Sounkaueecuantoco ees 26°5 | 25 BN IV. 4 Secale ata tae | EU 7-2 88 BV) Pa 88 87 Sula 4 metacarpal... 265 25 Ne Tail RE Aer enn eo Re Aare 13:5) 14. 178 J DraKiiGred Gee Ss eearsh ucagereaseadcnmocescnonenc: 148 14°8 165 DOO soo BF Ge MUST TEC 8 | 7:3 Skull, total length — Ot lh rear eee ante BS Y/ | eivemastoidiwidthie ane. ee nial | op WGUEI Ose LOMEMNAGCAEE 55000 cceovosce i | Pye omatic wid ulin eer peeeeeetrrr Tegel » supraorbital length . A | a width of nasal swellings .. seette 45 | 4. Mandible, IUSTeVER dle eeiepaudonendoe toeead scones 10 | 10:2 Winperstectly serw tegen siaserorenees a see 6 6 Mower teeth ei. foecs cae ese dacetenee 63 6-2 93, RHINOLOPHUS ACUMINATUS Peters. Diagnosis. Connecting process of the Jlepidus-type. Sella parallel-margined, Fore earm 47-51 mm. Details. This species, together with Rh. sumatranus and calypso described below, form a small, well-marked section of the lepidus group, which might conveniently be termed the acuwminatus section, confined to Java, Lombok, Sumatra, and Engano, and differing from all the foregoing species:—(1) in being very much larger; 7h. lepidus is in size like a Kh. hipposiderus ; Rh. swumatranus like a small Rh. ferrwm-equinum; (2) in being a trifle more advanced in dentition: there seems to be no “‘vacillation” in the position of p,. Sella in Rh. acwminatus practically parallel-margined ; on very close examination an extremely faint indication of an expansion below the middle can be traced. Lancet. strongly hastate. The rest of the nose-leaves, the mental grooves, the ears, the wing-structure, the length of the tail, and the insertion of the plagiopatagium (on the ankle, or slightly above or below) as in th. lepidus. 1905.] OF THE GENUS RHINOLOPHUS. 133 Skull. Very much larger than in lepidus. There is no essential difference in the shape *. Dentition (two skulls). p, external. p, and p, quite, or almost, in contact. p* in row; a minute cusp, pointing inwards. Measurements. On p. 134. Geographical races. There ave two forms of Rh. acwminatus, differing in size and in geographical habitat. 23 a. RHTNOLOPHUS ACUMINATUS Peters, TYPICUS. Rhinolophus acuminatus Peters, MB. Akad. Berlin, 1871, p- 308; Dobson, Cat. Chir. Brit. Mus. (1878) p. 113. Rhinolophus peterst (partim, nec Dobson 1872 et 1880) Dobson, op. cit. (1878) p. 114. Diagnosis. Larger: forearm 50°5-51 mm. Colour.—(1) Dark phase: ¢ ad., skin; teeth unworn. As Rh. refulgens. (2) Russet phase: 9 ad., in alcohol, unfaded; teeth unworn. ‘¢Cinnamon-rufous”” above; base of hairs of the same colour ; under side lighter. Distribution. Java. 23 6. RHINOLOPHUS ACUMINATUS AUDAX, subsp. n. Diagnosis. Smaller: forearm 47—49°5 mm. Colow. Two adult females, in alcohol, unfaded; teeth unworn, or worn. As Lh. refulgens. Type. Q ad. (in aleohol). Lombok. Collected by A. Everett, Esq. Brit. Mus. no. 97.4.18.16. Remarks. This form ought perhaps to be separated specifically from Lh. acuminatus. The mandible is markedly shorter, the teeth a trifle smaller, the nasal swellings slightly narrower, the geographical habitat quite isolated from that of RA. acuminatus. But the Bali form, still unknown, may perhaps connect them together. 24. RHINOLOPHUS SUMATRANUS, Sp. n. Rhinolophus petersi (non Dobson 1872 et 1878) Dobson, P. Z.8. 1880, p. 462 (specimen examined). Diagnosis. Acuminatus section, but sella very distinctly expanded below the middle. Width of horse-shoe 8-3 mm, Forearm 51—51-2 mm. Details. Chief characters :—(1) compared with acuminatus : the very different shape of the sella, as described above ; width at base, at expansion, and at summit: 2, 2°4, and 1-7 mm.; (2) compared with calypso : the much narrower horse-shoe. Colour. 3 ad., in alcohol, unfaded; teeth unworn. Upper * The skull of the species of the acwminatus section is much like that of RA. rouxt. It can, however, always be discriminated by the broader nasal swellings. The mandible is, proportionately, longer. 134 MR. K. ANDERSEN ON BATS [May 16, side darker than “ mars-brown,” lighter than “ burnt-umber” base of hairs scarcely differing in colour; under side “russet.” This looks like an intermediate stage between a “dark phase ” and a “russet phase.” A second specimen (Gottingen Museum) is, however, quite of the same colour. Skull. As in Rh. acuminatus. Dentition (one skull). p, external. p, and p, quite in contact. p? in row; a minute cusp, pointing inwards. The interspace between the upper canine and p* is narrower than in acwminatus. Measurements. Below. Type. & ad. (in alcohol). Lower Langkat, Sumatra; 1898. Presented by Herr Gustav Schneider. Brit. Mus. no. 4.4.1.1. 25. RHINOLOPHUS CALYPSO, sp.n. (Plate IV. fig. 19 a, 6, ¢.) Rhinolophus affinis (aon Horsf.) Thomas, Ann. Mus. Civ. Genova (2) xiv. (1894) p. 108. Diagnosis. Similar to Rh. swmatranus, but horse-shoe much broader: 10:2 mm.; ears longer and much broader. Forearm 52-52°3 mm. Measurements of Rh. acuminatus, sumatranus, and calypso. Rh. acuminatus. aes Rh. calypso. | f. typica. audax. 2 specimens, 2 specimens, 2 specimens, 2 specimens, 1 skull. 1 skull. 1 skull. | 1 skull. Sar eae = || a | Min. Max. Min. Max. Min. Max. | Min. Max. | mm. mm. mm. mm. mm. mm. mm. mm | Kars, length .. sels 18 19 187 19 I 19:5 2s » greatest breadth... | 14 14 145 143 14:3 163 168 Nose-leay es, total leng th ssceall . U4 14 =14°8 EG 16 =-16°8 8 breadth of horse-shoe a8 82 Sule e2 82 83 10:2 102 | Forearm | 505 51 47 495 Sl S152 O23 3rd metacarpal ... | 35:8 365 33°7 35°2 35'°2 36'8 37 883 | TIT.} 162 162 15 15 15:2 163 15 158 TIL 198 20°7 175 20 20 21" || 2079) Mais 4th metacar pal... 374 38:7 351 383 372 38 || 3882 39:3 LENSE oi 112 118 97 105 11 17 «|| 10:3 108 V2 ey 13 eu 12 13 13 13°6 12°8 13°8 Sth metacar pal... 3877 38°7 386-388 87°5 38'3 38:2 39°3 Vue ee trea. Selene Gas soma on mah MOG il OTS Tiles; Wes} 12°2 12:7 ily nil} Ven Ree ete vc ate tin ne RTOS 13 13°5 13°7 14:6 12°8 13°8 Tail poacsccvecs santas veboondaeoancacceal| LE 217 = 23°5 25:2 265 24-7 265 Wbowerslecwcrae gc ey ee aena eae ea a eee nee 21 217 22°5 22°5 22°5 23°2 INGOU co ocos 118 10°8 11 10°8 11 10°3 11 Skull, total length . Fea PIG tel Mma nie 21°6 » mastoid width . a3 Fea NI Ne Es ee Oe PATIOS 2 aes 102 » width of Braintcases co ine ie Bike OPS yall pea 92 » zygomatic width ©. .0......-.e.ce 114 C112 socal hemi coe | 109 » supraorbital length . eee 5 53 5 54. width of nasal swellings . Bae eee 6:2 6 62 || 683 Mandible, eae gs 16 148 15°8 15:2 Upper teeth . Ss ci Pea RU GN ha 88 8:2 88 87 Lower teeth ...........0 9°65 9 9°65 9:2 1905. ] OF THE GENUS RHINOLOPHUS. 135 Colour. $ ad.and 9 ad., in alcohol, unfaded; teeth unworn. As Lh. refulgens. Skull. As in Rh. swmatranus, but maxillar width, across the antero-external corners of m’*, narrower (8°1 mm., as against 8°6 in Kh. swmatranus). Dentition. Kssentially as in Rh. swmatranus, but the interspace between the upper canine and p‘ broader; p, and p, not quite in contact. Type. 3 ad. (in alcohol). Kifa Juc, Engano. Collected by Dr. E. Modigliani. Presented by Marquis G. Doria. Brit. Mus. no. 94.1.7.3. General Remarks on the Rhinolophus lepidus Group. The ancestral species—The ancestors of the simplex and lepidus groups were very closely related. The latter had a projecting connecting process, a slightly smaller skull and teeth. But the general shape of the skull, the dentition, the nose-leaves, apart from the process and a very slight difference in the shape of the sella, the ears, the wing-sirwcture, the length of the tail, and, we might even say, probably the size, were either identical or ex- tremely similar in both of these extinct Bats. The place of origin.—There can scarcely be any doubt that the lepidus group originated much farther westwards than the simplea group. If we regard Japan as a continental group of islands, and put aside Java, on account of its peculiar geological history, we still find, not only the most primitive, but in fact all the species of the /epidus section on the Continent. It is only the aewminatus section which has spread over the adjacent larger islands, one of which (Sumatra) has comparatively recently been con- tinental, while another (Java), probably in a more remote period, seems to have been connected with some part or other of Indo- China; and only one form, still so closely related to the Java species as hardly to be specifically different, has found its way so far eastwards as Lombok. The hypothesis, therefore, cannot be called unfounded, that of the two ancestral species, the ancient “simplex” and the ancient “ lepidus,” the former was Hastern in vange (Austro-Indo-Malayan), the latier Western (Oriental). Differentiation * —From a systematic point of view I found it convenient to divide the lepidus section into three “types”; I think that, phylogenetically speaking, there are two only: the lepidus and the minor type. The former, as coming nearest to simplex in the proportionate size of the skull and teeth, is, probably, the more primitive; it is now distributed over the Jndian Peninsula (/epidus), the Himalayas (monticola), and Malacca (refulgens). The latter, the minor-type, has spread from the Himalayas (minor) eastwards through 8. China to Japan (cornutus) ; it is represented on the now quite isolated Anambas Islands (“ minutus”); its occurrence in Java is not surprising, considering * Compare the diagram on p. 1388. 136 MR. K, ANDERSEN ON BATS [May 16, the faunistic attinities of that island; and it has established itself on the western coast of the Indian Peninsula (gracilis). I have but very little doubt that now, when attention has been called to the differences of all these forms of the mznor-type, it will be found algo in other parts of the Indian Peninsula. Tf any inference can be drawn from fragments of a skull and the external characters, the swbbadius-type would appear to be an offshoot of the minor-type: already in minor and cornutus the process is a little sharper-pointed than in lepidus; in swbbadius and monoceros this tendency is carried much further, The skull of the species of the acwminatus section (Java— Lombok, Sumatra—Engano) is of the lepidws-type; the process too; the colour remarkably like that of refulgens. This leads me to suppose that acuminatus and its allies (swmatranus, calypso) are scarcely more than giant representatives of the lepidus-type. Tt is the subbadius-type which, from a zoogeographical point of view, is by far the most interesting: it has spread southwest- wards over a vast part of the Ethiopian Region, and westwards over the Mediterranean countries :— (1) The empusa-type.—Rh. empusa* and blasti have progressed further on the way already indicated by Rh. subbadius. They have the small skull and the small teeth characteristic of m¢nor- subbadius ; in the shape of the skull there is no essential difference ; the dentition is identically the seme; the process is that of a swb- badius; the sella is deltoid, that is: the tendency, in the subbadius- sella (as emphasised above), towards assuming a subacute summit has been further developed; and we still see the constriction at the middle of the sella. But empusa and blasii are (as‘always the Ethiopian and W. Palearctic species) in several points more highly developed: III.? is lengthened (about, or more than, 14 the length of I11.); also IV. is very much longer (not far from twice the length of IV'.). 2h. empusa is, however, an inhabitant of Nyasa- land, far 8. of the Equator, Ah. blasii of the Mediterranean Subregion; thus, the two extremely closely allied species are now separated by an enormous tract, where no relative appears to occur. As we now know that they are descendants of the Oriental subbadius-type, the explanation seems to be quite clear: one branch spread southwestwards, into the Ethiopian Region, and developed into 2A. empusa (slightly more primitive dentition ; shorter ears, broader horse-shoe); another westwards into the Mediterranean countries, Rh. blasii. There is an instructive fact connected with these two Bats: I believe them to be compara- tively recent intruders into their areas; Ah. empusa is known from one specimen only, from the very Hast of Tropical Africa ; Rh. blasii is much more common in the Hastern Mediterranean tract, and still it does not seem to have reached Spain ’. * Andersen, Ann. & Mag. Nat. Hist. (7) xiv. (1904) p. 378 (there is a misprint on p. 380: the length of the mandible is 1271, not 13°1 mm.). + Not recorded in Cabrera Latorre’s “Quirépteros de Hspafa,” Mem. Soc. Espan. Hist. Nat. ii. (1904). Iam also not satisfied that there is any reliable record from the African coast of the Mediterranean. 1905. ] OF THE GENUS RHINOLOPHUS. 137 (2) The landeri-euryale type—The Ethiopian Rh. landeri (Fernando Po, Gaboon), Rh. lobatus (Lower Zambesi to Mombasa), and Rh. dobsoni* (Kordofan) have the small skull and the small teeth characteristic of minor-subbadius; the same shape of the skull; the same dentition (no vacillation in the position of p,); the process is that of a swbbadius. In so far there is no difference at all between this section and the former (empusa-blasii). But in the shape of the sella and in a certain peculiarity in the wing- structure they have taken a course of their own :— We have seen, in the simplex group, a progressive development from a sella constricted at the middle, through a parallel-margined stage, to a pandurate sella; we have seen in the lepidus group, too, the constricted sella (minor) modified into the parallel-margined (gracilis); the Hthiopian species here under consideration represent the third and final stage, the pandurate sella. In addition to this: in all of them IV. is peculiarly shortened: less than (extremely rarely, as a slight individual atavism, equal to) half the length of IV*. Asin Fh. empusa and blasii, II.’ is lengthened. Rh, euryale, from the Mediterranean Subregion, is so extremely closely allied to the above-named Ethiopian species that it shares with them ql essential characters (even the highly peculiar shortening of LV."), with one exception: i¢ has retained the parallel- margined sella. Summary.— When discussing the affinities of the Ethiopian species of the Ah. simplex group (above, pp. 117-20), I arrived at the conclusion that they are undoubtedly derived from Oriental types, and that, most probably, the ancestral species have spread directly from South Asia into the Ethiopian Region. As will be observed from this, a study of the Ethiopian repre- sentatives of the 2h. lepidus group leads to quite the same result: they have thei closest known allies in the Oriental Region, but they are, without exception, considerably more highly developed than any of their Oriental relatives. Bats of the subbadius-type have evidently spread from some part of South Asia southwestwards into the Ethiopian Region (empusa ; landeri, lobatus, dobsoni), and westwards over the Mediterranean countries (blasii; ewryale). Of all the species of the 2h. lepidus group only one has found its way to Lower Egypt, Rh. euryale. Tt is a species exclusively Mediterranean in range, and unusually lable to differentiation imto slightly differmg local forms’. Its presence in Lower Egypt is easily explained by invasion from the adjacent Asiatic coast of the Mediterranean, where it is very common (specimens from Lower Egypt are indistinguish- able from the Palestine form, Rh. e. judaicus) <. * Thomas, Ann. & Mag. Nat. Hist. (7) xiv. (1904) p. 156. + Andersen and Matschie, “ Ueber einige geographische Formen der Untergattung Euryalus” (SB. Ges. naturf. Fr. Berlin, 1904, pp. 71-83). { Although it is beyond the strict limits of the present paper, I propose to insert a tew words on the remaining Ethiopian species of the genus :—The ethiops section (Rh. ethiops, hildebrandti, and fumigatus) are very closely related to the Hima- layan Rh. macrotis, but much more highly developed in the dentition, the wing- 138 MR. K. ANDERSEN ON BATS [May 16, The probable affinities and phylogeny of the principal forms of the Rh. lepidws group are expressed in the subjoined diagram (Ethiopian types marked with an asterisk) :— ewryale. * landeri-type. Se % * py -tv *“empusa-type. _» (midas group.) es / ra y Baa subbadius-ty pe. ean, Pnenanee ae acuminatus-ty pe. Re PRS sch O ze —___—->(simplex group.) IIL. Tart RarwoLorxus uipAs GROUP. Diagnosis. Cochlee large, making the basioccipital, between them, extremely narrow (linear). Posterior connecting process very low and rounded off. 26. RuINOLOPHUS MiDAS, sp. n. (Plate IV. fig. 204, 6, ¢, d.) Diagnosis. Sella almost deltoid, summit rounded. Forearm 37°7 mm. Details. Horse-shoe as broad as the upper lip; no “tooth” on the sides of the median notch; no crenulation of the border. Lateral margins of sella converging from base to tip; breadth at base (2°3 mm.) much more than half the vertical height of the sella (3°5 mm.); a very slight (rather easily overlooked) constriction at the middle; summit rounded (breadth 1°6imm.). Connecting process very low, and broadly rounded off. Lancet long (4 min.) and cuneate. One mental groove only. Kars a little longer than in minor, outer margin immediately below the tip somewhat more emarginate; tip more distinctly pointed. Wing-structure, compared with that of mor, considerably structure, and the mental grooves (Andersen, Ann. & Mag. Nat. Hist. (7) xvi. Sept. 1905, pp. 291-92). Rh. maclaudi is an Ethiopian representative of the Rh. philip- pinensis group, but on a considerably higher stage of development in the same respects as the species just named (Id., tom. cit. Aug. 1905, pp. 254-55). This completes the account, showing that all the Ethiopian Rhinolophi, without exception, are of Oriental origin. 1905.] OF THE GENUS RHINOLOPHUS. 139 modified, chiefly in two respects:—(1) the third metacarpal is shortened ; but at the same time the fourth metacarpal has remained the longest (as in all primitive species of hinolophus) ; (2) IIT’, IV.’, and V.’?, that is all the distal phalanges, are lengthened. Compare the table of measurements of Rh. midas and hipposiderus on the one side, with those of minor, lepidus, and all their allies on the other (see p. 143). Tail rather long, 13 the length of the leg. Plagiopatagium inserted on the ankle-joint. Colour (somewhat faded in alcohol) probably as light as in Lh, blast. Skull, In all species of Ahinolophus the cochlee are large, making a narrow basioccipital (compare the genus Hipposiderus) ; but in Ah. midas and hipposiderus the peculiarity is carried to an extreme: the cochlew are so much increased in size as to reduce the basioccipital to a linear bridge of bone; in some individuals (of Lh. hipposiderus) the cochlee are almost in contact. This character alone makes the skull of these two species easily distinguishable, at a glance. But in every other respect, in the shape, the size, and the teeth, the skull is so exceedingly like that of Kh. minor, that there can scarcely be any doubt as to the very close relationship of the minor and midas types. Dentition. On the minor stage :—p, external. A very narrow interspace between p, and p,. p quite in row; a small cusp, pointing inwards. Upper canine and p* well separated. Type. 9 ad. (in alcohol). Jask, Persian Gulf. Presented by A. Butcher, Esq. Brit. Mus. no. 94.11.16.1. Remarks. The discovery of this highly interesting species seems to remove all doubt as to the close affinities of minor and hipposiderus. The sella of midas is intermediate between that of minor and hipposiderus ; it recalls that of empusa and blasii, which also are modifications of the minor-type; to the peculiarly long and cuneate lancet we have a parallel in one of the modifications of the minor-type described in this paper, viz. Rh. gracilis. The geographical habitat of midas is, too, rather intermediate between the Oriental minor and the W. Palearctic hipposiderus. Rh. midas is, of course, readily distinguishable from Zh. hipposiderus by the shape of the sella. In the width of the brain-case, as well as in external dimensions, it is like the southern, more primitive form of Aipposiderus (Lh. h. minimus). 27. RHINOLOPHUS HIPPOSIDERUS Bechst. Diagnosis. Sella cuneate; summit pointed. Forearm 34:7— 41°7 mm. Details. Breadth of sella at base never more, but generally less, than half its vertical height. Colow’. (1) Younger, but quite full-grown individuals; skins; Cyprus, 8. Carpathians, Switzerland. Very nearly ‘‘ mouse-grey ” above; horse-shoe patch faintly, or not at all, indicated; base 140 MR, K. ANDERSEN ON BATS [May 16, of hairs of the upper side and the whole of the under side “ drab- erey.” (2) Aged individuals; skins; Cyprus, Malta, Balearic Islands, Switzerland, Germany. Much browner. General colour above brownish “ drab,” with some individual variation in the shade of the colour: sometimes almost ‘“wood-brown” (lightest extreme), sometimes with a tinge of “ Prout’s brown” (darkest extreme) ; horse-shoe patch indicated, or quite obliterated; base of hairs “ecru-drab”; under side ‘“ ecru-drab,” sometimes with a tendency towards “ drab-grey.” Skull, As in Rh. midas. Dentition. AS in minor and midas. In the series of skulls examined (20; of all races) there is, of course, some variation in the position of p,; the general rule is: p, external, p, and p, almost or quite in contact; one extreme: p, almost in row (one skull), and p, and p,, therefore, well separated ; the other extreme: p, not only external, but hair-fine (four skulls; teeth unworn), or disappeared and the alveoli obliterated (two skulls; teeth unworn). Distribution. From Gilgit to Ireland; from the Baltic to Sennar, Geographical races. The series examined—95 examples, from almost the whole area occupied by the species—enables me to recognise three races of 2. hipposiderus. 'The first two of these would probably be called distinct species by other zoologists. 27a. RHINOLOPHUS HIPPOSIDERUS MINIMUS Heugl. Khinolophus minimus Heuglin, N. Act. Acad. Czs. Leop.-Car. xxix. (1861) p. 6. Rhinolophus hipposiderus minimus Andersen, Ann. & Mag. Nat. Hist. (7) xiv. (1904) p. 455. Diagnosis. Small: forearm 34:7-38 mm. Details. As lately pointed out by me elsewhere (1. s. c.), v. Heuglin’s 2h. mininws, first described from Keren in Erythrea (type in the Stuttgart Museum), is a well-marked geographical race of Kh. hipposiderus, differing from the Central Huropean form by its considerably smaller size. At the same time I mentioned that the British Museum possesses an example from Sennar indistinguishable from the type specimen of minimus. A subsequent examination of the whole series of Rh. hipposiderus preserved in the British Museum has revealed the rather surprising fact that Kh. h. minimus is by no means confined to Keren and Sennar, but generally distributed over the Mediterranean Subregion. It differs from the Central European form in being in every respect smaller ; in some respects, as it seems, absolutely smaller, in others at least on an average. I find the length of the forearm to be the best means for a ready discrimination: in minimus, 347-38 mm.; in the typical form, 39-41:7 mm. For other details, cf. the measurements on p. 143. 1905.] OF THE GENUS RHINOLOPHUS. 141 The skull is markedly smaller, the nasal swellings a trifle narrower, the teeth slightly smaller. Distribution. 32 specimens examined. As it is of some interest to have the range of this hitherto overlooked form exact] determined, I subjoin a list of the localities from which I have seen examples, together with measurements of the forearm; it might perhaps lead to further investigation :— Keren (1, the type*): forearm 36°3. Sennar (1): 36°5. Cyprus (6): 34°7-37-7. Smyrna (1): 37°5. Malta (8): 36-37. Middle Italy (Ostia 2): 35°7-36°8. Corsica (1): 37:7. Haute Savoie and Geneva (2): 37°7-38. Balearic Islands (7): 36°2-37°6. Seville tT (1): 37-7. Morocco (Tangiers 1): 37:7. Portugal (Cintra 1): 362. Summary of Distribution :—The Mediterranean Subregion, southeastwards to Sennar and Keren. Ee it noted: there is no record from Hgypt (and, very likely, it does not occur there: ¢f. remarks on p. 143). Remarks. In the whole series of Rh. hipposiderus examined (apart from the British specimens, of course) I have not found any individual which I could not easily refer either to the southern or the northern form. I have some reason to believe that in certain border districts (e.g. 8.W. Switzerland ; perhaps also Cyprus) the two forms occwi together, perhaps side by side, but intermediate examples I have never seen. They will probably be found. 27 6, RHINOLOPHUS HIPPOSIDERUS Bechst., TYPICUS. Vespertilio Ferrum equinum (partim) Schreber, Siugthiere, i. (1775) pp. 174, 188, pl. 62 (lower fig. only). Vespertilio equinus (partim) P. L. 8. Miller, Natursyst., Suppl. (1776) p. 20. Vespertilio Ferrum equinum, 6. minor, Gmelin, Linn, Syst. Nat. i. (1788) p. 50. Vespertilio Hippocrepis (partim) Schrank, Fauna Boica, 1. (1798) . 64, ‘ Vespertilio Hipposideros Bechstein, in Pennant’s Allg. Uebers. vierfiiss. Thiere, 11. (1800) p. 629, footnote (compare also pp. 615 and 736). Vespertilio hippocrepis Hermann, Obs. Zool. (1804) p. 18. Rhinolophus bi-hastatus Geoftvoy St.-Hilaire, Descr. de lKgypte, i. (1812) p. 132; id., Ann. Mus. d’Hist. Nat. xx. (1813) p. 259, pl. 5. * Wor the loan of this specimen I am indebted to Prof. Dr. Kurt Lampert, Director of the Royal Natural History Cabinet, Stuttgart. The type is a young, but apparently fullgrown, individual. 4/2 other examples otf hipposiderus, of allvaces, of which I give the measurements, are fully adult (distal epiphyses of metacarpals ossified). + As I have seen only one example frem Spain, I may mention that of the whole series examined by Cabrera Latorre, for his “‘Quirépteros de Espafia,’ no Spanish specimen had the forearm more than 37°5 mm. (Mem. Soc. Espan. Hist. Nat. ii. (1904) p. 252). IT am unacquainted with the Rh. phasma (allied to hipposiderus) described by Cabrera in the same paper. 142 MR. K. ANDERSEN ON BATS [May 16, Rhinolophus HHipposideros var. typus, alpinus, et pallidus (partim) Koch, Jahrb. Ver. Naturk. Nassau (1862-63) pp. 530— ol. Rhinolophus hipposideros (partim) Peters, MB. Akad. Berlin, 1871, p. 310; Dobson, Cat. Chir. Brit. Mus. (1878) p. 117. Rhinolophus bihastatus var. Kisnyiresiensis Daday, Orvos-Term. Ertes. x. pt. 3 (1885) p. 274. Rhinolophus hipposideros var. troglophilus Daday, Magy. tud. Akad. Ertekez. xvi. pt. 7 (1886) p. 8, figs. 1, 2. Rhinolophus euryale helvetica Bretscher, Vierteljahrsschr. naturf. Ges. Ziirich, xlix. (1904) p. 256 f. Diagnosis. Large: forearm 39-41-7 mm. Distribution. 33 specimens have been examined, from the following localities :— Gilgit (1): forearm 39°8. Urmi, N.W. Persia (1): 39°8. Van, Armenia (2): 39°2-39°3. Cyprus (1): 39°64. N. Bulgaria (1): 39. Roumania (13): 89-41:2. Transsylvania (2): 40-41. 8. Car- pathians (1): 39:3. Hungary (1): 41-7. Schlangenbad (2): 40-40-1. Strassbourg (3): 39-40°1. Thurgau and Vallais (5) : 40°2-41°7. Summary of Distribution :— From the extreme N.W. Himalayas, through N.W. Persia and Armenia, over the whole of Central Europe N. of the Balkans and the Alps. 27c. RHINOLOPHUS HIPPOSIDERUS MINUTUS Montagu. Vespertilio minutus Montagu, Trans. Linn. Soc. ix. (1808) p. 162, pl. 18. figs. 7-8. Diagnosis. Forearm 36°3--39 mm. Details. English and Irish individuals of 2h. hipposiderus differ from the Central European form in being on an average (and nearly always also absolutely) smaller. The length of the forearm yavies, in 30 adult specimens from England, Wales and Ireland, between 36°3 and 39 mm., the average being 37:6. In other words: the average size of the British race is considerably below the minimum of the typical form, and almost exactly like maximum of Rh. h. minimus. Distribution. England, Wales, Ireland §. Technical name. Till the close of the 18th century, the two Bats now called Rh. ferrwm-equinum and Lh. hipposiderus were ** Koch’s “varieties” are scarcely determinable, his descriptions being utterly vague and based upon such characters as are subject to individual variation or dependent on age: var. typus and alpinus belong, probably, to the Central Muropean form ; var. pallidus seems to be a mixture of this and the southern race. + A glance at the measurements in Bretscher’s paper is sufficient to show that what he takes to be “eine ausgesprochene Lokalform ” of Rh. ewryale is an ordinary, typical Rh. hipposiderus ! + Lought perhaps to mention that this example, the only typical hipposiderus T have seen from Cyprus, is a dealer’s specimen; a Cyprus series collected and presented by Miss Dorothy M. A. Bate (cf P. Z.S. 1908, ii. p. 342) are unquestionably of the Mediterranean form. § For details, cf. J. E. Kelsall, “The Distribution in Great Britain of the Lesser Horse-shoe Bat,” The Zoologist, xlv. (1887) p. 89. 1905.] OF THE GENUS RHINOLOPHUS, 143 regarded as a large and a small variety of one species. In 1808, Montagu pointed out some of their distinctive characters, and proposed for the smaller species the name Vespertilio minutus, being evidently unaware that the two Bats had already twice been specifically separated—by Bechstein in 1800, and by Hermann in 1804, Montagu’s name, as being antedated by “ hipposiderus,” was soon almost completely forgotten (it is not recorded in Dobson’s Catalogue). The original description of V, minutus being, however, based on English specimens, the name is now available for the British race ot hipposiderus. Remarks. We are now able to form a much clearer idea of the past history of Ah. hipposiderus. It originated from a Bat allied to Rh. minor, somewhere in Asia, most probably near the western border of (if not within) what is now called the Oriental Region. From there it spread southwestwards into Africa, west- wards through the Mediterranean countries to Central Europe and the British Islands. There is, to my knowledge, no record of Rh. hipposiderus from Egypt; if this is evidence that it does not occur, and has not occurred, there, it is at the same time a Measurements of Rh. midas and hipposiderus. | Rh. midas. Rh. hipposiderus. | | | Sad | minimus. | f. typica. minutus. | Typ ni 132 eee | 33 specimens,| 30 specimens, | : | 12 skulls. | 6 skulls. 2 skulls. Min. Max. | Min. Max. | Min. Max. mm. | mm. mm.| mm. mm.| mm. mm. Ear 8, leng Gee sono anddod so AGU CeO 1/ | 14 16 | 15 165 14:2 15°5 |, greatest breadth 00... ccc. 13 VoO)) ula esl dar 12:8) hasten Ties Nose- Teav es, total Teneth) .2...2.0...... 12°8 ) WOR TY) hale) 18) 10 118 3 breadth of horse-shoe ... 73 6 68 Goa 6 67 IM OTeaTITENe SS EY, hth readies emcee By Sule, Bis} |) BD) ae || BS) BD) rol AACA! — soasoonaceadbon soa sasusa x08 242, 22:2 25:7 | 248 27:3 | 29°38 249 TITEL EN Sh nee ea en a UI Pte WAG yp Ha 128 | 197 14-9 11:6 13:2 OLY. PE GASRIGoaEComtmchob dasa acne Noe | ists} |) dlse7/ US) ey USF 163 18°7 4th metacarpal — Eee eetcaenanaeeernces 27 25) e292 P28 3072) Zu 29 1A OU enon peacend aaacoe nen sosgmaacasnscs ors Wal IN, G8 a8 7 8 Gia 8 Tee i Sys UE cutie hn Semeachaa saan a 12 I 109 13:2 12 141 ile Wee 5th metacarpal — HES 2 Sener tN 25:8) || 2315 27e7 | 272 29:7" |" 2av7 98:2 Vv. Sa Ul Ae) Oe a Koe39 72 98 LR ED MMe Connect Bena Sao Hine aoc rR OGG 12:2 | ils) 1g) TBS} 233 12:5 1438 Tail PE SLs REA GaSe SEEM Nee ana CSS 2405 | 235 277 | 262 30:3 23°5 27 io werslera eae ee nce ieee: NO Gays} |) AS UD). | GSS. S55 Foot ....: Rees eythane cate ae 76 RD IY a RSS 7-5 R7 Skull, total length | Pe etree coke ccts ol 15:9 145 15° 16 16°2 16 16 » mastoid Rridthen eecet ane fae i ee ee “en ARS} 73 7:8 width of bram-case 07 00.-.0......| i il Gl GH | G8. Ge 68 68 5 “zygomatic width .......6:0......:. IAS eee EAD VRS 8 8 8 fy WaECIMENP GrAGhAN aac soo oodosscnavcs dee 58 Be Gs) | BS) BS 56 57 .», Supraorbital length . 45 4 45 | 42 5 43 AA width of nasal sw ellings .. 4-2 a7 83 | Be al 4, 4, Mandible, length .. Re on EL HORS it Srey TO. || NO 10°2 co | LOD Upper 1 aT REA Os Nang easel Si SS i Ba Bul Ar) By. i IDO WGP WEALD “ang ansooanoe soo ops eto-aboonn gaoa00 6-2 56 8658 58 6 AD 6 144 MR. K. ANDERSEN ON BATS [May 16, proof that it did not reach Erythrea and Sennar from the Mediter- ranean, by way of the Nile Valley, but via the formerly existing, broad land-connection between 8.W. Asia and N.E. Africa. The individuals which established themselves in Central Europe, N. of the Balkans and the Alps, gradually making their way as far north as the Baltic, developed into a distinct, larger race (Lh. h. typicus). The British colony, originally the extreme western off- shoot of the larger form, but soon cut off from communication with the Continental main stem, also developed into a distinct race (Rh. h, minutus); it got the not unusual stamp of an island form: the smaller size; and so it came to occupy, seemingly, but neither phylogenetically nor geographically, a somewhat intermediate position between the northern and southern races of hipposiderus, between its immediate and its more remote progenitors. It is worth noticing that Rh. hipposiderus is distributed over the whole of England, occurring also in several places in Ireland, whereas Rh. ferrum-equinum is confined to the extreme south of England, apparently not farther north than Hssex, Gloucester, and Pembroke, and has never reached Ireland. It may indicate that of these two comparatively recent immigrants into the British Islands, Rh. hipposiderus was the earlier comer. ‘This assumption seems strengthened by another fact. On the Continent Rh. hipposiderus goes farther northwards and considerably higher up on the mountains than ferrwm-equinum. Tt is but reasonable to suppose that the more hardy species was also the first to make its way to England. TV. SUMMARY. 1. A progressive evolution is pointed out from the Austro- Malayan 2h. simplex, through a long series of Oriental forms, to the Western Palearctic Rh. ferrum-equinum (pp. 76-120 ; résumé pp. 116-120). 2. A similar chain from the Oriental Ah. lepidus to the Western Palearctic Rh. blasii and Rh. euryale (pp. 123-138 ; résumé pp. 135-138). 3. The Western Palearctic Rh. hipposiderus has no closer known relative than Rh. midas from the coast of the Persian Gulf, which again can be traced back to the Oriental Ah. minor (pp. 138-144). 4, All the Ethiopian representatives of the genus Rhinolophus are of Oriental origin (pp. 117-120, 186-138). 5. The following 26 forms (14 species and 12 subspecies) are described as new, all of them Austro-Malayan, Oriental, or Asiatic-Palearctic :—h. simplex, p. 76; megaphyllus monachus, p- 80; nanus, p. 82; celebensis, p. 83; virgo, p. 88; nereis, p. 90; stheno, p. 91; rouwi sinicus, p. 98; thomasi, p. 100; affinis hima- layanus, p. 103; a. tener, p. 103; @ macrurus, p. 103; a. supe- rans, p. 104; a. nesiées, p. 104; a. princeps, p. 1065 ferrum- 1905. ] OF THE GENUS RHINOLOPHUS. 145 equinum regulus, p. 112; f. proximus, p. 112; monticola, p. 124; refulgens, p..124; cornutus pumilus, p. 127; gracilis, p. 129; monoceros, p. 131; acuminatus audax, p. 1383; swmatranus, p- 133; calypso, p. 134; midas, p. 138. 6. The following 10 forms, hitherto usually regarded: as iden- tical with other species, are shown to be distinct species or subspecies :—Fh. truncatus Peters, p. 80; borneensis Peters, 2 p. 84; roux: Temm., p. 93; (ferrwm-equinum) nippon Temm., p. 110; (f) tragatus Hodgs., p. 111; lepidus Blyth, ie SS cornutus Temm., p. 127; suwbbadius Blyth, p. 129 ; (hipposiderus) muumus Heugl., p. 140; (h.) minutus Mont., p. 142. 7. The following names, hitherto usually regarded as indicative of distinct species, are referred to the lists of synonyms :— Rh. petersi Dobson, p. 95; garoénsis Dobson, p. 181. EXPLANATION OF THE PLATES. Prats IIT. Rhinolophus simplex group; skulls; front views 2, all other figures te Fig.1. Rh. simplex (p. 76); Lombok; type of the species. Front view. 2a, b,c. Rh. megaphyllus f. typica (p. 79); Cooktown; B.M. no. 3.8.3.3. Upper, lateral, and front views. 3. Rh. nanus (p. 82); Goram; type. Front view. 4a, b. Rh. celebensis (p. 83); Makassar; type. Upper and front views. da, b,c. Rh. borneensis f. typica (p. 84); Labuan; topotype; B.M. no. 65.5.9.15. Upper, lateral, and front views. 6. Rh. malayanus (p. 89); Biserat; topotype; B.M. no. 3.2.6.84. Front view. 7a, b,c. Rh. nereis (p. 90); Siantan, Anambas; type. Upper, lateral, and front views. 8a, b. Rh. stheno (p. 91) ; Selangor ; topotype; B.M. no. 98.3.13.2. Lateral and front views. 9a, b, c,d. Rh. rouxit f. typica (p. 93); Ceylon. Upper, lower, lateral, and front views. 10, Rh. thomasi (p. 100); Taho, Karin Hills; topotype; B.M. no. 90.4.7.9. Upper view. lla, b. Rh. affinis himalayanus (p. 103); Nepal. Lower and front views. 12. Rh. a. tener (p. 103); Pegu; type. Upper view. 13. Kh. a. princeps (p. 106); Lombok; type. Upper view. Prats LV. Rhinolophus simplex, lepidus, and midas groups; skulls; front views 2, all other figures 2. Fig. 14a, 6, c,d. Rh. ferrum-equinum tragatus (p. 111); Nepal; one of the cotypes. Upper, lower, lateral, and front views. 15. Rh. f. proximus (p. 112); Gilgit; type. Upper view. 16a, b,c. Rh. refulgens (p. 124); Perak; type. Upper, lateral, and front views. 17a, 6, ¢. Rh. cornutus pumilus (p. 127); Loo-choo Isl. ; topotype; B.M. no. 2.10.7.2. Upper, lateral, and front views. 18 a, 6, c. Rh. gracilis (p. 129); Malabar coast; type. Upper, lateral, and front views. 19a, b,c. Rh. calypso (p. 134); Engano; type. Upper, lateral, and front views. 20a, b, e, d. RA. midas (p. 138); Jask, Persia; type. Upper, lower, lateral, and front views. Proc. Zoou. Soc.—1905, Vor. II. No. X. 10 146 DR. E. BERGROTH ON STRIDULATING [May 16, 4. On Stridulating Hemiptera of the Subfamily Halyine, with Descriptions of new Genera and new Species. By Dr. E. Bercrory, C.M.Z.8., Tammerfors, Finland. [Received April 1, 1905.} In his paper “ Zur Kenntniss der Stridulationsorgane bei den Rhynchoten,” Handlirsch * has described three different kinds of stridulatory organs in the Rhynchota : the prosternal furrow of the Reduviide ; thestrigose ventral patches of the Division Tetyraria of the Scutelleride ; and the, at that time, still incompletely known stridulating apparatus of the Corixide, of which Kirkaldy? has since given us a complete description and a probably correct interpretation. There is, however, one group of Rhynchota in which these organs have remained unknown to Handlirsch and all other zoologists, except the distinguished systematist Stal, who knew them without recognising their function. In his important paper “ Bidrag till Hemipterernas systematik” { he states that the Pentatomid genera Platycoris, Viarius, Alphenor, and On- cocoris have the following character in common: “‘segmentis ventris secundo et tertio latera versus vitta longitudinali nonnihil curvata, opaca, vix elevata, transversim densissime subtilissimeque strigosa, instructis.” Although Gilippus is described in the same paper, Stil seems to have overlooked the fact that this genus possesses the same structure, and in the systematic arrangement he places Oncocoris far apart from the other genera above mentioned. In a subsequent memoir$ he ascribes the same character to Mecidea, the species of which are mostly African, though it is also represented in India and the temperate parts of America. In a third work ||, finally, he gives “‘ventre anterius vitta laterali transversim strigosa vel rugosa instructo” to a group of genera comprising Jecidea Dall., Platycoris Guér., Niarius Stal, Oncocoris Mayr, Gilippus Stal, Alphenor Stal, and Caridophthalmus Assm. (Allocotus Mayr, preoce.). Mayr has also seen these organs in his Oncocoris punctatus, but simply mentions them as a “ schwache Erhéhung,” without having observed that they are strigose. That they were not quite unknown to Dallas and Walker, will be shown below. Distant** has lately removed J/ectdea from the Halyine, placing it together with a part of the genus Viphe Stal (Aenaria Dist. nec Stil) in a division which he names Mecidaria, without mentioning the transversely striolated ventral patches of Mecidea at all. This is, however, no systematic improvement, these two genera scarcely having anything in common except the * Ann. Naturhist. Hofmus. Wien, xv. (1900) pp. 127-141. + Entomologist, 1901, p. 9; Journ. Quekett Micr. Club, (2) viii. pp. 33-46 (1901). { @tv. Vet.-Ak. Forh. 1867, pp. 491-560. § Enum. Hem. ii. p. 17 (1872). || Enum. Hem. v. p. 34 (1876). @ Reise d. Novara, Hem. p. 46 (1866). «* Rhynch. Brit. India, i. p. 140 (1902). 1905. ] HEMIPTERA OF THE SUBFAMILY HALYIN&, 147 more or less ‘‘elongated body.” Berg* has described two new genera from the southern parts of South America, Proczelicus and Lobepomrs, which he says are allied to the genera Amawrochrous Stal and Oncocoris Mayr; but judging from the descriptions it seems very doubtful whether they belong here, the more so as Amaurochrous does not appertain to the Halyineatall. Distant > has also described an African genus, Crolliws, which he places near Platycoris, but as nothing is said in the description as to the presence or absence of the strigose ventral patches, its position is uncertain, It is said to have the “rostrum about reaching the anterior cox”; if this be correct, it scarcely belongs to the Halyine. Besides the seven genera referred to this group by Stal, there is one described genus which appertains to the same group, viz. Commius Stil. Stal overlooked the striolated ventral patches of this genus and therefore incorrectly placed it among the Pentatomine s. str., near Chaleocoris Dall. For this division of the Halyine I propose the name Platycoraria; all its genera, except Mecidea, are Australian, two of them (Oncocoris and Caridophthalmus) extending to the Austro-Malayan region (New Britain, New Guinea, Timor, Flores, Ceram). As Stal= calls the strigose ventral patches of the Tetyraria “ macule stridulatoric,” but only speaks of “ vitic strigose” in the Platycoraria, it is evident that he did not recognise their true nature in the latter division. A close examination of the ventral patches of the Platycoraria shows, however, that they are perfectly homologous to the strigose areas of the Tetyraria. As described by Handlirsch, the stridulatory organs of the Tetyraria are made up of two different parts—(1) the passive element, consisting of the strigose ventral patches ; and (2) the active element, consisting of a series of minute wart-shaped tubercles, bearing a subapical tooth or bristle and placed on the inner side of the tibie, The ventral patches are straight and situate on each side of the median line of the fourth and fifth segments, sometimes extending to the third or sixth segment; they converge behind and the striz are arranged longitudinally, beimg nearly parallel to the axis of the body. When the insect bends the tibia against the femur and again stretches it, the spinous tubercles of the tibia pass across the strigose surface of the venter, thus enabling the insect, by rapidly repeating these movements, to produce an audible sound. In the Platycoraria both the active and the passive parts of the stridulatory organ show the same structure as in the Tetyraria, but the ventral patches are usually comma- shaped, a little elevated and placed near the base of the venter, beginning with a rather broad base at the anterior margin of the second segment and proceeding, gradually tapering and curved inwardly, to the posterior margin of the third (rarely second) segment, where they end not far from the median line, The * An. Soc. Cient. Arg. xxxii. pp. 234 & 236 (1891). + Aun. & Mag. Nat. Hist. (7) vii. p. 21 (1901). ~ Enum. Hem. iii. p. 3 (1873). Os 148 DR. E. BERGROTH ON STRIDULATING [May 16, patches are transversely strigose in the basal part, but the striz gradually become more oblique and are often practically longi- tudinal at the narrow end. From the position of the patches it is clear that the tibize cannot come in contact with them, and the active part of the stridulatory organ must be sought for elsewhere. { have found it on the inner side of the hind femur, where it consists of a number of very small spinous knobs arranged either in a single regular row or in two or three irregular ones. | They are visible under a common pocket-lens, but under a compound microscope they present exactly the same structure as the tibial spinules of the Tetyraria. I propose to call them “ spicula stridulatoria.” The movements of the femur exactly correspond to the different direction of the striz of the ventral patches, these strie being always crossed at a right angle by the “‘ spicula.” By rubbing the inner side of the femur over the ventral patch I have experimentally produced a stridulating noise. Stal seems to have observed the “spicula stridulatoria” in Platycoris and Niarius, for in his above-quoted paper of 1867 he says they have the “femoribus posticis intus granulatis.” In his larger work of 1876 he has omitted to mention it. In his revision of the Hemiptera Heteroptera of the British Museum, Distant was apparently not satisfied with the state in which he left the genus Dictyotus. He says* : ‘‘ Dictyotus requires revision; all the species which Dallas included in his genus do not appear to be congeneric.” Without having seen Dallas’s types T had myself, in deterrnining some species of this genus, come to the same conclusion. In some of his specific descriptions Dallas speaks of “‘a curved raised line on each side of the second and third segments” of the venter. Walker has also described his Mormidea detersa as having the ‘‘abdomen beneath with a short curved smooth line on each side near the base.” I therefore suspected that these particular species belong to the genus Oncocoris, and this supposition has proved to be correct, my friend Mr. Distant having at my request kindly re-examined the types of the British authors previously referred by him to Dictyotus. It is therefore necessary to give a complete revised list of the species of Oncocoris, which follows here, and which is essentially founded upon the communications received from Mr. Distant. Oncocoris Mayr. Verh. zool.-bot. Ges. Wien, xvi. p. 362 (1866); Reise d. Novara, Hem. p. 44 (1866). Dictyotus Dall. List Hem. Brit. Mus. i. p. 139 (pro parte). Tarba Walk. Cat. Het. Hem. Brit. Mus. i. p. 236 (1867). 1, ONcocorRIs APICALIS Dall. West Australia. Dictyotus wpicalis Dall. List, i. p. 141 (1851). * Aun. & Mag. Nat. Hist. (7) v. p. 388 (1900). 1905. | HEMIPTERA OF THE SUBFAMILY HALYIN&. 149 2. Oncocoris ca@LeBs Fabr. Australia *. Cimex celebs Fabr. Ent. Syst. iv. p. 111 (1794). Oncocoris celebs Stal, Hem. Fabr. i. p. 23. 3, Oncocoris ConFrINIS Dall. Australia. Dictyotus confinis Dall. List, i. p. 143. 4, ONcocoRIS DETERSUS Walk. _ Ceram. Mormidea detersa Walk. Cat. ii. p. 554 (1868). Dictyotus detersus Dist. Ann, & Mag. Nat. Hist.-(7) v. p. 388 (1900). 5, Oncocoris DIMIDIATUS Mont. Victoria. O. dimidiatus Mont. Bull. Soc. Sc. Bucarest, xii. p. 291 (1903). 6. ONCocoRIS DIscoIDEUS Dall. North Australia. Dictyotus discoideus Dall. List, 1. p. 144. _ 7, ONCOCORIS FAVILLACEUS Walk. North Australia. Tarba favillacea Walk. Cat. i. p. 237. Dictyotus favillaceus Dist. Ann. & Mag, Nat, Hist. (7) ve p. 484 (1899). 8, OncocorIs GENICULATUS Dall. South Australia. Dictyotus geniculatus Dall. List, i. p. 142; Dist. l.c. Dictyotus lineatus Walk. Cat. i. p. 181. 9, ONCOCORIS INSULANUS Bergr. New Britain. O. insulanus Berger, Rev. d’Ent. x. p. 202 (1891). 10. OncocoRis LETHIERRYI Mont. Australia. 0. lethierryi Mont, Bull. Soc, Sc. Bucarest, xii. p. 294 (1903). 11, Oncocoris MopEstus Horvy. New South Wales. O. modestus Horv. Term. Fiiz. xxv. p. 601 (1902). 12, OncocorIS OVALIS Berer., infra, p. 153. Queensland. 13. Oncocoris PuNcTATUS Mayr. New South Wales, O. punctatus Mayr, Verh. zool.-bot. Ges, Wien, xvi. p, 362 (1866); Reise d. Novara, Hem. p. 46, tab. 1. fig. 6. 14. ONCOCORIS SEMIMARGINATUS Westw. West Australia. Pentatoma semimarginata Westw. Cat. Hem. Coll. Hope, 1. p. 42 (1837). Dictyotus semimarginatus Dist. P.Z.8. 1900, p. 810. 15. Oncocoris stmiuis Dall. Tasmania. Dictyotus similis Dall. List, i. p. 143; Dist. Ann. & Mag. Nat. Hist. (7) iv. p. 434 (1899). Pentatoma truncatula Walk. Cat. ii. p. 311 (1867). __% Fabricius gives no nearer habitat; I possess examples of the species from Queensland, 150 DR. E. BERGROTH ON STRIDULATING [May 16, 16. Oncocoris suBsIMILIS Mont. Victoria. O. subsimilis Mont. Bull. Soe. Se. Bucarest, xii. p. 293 (1903). 17. ONCOCORIS TRANSVERSUS Car'p. Murray Isl. Dictyotus transversus Carp. Proc. Roy. Dublin Soe. vu. p. 138, pl: xa £1 (US3p): 18. ONCOCORIS TRUNCATELLUS Walk. Australia. Eysarcoris truncatellus Walk. Cat. 111. p. 558. é Dictyotus truncatellus Dist. Ann. & Mag. Nat. Hist. (7) iv. p. 434 (1899). 19. ONCOCORIS VENTRALIS Walk. North Australia. Mormidea ventralis Walk. Cat. iii. p. 555. : Dictyotus ventralis Dist. Ann. & Mag. Nat. Hist. (7) iv. p. 434 (1899). The species of Oncocoris are very similar in facies to those of Dictyotus, but, apart from the stridulatory patches, are easily distinguished by having the antenniferous tubercles visible from above and the metasternal orifice prolonged in a keel. Before proceeding to describe some new forms of this group, I may remark that in Oncocoris and Commius the ‘“ spicula stridulatoria” are arranged in a single straight row, whilst in Platycoris and Niarius they are placed in two or three irregular rows. The other described genera of the group are unknown to me. NIARIUS TRYONI, sp. n. Ovatus, opacus, niger, subtiliter sat dense punctulatus, supra callulis minutis flavidis conspersus, vitta swperiore posteriore et margine angusto laterali capitis, basi hujus subtus, basi articuli primi, tertii quartique antennarum, rostro (articulo ultimo excepto), limbo laterali prothoracis et partis basalis corti, epipleuris, macula prope angulos basales et «pice seutelli, macula oblonga laterali intus rotundata segmentorum abdominalium, acetabulis pedibusque (annulo anteapicalt femorum excepto) flavo-ochraceis, ventre medio impunctato, dilute piceo, nitido. Articulus primus antennarum apicem capitis haud attingens, secundus duobus wultimis wiitis equilongus, tertius et quartus subeque longi. Pronotwm lateribus leviter sinwatum. Scutellum apice impunctatum. Hemelytra apicem abdominis attingentia, corio basin segmenti sextt connexivt superante. Segmentwum genitale maris medio segmentis tribus precedentibus unitis subeequilongum. (Pedes postici desunt.) Long. 5 10 mm. Queensland. Allied to WV. dluminatus Dist., but in Distant’s species the 1905. | HEMIPTERA OF THE SUBFAMILY HALYINA. 151 head, pronotum, and scutellum seem to be differently sculptured, and there is no annulation to the femora. LEVENNA, gen. nov. Caput longitudine latius, leviter convexum, apice rotundatum, marginibus anguste refleais, ante oculos magnos globosos sinuatis, tylo et jugis wque longis, illo postice elevato, his antice valde approuimatis, ocellis majusculis, in linea wneter marginem posticum oculorwm ducta positis, a linea media capitis quam ab oculis saltem duplo et dimidio longius distantibus, bucculis humilibus, rectis ; rostro coxas posticas attingente, articulo secundo duobus apicalibus wiitis breviore, tertio quarto longiore, tuberculis antenniferis @ supero distinguendis, extus spinula porrecta armatis ; antennis quadri-articulatis, articulo primo apicem capitis paullum superante, secundo longissimo. Pronotwm medio capite parum longius, marginibus lateralibus anticis acutis, leviter explanatis et reflexis, angulis lateralibus haud prominulis, anguste rotundatis, angulis basalibus latissime rotundatis, margine basalirecto. Scutellum pone medium leviter sinwatum. Mesosternum carinatum. Orificia metasternalia m rugam mediocrem continuata. Hemelytra apicem abdominis longe superantia, corio apicem abdorinis subattingente (3 ) vel hoc puullo breviore (2), margine apicali levissime sinuato, angulo upicali acuto, membrana venis circiter sex instructa, exterioribus tribus furcatis et pone medium vena transversa conjunctis. Ale apicemabdominis superantes, sed hemelytris paullo breviores. Abdomen hemelytris haud vel vix latins, subtus utringue prope basin vitta stridulatoria curvata usque ad apicem segmente tertii pertracta instructum, segmentis quingue primis ventralibus in mare medio valde retractis, quam lateribus fere triplo brevioribus, segmento sexto hujus sexus permagno, precedentibus unitis medio longiore, angulis apicalibus hujus segmentt in mare latissime rotundatis, fere deletis, margine apicali recto, in femina obtusiusculis, margine apicali late arcuato-sinuato, segmento genitali maris latissimo, angulis apicalibus productis, margine basali sub segmento sexto ventrali subocculto, medio processu liguliformi verticaliter recurvo instructo. Pedes longiusculi, femoribus posticis intus spiculis stridulatorus umseriatis instructis; tibtis omnibus femoribus subequilongis, superne sulcatis ; tarsis triarticulatis, articulo primo tarsorum posticorum tncrassato. This remarkable genus is to be placed near Commius Stal, but is at once distinguished by the four-jointed antenne, the very long hemelytra, and the enormously developed sixth ventral and genital segments in the male. The considerable length of the hemelytra is probably due to the necessity for preserving the inner parts of the very wide and open male genital segment from injury. = 152 DR. E, BERGROTH ON STRIDULATING [May 16, | LEVENNA SALAX, sp. n. Supra niger, sat dense et fortiter punctatus, linea longitudinali media plus minusve distincta verticis, vitta vel linealongitudinali media interdum medio late interrupta pronoti hujusque limbo laterali ab angulis apicalibus ultra medium interdumque etiam margine postico, macula parva ad angulos basales, maculis duabus mediis magnitudine variabilibus fasciaque subapicali scutelli, fascia corii ab angulo apicali interno ad marginen costalem ducta ibique dilatata segymentisque connewiri basi flavis, remote fusco-punctatis, venis membrane presertim apice sepe pallescentibus. Ale fusco-violacee. Caput subtus cum rostro et antennis violaceo-nigrum, bucculis scepeque macula utrinque adjacente flavidis. Pectus violaceo-nigrum, limbo laterali prosterni ab apice ultra medium, carina media et margine laterali mesosternit ac limbo postico metasterni, interdum etiam acetabulis omnibus et prosterno medio, flavis. Abdomen subtus flavo-testaceum, disco medio interdum rufo- piceotincto vel segmento sexto medio macula fusca notato, limbo laterali ventris violaceo, macula quadrata ad angulos basales segmentorum flavida cum disco ventris interdum confluente signato. Pedes nigro- vel fusco-violacei, femoribus posterioribus interdum basin versus flavescentibus. Caput subtus et pectus remote punctulata, articulo secundo antennarum tertio fere dimidio vel saltem tertia parte longiore, tertio et quarto subeque longis. Pronotum longitudine media fere duplo et dimidio latius, marginibus lateralibus rectis. Abdomen subtus impunctatum, appendicibus duabus internis (lateralibus, Sharp) segmenti genitalis maris longe subulatis. Long. 3 6-76 mm., cum membr. 8-9°6 mm. 2 7:8 mm., cum membr. 9°8 mm. Queensland ; South Australia (Yorketown). CoMMIUS MINOR, sp. n. | Flavus, capite (exceptis bast subtus, macula triangulari ab hac ad tubercula antennifera ducta bucculisque), maculis duabus magnis transversis triangularibus anticis pronoti, macula magna basali triangulari seutelli medium hayjus attingente maculaque ejusdem angusta elongata marginali pone medium, meso- et metasterno medio (carina illius excepia), macula magna pleurarum, maculis quingue ventralibus, una utrinque sublaterali segmenti quarti et quinti et wna magna media segmenti sexti, maculaque transversa segmenti genitalis maris violascenti- vel viride-ceneis, maculis duabus magnis basalibus bast contiguis pronoti fuscis, hemelytris subpurpureo-fuscis, opacis, levissime cenescentibus, corio fascia fava inter angulum apicalem internum et marginem costalem ornato; antennis, rostro pedibusque castaneis, his cenescentibus. Caput vertice medio et pone juga remote subtilissime punctulatum, jugis oblique strigosis, articulo secundo antennarum primo plus 1905. | HEMIPTERA OF THE SUBFAMILY HALYINA. 153 quam dimidio longiore, tertio secundo tertia parte longiore (cetert desunt). Pronotwm medio capiti subequilongwn, remotissime et subtilissime punctulatum, marginibus lateralibus anticis subrectis. Scutellum remotissime et subtilissime, pone medium latera versus fortius punctulatum. Hemelytra apicem abdominis attingentia, clavo et corio remote subtiliter punctulatis, margine apicali hujus exterius levissime sinuato. Pectus remote punctulatum, pleuris medio levibus. Abdomen impunctatwm, segmento sexto ventrali maris medio segmentis tribus antecedentibus unitis cquilongo, segmento genitalr maris margine apicali segmenti ultimi ventralis paullo latiore, appendicibus internis (lateralibus, Sharp) faleiformibus, dilute piceis (in elegante longioribus, nigris). aa 3b Imm. Queensland. Much smaller than C. elegans Don. and differently coloured, with the pronotum much shorter, the hemelytra also shorter, the apical margin of the corium much less sinuate near the apical angle, and the genital segment of the male much broader. ONCOCORIS OVALIS, sp. n. Ovalis, livido-testaceus, modice dense sat fortiter nigricanti- punctatus, ventre medio remote punctato, limbo laterali prosterni et ventris impunctato, articulo ultimo rostri, vitta angusta sublaterali paullo curvata dimidii antici pleurarum, macula oblongula ante angulos posticos prosterni, mesosterno medio (carina excepta), macula magna transversa basali ventris, macula media segmenti equs sexti maculaque parva ad angulum basalem et apicalem segmentorum connexivi et ventris nigrinis. Caput pronoto medio subequilongum, tylo et jugis eque longis, rostro coxas posticas superante ; (antennce desunt). Pronotwm longitudine media duplo et dimidio latius, marginibus lateralibus subrectis, angulis lateralibus levissime eminulis, obtusis. Scutellum maculis quingue minutis basalibus et summo apice impunctatum. Hemelytra apicem abdominis superantia, corio areolis aliquot impunctatis predito, membrana JSusea,venis albo-cinerascentibus. Abdomen hemelytris subeque latwm. Pedes maculis punctiformibus nigrinis conspersi, femoribus posticis medium segmenti sewti ventris paullum superantibus, spiculis stridulatoriis fusco-ferrugineis. Long. 2 8 mm. Queensland. HURYNANNUS, gen, noy, Corpus parvum, late breviter rotundato-ovale. Caput planius- culum, dimidio basali partis anteocularis antrorsum admodum angustato, lateribus obtusissimis, convexis, in tubercula antennifera continuo transeuntibus, dimidio ejus apical parallelo, apice late rotundato et medio levissime inciso, lateribus acutis, gugis tylo longioribus et ante hunc contiguis, 154 STRIDULATING HEMIPTERA OF THE HALYINE. [May 16, oculis minutis, brevissime stylatis, ocellis perminutis, ab oculis ac linea media capitis subceque longe distantibus, mox pone lineam inter marginem posteriorem oculorum fictam positis, vertice oculo circiter septuplo latiore, tuberculis antenniferis e supero distinguendis, antrorsum convergentibus, apice extus spinuloso-productis ; antennis ab oculis et ab apice capitis subeque longe insertis, quinque-articulatis, articulo primo apicem capitis haud attingente, bucculis humilibus, rectis ; rostro coxas posticas paullum superante, articulo secundo apicalibus duobus subceequilongis wunitis paullo breviore. Pronotum medio capiti subequilongum, marginibus lateralibus anticis leviter rotundatis, antice leviter sinwatis, margine postico rotundato, angulis lateralibus vix eminulis, obtusis. Scutellum subeque longum ac latum, parte apicali latiuscula, Srenis medium scutelli paullum superantibus. Sterna medio suleata ; orificia in rugam longiusculam oblique antrorsum producta. Hemelytra apicem abdominis paullum superantia, margine apicali corti levissime sinuato, membrana paucinervi. Abdomen hemelytris parwm latius, subtus prope basin utrinque vita stridulatoria curvata ad apicem segmenti tertii extensa instructum. Pedes mediocres, femoribus posticis intus spiculis stridulatorus wniseriatis instructis, tibiis femoribus sub- cequilongis, supra sulcatis, tarsis triarticulatis. Easily distinguished from Oncocoris by the structure of the head, the very small substylated eyes, and the short and broad body. HURYNANNUS LIPPUS, Sp. n. Ochraceus, opacus, sat dense ferrugineo-punctatus (disco ventris tamen remotissime punctato), wbique maculis minutis nigris remotissime adspersus, macula parva nigra ad angulos basales et upicales segmentorum abdominalium notatus. Caput longi- tudine paullo latius, rostro apice nigro ; antennis ochraceis, articulo tertio apice a duobus ultimis totis dilute Serrugineis, articulo secundo primo duplo longiore, tertio primo dimidio longuore, quarto secundo subequilongo, quinto quarto equilongo vel paullo longiore. Pronotwm capite duplo latius. Scutellum mox pone medium levissime sinuatum. Membrana cinerea, maculis minutis nigris remote conspersa. Abdomen (3) dorso apice arcuato-sinuatum, segmento ventrali sexto apice medio subrecto, latera versus levissime sinuato, medio segmento quinto duplo longiore, segmento genitali perpendiculari, ultra segmentum ultimum abdominale haud producto. Pedes maculis punctiformibus iigris remotissime conspersi, femoribus posticis apicem abdominis attingentibus, spiculis stridulatoriis ferrugineis. Long. 3 5:6 mm., lat. 4:3 mm. (Queensland. 1905. | ON THE ANATOMY OF LIMICOLINE BIRDS. 155 5. On the Anatomy of Limicoline Birds; with special Reference to the Correlation of Modifications. By P. Cwaumers Mrrcuent, M.A., D.Se. (Oxon.), Secretary to the Society. [Received May 16, 1905. | (Text-figures 23-28.) In this memoir I use the term Limicole in the sense of Gadow (3) as a major subdivision of the Order Charadriiformes. I have dissected examples of the following forms, and where, in this paper, I refer to family-characters, I must be understood as limiting my remarks to the birds I have myself dissected, unless I definitely state otherwise :— Suborder Limicoa. Family Charadriide ...... Charadrius pluvialis. Timantopus nigricollis. Vanellus vulgaris. Gallinago celestis. Rhynchea capensis. Scolopax rusticola. Chionide............ Chionis alba. Glareolide ......... Glareola pratincola. Thinocoride ...... Thinocorus species @ (Edicnemide ...... “dicnenus scolopax. Paaridee. apices. Hydrophasianus chirurgus. The greater part of the actual dissection was completed in 1902, in continuation of my work on Gruiform Birds (7); pressure of other duties has made it impossible to finish it sooner. JI am indebted to the facilities afforded by this Society in the prosectorium at the Gardens for the material, and to my friend Mr. F. E. Beddard, F.R.S., the Society’s Prosector, for much kindly interest. DIASTATAXY IN THE LIMICOLA. In the arrangement of the feathers on the wing, all Limicoline birds are closely similar. They are diastataxic in the most typical form. The condition in Chionis alba (text-fig. 23, p. 156) may serve as an example. Along the edge of the ulna, from the wrist towards the elbow, the great quills with their associated coverts are arranged in an orderly series, but after four of these rows, each headed by a quill, there is a row from which the quill is missing, forming the diastataxic gap (text-fig. 23, x, p. 156). The carpal remex and covert are present (C.R., C0. ), the covert, in most cases (although not in Chionis), beimg ‘conspicuously larger than the remex. These two feathers lie closer to the most pr oximal primary- 156 DR. P, CHALMERS MITCHELL ON THE © | May 16, quill than to the most distal secondar -quill in most Limicole, but the position varies, and that shown in the diagram is more primitive. As evidence of their association with the secondary series, there is to be taken into account first the fact that the covert crosses the remex as in the secondary rows, not lying distal to it as in the primaries, and, secondly, that a plica (pl.), to which I have called attention in other groups (4 and 5), unites the carpal remex with the most distal secondary remex. Text-fig. 23. A S Wing-structure of Chionis alba. V/-c. R C.C. Diagram of the distal secondary quills and coverts, showing the diastataxic arrangement. S. First secondary. P. First primary. «. Diastataxic gap. C.R. Carpal remex. C.C. Carpal covert. PI. Plica, binding carpal covert to first secondary. The condition of the wing in the Limicole is similar to that found in the greater number of the Columbe, but whereas in some Columbee (4) the eutaxic condition is found—or, as I have tried to show, has been attained,—it has not been attained by any of the Limicole, In the Gruiformes: a somewhat incoherent group certainly closely related to the Charadriiformes, both conditions of the wing are present (7). GUT-PATTERNS IN THE LIMICcOLa. T have already shown (6) that the pattern of the gut in Limicole is of considerable interest. It displays a configuration which differs from the pattern which is archecentrice for all birds in a fashion similar to the divergence shown by the Gruiform birds, 1905. | ANATOMY OF LIMICOLINE BIRDS. 157 and notably different from that of the Columbe, or, indeed, of any other group except the Lari. The duodenal loop is simple and definite (text-figs. 24 and 25, A-B); the portion of Meckel’s tract proximal to Meckel’s diverticulum (Div.) tends to be enlarged in such a way that the diverticulum is not at the apex of a loop as in Columbe or Passeres, but on the distal limb of a loop, which is short in the forms which are less specialised in this respect, such as the Chionide, Glareolidee, Thinocoride, Cidicnemide, and the simpler Charadriide (such as Mumeniws and Vanellus), Text-fig. 24. R Diagram of intestinal pattern of Rhynchea capensis. A. Cut proximal end of duodenal loop. B. Distal end of duodenal loop. Div. Meckel’s diverticulum. C. Origin of ceca. R. Cut end of rectum at cloaca. but which in other Charadriidz and Parride (such as Scolopaa, Himantopus, and Hydrophasianus) is elongated and _ spirally twisted. The portion of Meckel’s tract between this and the duodenal loop is ill-defined in the simpler forms, but in others tends to be thrown into a definite narrow loop. The portion pos- terior to the diverticulum is in close relation to the colic cxca, which are long in the simpler forms (text-fig. 24, C.), but become 158 DR. P. CHALMERS MITCHELL ON THE | May 16, almost atrophied in the more specialised types of gut-pattern— as, for instance, in Hydrophasianus (text-fig, 25, C.). With regard to these two features of their structure, the con- ditions of which in Birds generally are pretty well known, the position of the Limicole is easy to define. In wing-structure they are diastataxic, like all but the most specialised Columbe, and like many of the Gruiformes. In the gut-pattern they are not much modified from the archecentric condition, but the modification is definite, characteristic, and progressive, and, in its simpler form, Text-fig. 25. o a: Diagram of intestinal pattern of Hydrophasianus chirurgus. Lettering as in text-fig. 24. similar to that shown by the Gruiformes. In the more specialised types the elongation and spiral twisting of a portion of Meckel’s diverticulum brings about a superficial resemblance with the arrangement in the long-gutted Columb and Passeres, but the morphological condition is different, as a different portion of the intestinal tract is affected. I shall now endeavour to set out the chief modifications in muscular anatomy that I have found to be of interest in these birds. MuscuLar ANATOMY. Latissimus dorsi anterior et posterior.—The anterior division in all these birds is a broad flat muscular strap, without any peculiarities. 1905. ] ANATOMY OF LIMICOLINE BIRDS, 159 Its insertion is muscular, and just below that of the posterior division (text-fig. 26, L.A. and 3). The posterior division is absent in Scolopax, present in the others, and its proximal edge touches the distal edge of the anterior division in Wdicnemus (text-fig. 26, L.P. and 4) and Hydrophasianus, but not in the others. In Hydro- phasianus and @dienemus the two muscles are almost continuous, although they cross before insertion, at which point they are closely in contact ; whilst in the others the tendon of the posterior division is separated by a short gap from the muscular and more distal insertion to the humerus of the anterior division; this tendon is always in close association with the humeral anchor of the anconeeus. Text-fig. 26. Shoulder-muscles of Gdicnemus scolopazx. Left shoulder ; external view. 2, Tendon of supra-coracoideus. A.S. Anconeus scapularis, the reference lines pointing respectively to the humeral origin and the scapular anchor. S.A. Scapuli-humeralis anterior. §S.P. Scapuli-humeralis posterior. S. Expansor secundariorum. L.A. Latissimus dorsi anterior. L.P, Latissimus dorsi posterior. 3. Insertion of lat. dors. ant. 4. Common insertion of lat. dors. post. and humeral anchor of ancon. scapularis. The posterior division, where present, tends to spread backwards to reach the ilium and part of the ribs. In Chionis its origin is limited to the vertical anterior edge of the ilium, whilst the gap between it and the anterior muscle is wider than in any of the other birds. The archecentric, or most generalised, condition of these muscles in Birds appears to be the existence of an anterior and posterior division, fairly well separated at their origins and close together 160 DR. P, CHALMERS MITCHELL ON THE | May 16, at their insertions. Any well-marked deviation from this condition may be regarded as derivative. I have shown that in the more specialised Columbide the posterior division of the muscle tends to disappear (4); in the Kingfishers the anterior division sunilarly is in progressive diminution (5); in most of the Gruiform birds the anterior division is less strongly marked, whilst the posterior division tends to increase greatly in size and strength, whilst in the Crane and Bustard it is the posterior division which dis- appears (7). In the Limicoline birds generally the anterior division remains in the primitive condition, whilst the posterior tends to enlarge as it does in the Rails, the enlargement being specially a backward and downward extension of the origin. On the other hand, in the Woodcock, as an exception, there is a disappearance of the posterior division. Latissimus dorsi imetapatagialis. —This muscle is probably present in all these birds, but it is very slightly developed and apt to be removed in the process of skinning. Rhomboideus superficialis et profundus— OF these two muscles, the superficial is phylogenetically older. In all these birds it is the thinner of the two muscles, but is longer, being longest in Udienemus and Chionis, whilst it shows a general tendency to die away posteriorly. In Chionis it is nearly divided into a proximal and distal portion by a thin central area, a secondary cleavage which is well marked in the deep muscle of the eutaxic Kingfishers. The deep muscle in Hydrophasianus is almost of the same length as the superficial muscle and it is difficult to separate the two. In the others it is well separated by its greater thickness and by the slope of its fibres upwards and for wards from the scapula to the vertebre. Its origin begins at the extreme posterior end of the scapula and extends forwards under the origin of the superficial muscle, but never reaching so far forwards. In the condition of these muscles, then, the Limicolz are fairly homogeneous; the older superficial muscle is well developed, extending in front of the deep muscle, but, except in Scolopaw, leaving a portion of it exposed behind. The deep muscle has made comparatively little progress in forward extension along the line of the scapula and clavicle. Biceps brachialis——This muscle displays in Wdienemus the condition normal in the majority of birds; it arises by a narrow tendon from the acrocoracoid, and by a broad tendon from the proximal end of the humerus; the rounded belly runs down the arm and ends in a forked tendon, the thicker fork being inserted to the radius, the thinner to the ulna. No doubt, fleshy origins must have preceded tendinous origins, and there is considerable variation as to the relative size of the two origins and insertions in different birds; but the @dicnemus condition is a fairly central one, and it is interesting to notice that in this respect @dicenemus stands apart from other Limicoline birdsand might be associated with many other groups. In Hydrophasianus there is a comparatively slight deviation from the normal, consisting in the complete disappearance 1905. ] ANATOMY OF LIMICOLINE BIRDS. 161 of the humeral head. This has already been noted by Fiirbringer and Beddard (1 and 2), and the latter author states that he found a similar reduction in Rhynchea. In the example of Rhynchea that I dissected, however, I found a very different condition, the well-marked occurrence of a peculiar Limicoline deviation which occurs in a more or less modified form in all the other birds which form the subject of this paper. The Limicoline peculiarity of the biceps brachialis is well marked in Chionis (see text-fig. 27). The main mass of the Text-fig. 27. AG Biceps of Chionis alba. AC. Coracoid head. BI.P. Biceps patagialis. BI. 1. Chief portion of biceps. BI. 2. Accessory biceps. R. Radial insertion. U. Ulnar insertions. muscle arises by a tendon from the acrocoracoid (AC.) and by a large and fleshy head from the humerus (BI. 1); this tapers towards the lower end of the humerus and then divides into a Proc. Zoot. Soc.—1905, Vou. II. No. Xi. 11 162 DR. P. CHALMERS MITCHELL ON THE [| May 16, large tendon inserted to the radius (R.) and a very small tendon to the ulna (U.). There is also a second belly, smaller and rounder, arising almost wholly from the coracoid tendon of origin (BI.2), and towards the lower end of the humerus passing into a round tendon which is inserted to the ulna only (U.), distal to the insertion of the ulnar branch of the tendinous fork of Biceps 1. This doubled condition of the biceps is practically repeated in Scolopax, except that Bi. 1 appears to supply only the tendon to the radius, and the same state of affairs is present in Gallinago, Charadrius, Himantopus, Vanellus, and Rhynchea. It occurs also in Glareola and in Thinocorus, but in the latter the humeral head is degenerate although present. This complication of the biceps shows a link between the Charadriidee, through Chionis, with a more exaggerated peculiarity in some of the Gulls. The condition in Scolopax differs from that in Chionis practically only in the tendon of Bi. 1 in the latter being forked so as to be inserted both to the radius and ulna. If we suppose this fork in Chionis to be split up into the body of the muscle so as to separate the portion of the belly arising from the acrocoracoid tendon from the portion coming from the humerus, the Gull condition would be reached. In Larus, for instance, the tendon of origin arising from the acrocoracoid divides into two fleshy bellies, the one representing Bi. 2 in text-fig. and running to the ulna, the other, fused with Bi. 1 in the figure, running independently to the radius. As there is very strong evidence of other kinds for supposing that the Gulls are modified from a Gharadriiform stock, it would seem natural to suppose that here we have to deal with a case of progressive complexity, starting from the Scolopax condition and leading through Chionis to the Gull condition. But it is important to remember that, in cases of muscles and tendons of birds, the general morphological course is from the more complex to the simpler, and, to my mind, it is more probable that the Scolopax and general Limicoline condition is a simplification from the Gull condition, Chionts showing how the simplification may have come about. Deltoidis patagialis—This muscle is of moderate width in all these birds, and gives off the longus and brevis tendons from its relatively broad distal extremity, with not more than the slightest indication of division into peaks for the different tendons, and so far remaining in a primitive or archecentric condition. With regard to the tendons, Wdicnemus displays a condition markedly different from that found in all the others. The longus tendon is simple and slender ; it has an anchor to the humerus, and, after being joined by the biceps slip, gives off one or two very weak slips of fascia to the patagium, and then takes the usual course towards the wrist. The brevis tendon is simple, flat, and well marked ; it runs an undivided course towards the elbow, parallel with the biceps, and, close to its insertion, broadens out into a fan-shaped termination, which displays in a reduced condition the three slips named respectively a, 8, y by Firbringer. In all the 1905. ] ANATOMY OF LIMICOLINE BIRDS. 163 other birds on my list the condition is much more complex, but as the complexity is similar in all, it is unnecessary to add to the figures given on plate xxi. of Fiirbringer’s great monograph (2). The longus tendon has an anchor to the humerus in all: it is broad and partly doubled in Chionis, Scolopax, Vanellus, and Himantopus ; it is single in Glareola, Thinocorus, Hydrophasianus, Charadrius, Rhynchea, and Gallinago, although in these a greater width in the elastic portion shows a tendency to duplication. In all, from just below the middle of its course, it sends a tendinous anchor inwards and downwards to join with the a portion of the brevis tendon ; the width of this anchor and the exact point and mode of junction with the brevis tendon differ, but the details do not appear sufficiently important for individual description. The brevis tendon is doubled in all, the duplication being complete from origin to insertion. A well-defined tendon nearest to the biceps runs towards the elbow, where it is practically free from the second portion of the brevis; its extremity corresponds with y of Fiirbringer and it occasionally turns in towards the elbow, or may run a straight course towards the ulnar margin of the arm. The second division of the brevis is stronger and wider; it runs parallel with the latter, and nearer the longus tendon. At its distal extremity it divides into two well-marked slips—the proximal, being the # of Fiirbringer and spreading out into a fan running tonal ds the ulnar margin of the arm, whilst the more distal, the a of Fiirbringer, receives the anchor from the longus. Pectoralis propatagialis—This slip is present in all the birds on my list. Leaving the pectoralis major it joins the deltoides patagialis before the muscular part of that muscle has given off the longus and brevis tendons, but its fibres run towards the longus rather than to the brevis. It is weakest in Glareola and Thinocorus. Biceps patagialis.—This muscular slip, to the presence or absence of which Garrod attached so much importance, is present in all these birds and joins the longus. It is much weaker in Glareola and Hydrophasianus, where it is little more than a tendon. In Edicnemus it sends a slip to the patagium, recalling the arrange- ment which I have described in Heliornis (7, text-fig. 78, p. 640). Its length varies, it being much longest in Himantopus, where its distal end nearly reaches the radial margin of the arm—an extremely specialised condition. Where the biceps is double, the origin of the biceps patagialis is sometimes from both portions (text- fig. 27, BI. P. p. 161), but in other cases it comes from the acro- coracoid head only, and in others again from the humeral head. The condition of these alar muscles and tendons is of con- siderable interest. The first salient point is that Mdicnemus stands markedly apart from the others, showing in these structures, as in the biceps, an arrangement much more resembling that found in the Gruiformes, and, indeed, in many other birds, than the typical Limicoline condition. On the other hand, just as the complexity of the biceps in the Limicole recalls the similar ie 164 DR. P. CHALMERS MITCHELL ON THE [May 16, complexity in Gulls, so Gulls exhibit the doubling of the brevis tendon and the anchor from the longus to Fiirbringer’s a, which are the conspicuous features of the Limicoline alar complex. The general trend of change in the formation of the alar tendons seems to have been, first, the formation of distinct tendons from a series of scattered fasciz and cutaneous slips, and next a reduction of the complex tendons to a more and more simple form. The most ready interpretation of the facts appears to me to be that in the ancestors of the Laride and Limicole a complex and specialised alar series of tendons had been elaborated; this condition has been retained by the Gulls and by most of the Limicolous birds, whereas in the Gruiformes and in Wdicenemas it has more or less completely disappeared, leaving traces such as the separation of the distal fan of the brevis into the small divisions which can be recognised as the a, 3, y of Fiirbringer. Deltoides major et minor.—In all these birds both muscles are present and display little divergence. The minor is extremely small in Chionis, Gallinago, and Hydrophasianus; m the others it is normal. The major isa muscle which in many birds displays a progressive tendency to creep down the humerus. It is shortest in Hydrophasianus, not reaching more than three-eighths of the proximal end of the humerus, and is without the usual scapular anchor. In Gallinago it reaches rather less than halfway down the humerus, in Zhinocorus rather more; in the others nearly an exact half, the scapular anchor being well marked in all but Hydrophasianus. Scapuli-humerales anterior et posterior (text-fig. 26, p. 159).— The posterior muscle(S8.P.) is present in all these birds and is large and important, converging from an extensive origin occupying the greater part of the scapula toa rounded tendon inserted to the median process of the humerus. The anterior muscle (8.A.) is small and occupies the usual position across the angle between the scapula and the humerus. It is normal in Edienemus and Hydrophasianus, very small, merely a few fibres, in Glareola and Thinocorus. In Charadrius it is small but quite distinct, whilst in Himantopus it is represented by a narrow band of fibres. In Chionis, Vanellus, Rhynchea, Gallinago, and Scolopax it is absent. There seems little doubt but that the normal, or archecentrie, condition in Birds is for both divisions of the muscle to be present, whilst the anterior division is frequently absent. The Limicole obviously form a group with a marked tendency to the disap- pearance of this muscle, but there is no special correlation between specialisation in other directions and the degree au reduction of the muscle. Expansor secundariorum.—The specialised bined of the anconeus to which Garrod gave the name of ‘expansor secundariorum ” is a muscle in obvious course of disappearance in this group. It is present in a well-marked condition in Gdicnemus (text-fig. 26, 8., p. 159) and Hydrophasianus. Its 1905. | ANATOMY OF LIMICOLINE BIRDS. 165 proximal portion is well marked in Scolopax, Himantiopus, Gallinago, Rhynchea, Vanellus, Charadrius, Thinocorus, and Glareola, but it disappears before reaching the elbow. It 1s absent in Chionis. Ilio-tibialis internus seu sartorius.—This muscle is practically identical in all the birds on my list. It arises from the anterior edge and a narrow portion of the anterior dorsal extremity of the ilium, and has the usual insertion to the fascie over the knee-capsule. In most cases, it shows little sign of fusion with the anterior edge of the ilio-tibialis. Iio-tibialis.—In all these birds this muscle is large, the post- acetabular portion having a strong fleshy origin, whereas the anterior portion is more membranous. Iio-trochanterici posterior, anterior ct medius.—These muscles are all present in typical form in these birds, except that in Thinocorus, Edicnemus, and Hydrophasianus the anterior and medius are nearly fused, showing only a trace of separation at their tendon of insertion to the femur. Nio-trochantericus externus.—This variable muscle is present in all these birds, but is extremely small in Zhimocorus. Ambiens.—This important muscle is present in all the birds on my list, and, in the normal fashion, ends in a tendon which passes through the capsule of the knee-joint and is reinforced (except in Chionis) by a ligament from the head of the fibula, finally forming one of the heads of origin of the muscle complex which gives rise to the perforated flexors of the second, third, and fourth digits. Femori-tibiales sew Crureus and Vastus.—These muscles are alike in all the birds on my list, corresponding almost exactly with the condition I found in Gruiform birds (7), with the exception that in Thinocorus the femoro-tibialis externus is not developed as a separate slip. Caud-ilio-femoralis (Femoro-caudal and accessory F.-c.) (text- fig. 28, p. 166).—The condition of these muscles, to which the researches of Garrod, Forbes, and Beddard have given special importance, differs in @dicnemus from that found in the others. In @dicnemus, as in Otis and many Gruiform birds, the portion with a caudal origin (“ femoro-caudal ” of Garrod) is totally absent ; the portion arising from the ilium (“accessory femoro-caudal” of Garrod) is present and has the usual relations, but displays a considerable tendinous area in the middle of its muscular belly— an obvious sign of degeneration, to which I have already called attention (7). The condition in Chionis (text-fig. 28, p. 166) is more generalised. Both muscles are present, the caudal portion (CAUD. IL. F. 2) displaying a fairly large rounded belly, which tapers to the tendon of origin which is inserted to the femur just distad of the in- sertion of the iliac portion. The iliac portion (CAUD. IL. F. 1) has a fan-shaped origin from the ilium, displaying on its proximal border a well-marked area of tendinous degeneration (X), and is 166 DR. P. CHALMERS MITCHELL ON THE [May 16, inserted to the femur along a narrow vertical line. In Vanellus the caudal portion is very large, whilst the accessory portion 1s present, but minute and with a tendinous degeneration similar to that just described. In Thinocorus, Hydrophasianus, and Rhynchea both portions are present and large. In Glareola, Charadrius, and Himantopus the caudal portion is large, and the iliac is extremely minute, represented by not more than a few fibres. In Gallinago and Scolopaxc the caudal portion is of moderate size, the iliac portion completely absent. Text-fig. 28. IL.FIB (1) NK Wes fits Z. Wizz Vp UY 1 CAUD ILL, TRAE. FT. 6. 11 FIB (2) Thigh-muscles of Chionis alba. Right thigh, external view. IL.TR.E., P., M., A. Tlio-trochanterici externus, pesterior, medius et anterior. IL.FIB. (1). Origin of ilio-fibularis seu biceps, cut and reflected. IL.FIB. (2). Insertion of biceps, cut and reflected. I.F. Ischio-femoralis, seu obdurator externus. CAUD.IL.F. 1,2. Insertions of caud-ilio-femoralis (accessory femoro- caudal (1) and femoro-caudal (2)). CAUD.IL.FL. Caud-ilio-flexorius, cut, and origin reflected. IS.FL. Ischio-flexorius. P.I.F. Pub-ischio-femorales, seu adductores longus et magnus. The tendinous areas are dotted. X. Tendinous area on accessory femoro-caudal. F.T.E. Femoro-tibialis externus. (dicnemus in this respect, as in others, shows its wide diver- gence from the typical Limicoline condition. Of the others, Thinocorus, Hydrophasianus, and Rhynchea show what is probably the archecentric or generalised condition for birds, the presence of both muscles in a well-marked form. The remaining birds of the list show that the tendency of modification in the group is for the disappearance of the iliac portion (the ‘‘accessory ” of Garrod) ; and complete disappearance has been reached by Gallinago and Scolopax, two birds in other respects relatively highly specialised. Caud-ilio-flexorius (Semitendinosus and Accessory semitendi- nosus), Ischio-exorius (text-fig. 28).—In all these birds the three 1905. | ANATOMY OF LIMICOLINE BIRDS. 167 muscles are present, and, save that in Hydrophasianus and Himantopus the semitendinosus and its accessory or femoral head were very small, the conditions I did not find to differ from the generalised state found in Gruiform birds. Insertions of Caud-ilio-flexorius, Ischio-flexorius, and middle or posterior femoral head of Gastrocnemius—In a former communication to this Society (7) I described the differences that exist amongst Gruiform birds in this respect, and I grouped these divergences round four central types. The conditions in the Limicole are more uniform, and may be explained by comparison with the figure of the Otis type (7, text- fig. 83, p. 651). In all the birds the internal adductor muscle (Pub-ischio-femoralis internus) sends a strong slip to the middle head (internal femoral) of the gastrocnemius, or may be actually fused with it. The internal femoral head of the gastrocnemius at its Insertion to the femur is parallel with and ‘distad of the accessory or femoral attachment of the caud-ilio-flexorius; in Vanellus and Himantopus the edges of the two are in close contact, although they are not actually fused asin the Rallide. In all the other birds on my list they are quite asin Otis. From the raphe between the accessory and main portion of the caud-ilio- flexorius a strong fibrous band runs downwards fusing with the middle head of the gastrocnemius, whilst another band from the same point of origin runs across to be inserted into the tibia, under the tibial portion of the gastrocnemius, generally in association with the similar insertion of the ischio-flexorius. Gastrocnemius, external femoral head.—This is double in Vanellus, Himantopus, and Charadrius, single in all the others. The two heads unite before the muscle joins with the conjoined tibial and inner femoral portions. This recalls the similar doubling in Cariama, the three heads in Ofis and Hurypyga, and the enormous undivided head in AHeliornis. I have not in- formation as to the occurrence of a similar variation of the external head of the gastrocnemius in other groups. Ilio-fibularis (text-fig. 28, IL.FIB. (1) & (2))—This muscle, with its sling and connections, exhibits practically identical conditions, and these not differing from the state in the Grui- formes generally in all the birds on my list. The fleshy origin is unusually large. Pub-ischio-femorales (adductors).—These are both present in all the birds on the list. As I have mentioned above, the internal adductor has usually a strong connection with the middle head of the gastrocnemius. It is wider than the external adductor and shows traces of tendinous degeneration. Tibialis anticus and Solews.—These are present and normal in all the birds on the list, the tibialis anticus passing through a ligamentous ridge. ” Bectensor digitorum communis.—This has the normal arrange- ment and relations in all. Its tendon of insertion breaks up into two central slips for digit 3 and a single lateral slip at each 168 ON THE ANATOMY OF LIMICOLINE BIRDS. [May 16, side for digits 2 and 4 respectively, except in Glareola and Thinocorus, where it is a fan-shaped slip of fasciee common to the three digits, with the slightest trace of specialisation into tendons on the edges of the fan. Peroneus superficialis (with slip to perforated tendon of digit 3), Peroneus profundus.—These muscles are present, with one exception, in the normal or archecentric condition in all the birds on the list. The exception is the peroneus profundus in (dicnemus, in which bird it is practically absent, the absence being another point in which W@dicnemus differs from the Limicole and recalls many of the Gruiformes, such as Otis. Flexores perforantes et perforati.—These muscles and tendons, including the slip connecting the tendon of digit 3 with the corresponding tendon of the perforated flexor, all present a practically identical condition, which does not differ in any important respect from the condition in the majority of the Gruiformes. Flexores perforati.cThese muscles in all the birds on my list have the usual inter-relations and divide mto tendons for the three digits in customary fashion. The muscular mass has three heads: of these I have already described the ambiens head, which is similar throughout, except that there is no accessory ligament from the head of the fibula in Chionis. The external head is fleshy in Himantopus; it 1s small and tendinous in Chionis, Glareola, Thinocorus, Hydrophasianus, Charadrius, Rhynchea, and Gallinago. Itis absent in @dienemus and Scolopax. Pleaor profundus and Flexor longus hallucis.—In my communi- cation on the Gruiform birds I described various ways in which the tendons of these two muscles (which are similar in their origin in all the birds on my list) are united with one another and distributed to the toes. I suggested that probably the most primitive condition was such as is to be found in Hurypyga (7, text-fig. 85, VII), where the longus hallucis sends a slip to the hallux, and distad of this blends so completely with the profundus tendon that each tendon supplies each of the three digits. The condition in Chionis resembles this closely, except that, as in Rhinochetus, the hallucis tendon, after giving off its slip to the toe, is not so markedly spread out for the other toes. Mdicnemus shows a state practically identical with that of Otis; there is no great toe, and therefore no slip to it; the spreading out of the junction of the hallucis tendon with the profundus tendon has become obliterated. In Hydrophasianus the condition is exactly as in Hurypyga, except that, although there is a long great toe, there is no slip to it. In Rhynchea the condition is also the primitive one, except that the slip to the great toe comes off a considerable distance above the branching of the conjoined main tendons for the three other digits. In Scolopax the condition is similar to that in Rhynchea, but although there is a small great toe there is no slip to it, and the long junction of the two tendons is ossified. 1905. ] MR. R. I. POCOCK ON A HAINAN GIBBON. 169 Glareola and Thinocorus are like Scolopax, but have a slip to the great toe. Vanellus and Gallinago are exactly like Scolopas: ; Charadrius and Himantopus are also identical with it, except that there is no great toe. The conditions of these tendons in Limicole are similar and much alike, being not far removed from the condition that I take to be archecentric or primitive for Birds. But in the group there is a tendency to lose or reduce the great toe, and that loss or reduction has produced modifications which are similar in character and very easy to derive from the primitive type. SUMMARY. With the exception of Mdicnemus, the Limicoline birds examined, so far as relates to the characters dealt with, show a definite and coherent series of modifications. The group is moving, or has moved, along the same anatomical lines. The limits of its variations overlap in a special way the variations displayed by Gulls, and in a general way those exhibited by Gruiform birds. REFERENCES. (1) Bepparp, F. E.—The Structure and Classification of Birds. 1898. (2) Forprincer, M.—Untersuchungen zur Morphologie und Systematik der Vogel. 1888. (3) Gavow, H.—‘“ Aves” in Bronn’s Thierreich. (4) Mrroneii, P. Coatmers.—“ Quintocubitalism in the Wing of Birds.” Journ. Linn. Soc., Zool. vol. xxvii. p..237. (5) Mrrcnetn, P. Cuatmers.—“‘ Anatomy of Kingfishers.” The libris; SON p. 9%: (6) Mircuett, P. Coatmers.—“ On the Intestinal Tract of Birds.” Trans. Linn. Soc. ser. 2, Zool. vol. viii. p. 173. (7) MircHEett, P. Coatmers.—“ On the Anatomy of Gruiform Birds.” Proc. Zool. Soc. Lond. 1901, vol. ii. p. 629. 6. Observations upon a Female Specimen of the Hainan Gibbon (Hylobates hainanus), now living in the Society’s Gardens. By R. I. Pococs, F.L.S., F.Z.8., Super- intendent of the Gardens. [ Received May 16, 1905. ] (Plate V.*) Age at Maturity. On Jan. 26, 1904, the Society received on deposit a female specimen of the Hainan Gibbon, the property of Mr. E. H. de * For explanation of the Plate, see p. 180. 170 MR. R. I. POCOCK ON A HAINAN GIBBON. [May 16, St. Croix, who procured her in the island of Hainan on July 11th, 1897. She had thus been in captivity nearly six years and seven months. On the testimony of natives, her owner believed her to be about six weeks old at the time of capture; but since, as he affirms, she was already weaned and capable of fending for herself in the matter of food, it is probable that she was very much older than was supposed. On the assumption that she was at least six months old, it may be inferred that the beginning of 1897 was the approximate date of her birth. Menstruation set in at the end of the first week of December 1903; and taking this as the sign of maturity, coupled with the fact that she has not increased appreciably in size since her arrival in the Gardens, it may be assumed that she became adult when about seven years old. And in view of the close aftinity between the various species of Gibbons and the subequality in size of full-grown individuals, it may be further inferred that about seven years are required on an average for these animals to reach maturity * Menstruation. Very little appears to be known about the menstruation of Gibbons. In Chimpanzees, according to Dr. Keith (P. Z.8. 1899, p- 297), the discharge is sanguineous in colour, profuse, monthly in occurrence, and three days in duration. In our Hainan Gibbon it 1s also sanguineous, stains the floor of the cage, and, according to her keeper, Mansbridge, who also looked after the Society’s historic Chimpanzee ‘“ Sally,” is about the same in quantity relatively to the size of the animals as in that ape. The pudendal organs are always conspicuous by reason of their turgescence, and no very conspicuous change in their condition precedes the menstrual discharge. In this particular the Gibbon differs markedly from certain Cercopithecide (such as Baboons, Macaques, and Mangabeys), and also, to judge from published and verbal accounts, considerably, though to a lesser degree, from Chimpanzees. With the help of Mansbridge and Robertson, the two keepers of our Anthropoid Apes, I recorded the dates of the appearance of the discharge during the autumn, winter, and spring. The first noted was from Sept. 12 to 14, the second from Oct. 14 to 16, and the third from Noy. 19 to 21. During December the animal had a severe illness, beginning with an influenza cold and ending with diarrhoea, which was accompanied by extreme wasting and weak- ness. This illness extended over the time for menstruation, which did not appear in December. In January also there was no sign of it observable, although by the middle of that month she had apparently recovered her normal health. It is probable, I think, that the cessation for these two midwinter months was due to the illness. But it is by no means impossible that cessation during that time of the year is normal. The question can only be * Perhaps the Siamang (Symphalangus), which exceeds the other Gibbons in dimensions and differs from them in other respects, will be found to be an exception. 1905. | MR. R. I. POUOCK ON A HAINAN GIBBON. 171 decided by observing what happens in the ensuing winter, should the animal still be in the Gardens. Menstruation reappeared on Feb. 6 to 8, and has continued at tolerably regular monthly intervals since. Hence it may, I think, be laid down as an established fact that in Gibbons the interval between the men- strual discharges is a little over the calendar month and that the discharge continues for from two to three days. Determination of the Sex. When Mr. de St. Croix brought the specimen to the Gardens he informed me that she was a castrated male; and in support of his opinion drew my attention to the large size of the clitoris, which he most naturally mistook for the penis. The naked and turgid labia of the vulva he regarded as the unhealed wound caused by castration; and the menstrual discharge which first appeared in December of 19053, when the Ape was on her way to England, he attributed to normal bleeding induced by enforced sitting on the hard floor of her travelling-box. He also told me that it is commonly believed in Hainan that female specimens of the Gibbon are never brought to the coast and are practically unobtainable. There can be no doubt that this belief, coupled with the peniform clitoris of the Gibbon, misled Mr. de St. Croix as to the true sex of his animal, the castration of which, he admitted, he had not himself witnessed. And it seems probable that the belief itself is traceable to repeated mistakes on the part of Europeans in determining females as castrated males on account of the unusual length of the clitoris in these Apes as compared with the same organ in the Monkeys of the Old World generally. In this con- nection it is interesting to recall the fact that Dr. Harlan *, after dissection of the generative organs, described his specimen of Hylo- bates concolor as ‘an hermaphrodite Orang Outan.” It appears to me, however, that Lesson’s criticism of this opinion was perfectly justifiable and his decision that the specimen was an immature female undoubtedly correct. Pousargues, also, who evidently did not know Lesson’s paper, came independently to the same con- clusion, and stated that in the type of Hylobates nasutus, a young female, the clitoris was well developed and grooved below; and that the animal resembled in every particular, so far as the generative organs were concerned, the Gibbon determined as an hermaphrodite by Harlan. And since Harlan and two other doctors, presumably acquainted with human anatomy, who assisted at the dissection, were deceived as to the true sex of the specimen, in spite of the best possible opportunities for investigation, it is no wonder that the Europeans living in Hainan fall into a similar mistake. So far as can be seen, the clitoris of our Hainan Gibbon is like that of the specimen figured and described by Harlan, which resembled the penis of a Primate in a state of hypospadias. A * For Bibliography, see ixfra pp. 174-175. 172 MR. R. I. POCOCK ON A HAINAN GIBBON. [May 16, comparatively slight structural modification would convert such an organ into a closed tube for the passage of the urine—a fact perhaps of some significance in connection with the low position of the Gibbons in the Anthropomorphous series, seeing that in the Lemurs, the lowest of existing Primates, the clitoris is traversed by the urethral canal. Change of Color. IT am informed by Mr. de St. Croix that the young of both sexes of this species are alleged by the natives to be lighter- coloured at birth and for a short time afterwards than their parents. His animal, when first purchased, was a dark smoky grey, which, however, soon turned to black; and perfectly black she remained all the years she was in his possession. But within a few weeks of being brought to the Gardens she began to go grey, Mr. de St. Croix himself noticing a decided alteration in this respect when he visited her on March 8th, about six weeks after her arrival in London. During the spring and early summer the greyness progressed rapidly, but not quite uniformly all over the body. In midsummer, according to my notes, the head was black with a grey band extending on each side from the eyebrow over the ear; the beard was whitish and the nape of the neck blackish ; the greater part of the body was blackish grey, with a considerable quantity of blacker hair on the sides of the belly close to the thigh and a broad triangular black patch, narrower posteriorly, extending from the collar-bones on to the fore part of the belly and bordered on each side by a grey area paler in tone than the back; the thigh and upper arm were paler than the distal portion of the limbs. By this time she was not recognisable as the animal that reached the Gardens in January. Still the greyness continued to spread, the black pigment died out from the areas mentioned above, lasting longest upon the chest and the crown of the head. At this period she presented a decided similarity to the left-hand figure on the plate depicting H. pileatus Gray (P.Z.S8. 1861, p. 136, pl. xxi.), although the black pectoral area was smaller and the patch on the crown less sharply defined at the edges. In the early autumn she was a stone or silvery grey practically all over except for a black median band, fading away laterally and posteriorly, down the middle line of the head. At the present time (May 1905) she is brownish grey or silvery grey in colour, the tint varying according to the light. The black cap is still retained as a patch broadest and blackest between the ears, fading into brown upon the forehead and narrowing towards the nape of the neck, The hair on the chest has grown pale and thin, showing the blackish-grey tint of the underlying skin as a dark triangular shield. On the penultimate phalanges of the hands and feet the blackness of the hairs persists. The long hairs on the brows are also black. 1905. | MR. R. I. POCOCK ON A HAINAN GIBBON. 173 It is known that in some species of Gibbons, e. g. H. leuciscus, according to Mr. Hose, the individual variation in colour is con- siderable, like unto that which obtaims indeed in some Squirrels and Lemurs. But, so far as I am aware, it was not previously known that a given individual after reaching maturity may change in colour in the way exemplified by Mr. de St. Croix’s specimen. This change may be compared to that which takes place in the hair of the human head concomitantly as a rule with senescence or to that exhibited by some specimens of the Arctic Fox upon the approach of winter*. It is not accompanied by any replace- ment of coat, nor is it directly attributable to any change in the environment or to external agencies. The cause, whatever it may be, lies within the organism itself; it is constitutional or subjective, and as such may be distinguished by the term “ canescence,” from the decoloration or fading which is caused by exposure to sunlight or other bleaching agencies. In the case of the Hainan Gibbon it is important to note the coincidence between the appearance of menstruation and that of the colour-change. The former phenomenon began in December 1903, the latter about February 1904. Of the two specimens of this species in the British Museum (both of which are jet-black), one is only about two-thirds grown; the other, the type, as Mr. Oldfield Thomas states, not quite adult ; and since Mr. de St. Croix’s specimen is, on the contrary, full- grown, it might be inferred that it is characteristic of the species to change from black to grey upon reaching maturity. This, however, is not the case; for Mr. de St. Croix informs me that he was acquainted in the island of Hainan with another specimen, alleged to be a male, which was jet-black, like his own before coming to the Zoological Gardens, and had been in captivity sufficiently long to justify the belief that it was about twelve years old when he last saw it. Is the canescence, then, a matter of sex and exhibited only by mature females? The balance of evidence seems to be on the whole in favour of an affirmative reply to this question. For, apart from the change here recorded of the only adult female known, it must be remembered that Mr. Swinhoe, in his published account of all the information respecting the Hainan Gibbon he was able to gather, quotes from the Chinese gazetteer of the Kiung Shan district of the island a passage stating that the male is black and the female white (P.Z.8. 1870, p. 244, &e.). * There are two Arctic Foxes living in the Zoological Gardens at the present time. One remains dark-coated throughout the year; the other turns snow-white towards the winter. In both the winter coat, whether “white” or “blue,” is replaced in the summer by a darkish brown clothing of new hair, which is at its best in August, but becomes paler and loses to a large extent its richness of tint as it crows. In neither is there an autumn moult comparable in extent to that of the spring; and there is no doubt that in the animal which turns white the meta- morphosis is effected by the destruction of the pigment in the hairs themselves. This bears out Major Barrett-Hamilton’s statement as to what occurs in the Arctie Hare. 174 MR. R. I. POCOCK ON A HAINAN GIBBON. [May 16, The Name of the Species. The correct name for this species is still unsettled. The specimen now living in the Gardens is specifically identical with the type of H. hainanus Thos., and with the specimen previously exhibited in the Menagerie * and now in the British Museum, with both of which I have compared it. According to Matschie 7, however, hainanws 1s a synonym of concolor Harlan =. This opinion was based apparently upon the similarity in colour between the types of concolor and hainanus ; but it untortunately involves the assumption that the locality given for concolor, namely Borneo, is erroneous. It is also objectionable on the grounds that the hair of concolor was described as “thick, woolly, and frizzled.” The last two epithets are in no sense applicable to the hair of either of the three specimens of hainanus, comprising young and adult animals, available for examination. In these the hair, although thick, 1s smooth, depressed, relatively coarse, and quite unlike the hair of a young specimen of //. lar from Pahang, now in the Gardens, which is essentially rough and woolly ; and also equally unlike that of examples of H. agilis in the British Museum, which is beautifully silky and woolly. Furthermore, Trouessart § adopts for the species the name harlani, unlawfully proposed by Lesson || as a substitute for concolor Harl., alleging that concolor was first applied by Harlan in 1825 toa young specimen of H. (Symphalangus) syndactylus. _Concolor, therefore, falls as a synonym of syndactylus, and harlani comes in for the species described by Harlan in 1827, which Trouessart follows Matschie in identifying with hainanus. Trouessart, however, gives no reference to Harlan’s paper of 1825, and since T have failed to find it in the Royal Society’s Catalogue, and there is no suggestion in Harlan’s paper of 1827 (contained in a volume dated 1825), or in Lesson’s almost contemporaneous criticism of it, that the name concolor had been previously published, I must conclude that Trouessart has fallen into some error.. But in any case, since the specimen described by Harlan in 1827 as concolor and renamed harlani by Lesson in the same year and erroneously quoted as niger by Ogilby (P. Z. 8. 1840, p. 20) was definitely stated to have come from Borneo and to have had thick woolly frizzled hair, and since it 1s only known to have resembled the type of hatnanus in the matter of coloration, an admittedly variable feature in the genus and one in which it also resembles H. syndactylus 4, it 1s, in my opinion, premature to state without qualification that hainanus is a synonym of concolor. * Sclater, P. Z. S. 1892, p. 541. + SB. Ges. nat. Freunde Berlin, 1893, p. 211. t+ Jr. Acad. Sci. Philad. v. pt. 2, p. 231 (1827). Cat. Mamm. Suppl. 1904, p. 6. || Bull. Sei. Nat. xiii. p. 111 (1827). € Since Harlan states (loc. cit. p. 231) that concolor differs from HA. syndactylus and other species in being of a universal black colour, it is assumable that he did not know H. syndactylus. I do not, however, suggest that concolor is a synonym of syndactylus, because Harlan states that his specimen had no guttural sacs. Or 1905. | MR. R. I. POCOCK ON A HAINAN GIBBON. 175 Again, Pousargues * believed hainanus to be established upon a specimen of the same species as the type of H. nasutus, from Tonkin. This belief was also based upon resemblance in colour. Nothing else is known of the characters of nasutus except the alleged presence of a “ fine and delicate little nose,” whence the name was derived. But since hainanus is not distinguishable from other Gibbons by the fineness and delicacy of its nose, judgment on the synonymy suggested by Pousargues must be suspended until the type of nasutus has been re-examined and described. Trouessart, who may have seen the type, gives nasutus the rank of a subspecies of the Hainan form. No further justification need, I think, be sought for retaining the name hainanus for the subject-matter of these remarks. Description of the Species. Face, ears, palms of hands, soles of feet, and skin black, the face with a slightly brownish tinge; iris and exposed portion of eyeball blackish. Colour of hair either uniformly black, with shining tips, or grey, the roots of the hair being tinged with fawn or washed-out brown, their exposed portion shining with silver-grey lustre in reflected light, but of a more stone-grey in direct light. During the change from black to grey, the coloration is a mixture of the two, the black or the grey predominating according to the nearness of the time of observation to the incipience or com- pletion of the change. On the crown of the head a median longitudinal black patch with ill-defined edges and extending posteriorly as a narrow evanescent stripe persists. A few scanty hairs upon the penul- timate phalanx of the fingers and toes and the long hair on the brow also remain black. The hair on the body and limbs is longish, soft, and thick, but depressed and smooth. It is not woolly in the sense that the hair of our young Lar Gibbon is woolly, i. e. much resembling a Sheep’s fleece ; nor does ib exhibit the fine and silky woolliness of the skin of 1. agilis in the British Museum. On the forehead. and crown of the head the hair is shorter, fine, and close, and in the living specimen grows some- what &@ la Pompadour, being erect on the crown and almost porrect on the forehead, so that the head has the appearance of being very much higher than in our living example of the Hoolock (Hf. hoolock) and in adult skins of H. lar, H. pileatus, and H. leuciscus in the British Museum, in which the hair lies smoothly backwards. The difference may be briefly expressed by saying that in our Hainan Gibbon the hair looks as if it had been brushed up, whereas in the others it looks as if it had been * Bull. Mus. Paris, 1900, p. 272. Pousargues gave A. Milne-Edwards the credit of naming nasutus. Milne-Edwards, however, published no description of the species when the name was quoted (Le Naturaliste, 1884, p. 497). Hence if seems that Kunckel d’Herculais, who first associated the name with detinite characters, must be regarded as the author (Science et Nat. ii. no. 33, p. 86, 1884). 176 MR. R. I. POCOCK ON A HAINAN GIBBON. { May 16, brushed down. In the two skins of H. hainanus in the British Museum, however, the hair on the crown is not so markedly up- standing, nor so long, as in the living example. On the cheeks the direction of the hairs is, generally speaking, upwards. On the upper surface or back of the hand and on the corresponding surface of the forearm the points of the hairs lie towards the ulnar side of the limb, assuming a more and more elbowward direction as that joint is approached. On the palmar and radial side of the forearm, on the contrary, the hairs point for the most part towards the wrist. The palmar surface, however, is marked by a crest formed by the meeting of the two opposing streams of hair, the crest extending obliquely from the radial side of the elbow to the ulnar side of the wrist, the hairs on the ulnar side of it being directed proximally, those on the radial side distally. On the body the hairs lie backwards, except on the belly, where they incline towards the middle line and form a median longitudinal crest where the two streams meet. This is the area against which the inner sides of the thighs are pressed when the Gibbon is in a sitting posture. On the outer side of the thigh the direction of the hair is upwards (proximad) and backwards, below the knee it is downwards (distad). Additional Notes. The voice of our Hainan Gibbon is quite different from that of the Hoolock. It is a high-pitched trill all on the same note, and shriller even than the high note of the Hoolock’s cry. It consists of from about three to six distinct cries repeated in very rapid succession, suggesting almost production by vibration of the tongue, although, as a matter of fact, I believe the lips alone are instrumental in producing the effect. There is then a momentary pause, after which the cry is repeated. It may perhaps be represented in the following way:—hoo hoo hoo hoo—hoo hoo hoo—hoo hoo hoo hoo hoo—ke. The Hoolock, on the contrary, cries as follows:—hah, hoo, hah, hoo, hah, hah, hoo, hah. The “hoo” is on a lower note than the “hah,” with which the cry frequently ends. The ordinary expression of anger or remonstrance in the Hainan Gibbon is a prolonged and guttural grunt, which is repeated rapidly and often, and frequently interspersed with a kind of warble when the excitement rises. Both the Hoolock and the Lar Gibbon in the Gardens drink habitually by dipping the back of the hand and knuckles into the dish and licking the water off. They do not scoop it up, in the strict sense of the word, at all. Hence Col. Tickell’s generalisation to the effect that in its habit of scooping up water in its hands the Lar Gibbon differs from the Hoolock, which applies its lips directly to the fluid, is contradicted on both counts by our specimens of these species. The Hainan Gibbon, on the contrary, almost invariably drinks direct with her mouth, only very rarely using her left hand for the purpose. It is possible she may have 1905. | MR. R. I. POCOCK ON A HAINAN GIBBON. 177 abandoned the habit of employing the hand at the time when an injury deprived her of the use of her right arm. And since the. left is frequently occupied in supporting herself upon the bars or perches in the cage, she has no hand available for the purpose of drinking without quitting her hold. This method of hand-drinking, probably common to all Gibbons, may have arisen in connection with thei arboreal life. To avoid descending to the ground, they would naturally lick the rain-drops off the leaves near by, and their great stretch of arm would enable them to wipe the water off foliage hanging beyond reach of the mouth, the hairy back of the hand being clearly more fitted for the purpose than the smooth palm. In connection with this habit, it is interesting to recall the story told by Duvaucel of female Gibbons carrying their young to the waterside and washing their faces with their hands. This alleged proceeding, presumably witnessed in the jungle, can hardly, I think, be accepted without confirmation, on account of the absence of any obvious reason for the ablutions. If the young Gibbons of which the tale is told were hanging, as is their wont, to the breasts of the mothers, the action of hand-drinking by the latter might very easily be mistaken at a distance for the face-washing. Amongst ‘“ quadrumanous” Primates the Gibbons have no equals in proficiency in the use of the arms for arboreal and the legs for terrestrial progression. Moreover, within the limits of the entire order, they are only surpassed in bipedal activity by the specialised biped Man. Although able to stand and walk to a very limited extent, Monkeys are essentially quadrupedal and employ then arms and legs to an approximately equal extent in traversing the level ground, scaling rocks, or climbing trees. Generally speaking, the most active climbers are long non-prehensile tailed species, such as the Mangabeys*, in which the tail actsas a balancer, like the pole of a tight-rope dancer. Monkeys of this kind leap with great pre- cision and strength, and pass with speed from branch to branch in virtue of the great propelling power in their hind-quarters. ‘They ave specialised for that manner of progression, which only differs in degree of perfection from that of other Monkeys and Lemurs as a whole. The method, however, is entirely distinct from that practised by the Gibbons, which swing from branch to branch, with the legs tucked up out of harm’s way against the body, the motor power lying exclusively in the arms. Both groups have been specialised for arboreal progression, but along totally different lines; and it is as difficult to believe that the Gibbons, expert gymnasts though they be, have been derived from active long-tailed climbers, like the Mangabeys or Langurs for instance, as it is to believe that the tail-swimming Cetaceans have been derived from forms. like the flipper-swimming Seals. * T have never yet seen the Baboon or Macaque that could catch a Mangabey single-handed, given equal conditions as to health and age, in a large-sized cage. Proc. Zoou. Soc.—1905, Vou. Il. No. XII. 12 178 MR. R. I. POCOCK ON A HAINAN GIBBON, [May 16, This conclusion respecting the descent of the Gibbons may be inferred from their habits alone, quite apart from structure. If the Anthropoid Apes be ranged in series according to proficiency in bipedal locomotion, the order will be (1) Gibbons, (2) Gorillas, (3) Chimpanzees, (4) Orangs. Gibbons not only stand erect and habitually walk without putting the hands to the ground; they can even run with astonishing speed, a speed indeed comparable to that of Man, allowance being made for difference in size. Like Man they race away when scared; and, unlike the other Anthropoid Apes, they do not use their arms as crutches. Sometimes also, but rarely, they leap over the ground with both feet together *. Gorillas can stand and walk upright, but not with the ease of Gibbons, and it may be doubted if they ever run erect or leap, i.e. progress with both feet off the ground at one time; and they probably never run from danger, standing upright, as Man and Gibbons do. Their usual walk is quadrupedal. Chimpanzees, too, are essentially quadrupedal; and under ordinary conditions, and when in perfect health, almost always get over the ground on “all fours,” like a Baboon or Rhesus. In this respect, indeed, they more resemble the Cercopithecoid Monkeys than does any other Anthropoid Ape; and they are able to cover the ground with much greater speed than either Gorillas or Orang- Utans ; but I am unable to say if their quadrupedal method is so fast as the bipedal method of Gibbons. Like Baboons, they can stand erect and walk to a certain extent, but not with the facility of Gorillas. The gait of young Orang-Utans may be described as a clumsy quadrupedal shuffle, J never saw one stand unsupported by the arms. Weakness of leg and weight of body make exclusively bipedal action, if not an impossibility, at least so great an effort that it may be doubted if it is ever resorted to. Their whole organisation suggests unfitness for terrestrial locomotion. Thus, if the Apes be classified according to their quadrupedal activity on the ground, they will stand:—(1) Chimpanzees, (2) Gorillas, (3) Orangs, (4) Gibbons. It is interesting to compare this series with one based upon dexterity in climbing and addiction to arboreal life. It is: (1) Gibbons, (2) Orangs, (3) Chimpanzees, (4) Gorillas. The Gibbons stand quite alone both in method and expertness; the others differ inter se merely in degree. The foregoing results may be briefly summarised as follows :— The Gibbons are the most expert climbers and bipedal walkers, the least expert quadrupedal walkers. The Orangs rank second in climbing, third in quadrupedal and fourth and last in bipedal activity. The Gorillas take fourth place in climbing, second in bipedal and second in quadrupedal activity. The Chimpanzees * These and the following statements and reflections are based upon my own observations of the Anthropoid Apes that have come under my notice in the Society’s Gardens. 1905. | MR. R. I, POCOCK ON A HAINAN GIBBON, 179 stand third in climbing, third in bipedal and first in quadrupedal powers. Since, therefore, the action of Monkeys, whether Cercopithecide, Cebide, or Hapalide, and of Lemurs is essentially quadrupedal, the fore and hind limbs being used to an approximately equal extent, both .in terrestrial and arboreal locomotion, it may be inferred that the Chimpanzees have departed least in these respects from the primitive Primate stock; the Gorillas a little more in the line of bipedal erection and, concomitantly, loss of climbing power; the Orangs still more in the direction of loss of terrestrial activity and increase of arboreal expertness; the Gibbons most of all in the line of bipedal activity, dexterity in hand-climbing, and loss of quadrupedal power. This serial arrangement of the Apes is the exact opposite of the one prevalent in text-books, where the order adopted is based upon structure with Man placed first as the standard for comparison. It suggests that for the origin of Gibbons we must look not to forms resembling any known Cercopithecoid type, but to forms which had already acquired the Simiine or Authropo- morphine characteristics and had either lost or never learnt the method and skill in climbing found in the former group. They may have started from a type somewhat on a level with the Chimpanzees with respect to terrestrial and arboreal activity ; and to swing with greater facility from tree to tree and to obviate the risk of injury in ease of a fall, it is highly probable that they have become dwarfed in stature and grown lighter in build. Their muscularity and length of arm, slightness of body and strength of leg, all factors of importance in enabling them to traverse the jungle and, in case of 2 miss or a breaking branch, to drop lightly to the ground and run to the nearest tree for safety, were probably perfected concomitantly. That Gibbons are able to drop with safety a considerable distance is substantiated by the fact that Mr. de St. Croix has seen his specimen come to the ground without injury from a height of about 20 feet. When leaping to the ground Gibbons swiftly draw up the knees as the feet touch, exactly as a man does under similar circumstances, to break the shock. Another interesting feature connected with the habits of the Anthropoid Apes is the size of their ears. T have already suggested that the difference in size between the ears of the Orang and those of the Chimpanzee may be connected with the difference of habits of the two animals. ‘The Orang lives a more arboreal and. therefore a safer life than the Chimpanzee, which requires quick heaving to enable it to escape to the trees when feeding on the ground *, Gibbons also, which have relatively large ears, need auditory acuteness for the same purpose as Chimpanzees. This explanation, however, is not complete and appears at first sight to be contradicted by the case of the Gorillas, which have small ears * ‘Nature,’ Oct. 11th, p. 585 (1900). © jake 180 ; MR. OLDFIELD THOMAS ON [June 6, and yet are less arboreal in habit than other Anthropoids. It must be remembered, however, that they are far more capable of self-defence and much less liable to attack and therefore need less keenness of ear as an aid inavoiding enemies. In this connection it is important to note that of the two Apes inhabiting W. Africa, namely the Chimpanzee and the Gorilla, and of the two inhabiting the Hast Indies, namely, the Orang and the Gibbon, the larger and stronger has in each case small insignificant ears and the smaller and weaker large eaxs. EXPLANATION OF PLATE V. Hainan Gibbon (Hylobates hainanus), from the female specimen now living in the Society’s Menagerie. ‘The lower figure, taken from an obscure photograph by Mr. W. P. Dando, F'.Z.S, represents the Ape when she first came to the Gardens. The upper figure, modified from a photograph of another Gibbon, shows her as she has been since the change of colour took place. June 6, 1905. Dr, Henry Woopwarp, F.R.S., Vice-President, in the Chair. The Secretary read the following report on the additions that had been made to the Society’s Menagerie in May 1905 :— The registered additions to the Society’s Menagerie during the month of May were 367 in number. Of these 174 were acquired by presentation and 27 by purchase, 129 were received on deposit, 25 by exchange, and 12 were born in the Gardens. The total number of departures during the same period, by death and removals, was 185. Amongst the additions special attention may be directed to :— A Crowned Duiker (Cephalophus coronatus) from West Africa : deposited on May Ist. A Maxwell’s Duiker (Cephalophus maawelli) from W. Africa : presented by Lieut.-Col. Bartlett, R.A.M.C., on May 16th. A Nepalese Hornbill (Aceros nepalensis) from the Himalayas : received in exchange on May 18th. Two Sulphur-breasted Toucans (hamphastos carinatus) : purchased on May 13th and May 23rd respectively. Mr. Oldfield Thomas, F.R.S., exhibited a specimen of a Bush- buck which had been obtained by Mr. C. W. Haywood in British Kast Africa and which appeared to represent a new species of the group. It was described as follows :— 1905. ] A NEW BUSH-BUCK. 181 TRAGELAPHUS HAYWOooD! Thos.* Thos. Abstr. P. Z. 8. No. 21, p. 9, June 13, 1905. A large heavily-built member of the group of small species without a definite short-haired collar. Under surface darker than upper. Fur comparatively coarse and long throughout, the hairs of the back 35-40 mm. in length. General colour very dark, the nape black; the fore-quarters blackish brown (near ‘‘seal- brown”), passing into dark reddish brown (“ vandyke-brown”) on the middle back and deeper rufous (dark “ tawny”) on the rump. Sides gradually darkening downwards to the wholly black belly. Dorsal crest black as far as the withers, then whitish mixed with some black hairs. Three inconspicuous transverse whitish stripes on each side. No longitudinal bands, but a few white spots on the sides of the rump. Shoulders and proximal part of limbs deep black, succeeded by tawny below. (Feet unfortunately lost in the type.) Top of muzzle nearly black, with prominent interorbital whitish streaks nearly touching each other in the middle line. Forehead and crown deep ferruginous. Cheeks tawny ochraceous. ‘Two white spots on each side behind and below the eyes. Ears thinly haired, dull tawny brown with blackish edges; hairs of inner surface white. Chin and interramia white and a large throat-spot duller white; between this and the white chest-band the throat was glossy blackish, mixed with some tawny hairs. White axillary and inguinal patches present. Tail dark tawny, white below. Skull very large and heavy for one of the smaller members of the genus, much larger than in 7’. scriptus or sylvaticus. Median alatal notch rather farther forward than the lateral ones. Palatal foramina comparatively long. Horns also very powerful, thick and strongly ridged, much finer than those of any of the allied forms. skull dimensions of type :— Greatest length 265 mm.; basal length 247; greatest breadth 112; muzzle to orbit 154; muzzle to front of p* 77; length of palatal foramina 36. Length of upper tooth-series 72, of three upper premolars 31. Horns: length mm straight line 400; on anterior ridge 470; greatest basal diameter 59; basal circumference 171. Hab. Nyeri, Kenya District, British Hast Africa, Altitude 6000 feet. Type. Full-grown male. B.M. No. 5.5.16.3. Collected and presented by C. W. Haywood, Esq. - Mr. Oscar Neumann ~ had sorted the smaller species of Tragelaphus into two groups, characterised by the presence or absence of the peculiar collar of short hairs which had been so * [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. | + SB. Ges. nat. Fr. Berl. 1902, p. 98. 182 MR. OLDFIELD THOMAS ON [June 6, often noticed in Bushbucks, and which was evidently of definite systematic value. But, as Dr. Hinar Lonnberg had shown*, several of Mr. Neumann’s allocations were incorrect—certainly the Cape sylvaticus had a short-haired collar, and fell mto the seriptus eroup, while the Nilotie bor had a well-haired neck. Mr. Neumann had also stated that the forms with well-haired necks known to him did not have a darker underside, but since his paper was written Lonnberg’s Anutson?z and the present animal had both proved to present the combination of a hairy neck and a black belly. This combination therefore distinguished hayiwoodi from any of Mr. Neumann’s species, while from the Cameroon knutsont it was separated by its whitish dorsal crest, less numerous spotting, and other detailed characteristics. Tn company with this handsome animal, which Mr. Thomas had much pleasure in naming after its discoverer, Mr. Haywood had sent home to the National Museum two immature skins, with skeletons, of the recently described Forest-Pig (/Zylocherus meinertzhageni). It was hoped that an adult specimen suitable for mounting would soon be obtained, and_ this Mr. Thomas looked forward to exhibiting to the Society in due course. Mr. Oldfield Thomas also exhibited a series of Mammals and Birds from Japan as the first-fruits of an exploration of the islands of Eastern Asia conducted for the furtherance of science by the President, the Duke of Bedford, K.G., in order to show his Grace’s sympathy with the technical side of the Society’s work. The specimens obtained during this exploration would be laid before the Society from time to time, and papers would be read on them by various specialists, after which his Grace proposed to present them to the National Museum. Mr. Thomas commented on the immense value such a systematic exploration would be to science if it were carried on for some time, and instanced the revolution in our knowledge of the mammals of South Africa—a region supposed to be well- known—which had been effected by the similar exploration conducted by Mr. C. D. Rudd. The Japanese collection had been made by Mr. Malcolm P. Anderson, who had already proved his powers both during the Stone Expedition to Alaska and by the collections he had made in California. Of the specimens now laid before the meeting Mr. Thomas drew attention toa fine Marten, which appeared to be different from the ordinary Japanese Marten (M/ustela melampus) and which, as the first new mammal discovered on the expedition, he proposed to name in honour of the President :— * Arkiv for Zoologi, Stockholm, 1. 15 ( 905). 1905. ] MAMMALS AND BIRDS FROM JAPAN. =). MUusTELA MELAMPUS BEDFORDI Thos.* Thos. Abstr. P. Z. 8S. No. 21, p. 10, June 13, 1905. Size as in true melampus, or slightly larger. General colour above, in winter pelage, near “ isabella,” but rather darker and with an olivaceous tone, nearer to the yellowish brown of MM. m. isuensis * than to the golden yellow of melampus. Wool-hairs of back brown at base, then dark yellowish. Long hairs brown. Muzzle dark chocolate-brown, passing backwards, on the crown, into silvery greyish. Ears whitish both externally and internally. Nape more yellow than back. Sides of neck brilliantly yellow (* deep chrome ”), sharply contrasted with the upper colour along a line halfway up the neck, and in continuation with the deep orange ochraceous of the chest-patch. Lips pale brown, lighter than the top of the muzzle; sharply defined from the whitish interramia, which in turn passes without line of demarcation into the orange of the throat and chest. Belly brown, not unlike back, the throat-patch extending to the sternum and continued in some specimens as an irregular line of spots to the inguinal region. Limbs deep brownish black from halfway down the forearms and on the hind feet. Tail pale brown for the greater part of its length, the wnderfur dull yellowish as on the body ; tip sharply contrasted yellowish or cream-colour, forming a conspicuous terminal tuft. Skull as in ésuensis, slightly larger than in melampus so far as material for comparison existed, Dimensions of the type, measured in the flesh :—Head and body 425 mm.; tail 220; hind foot 87; ear 40. Skull—greatest length 84; basal length 75; zygomatic breadth 48; interorbital breadth 20; mastoid breadth 37°5; palatal length 42; length of upper p* on outer edge 9°5, Hab. Washikaguchi, Nava District, E. of Osaka, Southern Central Hondo, Japan. Type. Adult male. B.M. No, 5.5.30.3. Original number 123. Collected 13 January 1905 by Malcolm P. Anderson, and pre- sented by the Duke of Bedford. Four specimens. This very handsome Marten is conspicuously different from the yellowish M. melampus, and is curiously more similar in general colour to the I. m. tswensis of the Tsu-shima Islands. From both, however, it is readily distinguished by its brilliant yellowish throat and neck patches and its contrasted tail-tip. Mr. R. I. Pocock, F.L.S., the Superintendent of the Gardens, exhibited a female specimen of the Jamaican Scorpion, Cenirurus * [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.—EpD1To0R. | + Thos, Ann. Mag. N. H. (6) xix. p. 161 (1897). 184 LT.-COL, ©. DELME-RADCLIFFE ON THE NATURAL [June 6, insulanus, carrying its young on its back. The specimen had been presented to the Society by Mr. Henry Munt, F.Z.8. Dr. P. Chalmers Mitchell, the Secretary to the Society, read a paper, illustrated by lantern-shides, entitled “On the Intestinal Tract of Mammals.” This paper will be published entire in the ‘ Transactions.’ The following papers were read :— 1. Rough Notes on the Natural History of the Country West of Lake Victoria Nyanza. By Li.-Col. C. Deutm&- Rapcuirrse, M.V.O., F.Z.8. | Received June 6, 1905. ] These notes contain the general results of my observations on the Natural History of the region traversed by the Anglo-German Boundary Commission in the years 1902-4. Memoirs dealing more exactly with the collections that were made have already appeared in the ‘ Proceedings’ of the Zoological Society (P.Z.8. 1904, vol. i. pp. 371, 459) and ‘ The Ibis’ (1905, p. 199.). MAMMALS. Beginning with the larger mammals in the country under discussion, it may be stated that Elephants appear periodically in the swamps and forest near the mouth of the Kagera River on the northern side. These elephants stray in this direction, probably, at a time when it is dry in the interior. They come, no doubt, from the herds in northern Ankole and Toru. At no other point were traces of elephants seen except one single track going from north to south from the Koki hills towards the Busenya forest. In the west, a few elephants were noticed near the shores of Lake Albert Edward, also probably stragglers from the herds further north. There was no evidence of elephants crossing from south to north, or vice versd, along the Ist parallel south latitude. it may perhaps be assumed that the herds of elephants reported by H. 8. Grogan and other travellers in the Mfumbiro district belong to the forest-regions of the west. ‘The herds of elephants on the east of Lake Albert Edward and Ruwenzori probably do not wander into the Congo forests. It has been noticed that the elephants to the west of the great line indicated by Lake Tanganyika, Lake: Kivu, Lake Albert Edward, Lake Albert, &c., and the Nile differ in many particulars from those lying to the east of this line. At the same time, it must be remembered that large herds of elephants are in the habit of crossing the Nile to 1905.] ISTORY OF THE UGANDA ANGLO-GERMAN BOUNDARY. 185 the north of Lake Albert, and there seems no reason why they should not extend their wanderings into the Congo forests, although so far observation tends to show that these herds find their way back again, as a rule, to the countries east of the Nile. Hippopotami are not very numerous in the Victoria Nyanza near the mouth of the Kagera. The locality does not seem very well suited to them. In the Kagera River itself there are more, and parts of the river are infested by a number of very savage brutes that make navigation in canoes or small boats extremely dangerous. Lt. Weiss, of the German Commission, was repeatedly attacked when in a very large canoe. He was almost upset—one man was dragged out by the arm, but escaped. Finally his crew refused to go on and ran away with their paddles. The actual number of hippopotami cannot be considered large in comparison with the huge herds in the Nile north of Lake Albert. Probably in the great swamps of the Kagera, considerably to the south of the area traversed by the Boundary Commission, the hippopotam1 ave much more numerous. ‘The specimens secured in the Kagera were decidedly inferior in size and in development of ivory to those of the Nile. Rhinoceroses are extremely numerous on the right bank of the Kagera, especially in Karagwe. The number of these animals is quite remarkable, and, according to accounts received, they are to be met with in even greater numbers a little further south. It is a curious fact that no rhinoceroses are to be found on the lett bank of the Kagera. All those seen belonged to the common black African type. Stories were current of the existence of the White Rhinoceros on the right bank of the Kagera, but these rumours require confirmation. ‘The rhinoceroses appear to have no hesitation in frequenting the extremely steep and difficult hills of Karagwe. Their tracks and signs were seen up and down hills and on ridges which appeared more adapted to the habits of klipspringers or goats than of such bulky animals as rhinoceroses. In the virgin forest west of the lake near the mouth of the Kagera, in the swampy and open forest east of Koki, and in the Busenyi forest west of the Gambaizi group of hills, several herds of Buffaloes are to be found. These buffaloes are of a very interesting, new, large variety. They are, perhaps, the largest buffaloes in existence. In all, in the district referred to, there may be 400 or 500 buffaloes, and as their numbers are not likely to be interfered with, except by men armed with rifles, they may be considered to be firmly established again after the devastation caused by the great cattle-plague of some ten years back. In Bukanga the buffaloes wander in search of young grass, after the fires, as far as the hills of Ankole and Koki, from the forests which form their strongholds. There is one disadvantage, however, connected with the presence of the Buffaloes, of the Bland, and perhaps of other Antelopes. This is the tsetse-fly, and it is to be feared that as long as large herds of buffaloes and the greater antelopes exist, so long will the tsetse-fly make it 186 LT.-COL, C. DELM£-RADCLIFFE ON THE NATURAL [June 6, impossible for domestic cattle and horses to live in the same part of the country. I myself lost an Arab horse I had had for six years in Africa and was very fond of. He was bitten by tsetse- fly in Bukanga. Eland were met with at two points in Bukanga—near the Nyakafunzo swamp, and in the districts known as Mpororo and Rushenyi. In Bukanga there were herds amounting to, perhaps, 200 animals, and the uninhabited country surrounding the Nyakafunzo swamp seemed admirably suited to then needs. They were considerably preyed upon, unfortunately, by natives, who organised hunting-parties into this district both from the British and the German side. Still more unfortunately, the natives are sometimes armed with rifles. The result could be seen in many wounded animals observed from time to time, and in dead bodies found with bullets in them. Lions also take toll of the elands, but the natural decrease due to this cause is nothing compared to the damage inflicted by natives with firearms. Further west a herd of considerably over 300 elands was seen, and this, probably, is only an outlier of still greater herds in the open country further south. Jt seems, therefore, that this country is abundantly supplied at present with representatives of this magnificent antelope, which, I believe, might be made of great economic value. The meat is equal to the best English beef, and a bull eland weighs about 17 ewt. Zebras occurred coincidentally with the Hland in Bukanga, and they number, perhaps, 400 individuals. In Rushenyi another very large herd of zebras was seen ; and it may be remarked that in the Rushenyi herd a single zebra was seen almost entirely pure white in colour, a few stripes only appeared on the neck and hind-quarters. Another small herd of zebras, amounting, perhaps, to 150 individuals, was seen in the plains in southern Ruampara, on the left bank of the Kagera, just north of the point where the river turns from the south to east. Roan Antelope were encountered, a few at a time, in Bukanga, in the narrow valley of the Kagera, and in south-west Ruampara north of the bend of the Kagera just referred to. They were occasionally met with in Rushenyi and Mpororo, and appeared more numerous in Karagwe, where for some reason there appeared to be no Hland, no Zebras, and no Damaliscus. These last were the common hartebeeste throughout the area west of the lake. In Bukanga, Damaliscus were very numerous. The number in this part may be estimated at 1000 individuals. No other variety of hartebeeste made its appearance: 200 or 300 individuals were found with the herd of zebras in south-western Ruampara, and in Rushenyi and Mpororo the Damaliscus hartebeestes are very numerous. The Nile Valley variety of Water-buck (Kobus defassa) 1s common in Bukanga, and may be mét with in herds up to a dozen or fifteen individuals. They also appear fairly plentiful through- out the valley of the Kagera and in western Ruampara, but 1905.| HISTORY OF THE UGANDA ANGLO-GERMAN BOUNDARY. 187 apparently not further west. These water-buck have fine heads, as a rule considerably larger than A. ellipsiprymnus in Kast Africa, although the heads are not nearly so big as are found in the Semliki Valley. Jn the swamps near the mouth of the Kagera, on the shores of the lake, and on the islands of the Sesse group, Limnotragus spekei was fairly common, This animal, owing to its nocturnal and swamp-loving habits, is of course seen extremely rarely, but it is frequently hunted by natives with nets and packs of dogs. The horns are often to be met with in possession of natives. There is at present a doubt whether more than one species of this antelope is not found in the same district. In Bukanga, especially about the Nyakafunzo swamp and to the south of it, large herds of Mpala (Zpyceros melampus) may be met with. These beautiful antelopes are to be seen in herds of 200 or 300, and in the district referred to perhaps 1500 individuals exist at the present time. They were met with at no other point throughout the country traversed. Very common, although occurring only in ones and twos at a time, was a species of Reed-Buck (Cervicapra? sp.). This antelope was chiefly confined to the low-lying grassy country in Bukanga, along the banks of the Kagera, and in Ruampara. Another very common antelope was the Ovibi (Owrebia montana). This little antelope appeared almost everywhere on the low ground in the mountains, except in the highly cultivated parts of Ankole and the mountains in the west. In general, it may be said that no antelopes or game animals of any description were seen in the Ruchigga mountains and their northern and eastern extensions.- Bush-buck and Harnessed Antelope were seen at rare intervals in the valley of the Kagera. The latter appeared occasionally at the edges of the dense forests near the mouth of the Kagera, and in one or two places in the narrow valley of the Kagera between the mountains. Bush-buck were occasionally seen in .the Koki hills and the mountains of Ruampara, where the deep guilies choked with vegetation afforded them shelter, and the open grassy hillsides excellent feeding-grounds. On the steep hills of Ankole and Karagwe, Klipspringers were common wherever the ground suited them. The form in this country shows some differences when compared with the klip- springers of other parts of Africa, and may prove to be an intermediate variety. Of Monkeys, Colobus guereza was seen in the forests near the lake, The common grey African monkey was also observed in many places, and an interesting species, Cercocebus aterrimus, was also seen in the dense forests near the lake and in the dense forests round Minziro, The last-named monkey looks almost black and is very shy. Its ery is very loud and peculiar, reminding one slightly of the ery af the Chimpanzee. Baboons ave common, especially in the mountains in Ankole and Karagwe. 188 LT.-COL, C, DELME-RADCLIFFE ON THE NATURAL [June 6, Wart-Hogs (Phacocherus cethiopicus) were occasionally to be met with all along the valley of the Kagera, though nowhere very numerous. Aardvark were present, though of course never seen unless dug for, Their holes, however, were found in all directions in the low-lying country, and they are probably fairly common. Of beasts of prey Lions are fairly common in Bukanga, in the neighbourhood of the Nyakafunzo swamp. They also appear in western Ruampara and in Rushenyi and Mpororo. In general terms, it may be said that they are to be found wherever large herds of zebras and antelopes exist. In Bukanga, however, it appears that they have taken to man-eating fairly extensively. The natives in this part of the world have a wholesome dread of them, and during the short time the Boundary Commission was at work in Bukanga repeated instances occurred of lions attacking human beings. Leopards are also found throughout the whole area under discussion except the extreme western portion. Although they live principally on the small antelopes, monkeys, guinea-fowl, &e., they also take toll of the natives’ goats, &c., and thus become sometimes a great nuisance. At Mulema camp, for instance, a leopard took goats from one hut or the other almost every night for a month, and when Captain Laughlin, Dr. Bagshawe, and Mr. Doggett endevoured to kill him at the natives’ request, he wounded, more or less seriously, no less than thirteen men before being finally despatched. Cheetahs apparently do not exist in this part of the country west of the lake. Serval Cats were occasionally met with, and a smaller grey, rather long-tailed Wild Cat. Hyzenas appeared occasionally, but may be said to be rare. They were of the usual spotted variety. Otters are common in the lake. Two forms were met with, one very large, the other smaller. These two are stated also to be common in Lake Kivu. Among smaller mammals, interesting species were a Peacilogale doggetti, an extremely handsome, large, striped Stoat; Tatera fallax; Procavia bettoni: these three being new species. Another extremely interesting animal was Herpestes galera robustus, a fish-eating Mongoose. In all about 180 specimens of mammals were collected, and a large number have been described in the ‘ Proceedings’ of the Zoological Society of London, the most interesting being the new Buffalo (Bubalus caffer radcliffe), Pecilogale doggetti, Tatera fallaa and Procavia bettont. There is no doubt that a scientific inves- tigation would disclose a much larger number of small mammals than were secured for the collections of the Boundary Com- mission. All the region west of the lake abounds with species of great scientific interest, the interest increasing the further west one goes, and it is a matter for the greatest regret that a collection could not be made in the neighbourhood of the Mtumbiro Mountains. 1905.] HISTORY OF THE UGANDA ANGLO-GERMAN BOUNDARY. 189 Bibs. Lake Victoria is a disappointing sheet of water in bird-life as in fishes. Birds are of course present, but not in the vast numbers so extended a sheet of water in the heart of Africa might lead a naturalist to expect. The reason is probably to be found in the fact that the food-supply in the lake is very deficient for birds and fishes alike—for many species of birds in consequence of the poverty in fish. The White-headed Fish-Hagle (Haliaétus albicilla) is fairly common all round the lake-shore and up the Kagera River. This bird is invariably found in pairs, and appears to divide the districts into beats, each containing its pair of fish-eagles in possession, Their cheerful squalling, as described by Sir Harry Johnston, is one of the most familiar sounds near African river and lake. An Osprey may also occasionally be seen on Lake Victoria seizing fish on the surface of the water ina manner peculiar to its kind A species of Plotus is fairly common, and may be seen in small congregations at certain points where the ambatch or rocks afford a convenient spot for perching and hanging their wings out to dry. With them may also be seen the large Cormorants, which here and there form communities numbering several hundreds. The northern end of the island Usuwgwe and the small rocky Mwasambwa Islands and Dumo Point are favourite haunts of all these birds. A large Gull, resembling the black-backed gull at home when on wing, is also not uncommon close in-shore and especially in the rivers. The Pied Kingfisher is fairly common inland. The two varieties of brilliant-hued Kingfishers appear to forsake theix occupation of fishing to devote themselves entirely to the capture of insects. Herons are fairly plentiful in the swamps and at the edge of the lake. The most conspicuous amongst these is the Goliath Heron, a bird whose immense span of wing can be fully appreciated when, disturbed by a canoe, he flaps slowly across the Kagera River. The common Grey Heron is also a familiar sight, and at times flocks of the White Egret. Night-Heronsare fairly common along the Kagera River. In the lake, Egyptian Geese and Spur-winged Geese may be seen in small numbers; Pigmy Geese ave not wacommon near the lake- shore, where the open reeds afford them shelter. Yellow-billed Ducks are perhaps the commonest of the ducks on Lake Victoria. Throughout the course of the Kagera River no ducks and geese were observed until reaching the Rufua River, and especially the Karenge Lake. The latter seems a favourite haunt of wild-fowl, and for this region of Africa is very well supplied with water-birds of all descriptions. Huge flocks of Pelicans are to be seen, and large numbers of Pochards and Yellow-billed Duck ; also Egyptian Geese. 190 NATURAL HISTORY OF UGANDA ANGLO-GERMAN BOUNDARY. [June6, A few Pin-tailed Ducks were also seen, but no Mallard at any time. ; Teal are not uncommon; and in the Rufusa Stream and the swampy streams draining the Karenge Lake the Snipe were fairly numerous in December and January. Along the lake-shore, especially among the ambatch trees, were vast communities of Weaver-birds. No less than seventeen forms are represented in the collection of these, many belonging to the brilliant species found inland. Ibiges are not uncommon near the water. The Glossy Ibis is a common bird, and most travellers are familiar with its exasperating ery when disturbed. The Sacred Ibis, on the other hand, is much shyer, and confined to larger and remote sheets of water. Crowned Cranes are common, especially in the west. Bustards are not uncommon, especially in the open cattle- country in the west, about December. Denham’s Bustard, the large red-necked species, was frequently seen. Pigeons are not often seen, but the Doves in places were very numerous indeed, especially in Bukanga. Parrots were scarce, except the one small species collected. Grey Parrots, so common in Uganda, were never seen near the Kagera. Birds of Prey were represented by the Bateleur Eagle and another species which was frequently observed pursuing guinea- fowl. Vultures were rarely seen except in Bukanga, where the lions provided them with frequent meals. It is worth noting that in April a migration of Hobbies appears to pass through the country. Enormous numbers of this handsome little faleon were seen at the same time busily engaged in pursuing locusts, large clouds of which appear to make their appearance at the same time. In the neighbourhood of the lake Hornbills of two species are common, and Touracoes of two species make their appearance in the dense forests. An interesting bird was the Honey-guide, which in Bukanga and the narrow valley of the Kagera River very frequently provided us with honey by leading to the nests of wild bees. Goatsuckers are common, and in March, April, and May the ennant-winged species became very conspicuous, as when the long feathers are developed the bird has the appearance, when on the wing, of a toy Japanese kite. Three species of Bee-eaters were seen, but the Roseate Bee-eater of East Africa and the Nile countries did not make an appearance. The birds belonging to the scrub and open forest country, the Barbets, Woodpeckers, Pittas, Swallows, Flycatchers, Thrushes, Shrikes, Tits, and Finches, were never to be seen in large numbers, though appearing in isolated parties sufficiently often to preserve the district from the appearance of lifelessness, which is a disappointing feature in other parts. 1905. ] ON MEXICAN AMPHIBIANS AND REPTILES. 191 The Larks and Pipits were, on the whole, very scarce. An extremely handsome Glossy Starling was a very conspicuous bird, which seemed to like the neighbourhood of camps. The White- necked Crow and the fine Razor-billed Raven were especially common in the west; and it was in the cattle country and on the Ruchigga Mountains that the Tick-birds (Buphaga) were observed, although some were seen following large herds of elands in Bukanga and Mpororo, and the rhinoceros in Karagwe. Perhaps the most noticeable feature in the bird-life was the extraordinary number of Francolins of every species to be seen in Bukanga, very valley and almost every patch of dry grass appeared to contain a large number of these birds. In the evenings, when the grass had been burnt in patches, numbers of Francolins could be observed feeding in the open like pheasants outside a cover in September at home. It would have been easy to have shot forty or fifty brace a-day if time and cartridges had been available. 2. The Distribution of Mexican Amphibians and Reptiles. By Hans Gavow, F.R.S., F.Z.S. [Received May 17, 1905.] (Text-figures 29-32.) CoNnTENTS. Introduction, p. 191. List of Species collected by H. G., p. 198. Physical Features of the Localities, p. 196. Review of the Distribution of Cecilia, p. 199. = : Urodela, p. 200. Summary, p. 204. 39 29 Se F Anura, p. 205. ss p. 208. . os Crocodilia, p. 209. 3 p. 209. s i Chelonia, p. 209. 4 p. 210. - = Lacertilia, p. 211. 5 p. 220. Ophidia, p. 222, 55 p. 226. 33 3) Distribution according to Altitude, p. 227. List of Species occurring in High Altitudes, p. 232. General Conclusions :—Evolution of Middle America, p. 234.. Immigration and Spreading, p. 238. Northern and Southern Immigrants, p. 241. INTRODUCTION. These investigations are based upon a considerable material which it is convenient to enumerate :— 1. The volume on Reptilia and Batrachia, by Dr. Giinther, of the ‘ Biologia Centrali-Americana,’ with its thousands of references to localities, 2. Cope’s posthumous work, ‘The Crocodilia, Lizards, and Snakes of North America,’ Rep. U.S. National Museum for 1898. 3. Boulenger’s Catalogue, with the lists of ever-increasing additions, of the Collection in the British Museum of Natural History. 4. Collections made by Dr, Meek during his ichthyological 192 DR. H. GADOW ON MEXICAN [June 6, tours through many States of Mexico. These, and others, I have been able to examine owing to the courtesy of the officials of the Field Columbia Museum, Chicago. Dr. Meek has, moreover, given me much verbal information about the physical aspects of the places visited by him. 5. There is a fair number of native specimens in the Govern- mental Museums and other Institutions of various towns in Mexico; for instance, in Mexico City, Orizaba, and Oaxaca, but the labels vouchsafe at best no further trustworthy information than ‘“‘ Mexico” or “ La Republica.” 6. Lastly, the material which I have collected myself, or noted down, during two journeys in Mexico, notably in the Valley of Mexico, the States of Vera Cruz, Oaxaca, Guerrero, Morelos, and Puebla, and in the neighbourhood of Zapotlan s. Guzman in Jalisco, especially the Nevado de Colima. The features of the Central and Northern plateau, except the vicinity of El Paso, I know only from several rapid transits, quite enough, however, to gather the main aspects of this enormous stretch of country. Moreover, here Dr. Meek’s information has been especially welcome. Valuable for comparison, but of too short a time for serious collecting, were a few days passed in New Mexico, the Grand Cation of Arizona, the Californian Desert, and the neighbourhood of San Francisco. A few words are necessary as to the way in which I have marshalled the thousands of data. The reputed localities were marked down on an outlined map of the Republic, a separate map for each species. In this way alone generalisations could be formed, often at a glance, concerning the distribution of the species and genera. Many localities, at first suspicious, revealed themselves as very doubtful or as obviously erroneous on further reference to the original papers. Tt was also found that the number of different localities is astonishingly small, less than 100, although they now cover a fair portion of the whole country. With the exception of 20, all these localities lie south of the line Guadalajara, Guanajuato, Tampico. The whole State of Michoacan and the western half of Guerrero are still an almost absolute terra incognita, but to judge from what I have found in Middle Guerrero and what is known from Colima, the fauna seems to be rather continuous. However, the basin of the Lower Balsas and thence to Colima will in all probability yield much of interest to whoever will brave these inhospitable and positively unknown regions. Both Godman (introduction to the volume on Rhopalocera) and Giinther, in their statistical tables, have divided Mexico simply into Northern and Southern by an absolutely arbitrary line which runs from Mazatlan to Tampico right across the country! They have done this in spite of their correct statements about the main physical features of Mexico, the unmistakable continuation of North American forms over the Plateau, and the extension of 1905. } AMPHIBIANS AND REPTILES. 193 Southern or Central American forms northwards into the Pacific and the Atlantic borderlands embracing this Plateau. 'The two columns in these tables are of no use, they are even misleading. Giinther has properly taken off Yucatan as a separate district. Cope’s division (op. cit. p. 1206) into a Sonoran, Austroriparian, Text-fig, 29. © Revilla Gi pedo Tne Wy Qa Map of Mexico. and 'Tolteean subregion of Blanford’s Medi-Columbian region, and the Atlantic+ Pacific Tierra Caliente as belonging to the Neo- tropical region, is excellent when taken broadly ; but his sub- division of the Toltecan into an Oriental, Central, and Occidental province is a failure. List of Species collected by myself during the months of June to October 1902 and during 1904. Dermophis mexicanus. San Juan Evangelista. Amblystoma tigrinun. Lake Xochimilco. 5 altamirant. Dos Rios. Contreras, Sierra de Ajusco. Thorius pennatulus. Citlaltepetl, 9000’; Cerro de S. Felipe, Oaxaca, 8250’. Spelerpes orizabensis. Citlaltepet], 8000-12,500’. » leprosus. 5 8000-11,500’. %, chiropterus. is 9000-10,000’. 5 variegatus. Orizaba, Presidio S. of Cordoba, Tetela, 8. Juan Kvangelista. *. belli. Omilteme. Batrachoseps attenuatus. Nevado de Colima, 7000’. Scaphiopus dugesi. Totolapan, S. Oaxaca. Rhinophryne dorsalis. Presidio; Agua fria. Proce. Zoou. Soc.—1905, Vor. II. No. XIII. 13 194 Bufo valliceps. » marinus. 5 marmoreus. >, mtermedius. Hyla baudini. » eximia. » Sstaufferi. » coper. Phyllomedusa dacnicolor. Hylodes rhodopis. » beatae. Hupemphix gadovii. Leptodactylus albilabris. % caliginosus. Borborocetes mexicanus. Syrrhopus verruculatus. Paludicola mexicana. Engystoma ustum. Rana montezume. » halecina. 3 palmipes. Crocodilus americanus. Caiman sclerops. Cinosternum integrum. 5) effeldti. 99 leucostomum. sp pennsylvanicwn. Dermatemys mawi. Nicoria rubida. Chrysemys grayi. “6 ornata. Chelone viridis. Spherodactylus glaucus. Phyllodactylus tuberculosus. Coleonyx elegans. Holbrookia texana. Uta elegans. 5, bicarinata. » irregularis. Phrynosoma asio. 4 modestum. Sceloporus torquatus. x spinosus. - acanthinus. op Sormosus. os pyrrhocephalus. % encus. DR. H, GADOW ON MEXICAN [June 6, Orizaba, Presidio, Motzorongo, Tetela, Agua fria. Tetela, San Mateo del Mar, Tehuantepec; Iguala, Tierra Colorada; San Luis Allende. Salina Cruz; Cocoyul, Tierra Colorada, Cajones 3000’, Chilpancingo, Rio Balsas, Iguala. Tetela, Totolapan, Oaxaca; Omilteme, Chilpancingo, Presidio, Motzorongo, La Raya; Tierra Colorada. San Luis Allende. Tacubaya near M. C., Buena Vista. Motzorongo. Chilpancingo, Mazatlan, Cajones. Rio Balsas; San Luis Allende. Citlaltepetl, 8000-12,500' ; Motzorongo, Agua fria ; Nevado de Colima, 8000’. La Perla, North of Orizaba. San Mateo del Mar. Agua fria, Salma Cruz, Cocoyul. 8. Juan Evangelista; S. Mateo del Mar, Salma Cruz, Tequesixtlan ; Cocoyul, Pacific Camp, San Luis, Tierra Colorada. Omilteme, 7500’; Nevado de Colima, 8000’, Buena Vista, S. Guerrero. Chilpancingo. Presidio and Motzorongo. Xochimileo, Chalco, Zapotlan. Mexico, Orizaba, Motzorongo, La Raya, Agua fria, Salina Cruz, Tequesixtlan, near Totolapan ; Cajones, Buena Vista, Tierra Colorada, Limon ; Omilteme. Motzorongo, Tequesixtlan ; Cuernavaca. La Raya, Agua fria, Rio Balsas, Pacific Camp. Agua fria. San Mateo del Mar; S. Dionisio, Zapotlan. San Mateo, San Luis Allende. Tetela. Tetela, Agua fria, San Mateo del Mar. San Mateo. San Mateo. San Mateo; Pacific Camp. Salina Cruz. Totolapan ; Colorada. Cocoyul. Juarez, El] Paso. Juarez ;San Marcial, N.M.; Grand Canyon, Arizona. Salina Cruz, T'equesixtlan, San Bartolo, Totolapan, Rio Balsas, Iguala, Mesquititlan, Chilpancingo, Tierra Colorada, Cocoyul. Cocoyul, San Luis Allende. Salina Cruz, Tequesixtlan ; Rio Balsas. Juarez, Hl] Paso; San Marcial, N.M. Xochimilco; Chilpancingo; Zapotlan, Nevado de Colima. Totolapan. Cuernavaca, Iguala, Rio Balsas, Tierra Colorada, Ayutla. Oaxaca, Cerro 8. Felipe, 6000'; Omilteme. Oaxaca; Chilpancingo ; Nevado de Colima, 7000- 6000’. Teuala, Rio Balsas. Citlaltepetl, up to 18,700° ; Contreras, Pacific Camp, San Luis, Tierra 1905. | AMPHIBIANS AND REPTILES. 195 Sceloporus scalaris. variabilis. f siniferus. re melanorhinus. eS microlepidotus. op gadovia. Corythophanes hernandezi. Basiliscus vittatus. Iguana rhinolophus. Ctenosaura acanthura. Be quinquecarinata. Ameiva undulata. Cnemidophorus gularis. 3 mexicanus. s bocourti. : deppet. 3 striatus. guttatus. 39 Anolis sallei. 5, tropidonotus. > nebulosus. » gadovii. » liogaster. Gerrhonotus gramineus. 99 antauges. 55 deppet. 45 imbricatus. 5 liocephalus. Eumeces fuscirostris. % lynxe. Mabuia agilis. Lygosoma laterale. Anelytropsis papillosus. Chirotes canaliculatus. Glauconia albifrons. a; dulcis. Boa imperator. Tropidonotus melanogaster. i sealiger. “5 ordinatus, var. eques. var. marcianus. i scalaris. 3 chrysocephalus. Contreras. Chilpancingo, Tierra Colorada, Rio Balsas. Orizaba, La Perla; Mexico, Presidio, Motzorongo, la Raya, Tetela, Agua fria. Chilpancingo ; Nevado de Colima, 7000-6000. San Mateo del Mar, Tequesixtlan, San Carlos Yautepec. From Pacific Camp to S, Luis, Tierra Colorada, Buena Vista, up to southern slope of Los Cajones. Tierra Colorada, Cocoyul, San Luis. Citlaltepetl, up to 13,500’. Xochimilco, Contreras. Iguala; Omilteme, Nevado de Colima. Mesquititlan between Mescala and Chilpancingo. Motzorongo. Motzorongo, Agua fria, Huile; Tequesixtlan Tierra Colorada, Ayutla. Tetela, Agua fria ; Tequesixtlan, Totolapan. Yetela, Agua fria; ‘Tequesixtlan; Cuernavaca, Iguala, Tehuantepec, Salina Cruz, San Mateo. Rio Balsas, Tierra Colorado, Cocoyul, Pacitic Camp. Salina Cruz, Tequesixilan. Presidio, Agua fria; Cocoyul forest ; North and South below Los Cajones. Puente de Ixtla. Rio Balsas. Chilpancingo. Cuernavaca. Oaxaca. Oaxaca. San Juan Evangelista. Tehuantepec; Salina Cruz; Tequesixtlan; Totolapan; San Carlos. Rio Balsas. Tierra Colorada ; Cocoyul; Pacific Camp; San Luis Allende. Ayutla. Salina Cruz; Totolapan; ‘Tierra Colorada; Ayutla; San Luis; Cocoyul. Agua fria. La Perla near Orizaba. Motzorongo, La Raya. Cuernavaca. Tierra Colorada; Cocoyul, San Luis Allende, El Coquillo ; Nevado de Colima up to 7000’. Tierra Colorada. Omilteme, 7600’. Citlaltepetl, 8000-9000’. ¥ 12,000’. i 8000-10,000'. i 8000-10, ‘000 5 La Perla; near Mexico, 7800’; Nevado de Colima up to 11,000’. Omilteme, 8000’. Nevado de Colima. Omilteme 8000’. Salina Cruz, Tequesixtlan ; Tierra Colorada, Cum- bre del Coquillo, Buena Vista, Cocoyul. La Perla. Motzorongo. Rio Balsas. La Raya; Chilpancingo. Rio Balsas ; Chilpancingo. Motzorongo. North of Rio Balsas; Coquillo. Xochimilco. Nevado de Colima. Xochimilco. Xochimilco. San Mateo del Mar. Rincon, N.M. Citlaltepetl, 8000-12,000’, Omilteme, 13* 196 DR. H, GADOW ON MEXICAN [June 6, Contia nasus. Contreras. Zamenis mentovarius. San Mateo, Tequesixtlan. 35 mexicanus. Rio Balsas, Cocoyul. 3 pulcherrimus. Salina Cruz. 3 lineatus. Tguala. Coluber eorais. La Raya; Tequesixtlan; Ayutla. Coronella micropholis, var. B. Chilpancingo, San Luis Allende. Urotheca elapoides. La Raya. Drymobius margaritiferus. fotzorongo, Agua fria, San Juan Evangelista. 3 boddaerti. Motzorongo, La Raya. Leptophis mexicana. Motzorongo, La Raya. * diplotropis. San Mateo del Mar. Rhadinea decorata. Presidio. Hi clavata. Tetela. Ws vittata. La Raya, Salina Cruz, Tequesixtlan, Totolapan ; Omilteme, Chilpancingo. Streptophorus atratus. Presidio, Motzorongo, La Raya. Geophis semidoliatus. Orizaba. Geagras redimitus. San Mateo del Mar. Trimorphodon upsilon. Cuernavaca, Rio Balsas. is biscutatus. Tierra Colorada. Himantodes cenchoa. La Raya. Tieptodira personata. Rio Balsas, Ayutla. 45 albofusca. Cocoyul, San Luis ; Nevado de Colina. A guilleni. Rio Balsas. Conophis vittata. Salina Cruz, Tequesixtlan. Homalocraniwm miniatum. Tezonapan, N. of Ayutla. Manolepis putnami. Los Cajones. Petalognathus nebulatus. La Raya. Dryophis acuminatus. Motzorongo. laps fulvius. La Raya, San Juan Evangelista. Lachesis lanceolatus. Motzorongo, La Raya. » atroaw. Motzorongo. Crotalus terrificus. Tequesixtlan. 43 triseriatus. Citlaltepetl, 9000-12,500’ ; Nevado de Colima. PuysicAL FEATURES. A list of the names of the localities where collections have been made should be supplemented by @ short description of the chief physical features; without these it is of no more use than the bare mentioning of the name of the political country. The “altitude” is supposed to be all-sufficient; but this is a great mistake, since it conveys nothing without further informa- tion. For instance, 2000 feet on the Atlantic slope means typical tropical hot-country vegetation, while on the Pacific side (¢. 9. Oaxaca and @uerrero) the same elevation implies pine- and oak- forests, with a character devoid of tropical fauna and flora. Mexicans divide their country into the Tierra Caliente, Templada, and Fria, with a hot, temperate, and cool climate respectively ; the Tierra Templada corresponding on an average with an elevation of 3000 to 5000 feet. But the natives of the State of Vera Cruz draw the imaginary line at a level very different from that used by the citizens of Puebla. Chilpancingo, 4100 feet, in Guerrero, has a much cooler climate, with nothing tropical about its vegetation, than Oaxaca, 5060 feet, or even Orizaba at 4027 feet, which is in many respects subtropical. Valley of Mexico, 7600 feet. Alluvial; swampy meadows to west and south. Lake Texcoco brackish, Lakes Chaleo and 1905. ] AMPHIBIANS AND REPTILES. 197 Xochimileo freshwater, surrounded by meadows, wooded hills with streams. Sierra de Ajusco, volcanic, well-wooded mountains. Contreras, 8090'; Dos Rios, 8800’, pines. Orizaba, 4027’; on the east side of the slope of the plateau; valley with streams, pastures, and rich vegetation, on alluvial and hard-limestone terrain. Thence gradual ascent to the volcano Citlaltepetl; dense mixed forest, oak, arbutus, and pines, about 9000’; giving way entirely to pines. Tree-line about 13,500’; then tussocks of grass. Snow-line about 14,500’. Cordoba, 2700’. Dense tropical vegetation. Thence south- wards, through limestone terrain, along the foot of the slopes, which are covered up to the edge with luxurious forests ; east- wards bordered by savannahs. Permanent rivers with high banks ; lagoons in the forests and savannahs. Motzorongo, Presidio, and La Raya, about 1500’, in forest land. Tetela, about 900’, near the edge of the savannah. Agua fria, 100-200'; lagoons, swamps, and low forest; flat country, subject to inundations. The Rio Papaloapan, with its many tributaries, carries an enormous volume of thick, yellow water; much of the lower basin is for months under water, only island-like parts standing out, used as refuges, although by no means sanctuaries, by the game and other creatures. Then follows low, undulating, rolling, cattle-grazing land, with sandy subsoil. San Juan KEHvangelista, 100’. Tropical river-bed through savannahs, bordered by dense lowland forest. Isthmus of Tehuantepec. Eastern slope, dense humid ever- green forest; on the ridge, less than 1000’, open country with temporary stagnant lagoons; on the western slope prevails the dry Pacific type without continuous forests, but with more scattered patches of mostly deciduous trees. Tehuantepec, 120'. Sandy, varied terrain. San Mateo del Mar. Sandy, lagoons connected with the sea. Flat, scanty vegetation except in occasional swamps or near the lagoons, some of which are fringed with dense low brushwood and small trees. Salina Cruz. Porphyritic terrain; hilly, steep coast-range, varied by promontories and fresh- and salt-water lagoons. From Salina Cruz and Tehuantepec northwestwards to Oaxaca. The coast-range, averaging 2000-3000 feet, is covered with pines down to 2000 feet. Tequesixtlan, 560’, in a river-valley, shut off by the coast-range; varied, rather low vegetation, volcanic alluvial. Thence through mountainous country, across parallel ridges. Hills at first with tropical vegetation; oak-forests from 3000’ upwards, then pines prevailing, but nowhere continuous forests. General character of the country dry, most of the smaller rivers run dry except in the rainy reason. An intricate system of 198 DR. H. GADOW ON MEXICAN [June 6, hills, deep gorges with sandy river-beds and patches of xerophile vegetation, especially cactus. San Bartolo, 2800'; San Carlos, 2460', in more open country. Totolapan, 2800’, in broad, sandy river-valley. Thence steep ascent on to the plateau, the edge of which is 5300’. Looking back, southwards, over the many ridges of mountains and hills, the Tierra Caliente appears to be densely covered with wood, while towards the north stretches the flat, almost treeless Southern Plateau, here and there with outcropping low ridges which are barren when of volcanic, wooded when of Paleozoic formation. ; San Dionisioand Tlacolula, 5160’; Oaxaca, 5060’. The Paleeo- zoic terrain stretches from Oaxaca westwards. There is the forest-clad Cerro de San Felipe, 9000 feet high, with semialpine meadows on its slopes. From Cuernavaca southwards to the coast. Cuernavaca, 5000’. Fertile valleys of voleanic and limestone terrain, with little streams. Shut off from the Valley of Mexico by the high, voleanic, densely wooded Sierra de Ajusco, &c., averaging 10,000 feet. Puente de Ixtla to Tetecala. Pleasant, fertile, varied open country; limestone terrain, in which are the huge Caves of Cacahuimilpa. Tguala, 2400’, in a wide plain, surrounded by limestone hills, with scanty vegetation. Rio Balsas Station, 1500’; in the valley of the Balsas or Mescala River. Very mountainous ; tropical vegetation. High and low forest, interspersed with much brushwood. The river, during the rainy season, brings down floods of yellow or brown water with rather little sand, but much comminuted vegetable matter. The bed is rocky, limestone, the banks mostly steep, but there are many sandy shelves above the high- water mark. Mescala, 1700’. The Balsas basin is bordered on the southern side by a long and high range of mountains, parallel with the Pacific coast, attaining heights of 10,000 feet, densely wooded, intersected by very deep, steep, and long gorges, and the river-beds are the only available roads; here and there these river-beds broaden out into meadows. Mesquititlan, 2800’. Narrow, luxuriant gorges. Zumpango, 3400’. Open, sandy, meadows. Chilpancingo, 4100’. In a windswept, shallow depression of Cretaceous terrain, surrounded by sparsely wooded hills, and meadows on the top of the ridge. To the west, in the mountains the hamlet of Omilteme, 7100’, luxurious forests; at first oak, dwarf palms and pines; then oak, pine, and arbutus; then oak and pine; and lastly pines up to the highest summits, forming dense high forests, with the most luxurious underwood in the gorges, especially within the cloudbelt. Mazatlan, 4200’. Meadows and fields. 1905.] AMPHIBIANS AND REPTILES. 199 Cumbre de los Cajones; the pass at 3500’ over a ridge which, on the south, flank of the main range, marks the beginning of volcanic terrain. Oak and pines and ‘columnar cactus. Buena Vista, 2300’. Wide meadows, with pools in the rainy season ; low hills with oak and pines. Typical Tierra Caliente, with an essentially tropical aspect of flora and fauna, begins on the southern slopes of the main ridge, coinciding with what is officially and locally known as La Costa. Its upper ~ limit may be put at not higher than 1000 feet. In the depressions between the successive parallel ranges of ‘hills the type is absolutely tropical and southern, but the country loses this character at once on the ridges which rise higher than 1500 feet. Tierra Colorada, 990'; river valley, voleanic. Andesite overlaid with red rubble. Valley of the Omitlan River, 500’. Limestone, densest vege- tation on the slope which culminates in a ridge of 1600’, called Kl Cumbre de Coquillo. Coquillo to Chacalapan, about 700'; tropical life-—From here, across several smoothed down ridges and to the coast, the subsoil consists of gneiss and granite in rapid decompo- sition, fairly well wooded in clumps or large patches, often interrupted by meadows and natural stretches of pasture. Numerous small rivers, carrying much sand, but nearly always with clear water, but most of them are lable to run dry in the winter. Near the granite-bound coast are numerous lagoons, mostly of fresh water, and there is a broad belt of almost impene- trable high forest, which in many places touches the sea. The mangrove-swamps of course are permanent, but during the rainy season many parts of these forests are inundated. Limon, 1800’. Open, dry grass country. Teconapan, 1500’. Broad meadows. Ayutla, 1200’. Permanent river; rich vegetation. Cocoyul, 160’. Near the coast forest. Pacific Camp. Near the shore, 99° W., 16° 36’ N. Close to the forest ; large lagoons and swamps; granitic rocks and mangroves. San Luis Allende, oa0n Broad river-valley, with well- wooded hills which are covered with various kinds of oak, and from 1900' upwards chiefly with pines. Nearly the whole Coastal District is, during and after the rainy season, covered with a dense mass of tall herbs, which between the trees especially take the place of underwood. CACILIA. Dermophis mexicanus.—This is the only Ceecilian which extends into the Mexican Tierra Caliente. Previously known to range from “Tehuantepec” to Panama, I found it in the low woods near San Juan Evangelista. The American ancestral home of this circum-tropical family is South America, and none are known to occur on the Antilles or on the Galapagos. It is therefore 200 DR. H. GADOW ON MEXICAN [June 6, interesting that these burrowing, slowly moving worm-like creatures have managed to travel over at least 1500 miles of ground, covered with humus, since the close of the Miocene epoch, i.e. since the separation of the Antilles (cf p. 237). A not unreasonable computation of one million years carries us back into the Miocene epoch. The rate of spreading could in this case have been extremely slow, only about one mile in 700 years, and this works out at three yards a year. Of course this is mere speculation, but it may be as well to give even such an imaginary instance of slow spreading. The fact remains that Dermophis ~ has done it, and whether we double or treble the rate of progress, or increase the time two- or three-fold, the result remains within very reasonable possibility. URODELA. The Amblystomatine are a pre-eminently Eastern Palearctic group; only two out of eight genera occur in North America: Dicaumptodon ensatus in California, and Amblystoma, with some 16 species, on the North-American Continent, and one, A. persimile in Siam. In Mexico only two species occur. Amblystoma tigrinuwm, the larval form of which, when per- manent, is the famous Axolotl. This species has an enormous range, from the State of New York to Dakota and Colorado, whence, apparently now with wide gaps between, it extends through Mexico, as far south as the valley of Mexico City. But its dis- tribution in Mexico is, at least now, restricted to the western Sierra Madre and the southern part of the Mexican plateau. Well-ascertained localities of this species are the following :— West of Chihuahua Town; West of the town of Durango; Cumbre de los Arrastrados in Jalisco; somewhere N.W. of Guada- lajara ; district of Autlan in Jalisco ; Lake Patzcuaro in Michoacan, Valley of Mexico, notably Lakes Xochimileo and Zumpango (but not Lake Texcoco, to which alone Weismann’s dismal dream to account for the permanent Axolotl stage could apply!). Possibly there are Amblystoma, either metamorphosing or as Axolotls, in or near some of the other lakes of Michoacan and Jalisco, but they have as yet not been recorded from Lake Chapala ; and I found none in the Lakes of Zapotlan; nor were such creatures, or even the name Axolotl, known to the natives. A. altamirani.—This species, which metamorphoses regularly into a gill-less Newt, is known only from the streams of the mountains which border the western and south-western side of the Valley of Mexico. It was discovered in the Montes de las Cruzes, about 15 miles to the west of Mexico City, at an altitude of 8800 feet. In 1902 I found it also above Contreras, in the Sierra de Ajusco, some 12 miles south-southwest of the city, at an altitude from 8500 feet upwards to 8800 feet. Further up the rivulets are apparently too small. I stated in ‘ Nature,’ Feb. 5, 1903, that searching in the streams only a little above the City of Mexico, 1905. | AMPHIBIANS AND REPTILES. 201 which lies at an altitude of about 7600 feet, was fruitless. Inthe month of September 1904, however, when we revisited this district, I was able to ascertain that these Newts live regularly in the stream below Contreras (altitude 8090 feet) down to about 7900 feet, where the stream leaves the hills, and runs, still swiftly, in its stony bed through the Pedregal, or recent field of Java, then through rich evergreen meadows into Lake Xochimilco. Moreover, I can now add with certainty that A. altamirai is absolutely aquatic throughout its life. The natives (millers, field- labourers, and boys) knew the creatures well. They called them “axolotes sordos” (deaf, having no ears), and described them as axolotes sit aletas (without winglets, meaning gills); when I searched for them on land, on the bordering meadows, under stones, or amongst the trees, the people laughed at my ignorance of expecting to find “fishes” on dry land. ‘There are no fishes in that stream. But this, their “ fish,” they pronounced as no good, because these axolotes de cerro (Mountain Axolotl) are not eaten like the ‘ axolotes del lago.” During our last visit the mountain-streams were transformed into turbid roaring torrents, and it was only at a few spots that the Newts were visible, generally in some stiller water, in the shelter of some great boulder. There they stood, or rather were lying, on little patches of sandy bottom, the larvee working their gills vigorously, the adult motionless except for the undulating tail, and never rising to the surface to breathe. They were all extremely shy, quickly hiding beneath or between the stones. In the Montes de las Cruzes, close to the railway-station Dos Rios, the streams form here and there little swamps or ditches, with much watercress in the slowly-flowing water; there we found plenty of larve; the adult only in the running water. Not one of these mountain-streams runs dry. The lungs are well developed. The only specimen, a larya 100 mm. long, which I succeeded in bringing home alive in 1902, metamorphosed within 8 weeks, losing the fins and gills, and closing the gill-openings completely, but it died before losing the yellow and black piebald coloration. The distribution of Amblystoma in Mexico coincides absolutely with the large central and western portion of the country, which has been covered with volcanic masses, repeatedly or successively, since the Kocene epoch ; and the last outburst, which produced the Pedregal near Mexico, is known to have occurred after this part of the country was already inhabited by man. It was impossible for - Amphibia to live on such a terrain until it was weathered enough to sustain a permanent and moisture-loving vegetation. In fact every locality where A. tigrinwm is known to occur is on the Quaternary, mostly sandy, patches formed by the disintegrating debris of the voleanic masses; or it is found in the lakes, all of which are partially filled-up mountain valleys. We have to conclude that the Amblystomas are recent immigrants from the North. Where they have met such lakes, 202 DR. H. GADOW ON MEXICAN [June 6, these have become, or are becoming, too attractive for them, with the result that A. tégrinwm has sunk, or is sinking, into a more or less perennibranchiate state, the Axolotl. Typical Axolotl are those of Lake Xochimilco, the condition of which I have deseribed in ‘Nature, Feb. 5, 1903, and Lake Patzcuaro, which, with its rushes, weeds, and other abundance of vegetation, is very similai to the Mexican lake. Sexually ripe Axolotl are also known from Jalisco mountain tarns or lakes, and lastly from St. Mary’s Lake, Estes Park, Colorado. It is therefore the combination of certain favourable circumstances (permanence of water, abundance of food, shelter, equable temperature) which produces the ‘“ Axolotl.” Whoever has seen the very different conditions prevailing in Lake Zumpango, to the north of Mexico City, will easily credit Velasco's statement that A. ¢igrinwm metamorphoses into the normal gill- less Newt, as it does in the United States, and probably in various other parts of Mexico. Allthe more interesting is the fact that the other species, A. altamirani, the only one which lives in the streams of recent voleanic mountains, has been modified into a gill-less but permanently aquatic form. DESMOGNATHIN#.—The three species of Desmognathus inhabit the Kastern United States. Typhlotriton speleus is restricted to subterranean caves in Missouri. Thorius pennatulus, the only remaining member of this small group, and its sole representative in Mexico, points therefore unmistakably to the Eastern half of North America as the original home of the group, not of Thorius itself. This tiny Newt, less than two inches in length and thinner than a match, with weak limbs and reduced digits, shows a peculiar dimorphism of the size of the nostrils. They are very large and open in the males, much smaller in the females. The lungs are quite aborted as in Desmognathus and Spelerpes. Thorius has a very limited distribution. It was discovered on the south-western slope of the Pic of Orizaba. J found Thorius in abundance on the south slope, 9000-10,000 feet, in high, mixed forest, either on the ground beneath flat pieces of fallen bark, or on decaying logs of pine between the bark and the wood amongst the ‘““worm-meal” of boring beetles and maggots. Again I met them under exactly the same conditions on the Cerro de San Felipe, 8250 feet, near Oaxaca. These are the only two localities so far as we know at present. It is doubtful whether their distribution is now continuous; the watershed between the Atlantic and Pacific, to the west of a line drawn from Orizaba to Oaxaca, averages about 8000 feet in height, and it is well- wooded, but there are several deep transverse depressions in it. PLETHODONTINH.—This group, consisting of 5 genera with about 40 species, is entirely American, with the sole exception of Spelerpes fuscus in Kurope. 1905. | AMPHIBIANS AND REPTILES. 203 Spelerpes.—This large genus, composed of about 20 species, ranges from Massachusetts into North-western South America. At least 10 species live in Mexico, 9 of them south of a line drawn from Guadalajara to Tuxpan on the Atlantic; some of them extend into Guatemala and Costa Rica. S. yucatanicus in Yucatan. A few occur as far south as Peru; one, S. infwscatus, inhabits Hayti, and S. fuscus lives in Sardinia and Northern Italy. The distribution of the Mexican species is important. The Aztec name is “ Tlaconéte ” = little land creature. S. cephalicus, described by Cope from ‘“ North-eastern Mexico.” No Spelerpes seem to occur in Texas ; the nearest American species, S. multiplicatus, lives in Arkansas; S. orizabensis and S. lineolus, the latter with tiny, reduced limbs, are known only from the mountain of Orizaba, S. orizabensis ranging between 8000 to 12,000 feet. S. leprosus, of which gibbicaudus Blatchley is a not unfrequent individual variation, is common in the mixed and pine forests of the mountain of Orizaba, up to 12,000 feet. It has also been recorded from the north slope of Popocatepetl, 9000 feet, and from the mountains of Jalapa. S. morio from “ Jalapa,” and from Tlalpam, which lies between Mexico City and Lake Xochimilco, in flat, sandy, moist terrain, with meadows and willows. It appears again far in the south, in Guatemala and Costa Rica. S, chiropterus. Mountain of Orizaba, from the town, 4000 feet up to near 10,000 feet; ‘ Jalapa,” and Cuernavaca which has an elevation of 5000 feet. ‘ Vera Cruz” must be left as a doubtful locality. S. rufescens is recorded from ‘“ Orizaba,” Cordoba, Vera Cruz, Tehuantepec, Chiapas, and Tabasco; all in the Tierra Caliente, except the first locality. S. variegatus ranges from the Valley of Mexico, Orizaba (from 9000 feet downwards), Jalapa, Cordoba, right through the forest of the Tierra Caliente and through the whole of Central America tc Costa Rica. I found it on Orizaba mountain, as well as at San Juan Evangelista, which lies scarcely higher than 100 feet above the sea, in the same ground with Dermophis. S. untformis, with reduced limbs like S. lineolus, described from Costa Rica, elevation of 5000 feet, is said also to have come from “ Vera Cruz.” Lastly, S. belli: mountains of Jalapa, Orizaba, Mexico, Zacualtipan, Guanajuato, Guadalajara, Sierra de Nayarit; and at Omilteme, west of Chilpancingo. This species alone has found its way across the plateau, following the belt of alluvial deposits described elsewhere (p. 237). With the exception of this transverse belt, the distribution of Mexican Newts coincides closely with the broad band of Cretaceous limestone which extends from Nuevo Leon to the Isthmus, with intricate but almost continuous patches verging from Cordoba and Orizaba south-westwards to Chilpancingo. This limestone terrain was the only one available 204 DR. H. GADOW ON MEXICAN [June 6, for Newts during most of the Tertiary period: on the west the plateau suffering from the volcanic revolutions; on the east the sea still covering the present Atlantic Tierra Caliente. The Cretaceous parts formed so many oases where alone Newts could exist or survive. Later, when the volcanic ranges, even the voleanoes themselves, became covered with forests, the Newts spread onto them, just as they have spread into the moist hot- lands of the State of Vera Cruz. It may appear strange that the limestone should have been the means of their preservation, con- sidering that lime-water is, as a rule, not favourable to their development ; but here comes in the significant fact that most, if not all, the Mexican Spelerpes are viviparous, unless they deposit their eggs, like Batrachoseps, in hollow trees. Some of them, for instance S. orizabensis, lead a partially arboreal life. We found many on the pine-trees of Citlaltepetl, favourite hunting and hiding places being the epiphytic plants, especially the large TMlndstne. Orchids, and Philodendron in the hot country forests. The humus and moisture collecting in these growths, often many feet above the ground, swarm with insect life and with little Scolopendras, which seem to be the staple food of these Spelerpes. The apparent absence of Newts on the Northern plateau is most likely due to the dearth of permanent moisture, long-continued periods of drought, and dust. A more difficult question is the apparent absence of Newts on the terrain of gneiss and granite which covers so large a portion of Southern Guerrero and Oaxaca, and on the well-wooded mountains of the Sierra Madre. For months have we searched Guerrero during the rainy season (there are thousands of places which, if they were on the Eastern slope, would yield an abundance of Newts), but it was in vain. A few specimens of S. belli, from the mountain forest of Omilteme, are the only exception. Batrachoseps.—B. scutatus ranges from Illinois to Rhode Island and to the Gulf of Mexico; the other species live in the Pacific States, from California to Oregon. Quite unexpected was there- fore the occurrence of the Californian LB. attenwatus on the Nevado de Colima. I found a single, young specimen on the northern slope, at about 7000 feet elevation, in the stump of a decayed pine- tree. There are some patches of granite and of limestone in that district, but then comes an unbroken stretch of originally volcanic formation, for about 200 miles, until the gneiss is reached to the north-east of Mazatlan. We cannot well imagine that this species is an ancient survival; it must be a comparatively recent immigrant from the north-west, from California. Probably it occurs all through the slopes of the western Sierra Madre, which is mostly clad with pine-forests. Résumé of the Distribution of Mexican Urodela. All the American Urodela are of Nearctie origin, with their earliest centre in Old Sonoraland. At least the Amblysto- 1905. | AMPHIBIANS AND REPTILES. 205 matinee point to the long-continued land-connection with Eastern Asia. A later centre of dispersal lies in the Eastern half of North America, the old Appalachia, the Alleghany moun- tains, &e., whence Urodeles have spread, as Plethodontine and Desmognathine, over most of the Eastern and Southern States, also into and through Mexico’s eastern half. This spreading dates back to Miocene times, witness the existence of Spelerpes in Hayti, while others have reached even South America, and, lastly, the occurrence of a Plethodon somewhere in the La Plata basin. Much later immigrants, directly from the old north-western home, are Amblystoma and Batrachoseps: A. tigrinum and B. attenuatus being identical species in the States and in Mexico, only 4. alta- mirant being a new modification ; while Spelerpes has developed many species, different in the north, centre, south, and in Hayti. ANURA, PELOBATIDE.—Scaphiopus, the Spadefoot, closely allied to the European Pelobates, is the only American genus of this family, with about 7 or 8 species, two of which are restricted to the United States. The zone of sandy terrain of Texas, New Mexico, and California is richest in Spadefeet, whence they have extended over the Mexican plateau down to the Pacific and Atlantic coasts. S. dugest s.hammondi has the widest distribution: from Missouri and California through New Mexico and Texas, the mountains west of Chihuahua, in Guanajuato, and again on the southern slope of Oaxaca, where I found it at Totolapan, its most southern locality. The retiring habits of the Spadefoot no doubt account for the few scattered- records. The well-wooded mountain ranges which form the south-eastern, southern, and western borders of the Mexican plateau are a natural obstacle to a further southward spreading of this genus, Buronip#.—Central America and Mexico are one of the centres of Bufonide. Concerning Mexicans, they can be grouped as follows :— 1. Indigenous: Rhinophryne dorsalis, the only species of the genus, a toad specialised as an eater of Termites ; it is restricted to the moist Atlantic Tierra Caliente, from Tuxpan, north of Vera Cruz, through the Isthmus to Campeche and Guatemala. The light- coloured spots on the bluish-slaty black skin are either yellow or orange to red, varying in individuals from the same locality. They are very sluggish, rather nocturnal, and retire beneath a rotten stump or into a small self-dug hole in the moist humus. Aztec name “ Poche.” 2. Southerners : Hngystomops, the few species of which range from Venezuela and Ecuador northwards, but only 7. pustulosus reaches the Isthmus of Tehuantepec. Several of the 13 species of Swfo found in Mexico are southern forms: in their spread northwards they either stop short at the 206 DR. H, GADOW ON MEXICAN [June 6, Isthmus, B. coccifer and Bb, sternosignatus ; or they extend into the Atlantic hot-lands, canaliferus; or along the Western Sierra Madre into Jalisco, intermedius, with marmoreus peculiar to South-western Mexico; or they go as typical hot-country Toads into both the eastern and western Tierra Caliente, marinus, valliceps on the Atlantic side from Nicaragua to 'Texas, not on the plateau, but recorded from Jalisco and Presidio near Mazatlan ; lastly, B. stmus from Panama on to and over most of the Mexican plateau. 3. Northerners, chiefly at home in the South-western United States and in the northern half of Mexico, eventuallly extending south over the Central plateau: B. punctatus, debilis, compactilis. Of the Bufonide which are found in the Greater Antilles all are now separate, insular species, except B. marinus, which has probably been introduced. Hytipa.—The creative centre of this family is decidedly South America, Every one of the 14 genera of Hylide is found in America, and it is only by the large genus Hyla (incl. the slightly modified Mylella) that this family has attained its world-wide range with the remarkable exception of the whole Paleeo-tropical region. From North-western South America they have spread through Central America into the Antilles (about 7 or 8 species, mostly peculiar), and through Mexico into North America. Concerning Mexico they fall into the following groups :— 1. Genera peculiar to Mexico: Pternohyla, P. fodiens of Presidio near Mazatlan ; Zriprion, 7. petasatus of Yucatan; but Diaglena jordant of Ecuador and Corythomantis greeningi of Brazil point to the south as the old centre of these peculiar Mexicans. 2. Genera with preponderating numbers of species in Central and South America, while comparatively few have reached, or have been developed in, Mexico: Phyllomedusa with only P. dacnicolor on the Pacific side, Agalychnis callidryas and Nototrema oviferum in the Atlantic Tierra Caliente. P. dacnicolor is saturated green, often with the same white temporary patches or spots as happen so frequently in the Australian Hyla cerulea, They were pairing at Rio Balsas in the month of June. During the nightly thunderstorms the males kept up an incessant noise like the snarling bark of little dogs. The couples were sitting in low shrubs or amongst herbs, a foot above the ground, overhanging little ditches which led into a dirty stagnant pool. During the daytime the ditches were absolutely dry. The eggs are very small, very numerous, and of a light grass-green colour. 3. Hyla. About 15 species are recorded from Mexico, to which no less than 11 seem to be peculiar, but at least 8 of these have hitherto been found in single localities only. Our knowledge of the distribution of Tree-frogs is still very defective. Most of them inhabit the forest-regions of the Atlantic slope. They are dis- 1905. ] AMPHIBIANS AND REPTILES, 207 covered and caught by mere accident. For instance, I found one single specimen of /7. staufferi at Motzorongo, a species until then known only from Guatemala. H. bowcourti of the same country has been recorded once from Tepic, none from the enormous intervening stretch. H. miotympanum seems to range from the Isthmus through the mountainous parts of Vera Cruz, going up towards Puebla. H. venulosa is an eastern form, from South America to Tampico, decidedly Atlantic, but once recorded from near Mazatlan. HH. baudini, the commonest Tree-frog, ranges from fcuador right through Central America, and then spreads east and west through the hot countries of Mexico, absolutely avoiding the plateau, but reaching Texas. On July 4, 1902, when the rains were very irregular, we found H, baudini spawning, south of Cordoba. On a piece of inundated woodland meadow, about the size of a suburban lawn, were 45,000 frogs at a low computation, two-thirds of them in amplexus, the other males making a deafening din. Next day the pool was dried up completely, the grass glazed with the spawn, and there was not a single frog to be heard or seen in the neighbourhood. Hf, copei, known as ‘‘ Sapo blanco” or white toad, is a hill form. Known already from Texas, Chihuahua, Guanajuato, and Jalisco, I found it plentiful on the whitish calcareous terrain south of Chilpancingo, not in the trees but sometimes on rather barren and dry fields. Decidedly typical of the western and southern plateau and its Pacific slope, and very abundant, is H. eximia. Result.—Mexico has many Hylide in common with Central and even South America; but the majority are now peculiar to Mexico, and only two, H#. cope: and H. baudini, extend northwards into Texas. CysTiGNATHIDA, like the Hylide, of decidedly South-American origin, Of the 15 or 16 genera of this family only Leptodactylus *, Paludicola, Syrrhopus, Hylodes*, and Borborocetes occur also in Mexico, altogether with some 23 species. Those marked * are also Aoneilleen. ‘Not one reaches the United States : ; in fact the most northern record is made by H. calcitrans at Zacatecas. B, meai- canus is peculiar to the Central plateau and the high mountains of Jalisco, Colima, and Guerrero. Of the 9 or 10 species of Hylodes 6 are restricted to Mexico, but their recorded localities are still too few and scattered, The same applies to the six species of Syrrhopus ; the others range far south to Nicaragua and Costa Rica : H. palmatus is Pacific, HH, melanostictus Atlantic Mexican ; HT. rhodopis on either side. The last is the commonest species and seems to be an instance of a southerner which, although not going on to the plateau itself, ascends the high mountains on its eastern, southern, and western borders, e. g. Citlaltepetl up to 10,000 feet, Cerro de Oaxaca, Nevado de Colima; it also inhabits the hot low- lands of Agua fria in the State of Vera Cruz. Mostly of dark brown and reddish tints and living on or near the ground; how- ever, some specimens in the epiphytic Tillandsias, or on green 208 DR, H, GADOW ON MEXICAN [June 6, shrubs at the edge of a forest, were quite green, but they soon changed to reddish yellow and ultimately assumed the normal reddish colour. Of the 5 species of Leptodactylus, 3 are too little known, L. microtis from “‘ Guanajuato ” would be the only instance of the occurrence of a Leptodactylus on the plateau instead of in the lowlands. Only two species have a wide distribution : L. albilabris of South Guerrero, Oaxaca, and of Vera Cruz, also Antillean ; L. caliginosus from Paraguay northwards, in Mexico in the ‘Pacific Tierra Caliente as far as Mazatlan. ENGYSTOMATID, with an obviously South-American centre of dispersal; not Antillean. The small genus Hngystoma reaches through Mexico into the South-eastern United States (1. wstwm). Ranww#.—This family is essentially Paleo-tropical. Scarcely more than a dozen species, all belonging to Rana, occur in North America, only 6 in Mexico, and fewer still further south, in the north-western portion of South America. There, however, the Ranide have found a new congenial home, which has stimulated them into the development of 5 new genera, with about one dozen species, all arboreal, besides Dendrobates. The Ranide have not found their way into the Antilles. Of the six Mexican species, 2. forreri is restricted to the district between Durango and Mazatlan; 2. pustulosa to same district and Western Jalisco; &. omiltemana to Guerrero mountains: these three are peculiar to Mexico. &. halecina is the common river-frog of the country, both in the hot parts and on the plateau, extending from the United States through the whole of Mexico down to Costa Rica. palmipes ranges from southern tropical Mexico to South America. Lastly, 2. montezwme, the largest of all, is a lake-dweller, e. g. lakes of Jalisco, near Mexico City, Tehuantepec, extending south into Tabasco and Guatemala. Résumé of the Distribution of Mexican Anura. We can easily distinguish between northern and southern immigrants. 1. Northern, decidedly of Nearctic origin. Scaphiopus, scarcely reaching the Isthmus of Tehuantepec; and a few Rana, all water- frogs. Both genera are comparatively recent immigrants, non- Antillean, although Rana extends through and beyond Central America. 2. Southern, of obvious South-American origin.—a. With related, or identical, species in the Greater Antilles. Cyséz- gnathide, not reaching the United States. Wylide and Bufonide, each, especially the Hylide, with genera peculiar to Mexico, indicating ancient residence.—b. Non-Antillean, a few Hngysto- matine, scattered through Mexico, 1905. ] AMPHIBIANS AND REPTILES. 209 CROCODILIA. Crocodilus americanus is the commonest tropical American Crocodile, from Florida to Northern South America. In Mexico it is strictly confined to the Tierra Caliente, with Mazatlan as its north-western limit. It ascends the Rio Balsas at least up to Meseala, but this is not much more than 1700 feet above sea-level. Common in the lagoons on the coast of Guerrero and Oaxaca, except where it has recently been well-nigh exterminated by American skin-hunters. More exist in the river-systems of the State of Vera Cruz, ascending occasionally up to Motzorongo, i. e. 1500 feet. During the rainy season they often forsake the then turbid rivers, and roam at night through the forests in search of lagoons. CO. moreleti inhabits the Tierra Caliente from Tampico to Honduras. Caiman sclerops s. punctulatus has its centre in South America. In America it occurs only in the Atlantic hot-lands. I met with very large specimens (length of skull 20 inches) at Agua fria in the same lagoons and rivers as the Crocodile. Whilst the latter, anyhow not averse to brackish water, inhabits the Greater Antilles, the Caiman has found its way only into Trinidad and, if report is true, to Martinique. The Alligator of the southern United States does not seem to cross the Rio Grande. CHELONTA. It seems almost incredible that Chelydra has never been re- corded from Mexico, considering the wide range of Ch. serpentina in the United States and the existence of the other species, Oh. rossignoni, from Guatemala to Ecuador. The Papaloapan and 8. Juan Rivers of the State of Vera Cruz are certainly large enough, with pools and backwaters, but I could not ascertain the presence of a large, snappy species. DERMATEMYDID#.—The few species of this family are peculiar to Central America. Dermatemys mawi extends from Honduras into Yucatan and Vera Cruz; it occurs, for instance, in the pools of the forests and savannahs near Tetela, where it is known as the “Tortuga blanca.” Stawrotypus seems to have a similar range: S. triporcatus going up to Vera Cruz; S. biporcatus only up to the Isthmus. CINOSTERNIDH, with the sole genus Cinosternum. About 10 species in North and Central America, one extending to Guiana. Well represented in Mexico by 6 species. Of these, C. pennsyl- vanicum, previously recorded from the Valley of Mexico, was. found by myself in South Guerrero, at San Luis Allende. O. hirtipes ranges from Arizona and New Mexico along the Pacific side into Jalisco, and includes the Tres Marias Islands. Proc. Zoou. Soc.—1905, Vou. Il. No. XIV. 14 210 DR. H. GADOW ON MEXICAN [June 6, C. sonoriense in Sonora. C’. integrum (a variety of C. scorpioides) is likewise Pacific, from Jalisco, e. g. Lake Chapala, to which I can now add Zapotlan and the small rivers on the plateau south of Oaxaca and the swamps of San Mateo near Tehuantepec. C. leucostomum extends from New Orleans along the Atlantic side of Mexico through and beyond Central America. C. effeldti is known from the State of Vera Cruz, San Mateo del Mar, and Guatemala. TESTUDINIDH.—Cistudo with two species in North America and one in Yucatan, strictly terrestrial. (C. mexicana of Texas and New Mexico, e. g. San Marcial. How far it extends into Mexico is not known ; Tampico is quite possible, but I very much doubt “ Mexico City” and neighbourhood. Nicoria rutila I have met in swampy bush-land of the State of Vera Cruz and near Tehuantepec, and this seems to be its range ; allied species occur in Central America. Chrysemys extends from Canada to Argentina, but with a preponderance of northern forms. In Mexico restricted to the hot countries, and even there common only about the Isthmus, whence C’. grayi=wmobra and C. incisa go further south. C. ornata, from Panama to Tehuantepec, has been found by Forrer also near Mazatlan, with C. pulcherrima. I do not know of a single locality for Chrysemys on the plateau, or to the east of it, except for C. mobiliensis, which goes from Texas into the lowland of Nuevo Leon. ’ This scarcity of Water-Tortoises in Mexico is rather puzzling. On the plateau Cinosternwm alone is found, and these thick-shelled box-like creatures are, moreover, the only kind which can with- stand the buffeting to which they are subjected in the torrents into which the rivers of the slopes of the plateau are converted in the rainy season. The Tortoises hide then under the boulders in the stream. Chrysemys shuns such waters, and neither it nor Cinosternum occurs in those rivers which carry much sand. Chelone viridis was laying during July and August on the coast of Guerrero and Oaxaca. résumé of the Distribution of Mexican Chelonia. The Cinosternide, taken with the closely-allied Dermatemydidee and Chelydride, are autochthonous Americans; the first a Sonoran, the second obviously a southern group so far as the present distribution is concerned. Both Chelydrids and Dermatemyds are known from the Cretaceous of North America. The three together may well be regarded as originally northern and ancient. The same applies to the Testudinide, the only family which has, recently, sent a United States Chrysemys into the Antilles and a South American into the Windward Islands. The Testudinide, plentiful in North America, scarce in Central, and with still fewer species in South America, have clearly come from the 1905. ] - AMPHIBIANS AND REPTILES. 211 Northern continent. The earliest, probably all of the genus Testudo, have been found in the mid-Eocene of Wyoming and New Mexico; since Oligocene in Europe, still later in India. With this remote occurrence in ancient Sonoraland I couple the most important fact of the Galapagos Tortoises. They are a strong indication of the former, let us say Oligocene, extension of land considerably to the west and south of the present Central America. We shall find this idea supported by Iguanide. Now North America possesses but the single 7. polyphemus in the South-eastern States, and South America has only 7’. tabulata. Something has gone wrong with this genus, which has flourished in the Miocene of Dakota, Nebraska, and Oregon, as has been the case with so many mammals which started and flourished in the States and aie now restricted to the Old World. LACERTILIA, GecKoNID#.—The distribution of American Geckos is almost entirely tropical. The greatest number and diversity of species occur in the Antilles, in Northern South America and the adjoining Central America, whence few have spread into the warmer parts of Mexico, avoiding the plateau. North America has received only Spherodactylus notatus from the Antilles through the Bahamas into Florida, and Phyllodactylus tuberculosus into California; this species is the commonest Gecko in Mexico, ranging strictly along the Pacific slope to the Isthmus of Tehu- antepec and thence to Nicaragua. Spherodactylus sends only three species into Mexico: S. glawews to Salina Cruz and into the State of Vera Cruz; the Central American S. torguatus and the Antillean S. anthracinus are recorded from the same State, and S. torquatus has been described from Mazatlan. Gynno- dactylus sumichrastt reaches the Isthmus, and Zhecadactylus rapicauda, of Yucatan, Antilles, and southwards, is said by Cope to have been recorded from Guadalajara, a very doubtful locality. Phyllodactylus tuberculosus is common in the villages of Southern Oaxaca and Guerrero, where it is known as “ Pata de bueye,” i. e. ox-foot, because of its peculiar digits. The general name for Geckos is ‘‘Salamanqueza” or ‘‘Salamanquezca,” which name, how- ever, also applies to the slippery Mabuia and Humeces. I found the same Gecko on the trees of dense forests near the coast of Guerrero. Spherodactylus glaucus is typically xerophile. As in Spain and Portugal, all Geckos are considered extremely poisonous. EuBLEPHARIDE.—This small and very scattered family (in West Africa, Somaliland, India, Transcaspia, and Persia) 1s represented by three species in Mexico, a few others occurring in Panama and Ksuador. Hublepharis variegatus is the northern offshoot, from El Paso to the Gila River and California, probably also in Sinaloa. L. fasciatus is known from Ventanas, north-west of Mazatlan. These are apparently typically xerophile, like the 14* 212 DR. H. GADOW ON MEXICAN (June 6, two Asiatics of the same genus. But Coleonyx elegans is distinctly a forest form, I found it a few miles from the coast of Guerrero in a moist patch of thick lowland forest on the ground under stones and rotten stumps. A typical Central American, ranging through the Pacific and Atlantic Tierra Caliente of Mexico, from which country it had hitherto been recorded only by Sumichrast from near Orizaba; extending south to Costa Rica. Ieuanip# 7.—It is not profitable to speculate upon the or iginal home of this family. The overwhelming majority of genera and species is American, from Mexico to Brazil It is well known that the Galapagos possess the semimarine Amblyrhynchus and Cono- lophus, that a few species occur in Madagascar, and Brachylophus fasciatus in the Viti and Tonga Islands ; further, that an Iguanid allied to the genus Jguana existed in the Hocene or Oligocene of Europe, and that therefore attempts have been made to explain the present scattered distribution of the family by a formerly subuniversal range; in other words, they are a very ancient group. Concerning America, it is significant that only a few species of Sceloporus and Phrynosoma extend into the United States, although far northwards, Of the large genus Anolis, only A. carolinensis enters Texas to Carolina, but it is also found in Cuba. Mexico itself, Central America, and the Antilles are rich in genera and species. These Iguanide can be divided into two groups :— A Sonoran set, comprising genera which are essentially xero- phile and humivagous, with depressed bodies and short tails. None of these reaches far into Central America, and none has entered the Antilles. Crotaphytus, Holbrookia, Uta, Phrynosoma, Sceloporus, which, in the order mentioned, extend from California and Arizona southwards, with decidedly Pacific or Western pre- dilection ; only a few Sceloporus, those which have spread into the Atlantic Tierra Caliente, continue further into Central America. Nearly all these southern Sceloporus are fitted for arboreal life, less depressed in body, and suited to a moist climate, be this hot or cool. They lead thereby to the second set, which are essentially arboreal, mostly inhabitants of forests or of rocky bush-land; all southerners, with their centre in Central and South America, extending into the Mexican Tierra Caliente, with prevalence on the Atlantic side, and two* have allied genera or species in the Antilles : Anolis *, Iguana*, Basiliscus, Lemanctus, Corythophanes, and Ctenosaura. Of course there are transitional forms, for instance the genus + Iguana, or Guana, is a native word applied to-the Iguana; but where this does not oceur, the name is given to Ctenosaura, for instance at Cuernavaca. The Zapotec name of Ctenosaura is Tilcampo ; Basiliscus and Co1 rythophanes are called Teteréte. At Rio Balsas, scaly lizards, e.g. Sceloporus, are distinguished as Chintéte. 1905.] AMPHIBIANS AND REPTILES. 213 Ctenosaura, and we will not discuss the question which of the two groups is the more primitive ; apparently the latter, but this can be contested. Crotaphytus, a typical old Northern Sonoran genus with several species in Western United States. C. wisliseni ranging from Oregon and Nevada into Sonora and Chihuahua; C’. collaris also into Nuevo Leon. Holbrookia, from Texas and California into the dry parts of Northern Mexico. H. maculata into North Sonora, H. texana to Monterey and Lerdo near Torreon. I found it running about swiftly on the almost barren shaly ground near El Paso. A. pro- pinqua from Texas to Presidio near Mazatlan. Uta, with most species in South-western United States and in Lower California. U. elegans from Utah to Texas and Sonora; U. stansburiana from Utah to Torreon. U. lateralis from Presidio and Tres Marias Islands and JU. bicarinata are Mexican, from Presidio to Tehuantepec, and everywhere between these places. Otherwise strictly confined to the western side of the plateau and the coast, it has entered the plateau at Cuernavaca and Puebla. J have almost invariably found it on the stems and branches of low trees, upon which they flatten themselves like arboreal Sceloporus; rather remarkable, since the other species are so decidedly dwellers on the sandy or stony ground. Very important is the occurrence of a species, U. auriculata, on the Revilla Gigedo Islands, 280 miles south of Cape Lucas, Lower Cali- fornia, and nearly 350 miles from the coast of Jalisco. This genus is typically Sonoran, with its centre around the Gulf of California. Phrynosoma, “ Animal rey,” or “ Camaleén,” or ‘ Escorpién.” The original centre of this genus is undoubtedly Sonoraland, whence it extends now over most of the Central, South-western, and Western States of North America and over the whole of Mexico as far as Guatemala. Ph. cornutwm, modestum, and orbiculare are, in Mexico, scattered over the plateau. Ph. asio is the most southern and at the same time the largest and most handsome species, ranging from Colima to Guatemala. Stejneger and Cope have already remarked on the ‘‘metachrosis” of Ph, douglasi. I have found Ph, modestwm near El Paso of exactly the same delicate French-grey colour as the little slabs of Cretaceous limestone with which the hills are strewn; the same species at San Marcial and at Rincon in Mexico, on the red and sandy voleanic rubble of that hilly desert region, were of the same pronounced red tint. Examination with a magnifying- glass showed the spirit-specimens to be covered with the iron- stained red sand, but those which I have brought home alive show this same red colour also to be that of their genuine skin. Sceloporus may well be called the most characteristic genus of Mexican Lizards. Of the 34 species recognised by Boulenger, 28 occur in Mexico, between El] Paso and Tehuantepec. Only A live in the United States, and only 3 or 4 are found south of 214 DR. H. GADOW ON MEXICAN [June 6, the Isthmus of Tehuantepec, and are restricted to Central America. Some species hate a very wide, others a very limited dis- tribution. The majority combine humivagous with climbing habits, and show great adaptiveness to the nature of their surroundings ; for instance, S. scalaris and S. ceneus do not climb the trees beneath which they live, but prefer the grassy ground, and they are equally at home in the moist, clouded pine-forests and on the more barren, grassy and lava-strewn slopes up to the snow-line. S. variabilis prefers the wooded lowlands of the Atlantic side, and likewise does not climb, loving the banks of rivulets and well-herbed ravines. Others, e. g. S. torquatus, are found only on rocks, stone walls, and buildings; they are swift. Some—and these are the most depressed in body—are rather sluggish, e.g. S. spinosus, and spend most of their time on the ground between spiny growth of hedges and low trees, which they ascend a little way, in short rapid jerks, when alarmed. Lastly, S. microlepidotus is truly arboreal, ascending the trees in the morning, with the sun, right into the green tops, where they hunt for insects. This species has the greatest possible alti- tudinal range ; from the hot country of Southern Oaxaca, only a few hundred feet above the level of the sea, to the upper tree-line of Citlaltepetl, about 13,500 feet elevation. Many species are viviparous. According to my own observa- tions, the following: acanthinus, eneus, formosus, microlepidotus, scalaris—all gravid in the months of July to September. Iguana rhinolophus is interesting for various reasons. It is the largest Lizard in Mexico, attaining a total length of about 5 feet. Always arboreal and aquatic and truly tropical, it occurs in the whole of Central America, but in Mexico, north of the Isthmus, only in the States of Vera Cruz and Southern Oaxaca, everywhere strictly below the Plateau, and on the Pacific side it has been recorded only from Manzanillo near Colima and near Mazatlan; undoubtedly also near San Blas and in the lower reaches of the Balsas, but this information I have only from hearsay. I never found it in Guerrero. The creature requires permanent, rather sluggish rivers, or deep pools in the savannahs. They climb about in the trees, eating the succulent leaves, which they bolt without much chewing, for instance those of the guava tree. Favourite places for resting are the branches which over- hang the water, into which they plump with a loud splash, sinking at once and remaining at the bottom for many minutes. Whilst the adult are dusky, the young are grass-green and are frequently found in the tall grass at the edge of a pool. They at once take to the water and swim to the bottom, with their legs laid back and propelling themselves, like newts, by rapid undu- lating motions of the tail. The eggs are buried in the soft soil, among the roots of a tree, always near the water, in the month of May; by the end of July they are already hatched. They are known as Guanas or Iguanas. Ctenosaura acanthura is a common Lizard of the hot and warm 1905. | AMPHIBIANS AND REPTILES. 215 countries, from Yucatan to Tampico and on the Pacific side as far as Southern California. It does not take to the water, preferring rocky bush-land or savannahs. According to the locality, it makes its home in a hollow tree, in the roof of a house, or on the ground, where, among rocks or trees, it digs out a permanent burrow, heaping up the soil above and around it. This “Iguana,” or ‘‘Tileampo” of the Zapotecs, is very fierce, bites, and lashes out furiously with its tail. Its food is varied, from all kinds of lizards, snakes, and insects to grass and flowers ; in turn the Tileampo itself is much prized as an article of food, and in the markets fetches more than two fowls. The young are entirely vivid green ; in their second and third years the back and sides develop blackish patches upon the green ground, and in this stage they are often very beautiful. With approaching maturity the green colour disappears, being en- croached upon and then entirely suppressed by the spreading black and brown pigment. But in certain localities, where these lizards live amongst luscious growth of evergreen trees, many individuals retain their green livery throughout life. I caught a young Tilcampo, which belonged to a green family, as shown by the parents, at San Juan Evangelista, on the eastern side of the Isthmus, where the green colour was normal; within less than 18 months my captive had lost all the green, and had assumed the dusky brownish and patchy garb. Ctenosaura quinquecarinata.—This much smaller, brownish- yellow species is not arboreal, ranging from Honduras into the southern hot parts of Oaxaca. It becomes very tame, takes a varied diet, and defends itself in its burrow by sideward strokes of its spiny tail, much like the Indian Uromastix, which it greatly resembles in habits and outward appearance. Basiliscus vittatws.—Closely allied to the Central American B. americanus, ranges from Ecuador into the Tierra Caliente of Mexico, where it is, however, restricted to the southern part, not going further north than Cordoba. Until I found it at Teque- sixtlan and Tierra Colorada in the centre of Guerrero, it was not known from the Pacific side. The locality “ Orizaba” in the ‘ Biol. Centr.-Am.’ is erroneous; Sumichrast states clearly, and correctly, that this species extends only up to 3300 feet. The ‘“‘ Pasarios,” its universal Spanish name, lives always on the banks of rivers or pools. I generally found them busy on the ground close to the water’s edge, or upon a low overhanging branch. On the slightest alarm, they plunge or rush into the water, rapidly running over the surface in a slightly erect position, splashing the water with their long-toed hind limbs and the long wriggling tail, whilst the arms are adpressed to the body. They do not dive; arrived on the other side, they climb up the bank and hide in the tangled vegetation. The usual statement that they propel themselves by rapid strokes of the fore-limbs is erroneous, and the notion that the high dorsal and caudal crests, which adorn the male only, serve as a sail is a fable. Corythophanes hernandexi, ‘'Veteréte.,-—From Chiapas and 216 DR. H. GADOW ON MEXICAN [June 6, Yucatan to the State of Vera Cruz, absolutely confined to the Atlantic Tierra Caliente, in forest-land ; arboreal, or rather amongst shrub-like trees, the brown bark of which this curious-looking ~ gentle lizard looks to for protection. It feeds upon insects. Lemanctus.—The two Mexican species are excessively rare, perhaps because they live higher up in the trees, where it is then next to impossible to discover them. J. serratus is known from Campeche and the States of Vera Cruz and Oaxaca without localities. L. longipes, hitherto known from Jalapa only; all the more re- markable is the solitary specimen which I found amongst a collection sent to.the Field Columbian Museum from the State of Colima. Anolis, with at least a dozen species in the Hastern or Western Tierra Caliente. A. nebulosus has the widest range in Mexico, from Tehuantepec to Jalapa, and to Ventanas on the west; I found it not only on the coast of Guerrero, but also on the Nevado de Colima, up to at least 7600 feet, together with A. lio- gaster. A. gadovuw at Tierra Colorada, in bush-land. The Anolis seem to spend most of their time on the lower branches of shrubs and trees or amongst the rank herbaceous vegetation, waiting for insects, and trusting to not being seen when basking. Especially when they have become excited by being pursued, the males stretch out their mostly beautifully-coloured gular sac. None of the Mexican species which I have observed displays any marked change of colour like the A. carolinensis, the “‘ Chameleon” of the Americans. TEs1IDz* are clearly a Neotropical family, with several dozen genera in South America. Of all these, only Ameiva and the closely-allied Cnemidophorus extend through and beyond Central America: Amewa into the Eastern and Western hot-lands of Mexico and into the Antilles; Cnemidophorus through Mexico into the United States, where C. sealineatus has spread over nearly the whole Union. This genus is entirely terrestrial, preferring sandy districts with bush-land; only C. guttatus is a typical inhabitant of the lowland forests of Vera Cruz. The Mexican species avoid the high plateau, 5000 feet being about the upper limit. The only exception is made by C. guluris, which has been credibly recorded from Guanajuato, and of which I have examined specimens collected by Dr. Meek close to the town of Puebla, which lies at an altitude of more than 7600 feet, higher than the Valley of Mexico, where Cnemidophorus does not occur. Concerning distribution and variation, ef. my paper, “ Evolution of the Colour-pattern and Orthogenetic Variation in certain Mexican Species of Lizards, with adaptation to their surroundings,” Proc. Roy. Soe. vol. lxxii. p. 109 (1903). Ameiva wndulata, the only species in Mexico, is an inhabitant * From the Aztec “teco-ixin,” ¢. e. Rock-lizard, the name of Sceloporus torquatus, misspelt and misapplied. The Zapotecs and Mazatecs call Cnemidophorus and Ameiva Zumbichi and Cachumbo. 1905. ] AMPHIBIANS AND REPTILES. 217 of the hot, well-wooded parts of Guerrero, Oaxaca, and Vera Cruz, whence it extends far into Central America. It is far less quick than Cnemidophorus, and I have found it invariably in the vicinity of water. AnGuIDa,— Anguis, with its sole species fragilis, and two species of Ophisaurus s. Pseudopus (Morocco to Burma) are the only members of this family which are not American, and even the third species of Ophisaurus, O. ventralis, lives in the United States. The countries now richest in Anguide are Mexico, Central America, and the Antilles; a few extend into South, and a few, Gerrhonotus with the Ophisaurus, into North America, where the latter is widely distributed (also recorded from Jalapa). Diploglossus is peculiar to the mountainous regions of Mexico ; D. steindachnert from Orizaba, Jalapa, and Guatemala. The related genus Celestws in Antilles and Central America. Gerrhonotus is the main genus, eight species of which occur in Mexico, entirely in mountainous districts or on the plateau; they are consequently absent in the hot lowland forests, and references to Vera Cruz and Tehuantepec do not apply to such towns but to unknown places in the state or district. G. ceruleus has the widest range, from British Columbia and Colorado along the Pacific side of Mexico to Costa Rica. Most of the species live on the ground, in the oak- and pine-forests, preferring clearly a moist and by no means warm climate. G. antauges ascends Citlaltepetl to an altitude of more than 12,000 feet, in the pine-forests, or in the grass near little streams, and higher up amongst the tussocks of grass, basking on the top of such a tussock and making its home among the roots or in the mass of last year’s rotting blades. In such a place they disappear easily, although they are not quick. The same applies to G. inbricatus. G. gramineus, delicately light green above and yellow below, is arboreal, ascending the highest trees in search of insects and making its lair in hollow trees of oak, pines, and arbutus. They all are viviparous, live on insects and worms, and lose their shyness a few hours after having been caught and handled. XENOSAURIDH.— Nenosaurus grandis alone is recorded only from the mountains near Orizaba, Cordoba, and Oaxaca. HELODERMATID#.—The sole genus Heloderma, unless we include Lanthanotus of Borneo. H. suspectum of Arizona and New Mexico, and H, horridwm of Mexico. The notion that Heloderma is a dweller on arid mountains is quite erroneous. It is restricted to hot lowlands with sandy ground. Most of Arizona is high and dry tableland, and there is quite a trade in “Gila monsters,” but, so far as I could find out, they all came from such terribly hot and low sandy places as Yuma, on the lower reaches of the Gila river, and from similar localities in Sonora. H. horridum is 218 DR. H. GADOW ON MEXICAN [June 6, stated by Giinther to have probably a wide distribution in Mexico. The fact is that it has hitherto been recorded only from the following localities :—near Tehuantepec, and near Presidio by Forrer; and in the museum at Mexico is a specimen from Apatzingan in Michoacan. It is very local. In Guerrero and Oaxaca, Colima and Jalisco everybody speaks of the “ Escorpién.” “He is unkillable unless you crush him with a big stone. When at last secured in a cleft stick, his poison dropping to the ground causes all vegetation to wither for yards around. There are two kinds in Guerrero, one brown, the other black and yellow; nocturnal, hidden in the daytime beneath the stump of a tree or under a boulder; estivating during the dry season.” Hundreds of times have I offered much money, even for being taken to its lair, but all in vain. The only place where I personally know it to occur is Juchitan, not far to the north-east of Tehuantepec ; in the museum at Oaxaca is a stuffed specimen, a monster about 24 feet in length. At last I thought I had run the beast down, when at Zapotlan in Jalisco. The poison, the sluggish fierceness, difficulty in killing it, all this sounded favourable. We found the Escorpion, but it was the harmless, gentle Gerrhonotus, which for some unaccountable reason is feared as very poisonous! The Zapotecan name of Heloderma is “'Talachini”; the Aztecs called it “ Acaltetepon.” Hernandez states that “it is found in Cuernavaca and other hot districts.” But it does not occur anywhere near the State of Morelos, unless the huge figure of a lizard carved out of a rock near Cuernavaca is evidence! The last three families taken together form a very ancient group, which seems to have its original centre in the old Sonoraland, or let us say in the old Sonoran + Central American + Antillean landmass. The absence of Anguide in Eastern Asia suggests the spread from North America into Europe and Asia across the polar region, unless we prefer the problematic bridge across the Northern Atlantic from the Antilles (which possess their own genus Celestus with several species) towards the Mediterranean. Scincip£2—Of this large and almost cosmopolitan family America possesses the smallest number, and it is significant that the number of forms decreases from North to South. Mexico has about 10 species. They may perhaps be divided into a Northern lot, Hwmeces, which ranges from the middle of North America over the Mexican plateau and its bordering mountains; and into a Southern set, Jabuia and Lygosoma s. Mocoa, which love the hot country, extending far into tropical South America, with species in the Antilles, in Mexico restricted to the Southern States east and west. Mabuia agilis is fond of basking on shrubs and it even climbs trees, hiding under the bark. Like Lygosoma laterale it hunts in the dusk. Hwmeces, of which I have observed only lynwe and fuscirostris, prefer mountain forests, where they live on the 1905. | AMPHIBIANS AND REPTILES. 219 ground, basking on the fallen leaves, between which, and in the soft humus, they wriggle away with perplexing agility. ANELYTROPSIDA, an artificial assembly of a few degraded Scincoids in Madagascar, Tropical Africa, and montezume. ss validus . Sceloporus scalaris. Coluber triaspis. 5 microlepidotus. Crotalus horridus. Uia bicarinata. Gerrhonotus ceruleus. Cinosternum pennsylvamcum. IL. Essentially hot-country species which have ascended; and it is remarkable that most of these are not found on the plateau proper, although they ascend the surrounding mountains, up to an altitude equal to or surpassing that of the plateau. This fact seems to indicate that the respective species are still continuing their upward spreading, or that they have conquered these mountains comparatively recently. This fits well with the suggestion expressed on p. 244 that the Southern or tropical fauna of Mexico represents for the greater part the most recent immigrants. The Sierra Madre del Sur affords a good illustration. It is separated from the plateau by the depression of the basin of the Rio Balsas. Tropical species coming from the south can surge up to the Sierra, and they have ascended its higher mountains (e. g. those of Omilteme, Amula, Cerro de S. Felipe near Oaxaca), and the backbone itself is of no mean height; but then comes the descent into the hot basin, then again the ascent of the plateau. A tropical species, which has succeeded in acclimatising itself to life on the Sierra, will have to “undo” this hardening, become tropical again, and lastly once more ascend and accommodate itself to a cool climate. Of course all this can be done, but it takes time. The same applies to the fauna of the rather isolated Volcan and Nevado de Colima. The ranges of mountains which border 1905. } AMPHIBIANS AND REPTILES. 231 the great plateau are rather abrupt and in many parts are even higher than the plateau itself, so that to gain the latter would imply a descent. There are, as mentioned elsewhere, p. 240, regions which offer a gradual, easy entry, and they have facilitated the exchange of many species, but not of all, and of course not in other districts. Species found in the hot country and on high mountains ; those excluded from the plateau are marked * :— * Hylodes rhodopis. * Leptodira albofusca. * Anolis nebulosus. _* Xenodon rhabdocephalus (2). * Zamenis mentovarius. | Geophis chalybea. * Rhadinea vittata. | Trimorphodon upsiton. * Leptophis mexicana. | Llaps fulvius. * % diplotropis. | III. Lastly there are some species which are difficult to group, whether they have descended or ascended. For instance, most kinds of Hylodes live rather high up; they want permanent moisture, and this 1. rhodopis gets on the high mountains and in the hot forests of the Atlantic side; only a very few returns have been made from the truly temperate zone, and it isnot known from the plateau. Hyla eximia. | * Sceloporus formosus. * Bufo intermedius. | s acanthinus. * Hylodes palmatus. | ss spinosus (°@). Sceloporus variabilis. | Coronella niicropholis. The list (infra, pp. 232-233) contains 70 species, of which 8 (Diploglossus, Xenosaurus, 1 Zamenis, 3 Leptophis, 1 Drymobius, and 1 Sceloporus) may be deducted as probably not ascending beyond 6000 feet. The remaining 62 species, out of a probable total of 250 for Mexico from between the Isthmus of Tehuantepec and the line Mizantla to Mazatlan, represent about 25 per cent. Of these, again, 30-32 (13 per cent.) seem to be restricted to levels above 7000 feet. These have been marked with an asterisk (*). If we add to them the following 10 species, which seem to be restricted to the high plateau, 6000-8000 feet :— Spelerpes morio (also from | Tropidonotus variabilis, _ “ Jalapa”), : scaliger, Scaphiopus multiplicatus, Homalocranium bocourti, Bufo compactus, | Crotalus miliarius, - Lyla miotympanum (2), | 5 | salvar, Phrynosoma orbieulare, | we get a total of about 42 species, equalling 17 per cent., as restricted to the cold and cool zones (cf. p. 228). [June 6, DR. H. GADOW ON MEXICAN 232 * OOS8 * 0OS8 * 0008 3008 “OOILXO Tl Ivou ‘oosnly ap BILAIG | * 00¢8 O0¢8-O08FE 008-0007 | 0008 ‘UR, -yekep op BAILEIG | 0088-0064 | 00g8 *soperyselcy SOT ap arquing a | goes ee iia %000°TT-0064 * 0008 ee ae wee * * Beear OOlL 0828-0469 OOTL x 0068-0099 {| — oot, Gor era Pe x* ‘osuroued[iyg : FO ISOM BEAD) ‘QIPeT BITOIG OSEAN ‘auled UC, ‘osulued iy pus vovxtg Teamgjeq “Ing Top e«pey BLIALG “BTU 0008 “BOBXVO) qeau ‘adrpay *g ap O10) 000‘F1-000‘0T 00¢°E1T-0088 000‘0T-0006 008°6T-0008 009°6T-0008 0006 "yodezeTI1) ee we ee = Si afius 2 s2qugur.tna $2.UD1D9S SAUD Pee sadajouagoy : _snqopydajowoyut * snsowm.of” wadsp smuayqupaD wnoLwoh Hercueeesescsssssis? qupaang Tega sngpnb.woy sn.odojaag T= DUDUAAIUULO LL ee DUVIOI DY DUDT puna Diy oad bod pnO 90D Daynuy snjphjappowoy, 2 SNUDIVLAUL $AZ0I0L0G.L0G pond Ads TITEL sndoysihig SSgreH Camcntien ‘ snqpujod nuedde COGbb0 000 sedopoyt sapojvhyy snus smipauLsazur ofing TE UMM IID.» 20 oan soc unuribry puUojshi quip poe voc snqunuogyD sdasoyonuyng eds . "79q ~ snwagdowya snso.day poo oop sisuaqnzis10 sadwajady SINgDUUAd SNLLOY F [-popnput gou st ‘seooqvorz pue e[qeng ‘oorxeyy jo AoqeA *4 -a “neozeyq ous, | ‘QU0Z 100) ay) mYynM sIUMUNndT UNnwarT moLf pap.Loo—. soLoady fo ISU] AMPHIBIANS AND REPTILES. 233 1905.] “eon J, “COnOT, * 0098 “Bonyoy, 0088 008 00g8 K 000-0099 0002 0004-0089 OOOTT 0009 ¢ 0009 ¢ 0008 0084-0009 0006-0088 00S8 SnyDNpUN srsayooT aottand oon cananoda6 OG snoryittag SNYDUASULY SNIDZONY UNPIUDIOLID UWILUDLI0) NWO FL = DIDIPUL DOULP DY pomacran wojisdan Uuopoydsoweny, ae pasnfog py vAvpopdaryT snuafiquipbhinu snigowhiq PET AEG nts eRe puporeau snynaurrg sunjsoid povare ee sidouzojdip svydojdaTy pecpcoees ~ vadoa snyqnubouyosT mee yunugnd sidajoun ft ae DULOSO.LOLYD Laqnjon cress pagy vo ga sss pasa svydoay SnywawiIDM) SnovUmOLg Heiress emanaen) DON)OT, snspu 021910; “sip pydasopqny. wopouay Se aa ee SNwpaojzUaUL pap ongpRo ees wUnynIh suaung Bee as aupupob Hotere granulo messes ppasphounjaut ge snjpoydososhya SLLopwas snyouoprdo.wy, = sup snr.psoua “ 1uauyovpurags snssoj bo) dug Heeiercssststemgomgau os aoppb syour Hieeeeeesssss st 99.49 80.U10049 ee ee ee eee eee eee aowhy sana ee eee eee eee eee nee snpoyde00u DOO Song OMICS cngaeicnn snamartquir wee ee tee eee tee ee snauvwnsb —— sive she viele nse velo elses velsjue saddap sabnnyun snzouoy.t4a) SRMND, DUWOSOUN LY T 234 DR. H. GADOW ON MEXICAN [June 6, GENERAL CONCLUSIONS. 1. Hvolution of Middle America. We have seen in the review of the Aimphibian and Reptilian fauna of Mexico that it is composed of Northern and Southern immigrants; that a considerable number of the northern group can claim to be old, autochthonous Nearctics ; that some families, genera, or species have also representatives in the Antilles, and that most of these forms point unmistakably to Central America, or even further south, as their original home; lastly, that but few Antilleans belong to a northern stock. The explanation lies in the geological history of this part of the world. J restrict myself on purpose to this part, lest such an inquiry should lead to a discussion of the whole globe since the first dawn of Amphibian life in some Paleozoic country. Our present task limits itself to the Tertiary period. It is doubtful whether any of the genera in question are older than the Eocene, but not a few can be proved to have existed in our region in the mid-Miocene epoch; and it is surprising that they should date so far back. Lastly, there was no Central America in the Cretaceous period. The building up of Mexico and neighbouring countries seems to have taken place as follows, so far as I can gather from the writings of A. Agassiz, Suess, Lapparent, R. T. Hill, J. W. Spencer, J. W. Gregory, C. Sapper, and José G. Aguilera *. The accompanying consecutive series of maps illustrate my abstract conclusions, and only in this abstracted sense can claim originality. Mexico came into existence during the Lower Cretaceous epoch. To a nucleus of land, Sierra Nevada and California, were added the Rocky Mountains and the bulk of the Mexican Plateau. This large complex I call the Old Sonoraland, It is important to remember that it was separated, during the Upper Cretaceous epoch, by a broad belt of sea from the eastern and northern parts of North America. A third mass of land existed as Brazilialand. In the meantime appeared Antillean lands, and, possibly in sympathy with the east to west trending mountains of Honduras * SuEss.—Das Antlitz der Erde. Dr LapparEent.—Traité de Géologie. R. T. Hint.—*‘ The Geology and Physical Geography of Jamaica: Study of atype of Antillean development.” Bull. Mus. Comp. Zool. Harvard, xxiv. (1899) pp. 1-226. See also other papers in same Bulletin, xvi. (1895), and in Amer. Journ. Sci. vol. xlviii. (1894). J. W. Spencer.— Reconstruction of the Antillean Continent.” Bull. Geol. Soc. America, vol. vi. 1895 ; and Geolog. Mag. 1894, pp. 448-451. A. Agassiz.—Reports of the Results of Dredging ...... by the‘ Blake. Mem. Mus. Comp. Zool. x. (1883) no. 1, p. 79. J. W. Grecory.— Contributions to the Paleontology and Physical Geography of the West Indies.” Quart. Journ. Geol. Soc. vol. li. (1895) pp. 255-312. J. G. AGUILERA.—* Bosquejo Geolégico de Mexico.” Instituto Geolés. de Mexico, pt. 4 (1895) pp. 1-270, with maps. C. SAPPER.—“ Sobre la Geografia fisica y la geologia de la peninsulade Yucatan.” Inst. Geol. Mexico, pt. 3 (1896). 1905. | AMPHIBIANS AND REPTILES. 235 and Guatemala, also the Mexican Sierra Madre del Sur. These parts were in time annexed by Sonoraland. By the late Hocene, conditions were so far consolidated that there existed the present North American Continent, eastern and LAND DURING EAREY CRETACEOUS EPOCH. It Diagrams to illustrate the contours of Mexico at different geological ages. western halves joined, and the latter extending southwards as the present Mexico and part of Central America. Brazilia had grown into South America, but the two continents were still separated, the Atlantic and Pacific communicating across the present Isthmus of Panama and probably further north. 236 DR. H. GADOW ON MEXICAN [June 6, Late Eocene, or early Oligocene, times mark a period of considerable local subsidence which drowned the Antillean land, or islands, except their summits. Late Oligocene, or early Miocene, mark a period of considerable elevation with most important Text-fig. 32. DURING MIOCENE Diagrams to illustrate the contours of Mexico at different geological ages. results :—Establishment of the continuity of North and Central with South America, and a continuous mass of land from Central America, north and eastwards, comprising the Greater Antilles and the southern end of Florida. For this Central Land (Antilles + Central America proper, and adjoining parts of South 1905. | AMPHIBIANS AND REPTILES, 237 America, viz. Colombia and Venezuela) I use the name of Great Antillia, the term Antillia having already been used by others. The present Gulf of Mexico remained below the sea, and was larger than it is now, covering the Atlantic Tierra Caliente of Mexico, Yucatan, and, according to Hill, the main part of Florida. If correct, the latter point is important. It seems also probable that the Mexican-Central American land, during the Miocene epoch, extended considerably further west- wards than the present Pacific coast, taking in with almost certainty the Revilla Gigedo Islands. Late Miocene, or early Pliocene, comprise a time of subsidence, resulting in the present features. Severance of the Antilles into the present islands, which since have undergone comparatively unimportant changes of shape and extent; separation of Florida. Lower California became a peninsula, owing to the formation of the Gulf of California. The Revilla Gigedo Islands, still later the Tres Marias, are remnants of the subsiding land. Yucatan appears at the beginning of the Pliocene epoch*. The Isthmus of Panama is limited to its present narrow dimensions. A few words remain to be said about the volcanic activity and other changes affecting the configuration of the Mexican Plateau. A tremendous dislocation, at the latest in Kocene times, produced the Eastern Sierra Madre, composed entirely of Cretaceous lime- stones, raised up high, forming the elevated eastern rim of the plateau, and falling off abruptly towards the Atlantic lowlands. In the Eocene epoch began also the enormous outburst of volcanism, raising the Western Sierra Madre, piling up gigantic masses of igneous rocks, mostly andesite, and lavas, which con- tinued to spread over a vast part of the country during most of the Miocene epoch, and, more locally, even in historic times. Most of the plateau is now covered with the Quaternary debris, sand, d&e., which overlie the eruptive masses and the older calcareous or limestone formations. These accumulations of more or less sandy soil form plains, mostly treeless. They are of great extent, in the northern half, from Texas to Zacatecas. In the middle, say from Guadalajara to Puebla, exist a great number of smaller plains or “valles,” that is to say fertile plains, interrupted or partly surrounded by the outcropping hills of voleanic formation, and they contain a fair number of lakes. In the south of Mexico, in the States of Oaxaca and Guerrero, such plains are rare or absent. Trees are scarce or absent on the plateau ; it isan idle fable that it was well-wooded in historic times. The bordering high Sierras and their slopes are well- wooded, densest on the moist, Atlantic side. ‘The eastern, southern, and western Tierra Caliente is covered with luxurious growth, either forming continuous forests or showing the features of savannahs. The plateau is dry, verging towards prolonged droughts, interrupted by few, occasionally torrential, rains. The Atlantic * See footnote to p. 242, 238 DR, H. GADOW ON MEXICAN [June 6, hot-lands and the eastern slopes of the States of Vera Cruz and Chiapas are very wet, with a very long and abundant rainy season, interrupted by ashort dry time inthe winter. The Pacific side is much drier; the actual amount of annual rainfall is considerably less and the dry winter period is much longer. The plateau rises from less than 1000 feet near Laredo, and 3800 at El Paso, gradually to about 6000 at Aguas Calientes and Querétaro, and above 7000 at Mexico City and Puebla. The highest masses of mountains, bordering the plateau, lie in the south-east, south and west, culminating in the snow-capped peaks of Citlaltepetl or Volean de Orizaba, Popocatepetl, Nevado de Toluca, and Nevado de Colima, 2. Immigration and Spreading. Obviously these physical conditions influence the fauna now ; what they were like in bygone ages we can only surmise. Ranges of mountains are by no means alway s barriers ; on the contrar they help the dispersal along the lines of their long AXES. Regions covered by the sea are of course not available. The same applies to districts which are subject to voleanic eruptions. This is very important for Mexico. Not only the Western Sierra Madre with its continuations to Colima and thence towards Puebla, but also almost the whole of the plateau became covered with eruptive masses, and, considering the immense extent of this terrain, a long time must have elapsed before it became available for plants and animals. We may well ask, what remained of the country as suit- able for life. Of course, probably, there were archaic tracts standing out, not affected by these revolutions, but these gneisses, schists, and granites form scattered enclaves. I think it was the Pacifie strip—Sonora, Sinaloa, Tepic, and part of Jaliseo—which was not affected ; m fact, the Pacific slopes, together with the land which has since sunk below the Gulf of California. On the eastern side, part of the plateau did not suffer from eruptions, but the land was still narrowed; there was no Atlantic lowland, this being during the whole Miocene epoch, and even later, still below the sea. Consequently we have as available land the western strip as the least altered remnant of Old Sonoraland, and the present eastern limestone belt, beginning with a broad basis in Texas, and extending through Coahuila and Nuevo Leon southwards, narrow- ing down towards Oaxaca. These were the two belts of land available for spreading southwards. Obviously the Pacific belt is the older of the two, the north-east of Mexico, with Texas, being late Cretaceous terrain. Once arrived in the south of the plateau, there was the essentially granitic, gneissic, and older Cretaceous terrain of Guerrero and Oaxaca, not so much overlaid by volcanic masses. Thence the Great Antillia afforded easy access into the present Antilles. But it was a long way round from the North. The spreading from South America into this same Antillia was easier in this respect. Later immigrants from the North into Mexico are those of the 1905. ] AMPHIBIANS AND REPTILES, 239 plateau, which by climate and every other physical feature is a direct continuation of the more northern countries. Hence the imperceptible change from Arizona, New Mexico, and Texas southwards. The political frontier between Mexico and the United States is no boundary whatever for our purposes. For northern animals and plants the drier climate, not so much the annual mean temperature, of the plateau suggests this as a natural limit, but not a few northern forms, even the same species, have adapted themselves to life in the hot lowlands and have extended their range far south, even into South America. With the original natives of the latter continent, conditions are different. They could spread easily through Central America, but arrived in South Mexico the wedge of the plateau divides them into an Atlantic and a Pacific mass. They can go a long way north, and are still in Tierra Caliente, like the countries whence they came. But a sifting takes place. The Atlantic lowlands are hot and moist, whilst the Pacific slopes and much narrower lowlands are hot and rather dry, the dryness increasing rapidly towards the north. To people such divergent countries implies a severe sifting of the immigrants, or the necessity of changing, by adaptation to, or by, the new surroundings. This is well illustrated by the gradual change, from species to species, of essentially northern into slightly less northern, into almost tropical forms of the same genus; or, since a genus is most cases an imaginary abstract, of the same group of closely allied creatures. Still further south that particular genus comes in most cases to an end, ‘There may be a species or two which form outposts, straggling on, perhaps in actual process of successful adaptation ; however, after all the genus has found its mit. But it is there not met by the outposts of the southerners; they in their turn stand much further north. If it were otherwise, there would be a real boundary line, with a kind of neutral zone between North and South, and this neutral zone should contain compara- tively few species and genera. Emphatically this is not the case. Thé two faunas overlap broadly ; they commingle, except on the plateau, which seems to be a much more effective barrier to the southerners than is the descent from the plateau into the hot lowlands to the northern creatures. It seems to be easier for xerophile northern genera, and even species, to go south and to adapt themselves to life in a more equably hot and decidedly moister country with luxurious vegetation, than for hygrophile southerners to do the reverse. Be it noted, however, that this applies only to those terrestrial northerners which can adapt themselves to arboreal life; rattle- snakes cannot doit. Speaking broadly, xerophiles are essentially humivagous; hygrophiles either live on the ground which is rich in humus, grass, or herbaceous tangle and underwood, or they are arboreal. A favourite way of adaptation is arboreal life, whereby the xerophiles escape inundations, accumulation of humus, debris, 240 DR. H. GADOW ON MEXICAN [June 6, and the gloom of the underwood. In a desert or semidesert the amount and character of the scarce and precarious vegetation remain practically stabile; not so in the Pacificlowlands. During the rainy season grows up a dense mass of herbaceous plants covering the ground with a tangle of weeds, tall Salvias and Composites, stinging herbs and spiny creepers ; all this disappears, is burnt up, scattered during the dry season, and for months the ground may be bare, whilst many of the trees are leafless. In this Pacific type of Tierra Caliente we have periodical extremes, Difterent again is the moist Atlantic Tierra Caliente, and also the ranges of mountain forests of the Southern and South-eastern Tierra Templada. There are no extremes; the very opposite to arid tracts ; there is plenty of high and low vegetation all the year round. The important factor is not the temperature, nor the altitude as such, but the amount, or rather the distribution, of annual moisture. ‘Temperature: more than the northern half of the Mexican plateau belongs to one of the hottest regions of the world, the centre of heat being the State of Sonora. From May to July the mean temperature for Sonora is 36° C.=96°8° F.; for the rest of the northern plateau 30° C.=86° F., which is more than the summer average of South Mexico and Central America. But in the winter the North averages 16° C.=60°8° F., while the Tierra Caliente enjoys 25° C. In short, the Hot-land temperature averages from 25° to 28° C.=75° to 82° H.; the Northern plateau from 60° to 96° F., with additional extremes from frost and snow to unbearable broiling heat and drought. The overlapping, mentioned above, is much more generic than specific. There are, indeed, very few species which, although having a wide geographical range, are well established in stations of de- cidedly very different physical aspect. For instance, species on the higher mountains, or plateaux, and also in the Tierra Caliente : see p. 231. But of all these only very few, e.g. Hylodes rhodopis, Sceloporus scalaris, a Rattlesnake, and Tropidonotus ordinatus, can, in their indifference to physical conditions, be compared with the Puma, the Armadillo, Opossum, the Raven, and Turkey- Buzzard. Some species, natives of the plateau, descend from it down to the neighbouring coast (Bufo simus, Hypsiglena torquata, Zamenis grahami); others ascend from the hot countries on to the plateau, especially from the west by way of Guadalajara, and thence to Guanajuato and further east, the means being the alluvial plains spoken of before; or the ascent can be traced through the Balsas depression towards Iguala and Cuernavaca ; another opportunity seems to lead from the east side to Zacual- tipan in the State of Hidalgo. Such ascending species are Bufo marinus, B. valliceps, Hyla miotympanum, Engystoma wstwm, Phyllodactylus tuberculosus, Uta bicarinata, Zamenis mexicana. To another category belong those species which have a wide, but very scattered, discontinuous distribution, especially those 1905.} AMPHIBIANS AND REPTILES. 241 which, like most Gerrhonotus, ave now restricted to the higher mountains. Lastly, a considerable number of Southern species ascend from the hot lowlands high up onto mountains which rise isolated, or which fringe the plateau. Of course it is difficult, perhaps premature, to generalise in this respect, and sharp lines cannot be drawn between these categories. Not the least cause is the vagueness or doubtful nature of many of the reported localities. For instance, Cope had various corre- spondents in Mexico, and some of the alleged localities are quite impossible. Peters had a good correspondent resident in Puebla City, but the specimens which now figure as ‘“‘ Puebla” came from anywhere in that State, which has the most perplexing, intricate boundaries, and contains altitudes from 3500 to less than 100 metres! “ Vera Cruz” is another snare to the unwary. Others have bought specimens, even collections, in Mexico City. I myself found in a shop at Orizaba several large glass vessels full of well-preserved snakes for sale, but I left them alone since nobody knew where they came from. Sumichrast lived for many years in Tehuantepec and he travelled widely, all over the Isthmus and beyond. The town is situated on a plain, about 100 feet above the not distant sea; within a few hours’ ride are mountains, covered with pines, well above the Tierra Caliente, as typical of which every specimen labelled “Tehuantepec” is put down. I shall not, at least in this paper, go into the detail of the generic overlapping, a very important question. Suflice it to say, that Im many cases the species of a genus are so distributed that some are decidedly northern, living on the plateau, typical inhabitants of the Tierra Fria; another species lives in the ad- joining Tierra Templada, more often on the western than on the eastern slopes and descending more or less far into the lowlands ; while a third kind is confined to the typical tropical Tierra Caliente. Such cases are clearly illustrative of the evolution of species due to the prevailing physical conditions, especially when none of these species has a wide geographical range. Are we justified in calling a certain species ancient because it has a wide continuous range? For instance, Zropidonotus ordi- natus, Crotalus terrificus. It is rather doubtful, because these creatures are so indifferent to climatic conditions. With more right we consider those as ancient which have to be very par- ticular about their terrain, and which are now scattered, without the least chance of communication—as, for instance, Thorius, Chirotes, Heloderma, and other slow, or digging, creatures. 3. Northern and Southern Immigration. In the following table the Mexican Amphibia and Reptiles are divided into a Northern or Nearctic and a Southern or Neotropical mass, according to their presumable ancestral home or cerrttre of Proc. Zoot. Soc.—1905, Vor. Il. No. XVI. 16 242 DR. H. GADOW ON MEXICAN [June 6, Nearctic, extending into: NEOTROPICAL. Not beyond Mexico. Central America. Antilles. Pelobatidze Desmognathinee Amblystomatine Plethodontinze % ee Heloderma__ _ Xenosaurus__ _ Xantusiide : a. Sonoran 6. Southern 56 Anguidee x > Amphisbheenidee 8 fe Tguanidee: a. Xerophile b. Hygrophile Glauconiidee Boidee Crotalinee Colubrine Aglypha _ | Colubrinee | Opisthoglypha. __Hlapine. ———_——— SS SS es ee es HY ee Tejidee. = — ow“ — es | ee ee ae ene Ov suicomating ces ¥ Hylide. x“ Bufonide. Chelonia: Testudinidee BAG YB. Cinosternidee a mi ae Dermatemydidee__ fn ta0d Typhlopidee. a Geckonide. pees AOL Sea Beno ystomentce: dispersal so far as America is concerned. Those which have sent forms into the Greater Antilles are also indicated. The Greater Antilles have received their fauna* from Nearctic %* Gregory thinks it is “almost certain” that Yucatan was connected with Cuba. Other zoogeographers have likewise assumed this connection, and it looks very plausible on the map. If it ever existed, it must have been very transitory. Amphibia 1905. | AMPHIBIANS AND REPTILES. 243 and from Neotropical groups, of both Amphibia and Reptilia, but no northern group has contributed, unless it had spread well into Central or even into South America (witness the Plethodonta, Anguide, Amphisbenide, Scincide, Xantusiide, Aglyphous Colubri ine, Iguanidee). All these Nearctic, or Old-Sonoran, groups must have been there in Miocene times. The same age must be assigned to the outhern immigrants—the Cystignathide, Hylide, Bufonide, Tejidee, Typhlopide. On the other hand, the following must be considered as decidedly post-Miocene so far as their existence in the present Central America is concerned: from the North the Pelobatide, Desmognathinee, and Amblystomatins, none of which extend, southwards, beyond Mexico proper; from the South the Engy- stomatinee, Opisthoglypha, KElapinee, none or few of which go beyond Mexico into the United States. Lastly, the latest arriv: als in South America are the Crotalins, of which only Lachesis lanceolatus has entered the Lesser Antilles. Ancient Sonorans are //eloderma and Chirotes. The Testudinidee are also Old Sonorans. Still with fair numbers in Mexico, but ever decreasing southwards through Central into South America, Testudo has arrived in Central and South America too late for the Antilles, but in time for the Galapagos. This indicates that the Caribbean Sea and Gulf of Mexico connection was established before the disappearance of the western extent of Central American land. It is another hint that the Isthmus of Panama is but the last vestige of a former much broader land-connection between the two Continents. Concerning the Colubrine Snakes, they remind us in their dispersal southwards of the Iguanide, Anguide, and Boide. They have gone in detachments. The earliest migrants, when arrived in South America, have developed there, and since, into Opisthoglypha and the Aglypha part 3; and these are now surging back, northwards, post-Antillean. A second lot are the Aglypha part 2, many of which have entered the Antilles. Lastly, the last detachment of northerners passing through Mexico and Central America, too late for the Antilles, but still continuing their southward migration. If I am right in the conclusion that American Colubrinze gave vise to Opisthoglypha in South America, it follows that Opisthoglypha are not a natural group, those of the Old World, chiefly paleotropical, being an instance of collateral development, convergent, spreads, or whatever term Aer be ean. ‘ed. and mace do not support it; on the Pecan their present distribution is opposed to it. About 70 species are known from Yucatan. Its fauna is essentially that of the Atlantic Tierra Caliente; it differs from that of the Antilles apparently by the -absence of Xantusiide, Glauconiide, and Anguide. On the other hand, it is incon- ceivable why Tortoises, Pit-vipers, Opisthoglypha, and Cnemidophorus, all of which are plentiful in Yucatan, should not have crossed over into Cuba if a direct land- bridge had been ay ailable. 16* 244 ON MEXICAN AMPHIBIANS AND REPTILES. [June 6, Unless this conclusion be accepted, we have to resort to violent interpretations. Hither complete extinction all over North America, a measure which receives no support from actual distribution; or we must be prepared to assign to the Opisthoglypha a Cretaceous age, as a family not descended from North-American Colubrine ; or, lastly, if we should insist upon the Opisthoglypha as a natural group, the only explanation would be a land-connection across the Equatorial Atlantic, which with shifting modifications is supposed to have existed from Lower or Mid-Cretaceous into at least the Oligocene epoch. This bridging of the Atlantic is somewhat problematic. For our purposes we can discard the Cretaceous Brazil—Africa con- nection. Of more concern to periaretic distribution is the Europe— Greenland—North America continuity, which is supposed to have persisted well into the Tertiary period. But there was a third, more direct. bridge, although one of a curious and mysterious structure, which by its several advocates is dimly described as composed of a shallow sea interspersed with many islands; or as a solid land- belt; or, lastly, as a long archipelago with a continuous coast. This mysterious structure is supposed to account for the unmistakable similarity between the now extinct Antillean and Mediterranean coral-fauna, Old- World and Antillean land-mollusea, de. Obviously the corals require sea, the mollusca land. The apparent contradiction may be solved by the suggestion that there existed between Central America and the Mediterranean a sea (part of the Tethys of Suess and Ortmann, later their “ Great Mediterranean”), shallow during the Oligocene epoch, studded with islands, bordered by continuous land in the South (Brazilia to West Africa, or later between N. South America and West Africa, part of the Mesozonia of Ortmann) and in the North (Western Europe to Appalachia). Subsequently the Tethys increased to a big “bay” in Mid-Atlantic, this bay extending, spreading south and north, drowning first the southern land-belt, driving the northern land farther and farther north, with the ultimate result of a junction of the South with the North Atlantic; in other words, establishment of the whole Atlantic, Now these land-bridges, provided they existed long enough and at the right time and place, the Southern until at least the beginning of the Eocene, the Northern at least through the Oligocene epoch, would explain many a puzzle in geographical distribution ; for instance, that of the Aglossa, Boas, Podoenemis, Amphis- benide, Solenodon. The Northern bridge would throw light upon the Anguide and upon Spelerpes, a large American genus with a solitary species in Sardinia and Italy. But this is at present a land of dreams. With more claim to reality, we can conclude that Central America, although genetically part of the North-American continent, has received vis dominant, most characteristic fauna from South America, and this southern fauna has surged northwards chiefly to the east and west of the Mexican plateau. (HELLS WOOT SIPEOIND se IIUNOKGIAS) TSIENGUND YZ TL eo T rats) EMME TE AG SUASL WITT 39 ep Weezy f ra Meee eoers wa Bale & Danielsson ‘4 tmp lL SCE LOLORUS GADOVMIAD 2) LEP RODIRA ec Ullah J.Green delet hith. 1905. | ON NEW MEXICAN REPTILES. 245 3. Descriptions of new Reptiles discovered in Mexico by Dr. H. Gadow, F.R.S. By G. A. Boutencer, F.R.S., WaleeVaerse | Received May 17, 1908. | (Plates VI. & VII.*) ANOLIS GADOvII. (Plate VI. fig. 1.) Head once and two-thirds as long as broad, slightly longer than the tibia; forehead concave; frontal ridges distinct, divergent; upper head-scales rugose, not keeled ; scales on frontal ridges and supraorbital semicircles large, the latter in contact on the inter- orbital region; three large supracculars, forming together a disk separated from the supraorbital semicircle by two series of small scales; occipital large, a little larger than the ear-opening, separated from the supraorbital semicircles by two series of small scales; canthal scales four, loreal rows six; six or seven upper labials to below centre of eye; ear-opening large, vertically oval. Gular appendage very large, extending far back on the breast ; gular scales smooth. Body compressed; no dorso-nuchal fold. Dorsal scales small, smooth or faintly keeled, irregular, juxtaposed ; lateral scales minute, granular; ventral scales larger than dorsals, smooth, juxtaposed. The adpressed hind limb reaches the eye; tibia as long as the distance between the end of the snout and the ear; digits moderately dilated; 20 lamelle under phalanges TI and III of the fourth toe. Tail feebly compressed, not crested, once and three-fourths length of head and body. No enlarged postanal scales. Greyish above, with black wavy and vermicular lines; two parallel black lines on each side from shoulder to hip ; belly white; gular appendage bright red. Total length ...... 225 millim. Fore limb...... 37 millim. IEICE %GL Weare ee ZO ie% 5 ind linmiby ea) 63 iar, Whidthcot head). p512.' ., Dans, cos eseets 145_,, ISOON ee nde nea nee eee GO. This very distinct and handsomely marked Anolis is represented by a single male specimen, from Tierra Colorada, South Guerrero, ANOLIS LIOGASTER. (Plate VI. fig. 2.) Head once and a half as long as broad, longer than the tibia ; forehead deeply concave ; frontal ridges strong, short, divergent , upper head-scales smooth or feebly keeled; scales of the frontal ridges and supraorbital semicircles large, the latter in contact on the interorbital region or separated by one series of small scales ; three large, smooth or faintly keeled, transverse supraocular scales forming a single longitudinal series, in contact with the supraorbitals ov separated from them by one series of small scales; occipital larger than the ear-opening, separated from the supraorbitals by one or two series of scales; canthus rostralis sharp; canthal scales * For explanation of the Plates, see p 247. 246 MR. G. A. BOULENGER ON [June 6, three; loreal rows five; six upper labials te below the centre of the eye; ear-opening rather small, vertically oval. Gular appendage very large, extending far back on the breast, in the male, absent inthefemale; gular scales feebly keeled. Body feebly compressed ; no dorso-nuchal fold. Dorsal scales subrhomboidal, subimbricate, strongly keeled, passing gradually into the minute, granular scales of the sides; ventrals much larger than dorsals, rounded, imbricate, smooth. The adpressed hind limb reaches the eye or a litte beyond; digits moderately dilated; 16 lamelle under phalanges IJ and ITI of the fourth toe. Tail scarcely compressed, twice as long as head and body. Male with enlarged postanal scales. Reddish brown above, with a paler broad vertebral stripe, widening on the nape; this stripe edged with dark brown in the female; lower parts golden, the gular appendage bright red. e Total length ...... 150 millim. Fore limb....:. 23 millim. IE IGANG Raltan sar eaadieae TOMAS Jalgael ITM) gg By se Width of head... 10 ,, Hain Resse errs TOOT BOC N bosondeoaponn POETS Two specimens, male and female, from Omilteme, Guerrero, 7600 ft. The male is remarkable in the absence of the inner digit on the four limbs. Allied to A. nebulosus Wiegm. Distinguished principally by the smooth ventral scales. SCELOPORUS GADovI#. (Plate VII. fig. 1.) Head-shields smooth; frontal transversely divided, separated from the interparietal by a pair of frontoparietals; interparietal as long as broad ; parietals small, one pair on each side ; two canthal scales; five or six large transverse supraoculars, bordered inwards by one series of scales, outwards by one or two; five long pointed scales form a strong denticulation in front of the ear. Dorsal scales larger than ventrais, strongly keeled, pointed or shortly mucronate, forming oblique series converging towards the median line, passing gradually into the smaller scales of the sides; 73 to 77 scales between the interparietal shield and the base of the tail ; 19 or 20 scales, taken in the middle of the back, correspond to the length of the shielded part of the head. Ventral scales small, smooth, bicuspid. 75 to 80 scales round the middle of the body. The adpressed hind limb reaches the ear; tibia as long as the distance between the end of the snout and the ear; the distance between the base of the fifth toe and the extremity of the fourth exceeds the distance between the end of the snout and the posterior border of the ear. 28 to 33 femoral pores on each side, the two series narrowly separated on the preanalregion. Tail compressed; caudal scales a little larger than dorsals, strongly keeled, the two median upper series more strongly mucronate and forming a pair of serrated ridges. Male with slightly enlarged postanal scales. Greyish olive above, reddish on the sides, dotted with bluish 1905. | NEW MEXICAN REPTILES. 247 green ; limbs with rather indistinct dark bars; throat and belly dark blue; a narrow whitish median ventral streak. Total length ...... 147 millim. Fore limb...... 32 millim. iFleade hs. Sass DS” Bs Hindlimb ... 45 ,, Wadthvotheadit. sills ATMs cee see cect ae SO. BO. saris Snetsiteieee Deis Two male specimens from a ravine near Mesquititlan, north of Chilpancingo, Guerrero. This very remarkable species, which I take the liberty of naming after Mrs. Gadow, agrees with S. pyrrhocephalus Cope, in its distinctly compressed tail, but differs from it in having much smaller scales and more numerous femoral pores. No species of Sceloporus was hitherto known to have more than 25 femoral pores on each side. LEPTODIRA GUILLENI. (Plate VII. fig. 2.) Rostral twice and a half as broad as deep, scarcely visible from above; internasals a little longer than broad, little shorter than the preefrontals; frontal once and two-thirds as long as broad, a little longer than its distance from the end of the snout, a little shorter than the parietals; loreal as long as deep; one preocular, well separated from the frontal; two postoculars; a subocular below the preocular and another below the postoculars ; temporals 142; eight upper labials, fourth and fifth entering the eye; five lower labials in contact with the anterior chin-shields, which are much shorter than the posterior. Scales in 23 rows. Ventrals 189; anal divided; subcaudals 71. Above with eleven dark brown areas separated by narrow greyish-white bands; snout, interocular region, and temples brown, back of head and nape bright red with a dark brown median line; a light, dark-edged streak along the upper lip; lower parts white, the ventrals with a brown spot on each side; subeaudals brown, edged with whitish. Total length 530 millim.; tail 110. A single female specimen from the Rio Balsas, Guerrero. This species which, on the whole, is intermediate between L. nigrofasciata Gthr. and L. personata Cope, is named after Senor Don Manuel Guillen, Governer of the State of Guerrero, in recognition of valuable assistance rendered to Dr. Gadow. EXPLANATION OF THE PLATES. Prats VI. Fig. 1. Anolis gadovii, sp. n., p. 245. Vides, As ‘5 Upper view of head, x2. 2. » liogaster, sp. n., p. 245. Gs op 55 Upper view of head, x25. Puate VII. Fig.1. Scecloporus gadovie, sp. n., p. 246. Upper and lower views. la. 5 5s Upper view of head, x 25. 16. 5 a Side view of head, x 23. 2. Leptodira guilleni, sp. n., p. 247. Upper and side views of head and anterior part of body. 248 MR. G. A. BOULENGER ON SOUTH-AFRICAN [June 6, 4, On a Collection of Batrachians and Reptiles made in South Africa by Mr. C. H. B. Grant, and presented to the British Museum by Mr. C. D. Rudd. By G. A. BounEnGceER, F.R.S., V.P.Z.S. [Received May 29, 1905. ] The collections made within the last two years in South Africa by Mr. C. H. B. Grant and presented to the British Museum by Mr. C. D. Rudd, the Mammals of which have already afforded matter for two papers by Messrs. O. Thomas and H. Schwann, published in these ‘ Proceedings,’ included a good series of Batrachians and Reptiles, a list of which is here given. No new species were discovered, but the series is interesting for the sake of the localities, our knowledge of the exact distribution of these animals in South Africa being still very imperfect. A list of the localities is here given :— I, Cape. Durban Road, near Cape Town. This ‘‘ Durban” is a town about 15 miles N.E. of Cape Town. It. British Namaqualand. Port Nolloth, at mouth of Orange River. Klipfontein, a station on the railway between Port Nolloth and O’okiep, 54 miles from Port Nolloth. Altitude 3104 ft. Til. Zululand. Hluhluwe Stream, flows west into False Bay. or Umfolosi Station, on the railway, about 5 miles north of Umfolosi River. Eshowe, about 30 miles inland from Coast and Umbhalazi River. Altitude 1800 ft. Ngoye Hills, 15 miles E. of Eshowe, and 8 miles inland from Coast. Altitude 600-1000 ft. Sibudeni, about 60 miles inland from coast, at source of Umblatuzi River. Altitude 3500-5500 ft. Jususie River, close to Sibudeni. IV. Transvaal. Wakkerstroom, on the Natal border and at southern end of Drakenberg Range. Zuurbron, 20 miles East of Wakkerstroom. BATRACHIA. AGLOSSA. 1. Xenopus L&vis Daud. Durban Road, Umfolosi Station, Wakkerstroom, The largest specimen (2) measures 100 millim. from snout to vent. 1905. ] BATRACHIANS AND REPTILES. 249 In the present uncertainty as to the distinction of species in this genus, the distribution of _Y. levis is difficult to trace. This species appears to be found all over South Africa where there is water, and it extends as far north as Angola to the West and Abyssinia to the East, the British Museum possessing specimens, which I cannot separate from the typical form, from Lake Mweru, Uganda, and Senafe. Angola specimens (XY. petersii Bocage), which have been referred either to Y. levis or to Y. muelleri by Giinther, by Peters, and by myself, cannot be separated, by any character that I can detect, from XY. levis. I have examined eight specimens, one from Benguella, received from Prof. Barboza du Bocage himself, five from Pongo Andongo, obtained by Dr. Ansorge, and two from Dr. Welwitsch’s Angola collection. Bocage gives the length of the Angola specimens as not exceeding 65 millim. from snout to vent, but one of Welwitsch’s specimens measures 80. In the typical Y. levis from South Africa the subocular tentacle measures less than one-third the diameter of the eye, and is sometimes reduced to a mere tubercle, the inner metatarsal tubercle is very blunt and feebly prominent, never conical, and vomerine teeth are constantly absent. The true X. muelleri, as described and figured by Peters in his ‘ Reise nach Mossambique,’ vol. i11. (1882), has the tentacle more than half as long as the eye, the metatarsal tubercle more prominent, more conical than in YX. /e@vis, and vomerine teeth, first noticed by Tornier, are often present. In addition to Mozambique, whence it was first described, this species is found in Nyasaland and on Zanzibar and the opposite coast. To distinguish between Y. muellert and XY. levis is, however, not so easy as one might at first think, for the British Museum has received from Mr. C. 8. Betton three specimens from hot springs near Lake Nakuro, British East Africa, which agree with the former in the prominent, conical metatarsal tubercle, and with the latter in the short tentacle and the absence of vomerine teeth. AX. elivit described from Erythreea by Peracca, and obtained in numerous examples at Addis Ababa and Ashoofi, Abyssinia, by Mr. E. Degen, agrees with -Y. levis in the proportions, in the short tentacle, and in the absence of vomerine teeth, but is easily distinguished by the inner metatarsal tubercle being armed with a black claw, as in YX. calearatus, which inhabits Liberia, Lagos, Nigeria, Cameroon, the Gaboon, and the Congo. In the males of X. clivii the brown nuptial asperities, instead of being restricted to the inner side of the fore limbs, as in _Y. levis, extend as a large patch on each side of the breast. Two specimens from ‘“‘ West Africa,” collected by Mr. Fraser, therefore probably from Nigeria or Fernando Po, which have been referred by Dr. Giinther and by myself to Y. mwelleri in the British Museum Catalogue, agree with that species in the size of the eye, the length of the tentacle, and the presence of vomerine 250 MR. G. A. BOULENGER ON SOUTH-AFRICAN [ June 6, teeth (five in number)*, with X. eliwi and X. calearatus in the presence of a metatarsal “claw.” These specimens, the larger of which measures only 39 millim., no doubt indicate a distinct species, for which I propose the name XY. fraserv. PHANEROGLOSSA. 2. BuroO REGULARIS Reuss. Umfolosi Station, Hluhluwe Stream, Ngoye Hills, Wakker- stroom. : 3. Buro Granti Bler. Durban Road, Klipfontein. Since this species was described, in 1903, from numerous specimens obtained by Mr. Grant at Deelfontein, it has been rediscovered at Matjesfontein by Dr. W. F. Purcell, of the South African Museum. The male specimen which the British Museum has received from that institution measures 60 millim. from snout to vent and strikingly resembles a Bufo viridis. The interorbital space is as broad as the upper eyelid, the tympanum measures three-fifths the diameter of the eye, the first finger extends a little beyond the second, the tibio-tarsal articulation reaches the tympanum, and the subarticular tubercles under the toes are all single. The single male specimen found by Mr. Grant in a garden on Durban Road, near Cape Town, agrees very closely with the Matjesfontein Toad, but some of the subarticular tubercles under the toes are double. Another male, from Klipfontein, also has double subarticular tubercles. 4, Buro anausticers A. Smith. Durban Road. Several specimens, the largest measuring 46 millim. from snout to vent. The first finger never extends beyond the second, the fold along the inner side of the tarsus is more or less distinct, and the subarticular tubercles of the toes are usually single, although there are occasionally two between the last phalanges of the fourth toe. Bufo dombensis, from Dombe, Benguella, described by Barboza du Bocage in 1895 as a close ally of B. angusticeps, is more nearly related to Smith’s Bufo vertebralis, which, following Gunther, I have erroneously regarded as the young of B. carens. The examination of a small Toad found at Vredefort Road, Orange River Colony, by Major Barrett-Hamilton, and of which four specimens have been presented by him to the British Museum, has convinced me of my error. The breeding male, with large gular vocal sac, measures only 27 millim. from snout to vent, the female 35. In these specimens, the tympanum is close to the eye, * The vomer is single in X. levis, muelleri, and clivii, absent m X. calearatus, Hymenochirus, and Pipa. 1905. ] BATRACHIANS AND REPTILES. 251 and may measure three-fourths its diameter; the parotoids are flat and very indistinct, broken up into several glands; the subarticular tubercles are double, and there is no trace of a tarsal fold. The limbs are shorter than in B. carens and the white rhomboidal spot on the vertebral line, which does not exist in B. carens, appears to be constant; black spots are always present on the belly. 5. Rana DELALANDIT D. & B. Durban Road. 6. Rana Fuscicuuta D. & B. Klipfontein. 7. RANA ANGOLENSIS Bocage. Eshowe, Sibudeni, Wakkerstroom. The vocal sacs of the males form longitudinal folds on the sides of the throat. 8. RANA MASCARENIENSIS D, & B. Sibudeni. This species had not previously been recorded from South Africa. In the five specimens from Sibudeni the tibio-tarsal articulation reaches beyond the tip of the snout; a light vertebral stripe and a light line along the tibia are present. 9. Rana Grayi A. Smith. Durban Road, Klipfontein, Sibudeni, Ngoye Hills. 10. Rana rascrava Tsch. Sibudeni. The longitudinal folds and the dark stripes on the back are absent in the single specimen. 11. PHRYNOBATRACHUS NATALENSIS A. Smith. Sibudent. 12. ARTHROLEPTIS WAHLBERGII A. Smith. Sibudeni and Hluhluwe Stream. ‘The British Museum has also received a specimen from Pietermaritzburg, through Mr. Quekett. REPTILIA. CHELONIA., 1. STERNOTHARUS sInUATUS A. Smith. Umfulosi Station. A single half-grown specimen, the shell measuring 110 millim. As pointed out by me in 1896 *, this species is very variable and * Aun, Mus. Genova, (2) xvii. p. 15. 252 MR. G. A. BOULENGER ON SOUTH-AFRICAN [June 6, to distinguish it from S. nigricans is not without difficulties. In this specimen the cusps in the upper jaw are absent, the posterior border of the carapace is very distinctly serrated, the intergular shield is twice as long as broad, the length of the outer border of the pectoral shield slightly exceeds that of the humeral, and the suture between the abdominal shields is shorter than the front lobe of the plastron. Head pale brown above, with black vermiculations, white beneath, with blackish spots; plastron yellowish brown, bordered with black. 2, CINIXYS BELLIANA Gray. Umfolosi Station. This species had not previously been recorded from South Africa. In the specimens collected by Mr. Grant the shields of the carapace are marked with black radiating streaks. 3. Homopus sianatus Walb. Klipfontein. LACERTILIA. 4. LYGODACTYLUS CAPENSIS A Smith. Ngoye Hills. 5. PACHYDACTYLUS BIBRONIZ A. Smith. Klipfontein. 6. PACHYDACTYLUS MARIQUENSIS A. Smith. Klipfontein. 7, AGAMA BRACHYURA Bler. Klipfontein and Port Nolloth. This species was established on a single female specimen labelled ‘‘Cape of Good Hope,” from Si A. Smith’s collection. I have since examined four specimens collected at Deelfontein by Mr. Seimund,and presented to the British Museum by Col. Sloggett, and these, together with the six collected by Mr. Grant in British Namaqualand, enable me to give a revised description of this near ally of Agama hispida. Head convex, subcordiform, as long as broad. Nostril not tubular, lateral, pierced just below the canthus rostralis in a convex nasal. Scales on anterior part of head smooth or rugose, sometimes feebly keeled, often trihedral on middle of snout, on nack of head more or less strongly keeled, some erect and spinose ; occipital enlarged; head about the ears and neck with short erect spines. Body strongly depressed, covered with irregular, imbricate, strongly keeled scales intermixed with strongly enlarged, trihedral, spinose ones; a small nuchal crest, sometimes continued along the body ; ventral scales smooth or very feebly 1905. | BATRACHIANS AND REPTILES. 253 keeled. Limbs moderate, with scales very unequal in size; hind limb reaching between the shoulder and the ear; tibia as long as the skull to occiput; fingers short, third longest ; third and fourth toes equal, or fourth very slightly the longer, fifth not extending as far as first. Tail shorter or a little longer than head and body, cylindrical or slightly compressed, covered with strongly keeled scales. Male without gular pouch, with a single row of anal pores. Olive-brown or reddish brown above, with dark brown or blackish spots, the principal of which form a double series along the back, each pair separated on the vertebral line by a square or QC-shaped or f)}-shaped yellowish marking ; lower parts whitish or greyish, with a wide-marked grey or blackish network, which may disappear in adult males; the latter always have a bluish throat. sie millim, Motalllenethie way ssteee eset vote 235 160 Tes [era le eee Oe hee iat tee ook Ce 29 22. Wandin @it IN@ACl sccoccocecosanaoacccs 28 2) BOO iy tivciccweccatepiont «Pr suepeatings 81 63 1 ove’syil biivall orate in Gaerne uaa aaa aaa aes 53 40 iEEeaval linonoy) Oe Wahessennbesaeeecsadoncs 02 54 1] De ii bs eat estar RS UN CL A See BO a 125 75 A. brachyura differs from A. hispida principally in the fourth toe not being shorter than the third and in the absence of strong keels on the ventral scales. 8. AGAMA ARMATA Peters. Hluhluwe Stream. 9, AGAMA ATRA Daud. Klipfontein. Both A. micropholis Matschie (Zool. Jahrb., Syst. v. 1890, p- 607), and A. microterolepis Blgr. (Ann, & Mag. N. H. [6] xvii. 1896, p. 22), from the Transvaal, must be added to the synonymy of this species. 10. Zonurus potyzonus A. Smith. Port Nolloth, Klipfontein. 11. PsEuUDOCORDYLUS MICROLEPIDOTUS Cuv. Wakkerstroom. 12. CHAM@SAURA ANGUINA L. Umfolosi Station. 13. VARANUS ALBIGULARIS Daud. Umfolosi Station. 254 MR. @. A. BOULENGER ON SOUTH-AFRICAN [June 6, 14, Varanus nitoricus L. Ngoye Hills, Sibudeni, Jususie Valley. 15. NucRAS TESSELLATA A. Smith. Klipfontein. 16. Nucras DELALANDID M.-Edw. Sibudent. 17. IcHNoTROPIS CAPENSIS A, Smith. Umfolosi Station. The parietal shields sometimes form a short suture separating the interparietal from the occipital. The scales on the preanal region are much smaller in females than in males. 18. Scaprira KNoxit M.-Edw. Port Nolloth. 19. Scaprrra crenopactyLA A. Smith. Port Nolloth. ‘The femoral pores may number as many as 36 on each side. 90. MABUIA TRIVITTATA Cuv. Wakkerstroom. 21. Masur varia Peters. Klipfontein, Umfolosi Station. 99. MABUIA STRIATA Peters. Hluhluwe Stream, Umfolosi Station, Sibudeni, Ngoye Hills, Zuurbron, Wakkerstroom. 23. MABUIA SULCATA Peters. Klipfontein. 94, ScELoTES BEPES L. Durban Road. 95, ACONTIAS LINEATUS Peters. Port Nolloth, Klipfontein. RHIPTOGLOSSA. 96, CHAMELEON QUILENSIS Bocage. Jususie Valley. 27. CHAMALEON VENTRALIS Gray. Port Nolloth. 1995. j BATRACHIANS AND REPTILES. 255 OPHIDIA. 28. PyrHon sEBz Gin. Umfolosi Station. 29, ABLABOPHIS RUFULUS Licht. Sibudeni. 30. PSEUDASPIS CANA L. Wakkerstroom. 31, DASYPELTIS SCABRA L. Negoye Hills. Uniform brown (var. palmarum Leach). 23 scales across the body, Ventrals 218; caudals 75. 32. AMPLORHINUS MULTIMACULATUS A, Smith. Wakkerstioom. Uniform green, without spots, as in the specimens presented by Dr. Quain and mentioned in the British Museum Catalogue (ii. p. 125). Ventrals 138; caudals 76. 33. 'TRIMERORHINUS RHOMBEATUS L. Durban Road, Wakkerstroom, Klipfontein. 34. PSAMMOPHIS SIBILANS L. Umfolosi Station. The single specimen falls under Division F of the British Museum Catalogue (ili. p. 163). Ventrals 165; caudals 97. 35. DIsPHOLIDUS TYPUS A. Smith. Sibudeni. Green, the scales edged with black (Division D of British Museum Catalogue, 11. p. 189). Scales in 19 rows. Ventvrals 174; caudals 119, 36. ASPIDELAPS LUBRICUS Laur. Klipfontein. 37. DENDRASPIS ANGUSrICcEPS A. Smith. Negoye Hills. 38. Brivis ARIETANS Merr. Umfolosi Station, Hluhluwe Stream. 39. Brtts cornnutTaA Daud. Port Nolloth, Klipfontein. 40. Brris GAUDALIS A. Smith. Port Nolloth, 256 MR. F. E. BEDDARD ON tHE ANATOMY [June 6, 5. Some Notes upon the Anatomy of the Yellow-throated Lizard, Gerrhosaurus flavigularis. By F. EH. Bepparp, F.R.S., Prosector to the Society. [Received May 17, 1905.] (Text-figures 33-38.) Apart from osteology * and a few scattered notes, which will be veferred to in the course of the present communication, there does not appear to be a great deal known about the internal structure of Gerrhosaurus. Inasmuch as this Lizard is regarded, from the point of view of external characters and osteology, as being exactly intermediate between the Lacertidz and Scincide ‘7, it seemed to me interesting to attempt a criticism or confirmation of this view, while recording any new facts which an investigation of Gerrhosaurus flavigularis might bring to light. Jugal Ligament. Many, but not all, of the Lacertilia possess, as is well known, a jugal ligament, which Huxley compared to the bony lower temporal arcade of Hatteria. The exact relationships of this ligament have not, I believe, been described in some of the Lizards in which I shall now proceed to detail the arrangement. Tt is possible to recognise several stages in the conditions of the jugal ligament, which may represent evolutionary stages, though it is, of course, not implied that the genera to be mentioned are genetically connected in the order named. In Jguana tuberculata the ligament as a distinct structure 1s totally absent. On cutting through the skin covering the “ cheek,” the muscles and bones of this region of the skull are at once arrived at. It appeared to me, however, that the subcutaneous connective tissue, which is dense and white in most parts of the body, was rather denser and whiter in the region where the jugal ligament would be were it present. It is possible, in fact, that in this lizard an early stage is met with—that the ligament is not yet differentiated from the general connective tissue of the skin. On the other hand, it cannot be denied that the same fact may be explained on the theory that the ligament has disappeared. In any case, Gerrhosaurus offers an intermediate condition. In this reptile the ligament in question is anchored firmly to the quadrate behind, but in front it is not attached to the jugal bone but to one of the bony scales which cover the face in this region. That is to say, the ligament has not as yet completely detached itself from the skin. So, at any rate, the facts seem to indicate. It is important to notice in connection with the main object of the present communication, viz., to attempt to fix the systematic * Siebenrock, Ann. k. nat. Hofmus. Wien, vii. 1892. + Boulenger, Cat. of Lizards. 1905. ] OF THE YELLOW-THROATED LIZARD. 257 position of Gerrhosaurus, that this lizard agrees absolutely and in every detail, so far as the jugal ligament is concerned, with the skink Humeces. The final stage in the arrangement of the jugal ligament is shown in Physignathus. In this lizard the ligament is attached. firmly to the bones at either extremity, and has entirely lost its presumably original connection with the skin. Moreover, in the last-named lizard, the ligament is divisible into two regions. There is, first of all, a stronger narrow ligament which occupies exactly the position of the bony quadrato-jugal bar in Hatteria, and above this and in part overlapped by it is a thinner but still stout sheet of ligament which entirely fills up the temporal vacuity. This state of affairs does not exist at all in Gerrhosaurus and Humeces. It is distinctly suggestive of the complete obliteration of the lower temporal vacuity in certain Vertebrates. In view of the fact that bones in some cases can be shown to degenerate into ligaments, it is not certain that the stages sketched out above may not be read in the inverse order. For example, the lower part of the fibula is ligamentous in Birds; but it is not to be assumed that here there is anything but a degeneration of the bone into ligament. The facts which have been detailed above coneerning certain Lizards do not, however, appear to me to point to a reduction from a state of affairs such as is found in Hatteria. If we were only acquainted with the condition observable in Physignathus and Iguana, such a view might indeed be held. The bone, it would be urged, has degenerated into ligament in the one case, and has finally disappeared in the other. But the conditions to be seen in Gerrhosaurus and in Hwmeces would seem to negative such a supposition. Peritoneal Folds and Celom. Although the suspension of the alimentary tract and the other viscera contained in the ccelom is broadly like that of many other Lacertilia, there are some differences of detail which require attention. In the female example the line of attachment of the oviducal membrane, which diverges laterally on each side, marks off sharply the posterior pigmented area of the coelomic membrane from the anterior non-pigmented or less pigmentedarea. This is quite a common and well-known arrangement among the Lacertilia. The reason why I bring the matter forward here, is that Gerrho- saurus differs from Humeces, where there is no such differentiation of pigmented and non-pigmented areas*, and because the pig- * This is not, however, a distinctive mark of difference from the Scincide and of likeness to Lacerta. Lacerta shows this sharp demarcation; but there are varying degrees among the Scincide. In Hwmeces there is hardly any pigmentation ; in Tiliqua scincoides there is a moderate amount, but evenly spread through the body- cavity (in a male). In Macroscineus cocteaui (female), however, the oviducal membrane marks off two areas; but the posterior area is not so deeply pigmented as in Gerrhosaurus. Proc. Zoou. Soc.—1905, Vou. II. No. XVII. 17 258 MR. F. E. BEDDARD ON THE ANATOMY [June 6, mented area in the male Gerrhosaurus is distinctly greater than in the female example of that lizard, and there is no conspicuous fold of membrane continued forward from the gonad duct to serve as a demarcation between the two areas in the latter. The suspensory ligaments of the liver offer, as is well known, characteristic differences of arrangement in various Lacertilia. In both examples of Gerrhosawrus the falciform ligament of the liver is double posteriorly for about the last } of the total length of theliver. This double region of the umbilical or falciform ligament is a tent-like structure; that is, the two separate membranes converge ventrally to be inserted in common on to the ventral median line of the parietes. A partial duplication of the umbilical ligament of this kind is not uncommon in the Lacertilia. It occurs, for example, in Lacerta ocellata. The double condition of the umbilical ligament in the Scincide, originally discovered by John Hunter* and subsequently more fully dealt with by myself 7 and Prof. Cope t, seems to be merely an exaggeration of this, the union of the two, posteriorly separate, umbilical ligaments being deferred until at or near the anterior extremity of the liver. Furthermore, all of the members of the family Scincide are not thus characterised ; for in Macroscinus cocteawi the arrangement of the umbilical ligament is much like that of Gerrhosaurus. In the question of affinity, therefore, the disposition of these mesen- teries is not decisive. There are, however, one or two other points to be noted, In the first place, in Humeces algeriensis both the umbilical ligaments are thickly invaded by muscular tissue, especially the left-hand ligament. This is also noticeable in Macroscincus, though to a much less extent; and it will be remembered that Macroscincus cocteawi is a much larger lizard than is Humeces algeriensis, so that size in this case has nothing to do with the development of thickness and muscularity in the umbilical ligaments. It is plainly therefore of importance to note that in Gerrhosaurus these ligaments are not obviously muscular at all. In the accompanying figures (text-figs. 33, 34) of the ventral surface of the liver in Gerrhosaurus two other facts may be pointed out, In the first place, there are traces of a membrane which runs obliquely forward and ends in a notch in the left border of the liver. As this white seam (6 in text-figs. 33, 34) is much better developed in one example than in the other, I take it to represent a rudimentary structure, and it may represent the original course of the umbilical vein and thus correspond to a similar trace which Hochstetter has lately described § inj the Blind Worm (Anguis fragilis). %* Hssays and Observations, revised by Richard Owen, London, 1861, vol. 11. p. 369. “The liver [of Tiliqua] is attached forwards by two membranes, one to each lobe, which unite at top.” + P. Z.S. 1888, p. 98. { Proc. Acad. Sci. Philadelphia, 1896, p. 308. § Morph. Jahrb. xix. Taf. xvi. fig. 185 but the course of the seam is different m the two cases. 1905. ] OF THE YELLOW-THROATED LIZARD. 259 The second point concerns the relationship of the two umbilical ligaments to veins entering the liver. A dissection of both specimens of Gerrhosaurus shows that the anterior abdominal vein enters the liver in the region of the left umbilical ligament (c in text-figs. 33, 34), and that the epigastric vein is similarly connected with the right umbilical ligament. Precisely the same relationship holds for MJacroscincus cocteaui. Inasmuch as the anterior abdominal vein joins the portal vein, the latter might Text-fig. 33. Text-fig. 34. Ne WOE Ata, . Text-fig. 33.—Liver of Gerrhosaurus flavigularis, ventral aspect. a. Attachment of umbilical ligament; 6. Seam indicating course of embryonic umbilical vein (?); ce. Left half of umbilical ligament; Ant. Abd. Anterior abdominal vein; Hp. Epigastric vein; g.b. Gall-bladder. Text-fic. 34.—Liver of a second example of Gerrhosaurus flavigularis, ventral aspect. Lettering as in text-fig. 33. be regarded as fixing this point were it not for the conditions observable in Macroscincus cocteaui. In that lizard the portal vein, immediately in front of the region where it has, as have the portal yeins of other lizards, a spiral twist, divides into two branches, which enter the liver ina line with each part of the divided umbilical ligament. As to the relationship between the divided as 260 MR. F. E. BEDDARD ON THE ANATOMY [June 6, umbilical ligaments and blood-vessels, it is noteworthy that in Eumeces algeriensis two hepatic arteries are associated each with one of the two umbilical ligaments of that lizard. Gastrosplenic Omentum.—This mesentery is very conspicuous in Gerrhosaurus. It stands out asa free fold with the following relations :—It arises from the stomach close to the pylorus and passes obliquely downwards supporting the posterior extremity of the spleen, the rest of which lies upon the mesogastrium ; it is finally attached to the median dorsal line of the body-wall on a level with the left ovary. This arrangement is practically repeated in Macroscincus, where, however, owing to the position of the viscera, the omentum is shorter, but very strong and fibrous. Moreover the spleen does not even reach, let alone hang over, the edge, as is the case with Gerrhosaurus. In Humeces, however, the gastrosplenic omentum is identical in its relations with that of Gerrhosaurus, save that it is a little less pronounced as a free fold. I am not describing here a state of affairs which is merely Lacertilian; for in T’upi- nambis the course and relations of the apparently homologous fold are different and do not involve the spleen. Hepato-pulmonary Ligaments.—Gerrhosaurus agrees with the majority of Lizards in that the right lung is suspended by two mesenteries, viz., the hepato-pulmonary and dorsal pulmonary. Tt is noteworthy that the latter mesentery in the case of both lungs extends to the very tip of the organ; whereas in Hwmeces the mesenteries in question do not reach the extremity of the lungs. This is not, however, a characteristic of the Skinks as opposed to Gerrhosaurus, for in Tiligua the membrane is co- extensive with each lung as in Gerrhosaurus. Mr. Butler * observes that “certain Scincoid lizards are as to the relations of their right lungs and liver intermediate between the Teiide and other Lizards.” My own knowledge of the family Scincide enables me to confirm Mr. Butler; but his accurate statement requires expansion’. In Hwmeces, Macroscincus, and Tiliqua there is, in fact, attached to the right lung a pulmo-hepatic liga- ment which is not so extensive as in, e. g., Gerrhosaurus. In Macroscincus cocteawi this membrane extends rather more than halfway down the lung and ends off upon the dorsal pul- monary ligament, necessarily running in this region in a direction nearly at right angles to the longitudinal axis of the lung. Whereas in Gerrhosaurus flavigularis the two pulmonary mem- branes join behind the right lung, in both specimens which I examined. The fact that there is no ligamentous interval between the * “On the Subdivision of the Body-cavity in Snakes,” P. Z.S. 1892, p. 481. + And has also been expanded by Hochstetter quoted below. { Hochstetter (Morph. Jahrb. xxvii. p. 292) figures the same membranes in some other Skinks, where they appear to agree with those of the forms studied by myself. He says, however, of Gerrhosaurus madagascariensis that the “caudal end of the right lung commences to be isolated from the ligamentum hepato-cavo-pulmonale.” It is not soin G. flavigularis. 1905. ] OF THE YELLOW-THROATED LIZARD. 261 prolonged right lobe of the liver and the gonad, both male and female, does not bear upon the question of the affinities of Gerrho- saurus. For among the Skinks these organs may be in contact or separated by a ligamentous interval. Muscular fibres in Mesenteries—As is the case with other Saurians, Gerrhosaurus has bands of unstriped muscle in several of the mesenteries. The most important of these is a bundle of muscular fibres which accompanies the anterior abdominal vein (text-fig. 835, m) and runs into the gastro-hepatic ligament. It is a thick bundle of fibres, but after traversing the gastro-hepatic ligament for about half its extent it fans out into a fine bundle, - the individual fibres of which hardly reach the stomach. This bundle is represented in many lizards. But the conditions observable in Gerrhosaurus throw no light upon the affinities of Text fig. 35. Gastro-hepatic ligament of Gerrhosauwrus flavigularis, showing course of muscular bundle. A, Gastro-hepatic ligament ; Ant.Abd. Anterior abdominal vein; Gt. Stomach ; LZ. Left lobe of liver; m. Muscular band. that lizard. For though it differs from the arrangement found in the Scincide, it shows no likeness to what is found in Lacerta ocellata. In KHumeces, Macroscineus, and Lacerta ocellata the bundle of fibres is continued without fanning out to the stomach, where it forms a close investment of that organ for the greater part of its extent in MZacroscincus. Inasmuch as both specimens of Gerrhosaurus were identical in the characters of this muscle, it may, I think, be assumed that its condition is typical of the species. Pancreas.—The pancreas of Gerrhosaurus (text-fig. 36, p. 262) differs from that of Lacerta ocellata in the comparative stoutness of the branch which goes tothe spleen. It is, in fact, like Leydig’s figure of the pancreas of Lacerta agilis, expanding when it reaches the spleen. The pancreas of Gerrhosawrus furthermore differs: from that of Lacerta (at any rate ocellata) in that there is a 262 MR. F, E, BEDDARD ON THE ANATOMY [June 6, patch of the gland on the dorsal side of the pyloric angle from which the splenic limb arises, and which is continuous beneath the end of the stomach with the main body of the pancreas. In Lacerta ocellata the splenic limb arises from the main lobe of the pancreas further towards the gall-bladder. In these points the Text-fig. 36. Pancreas of Lacerta ocellata (left-hand figure) and of Gerrhosaurus flavigularis (right-hand figure). P. Pancreas ; py. Commencement of intestine; spl. Spleen; S¢. Stomach. pancreas of Gerrhosaurus agrees with that of the Skinks, in which, however, there is a tendency towards an enlargement of the dorsal lobe of the pancreas and a disappearance of the splenic lobe. I could not detect the latter in MJacroscincus, and it was very thin in Tiliqua. Arterial System. As one of the two specimens of Gerrhosawrus flavigularis which T have dissected was successfully injected, I am able to give some account of the arterial system, dealing particularly with those points which vary among the families of Lacertilia. The heart has the usual, but not universal, tag tying the apex of the ventricle to the pericardium. The pericardium extends forward beyond the trifurcation of the arteria innominata. A pair of arteries exist of very fair size, running one on each side of the trachea in the position occupied by the carotids in many Vertebrates, and they are like them quite close to the trachea. These arteries have, however, nothing whatever to do with the carotids, They are branches of the pulmonary arteries (P, text-fig. 37), and the existence of these arteries in what appears to be an unexpected place is possibly indicative of a former forward extension of lung-tissue. The branches of the carotid arch differ slightly from those of 1905. | OF THE YELLOW-THROATED LIZARD. 263 some other Lizards, though they agree, as might be expected, in their main features. The first branch given off is a hyoid (hy, text-fig. 37), which supplies the hyoid region generally; I have not followed its branches minutely. In this region the carotid is in close contact with the systemic arch. Further dorsally they part company, and, shortly after this separation has occurred, the Text-fig, 37. Aortic arches and first part of dorsal aorta of Gerrhosawrus flavigularis. Ca. Carotid; g. Gastric; I.c. Vertebral artery; hy. Hyoid artery ; 1. Muscular twig; os. @sophageal branches; P. Pulmonary arch; Se/. Subclavian. main trunk of the carotid arises. The trunks are here so twisted that the carotid stem is given off posteriorly and dives under the carotid arch to reappear on its anterior face. The rest of the carotid arch is to be regarded as ductus Botalli, From this section arise two arteries: the first is a small muscular twig; the 264 MR. F. E. BEDDARD ON THE ANATOMY [June 6, second is an important trunk which divides into two branches. One of these supplies the muscles of the shoulder-region (JZ); the Text-fig. 38. Ln Abdominal region of aorta of Gerrhosaurus flavigularis. g', g°. Gastric arteries; I.c. Intercostals; L.In¢. Artery of large intestine ; cs. Hisophageal artery ; Si. Artery of small intestine. other has a recurrent course and dives through the ring formed by the carotid and systemic arches to supply the cesophagus (@s). 1905. | OF THE YELLOW-THROATED LIZARD. 265 The left systemic arch gives off no branches at all that I could discover. From the right systemic arch (which joins the left at about the commencement of the lung) the two swbclavians (text- fig. 37, Sel., p. 263) are given off, nearly, if not exactly, opposite to each other. In front of this arises the vertebral artery (J.c.), which gives off an cesophageal branch before plunging into the thickness of the parietes. Behind the vertebral artery commences the series of intercostals. The first zntercostal artery arises just before the junction of the two aorte. It gives off a branch to the cesophagus. The next two intercostals have also cesophageal branches ; but it is to be noted that in all these the right inter- costal alone has this cesophageal branch. The left has none. The remaining intercostals have no cesophageal or gastric branches. Their arrangement is peculiar and agrees with that of the Skinks ; it differs from that of some other Lizards. In Tropidurus hispidus, for example, the regularly paired intercostals emerge from the dorsal aorta close to the articulation of successive vertebre, and plunge at once into the thickness of the parietes. In Gerrhosaurus the intercostal arteries emerge from the aorta at about the middle of each vertebra. In many cases, and the arrangement is roughly alternating, the intercostal of one or both sides divides at once into two branches; one of these plunges at once into the thickness of the parietes. The other passes obliquely forwards and runs superficially in close relation toarib. This, however, only occurs in the thoracic region, not in the lumbar. Precisely the same disposition of vessels is found in Huwmeces and some other Skinks, and the fact isa bond of union between the Gerrhosauride and Scincide. The next artery to arise from the aorta is a gastro-cesophageal (text-fig. 37, ws, g, p. 263, and text-fig. 38, ws, g', p. 264), which divides at once into a thin forwardly directed cesophageal, and a stout backwardly directed gastric. Between this and the large gastric artery (text-fig. 38, g°) are 6 pairs of intercostals. Two pairs intervene between this artery and that of the large intestine, and one pair between the latter and the artery of the small intestine. The ovarian and oviducal arteries present some features which are worthy of note. There are three pairs of oviducal arteries which are not symmetrical. The first of these is really mainly an ovarian artery, which gives off a thin and slender oviducal branch running along the anterior section of the oviduct. The two remaining oviducal arteries arise in common with an intercostal. They lie in front of the rectal artery. Venous System. Although the venous system of neither of the examples at my disposal was injected, most of the veins were beautifully displayed by their own turgescence. 266 ON THE ANATOMY OF THE YELLOW-THROATED LIZARD. [June 6, There are many differences in detail between the venous system of this lizard and that of other genera. Vena cava posterior.—It is interesting to note that Gerrhosaurus agrees with Zuliqgua in that the left vena cava posterior is very much thinner than the stout right vein. This is another of those numerous though individually perhaps small points of likeness between the genus whose anatomy is dealt with in the present communication and the Scincide. As in Z%liqua also *, the left cava or vena renalis revehens lies to the left side of the mesorectum and the right vein to the right side of that mesentery. The left vena revehens is large where it receives the three or four veins arranged in a fan-like fashion from the left ovary ; behind this point it dwindles immediately but can be easily traced to the kidney, where it becomes enlarged at its jnnction with the right vena renalis revehens. The left vena renalis revehens receives two intercostal veins before the ovarian veins join it, and on the right side also I observed two intercostals. I could only observe one, and that a slender, oviducal vein joining the left vena renalis revehens. J feel convinced, however, that no veins from the oviducts join the afferent renal veins, as is often the case in Lizards. The reason for this in the present species may be that the kidneys are unusually far back. Afferent Renal Veins.—'The caudal vein reaches the kidneys as an undivided vein. It runs between them and receives a cloacal vein before dividing. Immediately after division each half receives another cloacal vein. The cloacal artery runs exactly at the point of division between the two afferent renals. At about the end of the first third of the kidney each renal afferent vein turns at right angles and runs superficially over the kidney, giving off a large branch to the kidney itself at about the middle of the transverse diameter of that organ. There is no sign of any forward continuation of the renal afferent vein beyond the anterior border of the kidney such as occurs in Chameleon and Pygopus *. Where the renal afferent vein reaches the border of the hind leg it receives three veins, two from the hind limb and one from the median dorsal parietes. It there runs directly forwards parallel with the kidney, and ona level with the anterior end of that gland receives the femoral vein, and a small parietal on the opposite side which crosses the epigastric artery. The vein then continues its straight course forward,and before bending inwards and downwards to follow closely the inner margin of the fat-body gives off a short forwardly directed branch, which appears to me to be the equivalent of the lateral abdominal vein of other Lizards. Its shortness in Gerrhosaurus contrasts with its length in Tiliqua. Hepatic Portal Veins——The mode of entrance of the conjoined intestinal portal and anterior abdominal and of the epigastric vein * See Beddard, P. Z.S. 1904, vol. i. p. 445, fig. 93. + See P. Z.S. 1904, vol. ii. p. 15, fig. 4. 1905. ] ON THE BRAIN OF LIZARDS. 267 has already been noted. The epigastric springs from the anterior abdominal some way behind the liver (text-fig. 34, Hp., p. 259), and running along the umbilical ligament disappears in the substance of the liver some way behind the anterior end. It is reinforced by the usual branches from the median ventral parietes. These vary in number in what appears to me to be a remarkable way. In one specimen repeated examination has only enabled me to ascertain the presence of a single ventral parieto-hepatic vessel, which joins the epigastric at about the middle of the liver. In a second specimen, on the other hand, there were four of these ventral parieto-hepatic vessels (cf. text-figs. 33 and 34, p. 259). I am disposed to think that the fluctuation in number of these blood-vessels is related to fluctuation in the number and size of the dorsal parieto-hepatic veins. These differed in the two specimens which IJ have dissected, though not quite to so great an extent as the ventral parieto-hepatic veins. In the specimen with but one ventral parieto-hepatic vein, the dorsal parieto-hepatic veins were as follows :—a large vein accompanies the anterior edge of a fold of membrane which in this, as in many lizards, runs obliquely and binds the end of the right lobe of the liver to the parietes. This vein runs superficially for a short distance anteriorly alongside the aorta on the right side, and is clearly a fragment of the right posterior cardinal. It reaches the parietes on a level with and outside of one intercostal artery and disappears from view to the inside of the next intercostal artery in front; it resembles a large superficially running intercostal vein. Besides this there are three other dorsal parieto-hepatic veins lying behind it. In the second specimen, with numerous ventral pavieto-hepatic veins, I could find only three dorsal ones ; and the first of these was by no means so large as in the first described individual, I could find only one gastro-hepatiec portal, which was anterior in position. 6. On two Points in the Anatomy of the Lacertilian Brain. By F. E. Bepparp, F.R.S., Prosector to the Society. [Received May 17, 1905. ] (Text-figures 39 & 40.) (1) Note on the Cerebellum in Varanus exanthematicus. In the account of the Lacertilia in Bronn’s ‘Thierreichs’ * the following statement is made concerning the cerebellum of Varanus :— Das Cerebellum oder das Hinterhirn ist gewohnlich ein unpaarer, diner, steil und hoch aufsteigender Korper, der seitlich mit der Medulla oblongata fest zusammenhaingt. Bei * Bd. vi. p. 714. 268 MR. F. E, BEDDARD ON THE [June 6, manchen Gattungen, z. B., bei Varanus, Iguana, ist es nach den Angaben von Stannius zwar diinn, aber schildformig, vorne concav, hinten convex und zeigt Andeutungen einer Sonderung in eine mittlere und zwei seitliche Erhabenheiten, durch sehr schwache Vorragungen, zwischen denen Spuren von Furchen legen.” It seems plain from the above account that the cerebellum of Varanus is considered to be like that of Zgwana, and, presumably, of other Lacertilia. In one of the most recent works dealing with the brain of the Sauropsida, the Catalogue of the Museum of the College of Surgeons *, there is a description of the brain of Varanus and some incidental references to the brain in the Lacertilia. Of the brain of Varanus it is remarked that ‘the cerebellum is of moderate dimensions and has the plate-like form usual among Reptiles.” Elsewhere (p. 110) it is said that “the reptilian brain is narrow... , and, except in swimming forms, with insignificant cerebellum.” I have examined this specimen myself and agree with the description. None of these statements, as I think, does justice to the cerebellum of Varanus exanthematicus, which is not at all like that of Zguana, has not a plate-like form, and is not insignificant—comparatively speaking, at any rate. The accompanying figure shows the characteristics of the Text-fig. 39. Text-fig. 40. Chis eae Text-fig. 39.—Lateral view of brain of Varanus exanthematicus (upper figure) and of Tupinombis nigropunctatus (lower figure). Text-fig. 40.—Dorsal view of brain of Varanus exanthematicus. ce. Cerebrum ; ce. Cerebellum ; op. Optic lobes. cerebellum of the Teguexin Lizard (text-fig. 39), which appears to me to be quite typical of the Lacertilia and to bear out the above quoted statements. It isa plate-like disc convex posteriorly, which as it were lies up against the optic lobes and is propped up * Descriptive and Illustrated Catalogue of the Physiological Series contained in the Museum of the Royal College of Sur zeons of England, vol. ii. p. 113 (2nd ed.). 1905. ] BRAIN OF LIZARDS. 269 by them. It is faintly grooved in the middle line and laterally on each side is a flattened process extending backward rather beyond the rest of the cerebellum. Its insignificant proportions are shown by the fact that the transverse (antero-posterior) diameter of this thin plate is 2 mm., while the corresponding measurement of the optic lobe is 8 mm. As will be seen from text-figs. 39, 40 (p. 268), the cerebellum of Varanus exanthematicus is a much more important structure. Not only the actual but the relative size of the cerebellum is greater. The corresponding measurements to those given above for Tupinambis are for Varanus—diameter of cerebellum 4:5 mm., diameter of optic lobes 45 mm. They are thus equal. The difference in dimensions between the cerebella of the two Lacertilia is due to the exaggeration in Varanus of the boss-like eminence upon the cerebellum of Zupinambis and Iguana. Not only is the cerebellum of Varanus exanthematicus much greater in bulk than that of Tupinambis or Iguana, but it is more complicated in structure owing to furrows. The dorsal furrow, continuous with that dividing from each other the corpora bigemina, is more deeply marked in Varanws and more definitely circumscribed than in Zupinambis ; in Iguana I did not find any traces of it. Im the second place, the cerebellum of Varanus exanthematicus has an equally deeply marked lateral furrow, which runs obliquely upwards and forwards. Thirdly, the lateral process of the cerebellum is much more sharply marked off from the cerebellum itself than in Zupinambis, and runs downwards rather than backwards, thus distinctly suggesting the flocculus in the cerebellum of the higher forms. It is, indeed, not at all unlike the cerebellar flocculus in Alligator. It is plain therefore that the cerebellum of this Lizard is not ‘‘a mere transverse plate,” but an organ of some dimensions, and, indeed, not very far, in point of relative size, from that of the Crocodilia. A large cerebellum has been associated in reptiles with the swimming habit. And it is true that the Monitor Lizards are often largely aquatic in habit. Curiously enough, however, the present species, with its large cerebellum, is stated by Dr. Giinther* not to take to the water. More likely, as it appears to me, is this advance in structural complexity of the brain to be associated with the not only isolated but high position which the Monitors occupy among the Lacertilia. (2) On the Cerebral Hemispheres in Tropidurus hispidus. T imagine that I am right in believing that the brain of this TIguanoid Lizard has not up to the present been submitted to anatomical examination. I am able, therefore, to add a fact of * “On the Anatomy of Regenia ocellata,” P. Z.S. 1861, p. 60. 270 ON THE BRAIN OF LIZARDS, [June 6, some little interest to what isalready known about the Lacertilian brain, as a result of the examination of two brains of this Lizard. In the Lacertilian brain generally, so far as my own knowledge and the inspection of published figures* enable me to state, the optic lobes lie behind the cerebral hemispheres, the furrow between. them being practically vertical; there is, in fact, no trace of an overlap of the corpora bigemina by the hemispheres. In the Chelonia, on the other hand, it has been recognised that some forms show an overlap of the corpora bigemina by the cerebral hemispheres, IT have found this lobe very obvious in a brain of the large Testudo vicina, the vascular system of which I have recently described. The overlap, however, is lateral and not dorsal. It is quite different with Zropidurus. There isa very distinct overlap of the corpora bigemina by the hemispheres dorsally. The corpora bigemina are thus partly hidden when the entire brain is viewed on the dorsal aspect. A comparison of the measurements of the brain in this species and in Jguana tuberculata seems to throw some light upon the causation of this overgrowth of the cerebral hemispheres over the corpora bigemina, dorsally. The following are the measurements to which I desire to refer : Iguana. Tropidurus. mm. mm. Length of brain to end of cerebellum ... 16 11 Length of cerebral hemispheres ......... i) 6 Breadth of cerebral hemispheres ......... 115 6:5 It will be observed, from a comparison of these figures, that the proportions between the total length of the brain in the two Lizards, and both the breadth and length of the cerebral hemi- spheres, are about equal. It therefore results that the overlap of the hemispheres in Z'ropidwrus is rendered necessary by the skull formation and consequent lack of room for increased lateral growth of the hemispheres. By growing over the corpora bigemina, the hemispheres have been able to attain to the proper size necessary to the equilibrium of their possessor. These considerations may be regarded, perhaps, as discounting the morphological importance of the partial covering over of the corpora bigemina by an extension backwards of the cerebral hemispheres. Nevertheless, it is impossible to overlook the fact that there is an approximation in the brain of this Lizard, to whatever cause it may be due, to those of higher Vertebrates. * See Bronn’s Klassen u. Ordnungen des Thierreichs, Bd. vi., and Meyer, Zeitschr. wiss. Zool., Bd. lv. (1898). + Supra, p. 67. 1905. | ON SOUTH-AFRICAN COLEOPTERA. 271 7. On new Coleoptera from South Africa collected by Dr. H. Brauns and others—Serricornia, Endomyclide, Erotylide. By H. 8. Gornam, F.Z.8. [ Received May 19, 1905. ] This paper is a sequel to those published by me in the ‘ Annals and Magazine of Natural History ’ for 1900-1901. The material dealt with is similar in character to that contained in the col- lections sent to me by Mr. G. A. K. Marshall, but is without Coccinellidee or Languriide. The publication of the paper has unfortunately been much delayed, owing to domestic reasons. The Cleride are particularly well represented and indicate a very rich fauna in this subfamily as well as in the Melyride. Hedybius represents the European J/alachiws, and is evidently (with its allied genera) as rich in species. TELEPHORIDSE. TELEPHORUS VIRIDESCENS Fab, Telephorus viridescens Fab, Syst. Kleuth. i. p. 295 (Cantharis). Hab. Willowmore, Uitenhage, Cape Colony (Srawns). In the Munich Catalogue this species is given as a synonym of Cantharis smaragdulus Fab. Spec. Ins. p. 259, a Brazilian insect ; but Fabr. loc. cit. gives “Cap bon. spei” as the locality of the Cantharis viridescens, with which my specimens agree very well.* TELEPHORUS INcIsus Wied. Zool. Mag. ii. p. 71. Hab. Algoa Bay, Cape Colony (Lrauns). 2 examples. Smaller than 7’. viridescens, with a thoracic vitta from the front to the hind margin and without spots on the sides or at the base ; elytra less green, of a dull grey-black. TELEPHORUS zZoNATUS Gemm. Cat. Col. p. 1674. Telephorus vitticollis Bohem. Ins. Caffr. 1. 2, p. 453, nec Meénetr. Cat. Rais. p. 162. Hab. Algoa Bay, Cape Colony (brawns), 1 example. Described by Bohem. Joc. cit. from “ Limpopo” R., Caffraria. TELEPHORUS BIVITTATUS Fab. Nee 7. bivittatus Mars. Hab. Algoa Bay, Cape Colony (Brawns). 1 example. TELEPHORUS NiGRINUS Bohem. Ins. Caffr. i. 2, p. 457 (Cantharis) ? Hab. Algoa Bay, Cape Colony (Srauns). 2 examples, In one of the two examples sent the thorax is nearly twice as wide as long, in the other example (which, from the length of its antenne, appears to be a male) itisquadrate. This seems to me to agree with the insect described by me (Ann. & Mag. N. H. 1901, 272 REY. H. 8. GORHAM ON [June 6, vil. p. 351) as 7’. teter from Natal. From so few examples it is im- possible to say whether these two pertain to one or two species ; or what is the sex of the specimen with transverse thorax, which is also more shining and has shorter antenne. I incline to the belief that they represent two species, the latter being referable to C. nigrina Bohem. CLERID &. EucyMaroperA Schenkling. EucymatoderaSchenkling, Ann. Mus. Civ. Genova, 1899, p. 333; Genera Ins. fase. xiii. p. 19 (1903). EUCYMATODERA CINGULATA Klug, Cler. p. 273 (Zllus cingulatus). Hab. Algoa Bay. EUCYMATODERA HOTTENTOTA Knw. Ann. Soc. Ent. Belg. p. 463 (1893) ; Schenk. /. ¢. p. 20. Hab. Willowmore, Cape Colony (Lrauns). Cyxiprus Lat. Cylidrus Schenk. Genera Ins. fase. xiii. p. 5 CYLIDRUS BALTEATUS Klug, Cler, p. 263. Hab. Bothaville, Orange R. State (Lrauns). I had not seen this from 8. Africa before. Gyponyx Gorham. Gyponyz Gorham, Ann. Mus. Civ. Gen. 1883, p. 604; Schenk. lic. p. 45. GYPONYX CHINENSIS Fab. Gyponyx marmoratus Klug, Cler. p. 308 (Clerus), nota p. 379. Hab. Bothaville, Orange R. State (Brawis). GyPoNyX RETROCINCTUS Chevr. Rev. Mag. Zool. p. 283 (1874). Hab, Sunday River, Cape Colony (Lrauns). GYPONYX ALGOENSIS, sp. n. Oblongus, subparallelus, piceo-brunneus, nitidus; elytris basi dilutioribus, ultra medium fascia undulata et apice albis. Capite crebre prothorace parce punctatis, hoc pernitido, antice tenwiter constricto, postice coarctato; antennis et. palpis rufo- piceis ; elytris usque ad fasciam grosse seriatim punctatis, inde ad apicem fere levibus ; pedibus piceis, tarsis dilutioribus, metasterno punctato. Long. 11-14 millim. Hab. Algoa Bay (1. Brawns). The general colour of this species is dark pitchy brown, the antenne, palpi, tarsi, and the base of the elytra nearly as far as the white fascia are rufo- piceous, the elytra are blackish in an 1905. | SOUTH-AFRICAN COLEOPTERA. 273 indefinite way just before the very definite white fascia; this is narrowly interrupted at the suture; the apex is white, but with a fine blackish margin; the space between the fascia and the apex is very obsoletely punctured, as is the fascia, almost smooth externally. The eyes, head, legs, and body generally are clothed with long but fine hairs. One example of this insect received long ago from Dr. Baden is in my collection, but bad, unfortunately, no precise locality. It is the larger specimen. GYPONYX BRAUNSI, sp. 0. Hlongatus, nigro-piceus; ore, antennis palpisque, thoracis margine antico, corpore subtus, pedibus (geniculis tibiisque exceptis) dilutioribus, rufo-piceis. LHlytris basi indeterminate, fascia mediana (in marginem latissima postice bidentata) maculaque apicali obliqua testaceis. Thorace elongato, antice tenuiter, ad basin fortius constricto ; elytris ad basin punctato- lineatis, seriebus vin ad medium vectis, externe deficientibus. Long. 13 millim. Hab. Willowmore, Cape Colony (Brauwns). This species differs from any other known to me by its elongate thorax and its comparatively smooth and shining appearance, and also by the clear and distinct coloration and pattern of the elytra. The colour of the body, with the exception of the head and thorax, of the tips of the femora, and bases of the tibie, is a bright rusty red. The elytra have an oblique broad fascia, much indented, running backwards from below the callus to the suture, of a pitchy colour ; a much broader patch before the apex of a lighter pitchy brown, deeply indented twice on its upper edge, and once on its lower edge, so that it is narrowest in the middle. The lines of punctures scarcely pass the first brown fascia, only four or five punctures being on the yellow median wide patch. The thorax is nearly twice as long as wide, its front margin is rufous, the sides very little widened, not deeply constricted in front, the base coarctate and margined, the punctuation close and fine, the disk a little flat. I have at present only seen the example described, which I have pleasure in naming after its captor, Dr. H. Brauns. GRAPTOCLERUS Gorh. Graptoclerus Gorh. Ann. & Mag. N. Hist. ser. 7, vii. p. 351 (1901); Schenk. /. c. p. 48, nota. GRAPTOCLERUS QUADRIPUNCTATUS Gorh. J. c. p. 353. Hab. Grahamstown, Cape Colony (Dunkerbosh, Dr. Penther). Described by me from Natal. 1 example. TARSOSTENUS Spinola. TARSOSTENUS UNIVITTATUS Rossi. Hab. Willowmore (Brauns). Proc. Zoo, Soc.—1905, Vou, II. No. XVIII. 18 bo x ns REV. H. 8. GORHAM ON [June 6, DozocouLetus Chevy. Dozocolletus Schenk. J. ¢. p. 38. Pezoporus Klug, Clerii, p. 311. DozocoLLETUS BRUNNEUS Hintz, Deuts. ent. Zeit. 1902, p. 397 2 Hab, Willowmore, Algoa Bay (Brauwns). Bothaville, Orange R. Colony. T have not seen Hintz’s description, but I have little doubt that the insect taken in some numbers by Dr. Brauns is to be referred to it. This species has a large head, the thorax as wide in front as the head, the head and thorax deep pitchy brown, the elytra, legs, and body light brown, the punctuation and striz are obsolete. DozocoLLetus SORDIDUS, Sp. 0. Saturate brunneus ; antennis palpisque rufo-brunneis, lis quam caput et thorax brevioribus, articulis quarto ad octavum quad- ratis, tribus ultimis transversis, apice compresso; capite prothoraceque creberrime confluenter, elytris fortiter striato- punctatis, femoribus clavatis. Long. 5 millim. Hab. Algoa Bay (Brauns). Smaller than the species which I have assumed to be D. brunneus Hintz, and especially to be distinguished from it by the shorter antenne, which have much shorter joints; the second and third joints are a little longer than wide; the fourth to the eighth are about as long as wide, while the last three joints form a compact club and are transverse. The eyes are more prominent than in D. brunneus, the femora are strongly clavate. The thorax is as wide as the head in front, much narrowed to its base, obconic; the elytra small in proportion, elongate-ovate. TuHRIocERA Gorham. Thriocera Gorham, Trans. Ent. Soc. Lond. 1878, p. 156; Schenk. l.c. p. 117. THRIOCERA PECTORALIS Klug, Clerii, p. 348. a. Elytris unicoloribus fascia mediana plicata. Hab. Algoa Bay, Port Elizabeth (Lrauns). 6. Elytris bast rujis. Hab. Algoa Bay, Port Elizabeth (Brauwns). THRIOCERA BICINCTELLA, sp. 1. Nigra, nitida, tenwiter pubescens ; antennis, palpis pedibusque rufo-brunneis, illis basi dilutioribus ; prothorace brevi, antice tenuiter constricto, postice coarctato, lateribus rotundatis, disco ineequali impresso ; elytris sublevibus, fasciis duabus elevatis, eburatis, ad suturam interruptis, anteriore (juxta callum) intus abbreviatis. Long. 5 millim. Hab. Algoa Bay (Brawns). 1905. | SOUTH-AFRICAN COLEOPTERA. 275 Var., capite, antennis, palpis, elytris usque ad fasciam postervorem, corpore subtus cum pedibus rufo-ferrugineis. Hab. Port Elizabeth, Cape Colony (brauns). Antenne as in 7’. pectoralis, but less robust, the two basal joints stout, the intermediate joints longer than wide, the three apical forming a lax club. The thorax is notably shorter than in T. pectoralis and of a different form, being much more narrowed towards the base; its disk is also roughly punctured and uneven, finely pubescent with long hairs, but neither it nor the elytra are so thickly pubescent as in 7’. pectoralis. The elytra are black, or, as in the variety, red to the second fascia. The fasciz are raised and ivory-white, not clothed with silky-white hairs (as they are in 7. pectoralis), but clear shining white; the anterior one is shortened, so as to leave a space at the suture equal to its own length, nor does it reach the margin. There are two examples of the type form and two of the variety. NorostEnus Spinola. Notostenus Schenk. t.c. p. 114. Norosrenus viripis Thun. Nov. Ins. vol. i. p. 9 (1784). Hab, Algoa Bay (rawis). MELYRID2. ANTHOCOMUS Hrichs. ANTHOCOMUS CORTACEUS, Sp. Nn. Breviter oblongus, niger, obscure subviridescens, longe pubescens ; capite prothoraceque nitidis, hoc profunde parce punctatus ; elytris coriaceis; antenwis (apice exceptis), tibiis tarsisque ferrugineis. Long. 4 millim. Hab. Bothaville, Orange R. Colony (Lrauns). Black, densely clothed with long black pubescence. Head and thorax shining, the former very closely and finely, the latter very sparsely punctured. Clothed all over with long fine hairs. The mouth, palpi, antennz (excepting the extreme tip), the tibie, and tarsi are ferruginous red. The elytra are little shining, coriaceous, tubercles obscurely in rows and but little elevated, being rather uniform all over their surface. There is a slight greenish, but hardly perceptible, tint over the whole upper surface ; the body beneath is quite black. As there are only two specimens, and I cannot distinguish the sexes, it is impossible to say if thisis a true Anthocomus. "The lamellze of the tarsi are about as long as the claws. Hepystius Erichson. HEDYBIUS SCULPTICEPS, Sp. n. Nigro-subceruleus ; capite (basi excepta), antennis (articulis extus et apicem versus nigro notatis), prothoracis margine, pedibus Tigi 276 REV. H. 8. GORHAM ON [June 6, anticis et intermediis (eaterne nigris) testaccis ; elytris cewruleis, purpureo vel violaceo micantibus. Long. 55-65 millim. 3°. Mas, capitis basi profuwnde excavato-eroso, erosione ima tota nigra ; fronte elevata, in medio profunde sulcata. 480. From this point and passing over the edge of the placenta, and covering the free surface of the feetal side of the placenta, a flat- tened attenuated epithelium can be distinguished (text-fig. 45, H, p. 285). This becomes thicker and more cubical as it nears the point at which the yolk-sac wall is connected with the placenta, and here it passes into the decidedly cubical or columnar epithelium of the yolk-sac. This layer continued in the other direction would pass at some period into the distal wall of the yolk-sac, though whether this distal wall exists at the period under examination I cannot say. The rough surface of vascular attachment, so far as I can judge from the general character of the cells, is composed entirely of maternal tissue. This tissue is of that kind so frequently found where trophoblastic ingrowth is about to take place, and had been named by Hubrecht trophospongia (text-fig. 44, p. 284). Text-fig. 45 is a diagrammatic representation of a section passing through the centre of the placenta. The placenta, as 284 MR. R. ASSHETON ON THE F@TUS AND [June 6, stated above, had been separated from the wall of the uterus, so although shown im sié#w in the figure it must be understood that the two parts were not together in my specimen. The line 8 marks the boundary between the two. On the outside I have drawn in outline the muscle coats (M) of the wall of the uterus. Within this, and seen only near the centre, is the trophospongial tissue alluded to above (TS), while towards the peripheral parts the thin layer of detritus can be seen at D, composed probably of both maternal and foetal tissues. Text-fig. 44. A section through the junction between trophoblast and trophospongia of Acomys. T. Trophoblast. Ts. Trophospongia. EP. Pseudoepithelium of trophoblast. EP’. Pseudoepithelium of trophospongia. MCH. Maternal blood. Everything within this line formed by the detritus layer on the outside and the trophospongial layer at the centre is probably foetal in origin, except the maternal blood, which is extravasated and flowing in channels excavated in the ‘foetal trophoblast. It must of course be remembered that this description is an interpre- tation of a single stage based upon the known facts in closely allied forms (ref. Duval, Robinson, Jenkinson), and not upon the study of the actual development in this genus. 1905. ] PLACENTA OF THE SPINY MOUSE. 285 The advancing edge of the trophoblast is sharply defined from the maternal tissue (text-fig. 44, T). The trophoblast is composed of a compact mass of cells with large nuclei and generally fairly well-marked cell-boundaries. It has, in fact, the character more of a cytotrophoblast than a plasmodi-trophoblast (text-fig. 44). The nuclei tend to arrange themselves in pairs. Text-fig. 45. MA A diagram of a section taken through the centre of the placenta of Acomys. The maternal channels are not much exaggerated in size, but the fcetal capillaries are considerably magnified. The trophoblast laver in the region of the foetal capillaries is not nearly so much attenuated as it should be. The foetal vessels are dotted, the channels containing maternal blood are white. The deep black is trophoblast. AA. Allantoic artery. AV. Allantoic vein. D. Layer of cell-detritus probably foetal as well as maternal. FC. Fetal capillary. H. Hypoblast of the parietal wall of the yolk-sac. HH. Hypoblast of the proximal wall of the yolk-sac, very much folded. L. Lacune in trophoblast containing maternal blood. M. Muscle-coat of uterus. MA. Maternal artery. MES. Mesoblast of fcetus. MV. Maternal vem. S. Line along which the placenta had been detached from the uterus. SP. Splanchnopleur layer of mesoblast. ‘IT. Trophoblast. TS. Trophospongia. WV. Approximate portion of the main vitelline vessels. Y. Cavity of the yolk-sac. YC. Blood-vessels of the yolk-sac circulation. This trophoblastic tissue, which in the figure (text-fig. 45) is shown as a thick black mass (T), 1s honeycombed by channels containing maternal blood (LL), which channels become more broken and more numerous nearer to the fetal surface, and the trophoblast consequently more attenuated. I think there are possibly other cavities in the trophoblast-cells which are not blood-spaces. This mass of tissue formed entirely of trophoblast and maternal 286 MR. R. ASSHETON ON THE F@TUS AND [June 6, blood makes up nearly half the thickness of the placenta, and contains no foetal mesoblast or blood. The half of the placenta towards the feetal surface is made up of trophoblast (much attenuated) forming channels filled with maternal blood, which take a more or less sinuous course, and a network of fine fatal capillaries, with also the larger vessels and larger main maternal channels. This is shown diagrammatically in text- fig. 45, FC, p. 285). Text- fig. 43 (p. 283) is a drawing of an actual section of a piece of this region near the maternal surface. The great bulk is made up of the ‘channels (MCH) excavated in the trophoblast containing maternal blood. There are many leucocytes (LE), The walls of these channels are thin, though the large trophoblastic nuclei (T) are very conspicuous. The feetal capillaries are seen at F.BY. Nearer to the feetal surface the maternal channels become finer and the feetal capillaries perhaps rather more numerous. At places where the main feetal arteries penetrate the tissues of the placenta, a considerable quantity of foetal mesoblast tissue accompanies them. There are a few spherical masses of tissue within this region, which are not vascular, nor do they seem to be trophoblastic. They resemble in some respects Duval’s “ilots vésiculeux,” which, according to him, are pieces of the maternal sub-mucosa which have become enveloped by the advancing trophoblast layer. The main features of the vascular systems are fairly easily determinable. In this specimen the whole of the maternal arterial blood- supply arises from a single artery in the centre (MA), which opens into the large afferent channel which lies partly in the trophospongial tissue and partly in the trophoblast. This, like the other main channels, is lined by a flattened epithelium-like layer, which is probably a pseudo-epithelium of trophoblastic origin where the wall is trophoblast, and tropho- spongial origin where the wall is trophospongia. Duval has described the growth inwards along the maternal vessels of trophoblast cells to form a pseudo-epithelium. This is denied by Jenkinson*, who derives the pseudo-epithelium from the simple flattening of the adjacent cells. This is not a question which can be decided by reference to a single stage; but I may say that there is nothing in this specimen which supports in any way Duval’s account in the mouse. The afferent channel divides into two main branches, which diverge and then penetrate straight to the fcetal surface of the placenta. Here they break up into channels, which take a rather more sinuous course back again to the middle of the thickness of the placenta, where they collect into a number of efferent channels lying near the surface of the trophoblast and ultimately into two * Jenkinson, J. W., “ Observations on the Histology and Physiology of the Placenta of the Mouse,” Tijdschr. d. Ned. Dierk. Vereen. DI. vn. 1902. 1905. ] PLACENTA OF THE SPINY MOUSE. 287 large efferent vessels which presumably debouch into two maternal veins (MV). These large efferent maternal vessels are more peripheral in position than the afferent courses. The fcetal blood-supply is less easily followed. The allantoic arteries, on reaching the surface of the placenta, radiate and subdivide and give off branches which pass into the placenta at intervals over the more central part of the surface. These pass (as regards the main stream) straight through the fetal half of the placenta, but give off at frequent intervals small capillaries, which take a more sinuous and radial course, anastomosing, forming network, and collecting together again into veins, which I believe to run parallel with the arteries of the villus. A considerable amount of connective tissue accompanies these villi for the first part of their ingress into the placenta. Comparison with other Placentas. The placenta is clearly of the type which I have described elsewhere as cumulate as contrasted with the plicate type characteristic of such groups as the Ungulates. The general shape, the arrangement of the membranes and their character, resemble, according to the interpretation placed upon them above, the condition that pertains in J/us musculus, and, rather less closely, that of J/ws decumanus. Jenkinson has recently given an account of the development of the placenta of the Mouse, which differs in several important particulars from that of Duval, which, until that time, had been generally accepted. Jenkinson agrees with Duval as regards the general arrangement of membranes and in the main features of the development. He differs, however, in respect to the vascular system, and my account given above is in complete agreement with Jenkinson. I see also nothing to suggest the appearance of an ingrowth of the trophoblast into the maternal blood-vessels, and growth along the inside as described by Duval (“‘ plasmode endovasculaire ”). Another point of disagreement is with reference to the glycogenous tissue. Duval takes but little notice of this and does not seem to have found the maternal glycogenous mass, which, according to Jenkinson, degenerates, and the space occupied by it becomes subsequently filled with a second glycogenous tissue which is of foetal (trophoblastic) origin. This is a matter of very considerable interest. It is not to be expected that an isolated case in an allied genus can afford any conclusive evidence. Unfortunately the method of preservation (corrosive sublimate) is not suitable to the study of glycogen. Tested with iodine I find no trace of glycogen, but there are certain spaces in the trophoblast cells alluded to in the foregoing 288 ON THE FG@TUS AND PLACENTA OF THE SPINY MOUSE. [June 6, which may possibly have contained glycogen—so my evidence on this point is negative. Along the line which marks the internal limit of the fcetal mesoblast (that is to say, the limit of the capillary system of the foetal villi) a deposit of brown pigment occurs. This appears to be deposited in the walls of the capillaries, by the endothelium. It occurs nowhere else. Treated with the ferrocyanide test, it shows no trace of blue colour. In the detritus in the layer D (text-fig. 45, p. 285) there are indications of the presence of free iron. Summary. To recapitulate my interpretation of the single specimen I possess. The placenta of Acomys cahirinus is a compound structure of maternal and feetal tissues in which, excepting the blood, the feetal tissue largely preponderates. On the maternal side is a central area of attachment through which the maternal blood gains access to the placenta. Here a thin layer of maternal connective tissue surrounds the main afferent and efferent maternal blood-channels. Within this region comes a thick layer of tissue probably of feetal origin (the trophoblast, the cells of which are large, stain deeply, and have large nuclei), containing intercellular spaces, which are continuous with the expanded maternal vessels just named. These spaces are lined by an endothelium, as to the origin of which I can give noaccount. There is no feetal blood in this part of the placenta. These two regions, of which the latter is by far the larger, make up nearly one half of the whole placenta. The rest (that is, all towards the foetus) is composed of channels probably excavated in the trophoblast of the foetus and containing maternal blood interlacing with much branched tufts of feetal capillaries containing fcetal blood. These foetal capillaries are in parts thickly covered with fcetal mesoblastic tissue, but more often are separated from the maternal blood by their own endothelium and a single layer of trophoblast only. The maternal afferent channels penetrate to the feetal surface before undergoing much subdivision and are more central in position. The fcotal afferent vessels tend to penetrate the deeper layers, but begin to break up nearer to the surface of approach than is the case with the maternal afferent vessels. There is no such intimate connection between the yolk-sac and allantoic placenta as there is in the Rat, 1905. ] ON THE SUPPOSED CLAVICLE OF DIPLODOCUS. 289 9. Remarks on the supposed Clavicle of the Sauropodous Dinosaur Diplodocus. By Francis, Baron Nopcsa, EheD [Received June 6, 1905. ] (Text-figures 46-49.) It is still uncertain whether the extinct Dinosauria possessed clavicles. Considering the close relationship existing between these reptiles, the Rhynchocephalians, Parasuchians, and Birds—this last relationship being shown by the continuous tendency of Dinosaurs to specialize on most different occasions in bird-like manner—one is at first naturally imduced to believe that in Dinosaurs clavicles were present; but, as a matter of fact, bone after bone supposed to represent this element has had to be removed from this position. Hitherto only the family Ornithopodide is known to possess, in addition to scapula and coracoid, a curious further element in the shoulder-girdle, which was called clavicula, but may quite as well form only a part of the sternum (this double element being in one case united in the middle by bony matter). No other Saurischian or Orthopodous Dinosaur shows a clavicular ossi- fication. It is true that in the Sauropoda, besides scapula and coracoid, one or two flat bones are always present in the scapular region of the body: these, however, do not represent clavicule, but may with certainty be determined as ossifications of the sternum. ‘The discovery, therefore, of what may be called a supernumerary bone besides the sternal plates in two of the several Diplodocus skeletons known to science proves to be of quite exceptional interest. Hatcher, in his important Monographs of the Diplodocus skeletons Nos. 84 and 662 of the Carnegie Museum, describes this element as follows :—“ Throughout the greater portion of its length it is circular in cross-section, it is bifid at one extremity and slightly expanded at the other. It is strongly curved, especially toward the bifid extremity. It is asymmetrical.” In a more complete specimen (No. 662) than the former (84) it is ‘“‘somewhat expanded and spatulate; the flattened extremity presents a slightly rugose surface, as though it had been imbedded in cartilaginous or muscular tissue, and this together with the bifid nature of the other extremity has suggested the possibility that the bone might be an os penis.” After the description of this bone, however, its asymmetry is regarded by this eminent paleontologist as a weighty argument against its being an os penis, and therefore its identification with the clavicula is advocated. * Communicated by Dr. A. Smita Woopwarp, F.R.S., F.Z.8. Proc. Zoou. Soc.—1905, Vou. II. No. XIX. ig) 290 - BARON NOPCSA ON THE SUPPOSED [June 6, According to the figures given by Hatcher and reproduced here (text-figs. 46 & 47) the bone in question seems to present a great deal of what might be termed individual variation. Text-fig. 46. b SMUG Supposed clavicula of Diplodocus, No. 84. Text-fig. 47. Same bone of Diplodocus, No. 662. It seems_to fit fairly well into the shoulder-girdle, but still 1905. | CLAVICLE OF DIPLODOCUS. 291 there are several points to be brought forward against the theory of its clavicular nature. Firstly, it must be remembered that in one case this problematical bone, like the greater part of the skeleton, was displaced and that in the second skeleton, as pomted out by Dr. Holland, the femur In consequence of bears tooth- marks of carnivorous Dinosaurs. this the relative position of the bone cannot prove anything for or against its being an os penis; for the penis would be one of the first parts of the body to become displaced by decomposition and the first part that would be torn away if carnivorous animals were gnawing at the dead body. Text-fig, 48. "Ny, ca NUM COMM ag Hutte Wi] Diagram of penis of Struthio. Explanation of letters :—c.c., corpus cavernosuin; ¢.f., corpus fibrosum ; c.sk., coarse skin; g., gutter; gl., glans-like part; im., muscles. comprised this bone, only one example was present, and this one In other appeared to belong to the same side of the body. Secondly, in each of the two pretty complete skeletons that specimens of Diplodocus the element was altogether wanting, 292 BARON NOPCSA ON THE SUPPOSED [June 6, This highly remarkable coincidence suggests the probability that the bone in question represents an asymmetrical but nevertheless unpaired organ. So far as I am aware there is no known reptile, living or extinct, in which the clavicle is bifurcated at one end. Moreover, in most terrestrial and aquatic reptiles, when clavicles are present there is also an interclavicle, which has never been found in Sauropoda. It must also be remembered that these large herbivorous Dinosaurs were probably descended from the carni- vorous Theropoda, which are always destitute of a clavicular arch, Text-fig. 49. Os penis of European Otter. [ am therefore of opinion that the problematical bone of Diplodocus in question cannot be a clavicle, and it is necessary to consider Hatcher’s alternative suggestion that it is an os penis. The fact that existing birds and reptiles are destitute of an os penis does not necessarily imply that gigantic reptiles like Diplodocus similarly lacked the bone. Among Mammalia it is well known that the element occurs only sporadically, being present, for instance, in the Anthropoid Apes and absent in Man. Among the living reptiles we know two types of genital organs. The Squamata show what may be called a bifid penis, while the Crocodilia and Chelonia have the penis simple exteriorly, with a corpus fibrosum and frequently even a glans penis well developed. 1905. | CLAVICLE OF DIPLODOCUS, 293 In Chelonia the penis sometimes exhibits internally a partially bifid structure. For the purpose of this paper the penis of birds is of quite exceptional interest. In its origin it is not only traceable to the Crocodilian type, but shows a very great amount of asymmetry, and besides in the Ratite a distal bifurcation of the corpus fibrosum (text-fig. 48, p. 291). In Séruthio the distal part of the penis is changed into a glans-like organ, while in Rhea the corpus fibrosum consists of an exceedingly hard and nearly cartilaginous substance. A bifurcation like that observable in the problematical bone of Diplodocus is also frequently to be met with at the distal end of the mammalian os penis, which is often asymmetrical. The os penis of mammals always shows quite remarkable variability. For comparison with the bone of Diplodocus, side and hind views of the os penis of the European Otter (Zuéra lutra) are given (in text-fig. 49, p. 292), and one can see at a glance the well-rounded, smooth, condyle-like, distal ends, the proximal rugosities, and the lateral impressions for the attachment of the corpus fibrosum. In other mammals the corpus fibrosum is not attached laterally to the ossified element, but ends in a deep pit situated at the proximal end of the latter. We have therefore to consider the following propositions :— (1) That among the Mammalia it is the corpus fibrosum with which the os penis comes in close contact, forming the anterior prolongation into the glans penis, that the os penis ossifies from fibrous matter; that a corpus fibrosum is algo present among Reptilia, and that therefore an os penis in Dinosaurs can only have originated from the corpus fibrosum. (2) That in hea the corpus fibrosum is quite as hard as cartilage, and differs from this only by not possessing cartilage-cells. (3) That in Sauropsida a glans is frequently present. (4) That it is quite a common thing to find bird-like characters in various parts of the Dinosaurian skeleton. 5) That among the birds the Ratites show the most primitive and still the best-developed male genital organ. (6) That the shape and variation of the problematical bone in Diplodocus are well in accord with its being an os penis, while they militate against its determination as clavicular. (7) Lastly, that this so-called clavicula when present is always found only as an unpaired organ showing the same direction of curvature. Hence I am of opinion that it is at present advisable to remove the subject of this paper from the shoulder-girdle and determine it as the ossified axis of the penis. Further evidence and, especially, further discoveries are naturally necessary before so delicate a question can be regarded as definitely settled; but since Hatcher's single argument against the bone in question being an os penis (namely, its asymmetry) breaks down on reference to Struthio or even to Lwtra, the balance of the argument is at present in favour of this newer interpretation. Proc. Zoou. Soc.—1905, Vou. II. No, XX. 20 294 ON THE SUPPOSED CLAVICLE OF DiPLoDocus. [June 6, ~The existence of clavicles in Dinosauria must therefore still be considered doubtful. In conclusion, I wish to express my thanks to Mr. Boulenger, Dr. Forsyth Major, Mr. Pycraft, and Dr. A. 8. Woodward at the British Museum, and to Professor Stewart and Mr. R. H. Burne at the Royal College of Surgeons, for their kind help in studying so intricate a question. Literatwre. Gapow & SELENKA.—Birds: Bronn’s Klassen und Ordnungen d. Thierreiches, vol. vi. part iv. Gitper.—‘‘ Das Os priapi der Siiugethiere.” | Gegenbauer’s Morphologisches Jahrbuch, 1891. Harcupr. — “‘ Diplodocus carnegii.’ Memoirs of Carnegie Museum, Pittsburg, vol. 1. ——. ‘“Osteology of MHaplocanthosaurus, with Appendix on Diplodocus.” Loc. cit. vol. 1. [ Norcsa. |—Critical Review of Hatcher’s Paper on Haplocantho- saurus. Keilhack’s Geologisches Centralblatt, 1904. Weser.— Die Saugethiere. Leipzig, 1903. Wrepersuemm.—-Lehrbuch der vergleichenden Anatomie d. Wirbelthiere. Leipzig, 1902. 5, On the Anatonry of Limicoline Birds; with special Reference to the Correlation of 6. Contents (continued). May 16, 1905 (continued). : Pave Modifications. By P. Caaumers Mircnent, M.A., D.Sc. (Oxon.), Secretary to the Society: «15... BRED STDS Hinrtho > mcde orem Ele SiSis here owe PSIG Geico Welvlelsinrcisisiate cot Lae Observations upon a Female Specimen of the Hainan Gibbon (Hylobates hainanus), now living in the Society's Gardens. By R.I. Pocock, ¥.L.S., F.Z.8., Superintendent PPPOE IoeN eur Biter CEM MLE SVEN (iia wceia say cus ly «ty opealn dialet sy noe Sane SL ha _ 160 June 6, 1905. The Secretary. Report on the Additions to the Society’s Menagerie during the month of yi p Y 5 g Mr. Mr. Mr. Dr. bo Ng MISE Se ce es Oe apa gan Eins eee Bea et oe sake g0 Oldfield Thomas, F.R.8. Description of a new Bush-buck (Trag i haywoodt) RNsmAilarmaR MM MaSt ATTICA a 5:5 6 Alycia e stoselely arse el sua cl eg oS ainke Co eae Se gen tage 180 Oldfield Thomas, F.R.S. -Exhibition of specimens of Mammals and Birds from Japan and description of a new Marten (Mustela melampus bedfordt) ............ 182 R. I. Pocock, ¥.L.8. Exhibition of a Jamaican Scorpion (Centrurus insulanus) CAN GevOUmeTOnvItS: Hacky nt vskelseiaie atehc seine «\sta\isha’a a lalallategerate my alcarat evs al obais tebe eels 183 P. Chalmers Mitchell. Notice of a memoir entitled ‘‘On the Intestinal Tract of JWI aaa ES Spe ie ie tee Re a CR ALG RE ETM ee iD IND SNe mn US 184 . Rough Notes on the Natural History of the Country West of the Victoria Nyanza. By Lt.-Col. C, Duumn-Rapoutern, M.V.O., B.Z.8. 02... see ieee cree eee eee 184 . The Distribution of Mexican Amphibians and Reptiles. By Hans Gapow, E.R.S., . Descriptions of new Reptiles discovered in Mexico by Dr. H. Gadow, F.R.S. By G. A. IBOUUANGHR EE URuS Eva Asis (babes! Vie dor Vali. )ie se. store anal amereteneierayacnaieiscate tale 245 . On a Collection of Batrachians and Reptiles made in South Africa by Mr. C. H. B. Grant, and presented to the British Museum by Mr. C.D. Rudd. By G. A. SCR LONGI SE STU Sts Ver ZS anosaie aoc eu Biena ra aiaccPaca vat’. MyalWaay as onenlaip ft ea stata ue Zao ee a 248 . Some Notes upon the Anatomy of the Yellow-throated Lizard, Gerrhosaurus llavigularis. Byes Hy Banparp, WR,S:, Prosector to the Society? i... <.haee | iowa te same aoe 256 . On two Points in the Anatomy of the Lacertilian Brain. By F. E. Bepparp, F.RB.S., RHOSECEOT LOVGME SOCICL My «leit ake ciliaicial ta aieiaarsnctai Tous tesesiac ch soe payee eee ee SON Mott 267 . On new Coleoptera from South Africa collected by Dr. H. Brauns and others— Serricornia, Endomychide, Krotylide. By H.8..Gorwam, F.Z8. ........-....0 06 271 . On the Foetus and Placenta of the Spiny Mouse (Acomys cahirinus). By Ricwarp Assurton, M.A., F.Z.S., Lecturer in Biology in the Medical School of Guy’s ee) University of London ........ ans als REBUN Ornate le erda eS OG Reno p oe pote 280 . Remarks on the supposed Clavicle of the Sauropodous Dinosaur Déiplodocus. By RAR OLS, ARON: NOTOSA:)EIRSED Ste cuuss clase aincew ciety nce retire Gt” a) Siaioa aes MRE see. 289 LIST OF PLATES, 1905.—_VOL. II. PART ' Plate Page DN OLA TR PIL COMLOILL Was «c/erorutslc le eeioy ele teiattia cia sicher =e (eles eit eae iaie taeets 3 TI. Osteology of the Hurylemide ..........-. 20... cere cece ene 30 TIL.) skulle of Rhinolophi ...00+.0+0e0eceren Pry 825 75 V. Hylobates hainanus .........+ ‘el caches gencaen. aie 4 a(t ear 169 VI. 1. Anolis gadovii. 2. Anolis ogaster 1.0.0. sere cece ss caes \ 2455 VII. 1. Sceloporus gadovie. 2. Leptodira guillent .......+++..+.06 NOTICE. ‘The ‘ Proceedings’ for the year are issued in fowr parts, forming two volumes, as follows:— VOD. Part I. containing papers read in January and February, in June. ia i & », March and April, in August. VOL. II. 7 Part I. containing papers read in May and June, in October. 10 RS “Ns »» | November and December, in April. ‘ Proceedings,’ 1905, Vol. I. Part II. was published on August 10th, 1905. The Abstracts of the papers read at the Scientific Meetings in May and June are contained in this Part. PROCEEDINGS OF THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1905, vol. II. PARL IT; CONTAINING PAPERS READ IN NOVEMBER ann DECEMBER. APRIL 1906. PRINTED FOR THE SOCIETY,: SOLD AT THEIR HOUSE IN HANOVER SQUARE. LONDON : MESSRS. LONGMANS, GREEN, AND ae PATERNOSTER-ROW. Aa {Price Twelve Shillings.) 4 seme LIST OF CONTENTS 1905 Noe ae Part II. sv hae i November 14, 1905. : Page - The Seer ae Repcrt on the Additions to the Society’s Menagerie Ae the months of Jane, July, August, September, andi@ctober 90D 2.0 il citi slm ate taveee sllergnenstens » ) 2o8 Col. W. H. Broun. Exhibition of mounted heads of a White Woterbnck and two HUMGOCETOSAS > cosicvis cle = niv.oceia.a = 00: sicvminpdm caval Blete/e.a\oceein a a'o nelle telel nus a tae lel at 296 The Hon. Walter Rothschild, F.Z.S. Exhibition of specimens of a rare Marsupial CDG CLULOP SIG PAUPALOT) wc, viola nine o:0\e'nin sinlale/'s e's sient) « e'oie'hr)s cain ale els oct eae OG The Hon. Walter Rothschild, F.Z.S. Exhibition of two tusks from Abyssinia ......-... 297 Mr. A. 8. Hirst, F.Z.8. Exhibition of microscopic preparations of a new H:emosporidian (Hallieridiune Crumeniti) rau 2cccjeice eimlals «0 m+ ota savin genes + baie ene ele See 297 _ Dr. Walter Kidd, F.Z.8S. Exhibition of lantern-slides illustrating the Papillary Ridges iniMfammals .... see seSc 2 Sn de, Sule nth Ot. Aah cng aie 297 Dr. P. L. Sclater, F.R.S. Letter from Mr. W. Rodier on the Rabbit-pest in Australia .. 298 Mr, Henry Scherren, F.Z.8. Exhibition of lantern-slides of, and remarks upon, old pictures of Anthropoid Apes a Sle aa lejnise vousteyers ese teiblcatatenose eave sce s) ale .evensy tees te eter eafeic lle cee ee 295 . On a Collection of Mammals brought home by the Tibet Frontier Commissian: By J. imwis Bonnors, MyA., FLL.S.{cB.Z.8. sich eg cieels\aielehals o melinte ai inate ae 2 802 i 2. Notes on the Geographical Distribution of the Okapi. By Dr. ae Léxnsera, C.M.Z.8. 309 8 Notes on the Goral found in Burma. By Major G. H. Evans wer ois lofhoe aerate eta 811 4. On the Mammals of Orete. By Dorotnua M. A. Bate ... 2.02... -2 eee e eee eee ee olo November 28, 1905. Mr. J. T. Cunningham, M.A., F.2Z.S. Exhibition of photographs of, and rematle upon, a° horse bearing horn-like structures .. 1... 2: ee esse ees erect eeees Prana hos duis 323 Mr. Frank Slade, F.Z.8. Exhibition of photographs of a Sea-Anemone in the process of . PLAT STO Tbe ey oaks LARS ok Gatitne ra Toduyie iotevocelearel aib Ua eiale (tekele Cleeecs ol tye lelescjele: ofeicas ene aetna » 894 Mr. Douglas English. Exhibition of an albino Field-Vole ...--...s+..e+-s sees eeeeee 324 Mr. G. A. Boulenger, F.R.S. Exhibition of, and remarks upon, a melanistic specimen of the Wall-Lizard........-- Oa HE hehe beet eee eee rere cece reser ene eeeeceee 324 Capt. Albert Pam, F.ZS. Remarks upon a living specimen of the Violet-cheeked Humming-bird 2.0.2.5. cece ce cece ee eee te ect e cet e ene sid areteenteeer a) ‘signe aoe Mr. W. RB. Ogilvie-Grant, F.Z.8. Exhibition of a series of bird-skins from ee . 824 1. Colour Evolution in Guereza Monkeys. By R. Lypekker ..........++..--.. kate atte - 220 2 The White-maned Serow.. By R. Lyprxxer. (Plate VELL) its, eam “829 3. The Duke of Bedford’s Zoological Exploration in Eastern Asia.—I. List of Mammo obtained by Mr. M. P. Anderson in Japan. By Ouprizup Tuomas, F.R.S. (Plate IX.) 331 4. A Revision of the Fishes of the sn! Galaxtide. By ©. TatE Rzean, B. A., E.Z.S. (Plates X.~XTIL.) 2... 000. cee ee ee reece eee ee eee eee ee ee erect eee ee cc aa ee 363 © 5. The Mammalian Fauna of China.—Part I. Murine. By J. Lewis Bonnorn, M.A.,F.L.8. 384 6. Descriptions of new Species of Phytophagous Coleoptera of the Genera Homopheta, Asphera, and Oedionychis. By Martin Jacosy, F.E.S. (Plates XIV. & XV.) .. 398, 591 . Some Additions to the Knowledge of the Anatomy, principally of the Vascular System, of Hatteria, Crocodilus, and certain Jacertilia. By Frank a Bepparp, M.A,, F.R.S., Prosector to the Society -..-..-++-+sscsens POO OD OO oa Samia ene ol We ceauE SG 461 | Contents continued on page 3 of Wrapper. S 1905. | THE SECRETARY ON ADDITIONS TO THER MENAGERIE. 295 November 14, 1905. G. A. BouLenGceEr, Esq., F.R.S., Vice-President, im the Chair. The Secretary read the following reports on the additions that had been made to the Society’s Menagerie during the months of June, July, August, September, and Ogistban. 1905 :— The number of registered additions to the Society’s Menagerie during the month of June was 321. Of these 57 were acquired by presentation, 53 by birth, 14 by purchase, 195 were received on deposit and 2 in exchange. The number of departures during the same period, by death and removals, was 178. Among the additions special attention may be called to :— 1. An Orang-utan (Sima satyrus) from Deli, Sumatra, presented by Dr. J. C. Graham on June 25th. 2. A Wolt’s Monkey (Cercopithecus wwolfi) from the Congo, deposited on June 26th. 3. Eleven Kiwis (Apterya mantelli) from the North Island, New Zealand, seven being presented by the Government of New Zealand and Mr. H. C. Wilkie, F.Z.S8., on June 26th, and four by the Earl of Ranfurly, H.M.Z.S., on the same date. The Society is specially indebted to Mr. H. C. Wilkie, in whose care these Kiwis were successfully brought from New Zealand. The number of registered additions to the Society's Menagerie during the month of July was 274. Of these 96 were acquired by presentation and 17 by purchase, 92 were received on deposit, 1 by exchange, and 68 were bred in the Menagerie. The number of departures during the same period, by death and removals, was 184. nee the additions special attention may be called to :— nN enna Leopard (Felis purdus) from near Hong Kong, pr ee, by My. J. A. Bullin on July 26th. 2. Three Californian Sea-Lions (Otari wa gulespu) from Santa Barbara, purchased on July 11th. 3. A White-tailed Gnu (Connochetes gw) born in the Menagerie on July 25th. 4, A male Somali Ostrich (Struthio molybdophanes) from Se aaa purchased on July 14th. A Collection of Birds from British Guiana, including examples of ie species new to the Collection, presented by Mr. K. W. Harper, F.Z.8., on July 31st. The number of registered additions to the Society's Menagerie during the month of ‘Aug ust was 348. Of these 108 were acquired by presentation and 37 by purchase, 66 were born in the Gardens, 103 were received on deposit and 34 by exchange. The number of departures durmg the same period, by sleet and removals, was 255. Proc. Zoou. Soc.—1905, Vou. Il. No. XX. 21 296 A WHITE WATERBUCK AND HEADS OF RHINOCEROS. [ Noy. 14, Among the additions special attention may be called to :— 1, ASilky Marmoset (//apale chrysoleucos) from Brazil, deposited on Aug. 30th. 2. A pair of West-African Marsh-Bucks (Limnotragus gratus) from the Congo, purchased on Aug. 31st. 3. A Spot-billed Toucanet (Selenidera maculirostris) from Santos, purchased on Aug. 19th. 4. A Black-and-White Cobra (Vata melanolewca) from West Africa, deposited on Aug. 31st. The number of registered additions to the Society’s Menagerie during the month of September was 313. Of these 106 were acquired by presentation, 26 by purchase, 124 were received on deposit, 15 by exchange, and 42 were bred in the Gardens. The number of departures during the same period, by death and removals, was 245, Among the additions special attention may be called to :— 1. A male Orang-utan (Simia satyrus) from Sumatra, presented by Mr. H. N. Ridley on Sept. 7th. 2. Five Talapoin Guenons (Cercopithecus talapoin) trom Ubanghi, Upper Congo, deposited on Sept. 19th & 21st. 3. A Jaguarondi (felis jaguarondt) from 8. America, purchased on Sept. 11th. 4, A Binturong (Aretictis binturong) from Singapore, presented by Mr. H. N. Ridley on Sept. 7th. 5. A White-tailed Sea-Hagle (/aliaétus albicilia) from the Arctic, presented by the Duke of Orleans, F.Z.8., on Sept. 30th. 6. A Knob-nosed Lizard (Lyriocephalus scutatus) from Ceylon, presented by Mi. E. Ernest Green, F.E.S., on Sept. 26th. The registered additions to the Society's Menagerie during the month of October were 232 in number. Of these 74 were acquired by presentation and 79 by purchase, 3 were born in the Gardens, 53 were received on deposit and 23 in exchange. The total number of departures during the same period, by death and removals, was 216. Among the additions special attention may be called to :— 1. A White Oryx (Oryx leucoryx) frem Arabia, presented by Col. R. J. Seallon, C.B., D.S.O., on Oct. 30th. 2. A Cave-Rat (ZThryonomys swinderianus), a Bouvier’s Owl (Scotopelia bowvieri) (new to the Collection), and a Beautiful Wood- Hawk (Dryotriorchis spectabilis), from Lagos, presented by Dr. W. F. Macfarlane, F.Z.S., on Oct. 13th. 3. Three Antillean Boas (Boa diviniloqua) from St. Lucia, West Indies, presented by the Hon. KE. G. Bennett, K.C., on Oct. 28th. Col. W. H. Broun exhibited a mounted head and skin of a White Waterbuck (Aobus ellipsiprymnus) and two mounted heads 1905. | DR. WALTER KIDD ON PAPILLARY RIDGES IN MAMMALS. 297 of Rhinoceros (/thinoceros bicornis), and made the following remarks :— “The White Waterbuck was shot in July 1904 on the right bank of the Guaso Nyiro river, about 20 miles west of the en ian Swamp, British Hast Africa, lat. 1° N., alt. above sea 1000 feet. A white doe was alone with the buck. The ordinary Waterbuck seen there were all examples of Kobus ellipsiprymnus. The eyes of this buck were of the normal colour, not pink. “Of the two Rhinoceroses, one was a female and earried two normal and two rudimentary horns. She was shot in August 1904, in dense covert, west of the Jambeni Mountains north-east of Mount Kenia, at an elevation of 4150 feet above the sea. It was not seen till after death what an imteresting animal she was. One of the rudimentary horns was between the ears and the other about 4 inches further back. “The other individual was a male, and was shot in September 1904 north of Aberdare range, British East Africa; height above sea 9600 feet. The anterior horn showed abnormal growth due either to an old injury or excessive wearing away of the outer surface from the tip downwards.” The Hon. Walter Rothschild, F.Z.S., exhibited specimens of a very rare and interesting Marsupial, hitherto unique, in the Paris Museum, viz. Dactylopsila palpator Miulne-Edw., which differed from D. trivirgata in possessing an extremely thin, prolonged, second finger. Mr. Rothschild also exhibited two tusks which had been obtained by Baron Maurice de Rothschild during his recent expedition to Abyssinia. ‘They were so unlike the normal tusks of any known animal, that Mi. Rothschild was of opinion that they might belong to some new form. Mr. A. S, Hirst, F.Z.8., exhibited microscopic preparations of anew Hemosporidian from the blood of an African Stork (Lepto- ptilus crumeniferus). He poimted out that this parasite belonged to the genus Halteridium, but differed from H. danilewskyi in its ereater size (Stade moyen 7—10,;), and also in its method of sporulation, in which the merozoites were more numerous, smaller, and arranged in a ball-like rounded mass. The name Halteridium crumenium was proposed for the new species. Dr. Walter Kidd, F.Z.8., read a paper, illustrated by lantern- slides, ‘On the Papillary Ridges in Mammals, chiefly Primates.” The arrangements of the ridges on the hand and foot of 24 species were shown and described, and their functions discussed. Arguments were brought forward to show that their primary function was to increase the delicacy of the sense of touch. 21% 298 MR. H. SCHERREN ON OLD | Nov. 14, Dr. P. L. Sclater, F.R.S., read a letter addressed to him by Mr. Wilham Rodier, dated Tambua Station, Cobar, New South Wales, June 29th, 1905, in which it was stated that Mr. Rodier’s plan for combating the Rabbit-pest (‘ Nature,’ March 21st, 1889) was still proving a ‘wonderful success.” As there had been some good rains in the district the feed at Tambua was “ splendid, right up to the boundary netting-fence,” but on the other side there was “absolute starvation,” owing to the great numbers of rabbits. This summer, in Mr. Rodier’s opinion, would see the surrounding district quite ‘“‘eaten out” by the rabbits, which were there in millions, but were easily kept down at Tambua by his plan. Mr. Rodier’s plan, which was very simple, might be shortly described as follows :—Ferrets and nets are employed to catch the vabbits alive in the usual way, but while all the females captured are destroyed, the males are turned out uninjured. The results are that the male rabbits, so soon as they begin to predominate in numbers, persecute the females with their attentions and prevent them from breeding. They also kill the young rabbits, and, as Mr. Rodier declares, “ worry the remaining does to death.” My. Henry Scherven, F.Z.8., exhibited two lantern-slides of old pictures of Anthropoid Apes, and made the following remarks on the Satyrus indicus of Tulpius (text-fig. 50) :— There appeared to be in Dapper (‘ Beschreibung von Afrika,’ Amsterdam, 1670, p. 393) an early reference to a Gorilla. No figure was given, but the description, though of course inexact, seemed to fit the Gorilla better than the Chimpanzee, especially with regard to erect progression *, the folk-story of carrying off and ravishing women, and the supposed human origin. The passage 1s as follows :— “Hier [Quoja, north of Fernando Po] wird auch ein Tier gefunden, welches die Einwohner Quojas-Morrow oder Worow und die Portugallier Salvage dass ist ein Waldmann nennen. Es hat emen grossen Kopf, dicken Leib, fleischichte Arme, damit es in Ringen sehr starck ist, aber gantz keen Schwantz; und gehet zuweilen mit ausgerecktem gerade Leibe auf den Hinter- fiissen allen wie die Affen zuweilen auf allen vieren langst der Erde. Die Schwartzen sagen dass es von Menschen entsprossen ; aber durch das wilde Leben im Busche zum halben unverniinft- igen Tiere sei worden. Diese Tiere leben vom wilde Honige und die Friichten in den Biischen : auch fechten sie fort und fort nuit emander. Ja, sie diirfen nicht allein die Frauen ergreifen und nohtziichtigen, sondern auch die gewafnete Minner selbsten anfallen.” In December 1904 the Hon. Walter Rothschild, M.P., laid hefore the Society a valuable paper, entitled ‘ Notes on Anthiyo- poid Apes”, and exhibited what was undoubtedly the finest * R. I. Pocock, P. Z.S. 1905, vol. ii. p. 178. 7 P. ZS. 1904, vol. i. pp. 413-440. 1905. ] PICTURES OF ANTHROPOID APES. 299 collection of mounted specimens, skeletons, and skulls ever brought together. In his paper Mr. Rothschild spoke of the Satyrus indicus Tulp., and identified it with Simia satyrus Linn., claiming that the latter name must now be applied to a Chim- panzee—to quote his exact words, “the famous ‘Tschego’ proves to be the veritable Simia satyrus.” Text-fig. 50. The Satyrus indicus of Tulpius. Mr. Scherren then quoted the following description by Tulpius 300 MR. H. SCHERREN ON OLD [| Nov. 14, of the anthropoid presented at the end of the seventeenth century .to Prince Frederick Henry of Orange :— “ Quamvis extra forum anecHiean, attexam tamen huic tele Satyrum Indicum; nostra memoria ex Angola delatum: et Frederico Henrico Arausionensium Principi, dono datum. Erat autem hic Satyrus quadrupes: sed ab humana specie, quam prez se fert, vocatur Indis orang-outang: sive homo sylvestris, uti Africanis quoias morrou. Exprimens longitudine puerum trimum, ut crassitie sexennem. ‘““Corpore erat nec obeso nec gracili, sed quadrato habilissimo tamen, ac pernicissimo. Artubus verd tam strictis et musculisaded vastis: ut quidvis & auderet et posset. Anterius undique glaber : at pone hirsutus, ac nigris crinibus obsitus. Facies mentiebatur hominem: sed nares sime, & adunce, rugosam, et edentulam anum. “ Aures vero nihil discrepare, ab humana forma. Uti neque pectus; ornatum utrimque mamma. pretumida (erat enim sexus foamini) venter habebat umbilicum profundiorem ; et artus, cum superiores, tum inferiores tam exactam cum homine similitudi- nem, ut vix ovum ovo videris similius. “Nec cubito defuit requisita commissura: nec manibus digi- torum ordo: neque pollici figura. humana: vel cruribus sure, v vel pedi ealcis fulerum. Que concinna, ac decens membrorum fomine in caussa fuit, quod multoties incederet erectus : neque attolleret minus gravate quam transferret facile, qualecunque gravissimi oneris pondus. “ Bibiturus prehendebat canthari ansam, manu altera ; alteram vero vasis fundo supponens, abstergebat deide madorem labiis relictum, non minus adposite ac si delicatissimum vidisses aulicum. Quam eandem dexteritatem observabat utique eubitum iturus. Tbaclitieame quippe caput in pulvinar, & corpus stragulis convenienter operiens, velabat se haud aliter, ac si vel mollissimus illic decubuisset homo. “ Quin imo narravit aliquando affini nostro, Samueli Blomartio, Rex Sambacensis, Satyros hosce, preesertim mares, in Insula Borneo, tantam habere animi confidentiam, & tam validam musculorum compagem: ut non semel impetum fecerint, in viros armatos, nedum in imbellem, foeminarem, puellarumve sexum. “Quarum interdum tam ardenti flagrant desiderio : ut raptas non semel constuprarint. Summeé quippe in Venerem sunt pro- clives (quod ipsis, cum libidinosis veterum Satyris commune) immo interdum adeo pr otervi ac salaces: ut mulieres Indice propterea vitent cane pejus et angue saltus ac lustra, in quibus delitescunt impucica heee animalia.” Mr. Scherren called attention to the discrepancy between the Linnean diagnosis of Simia satyrus and the description of Tulpius with respect to form and size. Linneus wrote: “ Magnitudine puerl sexennis,” which differed widely from the words of Tulpius : “ Exprimens longitudine puerum trimum, ut crassitie sexennem.” It was also suggested that the expression ‘“‘ corpore quadrato ” suited a Gorilla rather than a Chimpanzee, and confirmation was 1905. } PICTURES OF ANTHROPOID APES. 301 sought in the pictures by Wolf of Wombwell’s Gorilla (from a daguerreotype) and a Chimpanzee from life which hung in the meeting-room. Sir Harry Johnston * had seen a reproduction of Tulpius’s figure in Tyson’s work on the Chimpanzee (London, Text-fig. 51. TM TTS SS SSSESS= Karly figure of Chimpanzee, from Astley’s ‘ Travels.’ 1699), and was struck by its resemblance to a Gorilla. Another picture (text-fig. 51) (with label, of which the following is a translation :—“ Chimpanzee, 21 months old, brought from Angola, * P. Z.S. 1905, vol. i. p. 72. 302 MR. J. L, BONHOTE ON MAMMALS FROM TIBET, [| Noy. 14, in 1738, 2 ft. 4 in. high”) was thrown on the screen to prove that as early. as 1746 the Satyr us mdicus was recognised as differing from the Chimpanzee. This picture was said £0 have been taken from life. It was usually said that the existence of an African anthropoid other than the Chimpanzee was not known till about the middle of the last century. This was not the opinion of J. E. Gray ; for ata scientific meeting of this Society *, in calling attention to Wombwell’s Gorilla, he alluded, but without quotation, to Bowdich’s ‘ Mission to Ashantee’” (London, 1819), where the “ African Orang (Pithecus Troglodites)” was compared with the Ingena. The following papers were read :— On a Collection of Mammals brought home by the Tibet Frontier Commission. By J. Lewis Bonnorr, M.A., abuses I /AgSi [ Received August 9, 1905. | (Text-figures 52 & 53.) The collection: of mammals brought home by the Tibet Mission, and collected by Capt. H. J. Walton of the Indian Medical Service, although not large in numbers contains several specimens of great interest, and there can be no doubt but that the region is full of mammalian treasures, only waiting time and opportunity for their discovery. Of the eight species of which examples were brought back, two, Microtus waltoni and Cricetulus lama, ave new to science, while the large red Fox of the country is sufficiently distinct to be entitled to subspecifie rank. In addition to these, I have been enabled for the first time to examine the skull of another Fox, Vulpes ferrilatus, described 63 years ago, but of which the skull- characters have hitherto remained entirely unknown, This skull shows features of great peculiarity, and proves the validity of Jerrilatus as a species, a matter hitherto considered doubtful by some writers. In addition to the specimens collected by Capt. Walton, the British Museum is indebted to Col. Waddell for two or three skins, an account of which has also been incorporated in this paper. FELIs MANUL Pall. Felis manul, Pall. Reise Russ. Reichs, ii p. 692 (1776); Blanf. Faun. Br. Ind., Mamm. p. 83 (1891). * P.Z.S. 1861, p. 278. + [The complete account of the new forms described in this communication appears here; but since the names and preliminary diagnoses were published in the ‘ Abstract,’ the former are distinguished by being underlined.—Hp1Tor. | TA sketch- -map giving all the localities j in which this collection was procured is published j in ‘The Ibis’ (1905, p. 57, pl. u.). 1905. | MR, J. L, BONHOVTE ON MAMMALS FROM TIBET. 303 a. 6. Yamdok Lake, alt. 15,000’, 28th Sept., 1904. This specimen, the only one procured, was brought home by Col. Waddell. VV ULPES VULPES WADDELLI. Vulpes vulpes waddelli Bonhote, Abstr. P.Z.S. No. 22, p. 14, Novy. 21, 1905. a. Khamba Jong, alt. 16,400’, 8th Oct., 1903 6b. Phari Jong, Upper Chumhi Valley, 11th Jam., 1904 (coll. Waddell). General colour above reddish fulvous, the median dorsal area from the occiput to the root of the tail being bright red, shading to pale buff on the flanks and hindquarters. The head rufous : the ears moderately large and pointed, bemg clothed with long white hairs on the inside and short black ones externally. Feet rufous along their margins and white or grey in the centre. Tail long, woolly, and very bushy, tipped w ith white; each hair being pale fulvous at its base, ath a long, bing terminal portion. Underparts pure white. The skill does not show any special characters by which it may be distinguished from that of the typical form. It is stout and well built, being short and broad in the muzzle and rather swollen in front of the orbits, but otherwise it shows no features of note. Dimensions of type (in flesh). Head and body 25 in. ; tail 16 in.; hind foot 6in. Height at shoulder 14°75 in. Weight 8 lbs. Skull. Greatest length 145 mm.; zygomatic breadth 72; width in line with ant. root of pm. 4, 37-5. (Further skull-dimensions are given under the next species.) Habitat. Khamba Jong, Tibet, alt. 16,400". Type. B.M. 5.4.6.1. Collected 8th October, 1903, by Capt. Walton. This race may be readily distinguished from V. v. flavescens by its much brighter coloration throughout, and especially by the deep red median dorsal area. In the true flavescens the back is much more uniform in colour, the median dorsal area being but very slightly darker than the surrounding parts and of a more brownish yellow, the red tint being entirely lacking. The local name is ‘“* Wamo.” VULPES FERRILATUS (Hodgs.). Vulpes ferrilatus Hodgs. J. A.S. B. xi. p. 278, pl.; Blanford, Fauna Br. Ind., Mamm. p. 155 (1891); Mivart, Mon. Can. p: 121 (1890). a. Karo-La Pass, alt. 16,600’, 30 miles E. of Gyangtse. The only specimen procured is a typical example of V. ferrilatus, but in very bad fur. This species may always be recognised by the underfur, which, besides being close and woolly, is fulvous to the base. The peculiar and woolly character of the fur through- out is quite sufficient to distinguish it from all other species. 304 MR. J. L, BONHOTE ON MAMMALS FROM 11BET. [| Nov. 14, Capt. Walton writes :—‘‘The small fox does not, I feel pretty sure, occur near Khamba Jong, all the foxes seen there being of the large species (V. v. waddelli, ante). I saw other foxes almost certainly of this species between Karo-La and the neighbourhood of the Yam Dok Cho (Lake Palti).” Accompanying the skin is a very fine adult skull (text-fig. 52), which, so far as I am aware, has never before been described. This skull is quite unlike that of any other species of Fox, and is characterised by the extreme slenderness and elongation of the muzzle and the great length of the upper canines. The brain-case Text-fig. 52. Rows é coher AUS (he orld ae AGS A, lateral, and B, upper view of the skull of Vulpes ferrilatus. % nat. size. and zygomata, on the other hand, do not show any signs of lateral compression or elongation, but are fairly normal in their dimen- sions and breadth. The supraorbital processes are stout and well developed, and the brain-case gradually widens out from imme- diately behind these processes and reaches its greatest breadth in line with the posterior roots of the zygomata. On the under side we may note the narrowness of the soft palate and the tendency of the pterygoids to approach each other posteriorly. The bull are more elongated and less rounded than usual. The dentition, which 1905. | MR. J. L. BONHOTE ON MAMMALS FROM TIBET. 305 is normal, except for the great length of the canines already noted, alls for but little comment. The spaces between the pr eniolars are large in correlation with the length of the muzzle, and the first upper molar is relatively small. Dimensions :— V. fervilatus. V.v. waddelli. mm. mim. Cueasest lemel teco-ccs fe sede c cers 155 145 Basalblen cular yan tas. 2 ate eee eee eet 138 130 JeeN eral) lieimead Bes Hoe eeanspacorsnddsovoec 78 al Length from post. end of palate to bastocenpitall Were ee eencats.s-eaueecseee 60 59 Length from last incisor to ant. root of Ist premolar (alveoli) ............ iS) 16 Length of premolar series ............ 48 39 Br eadth of brain-case immediately behind supraorbitals.................. 26 23 Greatest breadth of brain-case......... 5O°D AO Ax comeasicy Wr endl eye ses-ceeeen tes 84 12 Breadth of muzzle at ant. root of ai NOW OVRELTOO) EW, | NAGAR eek ABA Recs de saene 19 23 Length of upper canine along its AVMTeTNOIe TaN ACAD snanedabAnasounees 29 20 This comparison of dimensions will show more clearly than any description the main features in which this skull differs from that of the more typical “ Volpes”; and in spite of the doubt of Mivart * there can be no question that ferrilatus not only Is a good species, but is more differentiated than any other species in the region. The only other skull of ferrilatus known is a very young one collected by Mi. Hodgson ; it is, however, too young to show any of the specific characters enumerated above. Purortrus ALPINUS (Gebler). Mustela alpina Gebler, Mém. Soc. Lnp. Nat. Moscou, vi. p. 213 (1823). Putorius alpinus Blantord, Faun. Br. Ind., Mamm. p. 168 (1891), a. 6 ad. (in spirit). Gyangtse, alt. 12,900’, 1904. 6. 3. Khamba Jong, alt. 15,500, 11th Sept., 1904. CRICETULUS LAMA. Cricetulus lama Bonhote, Abstr. P. Z.8. No. 22, p. 14, Nov. 21, 1905. a. gad. Lhasa (skinned from spirit). 6. g ad. Lhasa (in spirit). The Cricetulus referred tor as ‘the little white mouse” is * Loe. eit. ante. + Percival Landon: Tibet,’ App. by H. J. Walton (1905). 306 MR. J. L. BONHOTE ON MAMMALS FROM TIBET. | Nov. 14, represented in the collection by two spirit-specimens, one of which has since been skinned. It appears, although closely related to Cricetulus pheeus, to have been hitherto undescribed. Size about that of C. pheus. General colour above pale fulvous erey, greyer than in C. pheus. Kach hair is slate-grey at its base, fulvous for about 3 of its distal end and with a black tip. Over the head and fore part of the body the fulvous portion of each hair is the more conspicuous, but on the hinder part of the back the dark tips predominate and a faint dark median dorsal line may be traced. The underparts are pure white, the hairs being slate-grey at their base. The line of demarcation between the upper and under parts, although abrupt, is very uneven in outline. The feet are but scantily clothed with hair and are white. The tail is moderately long and stout, well clothed with dark brown hairs above and white haus below ; the tip is white. The whiskers are for the most part black with a white tip, some shorter ones, however, being entirely white. The skull resembles somewhat “closely that of O. pheus, but is slightly larger and the brain-case more inflated and rounder. The chief points of difference, when viewed from below, are the ereater width of the basioceipital and the much flatter and smaller bullae in the new species. Above there is a slight, although very constant, difference in the hinder margin of the parietals, which are practically straight in outline; whereas in C. pheews there is a sharp turn backwards when about two-thirds of their length from the middle line. Dimensions (of type when in spirit). Head and body 87 mm. ; tail 40; hind foot 17; ear 16. Skull. Greatest length 28°5 mm.; basal length 24; palatal leneth from henselion 12; interorbital breadth 5; greatest breadth of brain-case 12°5; width of basioccipital at anterior end of auditory bull 3. Habitat. Lhasa, Tibet. Type. B.M. 5.4.6.4. Collected at Lhasa, Tibet, by Capt. H. J. Walton, I.M.S. The darker colour of the hinder part of the back combined with the general much greyer coloration, and in addition the somewhat longer and stouter tail, form characters by which this species may be distinguished from C. pheus. The animal, ac- cording to Capt. Walton, was extremely common, and was swarming in one of the shrines of the Jo Khang Cathedral at Lhasa. Microtus (PHAIOMYS) WALTON. Microtus (Phaiomys) waltoni Bonhote, Abstr. P.Z.S. No. 22, p. 14, Nov. 21, 1905. a. 2 ad. Lhasa, Tibet. Slightly smaller in size than Ph, blythi, to which it is by skull- 1905. | MR. J. L. BONHOTE ON MAMMALS FROM TIBET. 307 characters closely allied, though widely differmg in colour. General appearance above fulvous grey, slightly greyer over the anterior part of the body; below very pale buff. Each hair is slate-grey from its base and for the greater part of its length, with a fulvous subterminal portion and dark tip. On either side, between the limbs, the dark tips are absent, leaving a clear fulvous patch. Interspersed in the fur are a few thin black bristles. The feet ave whitish, both palms and soles are 5-tuberculate. The tail is bicolor. The ears small and sparsely covered with hairs similar in colour to those on the rest of the body. Mamie eight in number, four pectoral and four inguinal. Skull. The dental characters are practically identical with those of Ph, blythi. The skull itself is very similar in general outline, but slightly smaller; this is especially noticeable in the bulle, Text-fig. 53. A, upper, and B, lower right molar series of Microtus waltoni. which do not stand out so prominently from the rest of the cranium. Dimensions of type (fvom spirit), Head and body 98 imm. ; tail 30; hind foot 17; ear 10:5. Skull. Greatest length 28 mm.; basilar length 24:5; zygomatic breadth 16°5; interorbital breadth 4; length of nasals 7; dia- stema 8°2; palatal length 15; length of molar series (text-fig. 53) (alveolt) 7. Habitat. Lhasa, Tibet. Type. B.M. 5.4.6.5. 2 ad. Collected by Capt. H. J. Walton, 1. M.S. This interesting species cannot well be confused with any other, 308 MR. J. L. BONHOTE ON MAMMALS FROM TIBET. | Noy. 14, as the skull-characters clearly prove it to belong to the subgenus Phaiomys, and its colour is quite unlike any of the other species of that group. Lepus orostotus Hodgs. Lepus otostolus Hodgson, J. A.8. B. ix. p. 1186 (1840) ; Blan- ford, Faun. Br. Ind., Mamm. p- 452 (1891). a. WKhamba Jong, Oct. 1903. Apparently the common Hare of Tibet, two more skins having been brought home by Col. Waddell. Capt. Walton states that this hare, which, as a rule, avoids cultivated land and frequents bare and rocky hillsides, was very common at Khamba Jong and also at Tuna at the hand of the Chumbi Valley. It was, however, much scarcer, although still occurring, between Gyangtse and Lhasa. OcHOTONA CURZONLE (Hodgs.). Lagomys curzonie Hodgs. (nec Stoliczka) J.A.S. B. xxvi. p- 207 (1858) ; Blanford, Faun. Br. Ind., Mamm. p. ABT (1891); es P.Z.S. 1904, vol. 1. p. 214. a 2. Yamdok Lake, 14,800’, 29th Sept., 1904. b,c. No particulars d. Ad. imspuit. Tuna. The specimen from the Yamdok Lake had evidently just assumed its new winter pelage, the other two skins being in old and worn fur. Tuna, where the spirit-specimen was pr ocured, is only a few miles north of the Chumbi Valley, the type locality of the species. Capt. Walton writes of this species :—“ They are exceedingly common at Khamba Jong, Tuna, and in all the open bare country from Tuna to Gyangtse, as well as between Gyangtse and Lhasa. They, however, become less common as one approaches Lhasa, probably because the country is more cultivated. They ayer cultivated fields for the most part, and were always commonest in bare sandy country. They do not hibernate at ‘all, and on any sunny day i in the middle of winter they might be seen sunning themselves at the entrance to their burrows. I dug up a few burrows during the winter. The tunnel runs more or less vertically downwards for 1 or 2 feet and then somewhat hori- zontally for 4 to 6 feet. The passage is dilated at irregular intervals in some two or three places. At these spots and at the end of the burrow, which is also dilated, there is a certain amount of coarse grass collected to form a kind of nest. The ground in many places i is honeycombed with these burrows, which sometimes communicate with one another close to the OUthe but as a rule they are quite distinct. I never heard the animal utter a ery of any sort.” 1905.| ON THE GEOGRAPHICAL DISTRIBUTION OF THE OKAPI. 309 2. Notes on the Geographical Distribution of the Okapi. By Dr. E1yar Loénnpere, C.M.Z.8. [Received August 28, 1905. | I have had the pleasure recently of meeting my compatriot Lieutenant Karl Eriksson, who delivered to Sir Harry Johnston the first skull and skin of the Okapi. I took this opportunity of asking Lieutenant Eriksson about the distribution of this remark- able animal, and his reply was that he believed it to be distributed practically over the whole of the “ equatorial forest” of the Congo Free State. He showed me on the map of the Congo basin appended to Mr. Boulenger’s work, ‘ Les Poissons du Bassin du Congo’ *, the approximate limits of this area of distribution. Tf we begin at the River Ubangi in the west about midway between Mobena and Jmese, from there the limit extends north- east towards Businga at the River Likame or somewhat north of that place, and then more east to the River Uele just before it joins the River Ubangi. From that place and eastward the River Uele is the northern limit to a point about midway between Amadi and Suruaugo. From there it turns south-east, passing somewhat east of Mawambi, and continuing to a point a little west of Karimi. Not much south of this, the most eastern point of the great forest, the boundary-line turns westward again and crosses the great Congo River at Ponthierville, and continues westward a little south of Tschuapa River, but bends by-and-by a little north, so that it passes on the northern side of Bolondo towards Coquilhatville. It is evident that this is only a rough outline of the area of distribution of the Okapi, but it may hold good in a general way. Outside this boundary-line there are many forest-clad areas, but they are not extensive, and Lieutenant Eriksson does not believe that they are inhabited by the Okapi. It is an in- habitant of the great forest, but does not live everywhere in it. Tts regular pasture-grounds are open glades in the forest, where rivulets with shallow water expand and produce a rich growth of grass. This grass and the leaves of the bushes and undergrowth under the trees, which are especially luxuriant in such places, may form the principal food of the Okapi. Although a shallow sheet of water expands over the very flat ground to greater or less extent in these glades, there are noswamps. ‘The soil is hard and firm 7, which explains the shape of the hoofs of the animal. Lieutenant Eriksson has not seen the Okapi in a living state in its natural surroundings, as probably no white man ever has or is likely to do. But he has, while on his marches during the night, * Bruxelles, 1901. + Because gravel or hard red earth lies quite near the surface. 310 ON THE GEOGRAPHICAL DISTRIBUTION OF THE OKAPT. [ Nov. 14, many a time heard it run away when he passed such glades as described above. The Okapi is extremely wary and shy, and nocturnal in its habits, It lives singly or perhaps in pairs, never in herds. The negroes know very little about it, and, as a rule, it is only the Wambutti- dwarfs who ave able to kill it. These dwarfs are perhaps the most perfect of all hunting tribes and steal up near the animals, slaying them with spears. How little the negroes (not counting the dwarfs) know about the Okapi, may be concluded from the following ridiculous tale told and believed by them. They have observed that the Okapi is very cleanly, and even during the rainy season, when almost all other animals are more or less dirty, its skin is justas clean as ever. The negroes say then that the Okapi climbs up in the trees (!) to keep itself clean and to avoid the dirty muddy soil. I have used the name Okapi as that is the one known to the zoological world, and has become the nomen triviale of this interesting mammal. Lieutenant Eriksson informs me, however, that it was only a mere chance that it happened so. Okapi (with long-drawn a) is only used by the Wambobba tribe for signifying this animal. ‘The Wambobba language is hardly spoken by more than 300 persons, but it was Wambutti- dwarts, living in harmony with Wambobbas and speaking their language * , that brought the first remains of the Okapi, hence the name. (The first complete specimens were procured by another tribe of Wambuttis belonging to Wabira negroes, which use another name mentioned below.) But it is still worse, because the word ‘ Okapi” means simply in the Wambobba language cs donkey ” or “ass.” Strictly speaking, therefore, the latinised “ Okapia,” which became the second and permanent scientific generic name of this mammal, is not much better with regard to its original meaning than the first generic term “* Equus,” applied before anything but a piece of skin was known. The name by which the Okapi is known in most of the Congo languages 1s ‘* Dumba.” Tam glad to be able to add that the Okapi is protected by law, -so that it is forbidden to kill it without special permission. The Wambutti-dwarfs and the leopards do not, however, respect any laws, and therein lies the danger for the existence of this animal. Lieutenant Eriksson has also told me that in the great forest a kind of black wild hog is to be found, which may be the recently deseribed Hylocherus meinertahagent. These hogs are called by the negroes ‘ n’gulube bibi,” which means “ black hog,” whilst “youlube” = hog is the name of the common Red River-Hog (Potamochwrus porcus). * The Wambuttis always use the language of those negro tribes with which they live in symbiosis, and from which they obtain vegetables. for meat and honey from the forest. 1905. | ON THE GORAL FOUND IN BURMA, 311 3. Notes on the Goral found in Burma. By Major G. H. Evans *. [Received September 2, 19085. | The Himalayan range in Assam gives off a succession of spurs southward to form a tract of mountainous and, in many parts, almost impassable country extending into Arakan and Burma, and inhabited by numerous wild tribes. That portion of this tract lying between Assam and Manipur to the north, Chittagong and Tipperah on the west, Arakan on the south, and Burma on the east, is now known as the Chin-Lushai Hills. These so-called hills vary in their altitude from 1000 to 10,000 feet. T was employed in what was known as the Southern Chin Hills from November till June 1889-90, and during my stay visited several Chin villages. Like many others who have visited these people, I came to the conclusion that Chins generally, and their chiefs in particular, have one hobby at least, viz., collecting skulls. Outside and inside the villages, skulls were to be seen stuck on posts or kept in the houses. The finest collection I met with was in the house of a Boungshé chief, whose tribe is thus called by the Burmans, from the method in which they dress their long hair. The whole hair is done up ina large knot placed well forward on the top of the head, almost on the forehead, and round this ball of hair is wound, round and round, usually a white turban with a blue stripe through the centre. In the chief’s house was a collection of skulls, excellent as regards the number and variety. The heads ranged from those of elephants to palm-civets, and I doubt if there are many museums which could excel the collection of monkey skulls, at least numerically. The chief enjoyed the reputation of having been a mighty Nimrod in his youth, and I was informed that he had shot practically every head in the collection. I noticed one splendid gaur skull, three or four fine mythun or gayal, several sambar and serow, also some small heads which I concluded must be goral. Game throughout the hills was scarce, a matter not to be wondered at, inasmuch as every Chin had a gun of some sort, and in addition was always trapping and snaring. I was assured that the Goral heads had been obtained in the hills, but that now the animals were very scarce. I had no opportunity of verifying at this time the presence of Goral in these hills, and any attempt to do so would have been a matter of considerable risk owing to the most unfriendly attitude of the people. Many months later I happened to be in a Burmese village some hundred miles distant, but on the confines of the South Chin Hills, and there discovered in a house the skull of a Goral identical with those above mentioned. On enquiry from the Burmans I learned that it had been obtained from some Chinbéks, another tribe of Chins near Loungshé in the Yaw country. As * Communicated by R. LypEKKER, F.Z.S. Proc. Zoou. Soc.—1905, Vou. II. No. XXII. bo bo 312 MAJOR G. H. EVANS ON THE | Nov. 14, the Burmans dare not venture into Chin-land, they could afford no definite information beyond that the Chins had told them that there were several of these animals on a certain high mountain now known as Mount Victoria. Since then several Goral have been shot there by policemen on outpost and others. During the season of 1896-97 I visited the Arakan Hill-tracts, which are merely a southern continuation of the Chin Hills into the Akyab district of Arakan. Here again I came on askull and a skin (the latter in a very bad state of preservation) of this Goral. This animal, from the horns evidently a female, was shot in the hills at a place not very far distant, and local informants said that there were a fair number. Being unable to visit the place at that time, I told a friend of the ground, and asked him to find out if what I had heard was correct. He did so, and came across some six animals, of which he shot a couple. One of these, owing to the ground, it was impossible to recover. I sent a skull for identification, and was informed that it was a Himalayan Goral. I was unacquainted with the Indian Goral, but from the descriptions in books I was not quite satisfied that it was the same animal. Later on, while after Serow in the Shan range of hills to the east of the Inrawaddy, I was much surprised again to run across these animals. I was still more convinced that the beast was not the same as the Indian Goral, so much so, that I asked a friend to shoot an Indian Goral and send me a head and skin, which he very kindly did. On comparison my suspicions were confirmed. J was then most desirous to procure a specimen for the British Museum, but luck was against me, as it was a long time before I ran across them again. The following are the chief characteristics of these Goral :— General oie. —Goat-like with sturdy lmbs. Horns are present in both sexes: those of the female are shorter, thinner, and not so rough as those of the male. They are generally almost parallel, 7. e. only shghtly divergent, and have a slightly backward eurve. The coat is moderately long, close, and the hair rather coarse ; there is generally a well-marked underfur. The mammee are four in number. General colowr—A dark, more or less rat-grey, with an admixture of longish, dark, rufous-tipped hairs running through the coat, but mostly on back and upper surface of body. In an old buck the back, haunches, and upper portions of sides were dark pepper-and- salt ox evizzled grey. Ina young specimen the colour was generally lighter. There is no distinct dorsal stripe : in a young animal a very faint but distinct brownish line was traceable, extending from the nape to the dock, and in the skin of a female also, when held in a good light, a darker brownish median line could be discerned. The colour fades gradually on the side to a dirty reddish white under the abdomen. The colour about the back of the neck isa lighter grey than that of the body, and the hair is longer. A distinct crest of longer hair of blackish-brown colour extends from between horns to behind the 1905. ] GORAL FOUND IN BURMA. 313 ears. The hair surrounding base of horns is also long and of a rufous tint. The face is ruddy brown, passing into grey on the cheeks, and to a fainter and almost whitish colour around the eyes and lips. The throat is a yellowish white. The hair on the outer surface of the ears is rufous; whitish on the inner surface. The mufile is black in colour and naked. The tail is black or brownish black, and has a tuft of long hair of varying length. The colour of the iris 1s reddish brown. The Limbs: Fore legs—Outer aspect a dark brown or yellowish ved to just above the knees, and this colour is continued on the posterior aspect of lower limb to hoofs. The anterior aspect from below the knees, or in some cases just above the knees, 1s a yellowish white. Hind legs.—Outer aspect of thigh brownish, the posterior aspect of the hocks dark brown, continuing down posterior aspect of lower limb. The anterior aspect below the hocks is a dirty white. Horns.—Short, black in colour, conical, irregularly ringed, especially at the base in males. The annular markings extend for about three quarters of the total length on the posterior aspect of the horn; they appear to be rubbed off in front. The horns are set close, and in some cases are almost parallel. For the first inch or so from the base they are straight, then curved slightly backwards, and are slightly divergent towards the tips. Measurements of Horns. iL, 2. 3. 4. 5. | i inches. | inches. | inches. | inches. | inches. Right horn ....... vl 5 | 4 ae BE i al | | | Left horn ........... v2 Az 4 | L 38 | 3 (broken) ee 23 Peal dae e> | ares 23 Between horn-cores | 4 3 E i Between points .........) 22 13 oF 2 Dimensions. 3. ©. inches. inches. eiehtrat; Shoulders sce. cheen oe naoa cau 2 tOnZ a ONtORA Girth behind shoulder ........................ 294 PATE teniethy tom) nese tortalllyeee tease eee DOs 50 Tail: average of five specimens............... 41 Tuft: average of five specimens ............ 23 Tcenethiol carsisy cette ee thee i eds 32 to 43 Wemetiot Meadye recta me tee cee ae 102 10 TByRSAGH Eli AGIROSS OIA OWES Goo accseccascenccdadceddous 4 37 to037 Jere OCH ANOTAOIS. Gan neooosdodacooes oe cuonsdaesoons 5 4a Girtheotelronnshcn cen aeerrcs fen cen wnaeanies 2. to.22 2 22% 314 ON THE GORAL FOUND IN BURMA. [Nov. 14, Distribution.—So far as is at present known to me, in the localities noted, and at elevations above 3500 feet. These Goral appear to be rather localised, and I should say are uncommon, It is reasonable to expect, however, that when a more intimate knowledge of the higher ranges is gained, the distribution of these animals may be found to be more extensive. These Goral, I believe, extend into Siam and are to be found in suitable places on the Siamese side of the Thaungyin River, and also occur, but are more scarce, about the hills at the headwaters of the Me-Ping. Habits—As has been recorded in the case of the Indian form, these Goral live in parties of four, six, or even a dozen. They inhabit very steep ground and the more precipitous it is the better they seem to like it. They are never to be found at any distance from rugged, rocky ground, even though there may be forest near by. The only time they may be found away from dangerous ground is during the early hours of the morning and late in the evening, when they g graze on the grassy patches close by. No doubt when the sky is ‘overcast, as is the case during the rains, or in the cold weather when there is a heavy mist, they feed much later. Apparently they are inclined to remain always about any tavourite locality. Their sight seems to be extraordinarily good, and they appear to rely more on this sense than on smell or hearing. The day is usually passed lying on inaccessible ledges of rock about precipices. If a Goral is startled it Jumps up and makes a short: sharp hissing or sneezing noise, very often repeated at short intervals. It may be a note ‘of alas or a call to its mates, for as Sure as one calls, if there are any others about (and this is generally the ease), 1t 1s immediately answer ed. In Burma, at least, these Goats are not easily followed, unless by expert cragsmen; and in this category I do not include myself. Goral, when standing about these crags, afford fairly easy shots with high-velocity rifles, but the recovery of a carcase is, as a rule, by no means an easy matter. The shikaris and followers are generally anything but keen on a trip down one of these precipices, and I for one do not blame them, Though they may be adepts in woodcraft, they cannot be anything like the cragsmen (hill-shikaris) met with in the Himalayas. Goral-flesh is not at all bad, From December till May is the best season to hunt these animals, and morning and evening is the best time to find them, as they are then grazing or lying down in places more accessible. I sent specimens of the skin &e. of this Goral to Mr. Lydekker, by whom the animal has been named after myself, Urotragus evanst. I have to thank Captains Blakeway and Wood, R.E., and Mr. W. B. Tydd, of the Burma Civil Service, for their kindness in helping me in this matter, Rangoon, Ist June, 1905. 1905. | ON THE MAMMALS OF CRETE. old 4, On the Mammals of Crete. By DororHEea M. A. BATE*. ( Received September 6, 1905. | The following list of the wild mammals known to inhabit Crete is based on a small collection made in the island during a stay of four and a half months in the earlier part of last year (1904) This includes only sixteen species, but it is quite possible that a species of Crocidwra may have to be added to the number, for remains of a Shrew were found in more than one Pleistocene cave-deposit in the western part of the island, and it is not unlikely that it may yet survive. It is probable that a Roedeer still existed in the island during the earliest historical times. Four species seem to be here recorded for the first time from this locality ; these are Rhinolophus ferrum-equinum, R. hipposideros, Micromys sylwaticus hayi, and Acomys dimudiatus minous. In his work on Crete ¢ published in 1869, M. V. Raulin gives a list of thirteen species, amongst which is included the Polecat as well as the Beech-Marten and Weasel; however, no specimens appear to have been obtained, so that their occurrence may have been admitted on insufficient evidence or as the result of some confusion with regard to the other members of the group. Admiral Spratt t, in describing the country between Kremopoli and Palaikastro, mentions that Foxes occur there; but this was doubtless a slip, for elsewhere (vol. ii. p. 157), in reference to the safety of the flocks of sheep, he says that ‘ Crete has no wild animals but badger and weasels or martens.” Dr, Lorenz- Liburnau has written at some length on the Wild Goat of Crete ; and in 1903 Major Barrett-Hamilton described the Hare, and noticed the Badger and Beech-Marten, at the same time remark- ing on the paleness of the specimens from this locality. The same may be said of the Cretan Hedgehog and Rabbit, but is not the case with the Weasel and Spiny Mouse, which are both richly coloured forms. Crete has, in all probability, been isolated as an island for a considerable period, therefore it is not surprising to find that there are a number of localised forms amongst the Mammalia. Admiral Spratt, whose valuable researches were carried on in so many parts of the Mediterranean, was of opinion § that Crete was connected in earlier times with Europe (including Asia Minor), and not with the north coast of Africa as tradition would have us suppose ||. Suess §[ would also seem to link this island rather with the northern than the southern boundaries of the Medi- terranean, The mammalian fauna, as well as the recent land- * Communicated by OLrprreLp THomas, F.Z.S. + Description Physique de I’ Ile de Créte, 2 vols., Paris, 1869. £ Travels and Researches in Crete, 2 vols., London, 1868 (vol. 1. p. 205). § Op. cit. vol. 11. pp. 408-10. || Ibid. pp. 278-9. © la Face de la Terre, vol. 11. p. 713. 316 MISS D. M. A. BATE ON THE [Nov. 14, shells, of the island shows a decided preponderance of Kuropean types; the only suggestion of a North-African relationship bemg found in the Wild Cat and perhaps the Spiny Mouse. J should like to take this opportunity to express my thanks to Mr. Oldfield Thomas, who has again most courteously g given me every facility for working out my collection in his depar tment of the British Museum (Natural History). List of Species. 1. RuINOLOPHUS FERRUM-EQUINUM Schreb. In the latter part of March three large Horseshoe Bats were secured in a cave close to the sea, on the north-west coast of the island. RHINOLOPHUS HIPPOSIDEROS Bechst. Only one specimen of this Bat was obtained, from a cave in the hills south of Khania. Neither this nor the above mentioned species appears to have been previously recorded from Crete. 3. Myoris Myoris (Bechst.). When visiting the extraordinary underground quarry known as the Labyrinth, near Haghia Dekka in the south of the island, one of the galleries was found to be tenanted by hundreds of Bats belonging to this species. They were hanging from the roof in large clusters and became very noisy when approached. Four specimens were preserved, and these appear to be somewhat smaller than examples from the Continent. These underground galleries have evidently been inhabited by this species of Bat for many hundreds of years; their occurrence in the “ Labyrinth” was noticed by Tournefort as early as about 1700*. 4, MINIOPTERUS SCHREIBERSI (Natt.). Two examples of this species were also procured, and several others observed, in the so-called Labyrinth. These, however, occurred singly, and in galleries other than those occupied by Myotis myotis. 5, KRINACEUS EUROPUS NESIOTES, subsp. n. On comparing the three specimens obtained of the Cretan Hedgehog, these were found to differ from all the forms of E. europeus represented in the British Museum collection ; therefore this island race may be given subspecific rank. In external characters it seems to most closely resemble E. e. italicus Bary.-Ham.t, from which it may be distinguished by its slightly smaller size, dingy appearance, and the lighter * See Raulin, op. cit. vol. ii. p. 1033. + Annu. Mag. "Nat. Hist. ser. 7, vol. v. April 1900, p. 364. 1905. | MAMMALS OF GRETE, 317 colour of the fur. In one specimen (No, 17) this is almost pure white except on the face, hands, and feet. The spines are shorter and more slender, whilst both the short and long hair of the under- parts is much secantier. The following measurements (in millims.) were taken in the flesh :— Head Tail. Hind Ear. Basal length and body. foot. of skull. (oa) (Gyre) eeeeee 208 29 AQ) 29 51-5 INCRE AN (COD) SMe tans eee sits 34 37 28°5 A8 Ione), sate ee NG aye BS OE 5 BOE The skull differs from that of 1. e. italicus, and resembles that of EL. 2. roumanicus Barr.-Ham.* in having the frontal processes of the premaxille squared posteriorly, and further these only extend backward for less than half the length of the nasals. Jn Crete, Hedgehogs are common in the low country, but were not met with in the hills. In captivity they will eat oats freely as well as a more natural diet of eggs ce. 6. FELIS OCREATAYT AGRIUS, subsp. n. This species is the chief exception to the general European appearance of the mammalian fauna of the island, being unmis- takably African in type and belonging to the Felis ocreata group. The two specimens obtained were bought, at different times, in the bazaar at Khania, and therefore are unaccompanied by any measurements taken in the flesh, though they appear to have been large and robust. In one of these, No. 35, the type, which is in summer coat, the average length of fur on the back is about 32 mm., while in the other, No. 36, evidently a winter specimen, the fur is much thicker and longer, averaging 45 mm. in length on the back, and there is at the same time a corresponding difference in the intensity of the markings of the dorsal region. The Cretan race may be distinguished from specimens froni Abyssinia, the type locality, and Egypt, by their much more distinctly marked stripes, both longitudinal and transverse, and by the greater number of rings, or half-rings, on the tail, which is short. As Mr. de Winton has mentioned$, these markings of the dorsal region are more distinct in short-coated specimens ; and on comparing them it is found that even the /ong-haired Cretan skin is more strongly marked than short-haired ones from Abyssinia and Egypt in the British Museum collection, The same holds good in the case of a short-haired specimen from Machakos (B.M. 92.12.3.2.), which otherwise somewhat closely resembles the skin in winter pelage from Crete. It may also be mentioned that some specimens from Abyssinia show a * Op. cit. p. 365. + For use of this specific name see Mr. H. Schwann, “On Felis ocreata and its Subspecies,’ Ann. Mag. Nat. Hist. ser. 7, vol. xi. June 1904, pp. 421-2. From @ypeus, a hunter. Zoology of Egypt (Anderson), London, 1902, p. 173. Me 318 MISS D. M. A. BATE ON THE | Nov. 14, tendency towards a sandy colouring ; this is especially noticeable in a skin from Zoulla (B.M. 69.10.24.9.), in which the trans- verse dorsal bars are much broken up, causing a somewhat ‘“‘ spotty ” appearance. In the specimens from Crete the proximal portion of the fur is decidedly dark over almost the entire body ; this feature is hardly noticeable in those from Abyssinia, and is not so strongly marked in the examples examined from Egypt. The increased richness in colour of the Cretan race is no doubt chiefly due to climatic influences : a still further divergence in this particular direction is exemplified by the wild cat, /. 0. sarda Lataste, from the more westerly island of Sardinia. Hybrids between /. 0. agrius and the domestic cat of the island appear to be not uncommon, and this can easily be accounted for by the fact that formerly small villages were often totally deserted for a considerable time, or possibly entirely, during the insur- rections which occur so frequently in Crete, when the cats, as well as the villagers, are forced to take to a life in the hills. Skins of these hybrids, which are generally of large size like the true wild race, may often be seen hanging up in the bazaars at Khania and Candia, fF. 0. agrius was recorded by Raulin* as F’. catus. 7. MELES MELES MEDITERRANEUS Barr.-Ham. * Only two immature specimens of this Badger were obtained ; these came from an earth in a rocky mound, on the crown of which is perched one of the several monasteries of the Lassethe Plain. The local name for the Badger is “ Arkalos” (épxados); it is plentiful in the island, and is killed in some numbers by the natives, the richer of whom use the skins for saddle-cloths and for making into purses &e. 8. MusrELA FOINA BUNITES{, subsp. n. Five skins of the Cretan Beech-Marten were obtained, and have been carefully compared with those of MW. f. lewcolachnea Blanf., from Turkestan, with which two specimens from Crete, already in the British Museum collection, were formerly iden- tified§. However, the examples from these two localities are found to differ considerably and to be easily distinguishable ; therefore it is proposed that the island form be known by the above-given subspecifie name. In length and woolliness of coat J. f. bunites is intermediate between the typical M. foina and M. f. lewcolachnea, though in general appearance it most closely resembles the latter. From this it differs in its much duller and more uniform colouring, * Op. cit. vol. ii. p. 1038. y+ Ann. Mag. Nat. Hist. ser. 7, vol. iv. pp. 383-4. £ From Bovrirns, a dweller on hills. § Ann. Mag. Nat. Hist. ser. 7, vol. iv. Nov. 1899, p. 313. 1905. | MAMMALS OF CRETE. 319 which is paitly caused by the slighter contrast between the upper and under fur and by the lack of any gloss on the brown hairs, particularly on the paws and tail. The tail is very much less bushy and the fur shorter, in one specimen having an average length on the back of 25-26 mm., while in a skin from Vernoé, Turkestan (B.M. 83.4.21.2.), it is about 43 mm. The size and shape of the throat-patch seem to be even more variable in the Cretan race than it is in others; in one example of the former (No. 31) it is represented by only a few white hairs on either side of the throat close to the fore legs. The following measurements of the type (No. 34) were taken in the flesh :— Head and kody 403 mm., tail 255, hind foot 79, ear 39. The basal length of the skull is 75 mim., and the zygomatic breadth 58 mm. It is perhaps worth noting that MW. /. bunites also has much closer and shorter fur and a less bushy tail than the type of M. f. mediterranea Bary.-Ham.*, from Andalucia, from which it further differs in colour. The Beech-Marten is common in the island, both in the low ground and in the hills, where it is known to occur at Katharo, between 3000 and 4000 feet, though probably its range extends to a much greater height than this. It is killed in some numbers by the peasants, who bring the skins to the larger port-towns on the north coast, whence they are exported, chiefly to Trieste. The Cretans call this Marten “ Zouridha” (Zoupiéa), by which name it is also known in the neighbouring island of Karpathos’. 9. PUTORIUS NIVALIS GALINTHIAS, subsp. n. Only two specimens, without skulls or measurements taken in the flesh, were obtained of this Weasel, which is of large size. These I have been unable to identify with any one of the several races of Putorius nivalis represented in the collection of the British Museum. Therefore it seems necessary to regard it as a local form, which I propose to name after the mythological character changed into a weasel by the Moere and Ilithyie at the time of the buth of Herakles t It was somewhat unexpected to find that, among all the material which I have been able to examine, this island race most closely resembles in general appearance the type (the only specimen in the British Museum collection) of P. 7. atlas Barr.-Ham.§, from the Atlas Mountains, Morocco. Also there seems to be no ap- preciable difference in size between these two subspecies, which are amongst the largest of those belonging to the group of Weasels in which the colours of the upper and under surfaces are sharply divided. * Ann. Mag. Nat. Hist. ser. 7, vol. i. June 1898, p. 442. + “ Karpathos.” Etude géologique &c. Prof. C. de Stefani, Dr. C. I. Forsyth Major, and W. Barbey. Lausanne, 1895, p. 70. a Ibid. p. 65. § Ann. Mee Nat. Hist. ser. 7, vol. xiii. April 1904, p. 323. 320 MISS D. M. A. BATE ON THE [ Nov. 14, Considering the great distance by which the habitats of these two forms are separated, and that a number of other races occupy the intervening and neighbouring countries, the only plausible explanation of such a remarkable likeness seems to be that in this we havea striking case of similar characteristics independently acquired. This does not seem so improbable when it is remem- bered that among the Weasels variation acts only within very narrow limits; the chief points in which differences occur bemg in size, in the line of separation between the two colours, and in the presence and amount of white on the upper surfaces of the paws. In connection with the Cretan form it may be suggested that its large size is, at any rate partly, due to prolonged isolation in a locality where food is plentiful and competition not keen, owing to the absence of Stoats m the island. P. n. galinthias may be distinguished from P. . atlas by its richer colouring and in having only a scarcely perceptible “pencil” of darker hairs at the tip of the tail, which in one of the dried specimens measures 89 mm. exclusive of the terminal hairs. P.n. sicule Barv.-Ham.*, although differing from these species in size and colouring, agrees with them not only in the well- defined line of separation of the colours along the flanks, but also in having white on the upper surfaces of the hind as well as the fore paws. The colour of the under side in one of the skins from Crete (No. 33, ¢) is dirty white; while in the larger of the two (No. 15), the type, probably an old male, this colour is washed with buftish yellow. The “ white” extends im a narrow line along the upper lips to the base of the nose. This Weasel is common and frequently abroad in the daytime, when it may be seen running along the loosely-built stone dykes which are a noticeable feature of many parts of the country, being built for the purpose of ridding the ground of some of the overwhelming number of stones with which it is cumbered. It probably feeds largely on the lizards of various kinds that abound in the island: one day in an olive-grove at Phaestos a weasel was seen to spring out of some thick undergrowth at the edge of a stream and seize a large green lizard (Lacerta viridis major Blgr.), which, on becoming aware of my presence, it hurriedly carried off in its mouth. It is known to the natives by different names in different parts of the island: in the west it is called ‘“ Kalajannou,” in the east “ Sinteknaria” (currecvdapea 7), and in the Lassethe Mountains a modification of the former term which I neglected to make a note of at the time. 10. Mus rarrus Linn. This Rat is found in the port-towns on the northern coast, but * Ann. Mag. Nat. Hist. ser. 7, vol. v. Jan. 1900, p. 46. + “eadoouytexvapta, according to Dr. Forsyth Major, op. cit. p. 63. 1905. } MAMMALS OF CRETE. 321 no specimens were obtained in the country, although traps were frequently set for them in several localities. It is probably owing to the occurrence in considerable numbers of a weasel that this rat has not spread and increased in the interior of the island as it has already done in Cyprus. 11. Mus muscuxus Linn. This species like the last does not, so far as I am aware, occur beyond the limits of the larger towns. In connection with the restricted range of this Mouse, it is interesting to note that a form of Micromys sylvaticus is abundant in the island. 12. Mircromys syLyaricus HAyI (Waterh.). Of all the subspecies of J/. sylvaticus recognised by Major Barrett-Hamilton in his paper published in 1900*, the specimens from Crete seem to agree most closely with J/. s. hayi, though they are, if anything, slightly smaller. In colouring they cannot be distinguished from the darker examples of a series from Cintra, Portugal, in the British Museum collection. None of the Cretan skins shows any sign of a band of colour across the white of the throat. The following are the maximum and minimum measure- ments of the six specimens preserved ;— Head and body 80-88 mim., tail 86—89°5, hind foot 21-22, ear 16-17 ; total length of skull 25-26. This Mouse, which appears not to have been previously recorded from Crete, is plentiful in the island and easily trapped. ‘Two specimens, one of which (No. 11) is very dark, were caught not far from Khania in rocky ground close to some patches of culti- vated land; the remaining four are from Katharo, a small valley in the Lassethe Mountains nearly 4000 feet above sea-level. 15. ACOMYS DIMIDIATUS MINOUS T, subsp. n. The Cretan Spiny Mouse, a richly coloured form with fairly large ears and tail equalling or exceeding in length the head and body, i is evidently closely allied to 4. iniidiat ae It may be dis- tinguished from examples of this species in the British Museum collection from the vicinity of Aden, and one (somewhat faded) from Sinai, the typical locality, by the very restricted area occupied by the spines, which are exceptionally fine and have an average length of about 10°5 mm. Further, these are pigmented for a greater distance from the tip (about 45 mm.), which gives the spmous region a more richly coloured appearance owing to the proximal and semi-transparent portions of the spines “not showing on the surface. The colours of the upper and under surfaces do not intergrade, the line of separation along the flanks being very sharply defined. * “On Mus sylvaticus and its Allies,’ P.Z.S. 1900, p. 387. + “Minotis”’ was employed by the early poets as equivalent to Cretan. 322 ON THE MAMMALS OF CRETE. [ Noy. 14, The following measurements (in millimetres) of the three speci- mens pr eserved were taken in the flesh :-— Skull. aS Head fy Foeotic Sri adie Tail lind footae Wang ee eters length. breadth. Nome er ae ie ey). as No. 12(d) ...... Cae ce tet 15 VF No. 16 (9, PEC NT wote Dalila Wels gen 19:5 ane Tees of subspecies.) | The threespecimens obtained were trapped in the same locality — in rocky ground close to cultivated land between Khania and Suda. It was not known to any of the natives questioned on the subject. This discovery of an Acomys in Crete is interesting, being an extension, in a somewhat unexpected direction, of the recorded range of the genus. 14. Lepus EUROP2US CRETICUS Barr.-Ham. This Hare was described in 1903 by Major Barrett-Hamilton* but as no measurements accompanied the four skins received by him, the following dimensions of a single example (a ¢), taken in the flesh, may be of interest :— Head and body 514 mm., hind foot 123, ear 102. The basal length of the skull is 71 mm. Hares are found all over the island, even near the summit of Mount Ida, which attains a height of over 8000 feet, where Admiral Spratt mentions having disturbed a number out of their ‘‘ forms” in the open snow. The same author remarks that those seen on Mount Ida “seemed to be a smaller species than the Hare of the lowlands.” Unfortunately no specimens were obtained from this locality, so that this observation still awaits confirmation. Of late a close season has been instituted in the island, and the Hare is among the number of species so protected. It was recorded by Raulin = under the name of Z. timidus. 15. ORYCTOLAGUS CUNICULUS CNOSssIUS §, subsp. n. This Rabbit is paler and decidedly more uniformly grey in colour than the typical form; this lightness is partly caused by the paleness of the reddish area on the back of the neck, which more or less affects the greater part of the dorsal region, and further by the absence of a markedly dark ring between the smoky grey of the proximal portion of the hairs and the sub- terminal light band. In the one specimen preserved (a 2), which lived for some months in the Zoological Society’s Gardens, the hind paws are * Ann. Mag. Nat. Hist. ser. 7, vol. x1. Jan. 1903, p. 126. + Op. cit. vol. 1. p. 13. t Op. cit. § “Cnossius ” was employed by the early poets as equivalent to Cretan 1905. ] ON A HORSE BEARING HORN-LIKE STRUCTURES. 323 almost entirely white, and its dimensions, taken in the flesh, are as follows :— Head and body 341 mim., tail 65, hind foot 82, ear 70; weight 2 lbs. #0z. The skull’s greatest length 75 mm., basal length 57:5. It seems curious that this Rabbit does not occur on the main- land of Crete, and I have found no record of its having done so formerly. Raulin wrote* of it as being very plentiful in the small islands off the coast, and a man who brought me three from Dhia, off Candia, said that it is still found there in considerable numbers. 16. CAPRA #GAGRUS CRETENSIS Lorenz-Liburnau 7. The Cretan Wild Goat has been known from very early times, and has doubtless acquired an added interest on account of the legend of Zeus’ upbringing on Mount Ida by the goat Amalthea. It is still found in the three main mountain masses of the island— the Aspro Vouno, Mount Ida, and the Lassethe Mountains. One skin, that of a ¢, was forwarded to me in the spring of the present year (1905), it having been obtained during the winter in the Sphakia district. The horns indicate an animal of eight years old, and measure 605 mm. along the front curve, while the circumference at the base is 175mm. ‘The greatest length cf horn given by Dr. Lorenz-Liburnau £ for this subspecies is 81 em. (810 mm.), this being in a seven-year old specimen preserved in the Vienna Museum. November 28, 1905. Dr. Henry Woopwarp, F.R.S., Vice-President, in the Chair. Mr. J.T. Cunningham, M.A., F.Z.S., exhibited some photo- graphs of a Horse bearing structures that’ he interpreted as incipient horns, and made the following remarks :— The peculiarity of the horse represented in these photographs was described by Dr. G. W. Eustace, of Arundel, before the Linnean Society in 1903. The horse, the name of which is “¢ Domain,” was then in the stables of Mr. Alfred Day at ‘ The Hermitage’ near Arundel, and was still there when, by the ‘kindness of Mr. Day, these photographs were taken for me in October last. A few other similar cases have been recorded, but the pedigree of Domain contains no individuals which are known to have possessed the peculiarity, and it appears therefore to be a new variation, not a result of reversion or heredity. Dr. Eustace’s paper was illustrated by plaster casts of the fore- head of Domain which are now in the Natural History Museum, and Dr. Ridewood has presented to the Museum the frontal * Op. cit. vol. 1. p. 253. + ‘Die Wildziegen der Griechischen Inseln &c.,’ 1889. t Op. cit. p. 24. 324 MR. W. R. OGILVIE-GRANT ON BIRDS FROM JAPAN. [ Nov. 28, portion of the skull from another case whose history is unknown. ‘Domain’ was stated to be five years old in 1903, so that he was seven years of age when the photographs were taken. The horns are about 2 inch in length, the left slightly larger than the right. There can be no doubt that they are outgrowths of the frontal bone. They are covered by normal skin and hair. Mr. Frank Slade, F.Z.8., showed three photographs of the Sea-Anemone (Anemonia sulcata), which had been taken from life in the Horniman Museum at Forest Hill, in the process of division. The first photograph showed the Anemone at rest after having made the initial tear in the body-wall. The second showed the animal, two days later, straiming to increase the tear, whilst the third, taken after an interval of sixteen days, showed the division completed. Mr. Douglas English exhibited and made remarks upon a living albino Field-Vole (JMcrotus agrestis) which had been captured last July in Wales. My. G. A. Boulenger, F.R.S., exhibited a living Lizard, Lacerta muralis, from Brozzi, province Florence, which he had received from Dr. A. Banchi, through the mediaticn of Dr. J. de Bedriaga, C.M.Z.S8. The lizard belonged to the typical form of the Wall- Lizard, but was remarkable for its black coloration, above and below. Melanistic forms of the Wall-Lizard were well known on small islands in the Mediterranean, but, so far as Mr. Boulenger was aware, no black specimen had ever been recorded from the mainland. The scales across the body numbered: 58 and the lamellar scales under the fourth toe 25 in the specimen exhibited ; these two numbers being sufficient to distinguish the Brozzi lizard from the melanistic insulars previously described. Capt. Albert Pam, F.Z.S., made some remarks on a living specimen of the Violet-cheeked Humming-bird (Petasophora iolota) which he had recently brought home from Venezuela and presented to the Society’s Menagerie. He also gave a general account of the habits of these birds, as observed by him, in a wild and captive state, and notes on their management and feeding while in confinement. Mr. W. R. Ogilvie-Grant, F.Z.8., sent for exhibition a named set of the Birds collected in Japan by Mr. M. P. Anderson in connection with the Duke of Bedford’s Exploration in Eastern Asia. No new species were discovered, but several of the spe- cimens were of special interest as illustrating stages of plumage not represented in the British Museum. 1905. | MR. R, LYDEKKER ON GUEREZA MONKEYS. 325 The following papers were read :— 1. Colour Hyvolution in Guereza Monkeys. By R. LyDEKKER. [Received November 7, 1905. | (Text-figures 54-58.) An interesting example of the progressive evolution of specialised features in colouring (if we may thus term com-~ binations of black and white) is afforded by the black and black- and-white African long-haired monkeys included in the genus Colobus, and which may be collectively designated Guerezas, although the name “ guereza” refers properly only to the north- east African representative of the group. Beginning at one end of the series, we have the Black Guereza (Colobus satanas), of West Africa, which, as shown in text-fig. 54, Text-fig. 54. Black Guereza (Colobus satanas). is wholly black with tufts of long hair on each side of the face and throat, a pointed crest on the crown of the head, and the long tail short-haired from base to tip. Following on this we may take a variety of the Mantled Guereza from East Central Africa which I have recently described as Colobus palliatus cottoni, in which the face-tufts, chin, and narrow pendent tufts of long hair on the shoulders are white, while the terminal half of the tail is grey with a white tip, which shows a slight tendency to 326 MR. R. LYDEKKER ON GUEREZA MONKEYS. [ Nov. 28, expand into a brush. A further development is exhibited by the typical form of Colobus palliatus (text-fig. 55), from British Kast Africa and the neighbouring districts, in which the two lateral white face-tufts are connected by a white band across the brow, while the shoulder-tufts are of considerably larger size, a small whitish patch beneath the tail occupies the perineal region, and the tail has its terminal third whitish, the middle third grey, and the remainder black. A nearly allied type is found in the form of the Mountain Guereza (C. ruwenzorii), of the Ruwenzori dis- trict, in which the white perineal patch has assumed much larger proportions, although the extreme tip of the tail is alone greyish ; the latter feature placing the species, so far as the colouring of this appendage is concerned, next to the Black Guereza. Text-fig. 55. Mantled Guereza (Colobus palliatus). From the three foregoing black-and-white forms there is an easy transition to Sharpe’s Guereza (C. sharpei), of Nyasaland, in which, as shown in text-fig. 56, p. 327, the white brow-band, face-, throat-, and shoulder-tufts have become very long and pronounced, the hairs of the last hanging down the outer side of the fore-limbs. Moreover, the white terminal third of the tail has developed a dis- tinct tuft, not dissimilar in relative size and form to that of a lion’s tail. A step still further in advance is taken by the typical Guereza (C’. guereza) of Abyssinia and North-east Africa generally. In this handsome monkey the white shoulder-tufts extend back- wards to form a long mantle, falling down each side of the body 1905. ] MR. R. LYDEKKER ON GUEREZA MONKEYS. 327 Text-fig. 56. rete att Sa Sharpe’s Guereza (Colobus sharpei). Text-fig. 57. White-tailed Guereza (Colobus caudatus). Proc. Zoou. Soc.—1905, Vor. Il. No. XXITT. 23 328 MR. R. LYDEKKER ON GUEREZA MONKEYS. [Nov. 28, and uniting on the lower part of the back. . The culmination of this type of coloration is formed by the White-tailed Guereza (C. caudatus, or albocaudatus as it ought to have been called) of the Kilimanjaro district and other parts of Eastern Africa. Here, as we see from text-fig. 57, the beautiful pendent white mantle has become still longer, and the tail, which is wholly white except for a very small length at the root, is clothed with long pendent hair comparable to the “ flag” of a setter ; the cheek- and throat-tufts, however, have been completely lost, so that the head is wholly short-haired, with the face and throat white. The difference between the species last-named and the Black Guereza in the matter of colouring is enormous, and yet the transition from the one to the other in this respect is almost complete. In the case of the white-tailed species the excessive Text-fig. 58. White-thighed Guereza (Colobus vellerosus). length of the white hair forming the mantle and the tail-fringe appears to have been evolved in order to render the creature as in- conspicuous as possible amid the long pendent greyish-white lichens which clothe the branches of the trees of an East African forest. The evolution of such a type is, of course, easy to comprehend ; but, as in so many other cases, the difficulty comes in with regard to the purpose of the coloration in the intermediate types con- necting this species with the Black Guereza. What purpose do these incipient attempts at the development of a pied coat serve ? The line of evolution culminating in the white-tailed species by no means, however, brings us to the end of the modifications in the colour and local development of the hair in this group of monkeys, 1905. ] ON THE WHITE-MANED SEROW. 329 for the West African White-thighed Guereza (C. vellerosus), text- fig. 58, appears to exhibit a kind of retrograde development in these respects. The body, for instance, has entirely lost the mantle of long white hair and the tail its white “flag,” while the white of the perineal patch has spread on to the hinder and outer sides of the thighs. In this case we find, indeed, a practical reversion to the type of the Black Guereza, with the exception that the band on the forehead, the sides of the face and throat, the thighs, and almost the whole of the tail have become white, while the long hair has entirely disappeared from the face. That the colouring and special development of the long hair in the White-tailed Guereza form a protective modification, there seems to be little doubt. Whether, however, the colour-phases and hair-growth in the other forms are of a protective nature, or are merely due to what is commonly called sexual selection, must be left for those to decide who have the opportunity of seeing these beautiful monkeys in their native haunts. 2. The White-maned Serow. By R. LypEeKKeEr. [Received November 11, 1905. | (Plate VIII.*) In 1888 the very appropriate name of Vemorhedus argyrochetes was bestowed by the Rev. Pére Heude* on a large and strikingly coloured species of Serow inhabiting the mountains of Central China in the neighbourhood of Che-kiang in the Upper Yang-tse- kiang district. Later, a fuller notice, with a figure of the skull, was given by the same writer; while in 1890 Dr. A. Henry $ contributed a note on skins of the species which had come under his notice while in China. Hitherto, however, so far as I am aware, no coloured figure of the entire animal has appeared ; and since the colouring is of a very remarkable and striking type, somewhat different from that of the ordinary Serow, I think the opportunity ought to be taken of remedying this deficiency. This opportunity has heen afforded by the recent addition to the Collection of the British (Natural History) Museum of a mounted male specimen of this Serow and of the Tibetan Takin ~ (Budorcas taxicolor tibetana). They were acquired by Rowland Ward, Ltd., from a French dealer, by whom they were stated to have come from Tibet; but I should think that Sze-chuen, or thereabouts, is more probably their place of origin, unless, indeed, the Serow was procured still farther east. The two are, I believe, the first representatives of their respective kinds ever received in England, and it is quite probable that in the case of the Serow this statement may be extended to European museums in general. * For explanation of the Plate, see p. 331. + Mém. Hist. Nat. Emp. Chinois, vol. ii. p. 4, note (1888). t T. ce. p. 228, pl. xxxi. (1890). § Proc. Zool. Soc. London, 1880, p. 93. 23* 330 ON THE WHITE-MANED SEROW. [ Nov. 28, In his note of 1890, Dr. Henry described the White-maned Serow as being as tall as a cow, and employed by the natives of Central China for riding and as a beast of burden. This, I think, is somewhat exaggerating matters (unless a very small breed of cattle is referred to), and a good-sized donkey would seem to be a better standard of comparison. Although, from the circumstance that the skull still retains the last two premolars, and therefore indicates an animal not yet fully mature, it is possible that the specimen in the Museum does not quite represent the full height attained by the species, yet it certainly cannot fall very short of this, and, as mounted, the skin indicates an animal only about three inches taller than the ordinary Himalayan or (as it may well be called) Black-maned Serow. The general build and type of coloration approximate to those of the last-named animal, although im two respects there are marked peculiarities in the matter of colourmg. The horns are small and very thickly rmged for more than half then length, differing, I think, in these respects to some extent from those of the Himalayan animal, although, on account of the immaturity of the Museum specimen, I cannot be confident on these points. The ears certainly appear to be larger, but here again it is difficult to say that there may not have been shrinkage in the mounting of the Himalayan specimens. The general colour of the upper-parts is mingled black and white, but the face and outer surface of the ears are blackish brown, with an admixture of chestnut hairs on the upper part of the forehead and the sides of the upper lips. The insides of the ears and part of the muzzle are white, but the white area on the latter is of much smaller extent than in the Himalayan species, being confined to the margins of the upper lips, although wider on the lower lips, whence it extends as a streak on the sides of the jaws. A large patch on the throat, another on the chest, and the whole mane are dirty white. On the other hand, the lower portion of the buttocks, the whole hind limb (except a light streak inside), and the middle third of the tail are bright mahogany or ferruginous red. The fore legs from the knees downwards, and to some extent on their inner and outer sides above the latter, are of a lighter and more chestnut-red. The upward extension of the red of the legs and its deeper tone are features of this species as contrasted with the Himalayan and Sumatran Serows, which are best regarded as varieties of one species, the latter distinguished by the legs being chestnut in place of wholly white. It appears, however, that in some cases the Sumatran Serow has the mane white, as in the present animal. The prevalence of bright red, reddish yellow, and yellow in the colouring of mammals of the West Tibetan province, as exemplified by Rhinopithecus roellane, Budorcas taxicolor tibetana, and the present species, is very remarkable, and stands in need of explanation. The skull belonging to the same individual as the skin is in a 1905. | ON MAMMALS FROM JAPAN. 331 somewhat damaged condition, having a large portion of the parieto-frontal region cut away, and also lacking the nasal and premaxillary bones; it still, however, serves as a basis of com- parison between the present species and V. bubalinus. The third and fourth milk-molars are still retained, the second premolar is just piercing the gum, and the third molar has its summits slightly abraded by wear. The animal may therefore be con- sidered to have been sub-adult at the time of its death, and may perhaps have not quite attained its full stature. As it is, the skull is fully as large as that of an aged individual of V. bubalinus, but appears to have been of a relatively broader, deeper, and shorter type, although from its imperfection I cannot be sure on all these points. The palate is, however, evidently wider, the interval between the bases of the second molar being about a quarter of an inch more than in the Himalayan species. The basioccipital and basisphenoidal rostrum is also markedly wider and more tapering, with less prominence of the anterior tubercles for muscular attachment, which are, however, much larger. Perhaps the most important distinctive feature of the skull of the white-maned species is the much greater backward extent of the nasals on to the frontal region, in consequence of which the fronto-nasal suture is situated only a short distance in front of the vertical line formed by the anterior border of the orbit, instead of very considerably in advance of the same. The pit for the face-gland also occupies nearly the whole extent of the lachrymal bone, instead of leaving a large flat surface along the upper border of the same. The palatine bones likewise extend much further forward on the palate, so that the palato-maxillary foramina are situated on the line of the hinder lobe of the first molar instead of opposite the cleft between the two lobes of the second tooth of the same series. EXPLANATION OF PLATE VIII. The White-maned Serow (Nemorhedus argyrochetes), from the specimen in the British Museum. 3. The Duke of Bedford’s Zoological Exploration in Eastern Asia—I. List of Mammals obtained by Mr. M. P. Anderson in Japan. By Oiprietp THomas, F.R.S.* [Received October 9, 1905. } (Plate IX.7) AsT announced at the last Meeting of the Society, our President, His Grace the Duke of Bedford, K.G., has consented, in order to * [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 here by the name being underlined.— Eprror. | + For explanation of the Plate, see p. 363. 332 MR. OLDFIELD THOMAS ON [Nov. 28, further the cause of zoological exploration, to bear the cost of a collector working systematically through the islands of the Far East, and I now have to give an account of the Mammals obtained by that collector—Mr. Malcolm P. Anderson—in Japan, where he has begun his labours. The selection of Japan for the first field of Mr. Anderson’s work was almost a necessity, for practically nothing has been done with regard to the Mammalogy of that country since the time of Siebold and Burger, whose collections resulted in Temminck’s great work of 1842-45, while authentic modern specimens of the species then described were necessary before any further progress could be made. It is true that, by the liberality of the Leyden Museum, typical specimens of most of Temminck’s species were distributed to many European museums, our own National Museum receiving a very complete set, but these specimens, 60 to 70 years old, are all naturally much deteriorated by exposure to light, while scarcely any of them have any more exact locality recorded than “ Japan.” Under these circumstances no words can express the value of the fine series of specimens obtained by Mr. Anderson, who has proved himself a most successful collector, and one well able to do full justice to the liberality of our President. The specimens, which are of all classes, after being exhibited before the Society, are to be transferred by His Grace to our National Museum, where they will be most gratefully appreciated. Mr. Anderson landed at Yokohama in July 1904, and began collecting at once in Hondo (Central and North), and made two trips to Hokkaido in September and November. Later he went to the two large southern islands, Shikoku in March 1905, and Kiushiu in April and May, the specimens from this latter being particularly valuable, as the Dutch Factory, from which the Leyden Museum obtained its materials, was situated at Nagasaki, at the south-eastern corner of the island. Lastly, Mr. Anderson visited the Oki group of islands, to the north of Southern Hondo, and Tanegashima, south of Kiushiu, while his brother, Mr. Robert Anderson, went to Yakushima, still further south. The specimens from these separate island collections 1 have included in an Appendix to the paper. In all, the mammals: referred to in the present communication number about 600 examples, belonging to 50 species and sub- species. Of these I have found it necessary to give new names to twelve, besides describing one new Shrew discovered by Mr. Hawker in 1903. The fine new Marten, Mustela melampus bedfordi, now figured, but described last session, is also to be eredited to Mr. Anderson’s collection. As yet I am chary of making any general conclusions about the mammal-fauna of Japan. It is, however, evident from this collection that there is little faunistic difference between Shikoku, Kiushiu, and the main southern portion of Hondo, but that a number of species do not extend into the north of Hondo, where 1905. | MAMMALS FROM JAPAN, 333 such species as occur are sometimes subspecifically different from those of the south. Hokkaido, of course, is very different from Hondo, and the occurrence there of such typically Japanese species as Mieromys speciosus and geisha is somewhat unexpected ; possibly they are comparatively recent introductions, even though in each case subspecifically separable from their Hondo relatives. The following is a list of the species obtained in each ofithe four islands referred to :— Hokkaido : Sciurus vulgaris orients. Mus norvegicus. Micromys speciosus ainu, geisha hokkaidi. Evotomys mikado, bedfordie. Lepus timidus ainw. Hondo: Pinistrellus abramus. Sorex shinto, hawkeri. Crocidura dsi-nezumi chisar. Chimarrogale platycephala. Mogera wogura. Urotrichus talpoides pilirostris. Canis hodophylax. Nyctereutes viverrinus. Mustela melampus bedfordi. Putorvus ttatsr. Petaurista leucogenys. Sciuropterus momonga amygdali. Sciurus lis. Mus tanezumi, molossinus. Micromys speciosus, geisha. Microtus montebelli. Evotomys (Craseomys) andersont. ae (Phaulomys) smithit. Lepus brachyurus. Sus leucomystax. Nemorhedus crispus. Cervus sika. Shikoku : Macacus fuscatus. Crocidura dsi-nezumt. Mogera wogura kobee. Urotrichus talpordes. Putorius ttatst. Meies anakuma. Micromys speciosus, geisha, minutus japonicus. Evotomys (Phaulomys) smathir. Lepus brachyurus. 334 MR. OLDFIELD THOMAS ON [ Nov. 28, Kiushiu : Rhinolophus ferrum-equinum nippon. Myotis macrodactylus, natterert hombinus. Miniopterus schreibersi japonie. Crocidura cerulea, dsi-nezwmi. Urotrichus talpordes. Mustela melampus. Putorius itatst. Meles anakuma. Petaurista leucogenys. Micromys speciosus, geisha, minutus japonicus. Microtus montebelli. Evotomys (Phaulomys) smathii. Lepus brachyurus. In order to focus so far as possible the existing information about the mammals of Japan, I have prepared the following résumé of the literature, which, apart from 'Temminck’s fine work, is of a very fragmentary character. 1824. Siebold, G. T. de. Spicilegia Faun. Japon., in Dissertatio Hist. Nat. Japon. p. 13. Description of “ Wyoaus lineatus”—a Tamias, from Hokkaido. 1842-45. Temminck, C. J. Mammalia of P. F. de Siebold’s ‘Fauna Japonica,’ pp. 1-60, pls. 1.—xxx. A complete account of the Fauna, as known from the col- lections sent by Messrs. Siebold, Burger, and other Dutch naturalists to the Leyden Museum. One of the finest and most complete faunistic works ever published. Up to 1904, that is, for more than sixty years, the only valid species added to those contained in it were MWicrotus montebelli M.-Edw., Murina hilgendorfi Peters, and Talpa mizura Giinth. The following species are first described in this great work :— Macacus fuscatus (under the name of /nwits speciosus), Pteropus dasymallus, Rhinolophus nippon and cornutus, Pterygistes molossus, Myotis macrodactylus, Pipistrellus abramus and akakomuli, Talpa wogura, Urotrichus talpoides, Chimarrogale _platycephala, Crocidura dsi-nezumi and umbrina, Meles ana- kuma, Mustela melampus and brachyura, Putorius ttatsi, Nyctereutes viverrinus, Lepus brachyurus, Sciwrus lis, Pieromys leucogenys, Sciuropterus momonga, Mus erythronotus, argenteus, molossinus, tanezumi, and speciosus, Glirulus japonicus, Cervus sika, Nemorhedus crispus, and Sus leucomystax. 1857. Schlegel, H. Ursus japonicus, sp.u. Handl. Beoefening der Dierkunde, 1. p. 42; Sclater, P.Z.8. 1862, p. 261; Giinther, P.Z.8. 1880, p. 442. The U. torquatus of the ‘ Fauna Japonica.’ 1905. ] MAMMALS FROM JAPAN. 335 1862.. Gray, J. E. Leopardus japonensis, sp.nu. P.Z.S8. 1862, p. 262, pl. xxxiii. Based on a tanned Leopard-skin without exact locality. 1865. Gray, J. E. ; Martes japonica, sp. n. P.Z.8. 1865, p. 104; Cat. Carn. B.M. p. 82 (1869). ‘No doubt the summer form of Mustela melampus. 1867. Gray, J. E. Lutronectes whiteleyt, g. & sp. nov. P.Z.S. 1867, p. 181; Cat. Carn. B.M. p. 107 (1869). Based on young specimens of the Japanese Otter. 1868. Gray, J. E. Vulpes japonicus, sp.n. P.Z.S. 1868, p. 517; Cat..Carn. B.M. p. 204 (1869). “ Japan.” 1874. Milne-Edwards, A. Recherches Mammifeéres. Texte, p. 285. Description of Microtus montebelli, from Fusi-yama. 1875. Von Martens, E. Die Preussische Expedition nach Ost- Asien. Zoologische Abtheilung. I. Pt. 1, pp. 75 & 362. General account of Japanese mammal-fauna, and list of species obtained. (Determinations by W. Peters.) 1875. Rein, J. J. Notizen iiber die Verbreitung einiger Siuge- thiere auf Nippon. Zool. Gart. xvi. 1875, p. 55. Notes on habits, native names, and distribution of thirteen of the better-known species. In the same author’s ‘Japan,’ 1881, 1. p. 201, these notes are incorporated in a general popular account of the fauna. 1880. Giinther, A. Notes on some Japanese Mammalia. P. Z.S. 1880, p. 440. Notes on Urotrichus talpoides (with description of Vewrotrichus g.n. for the American U. gibbsi), Talpa mizura, sp. n., Ursus arctos, U. japonicus, and Calorhinus ursinus. The new Mole, Zalpa mizura, has not since been obtained. It is closely allied to the European 7’. europea. 1880. Peters, W. Ueber die von Hrn. Dr. F. Hilgendorf in Japan gesammelten Chiropteren. MB. Ak. Berl. 1880, p. 23. Records of 7 species, and description of Murina hilgendorfi from Yedo, near Tokyo. 1880, Thomas, O. On the Myoxus elegans of Temminck. P.Z.S. 1880, p. 40. See below, under Glirulus japonicus. 336 MR. OLDFIELD THOMAS ON [ Nov. 28, 1882. Doederlein, L. Ueber einige Japanische Saiugethiere. MT. Deutsch. Ges. Ostasiens, vol. 111. Heft 25, p. 210. (1) Existence of Fox in Shikoku. (2) A Changing Hare (“ Lepus variabilis”) in Japan. (3) On a small musky-smelling rodent (more likely a Shrew). 1886. True, F. W. Description of a new genus and species of Mole (Dymecodon pilirostris) from Japan. Pr. U.S. Nat. Mus. 1886, p. 97. From Yenoshima, near Tokyo. This Mole is probably an immature Urotrichus talpoides. 1900. Barrett-Hamilton, G. E. H. Lepus timidus ainu, sabsp.n. P.Z.S8. 1900, p. 90. From Hokkaido. 1904, Sasaki, C. A new Field-Mouse in Japan. Bull. Coll. Agric. Tokyo, vi. p. 51. Description of Arvicola hatanedzumi (= Microtus montebelli), from Tokyo. 1904. Beard, J. C. Nyctereutes albus, sp.n. Scientific American, 1904, p. 237. Based on a white specimen in the New York Zoological Park, said to be from Hokkaido. 1905. Thomas, O. On some new Japanese Mammals presented to the British Museum by Mr. R. Gordon Smith. Ann. & Mag. N. H. (7) xvi. p. 487. Descriptions of Mogera wogura kobew, Petaurista leucogenys nikkonis, oreas, and tose, Micromys geisha, and Hvotomys (Phaulomys, subg. n.) smithit. 1905. Thomas, O. Exhibition of Mammals from Japan. Abstr. IP /Ap tle USOS, jos OS IPs As tele IOS) ithe foe ssh, Description of Mustela melampus bedfordi. 1. Macacus Fuscatus Bly. Macacus fuscatus Bly. J. A. S. B. xliv. extra number, p. 6 (1875). 6. 304. ©. 303, 323. Jinrio, Tokushima Ken, Shikoku. 500’. This is the Jnwus speciosus of the ‘ Fauna Japonica,’ nec F. Cuv. ‘¢ Numbers of monkeys live in the forest surrounding certain large temples at a distance from Jinrio. I did not see them, but sent my servant in search of them, and through him secured these specimens. They are considered difficult to hunt, for they hide themselves very effectually in the high Cryptomeria trees. It is said that with the help of a dog they can easily be shot, as the monkey pays little heed to the man and his whole attention is absorbed in exhibiting his anger towards the barking dog. The flesh is commonly eaten by the natives, but on trying it I did not dee) hey) SL, J Al 1905. | MAMMALS FROM JAPAN, 337 2. RHINOLOPHUS FERRUM-EQUINUM NIPPON Temm. 3. 485. Tano, Miyasaki Ken, Kiushiu. 3. PIPISTRELLUS ABRAMUS Temm. g. 6. Takayu, near Yonezawa, Uzeu, N. Central Hondo. Dr. Jentink* has shown that Temminck’s Vespertilio akokomuli is the same species as his V. abramus. The type locality of both is Nagasaki, Kiushiu. “Caught, with the two succeeding species, in caves near the village.”— WM. P. A. 4, Myoris (LEvcoNoE) MAcRoDACTYLUS Temm. 3. 490,500. 9. 493, 494, 515, 516, 517. Tano, Miyasaki Ken, Kiushiu. 500’. Dimensions of an adult male :— Forearm 36 mm. Head and body 44; tail 35; ear 14:5. These specimens agree absolutely with Temminck’s description, and there can be no doubt that they belong to his species, in spite of Peters’s assertion t that macrodactylus resembled very closely the European J. capaccinii, to which these examples bear no resemblance whatever. Indeed, so great is the discrepancy, that I am tempted to suppose that Peters did not really see the specimens described by Temminck at all. I. macrodactylus in fact is more closely allied to WM, daubentoni. 5. MYoris NATTERERI BOMBINUS, subsp. n. @. 486, 487, 488, 489, 492. Tano, Miyasaki Ken, Kiushiu. 500". Similar in essential respects to the European J. natiereri, which it evidently represents in Japan. But the ear appears to be rather longer (judging from skins only), the tragus narrower and more boldly curved outwards, the skullis more abruptly and considerably inflated in the frontal region, and the colour is not quite the same. In true W. natterert the colour is paler and more uniform than in the other small European species of I/yotis, this being apparently due to the fact that the pale brown ends to the hairs are longer and therefore hide the blackish-grey of their bases. In bombinus, however, the coloration is more normal, a darker variegated brown, the blackish-grey bases of the hairs showing through. In a similar way below, the light ends to the hairs are shorter and less prominently white. Dimensions of the type :— Forearm 40 mm. Head and body 52; tail 44; ear 17. * Notes Leyd. Mus. ii. p. 37 (1879). + MB. Ak. Berl. 1866, p. 681. Dobson, on this statement, actually synonymised macrodactylus with capaccinii. 338 MR. OLDFIELD THOMAS ON [ Nov. 28, Skull—greatest length 15:5; basal length in middle line 11°8 ; front of canine to back of m® 5:9. Type. Old female. B.M. No.6.1.4.14. Original number 487. Collected 30 April, 1905. It is a matter of interest to find in the Far East this repre- sentative of JZ. natterert, which has hitherto only been known from Kurope. It is probable, also, that Mr. Miller’s I. thysanodes is the corresponding Bat in the N. American fauna. 6. MINIOPTERUS SCHREIBERSI JAPONL®, subsp. n. 3. 501, 502, 503, 505, 510, 511, 512, 513,514. 9.495, 496, 497, 904, 507, 508, 518, 519, 520, 521. And twoim alcohol. Tano, Miyasaki Ken, Kiushiu. Alt. 500 feet. Size rather large, uniformly larger than in the Liu-Kiu form, M. fuscus Bonh.* Colour of back between ‘“seal-brown” and dark ‘ Prout’s brown”; head and nape rather greyer, though the difference is perhaps due rather to the ends of the hairs being more glossy and so catching the light, than to any essential difference in colour. Under surface like the head. Dimensions of the type :— Forearm 47 mm. (range from 46 to 48). Head and body 57; tail 53; ear 12. Skull—greatest length 16, median basal length 12; front of canine to back of m* 6-4. Type. Adult male. B.M. No. 6.1.4.22. Original number 512. Collected 3 May, 1905. As was to be expected, the Japanese J/iniopterus is clearly different from the pale European one, nor does any form quite agreeing with it appear to have been described. Bonhote’s M. fuscus from the Liu-Kiu Is. is similar in colour (or slightly darker), but is uniformly smaller, the forearm rarely reaching 44mm. Seven additional examples of fwscus recently received from Mr. A. Owston confirm the characters derived from the three originally examined by Mr. Bonhote. The fine series obtained by Mr. Anderson is remarkably uniform both in colour and size. This Bat is of course the ‘“‘ Vespertilio blepotis” of the ‘ Fauna Japonica’; but that species was primarily described on examples from Java. i) 7. SoREX sHINTO Thos. Sorex shinto Thos. Abstr. P. Z.S. No. 23, p. 19, Dee. 5, 1905. 3.47. Makado, near Nohechi, Aomori Ken, N. Hondo. Alt. 400 ft. A small species with a long tail. Size as in S. macropygmeus Miller, though tail much longer. Fur of back slightly over 3mm. in length. General colour above * Nov. Zool. ix. p. 626 (1902). 1905. ] MAMMALS FROM JAPAN. 339 uniform brown (between ‘ seal-brown” and “ Prout’s brown”), quite similar in tone from head to rump. Sides not presenting a contrasted light area, being scarcely lighter than the back, and passing gradually without line of demarcation into the drab-washed belly. Chin and throat slightly more greyish. Hands and feet glossy brown, the hairs at the tips of the digits silvery. Tail long, nearly as long as the head and body, well-haired throughout, slightly pencilled at tip, blackish brown above, dull whitish below. Skull conspicuously larger than that of S. minutus, but of the same light and delicate build. Upper unicuspids subequal in transverse section, the last one not smaller than the rest or outof series. All the teeth liberally tipped with brown. Dimensions of the type, measured in the flesh :— Head and body 50 mm.; tail 49; hind foot 11:5; ear 7. Skull—greatest length 17°5; basal length 15; breadth of brain-case 8°5; length of upper tooth-series 7-6. Hab. as above. Type. Adult male, B.M. No. 6.1.4.30. Original number 47. Collected 28 September, 1904. This small Shrew belongs to the genus Sorex, which had not hitherto been recorded from Japan. It appears to be a member of the S. minutus group, but may be readily distinguished by its size and unusually long tail. A second specimen of it is in the Museum collection, obtained by Mr. Alan Owston near Tokyo. [SoREX HAWKERI, sp. n. 3. Inukawa, Yedo, Hondo. Size very small, about as in S. minutus. Fur, in summer, about 3 mm. long on the back. General colour of the dorsal area brown, near ‘‘ Prout’s brown” but considerably paler. Sides distinctly different from back, greyish broccoli-brown ; belly like sides, but paler. Upper surface of hands and feet pale drab. Tail short, much shorter than head and body, well-haired and pencilled, pale broccoli-brown above, rather lighter below. Skull crushed in the single specimen, but apparently even smaller than that of S. minutus, as indicated by the short tooth-row. Anterior incisor large, not very deeply notched. Five upper uni- cuspids broad, closely packed, in even slightly decreasing sequence to the penultimate, the fifth again slightly larger than the fourth, in the tooth-row, clearly visible from without. Brown on teeth about as in S. minutus. Dimensions of the type :— Head and body (measured in skin) 55 mm.; tail (measured in flesh) 30; hind foot 9. Skull—tip of anterior incisor to that of large premolar 2:8 ; to back of m* 5:2. Type. Male. B.M. No. 3.9.10.1. Collected 7 June, 1903, and presented by R. McD. Hawker, Esq. Although not collected by Mr. Anderson, I take this opportunity to describe a tiny Shrew obtained in Japan by Mr. Hawker. It 340 MR. OLDFIELD THOMAS ON [ Nov. 28, has no alliance with any of the European types of Shrews, but would seem to have relatives among some of the short-tailed Arctic American forms, with which it will no doubt prove to be linked by allied species from the East Siberian mainland. | 8. CROCIDURA CHRULEA Kerr. 4 in alcohol. Nagasaki; in houses. 9. CrocrpuRA (CR.) DsI-NEzZUMI Temm. S$. 263, 275, 293. Jinrio, Tokushima Ken, Shikoku. 500’. 9. 321. Tanano, Tokushima Ken, Shikoku. 250’. 3g. 420, 423. ©. 446. Takamori, Kumamoto Ken, Kiushiu. 1850’. 3. 463. Kawachi, Miyasaki Ken, Kiushiu. 1500’. Flesh-measurements of an adult male from Shikoku :—Head and body 67 mm.; tail 46; hind foot 13°5 ; ear 9. These specimens from the two southern islands are all of a uniform brownish grey, very lke the colour of the European C’. russula, of which this is evidently the Japanese representative. The type locality is presumably Kiushiu. But the corresponding Shrew found by Mr. Anderson in the north of Hondo is not quite the same, and may be regarded as a different subspecies. 10. GrociIDURA DSI-NEZUMI CHISAI*, subsp. n. ) Pp go. 51. @. 48. Tsunagi, near Morioka, N. Hondo. 9.93. Morioka, N. Hondo. Similar to dsi-nezumi in size, but tail rather and hind feet decidedly shorter. General colour much darker, near ‘ seal- brown,” but browner and less purple. Under surface ‘“ mouse- grey.” Tail uniformly dark brown. Skull rather narrower, less broadly flattened than in dsi- nezume. Measurements (in mm.) of three specimens in the flesh :— 51. (Type.) Head and body 68; tail 43; hind foot 12. 48. ” ” Og a) O03 ” ne 93. ” ” 3 oegy Oehs ” le Skull of type—Greatest length 18, basal length 15°4; greatest breadth 8°5; length of upper tooth-series 7:5. Type. Adult male. B.M. No. 6.1.4.43. Original number 51. Collected 3 October, 1904. Temmincek’s “ Sorex umbrinus” would appear to be similar to this Shrew in colour and size, but is distinguished by its very long tail, 54 mm. in length. 11. CHIMARROGALE PLATYCEPHALA ’emm. @. 168. Tajima, E. coast of Izu Peninsula, 8.E. Hondo. Flesh measurements :—Head and body 112 mm.; tail 90; hind foot 25°5; ear 9. * Chisai=small. 1905. |} MAMMALS FROM JAPAN. 341 12. Mocera wocura Temm. gd. 96. Tsunagi, near Morioka, Iwate Ken, N. Hondo. Flesh measurements :—Head and body 108 mm.; tail 20; hind foot 16°5. 13. MoGERA woGuRA KOBE Thos. 6. 336. Ochi, Kochi Ken, Shikoku. Flesh measurements :—Head and body 160 mm.; tail 25; hind foot 21. No specimens intermediate in size between these large Moles and the true wogura have as yet turned up. Possibly the two forms ought to be regarded as specifically distinct. 14. Urorricuus TALPoIDEs Temm. 3. 397, 398, 406, 429, 444. 9. 385, 396, 422, 430, 435. Takamori, Kumamoto Ken, Kiushiu. 1800’, 3. 471, 472, 476. 9. 447, 473, 474. Kawachi, Miyasaki Ken, Kiushiu. 1500’. 3. 274, 290. &. 288, 289, 310. Jinrio, Tokushima Ken, Shikoku. 500’. 2. 314. Fukuhara, Tokushima Ken, Shikoku. 750’. Q. 324. Ikeda, Tokushima Ken, Shikoku. 3. 339, 357. 2. 337, 338, 342, 343,358. Ochi, Kochi Ken, Shikoku. 1300’, ©. 329. Sakawa, Kochi Ken, Shikoku. 3. 370, 378, 379, 380. ©. 371, 377. Kuma, Ehime Ken, Shikoku. 1200’. Specimens from Kiushiu may be regarded as typical talpoides, for that island is the first locality mentioned in Temminck’s original account, besides being that which contains Nagasaki, where the factory of the early Dutch traders was situated. A pair of well-grown Kiushiu specimens measure (in mm.) as follows :— ©. Head and body 99; tail 34; hind foot 16. ” cB 102; 9 34 ; ” 16. The Shikoku specimens appear to be quite similar to those from Kiushiu.