ednrumes BND AE ee we fp DY aa es e “ . - ert ee oat oe ME ES, Pree tin ehien ne ¥ p A re ~ nan aS : : DA LAP PA a he i ial ae i eeleonenits P : 7 ° r 4 He 8aey " 5 RANA age DD aaa a TAN 8 Mtg a se emote naan net he Adel bye Lot ND ¥ Ave me tt St A A Pe $04 anne Ei aaah oH Sveti in Gtr Be BRE SG RAT eel AON RE TT FBTR Eve OEE AEN MAE EN CEL my RAE $ret i hoote teds, tents : oe KR 5 Nahe od eal ir wr aaa! Lhe Sepeteaaatad ey Pies r ees meteg ners E fy eeeapenceietand DELI De KW wee, Weth oth te Sey apse edi einguty anti cd 1 Ea) eae ae Alan age Cate ile a8 otal oS « 7 Bi t.8 Av eh at AUN Lt un PROCEEDINGS OF THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. 1921, pp. 447-887, . with 18 Phares AND 184 TEX'r-FIGURES. . ie 2s 723lp PRINTED FOR THE SOCIETY, SOLD AT ITS HOUSE IN REGENT’S PARK. LONDON: MESSRS. LONGMANS, GREEN, AND CO, PATERNOSTER ROW, fT Sab OF THE COUNCIL AND OFFICERS OF THE ZOOLOGICAL SOCIETY OF LONDON. LO ale Patron. His Masgestry Tue Kive. COUNCIL. His Grace Tar Duke or Beprorp, K.G., F.R.S., President. AurreD H. Cocks, Esq., M.A. CaarLtes Drummonp, Ksq., Treasurer. ALFRED Hizra, Esq., O.B.H. THe Ricutr Hon. 'Tur Viscount Grey, K.G., P.C. Sir Srpney Ff. Harmer, K.B.E., M.A., Se.D., F.R.S., Vice- President. Pror. JAMES P. Hitu, D.S8c., H.R.S., Vice-President. WILLIAM HuNntsMAN, Esq. Mayor THe Lorp ALASTAIR Roper? Innes-KeEr. Pror. G Cove Ge ik Cali: Ernest W. MacBripe, D.Sce., F.R.S., Vice-President. Con. Sir Henry McManon, E. G. B. Mrapr-Watpo, Esq., Vice-President. P. CHatmers MircHe.., Esq., C.B.H., M.A., D.Se., LL.D., F.R.S., Secretary. THe Haru or Onstow, O.B.E. Masor Arperr Pam, O.B.E. THe Lorp QUEENBOROUGH. His Grace THe DvuKE oF Rurtanp, K.G. Masor Ricuarp 8. Taytor. A. ‘Treyor- Batryz, Esq., M.A. Antuony H. WINGFIELD, EsqQ., Vice-President. A. Smrra Woopwarp, Esq., LL.D., F.R.S., Vice-Pre- sident. PRINCIPAL OFFICERS. P. Cuatmers Mircuett, C.B.M., Secretary. M.A., D.8e., LL.D., F.B.S., R. I. Pocock, F.B.S., F.L.8., Corator of Mammals and Resident Superintendent of the Gardens. D. SETH -Smiru, Curator of Birds and Inspector of Works. Epwarp G. BouLencer, Curator of Reptiles. Miss L, EK. Cuzxsman, F.E.S., Curator of Insects. Dr. C. F. Sonntag, Ch.B ., Anatomist. Dr. N. 8. Lucas, M.B., Ch.B., Pathologist. R. J. Orriterp, M.A.. Hon. Parasitologist. FE. Martin Duncan, F.R.M. Si Librarian. F. W. Bonn, Aeoouatant Ny Jal Coxe, Chief Clerk. LIST OF CONTENTS. 1921, pp. 447--887. EXHIBITIONS AND NOTICES. Sir 8S. F. Harmer, K.B.E., F.R.S. Exhibition of, and remarks upon,a mounted specimen of a young female TERMMOCCTOS SOHAMLCWS Vi. ae icchsiawiah «23 vias ak isd sb ols oF eMRaLN Sa, Mr. Tuos. Gerrarp, F.Z.S. Exhibition of, and remarks upon, photographs of the skeleton of an old male - Kiva Gorilla (Gorilla gorilla beringeri Matschie) ...... Lord Roruscuitp, F.R.S., F.Z.8. Exhibition of, and remarks upon, a mounted specimen of a male Kivu Gorilla (Gorilla gorilla beringert Matschie) ............ Mr. D. Seru-Smiru, F.Z.8. Exhibition of, and remarks upon, a series of skins and photographs of polyandrous EES HEEL 153100) A) oA OI GNOME calc tneche nla d dn datin do meee « Dr. Caas. F. Sonntag, F.Z.S. Exhibition of, and remarks upon, some abnorinalities in the Carnivora ............ ‘Mr. L. T. Hoasey, F.Z.S., and Mr. E. A. Spauu. Exhibi- tion of, and remarks upon, effect of Pineal Gland administration on Frog Tadpoles ...............:cceeeee The Secretary. Report on the Additions to the Society’s Menagerie during the month of April, 1921............ Miss L. E. Coresman, F.E.S. Exhibition of, and remarks upon, lantern-slides of the larval stages of some Hydropaychides i.) Fay ote riess Gs MOUNT NSUa eiteta:t Sua:clores Major Srantey S. Fiower, O.B.E., F.Z.8. 9 Exhibition of, and remarks upon, examples of three species of Higy phat Serbeas erates linseed ces yas edie eee ees 643 643 644 644 644 644 G44 iv Major Sranuey 8. rower, O.B.K., F.Z.8. Exhibition of, and remarks upon, living specimens of Zestudo leithit FE 00 i M0) Y7 eC RE nee BRIS HB NCR AE ME ayetea oe Arce Ar AS Dr, P. Coatmers MircHett, C.B.E., F.R.S., F.Z.S. Exhi- bition of, and remarks upon, photographs of the Death-mask of a young Gonilla iy Webs seer ee e-ee Dr. F. M. Cuarman. An account of the distribution of Bird-life in the Urubamba Valley of Peru ............ Prof. J. P. Hiuu, F.R.S8., V.P.Z.8.- Exhibition of, and remarks upon, a series of lantern-slides of Marsupial CM DR YOS i. cate meri tle rere Seems me am arena The Secretary. Report on Additions to the Society’s Menagerie during the months of May, June, July, Amoust, andyseptembers VO 2N cana enmeancone eee Mr. M. A. C. Hinton. Exhibition of an embryo African Mle plants eee 2 eee eek ee iat re eche Hastie ae As My. G. E. Boutencer, F.Z.S. Exhibition of photographs of a Common Grass Snake swallowing a Viper......... j Mr. F. Martin Duncan, F.Z.S. Exhibition of Sauna graph films of Marine Animals ............. Lelio ies laa Ale Mr. D. Seru-Suira, F.Z.S. Exhibition of photographs of the Sacred Ibis nesting in the Society’s Gardens...... 645 646 883 885 Prof. G. Eiutor Sirs, F.R.S., F.Z.S. An account of the- - habits of Zarsius, illustrated by lantern-slides......... Mr. Encar P. Caance, M.A., M.B.O.U. Exhibition of cinematograph films illustrating the laying-habits of, the Cuckoo (Cuculus canorus) Sand the life of the Mioume, © uclcooye same maa rence cere Ris tice shen aa Renee The Secretary. Report on the Additions to the Society’s Menagerie during the ronth of October, tee seoueane: 886 886 886 Mr. A. S. Le Sovir, C.M.Z.S. Exhibition of sistent ; illustrating the early life-history of Ornithorhynchus. Dr. A. Smira Woovwarp, F.R.S., V.P. Z.S. Exhibition ob a Human Skull and other emia from Broken Hl, North Rhodesia .......... Pe ee aun DS cvs gisele moretmuntnn te an 887 26. 30. PAPERS. . The Geographical Distribution of Orthopterous Insects in the Caucasus and in Western Asia. By B. P. PLO astern Aud etn: et pete Meee atc tneis dye ecrna pro anshehe eich moe aa . The Auditory Bulla and other Cranial Characters in the Mustelide. By R. I. Pocock, F.R.S., F.Z8. (Memmi umes MAS Sy Baia toes. sacsecns foetsonedec Genrenals el . On the Venous System of the Lizard Varanus bengal- ensis (Daud.). By Goxsinp Sineu Tuapar, M.Sc. (WexGeireranes Ay) ae AP aN ja luau wart « dete e Abnormalities in the Common Frog (Lana temporaria). By J. H. Luoyp, M.Se., F.Z.8. (Text-figure 1.)...... . Comparative Anatomy of the Tongues of the Mam- malia,—IV. Families 3 and 4. Cebide and Hapalide. By C. F. Sonnrac, M.D., F.Z.8., Anatomist to the mociety: (Rexti Heures 002.) ticsicseutiesctbieecuotes ee . Note on the Skull of Dinotherium gigantewm in the British Museum. By C. W. Anprews, D.Sc., F.B.S., EA Sat ( Verb mommecml an) ne. x00). ee eee ae shidog cubed . On the External Characters of some Species of Lutrinz (Otters). By R. I. Pococn, F.R.S., F.Z.8. (Text- HVC AUN ES ILS AL teh bis «oan at aa ane ee Ree Se The Comparative Anatomy of the Koala (Phascolarectos cinereus) and Vulpine Phalanger (TVrichosurus vul- pecula). By Cuas. F. Sonnrac, M.D., F.Z.S. (Plates Vie Vp ext etgesy ot Obs) oak jundena same ease. sien Jase . New Indian Drilid Beetles. By S. Mauuix, F.Z8. (Rextfie ures: 1S. icici scccescsrdnssarscsevsveresenns rie, Page A87 493 497 BAT 579 Vi 32. On some Abnormalities in-the Carnivora. By CHARLES F. Sonntac, M.D., F.Z.S. (Text-figures 64-€5.)...... 38. The External Characters of the Koala (Phascolarctos). By R.1. Pocock, F.RS8., F.Z.5. (Text-figures 22-26.) 34, The Nestling Feathers of the Mallard, with Observa- tions on the Composition, Origin, ond History of Feathers. By J. Cossar Hwart, ML. 1D Althel eee Aree (Plates I.-X.; Text-figures 1-14.) NE 2 ss are aa 35. On some new Genera and Species of Anomodont Rep- tiles from the Karroo Beds of South Africa. By R. Broom, E.RS., C.M.Z.S8. (Text-figures 28-45.) . 36. The Cichlid Fishes of Lake Nyassa. By C. Tare Recan, M.A., F.R.S., F.Z.8., Keeper of Zoology, British Museum (Na tural History). (Plates I.-VI.; Text- fiomares = 30.) 00% > traene ofa oor epee ee cee eee eee eee ee 37. Note on the Systematic Position and Distribution of the Actinian Sagartia lucie. By J. PLAYFAIR McMurreicn, C.M.Z.8. (Text-figures 1-4.)............ 38. The Comparative Anatomy of the Tongues of the Mammalia.—V. Lemuroidea and Tarsioidea. By Cuartes F. Sonnvrac, M.D., F.Z.S., Anatomist to the Society. (Text-figures 66-69.) .......... eee: 39. The Comparative Anatomy of the Tongues of the Mammalia.—VI. Summary and Classification of the Tongues of the Primates. By Cuarves F. Sonnvac, M. D,, E\Z.8., Anatomastito the Sociebyaerseas-se: ne 40. On some new Parasitic Mites. By Sranney Hirst, F.Z.S. (Text-figures 16-48.).......-...54. sees eee 41. On the External Characters and Classification of the Mustelide. By R. I. Pocock, F.R.S., F.2Z.8. (Text- fipumes 2L—O,) oo vaisn steele ae ane stele dale ce/feitelle Angier «eee 42. Megalohyraz Andrews and Titanohyrax, g. n.—A Revision of the Genera of Hyracoids from the Fayim, Egypt. By H. Marsumoro. (Text-figures DG), ese, «creat leteareiaeis cee rele) Se Ula eateatc a eee BSE \ Page 587 59] 609 647 675 769 803 839 vil . Page 43. Contributions to the Visceral Anatomy and Myology of the Marsupialia. By Cuarues F. Sonnrac, M.D., F.Z.8., Anatomist to the Society. (Text-figures ETI a 5 sealed sense ce sie sinjeiaisinin Mr. L. T. Hoesen, F.Z.8., and Mr. E. A. Spavu. Exhibition of, and remarks upon, effect of Pineal Gland administration on Frog Tadpoles........++..-...-+s- SEAN stg ciot The Secretary. Report on the Additions to the Society’s Menagerie during the month of ZN yall ea ey See ee arena sacs See ACO Gh Ore oC OG an Sako au Saa0 oA du ed bic Miss L. E. Cunnsman, F.W.S. Exhibition of, and remarks upon, lantern-slides of the larval stages of some Hydropsychide .....+....0+..0+---0-e aiuiele ainia)ee ictelelolafertete Major Srantny 8. Frowsr, O.B.E., F.Z.8. Exhibition of, and-remarks upon, examples of three species of Hoyptiam Jerboas 5.56 0c. sje je secs nie cles oe ce sis eelsin = olelelels « Major Stanuny 8. Fuowme, O.B.E., F.Z.8. Exhibition of, and remarks upon, living specimens of eee leithii and T. ibera . 643 643 644 644 644 oa 64 ihe 645 645 Dr. P. Cuaumurs Mitcnnt, C.B.H., F.R.S., F.Z.8. Exhibition of, and roa upon, — _ photographs of the Death-mask of a young Gorilla ...2........eese cece ees gdoce Dr. Frank M. Cuarman. An account of the distribution of Bird-life in the Urubamba Walley,ot Bernie ater oeleietelecweveleis esate, a alent efielaia Sasha Eaerecioe Boot balao bers mon Prof. J. P. Hitt, F.R.S., V.P.Z.S. Exhibition of, and remarks upon, a series of lantern- slides of Marsupial embryos 2.0.02 0sc. cen wee ceeemscenew as ous Saiawie Vayeretars ee 645 645 Contents continued on page 3 of Wrapper. PROCEEDINGS OF TIE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF THE ZOOLOGICAL SOCIETY OF LONDON. PAPERS. 23. The Geographical Distribution of Orthopterous Insects in the Caucasus and in Western Asia. By Val Be * Uvarov *. (With Map.) [Received October 25, 1920; Read March 8, 1921.] Introduction. In 1913 I undertook the systematic study of the Orthopteran fauna of the Caucasus and neighbouring countries (Asia Minor, Armenia, Northern Persia), based on the large collections of the Caucasian Museum in Tiflis and literary sources, as well as upon my own investigations in Northern Caucasus (in 1911-1914) and in Transcaucasia (in 1915-1919). This work offered great difficulties, as our knowledge concerning the systematic and geographical distribution of the Orthoptera, and those of Western Asia in particular, is as yet very incomplete. Besides, the war deprived me of the possibility of referring to Western European scientific centres and of obtaining from them the necessary literature and information. The materials which I have had before me, though not very extensive, give different impressions concerning the faunas of different districts, and some of the latter are still awaiting more careful investigation. I think, however, that it would be of some use for these future investigations if I gave a short summary of the chief zoogeographical results I have gained from my studies, incomplete and insufficient though they * Communicated by S. A. Nreave, F.Z.S, Proc. Zoou. Soc.— i921, No. XXXI. 31 448 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION may be. I hope these results will be of interest to every z0o- geographist studying the distribution of animals in the southern parts of the Palearctic region. Before investigating more minutely the distribution of Orthoptera in the territory under consideration, it is necessary to give some information concerning the zoogeographical division of the Palearctic region based on my studies of this group, for this division differs in some respects from those of other authors founded on the distribution of other animal groups. The subregions into which the Palearctic region may be divided are four innumber: Boreal, Steppe, Mediterranean, and Eremian. The chief characteristics of these are as follows :— The Boreal subregion includes the whole zone of the forests of Northern, Europe and Asia, but some representatives of its characteristic Orthopteran fauna.penetrate farther on north- wards—into the zone of the Arctic tundras which has not its own Orthopteran fauna. The Orthopteran fauna of the Boreal sub- region is very poor, both in species and in specimens; the suborders Mantodea and Phasmodea are entirely absent from it, and of the Gryllodea we only find the Gryllus domesticus here. As leading characteristics of Boreal fauna should be regarded : Chrysochraon dispar, Stenobothrus viridulus, Gomphocerus sibiricus, Mecostethus grossus, Psophus stridulus, and Podisma pedestris from Acridiodea, and Leptophyes punctatissima, Meconema thalassinum, and Olynthoscelis griseoaptera from Locustodea. The great distance between, our country and the Boreal sub- region causes the Boreal fauna to be of very little importance in the composition of our fauna. Nevertheless, in some districts of the Caucasus, as we shall see later on, the influence of the Boreal fauna is rather striking: in some mountainous districts we may find the typical boreal species, which are, at the same time. absent from the intermediate areas between the Caucasus and the Boreal subregion. These species with such discontinuous range of distribution are of great importance to the history of the fauna of the Caucasian mountains, as they give us a hint con- -cerning the former contact of this fauna with that of the Boreal subregion ; later on this contact was interrupted, but the cause of this interruption is still unknown to us with certainty; we can only suppose that this immediate connection between the Orthopteran faunas of the Caucasus and of the Boreal subregion took place during the Glacial period, and ceased after this period had given place to a warmer and drier one, when the Boreal elements of the fauna retreated to the North, leaving a few relics in the high mountainous districts of the Caucasus. The Orthopteran fauna of the Steppe subregion is rather rich and includes many typical forms. The most important character of this fauna is the presence of a large number of species of the genus Stenobothrus. The Steppe fauna in Europe shows dis- tinctly marked affinities with the Siberian, and we should suppose that its representatives have migrated into Europe from Asia. OF ORTHOPTERA IN THE CAUGASUS AND WESTERN ASIA, 449 The Steppe fanna penetrates into our country through the steppes of North-western Caucasus, which belong to the Steppe subregion, and through Daghestan some of its elements reach Transcaucasia, as well as the mountainous districts of Caucasus Minor, Armenia, Kurdistan, and Anatolia as far as the northern limits of the Hremian subregion. On the other hand, the Steppe fauna also penetrates into Anatolia through the Balkan Peninsula. It is evident that the influence of this fauna on the fauna of our country must be very great, and so it is, as we shall see later on. The Mediterranean subregion comprises the north-western mountainous extremity of Africa (Morocco, Algeria, and Tunis— but the mountains only and not the plains, which belong to the Kremian subregion), Spain, the southern coast of France, Italy, the shores of the Adriatic Sea, the Balkan Peninsula south of the Balkan Mountains, the islands of the Mediterranean Sea, and Anatolia. The Mediterranean fauna of Orthoptera is extremely rich, including about 16 Mantedea, 7 Phasmodea, 162 Acridiodea, 317 Locustodea, and 62 Gryllodea. esides its richness this fauna is remarkable for a large number of peculiar species: out of 564 species which are known of it, 424 or 75 per cent. are endemic. Some large families of Locustodea have their centre of development and distribution here, for~example Sagide, Decticide, Bradyporide, and, partly, Phaneropteride. The more detailed study of the Mediterranean fauna allows us to conclude that it may be divided into two very distinct groups of species: Western and Eastern. The first has its centre in Spain and North-western Africa, from where its representatives disperse to the East and North-east; while the second flourishes in the southern part of the Balkan Peninsula and in adjacent parts of Anatolia, sending its migrants into north-western, northern, north-eastern (into the Steppe subregion), and eastern directions. According to this distinction of the faunas we may divide the Mediterranean subregion into two zoogeographical provinces—Western or Tyrrhenian and Kastern or Balkano- Anatolian. The Tyrrhenian fauna is of no importance to us, as it cannot influence the composition of the fauna of the Caucasus, but of much greater value is the Balkano-Anatolian fauna. This fauna is very rich and has many characteristic points: here we find exceedingly rich Gee us of endemic Sagide, of apterous Phaneropteridee (Jsophya, Pwcilimon), of the genera Platycleis, Olynthoscelis, Drymadusa, Dolichopoda; some species of Stenobothrus and the genera Wocarodes (with six species) and Oallimenus (with two species) are peculiar to it. The Balkano-Anatolian fauna occupies the greatest part of the country which we are now studying. Through Asia Minor, which belongs entirely to this zoogeographical province, through Armenia, and along the southern aa aes n shores ot the Black Sea, the Balkano- ‘Anatolian fauna penetrates into the forest districts of Transcaucasia, giving place to the Eremian fauna in 31* 450 MR. B. P. UVAROY ON THE GHOGRAPHICAL DISTRIBUTION the desert plains of Eastern Transcaucasia and in the dry rocky | districts of Persia; we meet it again on the southern shore of the . Caspian Sea—in the district of Talysh, which is remarkable for its wet subtropical climate. Some typical Balkano-Anatolian faunistic elements also reach the Caucasus from the north, wandering from their native home along the western and northern shores of the Black Sea through the adjacent steppes of South Russia. To the south of the Mediterranean subregion lies the vast: Eremian subregion, which includes all the deserts of North Africa (Sahara, Libyan desert, Egypt), Sinai peninsula, Northern Arabia, Mesopotamia, Persia, the whole of the Aralo-Caspian impression; perhaps, also, the great deserts of Chinese Turkestan and Mongolia, the Orthopteran fauna of which is but little investigated as yet, belong to it. The Orthopteran fauna of the Eremian subregion has many peculiarities if compared with the Mediterranean, and I cannot agree with most of the zoogeo- graphists * who usually unite them together. In support of my opinion I give the following table :— E = . a S . | s 3 3 = 2 Ss 2 re) 9 cS: g a S iz £ 1h 2 iS E SN Gea (oi oe q a S) 5 Ss Sy | = ay < 4q o a | | Mediterranean fauna. | | | Total number of species..................| 16 7 162 | 317 62 564 | | Number of species which do not pene- | | trate into the Hremian subregion ... 9 | 7 130 | 299 ol 496 | bind evmnncShscccets cae select etter eetate eects salami an Mead, 108 | 259 | 4 424, | Percentage of endemism.................: 56 % | 100 Jp) 67 8/9 | 2% | 66% | 75 Oo | ' Common with the Eremian subregion.| 7 | 0 2 leds 12 69 | | Eremian fauna. | Total number of species..................| 53 $) 1387 | 46 31 275 | | Number of species which do not pene- | | | trate into the Mediterranean sub- [Fig RO STOMRY hs ve Gomtece Crea hae mace eRney Ace 9 105 Be) \ ae) 206 IMIS TY GLETANI GC Srtess step Wyataen GAL ae en rma eee ean (aes 9 99 24 | 16 190 | Percentage of endemism ........ ......| 79 %/o | 100%) 72% | 63%) | 51% | 70% | It is evident from this table that the affinity between the Eremian and the Mediterranean faunas is restricted to 69 species only. Out of these we must, however, not take into consider- ation 48 very widely distributed species (such as Varatettix meridionalis, Acrotylus insubricus, Conocephalus nitidulus, etc.) and 10 species with great power of flying, which may have wandered from one subregion to another in recent times ; there are, there- fore, only 11 species common to the EKremian and Mediterranean * W. L. Sclater was the first who recognised the difference between the Eremian and his “Europasian” (Boreal+Steppe+ Mediterranean) subregions, but he included in it the whole of Northern Africa, the western part of which (Morocco, Algeria, and Tunis) must be united with the Mediterranean subregion. ———— OF ORTHOPYTERA IN THE CAUCASUS AND WESTERN ASIA, Abd] faunas, 7. e. but 4 per cent. of the second and less than 2 per cent. -of the first of these. These eleven species are as follows: Fischeria beetica, Ameles abjecta, Stauronotus hauensteini, Wdipoda schochi, Nocarodes serricollis, Callimenus dilatatus, Paradrymadusa sordida, Olynthocelis punctifrons, Isophya triangularis, Gryllus (Gryllodes) kerkennensis, and Gr. lateralis. Only two ot them (both species of Gryllus) may be considered as having originated in the Kremian subregion, and they are both to be found in the Mediterranean subregion (in Spain, in Transcaucasia) in prox~- imity to its southern boundaries and on spots with clearly defined desert soil and vegetation ; we have the right to believe them to be comparatively recent invaders from the deserts of the Eremian subregion. ‘Theremaining nine species are of Balkano-Anatolian origin and do not penetrate deeply into the Eremian subregion, being restricted to its northern mountainous parts with mixed fauna. The relationship of the Mediterranean and of the Eremian faunas is, consequently, practically absent. On the contrary, the same table shows us that the difference between them is a very striking one; 496 species (out of the whole number 564) of Mediterranean Orthoptera do not reach the Eremian subregion, and 206 Eremian species (out of 275) do not go through the northern boundary of this subregion into the Mediterranean. ‘This difference is not numerical only; t the family Orthoderide is peculiar to the Eremian subregion*, where there are 35 species belonging to it; the family Pamphagide is represented in the Mediterranean subregion by 52 species, of which only five penetrate into the neighbouring parts of the Eremian subregion; the family Phaneropteride has move than 80 Mediterranean representatives, and only three of them are to be found among the Eremian fauna; two families of Locustodea Ephippigeridz (85 species) and Meconematide (4 species)—and three of Gryllodea—Gryllomorphide (7 species), Myrmecophilide (4 species), and Mogisoplistide (6 species), which are very im- portant in characterising the Mediterranean fauna, do not extend into the Eremian subregion at all. The generic and specific differences between these two faunas are yet more considerable, but I shall not go into details here, as I suppose the above mentioned facts are sufficient to support my statement that the, Eremian subregion is of the same zoogeographical value as the Mediterranean ‘. There are only two provinces of the large Eremian subregion which are particularly interesting to us: the Ivanian and the * With but one exception—Geomantis larvoides—which is Mediterranean endemic. + I even suppose that, when studying the distribution of Orthoptera, we are right in considering the Eremian subregion of the same value as the whole Palearctic region: this problem i is, however, too great a one to be discussed here, and I hope to return to it at some other time ; I am supported in my supposition by the state- ments of Mr. A. Birula, who, after his studies of the distribution of scorpions, made an Africano- Asiatic region nearly with the same limits as my Eremian subregion (see A. A. Bialynicki-Birula, Ayvachnoidea Arthrogastra Caucasica, Pars I. Scorpiones.—Mémoires du Musée du Caucase, sér, A, N. 5, 1917). 452 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION Turanian. The first of these has some peculiarities in its fauna which are of the greatest interest and value: there are among the rather numerous. endemics of the Iranian fauna some very ancient forms (Parudrymadusa bocquilloni, P. persa, P. pastuchovt, Tropidauchen, Platycleis persica, Olynthoscelis satunini, etc.), which belong to the groups characteristic of the Balkano- Anatolian province of the Mediterranean subregion. ‘The presence of these species in the Iranian fauna allows us to conclude that this fauna was in some ancient time in close connection with the Balkano-Anatolian fauna, but afterwards -(from the beginning of the dry climatic period in Iranian table-lands) this connection was interrupted, and the further development of the Balkano- Anatolian and of the Iranian fauna went in different directions : the remnants of the ‘“‘ Ancient Mediterranean” fauna in Iran partially died out, partially adapted themselves to the new conditions of life (the ‘“‘ desert” coloration of Paradrymadusa bocquilloni and P. persa, etc.); and the recent faunas of the Iranian and the Balkano-Anatolian provinces, being of the same origin, are entirely different and belong to the different subregions. The Iranian fauna occupies a rather large part of Transcaucasia reaching along the western shore of the Caspian Sea as far as the neighbourhood of Petirovsk, as we shall see later on. The Turanian province of the Eremian subregion comprises: Transcaspia and the southern steppes of Kir ghizes and Kalmyks along the northern and north-western shores of the Caspian Sea; its Orthopteran fauna is a rather recent derivate of the Tvanian which has migrated in northern and north-western directions, invading the parts of land from which the Caspian Sea has recently withdrawn. West of the Caspian Sea (in the: Ciscaucasia) the Turanian fauna spreads southwards, coming im contact with direct Iranian migrants somewhere near Petrovsk. To the south of the Eremian subregion lies the Indo-Ethiopian region, the fauna of which is of some importance for the com- position of the fauna of the country we are studying now, where we may find a rather large number of species of undoubtedly Indo-Ethiopian origin, for example: genera Gelastorrhinus, Hierodula, Duronia, Pyrgomorpha, Sphodromerus, and species— Tropidopola cylindrica, Liogryllus bimaculatus, ete. All these Indo-Hthiopian elements came into the country under consider- ation through the Eremian subregion, of which fauna they are very characteristic. The Zoogeographical Divisions of the Caucasus and neighbouring countries. Before continuing the study of zoogeographical districts into: which the country in question may be divided, I ought to point out that all my conclusions are based exclusively on the study of the geographical distribution of Orthoptera, though I have also taken into consideration the conclusions drawn by the late K. A. Satunin from his study of Caucasian mammals; by Nikolsky— OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA. 453 reptiles and amphibians; Birula (/. c.)—scorpions ; as well as by other zoologists. A full list of the Orthoptera, which are known to me from the Caucasus and the neighbouring countries, will be given by me elsewhere. 1. South Russian Steppe district (R.M.)*. The Orthopteran fauna of the open grassy steppes lying to the east of Azoy Sea has not yet been sutliciently investigated, and we know but 69 species (3 M7 + 32 A 4+ 27L+4+7G) from it. Notwithstanding, it is evident that this fauna cannot be separated from that of the steppes north of the said sea and of the Black Sea—the steppes of Southern Russia, and it has nothing to do with the fauna of the Caucasus proper. In fact, there is only one species in this fauna which is unknown from the South Russian steppes; it is Pezotetiix giornai, a Mediterranean species which I think has quite recently penetrated here from the neighbouring Novorossiisk district, and is only restricted to the south-western part of the Azov steppes. We are right, therefore, in regarding the latter as simply being the southern part of the vast South Russian Steppe district. The boundaries of this district, as far as they are lying within the limits of country we are interested in, are not quite defined yet; the southern boundary of it coincides with the northern mit of the forests growing on northern slopes of the Caucasus; its eastern limit is not so sharply marked and is dependent on the westward progression of the Avalo-Caspian (Turanian) flora and fauna along the Valley of Manytsh (see below). 2. The Kuban-Terek district (K.-T.). As I have had the opportunity of studying this district for a rather long time (1911-1914, 7.e. four years), its fauna is well known to me~t. The number of species known from this district is77 (2M 4-33 A+33 L+9G) and may be regarded as being very nearly exact. The bulk of this fauna—68 species—is common to it and to the foregoing district, which leads us to the conclusion thatthe Kuban-Terek district ought to be regarded as belonging to the Steppe subregion. The distinction between the South- Russian fauna and the fauna of this district is based on 13 species. Out of these seven are of well-defined boreal origin: Stenobothrus nigromaculatus, St. ventralis, St. scalaris, Gomphocerus variegatus, Psophus stridulus, Leptophyes punctatissima, and Olynthoscelis griseoaptera: they form, no doubt, a rearguard of the relic group of boreal species which retreated from South-Russian steppes, * The letters after the name of each district are the initials of their latin names as adopted in my zoogeographical map. + M=Mantodea, P=Phasmodea, A=Acridiodea, L=Locustodea, G=Gryllodea. + See my publication concerning the Orthopteran fauna of the province Stavropol (Bull. du Musée du Caucase, 1915), 454 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION way YN eueysy b (3p °d 90s ‘uoeurydxa 10,7) PV : sijatg SINE] G 2) puesewo quaqueg WSAOsIaqP 41yaqueig Wwinsess fe q velpuisug uepoulseyeyy ad A0jSOY *[ eInsg-9X9 7, FEHOLO- RISEWYO URSUIEO doug? N »SIISSOUONOKY Bouuebey jodoysenas OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA. A455 after the end of the Ice age, southwards into the mountains of the Caucasus (see p. 448). Of the remaining six species, four are of Mediterranean (resp. Balkano-Anatolian) origin; these are: Pecilimon similis, Paradrymadusa beckeri (peculiar to the Kuban-Terek district, but belonging to a Balkano-Anatolian genus), Platycleis fusca, and Myrmecophila ochracea, which are to be regarded as immigrants from Transcaucasia. The ways of this migration, doubtless, do not lead across the chain of the Great Caucasus, but I think they go on the west along the shores of the Black Sea, and on the east through the Somkheto-Kakhetian and Daghestan districts (see below). The remaining two species are: firstly, Ganypsocleis schelkovnikove, which has just been described and the zoogeographical physiognomy of which is as yet uncertain; secondly, Vemobius tartarus, which has recently penetrated into the Kuban-Yerek district from the neighbouring deserts of the Caspian Ciscaucasia. Thus the Orthopteran fauna of the Kuban-Terek district may be characterised as derived from the South-Russian Steppe fauna, with well-marked indicaticns of southern influences—from Caucasus (forms of boreal origin) and from Anatolia (Baikano- Anatolian species). The Kuban-Terek district comprises all the northern slopes of the Caucasian mountains, including the adjacent hilly country with insular forests; its northern boundary coincides with the northern limit of these forests on the west and of the grass steppe on the east; while the southern is formed by the upper lmnit of the forests on the main chain of the Caucasus. 3. The Daghestan district (D.). The fauna of this interesting district has as yet been very little investigated: the whole number of species known from it is only 43 (2M +23 A+12L+6G), which is, certainly, not more than one third of the real number. Notwithstanding, a eareful analysis of this fauna allows us to draw some very interesting conclusions as to 1ts composition and origin. The Daghestanian Orthopteran fauna is in direct contact with the faunas of four districts: Kuban-Terek, Somkheto-Kakhetian, Caspian Transcaucasia, and the Eastern Caucasus. But we find the closest resemblances between our fauna in the two first named districts, and as they belong to distinct subregions (Steppe and Mediterranean respectively), it is an interesting problem to be solved—in which of them the Daghestan should be included. Out of 39 Daghestanian Orthoptera 34 are common to Daghestan and to the Kuban-Terek district, and only five are distinct, as follows: Vocarodes serricollis, Orphania scutata zacharovi, Locusta caudata caudata, Decticus verrucivorus verrucivorus, and Platycleis daghestanica. The last named species is peculiar to the Daghestan and of no interest to us; Vocarodes serricollis is not to be considered as being characteristic of 456 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION Daghestan, since it is not distributed all over this district but confined to its eastern parts; finally, Locusta caudata caudata, Decticus verrucivorus verrucivorus, and Orphania seutata zacharovi are the typical mountain forms and, doubtless, came to Daghestan from the adjacent mountainous district of Eastern Caucasus. It is evident, therefore, that there exists but very little difference between the Daghestanian fauna and that of the Kuban-Terek district, and that this difference is of a recent date and of an accidental origin. Onthe contrary, the resemblances between them are far deeper, for nearly all the chief characteristic steppe forms (such es Arcyptera flavicosta flavicosta, Celes variabilis variabilis, Stauronotus brevicollis) range into Daghestan, but not farther southwards where (in the Somkheto-Kakhetian district) they are either entirely absent or replaced by other subspecies. ‘Thus, the steppe Arcyptera flavicosta flavicosta gives place in the Somkheto- Kakhetian district to the distinct race 4. flavicosta transcaucasica ; Celes variabilis variabilis is represented there by the subspecies C. variabilis carbonaria and so on. But the most striking difference of the Daghestanian fauna from the Somkheto- Kakhetian (resp. Balkano-Anatolian and even Mediterranean) is in the negative features of the first: the numerous non-flying Phaneropteride, which are very characteristic of the Somkheto- Kakhetian fauna, are strange to the Daghestan, where only three of them exist: one Orphania, one Pecilimon, and Leptophyes albovitiata — all three not characteristic of the Somkheto - Kakhetian district ; the numerous endemies of the latter do not range into Daghestan at all. All the above-mentioned facts lead us to the conclusion that the recent Orthopteran fauna of the Daghestan is in more intimate relation to the fauna of the Kuban-Terek distvict than to that of the Somkheto-Kakhetian. I think, therefore, it should be right to regard Daghestan as an independent zoogeographical district of the Steppe subregion, characterised by the purely steppe Orthopteran fauna with but slight admixture of mountainous forms and of endemicsas well as of some ‘‘ancient-Mediterranean”’ species {see p. 452), like WVocarodes serricollis, or an undescribed species of Paradrymadusa, known to me from Daghestan only in females. I think that further investieations of this interesting district may clear up some details concerning the composition and origin of its fauna but will not change the views expressed above. Turning to the establishment of the boundaries of this district, we can only definitely state the north-western and south-western ones, which coincide with the lower limits of the alpine district of the Hastern Caucasus. As for the eastern boundary of the Daghestan it should be presumed to go along the extreme eastern chains nearly parallel to the Caspian Sea shore, leaving a narrow strip along the shore itself bearing quite a different Iranian fauna of the district of Caspian Transcaucasia. The most obscure are the boundaries between Daghestan and the districts of Kuban-Terek and of Somkheto-Kakhetia. ira ‘ OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA. 45 4, The Western Anatolian district (A.M.). There are known to us from this district, little investigated though it is, as many as 103 forms of Orthoptera (7M + 387A + 48 L.+ 11G), which gives evidence that its fauna is a very rich one. As for the composition of this fauna, it may be regarded as the purest expression of the Balkano-Anatolian fauna, which has here its original home. The most characteristic families of Locustodea for this fauna are the Decticide and the Phaneropteride : here we find 23 species belonging to the first named family, and 19 to the second. Among the Phaneropteridx the flightless species are 15 in number; 16 species of Decticide are also flightless. It is not surprising, therefore, that we find many endemies in this district : 13 species are peculiar to it. Among these endemics there are no fewer than 6 species of Pecilimon, all belonging to the group with non-denticulate cerci, which group presents one of the most characteristic features of the Balkano-Anatolian fauna having its centre of origin and distribution in this and in the neigh- bouring Armenian district. One species of sophya (J. paveli) i 1s also peculiar to Western Anatolia. Of Decticida two species of Platycleis (P. iruncata and P. scherert), two Olynthoscelis (Ol. signata and Ol. prasina), Drymadusa spectabilis, and Gampsocleis recticauda are also Western Anatolian endemics. Considering all these endemiecs, 1t 1s noticeable that their specific features are very well marked anduer y constant, which gives us the evidence that these species are ancient and undoubtedly autochthonous forms. Thus we come to the conelusion that the Locustodean fauna of the Western Anatolia bears some peculiar features and is of great age. The composition of the other suborders of this fauna is of a rather mixed and indeterminate character. As for the more recent elements of the Western Anatolian Orthopteran fauna, we may distinguish amongst them the forms of the Steppe fauna (Stenobothrus spp., Stauronotus brevicollis, ete.) which came here through the Balkan peninsula, and, what is more interesting, some species characteristic for the Western Mediterranean (Tyrrhenian) province, which are six in number, as follows: Geomantis larvoides, Acrotylus longipes, Paracaloptenus caloptenoides, Platycleis nigrosignata, Olynthoscelis chabrieri, and Anterastes serbicus. All of them, except Aciotylus longipes, are wingless and doubtless very ancient in their origin; the careful study of their distribution shows that it is discontinuous, which allows us to think that a connection between the Tyrrhenian and Balkano-Anatolian has been ancient also and ceased long ago. The limits of this district are not yet sufficiently known to us. It is certain, however, that the north-western limit does not coincide with the recent natural limit of the Anatolia—with the Bosphorus and the Sea of Marmora, but it is to be looked for somewhere in the Balkanian peninsula. The Western Anatolian district occupies, conclusively, the more southern part of the last 458 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION named peninsula, the western part of Anatolia from the Mediter- ranean Sea to the western borders of the interior Anatolian plateau, extending along the southern shore of the Black Sea as far eastwards as the neighbourhood of Trebizond. 5. The Armenian district (At.). The whole number of species known from this district is larger than that of any other, being 134 (8 M + 62 A+ 54 L+ 10G), which indicates the exceptional richness of its fauna. The analysis of this fauna shows its affinity with the fauna of the preceding district, since 62 species are in common with the latter; it is of interest that many species are peculiar to both these districts, being unknown beyond their limits. This affinity is certainly due to the fact that they both belong to the same (Balkano-Anatolian) province. Far more interesting is the difference between them: out of 134 Armenian Orthoptera no fewer than 74 do not penetrate into Western Anatolia. This group is composed of very different elements. First of all we can distinguish in it an admixture of the more northern (steppe and boreal) forms, which are 12 in number, as follows: Parapleurus alliaceus. Stenobothrus pulvinatus. Stenobothrus fischeri. . scalaris. e nigromaculatus. Gomphocerus sibiricus caucasicus. a apricarius. Arcyptera fusca. bd macrocerus, Locusta caudata kolenatii. ie hemorrhoidalis, Olynthoscelis grisea. The absence of these forms from Western Anatolia may indicate that they did not come to Armenia through the Balkanian peninsula but across the Caucasian isthmus. Some of them, for example Arcyptera and Gomphocerus, ave absent from Western Anatolia simply because in this latter district there are no places (high mountains) suitable for their habitation. The same cause explains to us why the following four alpine species: Orphania scutata zacharovi, Psorodonotus brunnert, Ps. fieberi, and Ps. specsularis, do not penetrate into Western Anatolia. A very characteristic group of the Armenian Orthoptera is formed by the 13 species belonging to the true Eremian fauna. Their list is as follows: Eremiaphila genei. Sphingonotus balteatus. Acrida robusta. Sphodromerus ccelosyriensis. Stenobothrus simplex. Platycleis escalerai. Stauronotus anatolicus. Medecticus assimilis. Pallasiella truchmana. Gryllus tartarus obscurus. Pyrgodera armata. >» hebraus. (Edaleus mlokosiewitchi. Since all these species are good flyers it is evident that they came into Armenia in recent times from the neighbouring deserts of Persia and Mesopotamia; this view is supported by : : OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA. 459 the fact that most of them are restricted to the eastern parts of the Armenian district. But the majority in the group of Armenian Orthoptera which do not reach Western Anatolia belong to species peculiar to Armenia (or to both Armenia and Syrian Anatolia). They are 23 in number, as follows : Stauronotus hauensteini kurda. Isophya rodsjankoi. *Cuculligera maculinervis. Lone poltoratsky1. *Pamphagus yersini. Saga cappadocica. % is brunnerianus. *Drymadusa curvicercis. *Kunothrotes derjugini. * 33 recticauda. Peecilimonella armeniaca. : % re konowi. Pecilimon tschorochensis. *Olynthoscelis annulipes. rs kutahiensis. * 3 signata. 3 syriacus. * ye zebra. a concinnus. on kurda. *Kurdia nesterovi. *Troglophilus escalerai. *Phonochorion satunini. No fewer than one half of them should be regarded as the relics of the ‘‘ancient Mediterranean” fauna (these are marked with an asterisk), to which also belong the following 13 species whose area of distribution extends also beyond Armenian limits, as they are to be found in the neighbouring districts of Aderbeidzhan, Caspian Transcaucasia, and Somkheto-Kakhetia : (Edipoda schochi schochi. Tmethis escherichi. Thalpomena ledereri. py) lnlliivat Heliopteryx humeratis. Nocarodes serricollis. Tmethis saussurel. Pezotettix rugulosa. » carinatus. Platycleis squamiptera. > CiSti. Paradrymadusa sordida. » bilobus. We find, thus, in Armenia 27 relics of the ‘‘ ancient Mediter- ranean ” fauna which find here their western limit of distribution and do not reach Western Anatolia. The remaining 7 species of Armenian Orthoptera which do not range farther westwards (Zmpusa pennicornis, Acrida nasuta, Isophya triangularis, [. acuminata, Olynthoscelis indistincta, Dolichopoda euxina, and Gryllus frontalis) are partly of indeter- minate zoogeographical value, or their absence trom Western Anatolia may be explained as a result of insufficient investi- gations. Summarizing the results of our analysis of the Armenian fauna we may conclude that it is a Balkano-Anatolian fauna in its chief characters but well distinguished from it by (1) the well expressed influence of the Eremian fauna, (2) the great number of endemics and relics of the “ancient Mediterranean ” fauna, and (3) some ad- mixture of boreal and steppe forms. 1 think the frontiers of the Armenian district should be drawn in the following manner. 460 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION The northern boundary coincides with the chain of Pontus, coming on the east very near to Batoum and embracing a narrow strip of the Adzharian chain; from here it runs southwards along the Arsian chain as far as Arax, where the contact of Armenian and Aderbaidzhan fauna takes place; farther on the boundary turns eastwards along the chain of Aghridagh as far as Ararat, from where it goes southwards along the watershed of the basins of Urmiah and Tigris. The western boundary is presented by the margin of the central plateau of Anatolia, and the southern one goes in its western part along the Cilician Taurus and Antitaurus, not yet being satisfactorily known farther eastwards owing to the lack of investigation. I suppose it coincides with the Armenian Taurus. As the territory of this district is far more extensive than that of any other, it causes us to suppose that it should be divided into two or more separate districts. Some modifications of the Orthopteran fauna of different parts of Armenia support this conclusion, but I cannot offer any satisfactory division, as our knowledge of the fauna of southern and south-eastern parts of Armenia is extremely limited. Further investigations of this district should be, therefore, of the greatest zoogeographical interest. 6. The Syrian Anatolia (A.8.) This district belongs to those which have been less investigated, as is evident from the small number of species known of it, this being only 106 (9M+50 A+38 L+9G), though its southern position and dry climate offer the best conditions for the develop- ment of the richest Orthopteran fauna. As for the composition of its fauna, it may be regarded as very closely related to the fauna of Syria proper, all differences being of an accidental nature. On the contrary, the difference existing between this fauna and that of Armenia is rather well defined in the lack of the most characteristic Armenian endemics and in the evident influence of the Hremian fauna which isa typical feature of the fauna of Syria. The boundary between this district and the Armenian one coincides with the southern limit of the latter district following the chain of the Cilician Taurus; along the Euphrat valley this district penetrates into Armenia, as is to be seen on the map. All the other boundaries of this district lie beyond the limits of the country which we are studying now and are entirely unknown as yet. 7. The Pontian district (P.). The fauna of this district includes 59 species of Orthoptera (1M+27 A+20 L4+11G); having been well investigated it cannot be considered very rich. The analysis of the Pontian fauna shows us that it is very closely related to the fauna of Western Anatolia, including only OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA. A461 15 species which do not range into the latter district. Of these 15 species no fewer than three or four (Gryllus frontalis, Steno- bothrus macrocerus, Arachnocephalus vestitus and, perhaps, Doli- chopoda ewxina) should be regarded as not having been found as yet in Western Anatolia owing to the lack of investigations only ; one—Wdaleus mlokosiewitchi, being a very strong flyer, doubtless came to the Pontian district from the east in recent times; the presence of dipoda schochi schochi may be satisfactorily explained by the influence of the neighbourmg Armenia; and, finally, three species are peculiar to the Pontian district—Podisma koenigi (Pontian endemic ranging also into the adjacent western portion of the Somkheto-Kakhetian district), P. satunini, and Olynthoscelis kerketa. The remaining six species are of great interest’: three of them are definitely boreal in their origin and inhabit the alpine district of Western Caucasus, from whence they come into the Pontian district; this descendance of the representatives of alpine fauna to the sea-level is due to the great humidity of the Pontian climate, this fact being very characteristic for the fauna of this district. The last three species are Pecilimon schmidti, Isophya pyrencea, and Olynthoscelis fallax all northern Baikanian in their origin and sylvan in their habitation, which leads us to the conclusion that they come to the Pontian district from the north—through the Crimea. Thus, we may consider the Pontian fauna as an unpoverished Balkano-Anatolian one, with the admixture of peculiar and boreal forms and species of northern Balkanian origin which came here from the north. The eastern limit of the Pontian district is formed by the chain of Suram; the southern boundary goes along the Adzharo- Imeretian chain, approaching the Black Sea near Kobulety and turning from here westwards along the chain of Pontus; the western boundary is rather obscure and is to be looked for some- where near Trebizond; the north-eastern boundary coincides with the upper limit of the forests on the Western Caucasus; and the north-western separating the Pontian district from that of Novorossiisk is indefinite, as we shall see later on. 8. Zhe Novorossiisk district (N.). This district is one of the less investigated ones, the whole number of Orthoptera known from it being 50(4M-+21 A+ 18 L+7G). Analysing its fauna we observe the very close resemblance of it to that of the South Russian steppe fauna and, on the other hand, to that of the Pontian district. This intermediate character of the Novorossiisk fauna is to be explained by the geographical position of the Novorossiisk district between the steppe of South Russia and the district of Pontus. The differences of the Novorossiisk fauna from that of the South Russian steppes are expressed in seven species, three of them being boreal in their origin (Stenobothrus scalaris, Psophus siridulus, and Olynthoscelis 462 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION griseouptera) and penetrating hence from the mountains of Western Caucasus; one (Platycleis sepiunr) is a Balkano-Anatolian species coming from the Pontian district ; and three remaining ones (Parameles taurica, Olynthoscelis pontica, and Pezotettix giornai) are of special interest. The first two of them are peculiar to the southern part of the Crimea, and the Pezotettix is a characteristic Mediterranean species, unknown as yet in the Crimea, though doubtless present there. The presence in the Novorossiisk district of these three species, which are absent from all other districts of the Caucasus, indicates that this district was once in a direct connection with the south of the Crimean peninsula, the time and place of this connection being at present unknown to us. As to the boundaries of the Novorossisk district, its inter- mediate position and the transitional character of its fauna render them very obscure; I think they are not very markedly defined ; its southern boundary separating it from the Pontian district is to be looked for somewhere between 'Tuapse and Sotchi. 9. The Somkheto-Kakhetian district (S.-K.). The Orthopteran fauna of this extensive district, though fairl well investigated, is not yet fully known, as is evidenced by the fact of the recent description of some new species and subspecies. inhabiting it. The whole number of the known Somkheto- Kakhetian Orthoptera reaches 79(7 M+ 35 A + 241 4 138 G), which should be less than the real number by some 10-12 forms. The first problem to be solved is whether this district belongs to the Steppe or to the Balkano-Anatolian province. Let us consider its affinities to both of them. This district hasin common with the Steppe province (districts of Southern Russia and of Kuban-Terek) 69 species, the difference being 21. Temporarily setting aside eight Somkhete-Kakhetian endemics, the remaining 13 are as follows: *Empusa pennicornis. *Tsophya pyrenza. *Tettix depressa. * =, amplipennis. *Paratettix meridionalis. +) ee acumimahas *Thalpomena ledereri. *Paradrymadusa sordida. *Pcilimon distinctus. Nemobius heydeni. * 5 bosphoricus. Gryllus lateralis. *Tsophya adelungi. No fewer than eleven of these (marked with an asterisk) are Balkano-Anatolian species or belonging to the characteristic Balkano-Anatolian genera, and two Gryllids only are recent in- vaders from the adjacent deserts of the Caspian Transcaucasia. Thus it is evident that the difference between the Somkheto- Kakhetian and the Steppe fauna is very well expressed and in- dicates the entirely different sources of their origin. Turning to the relation of the Somkheto-Kakhetian fauna to the Balkano-Anatolian one we see, that only 16 species inhabiting OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA, 463 this district are foreign to other districts of the Balkano- Anatolian province, eight of them being peculiar Somkheto- Kakhetian forms. The remaining eight species are: Stenobothrus lineatus. Platycleis vittata. Pecilimon distinctus, Olynthoscelis fallax. Tsophya adelungi. Nemobius tartarus. » pyrenza,. Gryllus lateralis. Out of these only Stenobothrus lineatus and Platycleis vittata may be regarded as proofs of the influence of the steppe fauna; two Gryllids are of desert origin; and all others belong to the Balkano-Anatolian genera. The above is sufficient to enable us to come to the conclusion that this district may be regarded as a part of the Balkano- Anatolian province, a conclusion confirmed by the study of the Somkheto-Kakhetian endemics. They are eight in number, as follows : Arcyptera’‘flavicosta transcaucasica. Celes variabilis carbonaria. Tmethis zaitzevi. Nocarodes rimansone. Podisma koenigi (peculiar to this and Pontian district). Tsophya bivittata. Leptophyes nigrovittata. Olynthoscelis distincta. All these peculiar forms, with the exception of Arcyptera, Podisma, and Celes, belong to the Balkano-Anatolian genera, and are doubtless not recent invaders. This strongly supports my opinion as to the affinities of the Somkheto-Kakhetian district. Podisma koeinigi presents an evidence of the influence of the Pontian fauna, and two peculiar characteristic steppe species, Arcyptera flavicosta and Celes variabilis (as well as Stenobothrus lineatus and Platycleis vittata), are doubtless immigrants from the Kuban-Terek district (through the Daghestan), the somewhat different natural conditions of Transcaucasia causing the sub- specific differences between the Transcaucasian and the primary steppe forms. We ought, therefore, to consider the Somkheto-Kakhetia as a district of the Balkano-Anatolian province, bearing in its fauna some hints of an influence of the South Russian steppe fauna, migrating from the north around the eastern end of the Caucasus through the Daghestan. The outlines of the Somkheto-Kakhetian district are very complicated and circuitous. I include in it the southern forest-clad slopes and hills of the Great Caucasus from Svanetia on the west to the south-eastern extremity of this chain: here the district branches around this end on the northern s!opes, thus coming in contact with the Daghestanian district; through the Suram meridional chain the northern half of the Somkheto- Proc. Zoon, Soc,—1921, No, XXXII. 32 464 MR. BP. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION Kakhetian district is connected with the southern part, the district thus ranging all over the northern slopes of Minor Caucasus as far eastwards as Karabagh, where its fauna, becoming gradually poorer, comes in contact with the fauna of Aderbaidzhan. 10. Zhe Talysh district (T.). The fact that we only know 45 species of Orthoptera (3 M+ 24A+8L+10G) from this district is certainly due not to its poverty but to incomplete investigations. As regards the composition of the Talysh fauna, it gives the impression of bearing resemblance to that of the adjacent Caspian Transcaucasia, being distinguished from it by the presence of - five forms only, as follows: Acrida turrita turrita. Parapleurus alliaceus. Stenobothrus macrocerus. Epacromia strepens strepens. Platycleis capitata. The latter of these is an endemic species, while the presence of the four remaining ones indicates the close affinity of the Talysh fauna to that of the Balkano-Anatolian province. The most remarkable fact is that dAcrida turrita and Hpacromia strepens are represented in the Talysh district, not by the desert sub- species inhabiting the Caspian Transcaucasia and Aderbaidzhan, but by the same races that are met with in the districts helong- ing to the Balkano-Anatolian province, the range of these sub- species being discontinuous ‘The affinity of the Talysh fauna with the Balkano-Anatolian one is even more defined by the fact that its difference from the latter can be based upon a single subspecies (leaving the endemic Platycleis capitata aside)— Decticus verrucsivorus boldyrevi, which no doubt came here recently from the neighbouring deserts, The influence of the Eremian fauna on the fauna of Talysh is, generally speaking, very well marked, resulting in the presence of such forms as Thisoecetrus dorsatus, Platycleis escalerai, Liogryllus bimaculatus, ete., but it should be regarded as of secondary nature. The immediate connection of the Talysh with the other districts of the Balkano-Anatolian province is now absent, but it no doubt existed formerly ; I think it should be looked for in the south-eastern (Karabaghian) branch of the Somkheto-Kakhetian district which formerly used to reach the Talysh. In the district of Talysh I include only the rather narrow strip along the southern shore of the Caspian Sea, the southern bound- ary of this district being the upper limit of the forests on the northern slopes of the Talysh mountains. The north-western boundary delimitating Talysh from the adjoining deserts of Caspian Transcaucasia is rather indefinite; the north-eastern one is completely unknown, OF ORTHOPTERA IN THE CAUCASUS AND WESTERN ASIA. 465 ll. The district.of Aderbaidzhan (Ad.). This district possesses a very rich Orthopteran fauna, the number of species known being 125(9M -+1P+76A 4 24L + 15G); the real number should be considerably larger, not less, I believe, than 150-160. The most interesting features of this fauna are as follows. First of all, the presence of a representative of the suborder Phasmodea ( Gratidia bituberculata)—which is not to be met with in any other district, except Caspian Transcaucasia—clearly indicates that the Aderbaidzhan fauna belongs to a quite distinct zoogeographical division. This is supported by the large number of Mantodea, Acridiodea, and Gryllodea, while the number of Locustodea is comparatively small. Among the Acridiodea the large number of species belonging to the family Gidipodide is very conspicuous; there are thirty Cidipodids here against ten, for example, inhabiting the neighbouring Somkheto-Kakhetian district. But the fauna of Aderbaidzhan is as remarkable for what it lacks as for what it possesses. Out of these negative features the total absence of the genera Chrysochraon, Gompho- cerus, Arcyptera, Psophus, Celes, and Podisma is very remarkable ; they are all of northern (boreal or steppe) origin, and do not reach this district. Yet more interesting is the composition of the fauna of Locustodea: out of the whole family Phaneropteride, so well represented in the districts belonging to the Balkano- Anatolian province, we only find two here—Phaneroptera falcata and Tylopsis thynufolia, both very strong flyers and doubtless recent immigrants; the highly characteristic for the Balkano- Anatolian fauna family Sagide is represented in Aderbaidzhan by one species only, the most widely distributed Saga ephippi- gera; the majority of Locustodean fauna being thus formed by the Decticide, which are sixteen in number, mostly species of distinct ‘ancient Mediterranean” origin. The originality of the Aderbaidzhan fauna is most clearly demonstrated by the large number of peculiar species (some of them also ranging into Caspian Transcaucasia); nearly one third of them are not to be found in any other district of the country, being distributed beyond its limits, while eighteen are true endemics, as follows: Eremiaphila persica. Derocorys roseipennis lazurescens. Brunnerella mirabilis. Drymadusa grisea. Scintharista brunneri. Paradrymadusa pastuchovi. Thalpomena persa. =) satunini. Helioscirtus moseri tichomirovi. ae persa. Tmethis persa. cs longipes. ecaninanliss % boecquilloni. Nocarodes woronowi. Platycleis persica. schelkovnikovi. Olynthoscelis satunini. 33 Amongst these endemics a great percentage of the ‘ancient Mediterranean ” forms is evident, as, for example, all Paradry- madusa species, Drymadusa grisea, two species of MVocarodes, etc, 32* 466 MR. B. P. UVAROV ON THE GEOGRAPHICAL DISTRIBUTION Very characteristic of the Aderbaidzhan fauna are also nu- merous Hremian genera and species, for example: Hremiaphila, Oxythespis, LHelioscirtus, Sphingonotus, Derocorys, Sphodro- merus, ete. Summarising the results of our analysis, we may characterise the fauna of the Aderbaidzhan district as a true EKremian one, with a well-marked admixture of ‘‘ancient Mediterranean” forms and with very slight indications of the steppe fauna, which pene- trates here through the Somkheto-Kakhetian and Armenian districts. T include in the Aderbaidzhan district the Persian province bearing this name (but not the recently formed republic of Azerbaidzhan in Transcaucasia), as well as the valley of middle Arax from Kaghyzman as far as Migry. Thus, the northern boundary is to be drawn along the southern slopes of the Minor Caucasus at a somewhat considerable height, about 6000-7000 feet above the sea-level. ‘To this district also belongs (I am not yet sure whether partly or wholly) the Karabagh, where the Ader- baidzhan fauna comes in direct contact with the here already impoverished Somkheto-Kakhetian one. Farther eastwards the northern boundary of Aderbaidzhan goes along the eastern boundary of the Zangezur chain and near Migry ¢ goes over the Arax, turning eastwards nearly parallel with the latter river along the chain of Karadagh as far as the Talysh chain; farther on the boundary of Aderbaidzhan coincides with the latter chain. The western boundary is the same as the eastern limit of Armenia, which we have already considered above. ) 0h. 0. 7.012... Flower (3). (4) Three papille in a triangle (in 5 cases) ... Mtinch (5). THe WHITE-THROATED CapucHIN (C. hypoleucos). Tuckerman (9) described the macroscopic and microscopic appearances, and recorded three papillz in the form of a triangle. 502 DR. C. F. SONNTAG ON THE ANATOMY THe Hornep Carucuin (C. apella). F, J. C. Mayer (4) pointed out that the tongue has four vallate. papillz, and nine sulci in the lateral organs. Podwisotsky (7) described six papille arranged in the form ofa V. He also described the lingual glands which are well- developed. Boulart and Pilliet (10), Cuvier (2), and Minch (5) state that the triangular pattern is the usual arrangement of the vallate papille in the genus Cebus. It is present in the majority of my specimens, but the V-arrangement and double-pair type appear in the remainder. Genus ATELEs. THe GRizzLED Sprper Monkey (A. grisescens). Measurements :—total length 3°6 em.; length of the oral part ' 25 em.; length of the pharyngeal part 1-1 cm.; width between the lingual attachments of the palato-glossal folds 1-7 em. Pigmentation.—The apex and a small area of the dorsum behind it are unpigmented. Behind the latter the dorsum is yellowish- brown, with a white cross in the centre. The long limb of the cross is continuous with the colourless area behind the apex, and each of the short limbs lodges a vallate papilla (text-fig. 37). The frenal lamella is colourless, but the rest of the inferior surface is yellowish-brown. The Circumvallate Papille (text-fig. 41 A).—Three large vallate papille form an equilateral triangle with the apex behind. The left basal papilla is compound, and consists of two elements ; the right basal papilla is oval and excavated; and the ,posterior papilla is circular. The fosse are well-marked, and the vallums form clear zones. A fungiform papilla occupies the centre of the vallate area. The Fungiform Papille (text-fig. 41 A) form a dorsal bounding zone on which they have the usual arfangement. Those lying beneath the apex are large and closely aggregated, but the transverse rows extend far forwards on the dorsum. The Conical Papille (text-fig. 41 A) have the usual arrange- ment in clusters and rows; they increase in size in the usual manner, and their points run in the usual directions. On the pharyngeal part of the dorsum they ave small and irregular. On the lateral borders and inferior surface they have the usual arrangements. Ail the papille are small or of medium size. The Lateral Organs (text-fig. 41 C) extend from the dorsum to the lower parts of the edges of the tongue. Hach consists of elongated fusiform lamine running forwards and upwards, and the separating sulci are narrow. Each organ is surrounded by a diamond-shaped unpigmented area. The right organ is ‘8 em. long, and has 5 lamine and 6 sulci. The left organ is °7 cm OF THE TONGUES OF THE MAMMALIA, 503 long, and has 4 Jamine and 5 sulci. The dotted line in text- fig. 41 C represents the margin of the unpigmented area round the lateral organs. Lymphoid Tissue and Glands—The base of the tongue is delimited by a zone of large nodules with a concave anterior border. No duct orifices are visible even under a lens. The Frenal Lamella (text-fig. 41 B, a)—The large, flat, triangular lamella covers a large area of the inferior surface of the tongue, and is devoid of a ridge on its upper surface. The apex is divided into small sharp processes, and the plain edges extend back to the middle of the lateral organs. The Median Ventral Sulcews (text-fig. 41 B, b) lodges a well- marked crest. It opens posteriorly into a flat, shallow, unpig- mented triangular area. Text-figure 41. The tongue of Ateles grisescens ; ¢.t. tonsils. The frenwm is very short. ; The ventral papillary border is narrow, and its papille have already been described. Tur Buack-FAceD SpipER Monkey (4. ater). Tuckerman (9) described the tongue, and the following description is supplementary to his account. The measwrements of the larger of my two specimens are :—total length 4:9 cm.; length of the oral part 4:3 cm.; length of pharyngeal part °6 cm. ; width between the lingual attachments of the palato-glossal folds 2 cm. The Circwmvallate Papille (text-fig. 42 A & C).—Four large papille form a double-pair, and two small papille lie on the. 504 DR. C. F. SONNTAG ON THE ANATOMY lateral sides of the resulting rhomboid. Tuckerman states that there are three papillz within the vallate area, but believes that they are intermediate between the fungiform and vallate groups. All the papille are circular, the fosse are well-marked, and the vallums appear as clear zones. They vary in prominence, but the posterior papille (C, 5) project to a greater extent than the others. All are smooth and glistening, Only conical papille are seen within the vallate area. The papille of the anterior pair are level with the most anterior Jamine and sulci of the lateral organs. They may be excavated (text-fig. 42 C, a), Text-figure 42. Ye Dd. @.ve The tongue of Ateles ater. The Fungiform Papille (text-fig. 42 A & B), which are larger than in A. grisescens, form a narrow dorsal bounding zone on which they have the usual arrangement. They are numerous, but small. On the inferior surface they are large; round the under surface of the apex they are closely aggregated, but farther ‘back they are discrete (text-fig. 42 B). Tuckerman showed how the fungiform papille on the under surface of the apex are richly supplied with taste-buds. The Conical Papille (text-fig. 42 A) are all very small, those at the posterior part of the tongue, which are the largest, being OF THE TONGUES OF THE MAMMALIA. 505 just visible to the naked eye. They have the usual arrangement, and their distribution according to size takes the usual form. Their characters are shown in text-fig. 42 D. They resemble the papille of A. grisescens in that those within and behind the vallate area are not regularly arranged in oblique chains. Lymphoid Tissue and Glands.—The base of the tongue is delimited by a zone of prominent elevations, but no orifices of glands are visible. These nodules have not been mentioned by Tuckerman, who merely states that the base of the tongue is wrinkled. He describes the mucous glands, which are arranged in a small central and two lateral groups. The Lateral Organs (text-fig. 42 K).—The prominent lateral organs form crescentic bodies on the sides of the tongue, the concavities facing the lateral vallate papille, which are level with the anterior laminze and sulci. ‘The lamine and sulci are numerous, and are irregular in position for they have been ac- commodated withina small space. The right organ is ‘9 cm. long, and has 9 lamine and 10 sulci. The left organ is 1 cm. long, and has 10 lamine and 11 sulci. Many of the lamine have secondary sulci, and some of the primary sulci are very wide. Tuckerman described the histology of the lateral organs. The Frenal Lameila (text-fig. 42 B)—The triangular lamella forms a stout, thick plate with a notched apex and nodular sides. Tuckerman described a lamella, with a fimbriated edge, which possessed sensory (tactile?) end organs on its non-medullated nerves, which form a network. Gegenbaur (11) criticises Tucker- mau’s observations and thinks that the end organs are tactile in function. The long, narrow, deep ventral mesial sulcus opens posteriorly into a small deep, triangular fossa. It lodges a small, median, triangular crest derived from the short frenum. THe BLACK-HANDED SpipER Monkey (4. melanochir). Flower (3) describes the tongue as follows:—‘‘ The tongue is rather lancet-shaped and pointed at the apex. The lateral group of linear follicles at the base of the palato-glossal fold is very distinct and large. The papille generally are soft and small. Of cireumvallate papille there are only two, situated side by side, close to the middle line opposite the attachment of the palato- glossal folds. The frenum is attached very near (three-tenths of an inch) to the apex of the tongue, which has a stong median groove below. The sublingual papilla is rather small, soft, thick and pointed, with a slightly bifid apex.” THe Rep-raceD Sprper Monkey (4A, paniscus). Mayer (4) described three vallate papilla forming a triangle. Miinch (5) described the vallate papille of two specimens of A. vellerosus. In one there were seven papille on each limb, but 506 DR. CG. F. SONNTAG ON THE ANATOMY the central papilla was absent. The central papilla was present in his second example. Genus LAGoTHRIX. THE Smoky Wooutty Monkey (L. infumatus). IT examined three specimens, and the following measurements refer to the largest:—total length 4°6 em.; length of the oral Yext-figure 43. 1 C. The tongue of Lagothrix infumatus. part 3°8 cm.; length of the pharyngeal part °8 em.; width between the lingual attachments of the palato-glossal folds 1-9 em. Pigmentation (text-fig. 37).—The three specimens are bluish- black in colour, and the colourless fungiform and conical papille, frenal lamella, and mesial part of the inferior surface stand out prominently from the pigmental parts. The shape is variable in resh specimens, being conical or oval. OF THE TONGUES OF THE MAMMALIA. 5O7 The Circumvallate Papille (text-figs. 43 & 44 A & B).—Specimen No. 1 (text-fig. 43 A).—Five papille form a V with long arms. The two anterior papille are of medium size, the mesial papillee are small, and the posterior papilla is large. All are prominent, smooth and colourless, the fosse are well-marked, and the vallums appear as clear zones. The apex of the V is far from the epiglottis, and the anterior papille are close to the edges of the tongue. Text-figure 44. Om eo @® A B. LEX ADS (SI (Ss (a DOCS NG Wale / 4 9 ‘i Sa OV Gir at(D0o-o OMe E. iy Dh dees 12 Ma AD G. The tongue of Lagothria infumatus. Within the vallate area there are many conical papille, and the posterior rows of fungiform papille enter the mouth of the V. Specimen No. 2 (text-fig. 44 A).—Three papille form an isosceles triangle whose equal sides are long (2 cm.), and the posterior papilla is a considerable distance from the epiglottis. The anterior papille are small and circular, the fosse are well- marked, and the vallums are divided into lobes by fissures from the fosse. The posterior papilla is oval and surrounded by a prominent furrowed vallum. 508 DR, C. F. SONNTAG ON THE ANATOMY Specimen No. 3 (text-fig. 44 B).—Four vallate papille are arranged in the form of a Y, and all have well-marked fosse and clear vallums. I do not consider that the Y-form in this case corresponds to that of the Simiid, but I believe that the papille of the mesial limb are members of a double-pair put into line with the long axis of the tongue. The Fungiform Papille (text-figs. 43 A and 44 C).-— The papillae, which are of small and medium sizes, form a dorsal bounding zone, but the transverse rows extend as far as the apex, thereby eliminating an apical cluster. The rows contain many papillee. On the lateral borders and inferior surface the papille are small, discrete, and arranged in single chains. All are glistening and hemispherical, and are closely sur- rounded by shaggy conical papille. The Conical Papille (text-figs. 43. A and 44D) are closely- aggregated and increase in size in the usual manner, but have no definite arrangement on the oral part of the dorsum. They are arranged in chains on the pharyngeal part of the dorsum and on the inferior surface. All are filiform, and stand up round the fungiform papille when the tongue is dried. Glands and L ymphord Tissue,—The baseot the tongue Is occupied by a large number of nodules, but no orifices of ‘ducts and pits are visible. The nodules, however, are not aggregated together on a crescentic zone in front of the epiglottis. The submaxillary and sublingual glands are enormous (text- fig. 43 B), and play the greatest part in the lubrication and softening of the food in the buccal area, The Lateral Or ‘gans.—Specimen No. 1 (text-fig. 44 E’—The lamine and sulei vary greatly in size, and form irregular rows on the lateral borders. The spacing between them is also very variable, for the lamine are closely-packed posteriorly, and discrete anteriorly. Many of the lamine are oval and traversed by wide secondary sulci. The right organ is 1 em. long, and has 13 lamine and 14 sulci. ‘The left organ is ae em. long, and has 12 lamine and 13 sulci. Specimen No. 2 (text-fig. 44 F).—Both organs consist of long, pointed laminze running upwards and on wards towards the lateral vallate papillee (1. v.P), which are level with the anterior lamine. The right organ is 1:1 cm. long,and has 11 lamine and 12 sulci. The left organ is 1-1 cm. long, and has 12 lamine and 13 sulci. Specimen No. 3 (text-fig. 44 G)—The lamine and sulci all run from behind forwards and upwards, and their white colour makes them stand out prominently against the bluish-black background of the tongue. The anterior amine are level with the antero-lateral vallate papille, and the posterior laminz are level with the antero-mesial vallate papilla. The right organ is 1 em. long, and has 12 lamine and 13 sulci. The left organ is 1:1 em. long, and has 11 lamine and 12 sulci. OF THE TONGUES OF THE MAMMALIA. 509 The Frenal Lamella (text-fig. 43 Ba, Ca, & Da).—tThe tri- angular lamella may have an entire or bifid apex. In the latter the halves may be long and narrow, or wide and short. The edges are plain or have processes. The Ventral Papillary Border (text-fig. 43 Bb, Cb, & D b) is wider than in Cebus and Ateles. Its conical papille are in hori- zontal rows, but the fungiform papille form a single chain. The Ventral Mesial Sulcus (text-fig. 43 Be, Cc, De) may be narrow and deep, or wide and shallow. It has no mesial crest. Humporpr’s Wootty Monkey (ZL. humboldti), The tongue has three vallate papilie forming an isosceles triangle; of these the apical one is large. The fungiform papillz are large and conspicuous, and have the usual arrange- ment in clusters and rows. The triangular frenal lamella has a bifid apex. Genus SAImMIRIS (=CuHRYSOTHRIX), THE SqurrreL Monkey (S. sciwrea). Measurements :—total length 2:1 em.; length of the oral part 1-9 cm.; length of the pharyngeal part °4 em.; width between the lingual attachments of the palato-glossal fold 1:1 em. The Circumvallate Papille.—Three papille form an isosceles triangle with the apex behind. The right basal papilla is strongly retracted within its fossa, but the others are prominent, granular, and surrounded by nodulated vallums. The Conical and Fungiform Papille have the usual arrange- ment, and the latter form a dorsal bounding zone. Glands and Lymphoid Tissue ave present as usual on the base of the tomgue, but no duct orifices are visible. Both lateral organs are 5 cm. long, and have five primary sulci separating four lamine; the latter are traversed by secondary sulci. Each organ appears asa straight line of lamine and sulci. The frenal lamella, as shown by Pocock, is long, narrow, and has a slightly divided apex. The frenwm is long. Genus CALLICEBUS. THe Moxocw Treree (C. moloch), Three specimens were examined, and the measurements were very similar in all. The following figures refer to my largest example :—total length 2°55 cm.; length of the oral part 2°15 em.; length of the pharyngeal part -4 cm.; width between the lingual attachments of the palato-glossal folds 1-2 em. The tongue is bluish-black in colour on the entire oral part and on most of the pharyngeal part of the dorsum. The inferior surface has a bluish-black bounding zone. The frenal lamella, mesial part of the inferior surface, and cuneiform areas on the 510 DR. C. F. SONNTAG ON THE ANATOMY base of the tongue are unpigmented, and the vallate and fungi- form papille may or may not be colourless; in the former case the appearance of the dorsum resembles that of Lagothrix infumata (text-fig. 37). The Circumvallate Papille (text-fig. 45 A).—Three papille form an isosceles triangle with long sides, and an acute vertical angle. The posterior papilla is granular and _ projects considerably beyond the coarsely-nodulated vallum. The right basal papilla is granular, and the vallum is overlapped by conical papillz. The left basal papilla is round, granular, and pigmented. The Fungiform Papille (text-fig. 45 A & C) form a narrow dorsal bounding zone on which they have the usual arrangement, Text-figure 45. The tougue of Callicebus moloch. but the transverse rows extend far forwards. Those on the inferior surface are minute and form a single row. Some are overlapped by conical papille. The Oonical Papille (text-fig. 45 A & D) are all shaggy, and stand up around the fungiform papille when the tongue is dried. The rows are mostly transverse, but the distribution according to size is not according to the typical plan. All the papille are of the filiform types. Lymphoid Nodules and Glands.—The base of the tongue“is occupied by nodules, but no orifices of ducts of glands are visible. The Lateral Organs (text-fig. 45 E)—The wide primary sulci appear as incisions on the lateral borders, for the intervening lamin are flat. They form aslightly concave row whose posterior members are level with the lateral vallate papilla (1.y.P.). wo ee ee ee ee eS ee "a7 OF THE TONGUES OF THE MAMMALIA. 5y ial The entire organ is surrounded by an oval colourless area of the tongue, so they offer a marked contrast to the bluish-black lateral borders. The right organ is °6 em. long, and has 10 lamine and 11 sulci. The left organ is °5 em. long, and has 9 lamine and 10 sulci. The Ventral Papillary Zone is bluish-black and narrow. It has many small irregularly-placed conical papille, and a few fungiform papille in a single row posteriorly, The Ventral Mesial Sulcus is narrow and deep, and the frenal lamella is triangular and entire with lobulated edges, THE Brown TEEres (C. brunnea). In a hitherto unpublished paper Garrod states: ‘There are three circumvallate papillz in the form of a triangle. The fungi- form papille ere most numerous at the tip of the tongue, and there are a few in front of the circumvallate papilla and Mayer's Organ. There is a good bifid sublingua” (=frenal lamella). Owen (6) described a vallate triangle in C. personata, Genus Aorus (=NycriprrHects). Tue Fevine Dovrovucoeut (A. felinus). The two specimens which I examined were preserved, so no measurements were made, One of them has a single vallate papilla. The apex and lateral borders have the usual characters, and a well-marked mesial dorsal sulcus runs along the base of the tongue. The Circumvallate Papille (text-fig. 46 A & C).—Two pro- minent vallate papille lie near the lateral borders. Each is granular, and has a prominent nodular vallum. The fossa are. well-marked. Both are round on plan and on conical elevation. The Fungiform Papille (text-fig. 46 A & D) form a dorsal bounding zone on which they have the usual arrangement. Some are flush with the general surface of the tongue (0), but others stand up prominently (a & 6). They form a “single row on the ventral surface of the tongue, and some are overlapped by conical papille (d). The Conical Papille (text-fig. 46 A & E) have the usual dis- tribution according to size, and the usual direction of their points. They do not have the usual arrangement, however; on the oral part of the dorsum they are irregular, but on the base they form oblique chains. They are cylindrical or filiform, and may have central depressions. The Lateral Organs (text-fig. 46 F).—The lamine and deep narrow primary sulti lie on the edges and the inferior surface of the tongue. They run forwards and upwards, and the laminz are traversed by secondary sulei. The right organ is ‘9 cm. long, Proc, Zoou. Soc,— 1921, No, XX XV, 35 512 DR. C F. SONNTAG ON THE ANATOMY and has 5 lamine and 6 sulci. The left organ is ‘8 cm. long, and has 4 lamin and 5 sulci. The Inferior Surface (text-fig. 46 B)—The narrow ventral papillary zone has irregularly disposed conical papillee and a single row of discrete prominent fungiform papille; the latter are most numerous and closest together round the apex. Text-figure 46. Vad hs Elevation. Leve/ of surface of 2&- 6 SS a tongue. D Kee) COA [Ey The tongue of Aotus felinus. The frenal lamella is triangular. Tt has a bifid apex and plain sides. The ventral mesial sulcus is shallow and deep, and the frenum is of moderate length. Genus PIPHECIA. THe Buack Saki (LP. satanas). The specimen which I examined was preserved, so no measure- ments were made. The tongue has almost parallel sides, and the apex and lateral borders have the usual characters. The thickness rapidly diminishes from frenal lamella to apex. The Circumvallate Papille (text-fig. 47 A & C).—The vallate apillae form an isosceles triangle with the apex behind. All are slightly retracted, with granular surfaces, the fosse are well- marked, and the vallums are nodulated, 2 a ee ee ee ae ee OF THE TONGUES OF THE MAMMALIA. 513 The Fungiform Papille (text-fig. 47 A & B) form a dorsal bounding zone on which they have the usual arrangement, and the apical cluster is very well-marked. They form a single row on the inferior surface. They are all prominent, smooth, and hemispherical, and none are overlapped by conical papille. The Conical Papille (text-fig. 47 A & D) have the usual arrangement in clusters and rows of varying degrees of obliquity. The vertical direction of the rows on the lateral borders can only be made out by very careful examination, however. All belong to the filiform series. Text-figure 47. - The tongue of Pithecia satanas. The Lateral Organs (text-fig. 47 E) form triangular or concave bodies on the sides and inferior surface, and the lateral vallate papille are level with the apex. Several of the pointed laminz have secondary sulci. The right organ is ‘9 cm. long, and has 5 lamine and 6 sulci. The left organ is ‘8 cm. long, and has 4 lamine and 5 sulci. The Inferior Surface (text-fig. 47 B).—The broad bounding papillary zone is wide, and has papille disposed in the usual manner. The frenal lamella is a long plate with rounded apex. and the orifices of the Wharton’s Ducts open on the ventral surface near the base (text-fig. 47 B, a). The ventral mesial sulcus is wide and deep, and the frenwm is short. Mayer deseribed a vallate triangle in P. satanas (4). Bie 514 DR. C. F. SONNTAG ON THE ANATOMY Humpoupt’s Saxt (P. monachus). Flower (3) describes the tongue as follows :—‘“ The length of the tongue, from the base of the epiglottis to the tip, is 1-4 inch; its breadth *5 inch; its sides are parallel, the end square, with the corners shghtiy rounded. The sublingual organ is fleshy, except towards the end, which is sharp-pointed and cleft at the extreme tip; it is free to the extent of -25 inch, and its apex is ‘3 from the tip of the tongue.” In a hitherto unpublished note Garrod states that the tongue of Pithecia albinasa has a triangular vallate papillary area. Genus CAcagao. THe Batp OvaKari (C. calva). The specimen had been preserved, so no measurements were made. The apex and lateral borders have the usual characters, but fine lateral striz, only visible on high magnification, run in from them across the dorsum. Text-figure 48. The tongue of Cacajao calva. The Circumvallate Papille (text-fig. 48 A).— Three vallate papille form an isosceles triangle whose apical papilla is far from the epiglottis; the basal papillae are close to the edges of the tongue. The apical papilla is small and the basal ones large, the fosce are all well-marked and the vallums are nodular. The Fungiform Papille (text-fig. 48 A & B) form a dorsal bounding zone on which they have the usual arrangement in OF THH TONGUES OF THE MAMMALIA. 515 clusters and rows, but the apical cluster is small. Nowhere are they numerous, however. ‘They are absent from the inferior surface. The Conical Papille are irregularly arranged, but they increase in size in the usual manner and their points run in the usual directions. They form a narrow papillary zone on the inferior surface, but this is widest round the apex. Glands and Lymphoid Tisswe—The base of the tongue is finely nodular, and a few orifices are visible when it is highly magnified. The Lateral Organs (text-fig. 48 C) form concave bodies on the sides of the tongue, and the pointed or round Jamine have fungiform papille below their bases. The lateral vallate papillee are level with the centres of the organs. All rounded lamine have wide secondary sulci. The right organ is 1 cm. long, and has 7 lamine and 8 sulci. The left organ is -9 em. long, and has 6 lamine and 7 sulci. The Inferior Surface (text-fig. 48 B).—The ventral papillary zone is narrow, the ventral median sulcus is shallow, and the frenum is small. The frenal lamella is long and narrow, and terminates in two small points (text-fig. 48 B, a). Genus ALouarra (= Mycerss). Tue Brack Howrter Monkey (A. caraya). The specimen had been preserved, so no measurements were made. The Circwnvallate Papille (text-fig. 49 A & C).— Three papille form an isosceles triangle with long sides. The apical papilla is far from the epiglottis, and the basal papille are close to the lateral borders of the tongue. The length of the vallate triangle is one-third the length of the dorsum. ’ The apical papilla is large, oval, granular, and excavated in the centre, the fossa is well-marked, and the vallum striated. The Fungiform Papille (text fig. 49 A & B) are not numerous, but have the usual arrangement on the dorsum; and a few lie at the sides of the vallate triangle. They appear to stand on ridges of the mucosa when the tongue is highly magnified. Those on the inferior surface are very small. The Conical Papille (text-fig. 49 A, B, & D), which are closely- set, are irregularly disposed on the dorsum, but they increase in size in the usual manner, and their points run in the usual direction. They are filiform, cylindrical, or fusiform, and the number of points to each papilla varies. Some stand up so prominently round the fungiform papille that they conceal them entirely. Glands and Lymphoid Tissue (text fig. 49 A).—The base of the tongue is covered with nodules of various sizes, and many orifices of ducts are visible to the naked eye. The lens reveals the existence of many minute openings, and the base of the tongue is 516 DR. C. F. SONNTAG ON THE ANATOMY more pitted than in other species of the Cebide. Many nodules with central depressions are also present on the anterior surface of the epiglottis. The Lateral Organs (text-fig. 49 E).—On each side they form a straight row of lamine and “sulci. All the sulci and lamine are short ; the posterior ones are straight, and the anterior ones, which are level with the anterior vallate papille, are convex forwards. The right organ is ‘9 em. Jong, and has 10 lamine and 11 sulci. The left organ is ‘7 em. long, and has 8 lamine and 9 sulci. Text-figure 49. WG MP1!) yy )))) )) | I The tongue of Alouatta caraya; t.t. tonsils. The Inferior Surface (text-fig. 49 B).—The ventral papillary zone is narrow, but has papilla disposed i in the usual manner. The ventral mesial sulcus is short, but does not open up much when traced posteriorly. The frenum is short, and the frenal lamella is long, narrow, deeply incised at the apex, and has plain sides (text- fig. "49 B, a). Carus and Otto (1) described a vallate triangle (A. fuscus) closely resembling that of A. caraya, described above. —— ~~ ee ee ee SS, ee CE ew 5 OF THE TONGUES OF THE MAMMALIA. 517 Summary of the Tongues of the Cebide. 1. All tongues, except those of the Sakis, are conical, and some are pigmented. 2. The conical papille are irregularly disposed on the oral part of the dorsum in most of the Cebide, but they may be arranged in chains on the base of the tongue. 3. The fungiform papille always form a dorsal bounding zone. They may be absent from the inferior surface. 4, The vallate papillary patterns are—a single papilla, a pair, a triangle, a V, or a double-pair. 5. The vallate triangle is far from the epiglottis in many species, and its median may be very long (the median is the line drawn from the apex to the mid-point of the base). 6. The orifices of the ducts of glands are fewer than in the Cercopithecide. They are most marked in Alowatta. 7. The lateral organs are most frequently concave towards the epiglottis. 8. The frenal lamella is usually bifid at the apex. 9. The lingual characters are not sufliciently distinctive to be of value for the classification of species. Family HAPALID&. Genus HAPpALe. Tue Common Marmoszer (7. jacchus). ‘The comparatively long, slender tongue has the following measurements :—total length 2°1 em.; length of the oral part 1-9 cm.; length of the pharyngeal part *2 cm. ; width between the lingual attachments of the palato-glossal folds °65 cm. These figures were found to be the average in twelve specimens. The apex is flat and tuberculated by fungiform papilla; it is devoid of a notch. Running back for 1 cm. is a median dorsal sulcus from which transverse sulci pass to the sides of the tongue, but these are the result of preservation and are not shown in text-fig. 50. he lateral borders have the usual characters. The Circumvallate Papille (text-fig. 50 A).—Three vallate papill form an isosceles triangle with the apex behind, and the two basal papille are close to the edges of the tongue. ‘The vallums are covered by conical papillae which obscure the fossee, and the surfaces of the papillary bodies are finely granular. This pattern was present in all twelve specimens. Tuckerman (9) also described this as the type. The Fungiform Papille (text-fig. 50 A & B) form a dorsal bounding zone on which they have the usual arrangement, in clusters and rows, and they are surrounded by prominent upstanding conical papilla. They are few in number, however, on the dorsum, and are absent from the ventral bounding papillary zone. 518 DR. C. F. SONNTAG ON THE ANATOMY The Conical Papille (text-fig. 50 A, B, & C) have an irregular disposition, but they increase in size in the usual manner, and their points run in the usual directions. They appear as imbri- cated scales, and some stand up prominently round the fungiform and vallate papille, concealing the vallums of the latter (text-fig. 50 C, e). . Lymphoid Tissue and Glands—A few nodules are present among the conical papille on the base of the tongue, but no duct orifices are visible. Text-figure 50. Wy ints, * vty, a) ue 4 V tm Cc. The tongue of Hapale jacchus. The Lateral Organs (text-fig. 50 D).—These consist of a few fine lamine and sulci on the sides of the tongue. The right one has 6 lamine and 7 sulci, and the left one has 5 lamine and 6 sulci. Some of the lamine have small secondary sulci. The Inferior Surface (text-fig. 50 B).—A narrow ventral papillary border, possessing conical papille ‘only, is present, and the narrow, deep ventral mesial sulcus opens posteriorly into a triangular fossa. _ The frenal lamella is triangular, with a bifid apex and plain or nodulated sides, —=—; eT eee ee a ae —_ =. ee eee OF THE TONGUES OF DHE MAMMALIA. 519 Tse Buack-wARED Marmoser (ZZ. penicillata). The conical tongue, which is green in colour, has the following measurements :—total length 2°35 em.; length of the oral part 2-1 em.; length of the pharyngeal part -25 em.; width between the lingual attachments of the palato-glossal folds -9 em. The apex is pointed and tuberculated ; running backwards from 1:4 cm. is a median ridge, whence transverse ridges, which are convex forwards, run out to the lateral borders. The Circwmvallate Papille.—Three papille form an isosceles triangle. The apical papilla is close to the epiglottis, and the lateral papille are close to the edges of the tongue. The lateral papille are small, and their fossee and vallums are concealed by conical papille (text-fig. 51 A, 6). The posterior papilla, which is eval and excavated, has a nodulated vallum composed of two rows of lobules (text-fig. 51 A, c). The Fungiform Papilic form a wide dorsal bounding zone, and the area of the centre without papille is small. They have the usual arrangement, but the posterior rows form thick clusters in front of the vallate papille. They are hemisphericai or flat, and the surfaces are smooth and glistening. As in 1. jacchus, there are none on the inferior surface of the tongue. The Conical Papille have the usual arrangement, but their size remains the same from apex to base. They are scale-like and imbricating, as in H. jacchus. They form a narrow zone on the inferior surface of the tongue (text-fig. 51 A, a). Lymphoid Tissue and Glands—The area occupied by the lymphoid nodules is small, and no duct orifices are visible. The lateral organs consist of a few very fine laminz and sulci. It is difficult to detect these, even on high magnification, by a hand lens, so microscopical examination is necessary in order to settle their characters. The Inferior Surface (text-fig. 51 A, a).—A narrow ventral papillary border is present; the narrow, deep ventral mesial sulcus opens posteriorly into a triangular fossa; and the frenal lamella 1s triangular, with an entire apex and notched sides. Tur Pyemy Marmoser (41. pygmea). The tongue has the following characters :— 1. It is conical with a rounded apex. 2. The vallate papille form an isosceles triangle. 3. The fungiform papille resemble those of H. penicillata. . The conical papillae form imbricating scales. . The lateral organs are well-marked. . The ventral papillary zone is well-marked. . The ventral mesial suleus is long and deep. . The triangular frenal lamella has a bifid apex. OID Oe 520 DR. C. F. SONNTAG ON THE ANATOMY Genus LEONTOCEBUS. Tue Lion Marmoser (L. rosalia). The two specimens examined were preserved, so no measure- ments were taken. The tongue rapidly tapers from base to apex ; and the latter is pointed and covered with a dense cluster of small fungiform papille. The middle of the dorsum is occupied by a strip of mucosa covered by closely-aggregated conical papille. The Circumvallate Papille (text-fig. 51 B, 6)—Three papille form an isosceles triangle with two short sides and a long base. Text-figure 51. fi sf ANN a. (Ge ® @@ 6 ae c Wy © tGewe é. a. las 4, A 4 (e) Dd. Cc. B. The inferior surfaces of the tongue and yallate papillae of some Marmosets. The apical papilla is a short distance anterior to the epiglottis, and the basal papille are close to the edges of the tongue. All the papille are large and nodulated, the fosse are well- marked, and the vallums are clear zones or lobulated bands. The Fun giform Papille, which are prominent, form a well- marked dorsal bounding zone on which they have the usual arrangement in clusters and rows, and they increase in size in the usual manner. ‘T'wo rows of small elements connect the basal vallate papille. None are present on the posterior part of the ventral papillary zone. All are prominent, hemispherical, and smooth. The Conical Papille (text-fig. 51 B, c).—No apical cluster is present, for the transverse rows extend right forward to the apex. eS." OF THE TONGUES OF THE MAMMALIA. 521 Hach is leaf-shaped, and has several hair-like processes along its free edge, and all the points are directed backwards. Some of the posteriar papille may be umbilicated. They increase in size in the usual manner. Lymphoid Tissue and Glands.—The nodules form a narrow zone in front of the epiglottis, but no duct orifices are visible, even through the lens. The Lateral Organs.—The lateral organs consist of well-marked, but small, backwardly-directed Jamin separated by deep primary sulci. The right organ has 3 lamine and 4 sulci, and the left organ has 4 lamine and 5 sulci. The Inferior Surface (text-fig. 51 B, a & a'').—-No ventral papillary zone is present, but the outer part of the inferior surface incised by small fissures. The ventral mesial sulcus, which is narrow and deep, opens posteriorly into a large triangular space. The frenum is short, and the frenal lamella is triangular, with a bifid apex, or a deep notched lateral border. Genus Mystax. Tae Necro Tamarin (J. wrsulus). My fresh specimen has the following measwrements :—total length 2°8 em.; length of the oral part 2°4 em. ; length of the pharyngeal part ‘4 cm.; width between the lingual attachments of the palato-glossal folds 1:2 em. The apex is rounded, and a dorsal median sulcus, which is more marked than in any specimen belonging to the Cercopithecide, Cebide, or Hapalide, runs back to the apex of the vallate triangle. No transverse ridges or sulci are present. The Circumvallate Papille (text-fig. 52 A & C).—Four papille are arranged in the form of a V. Including the apical papilla the right limb has three papille and the left one has two. The posterior papilla (A, C a) is oval, with a well-marked fossa and nodular vallum. The left lateral papilla (A, C 3) lies in a wide fossa with three small ones. The right lateral papilla (A, C c) lies in a wide fossa with two small ones, and the right mesial papilla is circular. The Fungiform Papille (text-fig. 52 A & B).—Small, but well-marked papille form a dorsal bounding zone, but the trans- verse rows are maintained as far forwards as the apex. They are absent from the inferior surface. The Conical Papille,—No apical cluster is present, for the papille maintain their arrangement in rows right forward to the apex, and there is no great increase in size from before-backwards, or without-inwards. They have the usual arrangement on the ventral papillary zone. All are cylindrical and surrounded by zones of interpapillary dorsum. ; Lymphoid Tissue and Glands.—On each side of the epiglottis there is a large nodular mass, and the two are connected by a 522 DR. C. F. SONNTAG ON THE ANATOMY band of small nodules which also ‘passes towards the posterior vallate papilla. Some of the large nodules have small central pits. The Lateral Organs.—The lamine, which increase in size from before backwards, are separated by wide primary sulci. The right organ has 3 lamine and 4 sulci, and the left one has 4 lamine and 5 sulci. e The Inferior Surface (text-fig. 52 B).—The ventral papillary border, which is narrow, possesses conical papille only. Text-figure 62. ¢ 2pAt wx 27450 NS bo Al) t i Ya /| I 0 23 Voy 0 eS AS 10,4 (CA) \—w A. The tongue of Mystax ursulus. The ventral mesial sulcws, which is long, narrow, and deep, opens posteriorly into a Jarge triangular depression containing a well-marked median crest. The frenal lamella is a long, flat body with its anterior extremity deeply cleft into several long, thin club-like processes. Tue Movusracup Tamarin (I. mystaz). The tongue, whose form resembles those of the preceding species, has the following measurements :—total length 2-1 em. ; length of the oral part 1:8 cm.; length of the pharyngeal part -3 cm.; width between the lingual attachments of the palato- glossal folds °6 cm. A deep median dorsal sulews runs back for 1:1 em. from the apex, and transverse furrows run towards the lateral borders ; these separate broad ridges which are convex forwards and con- cave backwards. The Circumvallate Papille.—Three papille form an isosceles triangle with two short sides and a long base ; the apical papilla is close to the epiglottis, and the basal papille are close to the lateral borders of the tongue. eI ee eee eS ee eee ee ee ee eee Tie le ea a ett ene ee se es AS a OF THE TONGUES OF THE MAMMALIA, a8 All are round and finely-granular, the fosse are well-marked, and the vallums are lobulated. . The Fungiform Papille, which are all small but prominent, form a dorsal bounding zone on which they have the usual arrangement in clusters and rows, the usual number on each row being two. Most of them are equal in size, and are smooth and hemispherical. The Conical Papille, which are all closely-ageregated, have the usual arrangement in rows and clusters; they increase in size from without-inwards, but their increase in size from before- backwards is slight. They take the form of overlapping scales with hair-like processes. Lymphoid Tissue and Glands.—The nodular zone on the base of the tongue is narrow, and no orifices of ducts and pits are - visible, even when the tongue is highly magnified. The Lateral Organs ave small and oval: their centres are level with the lateral vallate papille. They consist of well-marked primary sulci separating comparatively broad lamine with pointed or rounded upper extremities. The secondary sulciare very thin. The right organ is *1 cm. long, and has 3 lamine and 4 sulci. The left organ is ‘1 cm. long, and has 4 lamine and 5 sulci. The Inferior Surface.—The narrow ventral papillary zone has only conical papillx, but is incised by transverse sulci running in from the dorsum. The ventral mesial sulcws, which is narrow and deep, opens into a large triangular space, which also receives the short frenum. The frenal lamella is triangular, with a bifid apex and fissured lateral borders. Genus CipIpomipas. Tue Pincnt Monkey (@. edipus). Flower (3) described the tongue and gave the following data :— 1. The tongue tapers gradually. 2. There is a median dorsal sulcus posteriorly. 3. The ventral median sulcus is deep. 4. The fungiform papille are regular, and cover the entire oral part of the dorsum. 5. Three vallate papilla form a triangle. 6. The lateral organs ave well-marked. 7. The frenal lamella has a dorsal ridge and cleft apex. Summary of the Tongues of the Hapalide. 1. The tongue is either long, narrow, and tapering slowly, or short, conical, and tapering rapidly from a wide base. 2. Mesial dorsal ridges are frequently present. 3. The tongue of Mystax ursulus has a better-marked mesial dorsal sulcus than those of all other Primates, 54 THE ANATOMY OF THE TONGUES OF THE MAMMALIA. 4, The vallate papille always form a triangle. 5. The fungiform papille always form a dorsal bounding zone. 6. Seale-like conieal papillz, with or without filamentous pro- cesses, are exceedingly common. 7. The ventral papillary zone is small or absent. 8. The frenal lamella is usually subdivided at the apex, but undivided lamelle occasionally appear. 9. The lateral organs are always small. 10. The lytta, foramen cecum, plice, fimbriate, and apical gland of Nuhn are absent. BIBLIOGRAPHY. 1. Carus and Orro.—Erliuterungstafeln zur verg]. Anatomie. Heft 4. Leipzig, 1835. . Cuvrer.—Legons d’Anatomie comparée. Paris: Crochard et Cie., 1835. . Frower, W. H.—Medical Times and Gazette, 1872, vol. i. pp. 507-509; and Proc. Zool. Soc. London, 1862, p. 331. . Mayer, F. J. C._—See Bibliography to paper No. 8. Mincu.—See Bibliography to paper No. 8. Owen, R.—The Comparative Anatomy of the Vertebrata, vol. 111.2 199: ”. Popwisorsky, V.—‘‘ Anat. Untersuch. iiber die Zungendriisen des Menschen und der Siiugethiere.” Diss. Dorpat, 1878. 8. Sonnrac, C. F.—“The Tongues of the Cercopithecide.” Proc. Zool. Soc. London, 1921, pp. 277-322. 9. TuckerMAN, F.—Journal of Anatomy and Physiology, 1892, pp. 391-392. 10. Bovunarr and Piniier.—See Bibliography to paper No. 8. 11. Gecensaur, C.—Morphol. Jahrbuch. 1894, pp. 1-18. ~ 12. Pocock, R. I.—Ann. Mag. Nat. Hist. (8) xx. pp. 247-258. Oop w »w ON THE SKULL OF DINOTHERIUM GIGANTEUM. 525 28. Note on the Skull of Dinotherium giganteun in the British Museum. By C. W. Anprews, D.Sc., F.R.S., F.Z.S. (Submitted for Publication by permission of the Trustees of the British Museum.) [ Received March 5, 1921: Read May 24, 1921.] (Text-figures 1-4.) One of the most important specimens in the Geological Depart- ment of the British Museum is the fine skull of Dinotheriwm giganteum, which seems to be the only even approximately complete example at present known. The history of this spe- cimen is of considerable interest. It was discovered in 1835 by Klipstein, about eighteen feet deep in the well-known ossiferous sands of Eppelsheim in Hesse-Darmstadt. An account of this discovery, including a description of the specimen and of the deposits in which it occurred, was published by Klipstein and ‘Kaup in 1836 under the title ‘“‘ Beschreibung und Abbildungen von dem in Rheinhessen gefundenen colossalen Schiidel des Dinothervi giganter.” : in the following year a French translation of this paper appeared. Kaup, in his part of the memoir, not only gives a description of the skull itself, but also'a very interesting account of the method of collecting the specimen. _It appears that it was found lying on its dorsal surface in a bed of clay and sand. This was partly removed, leaving the skull resting on six pillars of matrix. Next, iron bars were passed through the openings thus made, which were then filled up with plaster of Paris. Next, the remaining portions of the matrix were removed and their place filled with more plaster, so that finally the skull rested on a solid base of plaster strengthened by iron bars. A stout board was inserted beneath the whole mass, which was then lifted from the excavation by twenty-four men with ropes. A quaint picture of this operation is given on the cover of Klipstein and Kaup’s memoir. In this figure there is also shown a huge proboscidean femur, which is perhaps the original from which the cast of a Dinotherium femur now in the British Museum was made. In 1837 the skull was sent to be exhibited in Paris, part of the expenses of transport being borne by the French Academy. While there it was examined by de Blainville, whose description of the specimen is the best yet published ; but even he was not allowed any opportunity of examining the dorsal surface. On the same occasion Isidore Geoffroy, Straus, and others took the opportunity of giving their views as to the structure and affinities of Dinotheriwm. It was then intended to send the specimen to London, but whether this was done on this occasion is uncertain. Some years later, in 1849, it was certainly in London, and was offered to the British 526 DR. C. W. ANDREWS ON THE Museum for purchase ; it was then examined and reported on by Buckland and Owen. ‘The purchase was not completed, and the skull seems to have been sent back to Darmstadt. About 1866, Klipstein sold his collection to Dr. Oldham, the Director of the Geological Survey of India, and from him the British Museum acquired not only this, but also some other valuable specimens, including the skull of Dorcatheriwm nawi and the front of the skull of Z'apirus priscus. Dr. W. D. Mathew, when visiting the Museum recently, pointed out to me that there is a widely-spread idea on the Continent that the skull now under discussion was broken up on its journey to London. This mistake has even appeared in vint: thus Weinsheimer in his memoir on Dinotheriwm (Pal. Abhandl. Bd. i. 1883, p. 244), in speaking of the cast of this specimen, states that ‘“ The original of this cast no longer exists, since on its journey to London it was irretrievably broken into fragments.” Examination of the skuil shows that, in fact, at some time it has been much broken, but has been, on the whole, skilfully mended, the figures and cast no doubt having been followed in making the restoration. In addition to Kaup’s original descriptions and _ figures, de Blainville has given a very good account of the skull so far as he was able to examine it, and numerous other writers have referred to it. The most complete summary of the various views that have been held as to the affinities of Dinotherium is given by Stefanescu in his paper on Dinotherium gigantissimum (Anuarulu Museulni de Geologia, ete., Bukarest, 1894, p. 126). It is now voposed to give a short description of the skull as it now is, followed by some discussion of the reasons for its peculiarities and of the relationships of Dinotherium to the other Proboscidea. The occipital condyles (0.c.) are very large and prominent ; ventrally and laterally they are limited by a deep groove, but dorsally they seem to pass uninterruptedly into the occipital surface (oc.s.). The articular surface of the condyles from above down- wards makes rather more than a semicircle, so that the range of movement of the head up and down was extensive. Ventrally the condyles are separated by a broad, deep notch, in front of which the basioccipital runs downwards and forwards as a broad ridge, narrowing a little towards its junction with the basisphenoid, at which point. however, there is a prominence. The basisphenoid is not quite in the same straight lineas the basioccipital, but seems to have been directed a little more upwards. The basis cranii as a whole slopes rather steeply downwards, making an angle of about 135 degrees with the plane of the teeth and the posterior part of the palate, which is regarded as being horizontal. From the outer angle of each condyle a ridge (7.) runs outwards and a little upwards, terminating in the outer angle of the squamosal (a.s.), which projects outwards far beyond the level of the zygomatic processes. These ridges, which towards their outer ends rise into thick and prominent crests, separate the upper occipital surface SKULL OF DINOTHERIUM GIGANTEUM, 527 from the large post-tympanic flanges (pé.fl.), which appear to be formed partly by the exoccipitals and partly by the squamosal, though no suture can be seen; the angle between the post- tympanic flanges and the upper occipital surface is about 130 degrees. The ventral border of the flanges is gently convex, and is separated from the condyles on the inner side by a deep notch, which is partly occupied by a tongue-like process of bone, apparently the paroccipital process (p.p.) of the exoccipital. The anterior face of the flanges is concave from side to side, and is separated superiorly from the greatly elongated glenoid surface (gl.) for the mandible by a shallow transverse channel. Above Text-figure 1. i 1 a @ fl {| fi n ) Vien) ( y GAN) )y¢f\* rong | wey Hig ——= i Skull of Dinotherium giganteum. From right side: about js nat. size. af., antorbital foramen; a/.c., posterior opening of alisphenoid canal; ea.n., external nares; /.c., lambdoidal crest; m.s., depression for muscle attachment on preorbital prominence; o.c., occipital condyle; oc.s., occipital surface; o7b., orbit ; or.r., orbital ridge; pmx., premaxilla; p.orb., post-orbital process of frontal; pt.fl., post-tympanic flange; zyg., zygomatic process of squamosal. the transverse ridges just described the occipital surface slopes forwards to a remarkable degree, giving the skull a curious resemblance, in this region, to the skull of some types of Whales. -e. g. Balenopiera. The angle between the occipital surface and the basis cranii is about 535 degrees, so that the occiput is nearly parallel to the palatal surface. This extraordinary forward inclination of the occiput evidently limits the size of the cranial cavity, and de Blainville suggested that possibly the occipital surface had been crushed down towards the floor of the skull. Careful examination of the specimen, however, does not seem to support this idea, and it seems probable that, remarkable as the Proc. Zoou. Soc.—1921, No. XXXVI, 36 528 DR. C, W. ANDREWS ON THE structure of this region of the skull may appear, it represents nearly the condition in the living animal, In other Proboscidea the occiput slopes strongly forwards, but this inclination is masked, in the adult at least, by the enormous development of cellular bone that takes place in this region, except over the area of the insertion of the ligamentum nuche, which occupies in consequence a deep depression. In young individuals in which this development of the bone has not taken place, it can be seen that the occiput is inclined to the basis cranil at an angle about equal to that occurring in Dinotherium. Thus, in a figure of a median longitudinal section of the skull of a young African Elephant given by Flower (‘ Osteology of the Mammalia,’ ed. 2, p. 181, fig. 59), the somewhat convex occipital surface is inclined to the basis cranii at an angle of about 40 degrees—that is, at a more acute angle than in Dino- therium. In the skulls of two very young Indian Elephants this angle was about 50 degrees, and in Paleomastodon, in which in the adult the spongy bone is relatively little developed, the angle is only about 35-40 degrees in the middle line. Thus it appears that the forward slope of the occiput relatively to the basis cranii is not really greater in Dinotheriwm than in the later Proboscidea, but only appears so because, when the palate and tooth series in Dinotherium is placed horizontally, the basis crani slopes steeply downwards instead of being nearly horizontal. In Dinotherium no development of cellular bone seems to have taken place, the necessary surface for the attachment of the muscles supporting the heavy head being in this case supplied by the great widening of the occipital surface, which is carried out laterally on to the lateral extensions of the squamosals (a.s.). The pit for the nuchal ligament (7./.) is both wide and deep: behind it the occiput is convex from before backwards, while at the sides it is gently convex in all directions. Anteriorly the occiput is bounded by the transverse portion of the lambdoidal crest (U.c.), which is slightly convex forwards: in front of this the parieto- frontal surface makes an angle cf about 140 degrees with the occiput. Laterally the lambdoidal crest reaches the upper border of the temporal fossa at about its middle point, and thence is continued downwards and backwards along its border to the lateral processes of the squamosal. The lateral portion of the occiput thus bounded, projects far over the enormous temporal fossa, forming a kind of roof to its posterior portion. The fronto- parietal region of the roof is short from before backwards and in | its middle portion is flat or slightly concave. It first narrows a little, and then widens out and slopes down towards the post- orbital (p.orb.) processes, which are presumably borne by the frontals though no sutures are visible. A little in front of the level of the post-orbital processes the fronto-parietal region of the roof is bounded by a thickened border, which is concave forwards and forms the upper edge of the remarkably wide nasal opening (er.n.), There is no trace of the nasal bones; they may have been SKULL OF DINOTHERIUM GIGANTEUM. 529 lost, but in any case do not seem to have formed any median pro- jection over the nasal opening. The maxilla (ma.), which formed the sides of the nasal opening at least in part, is a very massive and greatly developed bone. Externally to its tooth-bearing and palatal region it is produced laterally into a shelf of bone the hinder border of which is opposite the hinder lobe of the first molar. Posteriorly it forms the floor of the orbit, and no Text-figure 2, ~ SS SS SSS = ——————— \ \ S Skull of Dinotherinm giganteum. From above: about 7s nat. size. a.s., outer angle of squamosal; ev.n., external nares; jug., jugal; U.c., lambdoidal crest; 2.f., nuchal fossa: 0.¢., occipital condyles; oc.s:, occipital surface; p.orb., postorbital process of frontal; pt,fl., post-tympanie flange; 7., ridge from outer angle of squamosal. doubt joined the jugal, on which the lower post-orbital process is borne. In front of the orbit the maxilla forms a great mass of bone, terminating laterally im a rugose somewhat concave surface for the attachment of a muscle (m.s.), probably the maxillo- 36* 530 DR. C. W. ANDREWS ON THE labialis superior. The Jachrymal bone perhaps took part in the formation of this pre-orbital mass, but its limits cannot be determined. The very large antorbital foramen (a,f.) opens on the lower surface of the maxilla, about mid-way between its prominent pre- orbital boss aud the premolar teeth. The point of union of the maxille with the premaxillz is probably marked by Text-figure 3. Cie Wed te i i tip. Phe Oe Skull of Dinotherium giganteun. From below: about 7's nat. size. a.f., anturbital foramen; ad.c., posterior opening of alisphenoid canal; a.s., outer angle of squamosal; aw. external auditory meatus; f.l.p., ? foramen lacerum posterius; j., jugal; g/., glenoid surface for mandible; m.8, third molar; m«., maxilla: oe.c., occipital condyles; pima., premaxilla ; pé.fl., post- tympanic flange; ¢.6., Ptympanie bulla; zyg., zygomatic process of the squamosal. the sudden narrowing of the snout about 40 centimetres in front of the orbit. The suture hetween the two bones can be traced for a short distance on the palatal surface. SKULL OF DINOTHERIUM GIGANTEUM. Dol The premaxille (pme.), though narrower than the maxille, ave still very massive and widely expanded. ‘Their upper surface is concave from side to side, the lower convex in the same direction. In the mid-ventral line they are separated in the actual skull by a cleft, probably the result of distortion through crushing: in the east, which seems to have been made before the skull had been broken, there is only a deep groove. From the posterior angles of this groove a pair of ridges run back on the palatal surface to the anterior end of the tooth series. The anterior ends of the premaxille are thickened, and terminate in a nearly flat surface in which there are some irregular asym- metrically arranged pits but no real trace of any incisor alveoli, though it seems not unlikely that incisors will be found to have been present in the young animal. The anterior portion of the skull in front of the tooth series is curved downwards, its palatal surface approximately following the curve of the downturned mandibular symphysis. The structure of the basal region of the skull is not very clear, probably in consequence of the crushing and fracturing it has undergone. On either side of the basioccipital there is a fairly weil-developed auditory bulla (¢.b.). External to this is a strong erest of bone, terminating posteriorly at the process described above as the paroccipital (p.p.):; Anteriorly it runs inwards to the sides of the basisphenoid. This crest, which seems—at least in front—to be formed by the pterygoid, turns downwards anteriorly, and forms the border of the narrow opening of the posterior nares (p.n.): ventrally it terminates just behind the tooth series in a boss of bone which forms the posterior angle of the hard palate. Dorsal to this boss there is a deep fossa, presumably in the palatine, passing upwards to the hinder opening of the ali- sphenoid canal (al.c.). Behind and a little to the outer side of this opening is the inner end of the extraordinarily elongated glenoid surface (g/.) for the mandible. This surface is very narrow from before backwards and gently convex in the same direction: posteriorly it is bounded by the depression separating it from the post-tympanic flange. Its outer end is immediately behind the base of the zygomatic process (zyg.) of the squamosal, and its anterior border, at least in its outer half, abuts on the temporal fossa. The auditory opening (aw.o.) is opposite the outer border of the zygomatic process: the external auditory meatus is greatly elongated, and its floor is formed by the roof of the channel behind the glenoid surface. Above the auditory opening the massive lateral angle of the squamosal (a.s.) projects outwards some fourteen or fifteen centimetres. The structure of a bony laby- rinth supposed to belong to Dinotheriwm has been described and figured by Claudius (“Das Gehorlabyrinth von Dinotheriwm giganteum,” Paleontographica, vol. xiii. (1864-66) p. 65). The opening of the optic foramen is indicated by a depression, from the upper edge of which a downwardly deflected ridge (o07.7".) runs upwards and forwards across the side wall of the skull and 532 DR. C. W. ANDREWS ON THE terminates in front in the post-orbital process (p.orb.). Towards its anterior end it becomes very prominent, sharply delimiting the eye-socket above. This ridge seems to be very characteristic of the Proboscidea: it occurs in Meritheriwm and probably in all the members of the group. The skull of Dinotheriwm is, in many respects, one of the most remarkable known. Although fundamentally its characters are clearly Proboscidean, nevertheless it differs widely from the skulls of the other members of the group and, indeed, in some respects from that of any other mammal. In the true Elephants and Text-figure 4. Skull of Dinotherium giganteum. From behind: about 33 nat. size. a.s., outer angle of squamosal; fm., foramen magnum; 2.f., nuchal fossa; 0.¢., occipital condyle; p.., posterior nares; p.p., paroccipital process; ptf. post-tympanic flange; 7., ridge from outer angle of squamosal. Mastodens the peculiar form of the skull is mainly due to the enormous development of cellular bone in the occipital region, increasing the area available for the attachment of the muscles necessary for the support of the heavy trunk and tusks. In Dinotherium, although the upper tusks are wanting, the trunk, judging from the large size of the nasal opening, must have been enormous, and the weight of the head was further increased by the great deflected mandibular symphysis with its large tusks. In this case, however, the area for the attachment of the support+ ing muscles was supplied by the widening out of the occipital surface, which was further increased laterally by the extension outwards of the squamosals. Little or no cellular bone seems to have been developed, the occipital surface above the post-tympanic flanges being nearly flat except for the depression for the nuchal SKULL OF DINOTHERIUM GIGANTEUM. 533 ligament. This flattening of the occiput, combined with its forward inclination, must have made it possible for the animal to move its head up and down through a large arc, a movement perhaps connected with the use of the downwardly directed lower tusks. The great width of the proboscis, probably rendered possible by the absence of upper tusks, led to the widening out of the skull in the orbital region, producing the shelf-like projection of the maxiliz above noticed, The great width of the glenoid surface for the mandible is a peculiarity for which it is dificult to account, unless it is correlated with the general widening of this region of the skull. The skeleton of Dinotheriwm is still very imperfectly known, but such bones as have been described show that the animal must have been quite Hlephantine in structure and appearance except as to its head, the legs being pillar-like and the neck short. A femur probably associated with the skull above described measured 150cm.in length. The numerous speculations as to the appearance and habits:of Dinotheriwm have been summarized by de Blainville and Stefanescu in their works referred to above. Most writers seem to suppose that the animal was chiefly fluviatile and aquatic in its habits, but there appears to be no good reason for believing that it was more so than the Elephants. The origin of Dinotheriwm is a question of much obscurity. The genus must have branched off from the main Proboscidean stem at a very early date, probably originating from an animal in much the same stage of evolution as Meritherium, the teeth being still simply bilophodont. The premolar series is more highly modified than in Meritheriwm, the series being reduced to pm. 3 and pm. 4, the latter being bilophodont like the posterior molars. This modification of the premolars, like the acquisition of the third ridge in m.1 and the reduction of the heel of m.3, may have arisen after the divergence from the majn stock had taken place. It is interesting to note that the last lower milk molar of Meritherium shows a strong tendency to trilophodonty, the posterior ridge being fairly distinct: in Dinotheriwm also this tooth is trilophodont. If, as seems almost certain, Dinotheriwm originated from some small comparatively generalized type, it is interesting to note how, after its separation from the main stock, the direction of evolution is almost the same as in the latter. Thus there was a_ general increase in stature, which, being accompanied by a shortening of the neck, necessitated, as in the Elephants and Mastodons, the development of an elongated snout ; so that it may be regarded as certain that Dinothervum passed through a longirostrine stage comparable to that of which the beginning is seen in Palwomastodon and the culmina- tion, so far as the elongation of the lower jaw is concerned, in Tetrabelodon angustidens. Subsequently the symphysis became somewhat shortened and was deflected, the lower incisors at the same time becoming enlarged. It is interesting to notice that a tendency to a similar type of modification of the mandible occurs 534 ON THE SKULL OF DINOTHERIUM GIGANTEUM. in that group of T'etrabelodonts to which Professor Osborn gives the name Rhynchorostrine, the most extreme case being L. dinotherioides. This deflexion of the mandibular symphysis led to the development of the flexible free trunk, just as the shortening of the symphysis did in the main stem, but while in this the upper incisors tended to increase in size, in the Dinotheres they disappeared. The dimensions of the skull above described (taken on the cast) are i— . cm. Length from occipital condyle to tip of premaxille. 121 Length from posterior border of palate to tip of prem axal les). eee! a OEE ea ie See 84 Width between outer angles of the squamosals ... 90 Width between outer borders of the zygomatic GOCESSOS sac CUT a tates Arua ei eave Ce ae oa, 74 Width between the openings of the ear ............ 68 Length of glenoid surface from before backwards. 6°3 Width of glenoid surface from side to side......... 24 Width of the occipital COMAVIES Ck A ee ee see 31 Width of skull just in front of the ant-orbital ROWAN Gini): Pare ps mle ol ioeule Ur asta ae aac 57 Width of the anterior end of the premaxille ...... 3] Length of molar-premolar series .........0.0,.020000+ AT Length of molar series ON THE EXTERNAL CHARACTERS OF LUTRIN A. 535 29. On the External Characters of some Species of Lutrinze (Otters). By R. I. Pocock, F.R.S., F.Z.8. [Received April 19, 1921: Read May 24, 1921.] (Text-figures 19-21.) CONTENTS. Page pis en EU na ani uunrieat, Metis we Pee, ac namareiy cena atelier Ey Tn JOB vere Wailea SE) gee nee bannices WecmabieGsae tuaanl Gna eo aL SMe Ce tReet ame te hhh Niels Mieniiane. setnmne ara The Anus and External Genitalia....................... 542 The Genera of Otters of the Old World............... 542 The Descent of the Lutring .............0.......000000... 544 Since the different species of Otters are remarkably alike in general appearance, and the published notes on their external characters, to which Gray in particular paid great attention more than half a century ago, were taken apparently from dried skins, the following illustrated descriptions ot the three species, Lutra lutra, L. maculicollis, and L. cinerea, exhibited during the past few years in the Zoological Gardens, supplemented by an examination of the softened skin of a stuffed example of donya capensis, and followed by some general conclusions with regard to the systematic status of the species examined and the descent of the Lutrine, may prove of interest. The fhinarvwm.—vhe extent to which the rhinarium is covered with hair has frequently been used since Gray’s time as one of the best external characters for distinguishing the species of Otters. In Lutra lutra it is entirely naked and a little wider than long ; but judging from descriptions is variable in shape. Gray, for example, described the lower edges as straight and the upper as “rather produced and angular.” Blanford corroborates this, saying that the ‘‘ upper edge is not straight but projects in the middle and is coneave on each side, running up considerably to the hinder edge of the nostril.” Miller, on the contrary, says the upper margin is “strongly convex at middle, the lower slightly so.” In the specimen here figured the upper edge agrees precisely with Blanford’s description, and the lower margin is angularly produced. The nostrils are widely separated and there is no philtrum, the upper lip being continuously hairy and about as deep as the height of the rhinarium. This species has a com- paratively long muzzle, as shown by the skull, and the plane of the rhinarium is inclined upwards and forwards at an angle of about 60° or less. In the Asiatic small-clawed Otter (Lutra cinerea), a short- 536 AHMaQeaeoanrPr MR. R. I. POCOCK ON THE EXTERNAL Text-figure 19. Nan WEN ff MM dN AN INS TASS F . Head of Amblonyz cinerea. . Khinarium and upper lip of the same from the front. Rhinarium of the same trom above. . Head of Hydrictis maculicoliis. . Rhinarium and lip of the same from the front. . Rhinarium of the same from above. - Rhinarium of Aonya capersis from the front (skin dried and softened), . Rhinarium of Lutra lutra from the front. . Rhinarium of Lutra lutra trom the side. . Har of Lutra lutra. All figures, except K, x4. CHARACTERS OF SOME SPECIES OF LUTRINA. 537 jawed form, the plane of the rhinarium is almost vertical. Otherwise it closely resembles that of the European species, except that its upper edge is straight transversely between the posterior edges of the nostrils and its inferior edge less produced. So far asit is possible to judge on the dried skin, the rhinarium of the African clawless Otter (Aonyx capensis) resembles that of the Indian form Z. cinerea, which is often associated with it. The other well-known African Otter (Lutra maculicollis) has a very different rhinarium, although as in JZ. cinerea its plane is vertical. It is very much reduced in size, being overgrown with hair to such an extent that little more than the edges of the nostrils are naked. The inferior border is quite straight and the infranarial portion on each side is shallow, while its upper edge is sinuously concave backwards, owing to the advancement of the hair in a curved line on each side from the posterior margin of the nostril to the middle line, where it nearly divides the rhinarium into two portions. From this it results that the middle of the rhinarium is less than one-fourth of the depth of the upper lip. The Facial Vibrisse.—The facial vibrisse in Otters are, as a rule, exceedingly coarse and numerous as in most predatory aquatic mammals. Two points to be noted, however, are that the mystacials, as compared with those of normal terrestrial mamnials, are relatively short, not longer that is to say than the genals, and that the gena! tufts, especially the inferior, are exceptionally well developed, the vibrissee being numerous and long. The material examined does not warrant the opinion that Lutra lutra, L. cinerea, and L. maculicollis differ in any way with respect to their vibrisse ; but the vibrissx of donyx capensis are decidedly less stout and stiff, especially perhaps those forming the upper genal tuft *. The Har.—The external ear in Otters is always small and oval, and comparatively simple in structure, the bursa being absent. In the European species (Zatra lutra) the tragus and antitragus are tolerably well developed, with a well-marked notch between them, and just inside the antitragus there is a strong ridge-like thickening forming the posterior rim of the cavity. The supra- tragus is a shelf-like ridge with well-developed fleshy lobe. In the Indian small-clawed Otter (Z. cinerea) the ear is very like that of Z. lutra, but relatively smaller, the cavity not extending so high above the supratragus; and the tragus and antitragus are inconspicuous. On the dried skin I could find nothing distinctive about the ear of Aonyx capensis. In L, macu- licollis the ear is simpler than in L. cinerea. The inferior edge of the cavity shows no trace of a notch, the tragal and antitragal thickenings being altogether suppressed, and the supratragus is represented merely by a rounded tubercle corresponding to the fleshy lobe of this thickening seen in L. lutra and cinerea. * Hinton (Ann. Mag. Nat. Hist. (9) vii. p. 195, 1921) has recently proposed the new generic name Paraonyx for two clawless Otters from the Congo which differ, amongst other things, from donyx in the complete suppression of the superciliary and upper genal tufts. 538 MR. BR. I, POCOCK ON THE EXTERNAL Text-figure 20. Be aaah NESGNAID Ween we Nate Nac y a 1 aa \ DA ee) Ni Olas walle V39,\ 4, Ny Sahel Pap 1 [ts % Gali (ee rhe yi ME ES at A. Right fore foot of Dutra lutra. B. Right hind foot of Dutra lutra. C. Right fore foot of Hydrictis maculicollis. D, Right hind foot of Hydrictis maculicollis. CHARACTERS OF SOME SPECIES OF LUTRINA. 539 The Feet *-—The feet of the Lutrine or Otters are so vari- able that it is hardly possible to mention a single character in which they differ from those of the other Mustelidz collectively, except that the fore paw is always much smaller than the hind. In some cases the digits are less fully webbed than in any other genus of the family except Lyncodon ; but in the typical forms the feet are more fully webbed because the digits are relatively longer and more widely separable, owing to the greater width and less emargination of the webs, especially of those joining the first and second, second and third, and fourth and fifth digits. In Lutra lutra, for example, the third and fourth digits of the fore foot are not more widely separated nor more fully webbed than in Charronia or Gulo; but the webs between the fourth and fifth, and third and second, and second and first digits extend with a nearly straight edge between points near the middle of the margins of the digital pads. The claws are well developed and project well beyond the tips of the digital pads, and the sole of the foot is naked back to the proximal margin of the single large, sometimes nearly hemispherical, sometimes transversely elongated carpal pad. The radial or internal element of the carpal pad is not separately distinguishable. There is a short, naked area between the carpal and the plantar pad; and the main part of the latter is well developed but narrow, that is to say about as wide as long, and distinctly three-lobed ; the two lateral lobes nearly meet behind the median, and the pollical element is small angl indistinctly defined from the posterior end of the adjoining lobe. The hair of the wrist extends right down to the carpal pad. Tt hind foot is markedly larger than the fore foot owing to the greater elongation of the digits, which have consequently deeper and wider webs. The plantar pad is remarkably modified. * Its median lobe is elongated and some- what heart-shaped; the external 1be, defined by a deep groove, extends far back behind the median. in contact throughout its leneth with the inner lobe, which is almost as long but has its free edge emarginate, the proximal or pollical element being large. The three main lobes just described are defined from each other by a Y-shaped groove. The hair of the metatarsus extends down to the proximal end of the plantar pad, and there is no trace of metatarsal pad. In Lutra maculicollis the digits are relatively longer than in L. lutra and the webs are better developed, extending to the distal end of the digital pads on their inner sides. The plantar pad of the fore foot is more nor mally shaped, being considerably wider than long and four-lobed., The median lobe is not narrowed behind, the _ posterior ends of the main lateral lobes are widely separated, and the pollical lobe is clearly defined. The single ulnar carpal pad is much smaller than in L. lutra; and apart from the pads which are naked, the underside of the foot is covered with fine, short scattered hairs. The hind foot differs * The carpal vibrisse are present in all the specimens described. MR. R. I, POCOCK ON THE EXTERNAL Text-figure 21, G A. Right fore foot of Aonyx capensis. B. Right hind foot of Aonyx capensis. C. Tip of digit with minute claw of hind foot of same. D. Right fore foot of Amblonya cinereus. E. Right hind foot of Amblonyx cinereus. F. External aspect of fore foot of same, showing claws. CHARACTERS OF SOME SPECIES OF LUTRINA. 54] from that of ZL. lutra in similar respects, except that the plantar pad is very poorly defined, its three lobes gradually fading away into the scantily hairy area of the metatarsus at the proximal end of the digits. The median lobe is tolerably broad in front and tapering behind, the external lateral lobe is scarcely differen- tiated from the base of the fifth digit, the internal lateral is a mere thickened ridge of skin covering the base of the second digit, and there is no trace of the internal or hallucal lobe. The skin of the metatarsus close to the proximal ends of the digits and of the plantar pad is scantily hairy, like the lower side of the webs and digits; but above it is thickly haired as in Z. lutra. In Lutra cinerea the feet differ from those of Lutra lutra and I. maculicollis in being narrower owing to the shortness of the digits and the comparative shallowness and narrowness of the webs, which reach only to the proximal end of the digital pads, those joining the second and third, and fourth and fifth digits being decidedly emarginate, although not quite to the same extent as in cursorial and fossorial Mustelide. The claws, more- over, are very short, almost erect, and quite concealed when the feet are examined from below. The plantar pads also are better developed and more normal in shape, being longer than wide, subsymmetrical and four-lobed, the inner main lobe being pro- longed by the pollical and hallucal elements. In the hind foot the outer main lobe is distinctly divided into an anterior and posterior portion by a transverse groove and almost in contact in the middle line with the internal lateral lobe. Above this pad the metatarsal area is mostly covered with naked wrinkled skin, the heel only being hairy. In the fore foot the carpal pad is relatively as large as in L. lutra, but it is defined above by a naked strip of skin over which the hair of the wrist does not extend. The lower sides of the webs and the grooves between the pads are beset with small hairs, but they are shorter and fewer than in L. maculicollis. In Aonyx capensis *, the African clawless Otter, the hind foot is very like that of L. cinerea in the development of the digits and webs; but the hair on the metatarsal area extends farther towards the plantar pad. The fore foot also is tolerably similar to that of ZL. cinerea apart from the great difference that the webs between the digits reach only to the distal end of the first phalange, the ends of the digits being free. This terminal freedom of the digits from webbing is not found in any of the cursorial or fossorial Mustelide. A. further peculiarity of this Otter is the suppression of the claws on the fore foot and on the first, second, and fifth digits of the hind foot. On the third and fourth digits of the hind foot they are merely retained as small flattish nails. * Tam greatly indebted to Mr. Fitzsimons, of the Port Elizabeth Museum, for very kindly sending me at my special request an example of this Otter from the collection under his charge. It was a mounted specimen, and the method of preservation made it impossible for me to ascertain the precise shape of the plantar and carpal pads. 542 MR. R. I. POCOCK ON THE EXTERNAL A further stage in the disappearance of the webs is shown in Paraonyz, which is described by Hinton as having unwebbed digits on the fore foot and the digits of the hind foot webbed to the base of the second phalange, apparently as in the fore foot of Aonyx. Anus and Haternal Genitalia.—About the anus and external genitalia there is very little to be said, since they closely resemble those of such typical Mustelide as Martes in having the anus encircled and the perinzum covered with hair and the two normal anal glands present, with the scent of the secretion compara- tively inoffensive. The penis of Lutra lutra is moderately long, and is supported by a stout, nearly straight baculum* about 60 mm. long, convex above, flatter and sometimes faintly grooved below, and gradually narrowed from the base up to the apex, where it expands into two short, deeply cleft, thick condyle-like branches, curved downwards. The branches are not alike, the right being a little thinner and Jess curved than the left. The urethra runs along the underside of the baculum, which is in- vested in soft vascular distensible tissue, and opens in the middle line below, just behind the condyle-like thickenings of the baculum. The baculum of Z. cinerea is very like that of L. lutra, but is of more even width throughout and the two terminal branches are not curved downwards. The Genera of Otters of the Old World. Apart from the incidental mention of Paraonya, the Otters discussed in the foregoing pages have been referred to the two genera, Lutra and Aonyx. Butif the nomenclature of this group is to be brought into line with that of other groups of Mammals, as the publication of Paraonyx seems to demand, it is certain that several generic and subgeneric terms long ago proposed but never as yet adopted will have to be restored to use. I have neither the material nor the time to go into the question beyond the limits of the species exhibited in the Zoological Gardens, of which skulls or skins, or both, are in the Society’s possession. Otters unfortunately, although so w.dely distributed, are very seldom in the live animal market, and it does not appear that more than the following species have been exhibited in the Gardens :—Lutra lutra, L. barang (entered as nair), L. cinerea, L. maculicollis, and L. platensis. Each of the four exotic species has been severed from Lutra, and received.a generic or subgeneric name. About platensis, named Nutria by Gray, I have nothing to say, the specimen having been sent unexamined to the British Museum. The Oriental species determined by Thomas as barang was referred « by Gray to the subgenus Lutrogale under the name monticola. I have no personal knowledge of the external characters ; but the * Described by Blainville, Pohl, and other authors. CHARACTERS OF SOME SPECIES OF LUTRINA. 543 skull appears to me to differ from that.of Zutra in characters which by the modern standard must be regarded as of generic value. It is much higher; the much larger orbit is thrust so far forwards that its anterior rim is above the front edge of pm. 3, thus greatly reducing the length of the muzzle, lengthening the floor of the orbit, bringing the frontal postorbital process over the middle of the upper carnassial, and lengthening the area of the cranium between those processes and the “ waist.” Added to this, the anterior nares are much less sloped, the infraorbital foramen is almost concealed in profile view by its upper bar, and the teeth are out of all proportion larger, ete. The long-clawed African Otter (maculicollis), the type of Gray’s genus Hydrogale, difters from typical Lutra in the reduction of the rhinarium and the simplification of the ear; in the rela- tively large, more fully webbed and hairy-soled feet with reduced plantar pads; and in many cranial characters, especially the shortness of the muzzle, length of the orbital floor, and the generally immature aspect of the skull owing to the feeble development of constrictions, crests, and prominences. Since the sum of the differences appear to me to be of generic value, I propose the name Hydrictis to replace Gray’s twice preoccupied name Hydrogale. ! In defence of the adoption of the name Aonyx for the African small-clawed Otter I need say nothing, since the genus appears to be admitted as valid. The Oriental small-clawed Otter (cinerea) was named Amblonya by Rafinesque in 1832. By modern writers it has been referred either to donyx or to Lutra, despite its differences from both. Tt resembles Lutra in the rhinarium, facial vibrisse, and the ear, but differs in the structure of the feet, which are narrower owing to the shorter digits being joimed by narrower webs, especially between the second and third and fourth and fifth digits, none of the webs extending beyond the proximal ends of the digital pads. The claws also are short and nearly erect or sometimes, according to Blanford, absent*; and only the calcaneal area of the hind foot is hairy. The skull also is very different from that of Zutra, and much resembles that of Zutro- gale barang in the character of the muzzle and the size of the teeth, but is altogether much shorter from the frontal postorbital processes and the molars back to the occipital crest and condyles ; the interorbital region is wider, the waist much shorter, and the orbital floor less extensive owing to the shortening of the tooth row. In external characters Amblonyx differs from Aonya, at all events, in the extension of the webbing of the fore feet to the digital pads and in the normal coarseness of the facial vibrisse. * The abbreviation of the claws of the front foot is accompanied in this Otter by extreme delicacy of the sense of touch in this extremity. I have seen one of these animals manipulating and playing with a marble in a manner recalling that of a conjuror juggling with a cricket-ball. Proc, Zoou. Soc.—1921, No, XXX VIT, 37 544 MR. R. I. POCOCK ON THE EXTERNAL Thus four genera of Otters occur in Africa: Latra lutra, north of the Sahara, Hydrogale maculicollis and Aonyx capensis, south of the Sahara, and Paraonyx congicus, etc., in the Congo area; and three in India, where they seem to be widely dis- tributed: Lutra lutra or allied species; Lutrogale barang; and Amblonyx cinerea *. Omitting Lutrogale barang, whose external characters are only known to me from descriptions which supply no evidence of difference from Lutra lutra, these Otters may be distinguished as foliows :— a. Fore feet with digits unwebbed ; hind feet webbed to proximal end of second peaeteys pete aal and nD EEE genal vibrisse sup- pressed (Hinton) . y: FI REEE CEMA e ECO CON is a’. Fore feet with ahete Hee leant Halearentede ond feet webbed at least to proximal end of third phalange; ‘superciliary and upper genal vibrisse retained. b. Fore feet webbed to proximal end of second phalange; facial vibrisse comparatively slender and soft . bapeanes Aonyx. b’. Fore feet webbed at least to proximal oil of weal ‘Gielianees facial vibrissee comparatively stout and stiff. c. Claws small or vestigial; digits short, webs narrow, reaching to proximal end of digital pads; hairs on metatarsus restricted to calcaneum ; skull short, especially behind zygomatic arches. Amblonya. ce’. Claws long, digits longer, webs wider, reaching at least half- way along digital pads; metatarsus hairy to plantar pad ; skull long behind zygomatic arch. d. Digits longer, more fully webbed, plantar pads voorly developed, especially on hind foot, carpal pad small, webs hairy below; rhinarium nearly covered with hair; ear simplified; skull with muzzle and Hebert fossa JNORH “ahheacoser . Hydrictis. . Digits short, hes fay, Ww seueat inte a ae Gell Pas Hopedi carpal pad large, webs naked below; rhinarium naked ; ear with well-developed tragus, antitragus, and supra- tragus ; skull with muzzle and mesopterygoid fossa long... Dutra. The Descent of the Lutrine. So far as I am aware, Miller is the only author who has attempted to affiliate the Lutrine definitely with one of the other subfamilies of Mustelide. ‘This is embodied in the suggestion that they “appear to be essentially badgers modified for semi- aquatic life.” Even granting the widest possible meaning to the word “badgers,” I am quite unable to agree with this view. Miller, I think, placed too much reliance upon the teeth; but it does not appear to me that the teeth of Lutra lutra, perhaps the least modified of all the Otters, are more like those of Meles than of Martes. Again, in the Otters the shape of the skull, with its flat or slightly arched crown, its short muzzle, long back, and forwardly set posterior palatine foramina, forcibly recalls that of Mustela or Grison, and is quite unlike that of J/eles, or indeed of any of the genera usually referred to the Meline, in all of which the skull has a long jaw and short back and, except in the * T have not examined the type of Gray’s genus Barangia, namely sumatrana, which is, I believe, the only other admitted species of Old World Otter, # ee ee ee Se a Se Ss ee ee ee a ee CHARACTERS OF SOME SPECIES OF LUTRIN®. 545 Skunks, the palatine foramina in question placed on the maxillo- palatine suture. Both Mustela and Grison are predatory terrestrial forms; and on @ priort grounds it seems to be much more probable that the Otters are the descendants, modified for aquatic life, of some group of active, predatory terrestrial Mustelide than of com- paratively inactive, specialized diggers like the Badgers. This view is supported by the long, slender, flexible body, the long tail, the broadly webbed, well spread digits, small claws and small front paws. These characters are found associated only in the genera currently referred to the Musteline, but not in any of the so-called Melinz, which have a short thick body, short tail, closely tied toes, powerful claws, especially on the fore feet, and the latter are larger than the hind feet. Another point worth considering is the presence in a few unrelated species—i.e. Lutra canadensis, according to Gray, and in Hydrictis maculicollis and Amblonyx cinerea—of hairs upon the lower side of the webs of the feet. It seems to me to be difficult to explain their occurrence on the hypothesis of the descent of the Otters from naked footed Meline burrowers. The hairs seem to be quite functionless, and they have all the appearance of harmless vestigial structures suggesting the descent of Otters from hairy-footed forms. And since the Martens, Wolverenes, and Weasels, in the broad sense of the word, are the only members of the Mustelide, apart from the Otters just mentioned, which have hair upon the webs below, these hairs may be regarded as an additional item of evidence pointing to those active predatory Mustelide as standing nearest to the extinct ancestors of the Otters. Tt is not of course suggested that the Otters are the descendants of any of the Martens or Weasels known to us, but that they are an offshoot from that line of descent; and I visualize their ancestor as an active terrestrial, predatory Mustelid with a well-developed tail and feet with the longish digits provided with short, sharp claws and united by wide webs, hairy below but not so hairy as in Martes or Mustela, and with the pads probably as large as in Charronia, with the skull shaped like that of Mustela or G'rison, the auditory bulla like that of MJartes and also the dentition, except that the upper carnassial had a larger inner lobe such as is seen in Grison. The enlargement of the hind feet as compared with the fore feet, the thickening of the facial vibrisse, with the correlated expansion of the upper lip and enlargement of the suborbital foramen, the reduction and simplification of the ear, the texture of the fur, and the muscular development of the tail, are special adaptions to aquatic life. Similarly, the enlargement of the back teeth, recalling in a measure that of the Melinz, may be looked upon as an adaptation for breaking up fish-bones and rendering them harmless to swallow. The explanation of the great difference in dentition between 37* 546 ON THE EXTERNAL CHARACTERS OF LUTRINA. the Otters and the Seals may be explained in this way. Otters masticate their fish and require broad, flat-crowned teeth for the purpose. Seals swallow them entire, and the teeth are modified for catching and holding the slippery prey. ; [Nore.] After this paper was written, Mr. Thomas drew my attention to the elaborate monograph of the Lutrinz published by Pohle in November 1920 (Arch. f. Naturg. 1919, pt. 9, pp. 1- 247). So far as the species I have discussed are concerned, Pohle’s systematic results, based upon skulls, teeth, and dried skins, differ im two points only from mine. While admitting the genera Amblony« and Aonyax* as distinct from Lutra, he granted Latro- gale barang only subgeneric rank under Lutra, and dismissed maculicollis and two related species as ‘‘ Die maculicollis-Gruppe” of Lutrat. Since the properly constituted subgenera and “oyoups” of species of one generation of workers almost invariably become the genera of the next, the fulfilment of their destiny in that respect by Lutrogale and the maculicollis-group may, I think, be confidently predicted. I see no reason, therefore, for making any change in the text of this communication. * Pohle missed Paraonyx although he had the description of congica before him and apparently plenty of material of the genus in the skins and skulls of the species he described as Aonya microdon. This remark, however, is not intended as any disparagement of Pohle’s work, which appears to me to be exceptionally good. In addition to the genera and species, it deals with geographical distribution, cranial changes with growth, milk-dentition, and so forth. + Similarly, Pohle established “ Die Swmatrana-Gruppe” [= Barangia Gray} and “Die platensis-Gruppe” for the American Otters canadensis, annectens, enudris, platensis, provocax and felina—a group tor which the name Lontra Gray appears to be the oldest, with Latawina and Nutria as synonyms. ‘ % ‘ ‘ / i / " \ * . t ; : \ fi r a . 6 2 é i, ‘ er i ‘ or y ‘ ; ‘ i 4 : ? z ao : : : i f * ’ * oe / ’ » 3) a \y a7 o e ‘a ‘ aes j . ts * PP, 4, S. igi, SOWNWAG, iil, WwW. COMPARATIVE ANATOMY OF THE KOALA (PHASCOLARCTOS) IP Zo So Ii, SOI TAG, Il, Wil a eat | ’ \ COMPARATIVE ANATOMY OF THE KOALA (PHASCOLARCTOS) : ON THE ANATOMY OF THE KOALA AND VULPINE PHALANGER. 547 30. The Comparative Anatomy of the Koala (Phascolarctos cinereus) and Vulpine Phalanger (Zrichosurus vulpe- cula), By CuHartes F. Sonntag, M.D., F.Z.58., Anatomist to the Society. [Received May 7, 1921: Read June 7, 1921.] (Plates V., VI.; Text-figures 53-63.) The observations recorded here are based on the examination of an adult male Zrichosurus vulpecula which died in the Society’s Gardens, and a young female Phascolarctos cinereus which Mr. A. F. Richardson, of Hull, sent to us for examination after death. The measurements of the Koala are as follows :— Centimetres. Length from tip of nose to root of tail......... 50 i CUNT AVELICG) po AU" cet pre aa i aR Pee Sa he RO Ae 12°5 Oly tami a Pans er AM kas se ee alts 2°5 Greatest, widtheotimead, 22, e204. 0Gasecceeeacec es 75 Length of fore limb (arm, 9°5; forearm, 13:5; TAVATATES AO) acs eee seed ashe Wri kee HSER Sa. e 5:0 275 Length of hind limb (thigh, 11°5; leg, 12; [OSIRIS ids. d Sus BO ROSE AnH Oey cnr ar ries ee eee 28:5 The external characters presented nothing new. Sir Everard Home, who wrote the first account of the internal anatomy of the Koala, in 1808, described the character of the glandular patch of the lesser curvature of the stomach, and the prostate gland which resembles that of the Kangaroo; he also recorded the absence of the patella (5). Highteen years later Dr. Robert’ Knox, of Edinburgh, described the long cecum for the first time, and named the Koala the ‘‘ Wombat of Flinders ” in honour of its discoverer. He also criticised the writings of Home, Illiger, Cuvier, Desmarest, and De Blainville. Since then accounts of the anatomy of the entire animal have been written by Martin (8), Forbes (4), and Young (18), and indi- vidual organs have been described by Flower (3), Owen (11), Macalister (7), and Chalmers Mitchell (9). The histology of different parts has been described by Oppel (10) and Braus (1). THe ANATOMY OF THE Heap AND NECK. Phascolarclos cinereus. When the skin of the head is removed it is seen how the platysma myoides, muscles of mastication, auricular muscles, and those round the orbit and mouth are strongly developed. The platysma myoides avises posteriorly from a strong panni- culus carnosus. It has a clearly-defined line of attachment to 548 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY the lips, and some of its fibres fuse with those of the muscles of the angle of the mouth and orbit. It covers the anterior part of the masseter muscle and the facial vessels. It has been divided along the lines 7.7. in text-fig. 53. The masseter muscle (text-fig. 53) has been described by Macalister. It has a tendinous lower border (6), from under cover of which the facial artery (a) emerges and runs antero- superiorly ; it gives off an external nasal branch (c) which runs along the antero-syperior border of the cheek-pouch (CP). The masseter is crossed by the anterior facial vein (d), transverse facial artery (f), and Stensen’s Duct (e). Text-figure 53. The side of the head of Phascolarctos cinereus. j.k. posterior auricular vessels; 7. parotid gland. Other letters in text. Macalister has described the attolens, attrahens, and retrahens muscles which act on the large auricles. The anterior facial vein has the usual origin, course, relations, and branches. It unites with the posterior facial vein on the surface of the exoccipital process (g) to form the external jugular vein. No part of it is seen in the interramal space of the mandible, and the vena transversa (text-fig. 54, 12), which is long and V-shaped, connects the external jugular veins (11) instead of the anterior facials; 10 receives lingual, muscular (3, 18), sub-maxillary (21), and a single submental vein (2), but no anterior jugular vein connects it to the internal jugular vein. The Parotid Gland is large, and forms a prominent mass in the face and neck. Running on its surface are the posterior auricular vessels. The neck is divided, as usual, into mesial and lateral triangles by the sterno-mastoid muscle (text-fig. 54, 10), and these have the usual boundaries. The base of the lateral triangle is formed OF THE KOALA AND VULPINE PHALANGER. 5A9 by the clavicle (17) as in all Marsupials except the Peramelide, which have no clavicles. In the space between the mesial borders of the sternomastoids there is a thin muscular sheet formed by the mylo-hyoid (1) anteriorly, the digastrics (9) antero-laterally, and the sterno- hyoids (13) posteriorly ; the sterno-hyoids have become separated from the hyoid bone in this specimen, but the whole sheet plays over it. Anterior to the digastrics are the submaxillary glands (4). Text-figure 54. The superficial anatomy of the neck of Phascolarctos cinereus. 5 and 6. anterior and posterior facial veins; 7, 20. exoccipital process; 8. posterior auricular vein; 11. external jugular vein; 16. supraclavicular nerves; 17. clavicles; 19. masseter. Other numbers in text. The space between the sheet and the sterno-mastoids is bridged over by a triangular fascia (14) which covers the common carotid artery, vago-sympathetic nerve-trunk, and the nerve to the pretracheal muscles. It does not cover the internal jugular vein, which is under cover of the sterno-mastoid. Trichosurus vulpecula (text-fig. 55). The superficial anatomy differs from that of Phascolarctos in the following respects :— 1. The platysma is not so well developed. 2. The external nasal artery does not follow a cheek-pouch. 3. The auricular 550 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY muscles are not so powerful. 4. The vena transversa is horizontal, and runs between the anterior facial veins. It is connected to the internal jugular veins (#) and receives two submental veins ; a small part of it loops into the interramal space. ‘5. The sterno-hyoid muscles are attached to the hyoid bone. 6. The superficial sheet between the sternomastoids of Phascolarctos is absent. Neither in Phascolarctos nor in Trichosurus does the vena transversa receive laryngeal veins. Text-figure 59. The superficial anatomy of the neck of Trichosurus vulpecula. Numbers as in previous figure. The deep anatomy of the neck differs greatly in the two species :— Phascolarctos cinereus. When the muscular sheet described above is removed the conditions shown in text-fig. 56 make their appearance: The walls of the mouth (/) and pharynx (g) lie immediately under cover of the muscular sheet, and the following structures are seen to lie on their external surface :— 1. Two long, narrow strap-like genio-hyoid muscles (a, @’) running from the symphysis menti to be attached to the hyoid bone (m) and larynx (/). ‘They cross the lingual arteries (c). 2. The lingual arteries. 3. The hypoglossal nerves (d). 4. The pharyngeal branches of the vagi (e), which are very tortuous. OF THE KOALA AND VULPINE PHALANGER. 551 A short interval separates the insertions of the genio-hyoid and sterno-thyroid muscles (7), which arise from the back of the sternum and third costal cartilages. At the posterior part of Text-figure 56. The deep anatomy of the neck of Phascolarctos cinereus. J. level of opening of nasal tube into the pharynx; &. thyro-hyal capping the shaded basi-hyal (m). Other letters in text. the neck the lateral lobes of the thyroid gland (A. A) project from under cover of the sterno-thyroids, and extend for a short distance dorsal to the clavicles, 552 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY Macalister (7) described the styloid group of Biaees as follows :—“ The styloid group form a single sheet, the hind fibres of which pass to the pharynx, the middle to the ceratohyal under the stylohyoid ligament; the anterior pass to the side of the tongue.” The digastrics have the usual attachment, and they are united by muscular fibres to other muscles around. The stylo-hyoid ligament is loose (text-fig. 56, h). The genio- hyoids are connected to the digastrics by Hinsemlnn slips (b). Text-figure 57. The deep anatomy of the neck of Trichosurus vulpecula. D. sterno-mastoid; E. digastric; k. hyoid bone; m. common carotid artery; m. external jugular vein; o,, vena transversa; qg. anterior jugular vein ; 7. s. thyroid and laryngeal arteries. Other letters in text. The cervical lymphatic glands form two groups. The first, consisting of a few large giands, is a little posterior to the prominent alisphenoid bullae The small glands of the second group are around the great veins at the root of the neck. The distance from the symphysis menti to the level of the angles of the mandible is 6 em., from the angles of the mandible to the hyoid bone 1:2 cm., and from the hyoid bone to the sternum 6 cm, OF THE KOALA AND VULPINE PHALANGER. 553 Trichosurus vulpecula (text-fig. 57). The deep anatomy of the neck differs from that of Phasco- larctos in the following respects :—1. The walls of the mouth and pharynx are more deeply placed under muscles. 2. True hyoglossus muscles are present (B), so the lingual artery (¢) and hypoglossal nerve (¢) are more deeply placed. 3. The thyroid gland is applied to the posterior part of the larynx and is far anterior; it does not pass dorsal to the clavicles. 4. The alisphenoid bulle are not so prominent. The exoccipital pro- cesses are not visible. 5. The external jugular vein is formed on the surface of the sternomastoid muscle. THe RESPIRATORY ORGANS. In Phascolarctos the epiglottis is broad and entire, with an emarginate apex, and each lung contains two lobes; Forbes (4) described two left and three right lobes. No azygos lobe is present. In Zvrichoswrus the epiglottis is notched ; the left lung has two lobes and the right cne has four, of which one is azygos. The tracheal rings are wide, but deficient dorsally. The nasal tube of Phascolarctos has no median dorsal ridge, and does not grip the larynx, whose epiglottis is so large that it could only be accommodated by considerable stretching of the orifice of the tube. In Trichosurus the nasal tube has a median dorsal ridge, and it gripped the larynx firmly in my specimen. Sir Everard Home states that the Koala utters a shrill yell when hungry or hunted, but Owen says that he never heard a note of any kind from the Phalangers. THe Moura. Phascolarctos cinereus. The Vestibule.—Fleshy lips are connected to the gums by lax frenums, of which the upper is ‘6 and the lower is 1:2 em. long. The cheeks are attached to the gums opposite the canine tooth and posterior border of the premolar of the upper jaw on each side; between these lie the mouths of the cheek-pouches (Plate V.,A) already described by Owen (11), Young (13), and Forbes (4). Forbes states that “the mucosa of the cheeks is smooth throughout”; but in my specimen it is covered with rounded papille, which are all visible to the naked eye. They extend on to the fauces, and for a short distance along the outer walls of the cheek-pouches, but they are absent from the lips whose mucosa is quite smooth. They are not regularly arranged like the papille of the cheeks of the Ungulata. 554 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY The vestibule and cavum oris communicate through the diastema anteriorly, and behind the last pairs of molar teeth posteriorly ; the former of these is wide and the latter narrow. The Palate (Plate V., A).—The upper incisor and canine teeth surround a small area on which there are six small tubercles, — but these do not cover prominences on the premaxille. Imme- diately behind these lies the first of the nine palatal ridges, which are most marked anteriorly. Forbes also described nine ridges, and Young recorded eight. Martin described eight furrows. ‘The anterior ridges are transverse, and are complete or incomplete, but some of the posterior ridges are curved and bear small nodules on their anterior borders. When the mucous membrane is removed the bony palate is seen to have a sinuous outline when viewed laterally (text- fig. 59, HK). The soft palate is long, smooth, and devoid of a uvula. The isthinus faucium (text-fig. 59, f) is a circular orifice which admits the tip of the index finger ; and the fauces form a septum without the tonsils, which lie far back in the pharynx at a distance of 6 cm. from the incisor teeth. Trichosurus vulpecula (Plate V., B). Cheek-pouches are absent, and the vestibule and mouth com- municate as in the above. On the palatal area surrounded by the incisor and canine teeth, which form a more perfect arcade than in Phascolarctos, there is a median ridge and several small, hard papillee. The palatal ridges, of which there are six, are sharper than in Phascolarcios, and the anterior ones are convex forwards, with the free edges directed backwards. Between several of the ridges the mucosa is covered with papille. The tonsils have the usual situation in the fauces. Pseudochirus peregrimus (Plate V., C). There are eight palatal ridges, most prominent anteriorly, and many papille crowd the mucosa between them. Cheek-pouches are absent. Tue Toncve (text-fig. 58). Martin (8) states that the tongue of Phascolarctos rises abruptly from a furrow SuLWe wag its base, and the distance from its root to the epiglottis is $ of an inch. I found, however, that it is much farther removed, the distance being slightly more than 13 inches, and the mucosa of its dorsum is continuous with that of the ventral wall of the pharynx (text-fig. 59, A,@). Measurements :—Total length 5:5 em.; length of the oral OF THE KOALA AND VULPINE PHALANGER. 555 part 5 em.; length of the pharyngeal part ‘5 cm.; greatest width 15cm. The slightly spatulate narrow tongue does not narrow greatly from base to apex. The Papille.—A single prominent vallate papilla (text-fig. 58, A) lies in the mid line; it is circular on plan and conical on elevation, with the base of the cone projecting beyond the well-marked fossa ; the vallum is rough, granular, and not easily distinguished from the rough surface of the tongue. Martin, Owen, and Forbes each describe a specimen with one papilla. The fungiform papille ave small, and form a dorsal bounding zone on which they have the usual arrangement in clusters and rows; none are overlapped by the conical papille, which are small, pointed, and irregular in distribution. On the ventral papillary zone only conical papille are found. Although there are no lateral organs similar to those of the Primates, the posterior extremities of the corrugated lateral borders have circular nodules with well-marked central orifices (text-fig. 58, I); some of these are single and isolated, but others are in small chains. Poulton (16) has described similar bodies in Halmaturus walabatus, Macropus melanops, and Petro- gale wanthops, but points out that their arrangement varies ; they may be irregular or arranged in curves. He draws attention to their appearances being similar to those of gland-ducts, and sugeests that the lateral organs arise from glands; “ this view,” he states, ‘is confirmed by a study of minute structure.” No Apical Gland of Nuhn is present, but the glandular mass, as described and figured by Oppel (10), sends two long processes forward almost as far as the apex of the tongue. He believes that the Gland of Nuhn, which is present in the tongue of Man and .the Orang-Outan, are pieces of the basal glandular mass which have become separated. Perhaps they have become separated off from a prolongation similar to those of the Koala and other Marsupials. The trferior surface is wrinkled and fissured. In the mid line in front there is a well-marked rough area, bounded laterally by fissures. Posterior and lateral to the central area are crenated, transversely-furrowed plice fimbriate (text-fig. 58, B), which increase in width from before backwards. Many Marsupialia have a sublingua which differs from that of the Lemuroidea in being firmly fixed by its apex to the under surface of the tongue. It varies in size in different species, but that of the Koala is the smallest; it has been reduced to the condition of plice fimbriat in the same way as the sublingna of the Lemuroidea gives way to the plice of the Simiide. A few illustrations of the degrees of development in the Marsupialia are shown in text-fig. 58, B~-H. The sublingua also becomes softer as well as reduced in size, as in the Primates. Thus there is the complete horny sublingua as in Lemur, the softer and smaller sublingua as in Varsiws, and the plice fimbriate as in -Anthropopithecus. 556 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY Text-figure 58. Trichosurus Pseudochirus vulpecu/a. PETegrinUs. i Dd. Didelphys Dasyurus Perameles. Dendrolagus. wirgimana. hellucatus. gunn. inustus. (Ei is G H, P2060 ea 200 02° anc J. A. /. The tongues of some Marsupials. A. and B. Phascolarctos cinereus; C-H. inferior surfaces of the tongues; I-K. lateral organs of Phascolarctos, Trichosurus, and Pseudochirus respectively ; L-N. vertical sections of the tongues of Phaseolaretos, Trichosurus, and Pseudochirus showing degrees of freedom of the sublingua. OF THE KOALA AND VULPINE PHALANGER, 557 Trichosurus vulpecula. Poulton (16) described the tongue under the old name of Phalangista vulpina, so I shall only mention the differences between it and that of Phascolarctos :—1. The epiglottis overlaps the base. 2. Three papille form a vallate triangle. 3. The lateral organs consist of fissures and lamine. 4. The sublingua is a flat horny plate with a strong median ridge and a denticu- lated apex bound down to the under surface of the tongue (text-fig. 58, C). 5. The two forwardly-directed glandular processes are stouter and relatively shorter. Pseudcchirus peregrinus. The pavrallel-sided tongue has the following dimensions :— Total length 4:2 em.; length of the oral part 3:7 em.; length of the pharyngeal part °5 em.; width between the lingual attach- ments of the palato-glossal folds ‘4 cm. The Papille.—Three circular vallate papille form an equilateral triangle with the apex behind; the basal papille are prominent, but the apical one is strongly retracted. The fungiform papille, which are small and few in number, form a dorsal bounding zone, but they have no definite arrange- ment in transverse rows and clusters; they are surrounded by sharp conical papille, which form transverse rows almost up to the apex of the tongue. Only conical papille are present on the ventral bounding papillary zone. No orvfices of ducts and glands are seen on the base of the tongue. The Lateral Organs (text-fig. 58, K).—On each lateral border of the tongue there is a row of circular elevations with central slit-like orifices, so they are not quite similar in appearance to those of Phascolarctos. The slits run upwards and backwards. Each organ consists of six areas. On the surrounding mucosa there are many nodules and papille. The Sublingua (text-fig. 58, D).—The most pronounced feature of the tongue is the large, soft sublingua. It is firmly bound to the under surface of the apex, as in all Marsupialia in which it is complete, but its under surface is deeply fissured and separated by a deep gap from the tongue (text-fig. 58, D and N). Running backwards in the mid line is a strongly-developed median crest, which is bounded behind by a fissure. There are great differences, therefore, between the lateral organs, ventral surfaces, and vertical sections of the tongues of these three species, as seen in text-fig. 58, B-D and I-N. The essential differences between the sublingue of the Marsupialia and Primates lie in the apices—in the former the apex is fixed to, but in the latter it is free from, the under surface of the tongue, 558 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY THe PHARYNX. Phascolarctos cinereus. The distance from the fauces to the beginning of the cesophagus is 4°5 cm., and the greatest width of the pharynx, which narrows from before backwards, 1s 2°2 em. The mucosa is smooth, and two ridges run along the sides of the mid-dorsal line (text- fig.) OO He): Text-figure 59. Mouth Fauces. Pharyn x. | Ocesophagus The pharynx of Phascolarctos cinereus. Explanation of letters in text. The éonsils (text-fig. 59, c) are small oval bodies in the dorso- lateral aspect of the pharynx, a little distance anterior to the hyoid bone. They appear as pits on the interior of the pharynx, and the right one is surrounded by a prominent ridge. The naso-pharynx.—The nasal fosse are prolonged backwards by a conical tube, which opens on the mid-dorsal wall of the pharynx by a small circular orifice guarded by two fleshy cushions (text-fig. 59, Ad and C). It has no mesial dorsal ridge as in OF THE KOALA AND VULPINE PHALANGER. . 559 Trichosurus and Pseudechirus. The opening lies within the hyoid ring about 5 mm. anterior to the beginning of the cesophagus, so the naso-pharynx is exceedingly small. his is a remnant of the condition present in the new-born animal. Owen (11) describes how the larynx is gripped by the nasal tube in the mammary foetus of the Kangaroo, and the milk stream is divided by the larynx, which crosses the pharynx; he gives illustrations. Sections through the head and neck at different levels show how the shape of the nasal tube alters from before backwards (text-fig. 59, F and G). The Hustachian Tubes (text-fig. 59, b).—The lower orifices of the tubes are slit-like, and lie on the infero-lateral aspects of the nasal tube at a distance of 2:9 cm. posterior to the hard palate, and at a point about 1 em. posterior to the fauces. In Plate V. pins are inserted into the inferior orifices of the Eustachian tubes, Trichosurus vulpecula. The pharynx differs from that of Phascolarctos as follows :— 1. The tonsils are in the fauces. 2. There is a ridge on the dorsal wall of the nasal tube, but none in the pharynx. 3. The larynx crosses the pharynx and enters the nasal tube. The nasal tube of Pseudochirws has a median dorsal ridge (Plate V., C). THE CisopHaaus. Phascolarctos cinereus. The cervical part (4:3 cm. long) is entirely concealed by the larynx and trachea, and its mucosa is smooth throughout. The thoracic part (7:8 em. long) has small longitudinal folds on its inner surface. The abdominal part (1:8 cm. iy is lined by mucosa thrown into prominent longitudinal folds. At the point where it emerges from the posterior surface of the liver it is in contact dorsally with the right suprarenal capsule. Trichosurus vulpecula. The esophagus projects beyond the left border of the trachea in the neck, and it has a longer course (=3 cm.) in the abdomen than has that of Phascolarctos. It is not in contact with the right suprarenal capsule. In some examples its abdominal part is much shorter, however. Tur PERITONEUM. Phascolarctos cinereus. The great omentum is devoid of fat, but contains a well-marked arterial arcade formed by a branch of the splenic artery and a Proc, Zoot, Soc,—1921, No, XX XVIII, 38 560 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY twig from the arterial ring round the pyloro-duodenal junction (text-fig. 60, B, a’). There is nothing new to record about its attachments. The lesser omentnm has a T-shaped attachment to liver and stomach (text-fig. 60, A, @). The horizontal limb lies along the lesser curvature, and the vertical limb runs back along the ventral surface of the stomach. The great omentum of Trichosurus vulpecula is well laden with fat. THE STOMACH. Phascolarctos cinereus. The observations recorded here are supplementary to the descriptions by Martin (8), Young (18), Forbes (4), Edelmann (2), Flower (3), and Home (8). The long axis of the sac-like stomach, which lies in the epigastrium and left hypochondrium. runs posteriorly, ventrally, and to the right, and the well-marked fundus is directed forwards and to the left. Projecting from the greater curvature imme- diately below the pyloric cylinder is a sac whose walls are thinner than those of any other part of the stomach. Relations :— Anteriorly—iiver and right suprarenal capsule. Posteriorly—spleen, intestines, solar plexus, left kidney, and suprarenal capsule. Right—duodenum, gall-bladder. Left and ventral—abdominal parietes. Dorsal—crura of the diaphragm, aorta, pancreas, right kidney, inferior vena cava. Measurements :— Nobtal lengthy. 255%. trades tees aeeenere a eng th sot fumals: ficctseteseus Aen ee 2 Width between curvatures ............ Be 2 1 Arteries (text-fig. 60).—The stomach receives its blood supply from the three branches—gastric (@’), hepatic (6), and splenic (c) —of the celiac division of the cceliaco-mesenteric artery (see page 570), and each vessel is accompanied by fine nerves derived from the vagus and solar plexus (text-fig. 60, A). The gastric artery divides into a large ventral branch (d) and two dorsal branches (e)—large and small. The former, after a straight course towards the stomach, breaks up into many tortuous vessels ramifying over the fundus, and some can be traced into the left side of the gland-patch. The latter make a similar vascular arrangement on the dorsal aspect of the f undus, and supply the patch as well. OF THE KOALA AND VULPINE PHALANGER, 561 The hepatic artery gives off a large gastric branch (G. A.) which divides into four—two dorsal (g.g) and two ventral— branches (h. 2). Each pair supplies one surface of the stomach, Text-figure 60. Ovesophagus. Saccu/us, CA. Oesophagus. The ventral (A) and dorsal (B) surfaces of the stomach of Phascolarctos cinereus with arteries. 38" 562 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY and forms half of a vascular ring encircling the pyloro- duodenal junction. They supply the gastric wall and gland-patch. The pyloro-duodenal ring communicates with the mesenteric arteries, and a twig forms a long arcade (777) in the great omentum with a branch of the splenic artery ; this arch gives off vasa brevia, which enter the stomach-wall vertically (text-fig. 60, B). There is no continuous vessel or arcade on the lesser curvature, for the gland-patch interrupts it. The veins enter the portal and splenic veins. The gastric nerves (text-fig. 63, V.G.P.) form a complex plexus on the left side of the gland-patch, and are intermingled with branches of the gastric artery, and fine nerves reach the stomach along branches of the hepatic artery. They will be described later along with the solar plexus and vagus nerves. The nerves reach the interior along the course of the arteries. The interior of the stomach (Plate VI., A) was briefly described by Martin (8). Apart from the enumeration of the orifices of the gland-patch by several authors, no other description of the interior has been published. The long axis of the oval gland-patch, which lies along the lesser curvature of the stomach, is 2 cm. long, and the width is 1:7 em. Its inner surface has thirty orifices of different sizes, arranged in rows at right angles to the long axis of the patch. In addition to these the lens reveals how the surface of the patch is finely-pitted. The remainder of the inner surface is traversed by ruge radiating from the edges of the gland-patch like the spokes of a wheel. Those running towards the lower end of the cesophagus are short and Y-shaped; those passing into the fundus and body of the stomach are at first straight but undulating later ; none are present in the sacculus on the greater curvature ; and those passing to the pylorus are few in number, but stronger than in other parts. The musculature of the pylorus thins out on the walls of the sacculus, and that of the ceesophagus thins out on the fundus and lesser curvature. The gastrie contents consisted of a viscid creamy chyme, and they resembled the contents of the stomachs of two Langurs which I examined. In the Sloth, on the other hand,as I showed in a previous paper, the stomach contained many hard, almost entire leaves. Special gastric glandular apparatus is also present in the Wombat, Beaver, Dormouse, Manatee, and Dugong. It was also present in the now exterminated Steller’s Sea-Cow (Rhytina stelleri). Knox thought that the apparatus in the Sirenia was an electric organ, and traced the two vagus nerves into it. The pylorus projects into the duodenum (Plate VI., B), as in Man. The human form has been likened to the projection of the cervix uteri into the vagina, The free surface is rough and villous, ‘ OF THE KOALA AND VULPINE PHALANGER. 563 Trichosurus vulpecula. The stomach differs from that of Phascolarctos in the following respects :— 1, It is more globular. 2. There is no sacculus on the greater curvature, but the stomach projects to the right of the pylorus. 3. The gland-patch is not present on the lesser curvature. 4, The pylorus is not such a well-defined cylinder, and does not project into the duodenum as in Man and Phascolarctos. 5. The walls of the stomach are relatively thinner. 6. The mucous membrane is rough, but has no ruge. In a hitherto unpublished note Garrod states that the fundus of Cuscus maculatus has a reticulum of ruge. Tur INTESTINES. Phascolarctos cinereus. The external and internal characters have been fully described by Forbes and Chalmers Mitchell (9), and I have nothing to add to their accounts. My measurements differ greatly, however, from those recorded by various authors. The figures in the following table represent inches, and the asterisks denote that fresh specimens were examined. Lonnberg’s specimen was a mammary fetus*. It due allowance is made for shrinking induced by preserving fluids, it will be seen that the lengths of the different parts of the intestinal tube vary greatly in indi- viduals, and have no relation to the length of the animal :— ache Ae z es ae bith Seen es. SET * Knox (1826) ...... 23 92 128 17 Martin (1836) 2. not given not given 76 50 Owen (1837)...... 23 92 125 a *Horbes (188). oF 115 93°25 46°75 Forbes (1881) 9. 20°5 st) 160°8 66 Lonnberg (1902). 3°5 14°55 11°45 34 *Sonntag (1921)... 21 54:5 (5) 36 Lonnberg (17) compared the measurements of the different parts of the intestinal tract of Phascolurctos and several other Marsupials, and he also compared those of the feetal and adult intestines, taking Forbes’s figures as the standard for the latter. He came to the following conclusions :—“. ... the difference is quite striking with regard to the cecum and the large intestine. The former is proportionately only about a third as long in the foetus as in the full-grown animal, and the latter less than half * The mammary foetus is the new-born animal fixed to the nipples in the pouch, 564 DR. C. F, SONNTAG ON THE COMPARATIVE ANATOMY as large in the fetus as in the adult. The difference of the small intestine in both stages is not so great, that of the foetus being about four-fifths of the same in the adult. It is also to be remarked that in the feetus the small intestine is considerably longer than the colon, but in the adult the reverse is the case. The differences can of course be ascribed to the difference of the diet of both stages. The milk food of the mammary fetus is chiefly or completely digested in the small intestine, but the vegetable diet of the adult needs a greatly developed cecum and colon, The longitudinal folds of the cecum and colon are, however, already developed in the feetus.” Tt is possible that the growth of the intestinal tract in my specimen has been considerably retarded. Ellenberger, Tullberg, and Lonnberg believe that the large size of the cecum in many animals, such as the Horse, Koala, and some Rodents, is brought about by the presence of a large amount of cellulose in the diet; and when the small intestine deals with all the elements of the food, there is no specialisation in the cecum and colon. Lonnberg also speculates about the great length of the small intestine in phyllophagous Mammals. He states that the leaves consist of cellulose and certam proto- plasmic, amylaceous, and other substances contained within resistant envelopes. Some time must elapse, therefore, before these are broken down and their contents become available for digestion and absorption; the small intestine is, consequently, long. Lonnberg’s views, however, will not account for the conditions present in the intestinal tract of the phyllophagous Three-toed Sloth (Bradypus tridactylus). In my paper dealing with its anatomy (18), I showed that the intestinal tract is short, the cecal pouch is rudimentary, and the interior of the colon and rectum offers a large absorptive area, due to the excavation and folding of the mucous membrane; it possesses sphincters which are similar in character-to, but larger than, those of the cecum of Trichosurus vulpecula, and its muscular coat, which thickens from before backwards, may have a similar function to that of the cecum of 7’. vulpecula. Although the stomach of 86. tridactylus contains many hard, almost entire leaves, the intestinal tract contains none, so the cellulose and hard materials have been completely comminuted and reduced to the state of chyme; Rapp’s analogies are interest- ing in this connection, for he likens the paunch and pylorus to a crop and gizzard. This thorough digestive process must throw less work on the small intestines, which, in consequence, are short. The chyme yields a number of substances for absorption in the small intestines, and the remainder is absorbed from the large intestines, whose pockets and sphincters retain it in the colon and rectum, and the muscular coat moves it backwards and forwards; and the sphincter ani has ‘a retaining function similar to that of the septa higher up in the colon. OF THE KOALA AND VULPINE PHALANGER. 56d The differences between these two phyllophagous Mammals must therefore be due to complexity of the stomach and the different varieties of specialisation of the interior of the colon and rectum, but the increased internal surface of the colon and rectum of the Sloth is compensated for by the large size and great length of those of the Koala. The Langurs are inter- mediate in the complexity of their intestinal tract between the Sloth and Koala. In Phascolomys mitchelli the cecum is considerably reduced, and Lénnberg believes that the colon has become considerably enlarged to take on the cellulose-digesting function of the ceecum. The stomach is simple as in the Koala, so it has not had any modifying influence on the complexity of the intestinal tract as in the Sloth. The Wombat has external sacculi, but the Sloth has internal cavities, so there is a greater similarity between their colons than exists between the colon of either of these animals and that of the Koala. I have already shown how the complexity of the stomach is also influenced by the degree of development of the buccal organs (19). One must take a wide view of the whole alimentary canal from mouth to anus in framing an explanation of the physiology of any one part. THe Urinary Orcans (text-fig. 61). Phascolarctos cinereus. The right kidney is 3:2 cm. long, 1:8 cm. wide, and 1:6 em. thick ; it lies entirely anterior to the left kidney, whose corre- sponding measurements are 3°7, 2°4, and 15 cm.; ts hilum is ventral in position, whereas that of the left kidney is mesial. The structures in the hilus have the usual disposition in regard to one another. On section each kidney is seen to possess a thin cortex, and a single large papilla. The surface of each renal pelvis and hilus is covered by a plexus of small veins. These communicate posteriorly with the veins ascending along the ureter, and they open anteriorly into the renal vein. The right ureter lies parallel to, but 1 em. to the right of, the vena cava inferior. It passes ventral to the iliac vessels and runs posteriorly along the dorsal pelvic wall. It opens into the lower and back part of the bladder. It is accompanied by small veins from the pelvis; the latter are joined by veins from the ovaries, after which a Jadder-like arrangement of veins runs up to the venous plexus on the ventral surface of the renal pelvis (text-fig. 61 C, wv). The left wreter runs along the left border of the abdominal aorta, crosses the iliac véssels, and courses through the pelvis as does the right one. The veins accompanying it are numerous 566 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY and some of them ramify; so dense is the venous plexus that the ureter is concealed by them. At first sight it looks like a cord of veins. John Hunter (22) showed that the relative positions of the kidneys of Petaurus taguanoides are similar to the above. Text-figure 61. A. heart and aorta of Phascolarctos cinereus with origins of right subclavian, right common carotid, left common carotid, and left subclavian arteries, from» right to left, and described in text as a, b, c, d: B. heart and aorta of Trichosurus vulpecula: C. kidneys and some of the abdominal veins: ~ L.S.R.C. and R.S.R.C., suprarenal capsules; P.L., peritoneal ligament: D and E, origins of vertebral arteries in Phascolarctos cinereus and Trichosurus vulpecula respectively: F and G. spleens of Trichosurus vulpecula and Phascolarctos cinereus respectively. ——- OF THE KOALA AND VULPINE PHALANGER. 567 The bladder is pyriform and firm, but no urachus is seen. Lateral peritoneal folds connect it to the pelvic wall, but no other ligaments are present. Trichosurus vulpecula. The anterior half of the right kidney. is received into a deep recess in the liver, and its posterior pole is level with the hilus of the left kidney. The bladder is transversely corrugated on the surface and no urachus is present. Only one wide renal papilla is present in each kidney. GENERATIVE ORGANS. I have nothing to add to existing accounts of the generative organs of either species. Tur PANCREAS. The duct is dead-white in Phascolarctos and shines through the translucent brown glandular tissue. The latter in my speci- men has the same appearances as those described by Martin (8). Tue Liver. . r Phascolarctos cinereus. Forbes (4) described and figured the liver, and Braus (1) described its histology. The hepatic artery is very convoluted, and divides into two terminal convoluted branches which diverge at the sides of the gall-bladder. From the left branch a straight, slender cystic artery runs in the free edge of the vesical mesentery to the gall- bladder, which it reaches about its middle; then it runs poste- riorly along the surface of the gall-bladder as far as the fundus. The postcaval vein is never deeply embedded in the hepatic substance, and receives three hepatic veins at its emergence from the anterior surface of the liver. A fourth vein enters it within the organ. The common bile-duct does not exhibit cysts of its mucous membrane similar to those of some Marsupials described by Owen (11). Trichosurus vulpecula. _ The liver differs from that of Phascolarctos as follows :—1. It is not so compressed antero-posteriorly. 2. Its edges are not notched by many small fissures. 3. The hepatic tissue is not so subdivided. 4. The right lobe recesses the anterior half of the right kidney. 5. The gall-bladder is shorter and broader, and only projects for a short distance into the abdominal cavity 568 DR. C, F, SONNTAG ON THE COMPARATIVE ANATOMY beyond the posterior edge of the right lobe when fully distended with bile. THe Ductuess GLANDS. Phascolaretos cinereus. The Thyroid Gland (text-fig. 56, A, A).—The two lateral lobes are unconnected, and vary in shape and extent. The lefé lobe is pyriform, with the broad extremity posterior and slightly within the thorax. It is 2 em. long, *8 em. wide, and °3 cm. thick. It is separated from the larynx by the first three tracheal rings. The right lobe is elongated and ovoid, and measures 27x'‘5x:3 em. Its anterior extremity overlaps the lower end of the larynx, and its posterior pole stops anterior to the thorax. The arterial supply comes from the thyroid and laryngeal branches of the common carotid artery (m), and the veins run to the internal jugular vein (7). I did not observe any structures which could be definitely described as parathyroid bodies, but they may have been lying among the small glands which are numerous at the root of the neck. The Thymus Gland has been described hy Symington (15). The Suprarenal Capsules (text-fig. 61, s..c.). —The left capsule is a round body, of the size of a large pea, lying on the mesial aspect of the anterior pole of the left kidney, to which it 1s attached by a fold of peritoneum. The right capsule is oval, and has a peculiar position; it lies in contact with the posterior surface of the liver, and is situated between the inferior vena cava to the right, the right crus of the diaphragm to the left, and the esophagus ventrally. The greater part of each capsule is composed of brown medulla. The spleen has been described by Martin (8). My specimen is triangular and measures 7°4 cm. long, but that of Zrichosurus vulpecula has a lateral process (text-fig. 61, Fand G). In Phascolomys it is an equilateral triangle. Trichosurus vulpecula. The lateral thyroid lobes (text-fig. 57 A) measure 1:3 cm. long and *3 cm. wide. ‘They lie at the sides of the posterior extremity of the larynx and anterior seven tracheal rings. The connecting isthmus crosses rings 5, 6, and 7. The thyroid artery describes a curve, with the convexity forwards, and runs along the anterior five-sixths of the mesial border of its corresponding lateral lobe. The thyroid vein runs to the internal jugular vein(p). Thisarrangement of the thyroid vessels is seen in animals belonging to several of the Mammalian orders. OF THE KOALA AND VULPINE PHALANGER. 569 Tur ORGANS OF CIRCULATION. Phascolarctos cinereus. The Heart (text-fig. 61, A).—The fibrous pericardium is firmly bound to the dorsal Surface of the sternum and left ribs by connective tissue. It narrows out and is lost on the great vessels. These conditions are also present in J'richosurus vulpecula. The area of adhesion is greater than that of Dendrolagus ursinus. The long axis of the heart is antero-posterior and parallel to the left side of the sternum, and the apex is level with the posterior border of the fourth left costal cartilage. Its apex is more obtuse than that of 7'richosurus vulpecula * Measurements :— Phascolarctos. Trichosurus. Meme this. 2 uay hehe 5: 4-6 cm. ae em. Greatest width ...... DWART os Dahan Greatest thickness ... 1:9 ,, Toes. The characters agree with those already described for the Marsupial heart. The Ascending Aorta (text-fig. 61 A) is embraced by the two parts of the right auricular appendix, and is parallel to the right precaval vein. It quickly gives way to the aortic arch, whose most anterior point is level with the anterior end of the first intercostal space. At the level of the fourth dorsal vertebra it gives way to the descending aorta; this is the commonest level for the transition in the Mammalia. The relations of the arch are :— Ventral.—Left lung, left precaval vein, left azygos vein, left vagus nerve, left phrenic nerve. ‘Dor sal.—Trachea, cesophagus. Posterior.— Root of left lung. _ Anterior. — Great vessels of head, neck, and fore limbs. Branches :—The right subclavian (a), right common carotid (4), and left common carotid (c) arise directly, and close to one another, from the aortic arch, so there is no innominate artery. After a short interval the left subclavian artery is given off (d). Martin (8) and Owen (11) describe an arrangement similar to that of Man, but Forbes (4) states that the innominate artery gives off three vessels close together from a common trunk. In Trichosurus vulpecula the aortic arch gives off the innominate and left subclavian arteries, and the former, after giving off the left common carotid, divides into right common carotid and right subclavian arteries; the left common carotid arises almost at the beginning of the innominate ar tery. The Common Carotid Arteries give off fine entocarotids at the level of the anterior border of the hyoid bone, and are continued * Some specimens of Trichosurus vulpecula have a more obtuse cardiac apex than the above. 570 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY as stout ectocarotids. The former give off no branches in the neck, and enter the carotid foramen in the basisphenoid. The latter pass antero-laterally between the exoccipital process and the mandible, and the former separates it from the external jugular vein. In Trichosurus vulpecula the exoccipital process does not separate these vessels. The Descending Thoracic Aorta lies first on the right half of the body of the fifth dorsal vertebra, but passes more to the right as it runs backwards through the thorax. It is connected to the esophagus by a triangular fold of pleura, whose base is attached to the diaphragmatic pleura. It has the usual relations and gives off the usual branches. The vertebral arteries come off from the subclavians, but in my specimen of Trichosurus they arise from axes (text-fig. 61, E). The Abdominal Aorta differs from the common Marsupial condition in that it is entirely to the left of the postcaval vein. Hochstetter (20) examined several species, and found that the postcaval vein concealed the lower end of the aorta in all but Petaurus taguanoides, but he did not include Phascolarctos cinereus in his list. Beddard (21) also found the posteaval vein covering the aorta in all his species. My observations on 7'richo- surus vulpecula and Dendrolagus ursinus confirm those of Hoch- stetter and Beddard. Relations.—At the point where the abdominal aorta reaches the level of the posterior surface of the liver, it is separated from the right suprarenal capsule by the right crus of the diaphragm. These structures separate it from the postcaval vein. A little posterior to the capsule the solar plexus lies between the aorta on the left, the postcaval vein on the right, and the coeliaco-mesenteric artery ventrally. Posterior to the solar plexus the postcaval vein lies immediately to the right side of the aorta till the latter divides into the internal iliac and mid-sacral arteries, and then the vein crosses the surface of the aorta. The whole trunk of the aorta is visible. The relations to other structures show nothing unusual. Branches.—With the exception of the four pairs of lumbar arteries which arise at regular intervals from the sides, the following is the order of the branches from before backwards :— Phrenics, cceliaco-mesenteric, suprarenals, renals, ovarians, external iliacs, internal iliacs, and middle sacral (caudal). After giving off the external iliac arteries, the trunk of the aorta diminishes in calibre to a little more than half of that of either of the external iliac arteries. The Celiaco-mesenteric Artery divides into celiac axis and anterior mesenteric artery, and a branch of the latter replaces the posterior mesenteric branch of the abdominal aorta of the higher Mammalia. The celiac axis divides into gastric, hepatic, and splenic arteries, whose distribution has already been described. OF THE KOALA AND VULPINE PHALANGER. 571 The anterior mesenteric artery describes a curve, convex forwards and to the right, and terminates in the cecum; it occupies the free edge of the cecal mesentery in the latter part of its course. The branches supplying the large intestines arise from the convexity of the curve, and those destined for the small intestines are given off from the concavity. The highest branch anasto- moses with the duodenal branches of the hepatic artery, and the lowest one anastomoses in the pelvis with the hemorrhoidal vessels. The lowest branch to the ileum anastomoses with the branches to the cecum. The mesenteric areas between the branches of the mesenteric artery contain rich networks of fine lymphatic vessels, and fine branches of the solar plexus accompany the arteries. The mesenteric veins have been figured already by Chalmers Mitchell (9). In Trichosurus vulpecula the celiac axis arises separately. Tur VEINS. The Anterior and Posterior Facial Veins have the usual forma- tion, course, and tributaries. They unite to form the external jugular vein on the surface of the exoccipital process. The latter runs postero-mesially across the sternomastoid to reach its lateral border in the posterior sixth of the neck. It passes dorsal to the posterior end of the sterno-mastoid and clavicle, and, anterior to the neck of the first rib, it unites with the axillary vein to form the precaval vein. Both veins are united by a long V-shaped vena transversa. In Trichosurus vulpecula the vein is more transverse (text-fig. 55). Tributaries.— Auriculo-tem poral, acromial, transverse scapular, muscular, ascending cervical and internal jugular veins. ‘hese vessels have the usual distribution. No anterior jugular vein is present. _ The vena transversa veceives veins from the tongue, sub- maxillary glands, and muscles of the mylo-hyoid region. At the point where the internal and external jugular veins unite there are several small glands. The innominate and precaval veins follow the usual course of the Marsupial type, and the azygos veins in my specimen are similar to those described by Forbes (4). The intrathoracic part of the postcaval vein measures 1-5 em. The postcaval vein is formed in the usual manner, but the only part lying ventrally superficial to the abdominal aorta crosses the termination of the latter. This arrangement is an exception to Beddard’s statement that “the Marsupials show a very constant condition in that the postcava lies medianly ventral to the aorta, so that on dissection the aorta is not seen, being completely covered by the postcava ” (21). It lies to the right side of the aorta, and close to it till it yf DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY reaches the right renal vein. After that it bends to the right and enters the liver. At the point where it disappears from view it is in contact with the right suprarenal capsule, and is accompanied by twigs from the right half of the solar plexus. The left ovarian vein (text-fig. 61 C, ov.) joins the thick venous plexus running along the ureter; therein it is united to veins ascending from the pelvis. The right ovarian vein communicates with the right ureteric veins by a stout trunk, and then runs forwards to join the postcaval vein. The ureteric veins (uv. and wv.) form a plexus on the ventral surfaces of the renal pelves and open into the renal veins (rv.). Tue Neryous System. The brain has been described by Owen (11), Forbes (4), Elliot Smith (12), and Ziehen (14), and I have nothing to add to their accounts, Tur Vacus Nerves (text-figs. 62 and 63). Phascolarctos cinereus. A short distance posterior to the base of the skull the vagus nerve on each side receives the hypoglossal (XIT.) nerve, and their fibres intermingle in a very thorough manner. Soon they separate again, and the vagus runs postero-mesially as far as the anterior extremity of the thyroid cartilage, where it unites with the cervical sympathetic cord(C.5.) to form a vago-sympathetie (V-S.). A short distance anterior to the clavicle it separates again into vagus (V.) and sympathetic (8.). This arrangement is present in most of the Carnivora. It differs from that of Dendrolagus, in which the vagi and sympa- thetics remain separate. There is no trace of the ganglion of the trunk of the left vagus (ganglion nodosum) in the neck. The right nerve has one (G.N.). Branches in the neck.—Pharyngeal (a), anterior laryngeal (6), communicating to the first cervical nerve (c), and recurrent branch of the right vagus (d). The relations of the vagiin the thorax differ. The left nerve erosses superficial to the left subclavian artery and aortic arch, lying on the inner side of the left precaval vein and left phrenic nerve, and on the outer side of the trachea, cesophagus, and left common carotid artery. It passes dorsal to the root of the left lung and gains the dorsal aspect of the cesophagus. The right nerve crosses the right subclavian artery under cover of the right precaval vein, runs down the side of the right main bronchus, and gains the dorsal aspect of the root of the right lung; lastly it passes to the dorsal aspect of the cesophagus, where it fuses with the left nerve to form the combined trunk of the vagi (C.T.V.). OF THE KOALA AND VULPINE PHALANGER, 573 Branches of the separate nerves in the thorax.—Left recurrent laryngeal nerve (e), cardiac (f), pulmonary (g), and a branch from each nerve which unites with its neighbour to form a fine cord, which runs posteriorly along the ventral surface of the cesophagus (/). The combined trunk of the vagi enters the abdomen dorsal to the cesophagus, runs dorsal to the stomach, and ends in the left Text-figure 62. a--2- + se e- Qs te ee b.=)= =i ‘ ‘ : b= 22 ine The vagus nerves of Phascolarctos cinereus in neck and thorax. half of the solar plexus. It gives off ventral gastric nerves in the thorax (7), and stout ventral (j), and strong dorsal (4), gastric nerves in the abdomen. It is possible that sympathetic filaments run anteriorly from the solar plexus to the stomach in the combined trunk of the vagus, for it has been shown that sympathetic filaments run in other nerves (¢. g..the sciatic). 574 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY The two halves of the solar plexus (text-fig. 63) he between the abdominal aorta and postcaval vein under cover of the origin of the cceliaco-mesenteric artery. The right half (A) is a fusiform ganglion, but the left half (B) consists of a nerve Text-figure 63. The vagus nerves and abdominal sympathetic of Phascolarctos cinereus. t=} J plexus with a small round ganglion. The combined trunk of the vagi (C) enters the left half of the plexus by a brushwork of nerves. ‘The right and left sympathetic nerves (D and K) end in the right half of the plexus, but the left sympathetic gives off OF THE KOALA AND VULPINE PHALANGER. 575 a fine nerve (F), which enters into the formation of the aortic plexus farther back (G). Fine fibres (H) connect the halves of the plexus. Branches of distribution :— 1. Three nerves which course along the posteaval vein and embrace the right suprarenal capsule (1). 2. Hepatic plexus (J). 3. Splenic plexus (K). 4, Right renal plexus (L). 5. Two cords (M and N) which emerge from the posterior end of the right half of the plexus and form a loop lying dorsal to the postcaval vein. The left one breaks up into a number of fibres which reunite later with the right one, and it gives off the left renal plexus (QO). The loop unites with the left sympathetic by a T-shaped junction (P), and fine filaments (Q, Q, Q, Q) also run between them. Only one cervical sympathetic ganglion one—exists, namely the superior. SUMMARY. The differences between Phascolarctos and T'richosuruws may be summarised as follows :— Phascolarctos. ' Trichosurus. 1. Platysma myoides strong. 1, Platysma weaker. 2. No true hyoglossus muscle. 2. Hyoglossus present. 3. Vena transversa A-shaped. 3. Horizontal. 4, A. 5 5 . Absent. . Anterior jugular vein present. . Lateral organs of tongue look like . Composed of fissures and lamine. elands. 6. One vallate papilla. 6. A vallate triangle. 7. Plice fimbriatz present. 7. Sublingua well developed. 8. Cheek-pouches ‘present. 8. Absent. 9. All anterior palatal ridges trans- 9. Anterior palatal ridges convex for- verse. wards. 10. Tongue far from epiglottis. 10. Close to epiglottis. 11. Epiglottis broad and entire. 11. Narrow and notched. 12. Gland-patch in stomach. 12. Absent. 13. Pylorus projects into the duodenum. | 13. Pylorus does not project. 14, Tonsils far back in the pharynx. 14. Tonsils in fauces. 15. Pharynx has two dorsal ridges. 15. Absent. 16. Absent. 16. Dorsal ridge in nasal tube. 17. Cardiac apex obtuse. 17. More acute. 18. Absent in my specimen. 18. Innominate artery present. 19. Postcava crosses end of abdominal | 19. Postcava conceals posterior half of aorta. aorta. 20. Liver more subdivided. 21. Gall-bladder long and narrow. 21. Short and wide. 22. Right kidney posterior to liver, and | 22. Sunk in liver, and each kidney has each kidney has one conical papilla. one broad, flat papilla. 23. Thyroid gland in posterior part of | 23. In anterior part. neck. Proc, Zoot, Soc.—1921, No. XX XIX, 39 576 DR..C. F. SONNTAG ON THE COMPARATIVE ANATOMY Phascolarctos. Trichosurus. 24. No thyroid isthmus. 24. Isthmus present. 25. Patella absent. . Patella thin and scale-like. 25 26. Absent. 26. Right lung has an azygos lobe. 27. Two lobes in each lung. 27. Left lung has two, and right has four lobes. 28. Spleen triangular, and has no lateral | 28. Spleen has a lateral piece. process. 29. Vertebral arteries arise from sub- | 29. Vertebral arteries arise from an clavian arteries. axis on each side. 30. Intestinal measurements much greater. Owen stated that the Koala is the Marsupial which has the greatest claim to typical pre-eminence. MacAlister, on the other hand, showed that it has more individual peculiarities in its myology than has any other Didelphian; and the remarks made in this paper show how many of the appearances of its internal organs are peculiar to itself, for several of those described for Prichosurus vulpecula are similar to the conditions present in many Marsupialia. There are also many striking differences between the external and skeletal characters of Phascolarctos and those of Tvichosurus, but I have nothing new to add to existing descriptions. BIBLIOGRAPHY. 1. Braus, H.—Jen. Denkschr. Bd. 5 (Semon, Zool. Forschungs- reisen, Bd. 2). Jena. G. Fischer, 1896. 2. Epe~Mann, T.—Deut. Zeitschr. f. Thiermed. Bd. xv. pp. 165-214. 1899. 3. Frower, W. H.—‘ The Comparative Anatomy of the Organs of Digestion of the Mammalia.” Medical Times and Gazette, 1872, p. 645. 4. Forpes, W. A.—‘‘On some Points in the Anatomy of the Koala.” Proce. Zool. Soc. London, 1881, pp. 180-195. 5. Homs, E.—‘‘ An Account of some Peculiarities in the Anatomical Structure of the Wombat.” Phil. Trans. Roy. Soc. 1808, pp. 304-312. 6. Knox, R.—‘‘ On the Wombat of Flinders.” Edinburgh New Philosophical Journal, vol. i. p. 104. 1826. . 7. MacAutstrer, A._—‘‘ The Muscular Anatomy of the Koala.” Annals and Magazine of Natural History, 1872, pp. 127- 134. 8. Martin, W.—“ Notes on the Anatomy of the Koala.” Proc. Zool. Soc. London, 1836, pp. 109-113. 9. MircuEett, P. CHoatmers.—‘ Further Observations on the Intestinal Tract of Mammals.” P.Z.S8. 1916, pp. 195- 19%. 10. Orret, A.—Text-book of Comparative Microscopic Anatomy, vol. i. p. 291, and vol. i. p. 308. 11. Owen, R. Art. « Marsupialia” in Todd’s ‘ Cyclopedia of Anatomy and Physiology,’ vol. i. ‘ ! q : : ; 12. 13. 14. 15. 16. re 18. 19: 20. 21. 22. A. Qa OF THE KOALA AND VULPINE PHALANGER. iy Suirn, G. Exxror.—Journal of Anatomy and Physiology, EXxili. 1898, p. 30. Youne, A. H.—Journal of Anatomy and Physiology, xii. 1880, pp. 305-317. ZienEN, T'.—Jenaische Denkschr., Bd. vi. 1897, p. 98. Symineron, J.—Journal of Anatomy and Physiology, xxxiv. p. 226. Pouttron, E. P.—‘On the Tongues of the Marsupialia.” Proc. Zool. Soc. London, 1883, pp. 599-628. Lonnperc, E.—‘‘On Digestive Adaptations in the Marsu- plata. Pawns. 1902.1. p. 12. Sonnrac, C. F.—‘‘ A Contribution to the Anatomy of the Three-toed Sloth.” P.Z.S8. 1921, pp. 157-177. Sonnrac, C. F.—‘‘The Tongues of the Cercopithecide.” P.Z.S. 1921, pp. 280-281. Hocusretrer, F.—Morphol. Jahrb. xx. 1893, p. 626. BepparbD, F. E.—‘“‘ On the Postcaval Vein and its Branches in certain Mammals.” P.Z.S. 1909, pp. 496-526. Hunter, Joun.—‘ Essays and Observations. Edited by Richard Owen. 1861, p. 257. EXPLANATION OF THE PLATES. Prate VY. The Palate of Phascolarctos cinereus with the cheek-pouches packed with wool and pins in the Eustachian Tubes. . The Palate of Trichosurus vulpecula. . The Palate of Pseudochirus peregrinus. Prate VI. The Stomach of Phascolarctos cinereus. . The interior of the stomach. . The projection of the pylorus into the duodenum. cr “I Je} ON NEW INDIAN DRILID BEETLES. 31. New Indian Drilid Beetles. By 8. Mavuix, F.Z.S., Professor of Zoology in the University of Calcutta. [Received May 11, 1921: Read June 7, 1921. | (Text-figures 1-5.) In the present paper four beetles from India of the Family Drilide are described. They are all characterized by the pos- session of extraordinarily developed lamellate terminal joints of both the maxillary and labial palpi. These joints, of both pairs of palpi, are similar in structure, and are of almost equal size. They naturally fall into two genera, and can be distinguished thus :— Head almost as long as the pronotum, so that the eyes are distant from the front edge of the pronotum. Antennee distant ............ecsssececeecsesetereeesasess, Lamellipalpus, gen. 0. Head short, the eyes almost touching the front edge of the pronotum. Antennze more or less approxi- THU C Meee sneer aee ven aa eee Re EBs esac LOMeELlipalpodes, Senn Ml: LAMELLIPALPUS, gen. n. The type of this genus is Hugeusis nigripennis Pascoe, from Burma. At the end of his description of this insect, (Ann. Mag. Nat. Hist. xx. 1887, p. 10, t.i.f.7) the author remarks: ‘“ The description is from a female; the male has probably flabellate antenne, as in #7. palpator Westwood” (the type of the genus Hugeusis). Ihave examined Pascoe’s type; it is not a female, therefore his conclusion that the male has flabellate antenne is not correct. Under this assumption he included his nigripennis in Westwood’s genus Hugeusis, of which I have seen a specimen in the collection of the British Museum, and with which it is not congeneric. It is necessary, therefore, to erect a new genus for the reception of this species as well as two others in the collection of the British Museum. Generic characters.—The antenne are 11-jointed, filiform; sometimes each joint is slightly expanded on the inner side. They are separated from each other by the whole breadth of the head, and are situated in front of the eyes. The eyes are convex and prominent. The mandibles are large, curved, and pointed; in repose they lie crossed. The maxillary and labial palpi are 4- and 3-jointed respectively, the terminal joint in each being very large, elongate, and compressed. The front coxe are large and exserted, having a large trochanter ; the intermediate and hind coxe are contiguous, but not so large as the front ones. The tarsi are 5-jointed, the fourth being bilobed. The elytra are not of a membranous texture and the suture is closely fitting. 580 PROF. S. MAULIK ON LAMELLIPALPUS NIGRIPENNIS Pascoe. The following description and illustrations are from a fresh specimen taken “by Mrs. Kemp in Assam :— Colour testaceous brown; underside paler; elytra blackish, except an ill-defined basal area, the suture and the Perel margins; the antenne except the two basal joints blackish ; eyes deep black. The upper side of the head and prothorax is shining, the rest of the body being subnitid. The whole body of the insect is covered with fine brown pubescence. Text-figure 1. Lamellipalpus nigripennis Pascoe. Head almost as broad as long, slightly narrowed towards the neck, shallowly depressed in front; upper surface smooth, shining, spar sely covered with fine brown hei Viewed from aber e, the extraordinarily developed terminal joints of the palpi and the large mandibles are conspicuous. The joints of the antenne are not at all expanded on the inner side. Pronotwm broader than long, front margin straight, lateral and posterior margins slightly sinuate, anterior lateral angles rounded, posterior ones very acute. The upper surface is gently convex, shining, covered with fine brown pubescence, more so on the sides than on the middle area, deeply depressed at the base on either side of the longi- | : NEW INDIAN DRILID BEETLES. 581 tudinal middle line. Zlytra as broad at the base as the pro- thorax; the humeral angles are rounded. Corresponding to the basal depressions of the pronotum, the opposite elytral areas at the base are also depressed. The surface is punctate, covered with brown pubescence. Lach elytron has four feeble coste, the two inner ones—z. ¢e., those nearer the suture—being more prominent than the outer ones; the two inner ones commence at the base and terminate before reaching the apex; the third, commencing posterior to the humerus, terminates about the middle; the fourth, commencing below the humerus, terminates Text-figure 2. Underside of Lamellipalpus nigripennis Pascoe. just beyond the middle. Of the four cost the third is the feeblest ; to a certain extent they converge towards the apex. Scutellum triangular, light brown, pubescent, impunctate, sub- nitid. Underside brown, but lighter than the colour of the upperside; the abdominal segments have a slight suffusion of black. The surface is sparsely covered with fine brown hairs, their roots give the surface an appearance of being finely punc- tate. The abdominal segments are at a lower level than the metathoracic segment; each is convex in the middle and de- pressed at the sides, these lateral depressions forming an oblique line on either side. The underside of the femora and tibie is not suleate. Length 10 mm., breadth 43 mm. Length of head 2 mm., 589 PROF. Ss MAULIK ON breadth of head across the eyes 3 mm., without the eyes 23 mm. Length of pronotum 23 mm. Assam, Garo Hills; Tura, 1400 ft.; Oct. 1917 (Mrs. Kemp) (Indian Museum). The type specimen (Brit. Mus.) is from Burma. The brown parts in the type are lighter than Mrs. Kemp’s specimen ; also in the type the lighter colour of the suture and the lateral margins of the elytra is not so well marked. LAMELLIPALPUS BOMBAYENSIS, Sp. n. Upperside shining, dark brown, in one specimen of the two lighter brown; the antennal joints from the third joint to apex Text-figure 3. Lametlipalpus bombayensis Maulik. brownish-black, and the apical area of the elytra obliquely con- tinuing to the middle along the side smoky black; eyes deep black. Underside shining light to dark brown. The whole insect, both on the upper and lower sides, is covered with light brown pubescence. Head almost as long as broad, depressed in the middle in front but convex round the roots of the antenne ; surface punctate, each puncture having a stiff hair. The eyes strongly convex. NEW INDIAN DRILID BEETLES. 583 The first joint of the antenne club-shaped, the second very short, the third to the tenth almost equal in size and slightly expanded triangularly towards the inner side, the last joimt bluntly rounded. In the expanded character of the several joints of the antenne it bears a certain amount of resemblance to Hugeusis palpator Westwood, in which these points are extraordinarily expanded into rami. Prothorax about as broad as long, slightly narrowed towards the front; the posterior lateral angles pro- duced to acute points; surface convex, with a faint longitudinal impression along the centre and deeply sloped on each side ; this, with a similar slope on the elytra in front of the humerus, forms a deep depression ; closely punctate. Scutellum triangular, punc- tate. Hlytra as broad at the base as the prothorax. The lateral margins and suture costate. On each elytron there are three more or less prominent coste; one runs close to the suture, the second along the longitudinal middle line, and the one nearer the margin starts from a point below the humerus, a feeble one, corresponding to the third in Lamellipalpus nigripennis Pascoe, is observable in certain lights. These coste approach one another towards the apical area, where they disappear. The surface is irregularly punctate. Underside shining, smoother than the upperside. : Length 9-10 mm. Bombay. Described from two examples, the larger of which I have made the type specimen. It has the apical black patch on the elytra diffused, while in the other it is more sharply defined. In the type specimen the lamellate joints of the palpi are broader, with apex slightly narrowed, while in the smaller specimen they are uniformly narrower. Type in the British Museum. LAMELLIPALPUS MANIPURENSIS, Sp. n. Upperside subnitid, thorax and head more shining than the elytra ; elytra and eyes black, the four lamellate terminal joints of the maxillary and labial palpi, and the antennal joints except a part of the two basal joints, smoky black ; the rest of the body, including the underside, brown. The lamellate joints of the palpi are narrower and more elongate in this species than in the two others of this genus. Head almost as broad as long; interantennal space shallowly depressed, covered with light brown pubescence. The first joint of the antennz is almost as long as the third joint but very slightly thicker, second joint small and rounded, third joint slightly club- - shaped, fourth to tenth almost equal to each other in length and thickness, the eleventh slightly lorger than each of the preceding ones and rounded at the extremity; the joints of the antenne are not at all expanded. Prothorax slightly broader than long, sides and front margin straight, basal margin sinuate; the 584 PROF. 8. MAULIK ON . posterior exterior angles acutely produced, the extreme edges of the produced part sharply depressed. The pronotum is covered with light brown erect pubescence. Scutellwm brown, triangular, covered with light brown pubescence. lytra almost as broad at the base as the prothorax. The surface is confusedly punctate and covered with light brown pubescence. The three cost on each elytron are very feeble though distinctly recognizable. Underside: the elytra project to a certain extent beyond the abdomen, shining, covered with stiff brown hairs. Length 7:5 mm. Manipur (Doherty). Deseribed from one example. Type in the British Museum. LAMELLIPALPODES, gen. n. The genus differs from Lamellipalpus in the length of the head and in the approximation of the antenne, as has been indicated in the key given above. The type of the genus is the following species. Although I include the two new genera described here in the family Drilide, I am not sure about the position of Lamelli- palpodes, gen.n. When I examined Lamellipalpus nigripennis Pascoe (Garo Hills, Mrs. Kemp) and Lamellipalpodes annandalet Maulik at Calcutta, I was impressed by the fact that they both possessed the remarkable lamellate palpal joints but did not appear to. belong to the same family. Dr. C. J. Gahan thinks that Lamellipalpodes Maulik does not belong to the Drilide, but is more allied to the Rhagophthalmide, a neighbouring family of about three genera, including nine species. Without further research it is‘not possible to decide this point. LAMELLIPALPODES ANNANDALEI, sp. n. Upperside yellow-brown, underside lighter; antennz except the two basal joints, and the elytra except an ill-defined basal. portion (which shares the colour of the pronotum) smoky black. Eyes deep black. The whole body is covered with brownish pubescence. Head narrower than prothorax, the interocular space piceous. The first joint of antenne stout and large, second joint very small and rounded, the third to the tenth similar in structure but of gradually diminishing length, the eleventh bluntly rounded. Prothorax broader than long, with a longitudinal median sulca- tion; posterior angles acute, pointed, anterior ones rounded ; basal area depressed on either side of the middle longitudinal line, basal margin bisinuate. Scutellwm triangular, with the apex rounded, brown, smooth. Zlyira as broad at the base as the prothorax. Corresponding to the prothoracie depressions, the elytra are also depressed at the base. The surface neither | | : = NEW INDIAN DRILID BEETLES. 585 smooth nor regularly punctate-striate. Each elytron is feebly tricostate; the costz are more visible if the insect is held at certain angles; the first—that which is nearest to the suture— proceeds from the apex of the basal declivity of the elytra, the _Text-figure 4. Lamellipalpodes annandelei Maulik. Text-figure 5. 5 Bagche: Underside of Lamellipalpodes annandelei Maulik. second commences from the edge of the elytral black area, the next commences about the middle of the elytron, each of these terminating on the apical area of the elytron- Length 53 mm. 586 ON NEW INDIAN DRILID BEETLES. ' Described from five examples. Type in the British Museum. - Cotype in the Indian Museum. Of the five examples four belong to the British Museum, and these have only ‘‘ Bengal” on the locality labels; the specimen belonging to the Indian Museum was taken by Dr. N. Annandaie from the following locality :— Barkuda Island, Chilka Lake, Ganjam district, Madras Pre- sidency, 22. vii. 1920; found on bath-room window. In this specimen the cost of the elytra are more pronounced, In preparing this paper I have had the advantage of seeing specimens in the British Museum collection. I wish to thank Dr. C. J. Gahan and Mr. K. G. Blair for their courtesy, valuable suggestions, and criticisms. I wish to thank also Dr. N. Annan- dale and Dr. and Mrs. Kemp for kindly letting me examine their respective insects. The drawings of two insects were made by Mr. 8. C. Bagchi, of the Indian Museum, Calcutta, and of one by Mr. A. J. E. Terzi; to these artists my thanks are also due. The preliminary study of these interesting beetles was made in the Indian Museum, Calcutta, while I held charge of its Entomological Section. ON SOME ABNORMALITIES IN THE CARNIVORA, 587 | 32. On some Abnormalities in the Carnivora. By CHarEs F. Sonntac, M.D., F.Z.8., Anatomist to the Society. [Received May 10, 1921: Read May 10, 1921.] (Text-figures 64-65.) The following abnormalities were seen in animals dissected in the Society’s Prosectorium during the past two years, They consist of :— 1. Elongation and contortion of vessels and nerves in the neck of a Bengal Fox (Canis bengalensis). : 2. The occurrence of two separate precaval veins in a Panda (Ailurus fulgens). 3. Abnormal modes of termination of the vertebral veins of a Common Badger (J/eles meles). These conditions were compared with the normal state of the vessels, which is identical in all three animals. The Normal Blood-vessels. The aortic arch gives off the innominate and left subclavian arteries, and the former divides into the right subclavian and right common carotid arteries after giving off the left common carotid. The common carotid arteries run antero-laterally, and then straight forwards. The internal jugular veins are on their lateral aspects, and the vago-sympathetic nerve-cords lie on their dorsal surfaces. The external jugular veins unite with the subclavian veins to form the innominate veins, of which the left one runs obliquely across the anterior part of the thorax and joins the more vertical right vein to form the precaval vein, The vertebral veins open by single trunks into the innominate veins. The vena azygos major opens into the right side of the precaval vein immediately anterior to the root of the right lung. Abnormalities in a Bengal Fou (text-fig. 64). The animal, which was young, poorly-nourished, and rachitic, had a symmetrical and uniform enlargement of both lateral thyroid lobes (text-fig. 64, A.A), and the thymus gland was well- developed. The former were about twice the size of the lobes in a normal adult male. The isthmus of the thyroid gland was normal, The innominate artery (B) divided in the normal manner, but the right subclavian artery was concealed by the right innominate vein (C). The left common carotid artery (D) has a more horizontal course than the right one (EK), and describes a very acute flexure with the convexity to the left. The right common carotid artery is not only flexed, but recurved, with the result that there is an S-shaped flexure. The left vago-sympathetic cord runs a 588 Dh. G. F, SONNTAG ON SOME straight course on the dorsal surface of the left artery, but the right cord (F) follows the right carotid in its wanderings, and appears on the surface for a short distance. No growths or adhesions are present in the root of the neck which could be responsible for distortion of the vessels and Text-figure 64. aa st RUNEUEEH SESS —s The vessels of the neck and thorax of Canis bengalensis. For explanation of letters see Text-figure 65. nerves. All other vessels and nerves are normal as regards their course. The precaval vein (1) receives internal mammary (N), peri- cardiac (J), and azygos veins (M), and a vein from the diaphragm running along the surface of the left phrenic nerve (1). No phrenic vein accompanied the right phrenic nerve. 1 ; ABNORMALITIES IN THE CARNIVORA,. 589 The Occurrence of T'wo Precaval Veins in a Panda (text- fig. 65 A). The innominate veins were formed as usual, but soon gave way to two precaval veins (L.L). The latter at first ran posteriorly and slightly mesially to points level with the anterior limit of the heart, when they turned directly inwards and ran horizontally ; in this part of their course each vein receives an azygos vein (M). The orifices of the two precaval veins and the wide postcaval vein (‘I’) form the angles of a triangle with the apex posterior. This arrangement of the precaval and postcaval veins is the rule in Marsupials, but it occasionally occurs in higher. Mammals. The chief difference between the condition described above and Text-figure 65. The veins of Ailurus fulgens (A) and Meles meles (B). Explanation of Letters. G. subclavian veins; H. intercostal vein; Ic.1 and Ic.2. intercostal veins; O. esophagus: P. trachea; Q. right recurrent laryngeal nerve; R. internal jugular vein; 8. inferior thyroid vein; U. internal jugular vein; V. vertebral vein; W. externa] jugular vein. Other letters in text. that of the Marsupials lies in the nature of the terminal orifices. In the latter the openings are closer together, and the postcaval and right precaval veins are only separated by a crescentic fold. In the former they are wider apart, and the crescentic fold is not present. In the process of development the anterior cardinal veins from the head and the posterior cardinal veins from the body unite to form the Ducts of Cuvier, which open into the sinus venosus. The anterior cardinal veins become united by a transverse vessel, which takes a gradually increasing share in returning the blood from the left side of the head. Kventually the part of the left anterior cardinal vein between this vein and the sinus venosus 590 ON SOME ABNORMALITIES IN THE CARNIVORA. atrophies, and the blood is conveyed entirely to the vein of the right side. The part of the right anterior cardinal vein between the cross vein and the sinus venosus becomes the precaval vein of the normal animal. I believe that the persistence of two precaval veins in this animal is due to the non-formation of the cross vein. No abnormality was found in the postcaval vein. The veins of the father of this animal were quite normal. Abnormalities in the Vertebral Veins of a Badger (text- fig. 65 B). The left vertebral vein bifurcates in the lateral triangle of the neck; one of the divisions opens into the left external jugular vein, and the other one opens into the left innominate vein. The right vertebral vein runs posteriorly dorsal to the right subclavian vessels, collects the veins from the first and second right intercostal spaces, and opens into the precaval vein. ON THE EXTERNAL CHARACTERS OF THE KOALA. 597 33. The External Characters of the Koala (Phascolarctos) and some related. Marsupials. By R. I. Pocock, FLR.S., F.Z.8: [Received May 10, 1921: Read June 7, 1921.] (Text-figures 22-26.) CoNnTENTS. Page MtVOOUCTIONM Nee Cor arnt er Seen we ar OL AHR HTMar MING A eet Lae een ee Cee eta Dee rep hacialeValbnisscocecs det cee eee ee eee OG AHeMMiouibligye eco ete een ean oa eats L1G Teh B aie ae ee SS POP LCE Moh Se er OG RhepMore/ Hootie sass eas ee eas Sa LOS MhegtHan dub oO ten 4-8 tyes sees Bat danke Pere ie ee (GOL hewRouchvwas eh. cies Dettat nes LOO: The Classification of crete. eratgdants, ees 606 Introduction. The body of the female Phascolarctos described in the following paper was very kindly sent to me by Mr. A. F. Richardson when I explained to him the zoological importance of the animal and the infrequeney with which fresh examples come into the hands of anatomists in this country. The need for renewed examination of Phascolarctos is especially shown by the disagreement amongst modern zoologists regarding the position and status to be assigned to it in the classification of Diprotodont Marsupials. Thomas, for instance (Uat. Marsupialia and Monotremata in Brit. Mus. 1888), divided the suborder into three families: (1) Macropodide with the subfamilies Macro- podinz, Potoroine, Hypsiprymnodontine ; (2) Phalangeride with the subfamilies Tarsepedine (Varsipes), Phalangerine (Plelane ger, Trichosurus, Pseudochirus, Petauroides, etc.), and Phascolarctine (Phascolar hos); and (3) Phascolomyid (Phascolonys). In connection with the Macropodide and Phalangeride, Thomas pointed out that the characters of Hypsiprymnodon are inter- mediate between those of the two families, and render the division between them by no means so sharp and well-defined as it has been usually considered to be. Nevertheless, his arrangement suggests that Phascolarctos is an aberrant member of the Phalangeridez more nearly related to Phalanger than the latter is to Macropus. Thomas's classification has been adopted by English zoologists and by Trouessart in his Catalogue (1912). Bensley (Amer. Nat, xxxv. pp. 117-138 and 245-269, 1901; and Tr. Linn. Soc. Lond. (2) ix. pt. iii. pp. 83-214, 1903) followed Thomas in admitting the family Phascolomyide and in regarding Phascolarctos as the representative of a subfamily of the Phalan- geride; but he made the important modification of transferring Proc. Zoou. Soc.—1921, No. XL. 40 592 MR. R. I. POCOCK ON THE Pseudochirus and Petawroides from the Phalangerine to the Phascolarctinz because of the crescentic (subselenodont) pattern of the molar teeth in the three genera. Bensley’s classification was’ adopted, presumably with approval, by W. K. Gregory (Bull. Amer. Mus. Nat. Hist. xxvii. pp._215-216, 1910), who was acquainted with but rejected the classification by Winge quoted below. Thomas, of course, was quite familiar with the dental character to which Bensley and Gregory attached so much importance. On p. 167 of his Catalogue he pointed out that by the compli- cated subselenodont character of its molars, Psewdochirus, with its close ally Petawroides, stands somewhat apart from most of the other Phalangers, and approaches Phascolarctos, in which a similar but simpler modification is observable. But his reasons for attaching to it subordinate systematic value lay apparently in the circumstance that in young examples of Phalanger a tendency towards the same structure is visible, but the crests on the molars - goon wear off, leaving little difference between them and the simple quadricuspid molars characteristic of typical genera of Phalangeride. The obvious, but not on that account necessarily true, inference to be drawn from this fact is that the tendency towards the subselenodont molar pattern exhibited by Phalanger is a primitive character of the family Phalangeridee, which is tem- porarily retained in Phalanger, lost in Trichoswrus, Dactylopsila, and others, and elaborated in Pseudochirus and Phascolarctos. Winge (E Museo Lundii, viii. pt. 1, 1893) held very different views. He adopted two families: (1) Phalangistide |= Phalan- geride] with the subfamilies Pseudochirini for Psewdochirus and Petauroides, and the Phalangistine [= Phalangerine; for the two groups Phalangistee (Phulanger, Trichosurus, Petaurus, Tarsipes, etc.), and Macropodes (Macropus, Hypsiprymnodon, ete.); (2) Phascolarctide with the subfamilies Phascolarctini (Phascolarctos) and Phascolomyini (Phascolomys) oe Winge took as the basis for his classification the degree of extension of the tympanic process of the alisphenoid, which in the Phascolarctidee (Phascolomys, Phascolarctos) is small, does not envelope the tympanic cavity, and fails to reach the paroccipital process, whereas in the Phalangeridw, comprising the rest of the genera of Diprotodont Marsupials, the bone in question is large, envelops the tympanic cavity, and reaches the paroccipital process. As accessory characters, the vestigial tail, the presence of a cardiac gland in the stemach, and the loss of one of the two normal pairs of teats further serve to distinguish the Phascolarctide from the Phalangeride 7. %* To these Winge added Thylacoleontini (Thylacoleo) related to Phascolarctini and Diprotodontini (Diprotodon, Nototherium) related to Phascolomyini. + Winge’s valuable paper is unfortunately written im Danish. The statements about the extension of the alisphenoid is taken from Max Weber’s work. I have not been able to verify it im the case of the two skulls of Phascolarctos available for examination. EXTERNAL CHARACTERS OF THE KOALA. 593 Attention may also be drawn to Winge’s severance of the erescent-toothed or subselenodont Phalangers (Pseudochirus and Petauroides) from the rest as representing a special subfamily Pseudochirini equivalent to the’ Phalangerine containing not _ only the rest of the genera Thomas assigned to that subfamily, but all of those constituting the family Macropodide of that author. Max Weber (Die Siug. p. 348, 1904) followed Winge in admitting the two families Phalangeride and Phascolarctide as he defined them; but he did not admit the subfamily groups Phascolarctine and Phascolomyine for Phascolarctos and Phasco- Jomys respectively, and in the case of the Phalangeride he made a compromise between Thomas’s and Winge’s systems by dividing the family into three subfamilies : se Mlk n peas, Hypsiprymno- dontine, and Macropodine. It will be noticed that neither Winge nor Max Weber attaches any particular importance to the structural peculiarities of Tarsipes. ‘The divergence of opinion between the authors quoted may be briefly and more clearly expressed by the following tabulation of their classifications :— Thomas. Family MACROPODID. Subfamily Macroropin# (Macropus, Dendrolagus, etc.). Potoroinz (Potorous, Bettongia, etc.). 55 HyYPsiPRYMNODONTIN® (Hypsiprymnodon)- Family PHALANGERID A. Subfamily Tarsrpepinm (Tarsipes). PHALANGERIN® (Phalanger, Trichosurus, Pseudo- chirus, Petauroides, etc.). x PHASCOLARCTIN® (Phascolarctos). Family PHASCOLOMYID (Phascolomys). 39 32 Bensley. Family MACROPODID (Macropus, Hypsiprymnodon, etc.). >» PHALANGERIDA. Subfamily Tarsrpepinz (Tarsipes). PHALANGERINE (Phalanger, Trichosurus, etc.). PHASCOLARCTINE (Pseudochirus, Petauroides, Phascolarctos). Family PHASCOLOMYID (Phascolomys). Winge. Family PHALANGERIDE*. Subfamily PHALANGERINE. a. Macropodes (Macropus, Hypsiprynnodon, etc.). b. Phalangeri (Larsipes, Phalanger, Trichosurus). Subfamily PspeupocurIrinw (Pseudochirus, Petauroides). Family PHASCOLARCTID As. Subfamily PuascoLarcrinz (Phascolarctos). PHAscoLoMYinm® (Phascolomys). 39 2) 22 * Por the sake of clearness in comparison, I have here altered the name Phalan- gista and its derivatives used by Winge to Phalanger and its derivatives. 40* 59A MR. R. I. POCOCK ON THE Weber. Family PHALANGERIDA. C Subfamily Macropopinm (Macropus, Dendrolagus, Potovous, Bettongia, etc.). _ HypsIPRYMNODONTIN® (Hypsiprymnodon). 5 PHALANGERINE (TLarsipes, Phalanger, Trichosurus, Pseudochirus, Petauroides). Family PHASCOLARCTID A (Phascolarctos, Phascolomys). Kuternal Characters. The Rhinarium.—In Trichosurus the rhinarium is large, naked, and convex above; the infranarial portions are complete laterally and narrow; there is a median groove extending between the nostrils on to the philtruam, which is divided into two narrow strips, with a little process, overlying the gum between the median incisors, at the apex of the angular excision above. The nostrils are of the typical form, consisting of a rounded orifice in front, moderately widely separated iran its fellow of the opposite side, and of a lateral aud posterior narrow slit. In Phalanger the rhinarium is in a general way similar to that of Trichosurus, but the nostrils ave more widely separated, the infranarial portion is deeper in front, and the philtrum 1S eX- ceedingly wide, and, although gradually narrowed below, is quite wide where it terminates on the edge of the upper lip. There is, moreover, no angular excision such as is seen in Z'richosurus, but the median groove broadens below, its floor terminating in a median process which projects at least as low as the lateral portions of the philtrum. In Pseudochirus the rhinarium is more like that of T'richo- surus, but the infranarial portions are wider in front, with the inferior edge more steeply inclined ; the median groove is con- tinued between the nostrils to the summit of the rhinarium; and the inferior portion of the philtrum is quite narrow, with a small median notch. In Phascolarctos the nose is widely different. The muzzle is abruptly sloped downwards about an inch above the nostrils, and the whole of this inclined area is covered with very fine, short, scattered hairs, leaving a narrow, naked, thickened 11m round the nostrils. Thus there is no true rhinarium. The nostrils them- selves are also peculiar. Hach consists of an ovally elongated valvular orifice, with a thickened superior and lateral rim, but. with the inferior rim hardly defined from the upper lip; all trace of differentiation into lateral slit and anterior orifice has disappeared ; the septum is very narrow, and is continued in- feriorly between the two halves of the upper lip, ending below in a process over the gum between the median incisors. In Phascolomys ursinus* the rhinavium is quite unlike that of Phascolarctos, and resembles, broadly speaking, the rhinarium * The rhinarium of Lasiorhinus latifrons, the hairy-nosed Wombat, which should, I think, rank as a distinct genus, is unknown to me except from descriptions and a dried skin. It appears to differ solely from that of Phascolomys in being covered with short hair, the nostrils being normally formed. EXTERNAL CHARACTERS OF THE KOALA. 5G5 of the three genera of Phalangers above described. It extends dorsally, however, further backwards in the middle line, and the hair of the muzzle encroaches over the well-developed slit of the nostrils, the expanded portions of which are widely separated in the middle line. The infranarial portions are moderately deep, and reach almost to the posterior end of the nostril-slits. Text-figure 22. F . Head of Pseudochirus peregrinus, with pocket of ear (a) seen from above. . Rhinarium of the same. . Ear of Trichosurus vulpecula. . Rhinarium of the same, from the front. . Rhinarium of the same, from the side. . Head of Phalanger maculatus, the ear stripped of hair. Rhinarium of the same, from the front. Qertoawe Their inferior edge slopes obliquely downwards and inwards to the gum of the incisors, the two lobes of the upper lip being 596 MR. R. I. POCOCK ON THE tolerably widely separable at this point. There is no median excision in the inferior edge of the rhinarium above the teeth, and there is no median groove, but a groove runs on each side from the nostril downwards and inwards to the gum, defining the infranarial portion from the rest of the rhinarium. Facial Vibrisse.—The facial vibrissee are well and normally developed in Z'richosurus, Phalanger, Pseudochirus, and Phasco- lomys, being represented by numerous long mystacials, super- ciliaries, genals—one tuft,—submentals, and Unterramals. In the examples of Phalanger and Pseudochirus examined, however, the interramals were few in number and slender. In Phascolarctos the vibrisse are comparatively poorly developed, the mystacials. especially being few and short and the interramal tuft wanting. The Mouth—There are only two points for me to notice in connection with the mouth. The first is the presence, with which everyone is familiar, of cheek-pouches in Phascolarctos. These lie alongside the gum of the upper jaw in front, the orifice looking downwards iit the cavity extending upwards and backwards. towards the eye. No such pouches occur in 7 richosurus, Pha- langer, Pesnialoaiar us, or Phascolomys. The second is the presence in Phascolomys of a flap of naked skin extending into the mouth from the inner surface of the lips laterally. This 1s analogous to- the similar ingrowths well known in the Rodentia, where they serve to close the throat during gnawing. Their function is no doubt the same in Phascolomys, Ghich has rodent’ incisor teeth. These oral flaps of skin are not developed in the other genera here discussed, nor so far as I am aware in any other Marsupial. a. A flap of skin jutting inwards from the lower ip and lying alongside the tongue on each side .............. veseseeueees. Phascolomys. a’. No flap of skin jutting into mouth from the ie 6. A cheek-pouch on each side of the upper jaw .................. Phascolarctos. bP Niovcheek-pouchessers: 15 seer eee eek eeeees ee .... Pseudochirus, ~ Phalanger, Dr PenGR cee f The Ear.—In Trichosurus the ear is high, but tolerably narrow and nearly naked internally. The supratragus (metatragus) is. a well-developed thick, semicircular, slightly upturned disk. Above it there is a low curved ridge. In front of it the cavity of the ear is defined by a strong oblique ridge passing upwards. slightly above the low ridge above described and inferiorly curving backwards beneath the supratragus and passing thence downwards on the inner side of the well-developed tragal thickening forming the anterior rim of the inferior notch (aditus inferior), the posterior border of which is defined by a strong thick antitragal ridge, capable of being turned outwards and backwards to a certain extent; but it exhibits no pocket-like depression. In Pseudochirus peregrinus the ear is nearly as naked intern- ally, and relatively nearly as high as in Jr ichosurus but somewhat broader. ‘The supratragus is as well developed, but forms a more distinct roof to the deep lower portion of the cavity. The low EXTERNAL CHARACTERS OF THE KOALA. 597 ridge seen in Zvrichosurus is represented by a lobate thickening, and there is a similar thickening lower down behind, above the posterior end of the supratragus. The anterior ridge forms a lobate expansion above the anterior end of the supratragus, and the tragus itself is less well defined. The ridge behind the Text-figure 23. A. Head of Phascolarctos cinereus, with position of cheek-pouch (p) dotted in and pocket of ear (a) seen from above. B. Nose and part of palate of the same, showing the nostrils and the orifices of the cheek-pouches (p). C. Nose and mouth of Phascolomys ursinus, with lower incisors cut short and flap of skin (f) projecting into mouth. D. Side view of rhinarium of the same. E. Ear of Lasiorhinus latifrons, from dried skin softened in water. 598 MR. R. I. POCOCK ON THE inferior notch turns inwards above towards the posterior end of the supratragus, and it is provided with a shallow pouch defined externally by a definite thickened ridge. The ear of Phalanger maculatus is greatly reduced, thickly covered with hair externally, and scarcely projects above the level of the head. The supratragus is large and valvular, and the cavity of the ear above it is deep and roofed over by a ridge resulting possibly from the confluence of lobate thickenings homologous to the two seen in Phalanger. The anterior ridge does not extend so high above the supratragus as in the two genera already discussed, and there is a trace of the pouch on the posterior ridge, although it is not so well defined as in Pseudochirus. In Phascolarctos the ear is more expanded even than in Pseudochirus, but it differs from the ear of that genus, 7'richo- surus, and Phalanger in having the supratragus reduced to a low curved ridge without any trace of the lobe*. There is a low ridge above it as in Z'richosurus, and the anterior ridge bounding the cavity of the ear in front is as in that genus; the tragus, however, is much smaller. The pocket on the posterior ridge is present as in Pseudochirus, but is deeper and has a more sharply defined posterior rim. In Phascolomys the antero-internal ridge, continuous above with the slightly overfolded anterior margin of the pinna, is produced inferiorly into a large compressed lamina, jutting back- wards into the cavity of the ear and descending to the auditory orifice. Anteriorly and externally this lamina is marked by a deep groove which descends and passes into a low ridge defining the anterior border of the inferior notch, but there is no definite tragal thickening. The antitragal ridge is also simple. It ascends and forms the posterior edge of the cavity of the pinna, and is itself marked externally and posteriorly by a groove. The supratragus appears to be represented by a low ridge visible above the laminate expansion of the antero-internal ridge. The ear of Lasiorhinus latifrons is very like that of Phascolomys, but is longer. On askin softened in water I could find no trace of the supratragus. The ear of Phascolomys resembles that of Phascolarctos in the complete or almost complete disappearance of the supratragus— an important character considering the constancy in the deve- lopment of this ridge in many Mammals. But in -the simple structure of the posterior antitragal ridge it is more like that of Trichosurus. The differential characters of the ear may be used to define and classify the five genera as follows :— a. Supratragus represented by a strong ridge with well-developed lobate thickening. 6. No pocket above the antitragal ridge ..........seccceuseee wees Trichosurus. * Thomas mentioned this peculiarity and wrote ‘‘metatragus almost obsolete.” EXTERNAL CHARACTERS OF THE KOALA. 599 b’. A pocket above the antitragal ridge. c. Pinna greatly reduced; pocket small ...............0.05. Phalanger. ce’, Pinna large; pocket better developed ............ 06000000000 Pseudochirus. a’. Supratragus at most represented by a low curved ridge; without trace of lebate thickening. d. Antero-internal ridge normally developed; a large pocket abovethe antiinaral rid cert comes cer slavaeneitntentnse os eietan le) Phascolarctos. d'. Antero-internal ridge large and laminate; no pocket above amtifvagall tGee wes cia sud sdoves cocclugeer Gandy ced tatebcrahadcss sun mR ASCOLOMYS. The Fore Foot.—In Trichosurus, Phalanger, and Pseudochirus there is a tuft of long carpal vibrisse. In Phascolarctos and Phascolomys these tactile bristles are absent. In Trichosurus the fore foot presents no special modifications. The five digits are free of webbing down to the plantar pads and are nearly evenly spaced, the third and fourth being the longest and subequal, the second and fifth a little shorter but subequal and rising at approximately the same level, and the first or pollex much the shortest, about two-thirds the length of the third or fourth, and not opposable but closing obliquely backwards and inwards across the sole (palm). The claws of all the digits are compressed, curved, and pointed, and the integument of the lower side of the digits is transversely grooved. The uniformly granular plantar pad is four-lobed, the three main lobes being in contact. The inner or pollical lobe, sometimes separated from the main lobes, is fused with the inner element of the carpal pad into a longitudinal mass, longer than wide. The external element of the carpal pad is elliptical, narrowly separated in front from the external lobe of the plantar pad, and, like the internal element, passes imperceptibly proximally into a narrow naked area of skin between them and the hair of the wrist. The centre of the sole is depressed and granular like the pads and the depressions between them. In Phalunger the fore foot is considerably modified from the type seen in Zrichosurus. The third, fourth, and fifth digits are evenly spaced; but the second is capable of being separated from the third by a much greater space than that between the third and fourth. With the pollex it is capable of being extended almost at right angles to the axis of the foot and of being closed transversely upon the sole. The pads are transversely striate, with the spaces between them granular, the space or groove between the second and third being deeper than that between the third and fourth. The first, the internal or pollical lobe, forms a continuous subtriangular mass, directed obliquely backwards and inwards when the pollex is drawn back, and the external moiety of the carpal pad is larger than in Zrichosurus. The capacity for co-operative movement of the first and second digits in a plane nearly at right angles to that of the third, fourth, and fifth seems to have been overlooked in Phalanger maculatus. Bensley, at all events, records the similar phe- nomenon in Pseudochirus and Phascolarctos as peculiar to these two genera. Nevertheless, the modification in question may be 600 MR. R. I. POCOCK ON THE ; Text-figure 24, . Right fore and hind foot of Trichosurus vulpecula. . aie % Phalanger maculatus. 5 oy i) Pseudochirus peregrinus. x F- ES EXTERNAL CHARACTERS OF THE KOALA. 601 clearly seen even on dried skins; and it is shown in the figures of the Spotted Cuscus in the Royal Natural History, iii. p. 257, 1894, and Harmsworth’s Natural History, ii. p. 888, 1910. These figures bear the impress of having been drawn from life by Mitzel and Kiihnert respectively. The fore foot of Pseudochirus resembles tolerably closely that of Phalanger, except that the third digit is not lengthened and the lobes of the plantar and carpal pads are separated and smaller, the fused pollical element of the plantar pad and the inner moiety of the carpal pad forming an oblique transverse mass about twice as wide as long. The fore foot of Phascclarctos is an extreme exaggeration of the type seen in Phalanger and Pseudochirus, although the pads and intervening spaces are areolated or granular, not striated. The sole is much longer as compared with its width: the first and second digits are completely isolated from the rest, and rise close together from the postero-internal angle of the foot, at right angles to its long axis and close transversely across the proximal half of the sole, the posterior border of the pollex when drawn back being approximately in the same transverse line as the posterior border of the sole. There is a large three-lobed plantar pad at the base of the third, fourth, and fifth digits, and there is a similar but smaller lobe upon the base of each of the second and first digits. The one on the pollex probably represents the inner moiety of the carpal pad, and the one on the base of the second digit the first or external element of the plantar pad, widely severed from the lobe in front of it and altogether discon- nected from the pollex behind it. The external moiety of the carpal pad is small, restricted to the postero-external angle of the foot and widely separated from the plantar pad. In Phascolomys the fore foot, modified for terrestrial progres- sion and digging, is very different from that of the preceding genera, especially from that of Phascotarctos. It may be derived from the type seen in Zrichosurus by the shortening and widening of the digits and sole, the obliteration of nearly all trace of the individual elements of the pads, and by the straightening, lengthening, and blunting of the claws. The Hind Foot.—In Trichosurus*, Phalanger, and Pseudochirus there are one or two tactile vibrisse, similar to the carpal vibrisse, on the inner side of the heel behind the hallux. These are absent in Phascolarctos and Phascolomys. The hind feet of the four arboreal genera—Trichosurus, Phalanger, Pseudochirus, and Phascolarctos—difter from each other much less than the fore feet. Their general structure is well known. Hence only the comparatively minor points of difference need be noticed. * Represented in the figure of the hind foot of Psewdochirus cooki (pl. 20. fig. 4, of Waterhouse’s ‘ Marsupiata’), but apparently unnoticed in the text. 602 MR. R. I. POCOCK ON THE In Trichosurus the conjoined second and third digits* are much shorter than the fourth. The three lobes of the plantar pad at the base of the terminal digits are in contact, the outer lobe being larger than the other two taken together. The large lobe of the hallux shows no trace of division, and there is a tolerably well-defined, long external metatarsal pad reaching from the heel almost to the plantar pad. The digits of Phalanger are approximately as in Trichosurus; but the external lobe of the plantar pad is separated to a certain extent in front from the other lobes, which are fused although defined by a groove, and are together larger than the outer lobe. The entire hallux is relatively larger than in T’richosurus, and its great pad is indistinctly divided; there is no distinct external metatarsal pad and the heel is shorter than in 7’richosurus. In Pseudochirus the digits are as in the foregoing genera, but all the elements of the striated pads on the sole are better defined even than in 7richosurus. The outer lobe of the plantar pad is completely isolated from the two conjoined inner lobes and about half their size. The great pad of the hallux is distinctly divided, into a distal and a proximal portion, the latter representing the inner metatarsal pad; the outer metatarsal pad is well developed, sharply defined, and about twice as long as wide. In the foregoing genera the hind foot is larger in every way than the fore foot. In Phascolarctos the hind foot is not larger than the fore foot. It differs from the hind foot of the other genera in having the united second and third digits relatively much longer and only a little shorter than the fourth; the Jobes of the plantar pad are feebly differentiated, being fused to form a transverse cushion, with irregularly convex anterior and irregularly concave posterior border. The great pad of the hallux is undivided, and there is no distinctly defined external metatarsal pad. Thus in the development of the pads the foot of this genus differs more from that of Pseudochirus than from that of Trichosurus. The hind foot of Phascolomys, as has often been pointed out, is a fossorial modification of the Phalangerine scansorial hind foot, resulting from the conversion of the sharp, curved claws of the latter into longer, stronger, and straighter claws, from the approximate equality in length between the fourth and the united second and third digits, the reduction in bulk of the great pad of the hallux so that it projects only slightly beyond the inner margin of the foot, with the terminal segment of the digit planted like a hemispherical tubercle upon it, and from the practical * These two united digits act as a fur-comb in Marsupials. Possibly they were primarily modified for that: function. Generally they are too small in arboreal forms to be of much use fer grasping. But this cannot be maintained in the case of Phascolarctos; and in Phascolomys they are large enough to be subservient to digging. But in the Kangaroos they appear to be retained solely for the purpose mentioned, and may frequently be seen to be so employed. Moreover, it is signifi- cant that these are the only digits in Tarsipes which have other than rudimentary claws. EXTERNAL CHARACTERS OF THE KOALA, 603 obliteration of the pads with expansion of the heel; but, as in the Phalangerine foot, the fifth digit is more widely separated from the fourth than the latter is from the third, although the digit in question is relatively shorter and the space above mentioned somewhat wider. Text-figure 25. \ \ \ B. Right fore and hind foot of Phascolarctos cinereus. D. 53 Ss Phascolomys ursinus. 604 MR. R. I. POCOCK ON THE By their feet the genera may be distinguished as follows :— a. Feet fossorial. with digits short, claws long, blunt and slightly curved, and the granular pads but itttle differen- tiated ; hallux short, its distal phalange reduced to a button- like excrescence on the shortened lobe, which, however, still retains the capacity for movement in the opposable plane ... Phascolomys. a'. Feet scansorial and preensile, with longer digits and sharp, curved claws, and differentiated pads; hallux very large and opposable, with well-developed terminal phalange. b. Fore foot as large as hind foot, with comparatively long and narrow sole; the inner Icbe of the plantar pad small aud attached to the base cf the second digit, which is separated by a long space from the third; the imuer lobe of the carpal pad small, isolated from the outer and attached to the base of the first digit (pollex) ; second and thivd digits of hind foot comparatively large, strongly prekensile ..... Phascolarctos. b', Fore foot smaller than hind foot, with shorter, broader sole ; the inner lobe of the plantar pad large, in contact, or nearly so, with the rest of the pad and fused into one mass with the inner lobe of the carpal pad, there being no differ- tiated pad at the base of the first digit (pollex) ; second and third digits of hind foot short, weakly prehensile. c. A comparatively long space between the third and second digits of the fore foot, the second actimg in unison with the firzt like a double opposable pollex; pads striated ... Phalanger and Pseudochirus. c/. The five digits of the fore foot evenly spaced, the second in no respect opposable to the next, and the first or pollex only slightly so; pads granular..............:0::-.eee Trichosurus. The Pouch.—I\ have seen no fresh female examples of Phalanger and Pseudochirus. In Trichosurus the pouch, as in J/acropus, is deeper than wide, its orifice has a well-developed lateral and posterior rim but no overhanging anterior rim; it therefore leoks forwards * ; and in the specimen examined there were only two teats, not four, as stated by Winge to be characteristic of the Phalangeride. In Phascolomys the pouch is a little wider than long, being deeper laterally, especially anteriorly, than behind or mesially in front, and a little deeper behind than mesially in front, the muscular rim overhanging the cavity all round; and there is a single pair of teats. In Phascolarctos the pouch, as recorded by Forbes, is much wider than long, being extended, gradually narrowing as it goes, along the depression between the muscles of the thigh and abdomen nearly as far as the edge of the flap of integument joining the hind leg to the body, its width being about three times its length. The orifice looks slightly backwards, the muscular rim overhanging the cavity laterally and to a slight extent in the middle line in front, but ceasing on each side close to the position of the single teat. The orifice therefore looks backwards. % In his volume on Marsupials (Allen’s Naturalists’ Library, p. 76, 1894), Lydekker wrongly states that the orifice of the pouch is directed backwards in Phalangeride. EXTERNAL CHARACTERS OF THE KOALA. 605 From the description of these three pouches it is evident that the pouch of Phascolomys is structurally intermediate in many respects between those of Vrichoswrus and Phascolarctos. The differences between the pouches of the last two genera, consider- ing that both are arboreal forms, is remarkable, especially in view of the theory that has been advanced that the anterior upward aspect of the orifice is an adaptation to arboreal life or to the Text figure 26. K » wie chy Way Nes ig "| oy ae WNL CY a CH iohss NHS § th TR sant . (<— m %. Rijs F at — CI mee enon ‘ ( = Sen ere ar ry QJ NY)? W)C ———— DAS = RSs Raia y XS HS } ~7r :~ ger 7 a te (PA QIOR LN? hea Se 2 ia ‘o “2G ec a ol wo 4/)) wy 4 Fe al Gat wea & (Ss thd et b t A 7 = Se SS = AAA) (7 Sis ZS k roan JE NN Wesnes SAS7 ae Ss VU WAY 528% (277 i IN = Sor af SSS MARA M4 SCO mes RN DA YANN : mes BY TATAY YAN Z IN) WA A UO MINAIN og MAA (NN RN NY Vay WA ae } AM YY WW) | --~-...sh més Cx Me Nad BOUT wy RYAN ie: af RNA a Sete to ANN met MN af = Ae 2 1 1 1 1 1 1 ' 7 1 ' i os) (mr ees I Qa a eB. ies Diagrams indicating the relation of the mesoptile to the protoptile and to the first true feather of a Ringed Penguin :—pro., the protoptile, consisting of four outer barbs representing a shaft and of four inner barbs representing an after- shaft. sh.mes., shaft of mesoptile continuous distally with protoptile and proximally with the shaft of the true feather. afmes., mesoptile aftershaft also continuous with the protoptile and the true feather. cé., connection with pro- -toptile. c¢., connection with shaft of true feather. ea., connection with after- shaft of true feather. cal., calamus. NESTLING FEATHERS OF THE MALLARD. 615: and in the following year an important paper by Mr. Pycraft on the Emperor and Adélie Penguins was published which deals with the composition and sequences as well as with the structure of nestling feathers. Dr. @lanke refrained from stating whether the silky-white protoptile coat, or the grey fur-like mesoptile coat, of the Ringed Penguin, corr esponds to the nestling coat worn by newly -hatched chicks and ducklings. But Mr. Pycraft, in discussing the sequences of nestling-feathers, states that there ave good reasons for assuming that in most of our common birds the first or protoptile coat has been lost, and adds that the feathers forming the adult coat of the Emu might possibly consist of mesoptiles. Whether ornithologists have, as a rule, adopted Mr. Pycraft’s views it is impossible to say, but ib may be mentioned that Mr. Ingram, in the ‘Ibis’ paper already referred to, points out that a study of the nestling plumage of the Raptores suggests. that in some cases, at any abe, it is the second and not the first generation of nestling ‘‘ down ” that has been suppressed (1). When, some years ago, I was led to study the feathers of Penguins, I assumed that ‘birds with only one nestling coat had lost the first or protoptile coat; but the examination of a number of nestlings soon made it evident that when in the Galli and Anseres there is only one generation of nestling feathers, it is owing, not to the absence of the protoptiles, but to the suppression of the mesoptiles. In the account of the nestling feathers of the Emperor Penguin, Pycraft states that the mesoptiles are umbelliform, and that in the Adélie Penguin the mesoptile is in part attached to the main shaft, but mainly to the aftershaft. I have not had an opportunity of studying the nestling feathers of either the Emperor or Adélie Penguins, but from material placed at my disposal by Dr. Eagle Clarke I have worked out the structure of the mesoptiles of the Ringed Penguin. In this species the mesoptile is extremely complex: it consists (1) of an outer small series of barbs, which extends between the protoptile and the tip of the true feather, and hence occupies the position of, and doubtless represents, a shaft; and (2) of an inner series of barbs (text-fig. 3), arranged to form two or three bundles, which as obviously represent an aftershaft. ‘The chief connections of the mesoptile in the Ringed Penguin are diagrammatically repre- sented in text-fig. 4, which indicates that the mesoptile consists of a simple shaft extending between the protoptile and the tip of the true feather, and a complex aftershaft connected with the shaft as well as with the aftershaft of the true feather. In a Mallard duckling at the end of the fifth week the mesoptile looks as if it would develop into a simple umbel not unlike the umbelliform protoptiles of Penguins. But this appearance is due to the fact that nestling feathers, like true feathers, are developed, not out of a scale- like plate, but out of an epidermic tube surrounding a highty-vascular dermic pulp and Text-figure 5. + te}. Ady) Yy yy ps Diagram indicating the structure and relationships of the mesoptile with the pro- teptile and with the true feather in the Chinese Goose. The mesoptile shaft (sh.mes.), which lies between the protoptile shaft (sk.pro.) and the tip of a true feather, consists of numerous barbs and of a band representing a rhachis. Con- tinuous with the rhachis is a perforated band from which slender cords (c.) proceed to the barbs forming the tip of the true feather (¢eZ.). The mesoptile aftershaft is also continuous with a perforated band which ends in a vestige of an aftershaft of the true feather (aft.). From the perforated band slender cords (¢1) proceed to barbs near the tip of the true feather. 1¢ will be ob- served that the aftershaft of the protoptile consists of twelve well-developed barbs armed with numerous barbules. NESTLING FEATHERS OF THE MALLARD. 617 protected by a thin epidermic sheath. When the development is carried further, it becomes evident that in the Mallard the mesoptile consists of two sets of fibres, an outer set representing a shaft, and an inner, an aftershaft, each provided with a more or less distinct rhachis. Though the mesoptiles in Ducks may reach a considerable size, it is easier to make out their develop- ment and structure in Geese. The wing-quill mesoptiles, though well-developed in the Mallard, are not present in the Emden goose but as Pl. I. fig. 3 shows, the major wing-covert mesoptiles may reach a considerable size in Geese. To start with, the mesoptile of a Chinese gosling looks as if it possessed a calamus (Pl. IV. fig. 13); the mesoptile barbs seem to proceed from a tube lying immediately in front of the tip of the true feather. That this tubular part represents a calamus is supported by the fact that it contains several ‘“ cones,” but ugainst the view that it is a true calamus is the fact that it is in the act of splitting longitudinally into numerous slender cords. In mesoptiles from older goslings part of the tubular portion is represented by two perforated bands, from which proceed numerous slender cords towards the wing-covert. Sometimes the mesoptile in goslings splits into two portions (Pl. IV. fig. 14, & text-fig. 5), an outer (the shaft) extending between the shaft of the protoptile and the tip of the wing-quill, and an inner (the aftershaft) extending between the protoptile aftershaft and the vestigial aftershaft of the wing-quill. It may be mentioned that in the Anseres the mesoptiles vary greatly in structure—the proximal part of the aftershaft may be perforated and assume the form of a section of a relatively large tube (PI. IV, fig. 14), or, like the rhachis of the shaft, be represented by a simple narrow band. Though in the Mallard and Chinese goslings the aftershaft, as well as the shaft of the mesoptile, has usually a distinct rhachis, in Emden goslings the whole of the mesoptile, as Pl. 1. fig. 3 suggests, usually consists of simple barbs, one or two of which are continuous with the rhachis of the wing-covert shaft, while two or more end in a vestigial aftershaft. The barbules of the mesoptile, like those of tlie protoptile aftershaft, are narrow and provided with well- developed cilia. In the Ringed Penguin the mesoptiles closely agree in structure, but in the Ducks and Geese examined no two mesoptiles were alike in structure. The mesoptile coat is probably as useful now to Penguin chicks hatched within or near the Antarctic Circle as it was when originally acquired, during perhaps a cold phase of a glacial epoch. In the Anseres, though the protoptile coat still plays an important part, the mesoptile coat probably no longer counts in the struggle for existence. If this is the case there is no difficulty in accounting for its being well- developed in some parts but entirely absent in others, and for the marked variation amongst the mesoptiles which still persist. The Mallard has wing-covert, as well as wing-quill, mesoptiles. The mesoptiles which precede the wing-coverts in the Mallard 618 _, PROF. J. COSSAR EWARY ON THE are especially interesting, because they indicate how the suppres- sion of the mesoptiles was gradually accomplished. In the case of the disappearing protoptiles of the King Penguin (PI. I. fie. 4) the size is gradually reduced until they are smaller than the minute bunch of barbs which precede the filoplumes of ducklings. In the case of the vanishing mesoptiles two sets. of factors are evidently at work ; one set arrests the development of the mesoptile barbs, another set accelerates the development of the barbs of the true feather, with the result that in course of time the tip of the true feather is found projecting into the base of the protoptile calamus. Though in most cases the mesoptiles are gradually reduced in size, in some eases the length is maintained and either a few barbs with barbules are developed, or the simple epidermic tube splits into two or more bands, which for a time connect the protoptile to the tip of the true feather. It may be mentioned that though all the wing-quill protoptiles and mesoptiles may be present in the Mallard duckling at the end of the sixth week (PI. III. fig. 12), some of the mesoptiles. are usually shed before the end of the seventh week, and all are usually lost before the middle of the eighth week. But though the wing-quills lose all their nestling feathers during the eighth week, some of the feathers of the humeral track may retain protoptiles and mesoptiles to the end of the eleventh week. The existence of the wing-quill mesoptiles in the Mallard proves conclusively that the nestling coat worn by newly-hatched Ducks and Geese corresponds, not as Pyeraft suggested to the second generation of prepenne in Penguins, but to the first or protoptile generation. (3) The Tail-Quill Protoptiles. The tail varies greatly in young aquatic birds. In Penguins the tail protoptiles have almost disappeared and the mesoptiles are only represented by a few simple barbs (Pl. I. fig. 4). In the Mallard the mesoptiles of sixteen of the eighteen tail-quills have completely disappeared, but the protoptiles of all the eighteen quills are larger and more complex than in the plumose (meta- ptile) feather represented in P]. I. fig. 1. In Penguins steps are soon taken to develop the tail-quills with a view to their forming with the hind limbs a tripod useful in maintaining the erect attitude. In the Mallard the protoptiles form a ‘nestling tail” which plays an important part during the earlier portion of the period that Mallard ducklings behave like diving ducks™. In a ten days’ duck embryo the tail-quill papille ? are easily * The adult Mallard, Mr. G. J. Millais informs me, “hardly ever dives except when in play during the love chase, or to escape when wounded.” + There ave twenty tail-quill papilla, but only eighteen of them develop into feathers large enough to rank as tail-quills. A paper “On the development of the Feathers of the Duck during the Incubation Period” was recently communicated to the Royal Society of Edinburgh by Augusta Lamont, B.Sc. NESTLING FEATHERS OF THE. MALLARD. 619 identified (Pl. IL. fig. 6), and by the twenty-fifth day of incubation they have grown into long filaments out of which escape imme- diately after hatching highly specialized protoptiles. During the first week the nestling tail consists only of protoptiles (PI.IV. fig. 15), but during the second week the protoptiles are pushed from the skin by the tail-quills, or by vestigial mesoptiles. At Text-figure 6. , HA SS if vy ve 7 his ast mK Dy ends in a calamus containing “cones,” that this calamus is directly continuous with the expanded distal portion of the tail-quill rhachis, and that the connection of the protoptile with the teleoptile is strengthened by barbs at the tip of the tail-quill ending in the proxinilal portion of the protoptile calamus. the end of the fourth week the tail-quills may be nearly an inch in length, and they are still longer at the middle of the fifth week (Pl. [V. fig. 16). During the latter part of the fifth week the protoptiles begin to break off, and with the exception of the ninth from the middle line at each side, they are usually all shed by the middle of the sixth week. 620 PROF. J. COSSAR EWART ON THE The calamus of the protoptiles is strong and rigid and firmly fixed to the expanded tip of the tail-quill rhachis. By way of strengthening the connection between the protoptile and the tail- quill, four or more pairs of the terminal barbs of the tail-quill are attached to the protoptile calamus (text-fig. 6). At the end of the second week the “* nestling ” tail and the other structures. concerned are sufficiently developed to admit of young Mallards. Text-figure 7. The proximal portion of the calamus of a tail-quill protoptile containing “ cones.’” As the calamus approaches the tip of the tail-quill it splits into an outer ‘portion continuous with the expanded rhachis of the quill and a more slender inner portion continuous with a vestigial tail-quill aftershaft. c., protoptile: calamus with “‘cones.” sf., tail-quill shaft. af., vestige of tail-quill after-. shaft. obtaining the greater part of their food by diving*. Notwith- standing the loss of the protoptiles during the sixth week, young Mallards usually continue to behave like diving ducks up to the end of the eighth week. From what has been said it is evident that the “ nestling tail” of the Mallard duckling is not from first. to last entirely composed of nestling feathers. It is a true nestling tail up to the middle of the second week, but from the * Two of nine Mallard ducklings I had under observation began to dive on the- thirteenth day: they were all diving on the fourteenth day. NESTLING FEATHERS OF THE MALLARD. 621 beginning of the third to the end of the fifth week the true tail- quills take an ever increasing part in forming the “ nestling tail.” Hence nestling feathers only function during three of the six weeks that Mallard ducklings feed and disport themselves like- diving ducks. When the protoptiles are shed, the tail-quills, owing to the rhachis being short and incomplete, look unfinished. Perhaps for this reason the first set of tail-quills is ere long superseded by larger and more perfect quills in which the rhachis, as in adult ducks, ends in a fine point. Usually the first set of tail-quills begins to drop out during the fourteenth week, and a new tail consisting of feathers of the adult type is eventually developed. The Structure of the Tail Protoptiles—The calamus of the tail-- quill, like that of the wing-quill, protoptile, continues to grow after hatching. In the case of the protoptiles at each side of the: middle line the calamus only reaches a length of 4 or 5 mm. and may have only three or four “cones”; but the calamus of the: outer tail protoptiles may eventually measure 15 mm. and contain over twenty ‘‘cones.” I expected the entire protoptile calamus: to be directly continuous with the expanded tip of the tail-quill rhachis, but I found that, as text-fig. 7 shows, the calamus splits. into an outer segment continuous with the tail-quill rhachis and an inner segment continuous with the rhachis of a vestigial aftershaft, the presence of this unexpected vestige of an after- shaft indicates that even the highly specialized tail quills. originally consisted of two shafts, and hence were constructed on the same plan as the double feathers of the Emu. The shaft of the tail-quill protoptiles of a Mallard duckling as a rule consists of fourteen or more pairs of barbs, each pro- vided with barbules. As already mentioned, the majority of the- barbs of the wing-quill protoptiles are long and pointed and have only a limited number of slender barbules (PI. II. fig. 8), but some of the barbs of the tail-quill protoptiles are ribbon-shaped (text-fig. 9), and they all have barbules along their whole length.. The majority of the barbules have simple cilia along both margins; but the six or seven pairs of ribbon-shaped barbules at the tip of the protoptile have large curved cilia along one margin, which differ but little from the hooklets on the barbules forming the tip of the true tail-quills. The difference between the specialized barbules with hook-like cilia at the tip of the pro- toptile and the slender proximal barbules with small cilia will be- evident if text-fig. 8 is compared with text-fig.9. When the broad barbules with hook-like cilia on one edge cross each other the approximation to a true teleoptile is especially marked.. What Mallard ducklings gain by having the tips of the tail-quill protoptiles highly specialized is not very obvious. The after- shaft of the Mallard’s tail quill consists of eight long barbs (Plate III. fig. 11) bearing slender almost straight barbules, each with two rows of small cilia. It may be mentioned that in the case of the goslings of Chinese- 622 PROF. J. COSSAR EWART ON THE> and Emden geese the tail-quill protoptiles—though possessing a long, strong calamus firmly connected with the rhachis of the tail- quill, a well-developed shaft, and an aftershaft made up of twelve barbs—are not provided with broad distal barbules armed with hook-like cilia. In goslings, as in ducklings, the inner segment of the calamus of the tail-quill protoptile is continuous with the rhachis of a vestigial tail-quill aftershaft. Text-figure 8. Two of the proximal barbules of a tail-quill protoptile; they are narrow and armed with small cilia. x 64. Text-figure 9, ! ' ' ' 1 ' 1 ' ‘Three of the broad distal tail-quill barbules armed along one edge with tooth-like cilia (hamuli) of true feathers. X 64. The difference between the simple ‘proximal and the specialized distal barbules is also indicated in text-fig. 6 (4) The Tail- Quall ehante In sixteen of the tail-quill protoptiles of the Mallard the mesoptiles are completely suppressed, but in the remaining two— the ninth from the middle line at each side—the mesoptile is represented by a simple tube from 10 to 12 mm. in length ex- tending between the protoptile calamus and the tip of the tail- quill. This tube, though complete to start with, sooner or later splits into two ov more bands, the outer of which is con- tinuous with the rhachis of the tail- quill, while the inner ends in the rhachis of a vestigial tail-quill aftershaft. As these bands have no barbules they evidently do not represent barbs. Though there are only vestiges of two tail-quill mesoptiles in the Mallard, there are sometimes vestiges of all the mesoptiles in cross- bred NESTLING FEATHERS OF THE MALLARD, 628 domestic ducks. In cross-bred, as in Mallard ducklings, the feathers at each side of the true tail-quills approach in structure the wing-quill mesoptiles. In Hider ducklings the wing-quill, as well as the tail-quill, mesoptiles are either absent or represented by minute vestiges, hence the Hider supports the view that in the Anseres the mesoptiles no longer count in the struggle for existence. (5) The Prepenne of the Trunk. About the prepennz which precede the feathers of the head, neck, trunk, and hind limbs, very little need be said. In many cases the protoptiles closely resemble, and may be as large as, the wing-quill protoptiles, but they are sometimes so small and poorly developed that it is difficult to distinguish them from pre- plumule. ‘The protoptile from the dorsal track of a thirty-eight days’ Mallard duckling represented in PI. I. fig. 1 is only half the size of the wing-quill protoptile in Pl. IT. fig. 8; but some of the protoptiles of the wing-coverts are smaller than the majority of the preplumule. Usually the protoptiles of the trunk are shed about the same time as the wing-quill protoptiles, but in a small chequered area at each side of the body near the tail well- developed protoptiles are still present at the end of the eleventh week. Though mesoptiles precede some of the trunk penne, they are never as well developed as the wing-quill mesoptiles. The examination of the mesoptiles of ducks and penguins lends support to the view that the mesoptile coat was specially acquired to meet a marked change of climate—a change perhaps from cold dry to cold damp conditions—and that in most cases as the climate improved the fur-like second coat was more or less com- pletely suppressed, and a new coat, made up of feathers con- structed on the protoptile rather than the mesoptile plan, was subsequently evolved. It is conceivable that some of our modern birds, e. g., the Emu and Cassowary, are descended from ancestors which never acquired a mesoptile plumage. 2. The Preplumule and Plumule. (1) Zhe Preplumule. In many birds the nestling coat sooner or later consists of pre- plumule as well as prepenne. In the Gannet the preplumule appear before the prepenne, and grow so rapidly that even when the prepenne are well developed they are difficult to find; but in the Mallard in most regions the preplumule are hidden by a relatively dense coat of protoptiles. Though the preplumule papille in the Mallard appear three or four days later than the protoptile papille, the majority of the prepiumule are ready to escape from their filaments as soon as the duckling is hatched. The relative size of the protoptile and preplumule filaments of the Mallard is indicated in Pl. IIT. fig. 9 and in text-fig. 10. The difference in size of the protoptiles, preplumule, and prefiloplume Proc. Zoou. Soc.—1$21, No. XLII. 42 624 PROF. J. COSSAR EWART ON THE is made evident by Pl. V. fig. 17, which shows a protoptile, three plumule, each with a preplumula on its tip, and, on the left of the protoptile, a prefilopluma on the tip of a developing filoplume. The preplumula consists of a short conical calamus, a shaft often consisting of a short rhachis and four barbs, and an after- shaft also as a rule consisting of four barbs. It is sometimes difficult to distinguish the small protoptiles which precede small true feathers from preplumule, but when in ducklings the Text-figure 10. Filaments from a Mallard embryo. P, proximal part of a protoptile filament. P.1, a preplumula filament. P.2, a prefiloplume filament. S., skin. barbules are characterized by nodular swellings and very few cilia it may be safely assumed that the nestling feather under examination is a preplumula. (2) The Plumule. At the beginning of the Jurassic Age the coat of birds may have consisted only of protoptiles and preplumule; but in many recent birds the plumule form a very important part of the nestling plumage. From Pl. V. fig. 17 it may be inferred that during the fifth week the plumule form an inner coat as effective in preventing a loss of heat as the woolly under coat of wild sheep. s , ; 4 SS ee ee eee NESTLING FEATHERS OF THE MALLARD, 625 Like plumose cr metaptile feathers, the plumule vary in size, colour, and structure, but they, asa rule, consist of a calamus, a shaft, and an aftershaft. ‘lhe calamus, especially well developed in the Hider Duck, is barvel-shaped and contains relatively large cones. As it leaves the skin it divides into two portions, an outer which forms the short rhachis of the shaft, and an inner which forms the still shorter rhachis of the aftershaft( Pl. V. fig. 18). The shaft consists of a dozen or more pairs of very long slender barbs from which spring numerous slender barbules, the proximal portions of which are characterized by pyramidal swellings, while the distal portions like the preplumule have nodular swellings. The aftershaft (Pl. V. fig. 18)is always well developed, and like the shaft is largely made up of long barbules having the characteristic pyramidal swellings. When a plumula is shed the tip of the sheath of the growing new plumula is often found attached to the calamus. Evidence of this we have in Pl. VY. fig. 19. (3) Prefiloplume and Filoplume. Nitzsch, in his account of filoplume, says they are strikingly distinguished from the other three kinds of feathers (contour, down, and semiplume feathers) by their peculiar position, their extraordinary slenderness, and by the entire deficiency or very small size of their vane; but he states subsequently that he believes the narrow white downy feathers of the Cormorant must be regarded as filoplumes, even though they are furnished with perfect vanes. Nitzsch further mentions that filoplumes are “associated with the contour feathers, one or even two filoplumes standing quite close to every contour feather of the head, neck, and trank, apparently issuing almost out of the same pouch of the skin.” Fuiloplumes are usually regarded as degenerate feathers because during development they have barbs arranged after the fashion of a down (plumula) feather. In the Mallard during development there are usually two or more minute filaments at each side of the filaments containing the wing-quill protoptiles. These minute filaments contain prefiloplume. In text-fig. 10 the two short filaments (P. 2) lying in contact with the large protoptile filament are prefilo- plume filaments. ; In the newly-hatched duckling the minute prefiloplume and the large wing-quill protoptiles escape simultaneously from their sheaths—being barely visible to the naked eye, the prefiloplume have apparently hitherto escaped the notice of ornithologists. In course of time, as the protoptiles are pushed from the skin by mesoptiles, the prefiloplume are pushed from the skin by filoplume. The wing-quill protoptile of a thirty-eight days’ duckling represented in PI. II. fig. 8 is attached to a developing mesoptile. At the base of the mesoptile are three young filo- plumes, two of which support prefiloplumes—a young filoplume with a prefiloplume attached to its tip is also seen in. Pl. V. fig. 17. 42* 626 PROF. J. COSSAR EWANT ON THE Notwithstanding their small size, the prefiloplume have the structure of prepenne. They are hence not dwarfed pre- plumule, as their name suggests, but dwarfed protoptiles. It is important to note that, though several developing filoplume may be intimately related to a protoptile, they are not imbedded in its sheath; each projects from a separate pit in the skin. At the end of the seventh week, when the filoplumes have reached a length of about 20 mm., they are still found in contact with the developing true feather. A filoplume from a forty-eight days’ duckling, still carrying the minute prefiloplume on its tip, is represented in-Pl, V. fig. 20. At this stage the filoplume still con- sists of a calamus, a shaft and an aftershaft; but as the duckling increases in size the filoplume in connection with the wing-quills degenerate. ach loses its prefiloplume, its aftershaft, and most of the barbs of its shaft, with the result that, as a rule, only the rhachis of the shaft and two or three vestigial terminal barbs are left. That the filoplumes are degenerate penne is supported by a study of the foot-feathers of the Barn Owl (Striw flammea). In a newly-hatched Barn Owl one large and two small filaments are seen projecting from under the foot-scales. Soon a well de- veloped protoptile escapes from the large mesial filament and a minute protoptile from each of the two small filaments. In course of time a simple plumose feather (metaptile), with a large aftershaft, but with a rhachis which suggests a filoplume, succeeds the large protoptile, and typical hair-like filoplumes succeed the small protoptiles. Tf further proof were needed in support of the view that filoplumes are degenerate penne, it is obtained by the examina- tion of the white neck feathers which form part of the nuptial plumage of Cormorants, z. e., the feathers Nitzsch assumed were tiloplumes. Instead of regarding the narrow neck feathers of Cormorants as filoplumes, it would be better to regard them as true feathers (metaptiles) which have undergone degeneration. It may be added that, notwithstanding the small size of the prefiloplume, a microscopic examination clearly proves that they are more intimately related to protoptiles than to preplumule. Seeing that the filoplumes are degenerate penne, it might be better if the name filopenna were substituted for filopluma. Ti. THe Compostrion oF FEATHERS. The feathers forming the coat of the adult Emu consist of a calamus and two blades, an outer, the shaft, and an inner, generally known as the aftershaft; the wing- and tail-quills of ducks consist of a calamus and a well-developed shaft, but there is only a minute vestige of an aftershaft. Hitherto it has been taken for granted that a feather with only a vestige of an after- shaft is more primitive than a feather with an aftershaft as long and as complex as the shaft, that, in fact, the aftershaft is not a a ee eed a NESYLING FEATHERS OF THE MALLARD. 627 primitive but a secondarily acquired feature. The chief reasons given for regarding the aftershaft as an accessory and secondarily acquired structure are: (1) that the aftershaft is developed from a forward elongation of the calamus, and (2) that the tip of the aftershaft of the new feather is never attached to the calamus of the feather about to be shed—that, for example, in the Emu the tip of the aftershaft of the feather of the second generation is from first to last free. If I succeed in showing that the aftershaft, like the shaft, grows from apex to base and is completed before the calamus 1s formed and that, during development, the tip of the aftershaft is connected with the calamus of the feather about to be shed is, in other words, for a time attached to the old feather—it may be safely assumed that the aftershaft is a primitive feature, and that a complete true feather (penna) like a down feather (plumula) consists of two shafts or blades. Owen evidently assumed that a true feather was made up of a calamus and a single shaft, for he states that “‘ besides the parts which constitute the perfect feather there is an appendage attached to the upper umbilicus called the accessory plume” (4). According to Nitzsch, the aftershaft ‘‘ originates from the under- side of the umbiliciform pit” (5). Gadow, in discussing the aftershaft, states that when present it is developed as ‘“‘a ventral elongation of the calamus,” and points out that if we consider the various types of nestling feathers ‘‘ with reference to the presence or absence of an aftershaft in the Teleoptiles, we are led to conclude that this appendage and eonsequently also the double feathers of certain Ratite are secondarily acquired not primitive features” (6). That the aftershaft is a primary feature was suggested in 1903 by Degen in a paper on Ecdysis. In the chapter on “ Adven- titious Plumage,” Degen writes: ‘‘I may mention here that owing to the yet ‘more primitive conditions of the feathers of the trunk in some Ratite, the extreme tips of the aftershafts, which in the Cassowary and the Emu attain a length equal to that of their main shafts, jointly support the new-growth feather with the latter” (7). But Degen, following Gadow, seems to adopt the view that the calamus is concerned with the formation of the aftershaft, for he assumes that while the one half of the calamus “deposits substantia rhachidis evterna for the main shaft, the other half deposits” the same substance required in the building up of the other shaft! Obviously the best way to throw light on the status of the aftershaft is to study its development. 1. The Development of the Aftershaft of Plumule.—tIn the Mallard a preplumula consists of a small conical calamus, a shaft usually made up of four barbs, two of which coalesce to form a short rhachis, and of an aftershaft also made up of four barbs. Soon after hatching the preplumule are pushed from the skin by plumule. Some of the plumule at the end of the fifth week, though still bearing preplumule on their tips, are 15mm. in 628 PROF. J. COSSAR EWART ON ‘tHE length and consist of two distinct bundles of barbs—a large outer bundle in a line with the shaft of the preplumula and a smaller bundle in a line with the aftershaft. In the plumula from a five weeks’ duckling represented in IBIS AY Ihe fig. 21, the outer bundle already consists of a rhachis and ~ eight pairs of barbs, but the inner or aftershaft bundle only consists of five pairs of barbs. By the end of the sixth week the development of both the shaft and aftershaft is complete, and the unsplit portion of the epidermic tube is giving rise to a calamus. In the completed plumula (Pl. V. fig. 18) the shaft consists of a rhachis and over twenty pairs of barbs; and the aftershaft of a rhachis with about a dozen pairs of barbs. From Pl. VI. fig. 21 it is evident that the aftershaft of a plumula in its development and structure agrees with the shaft ; as it is developed before the calamus it cannot be regarded as a secondarily acquired appendage produced from a forward elongation of the calamus. 2. The Development of the Aftershaft of Emu Feathers.—It has hitherto been often assumed that in the Emu the proto- ptile has ‘no distinct aftershaft.” That there is anaftershaft in the first, as wellas in the second and third Emu feathers, is indi- cated by the diagram, text-fig. 11. In the young Emus I have examined the protoptile aftershaft has, with rare exceptions, consisted of seven or more barbs longer than, or at least as long as, the barbs of the shaft. In the Emu protoptile given in Pl. VI. fig. 22, the aftershaft has a short rhachis. It is especially noteworthy that in the Emu the rhachis of the protoptile shaft, instead of ending in a hollow calamus containing ‘‘ cones,” is connected with the relatively thick rhachis of the shaft of the feather of the second generation by a solid band (Pl. VI. fig. 23). In a like manner the rhachis of the protoptile aftershaft is connected with the aftershaft rhachis of the second feather. When, as sometimes happens, the connecting bands occupying the position of a calamus separate from each other (Pl. VI. fig. 23), it becomes obvious that the tip of the aftershaft of the second feather is not free as generally assumed, but from the outset continuous with the protoptile aftershaft. Though as in plumule the rhachis of the aftershaft is formed by the fusion of two barbs, the distal portion of the rhachis of the shaft is simply.a continuation of the band proceeding backwards from the protoptile shaft. That the aftershaft of the feathers of the second generation is not ‘‘ developed as a ventral elongation of the calamus,” but develops from apex to base like the shaft, is made clear by Pls. VI. and VII. figs. 23, 24, & 25, which show the second feather at various stages of growth. As the destined length of the second-generation feather is reached the rhachis of the shaft expands, gives off six or more pairs of relatively long barbs, and thereafter becomes continuous with the rhachis of the shaft of the feather of the third generation (Pl. VII. fig. 26). Im a like manner the proximal end of the NESTLING FEATHERS OF THE MALLARD, 629 a rhachis of the aftershaft, after giving off several pairs of barbs but without appreciably increasing in width, becomes continuous with the rhachis of the aftershaft of the feather of the third generation (P]. VII. fig. 26). When (as was the case in the specimen represented in Pl. VII. fig. 27) the two connecting bands naturally separate from each other, it is impossible to escape Text-figure 11. Eee In this diagram the structure and relations of the feathers forming the first three generations in the Emu are indicated. sh.1, shaft, and af. 1, aftershaft of pro- toptile. sh.2, shaft, and af.2, aftershaft of a feather of the second generation. sh. 3, shaft, and af. 3, after shaft of a feather of the third generation. The aftershaft of the protoptile i is from the outset free, but the tip of the aftershaft of the feather of the second generation is, to start with, continuous with the protoptile aftershaft, and the aftershaft of the feather of the third generation is also for a time continuous with the aftershaft of the second generation feather— according to Gadow the tip of the aftershaft of the second generation feather is a forward prolongation from the calamus which is never during development connected by its tip with the “calamus” of the protoptile or first generation feather. from the conclusion that, just as the aftershaft of the second feather of the Emu is connected at the outset with the proto- ptile, the aftershaft of the feather of the third generation is connected with the feather of the second generation. The feathers represented in Pl. VII. figs. 26 & 27 (from a seven 630 _ bRor. J, COSSAR EWART ON THH months’ Emu bred in the Gardens of the Zoological Society of London) prove conclusively that in the Emu the aftershaft in origin and structure agrees with the shaft, while the feathers represented in Pl. VII. figs. 24 & 25 (from a much younger Emu in the Museum of the Philosophical and Literary Institute, Leeds) as clearly indicate that as the aftershaft is developed before the calamus it cannot be an outgrowth from the calamus. The first three generations of the feathers of an Emu are diagrammatically represented in text-fig. 11. Before dealing with the aftershaft in the Carinate, attention may be directed to the remarkable Emu feather of the second generation represented in P]. VIII. fig.28. In this growing feather the aftershaft, instead of resembling the shaft, simply consists of two barbs bearing barbules. An Emu feather with an aftershaft represented by two simple barbs is interesting, because in true feathers near the oil-gland of ducks two long barbs sometimes occupy the position of an aftershaft. That the two long barbs in this Emu feather represent a phase in the degeneration of the aftershaft is suggested by Pl. VIII. fig. 29, a developing true feather with the protoptile intact from the rump of an eighteen days’ Mallard duckling. Connected with the protoptile calamus are two bundles of barbs—a large bundle with a distinct rhachis which is obviously a developing shaft and a smaller bundle repre- senting an aftershaft. The smaller or aftershaft bundle is con- nected to the inner aspect of the protoptile calamus, as in the Emu feather, by two long barbs. In course of time the large bundle in Pl. VIII. fig. 29 would have developed into a shaft like the one represented in Pl. I. fig. 1, and the small bundle into an aftershaft with a short rhachis. Possibly long barbs which represent an aftershaft will be met with in the feathers of other Ratite. The two long barbs in the Emu feather in Pl. VIII. fig. 28 support the view that the coat of the adult Emu consists of true feathers and not, as Pycraft suggested, of mesoptiles. 3. The Development of the Aftershaft of the True Feathers of Penguins.—In the Ringed Penguin the protoptiles though um- belliform may be said to consist of two sets of barbs—an outer representing a shaft and an inner occupying the position of an aftershaft. Though in ducks and geese there is a well-developed protoptile calamus containing “cones,” in penguins as in the Emu a true protoptile calamus is never developed. Soon after the protoptile escapes from its sheath the part of the epidermic tube representing a calamus splits into slender cords, which connect the barbs of the protoptile with barbs of the developing mesoptile. The mesoptile consists of a shaft made up of seven or more simple barbs which end in the tip of the true feather, and of a complex aftershaft connected with the aftershaft of the true feather. In text-figs. 3 & 4 the mesoptile aftershaft is con- nected to the aftershaft of the true feather by a narrow band formed by the fusion of the distal portions of the barbs of the NESTLING FEATHERS OF Tite MALLARD. 631 aftershaft of the true feather. Pycraft considered the “ ribbon- shaped stem” connecting the mesoptile to the aftershaft of the true feathers as somewhat remarkable, because ‘‘in all other instances yet known the nestling-down when attached to contour feathers is attached to the rami (barbs) of the main shaft” (9). When it is realised that in Penguins the mesoptile may have a well-developed aftershaft the connection of the mesoptile with the aftershaft of the true feathers ceases te be remarkable. Though in the Ringed Penguin the mesoptile aftershaft is de-’ cidedly larger and more complex than the shaft, it is developed at the same time as the shaft, and is directly continuous with the aftershaft of the true feathers. That in Penguins the aftershaft of the true feathers has the same origin as the shaft is made abundantly evident when the development of the feathers of the fourth generation is studied. In P]. VIII. figs. 30 & 31 we have feathers from a moulting King Penguin in the Scottish Zoological Park. From fig. 30 it is evident the new feather has from the first been double, consisted of outer barbs representing a shaft and of inner barbs in the position of an aftershaft. Had the develop- ment proceeded the tip of the aftershaft would have been set free when the old feather was shed. In fig. 31, the shaft having been removed, the aftershaft is seen emerging from the tip of the sheath which invested the growing feather. These figures show that in Penguins the aftershaft of a feather of the fourth genera- tion is for a time as complete as in the Emu. The aftershaft of , the mesoptile and of the first true feather of a Ringed Penguin are diagrammatically represented in text-figs. 3 & 4. 4. The Development of the Aftershaft in the Ptarmigan, the Common Fowl, and the Guillemot.—As already mentioned, it has been suggested that in Grouse, Fowls, and other Galli the feathers forming the first nestling coat may correspond to the mesoptiles forming the second nestling coat of Penguins. Against this view it may be pointed out (1) that unlike the mesoptiles, but like the protoptiles of Penguins and Ducks, the first nestling feathers of the Galli are developed in filaments; (2) that in structure they are intermediate between the protoptiles of Penguins and Ducks; and (3) that they are sometimes succeeded by vestiges of mesoptiles. Degen mentions that in Grouse the aftershaft may be two-thirds the length of the shaft but, as Pl. IX. fig. 32 shows, the aftershaft in Grouse may be as long and as complex as the shaft—the only difference in the feather figured apart from its width is that, as in the Emu, the rhachis of the aftershaft is slightly shorter than the rhachis of the shaft. In the Ptarmigan the protoptile often con- sists of fifteen barbs, eight representing a shaft with a short rhachis, and seven forming an aftershaft also with a short rhachis. From the protoptile calamus there proceed about twenty barbs, which in some cases readily separate into three distinct bundles—a large one destined to form the tip of the shaft, and two small ones which later unite to form the tip of the 632 PROF. J. COSSAR EWART ON THE aftershaft. In some metaptiles the rhachis of the shaft extends to the tip of the feather, in others it ends in two slender barbs some distance from the tip. Though at the outset in Grouse both the shaft and the aftershaft of the true feather may consist of several more or less distinct bundles, sooner or later the outer barbs give rise to the shaft, while the inner form an aftershaft. When the destined number of barbs for the shaft and aftershaft have been produced the splitting of the epidermic tube is arrested, and steps are taken to form a calamus in which “cones” make their appearance. Evidently in Grouse, as in Penguins, the aftershaft is not a forward elongation of the calamus. An interesting stage in the development of a Ptarmigan true feather is given in Pl. IX. fig. 33. It has been asserted once and again that ‘in the great feathers which form the ‘quill’ or‘ flight’ feathers (remiges) of the wing and the tail-quills (rectrices) the aftershaft is in- variably wanting” *, Assuming that in the case of the wing- quills the aftershaft is invariably absent, it might be argued, that these all-important feathers differed in their origin and history from the other true feathers, that while the trunk feathers may possibly have developed from an epidermic tube, the wing- and tail-quills were formed by the splitting of elongated scales. Attention has already been directed to the fact that in the Mallard during the earlier stagesin the development of the wing- and tail- quills, there is a vestige of the distal portion of an aftershaft (text- figs. 5 & 7). That in wing- and’ tail-quills there may also be a vestige of the proximal portion of an aftershaft is made evident by Pl. IX. fig. 34, which represents the aftershaft of a wing-quill from a full-grown Grouse. Fowls are not supposed to have the aftershaft as well developed as Grouse, nevertheless, as Pl. 1X. fig. 35 shows, the aftershaft in the silky breed may be as long as the shaft. In its development the aftershaft in the silky breed follows the same course as in Ptarmigan. _ In the Guillemot the aftershaft is, to start with, sometimes as long as the shaft. A Guillemot feather from the side of the trunk is represented in Pl. X. figs. 36 and 37. In this case the protoptile consists of seventeen barbs. Of the seventeen barbs nine end in the tip of the shaft of the true feather, four end in barbs of the shaft which join the rhachis a considerable distance from the tip, and four, which represent the protoptile aftershaft, are connected with the aftershaft of the true feather (Pl. X. fig. 37). Bvidently the history of the aftershaft in Grouse, Fowls, and Guillemots affords no support to the view that the aftershaft is a secondarily acquired appendage, and may be said to fully estab- lish the view that a complete true feather like a plumula consists of a calamus and two shafts. * Pycraft, ‘A History of Birds,’ p. 9. NESTLING FEATHERS OF THE MALLARD. 633 III. Ortein anp History oF FEATHERS. It seems to be generally assumed by naturalists and others interested in birds that feathers are modified scales, and were originally acquired to enable birds to fly. When evidence of the scale origin of feathers is asked for, the reply usually is, ‘If you accept the view that birds were evolved from Reptile-like ancestors you may safely also accept the view that feathers are specialized scales.” Having adopted the view that feathers are modified scales, investigators interested in the coat of birds proceeded to inquire if each feather corresponds to an entire scale or to part of a scale, Wiedersheim favoured the view that each feather represents a complete scale, but Bornstein (13) and others came to the conclusion that a feather papilla only repre- sents a fragment of a scale. How scales were transformed into feathers has been indicated by various writers. Baron Frances Nopcsa, in a paper “ On the Origin of Flight,” states that we may quite well suppose “ birds originated from bipedai long-tailed cursorial reptiles which during running oared along in the air by flapping their free anterior extremities,’ and added that the double running and flapping action would subsequently easily lead to an enlargement of the posterior marginal scales of the antebrachium, and even- tually to their development into actual feathers(10). Lull, another believer in the transmission of acquired somatic charac- ters, points out that the anterior extremities of birds when in the making would be more effective if their breadth was increased, and that an increase of the size of the scales along the arm margin would be a ready means to this end. Lull then proceeds to say ‘the “scales would extend, lighten, and ultimately evolve into feathers which would not only subserve the function of flight, but acting as clothing retain and aid in the increase of tempera- ture” (11). Similar views are held by W. K. Gregory, who tells us the Pro-Aves jumped lightly from branch to branch and finally from tree to tree partly sustained by the long scale-like feathers of the pectoral and pelvic wings. The view that feathers are modified scales has long been held by Pyeraft. In a work published in 1910 Pycraft points ont that, to start with, the body clothing of birds probably consisted of relatively large scales, ‘those covering the hinder border of the incipient wing growing longer would still retain their original overlapping arrangement and along its hinder border would in their arrangement, appearance, and function simulate the quills of modern birds; as their length increased they became also fimbriated and more and more efficient in the work of carrying the body through space.” It is now evident that the wing-quills of birds are preceded by nestling feathers, and may be regarded as modified protoptiles: hence the question at issue is, Are the protoptiles modified scales? 634 PROF. J. CoSsAR EWART ON THE It is inconceivable that the small protoptile forerunrers of the wing-quills of Ducks were evolved from enlarged scales, which as they increased in length became fimbriated. The more the history of feathers is studied the more untenable becomes the belief in their origin from scales. When special attention 1s directed to the phases through which feathers pass during their development, it becomes evident that in an attempt to trace their origin we are not concerned with the wing-quills or other contour feathers, but with the feather papille, and still more with the filaments, in which the nestling feathers are developed. It is conceivable that by the appearance of new and dominant factors in the germ-plasm, scale papille, instead of developing into scales, developed into feather filaments, but it is difficult to imagine how scales by growing longer and splitting were transformed into double-shafted nestling feathers. There is little chance of the geological record ever throwing any light on the origin of either feather papille or feather filaments, hence we must make the most of the embryologicai record. Thus far the embryological record has afforded no support to the view that scales in whole or in part were concerned in the making of feathers. When birds first acquired feather filaments it is impossible to say, but we may assume that when the remote ancestors of birds and mammals became warra-blooded any outgrowths from the skin which prevented a loss of heat would from the first count in the struggle for existence. A hint as to the nature of the coat in the progenitors of mammals we have from Armadillos, and an indication of the kind of coat worn by the Pro-Aves is obtained by studying the deve- lopment of nestling feathers, more especially of the protoptiles of feather-footed birds, The Six-banded Armadillo has an extensive scaly covering, but only a sparse coat of hair; birds, with few exceptions, have during development a more or less complete coat of filaments*, and probably in Owls the foot-feathers still bear the same relation to scales as in the primeval birds. In the developing Armadillo a stage is eventually reached when hairs are seen projecting from under the developing scales or from the skin between the scales. ‘The number of hairs developed in con- nection with scales varies. There are often groups of three hairs projecting from under the tail scales of mammals; in the Six- banded Armadillo there are often three or four hairs to a scale, but sometimes only one; in the small Argentine Armadillo Chlamy- dophorus, as text-fig. 12 shows, fifteen hairs may be developed in connection with one foot-scale. In Armadillos and other mammals having scales as well as hairs there is no evidence that the hairs are ever developed from scales. * All the birds I have examined, with the exception of the House Sparrow, had a more or less complete coat of protoptiles—in the Rook a nestling coat is said to be absent, but even in Rooks there are a few well-developed nestling feathers, it. - Wy i oY ih ott NESLLING FEATHERS OF THE MALLARD. 635 That in ancient birds feathers were intimately associated with scales is suggested by the feather papille having approximately the same arrangement as the scales of reptiles, by feathers emerging from under scales above the tarsal joints of the Wood- cock, and by chicks and ducklings having vestiges of scales in connection with the feather filaments of the legs. In the case of Grouse and other birds with only one feather emerging from under a foot-scale, we are tempted to assume that each foot- feather is developed from part of the scale it is associated with. But when we find three feathers projecting from under the foot- scales of the Barn Owl (Séria flammea) and seven or more feather filaments occupying an area equivalent to a foot-scale, we Text-figure 12. Scale from the foot of the small Argentine Armadillo (Chlamydophorus) from under which proceed a number of hairs developed from the skin underlying the scale. The three filaments proceeding from under the foot-scales of Barn Owl embryos like the hairs associated with the scales of mammals are developed from the skin under the scales and not by a splitting of the scales or from fragments of the scales. are less inclined to admit that each of the foot feathers of Grouse corresponds to part of a scale. When the development cf the foot feathers of the Barn Owl is studied, one large and two small filaments are found projecting from many of the scales; the large filament contains a protoptile, and each small filament contains a prefilopluma. Similar groups of filaments which occur above the tarsal jot are overlapped by scale vestiges, but in the three feather groups on the head scales are conspicuous by their absence. The well-developed scales in connection with the foot- feathers and the vestiges of scales in connection with feathers above the tarsal joints support the view that birds, to staré with, had a complete scaly covering and a sparse coat of feather filaments. 636 PROF, J. COSSAR EWART ON THE As in feather-footed birds the feather papille are from the first as distinct from the scale papille as the hair rudiments are distinct from scales, there seems no escape from the conclusion that the foot feathers are not modified scales. Further, when it is realized that the foot feathers often belong to a simple primitive type, have the aftershaft as long as the shaft, and that in many birds there are seven or more feathers occupying the position of an ancestral scale, we are justified in assuming that feathers, like hairs, were originally intimately related to, but not derived from, scales. As a matter of fact, believers in the scale origin of feathers admit that the foot feathers of recent birds are not derived from the scales from under which they project. Just as it has been assumed that the aftershaft in Emu feathers is a secondarily- acquired structure, it is assumed the scales on the feet of birds are secondary structures. That the scales overlapping the feathers of modern birds were secondarily acquired and that the foot- feathers were formed from the original foot-scales, are gratuitous assumptions wholly unsupported by facts. How or when feather filaments made their appearance it is impossible to say, but we may safely assume that many centuries elapsed before there was evolved in each filament a more or less complex feather; needless to say that unless a filament in which a feather was evolving was of more use than a simple filament, natural selection took no part in providing birds with their original coat of what are now familiarly known as nestling feathers. On the other hand, when one bears in mind that blood- vessels extend right to the tip of a feather filament, and not merely a short distance into its root as in a hair, it is probable that early specialization of the filaments was well-nigh inevitable. Given a constant flow of blood through the axis (pulp) of the filament, an inner epidermic layer capable of rapidly adding new cells to the middle layer, an outer epidermic layer forming a protecting sheath and, in addition, an ever increasing demand for a coat capable of arresting the flow of heat from the skin, changes were almost bound to take place in the bristle-lke outgrowths pro- jecting from under or between the scales of the remote ancestors of birds. Judging by what takes place to-day during the development of the simple protoptiles of pigeons, the first step in the making of a feather probably consisted in the splitting of the middle epidermic layer to form hair-like barbs. The result of this splitting of the hollow cone formed from the middle epidermic layer of a pigeon ‘ filament is represented in text-fig. 13. The appearance of a completed pigeon protoptile after escaping from its protecting sheath is represented in Pl. X. fig. 38. The protoptiles of Penguins represent a second stage in the evolution of true feathers (text-fig. 3). Like the pigeon protoptiles all the barbs are alike (form a tassel or umbel), but by sprouting they have acquired barbules, te tt \ eae NESTLING FEATHERS OF THE MALLARD. 637 A series of links connect the relatively simple umbelliform protoptiles of Penguins with the highly-specialized protoptiles of Ducks and Emus. There are good reasons for believing that the warm-blooded progenitors of birds and mammals were evolved out of reptile-like ancestors during a period when desert con- ditions prevailed. How long this period lasted, and to what extent feathers had been specialized when it came to an end it is impos- sible to say, but judging by what happens during the development of the wing-quills of the Mallard, it may be assumed the period of aridity came to an end, aiid a glacial period had fairly set in long before wing-quills and other true feathers were evolved. In other words, before protoptiles had been modified to form meta- ptiles and teleoptiles, progress in the evolution of true feathers was Text-figure 13. Section through the filament of a Pigeon in which the middle epidermic layer has been split longitudinally to form barbs. s., the thin sheath which disintegrates immediately after hatching. 6., one of the eight barbs formed by the splitting of the hollow cone formed by the middle epidermic layer. p., the pulp con- taining blood vessels, 7. e., the dermis or true skin which extends during development to the tip of the protoptile. After Davies. arrested owing to the necessity of providing birds with a fur- like coat as capable of arresting the flow of heat from the skin as the dense coat now worn by Polar Bears and other Arctic mammals. The discovery by Dr. Eagle Clarke and Mr. Pycraft of a mesoptile coat in penguins has profoundly modified our views about the plumage of birds. We do not yet know if the remote ancestors of all our modern birds acquired a mesoptile coat — probably in the case of the ancestors of the Ratite this was unnecessary—but the more the nestling feathers are studied the evidence of the existence of two coats of prepenne in the ancestors of modern birds is increased. The fur-like mesoptile coat is probably as well developed to-day in penguins as it was in their remote ancestors of the Jurassic Age; relatively large mesoptiles _ still occur in ducks and geese, and I have found recently distinct vestiges of a mesoptile coat in pigeons, How long birds, or most 638 PROF. J. COSSAR EWART ON THE of them, in order to survive required a fur-like coat throughout life we have no means of knowing; but this much is certain, that in most recent species of birds the mesoptile coat is in the act of being suppressed, or has already disappeared—even in the King Penguin chicks it is probably never worn in the Antarctic for more than 9 or 10 months. From what we know about the mesoptile coat it seems to have been originally produced partly by lengthening the protoptiles and partly by the development of Text-figure 14. A developing mesoptile of an Adélie Penguin. In penguins the mesoptiles consist, as a rule, of numerous barbs which in some areas may reach a length of over three inches. A true protoptile calamus containing “cones” though present in ducks is seldom mei with in other birds. As the protoptile is pushed from the skin the inner ends of the proteptile barbs, destitute of barbules, separate from each other and are subsequently seen to be continuous with mesoptile barbs. mb., one of the numerous barbs of the developing mesoptile. pé., one of the nine protoptile barbs, each of which is connected with a mesoptile barb. b., one of the nine slender cords which together for a time playeu the part of a protoptile calamus. new barbs. This is suggested by text-fig. 14, which represents a developing mesoptile feather from the breast of an Adeélie Penguin. The nine barbs forming the protoptile are directly continuous with nine mesoptile barbs; the rest of the mesoptile consists of barbs having no connection with the protoptile. As Pyeraft points out, the second coat of the Adélie Penguin is “a 4 ; i i ‘a NESTLING FEATHERS OF THE MALLARD. 639 long, dense woolly covering, so dense that it is difficult to expose the skin.” In course of time the protoptile coat is shed and the mesoptile coat is pushed from the skin by the true feathers. The long fur-like mesoptile coat of the King Penguin suggests that during the Ice Age, which succeeded the period of desic- cation, the evolution of true feathers was arrested, and that the mesoptiles cannot be regarded as representing a stage in the evolution of true feathers. Though to the naked eye the meso- ptile barbules seem to be identical with the barbules forming the protoptiles, they are seen to differ when examined under the microscope—instead of being nearly straight they are spirally twisted at their origin. In the case of King Penguin chicks the mesoptiles differ from the protoptiles more than in the Adélie chicks, and, as already pointed out, the mesoptiles of the Ringed Penguin decidedly differ from both the protoptiles and the metaptiles. In all the three Penguins mentioned the first true feathers differ from the prepennze which precede them. Hence, though the prepenne and all the true feathers which succeed them must be regarded as links in a chain, it should be borne in mind that the first link differs both in function and structure from the second, and the second from the third. Further, it should be borne in mind that there are good reasons for believing that, but for the advent of an Ice Age, the protoptiles at a comparatively early period would have been succeeded by true feathers, and that probably long before Archeopteryx made its appearance on the scene some millions of years ago, steps would have been taken in most birds to suppress the fur-like coat of mesoptiles, now apparently only well developed in Penguins. In conclusion, it may be stated that a study of the develop- ment of birds suggests: (1) that their coat, to start with, consisted of various kinds of simple filaments; (2) that in course of time the hollow epidermic cone in each filament, in some incompre- hensible way, was converted into a protoptile or preplumula: (3) that the protoptiles in some cases soon acquired the chief characteristics of true feathers ; (4) that for a time progress was arrested in order to provide birds (or most of them) with a fur- like (mesoptile) coat, by way of giving them a chance of surviving during the cold phases of an Ice Age: (5) that as the climate improved the mesoptile coat was in many cases superseded by a coat of true feathers ; and (6) that by the specialization of feathers along the posterior margin of the hands and forearms and the sides of the tail, birds of the Archzopteryx type were eventually evolved capable of flight, or at least of gliding easily from tree to tree. Text-figures 1, 2, 8, 9, 12 are from drawings by Miss Augusta Lamont, B.Se., text-figures 3-7 and 10, 11, 13, 14 are from drawings by the author, all the other figures are from photographs by Mr. John Chisholm. Towards defraying the cost of the drawings and photographs a grant was obtained from the Moray Fund of the University of Proc. Zoou. Soc,— 1921, No. X LIII. 43 640 PROF. J. COSSAR EWART ON THE Edinburgh and a grant has been voted from the Carnegie Trust for the Universities of Scotland, towards the cost of publishing this paper. In addition to material in the form of ducklings, goslings &c., bred at Craigybield, Penicuik, it ought to be mentioned that valuable contributions for the study of nestling feathers were provided by the Director of the Zoological Society of Scotland, and that but for the Institution of the Scottish Zoological Park I would not have been led to undertake research work on the Structure and History of Feathers. further, it ought to be mentioned that during the investigation I received valuable material from the Zoological Society of London, the Royal Scottish Museum, and the Philosophical and Literary Institute of Leeds, and that I am especially indebted for specimens that proved most useful to Colonel Ashley, M.P., Miss Clerk of Penicuik, William J. Gordon, Esq., of Windhouse, Yell, Shetland, Professor Meakins, the University of Edinburgh, and Professor Philip J. White, the University of North Wales. References to Literature cited in Paper. 1. Iveram, CoLtitinewoop.—‘ The Ibis,’ October 1920, p. 856. 2. Gavow, Hans F.—Newton’s ‘ Dictionary of Birds,’ p. 243. 3. Pycrart, W. P.—‘‘ The Emperor and Adélie Penguins,” Nat. Ant. Expedition, Zoology, vol. it. p. 12. 1907. 4. Owren, R.—Comparative Anat. vol. il. p. 235. 5. Nirzscu.— Pterylography,’ English Translation, p. 8. 1867. 6. Gapow, Hans F.—Newton’s ‘ Dictionary of Birds,’ p. 245. 7. Decen.—Trans. Zool. Soe. vol. xvi. pp. 362 & 363. 8. Pycrarr, W. P.—-Trans. Zool. Soe. vol. xv. p. 164. 9. an Nat. Ant. Expedition, Zool. vol. ii. pp. 7 & 8. 10. Norcsa.—Proe, Zool. Soc. 1907. 11. Luti.— Organic Evolution,’ pp. 533 & 534, 1917. 12. Pycrarr.— A History of Birds,’ p. 39. 1910. 13. Bornsrey.—Archiv f. Naturgesch., Band I. Supplement, heft 4. EXPLANATION OF THE PLATES. PATE I. Fig. 1. Protoptile and metaptile from vicinity of preen gland of a 38 days’ Mallard. The mesoptile has been suppressed. T he] long barbs proceeding from the inner aspect of the protoptile calamus towards the calamus of the metaptile are vestiges of a metaptile aftershaft. X 2. 2. Protoptile, mesoptile, and tip of a wing-quill of a 5 weeks’ Indian Runner duckling. 3. Protoptile, mesoptile, and tip of a wing-covert of a 6 weeks’ Emden gosling. 4, Tip of tail-quill of a young Ringed Penguin (u. s.). ‘Prats II. .5. Tail-quill and protoptile of a 5 weeks’ Mallard duckling (n. s.). 6. A 10 days’ Mallard embryo. The tail-quill papille are well developed, but there is tae indication of the wing-quill papille. 7. Wing of a 25 days’ Mallard duckling showing the wing-quill protoptiles (n.s.). 8. Wing- quill protoptile from a five ‘weeks’ Mallard duckling attached to tip of developing mesoptile, ee 10. ie 12. Fig. 13. 14. NESTLING FEATHERS OF THE MALLARD. 641 Prats III. . Filaments from a 13 days’ Mallard embryo; the long filaments contain developing protoptiles, the small ones preplumule or prefiloplume. x 2. Wing-quill protoptile artificially removed from a filament of a 26 days’ Mallard embryo. Aftershaft of the protoptile of a duckling. Wing of a 38 days’ Mallard duckling showing mesoptiles between the pro- toptiles and the developing primary and secondary wing-quills. Lying between developing wing-quills (each still invested by a sheath) are wing- coverts in which the mesoptile is vestigial. Prats IV. Protoptile, mesoptile, and tip of wing-quill of a Chinese gosling. Though the mesoptile seems to be umbelliform and to end in a calamus, it really consists of a shaft and an aftershaft from which slender threads proceed to the tip of the true feather. : In this figure the proximal portion of the mesoptile is represented. It con- sists of a narrow, perforated band representing the rhachis of the meso- ptile shaft, of a broad perforated band representing the rhachis of the aftershaft, and of numerous slender threads which connect these bands with the shaft and the vestigial aftershaft of the true feather. 15. Nestling tail of a 7 days’ Mallard duckling. The protoptiles at each side of the middle line have beeh pushed from the skin by the developing true tail-quills. 16. The “nestling tail”’ at the middle of the fifth week. It now consists of growing true tail-quills as well as of protoptiles, usually with the excep- tion of the ninth from the middle line at each side the protoptiles are directly continuous with the tail-quills—in other words, in the tail of the Mallard duckling nearly all the mesoptiles have been completely suppressed. Prats V. Fig. 17. Nestling feathers from the back of a 33 days’ Mallard duckling. The large protoptile has a distinct calamus containing “cones.” ‘To the left of the calamus are: (1) a filoplume with a minute prefiloplume on its: tip; and (2) two plumule each carrying a preplumula, to the right a plumula with a preplumula on its tip. X 5. 18. A plumula from a young Aylesbury duck showing a large shaft consisting of long, slender barbs springing from a relatively short rhachis and a well-developed aftershaft with a short rhachis. 19. The calamus of a naturally shed plumula with part of the sheath of the growing plumula attached. 20. A developing filoplume with a prefiloplume at its tip from a 7 weeks’ Aylesbury duckling. X 2°5. Prate VI. Fig. 21. Developing plumula from a 5 weeks’ duckling. The aftershaft is developing from apex to base in exactly the same way as the shaft. X 4. 22. A protoptile from a young Emu showing an aftershaft with a short rhachis. 23. Emu feathers of the first (protoptile) and second generation. The protoptile shaft is directly continuous with the shaft of the second feather, and the protoptile aftershaft is continuous with the aftershaft of the second feather. Pirate VII. Figs. 24 & 25. In 24,we have an early stage in the development of an Emu feather of the second generation. In 25 the development of the second generation feather has been nearly completed. In both cases the aftershaft is deve- loped in the same way as the shaft—the tip of the aftershaft of the second feather is continuous with the protoptile aftershaft and the tip of the shaft is continuous with the protoptile shaft, 642 ON THE NESTLING FEATHERS OF THE MALLARD. Fig. 26. Emu feathers of the first, second, and third generations. The tip of the aftershaft of the feather of the third generation is continuous with the rhachis of the aftershaft of the feather of the second generation. 27. In this specimen the tip of the aftershaft of the second generation feather had lost its connection with the protoptile, and the tip of the aftershaft ot the third generation feather had lost its connection with the “ ealamus ” of the feather of the second generation. Pratt VIII. Fig. 28. Emu protoptile and a developing feather of the second generation, the aftershatt of which is represented by two barbs. X 5. 29. Developing feather from rump of 18 days’ Mallard duckling. This feather consists of two portions, one, the larger, representing a shaft, and one representing an aftershaft, connected like the Hmu feather given in fig. 28, to the protoptile hy two barbs. 30. Breast feathers from an adult moulting King Penguin. The growing feather consists of a shaft and an aftershaft connected by their tips with the calamus of the feather about to be shed. 31. In this figure the shaft of the new feather has been removed to show the connection between the developing aftershaft and the old feather. Prate DX: Fig. 32. Grouse feather with an aftershaft as long and as complex as the shaft. 33. A protoptile and a developing true feather from a young Ptarmigan. The part of the developing feather with a distinct rhachis represents the shaft, the part m which a rhachis has not yet appeared represents an aftershaft. 34. Calamas and aftershaft of the wing-quill of a Grouse. 35. Feather of a fowl of the Silky breed with an aftershaft as long as the shaft. Prate X. Figs. 36 & 37. Feather from a nestling Black Guillemot. 36, represents the shaft, and thirteen barbs of the protoptile. 87, the four remaining protoptile barbs connected with the rhachis of the aftershaft which was detached along with the calamus. 38. Protoptile of a Wood Pigeon connected to the tip of a true feather by a vestige of a mesoptile. The pulp projects from the tip of the teleoptile. Note that the barbs resemble hairs, have no barbules. SS: ON A MALE KIVU GORILLA. 643 EXHIBITIONS AND NOTICES. May 10th, 1921. Dr. A. Smirx Woopwarp, LL.D., F.R.S., Vice-President, in the Chair. Sir S. F. Harmer, K.B.E., F.R.S8., exhibited a mounted spe- cimen of a young female /hinoceros sondaicus, the one-horned ** Javan Rhinoceros,’ presented by T. R. Hubback, who shot it in Lower Tenasserim, to the British Museum (Natural History). Attention was called to the absence of a horn in the female, a character which Mr. Hubback believes to be normal. A young male of this species, from Java, was formerly exhibited in the Society’s Gardens, and was described by Dr. P. L. Sclater, F.R.S. (P. Z. 8. 1874, p. 182; Trans. Z. S.ix. p. 649). Although osteo- logical material is to be found in various Museums, there appear to be extremely few mounted specimens in existence. The Javan Rhinoceros, which formerly ranged from Assam to Lower Bengal, the Malay Peninsula, and Java, has become increasingly rare during recent years, and it is in urgent need of protection, Its occurrence in Sumatra and Borneo has been questioned, but reference was made to the statements of W. Volz. (‘ Nord- Sumatra,’ vol. ii. p. 372, Berlin, 1912), as proving its occurrence, in limited numbers, in North Sumatra. Mr. THos. Gerrarb, F.Z.8., exhibited, and made remarks upon, a series of photographs of the skeleton of an old male Kivu Gorilla (Gorilla gorilla beringert Matschie), the property of Lord Rothschild, and also a series of photographs showing the milk- dentition in the skull of a young African Elephant. On behalf of Lord Rothschild, Mr. Gerrarp exhibited a speci- men of a male Kivu Gorilla (Gorilla gorilla beringert Matschie) mounted by Messrs. Rowland Ward, and stated that it differed from the other races of Gorilla externally by its long, thick pelage, intense black colouring, and the large fleshy protuber- ance on the top of the head, which often measured 6 inches by 6 inches (7. ¢. width and height). The skull was much wider at the occiput even than in G. gorilla matschiet Rothschild. The specimen was a very aged male, long past the age of repro- duction, and was found solitary, several miles from the two herds inhabiting the region. The specimen was of great pathological 644 ON SNARE-BUILDING CADDIS WORMS. interest, as it was the first instance known of pyorrhea alveolaris in a Gorilla. Habitat Mikeno Volcano, Kivu District. Mr. D. Seru-Smiru, F.Z.8., exhibited a series of skins and photographs of some polyandrous Birds, and described their mating habits. Dr. Cuas. F. Sonnrac, F.Z.8., exhibited and made remarks upon, some abnormalities in the Carnivora. Mr. L. T. Hoesen, F.Z.8., and Mr. HE. A. Spauu exhibited, and made remarks upon, the effect of Pineal Gland administra- tion on the pigment-cells of Frog Tadpoles. May 24th, 1921. Prof. E. W. MacBrips, D.Sc., F.R.S., Vice-President, in the Chair. The Srecrerary read the following Report on the Additions to the Society’s Menagerie during the month of April ;— The registered additions to the Society’s Menagerie during the month of April were 198 in number. Of these 81 were acquired by presentation, 27 were deposited, 84 were purchased, and 6 were born in the Menagerie. The following may be specially mentioned :— A pair of Lion Cubs (/elis leo), from Somaliland, presented by Sir Geoffrey Archer, K.C.M.G., on April 21st. 2 Sloth-Bear Cubs (Melursus ursinus), from the Central Pro- vinces of India, presented by Capt. R. C. Ansdell on April 5th, 2 King Vultures (Gypagus papa), from Tropical America, purchased on April 6th. 2 Spot-billed Toucanets (Selenidera maculirostris), from Brazil, purchased on April 28th. A collection of Tropical Fish, including several species new to the Collection, purchased on April 6th. Miss L. E. Cunssman, F.E.S., exhibited, and made remarks upon, a series of lantern-slides illustrating the habi's, in the . larval stage, of the Hydropsychide (Snare-building Caddis Worms). DISTRIBUTION OF BIRD-LIFE IN PERU. 645 Major Stanuey 8. Frower, O.B.E., F.Z.S., exhibited examples of three species of Egyptian Jerboas—Jaculus jaculus, Jaculus orientalis, and Scarturas tetradactylus, the latter being the first living specimens ever brought to England; and made remarks upon the distribution of the three species in North Africa. June 7th, 1921. Prof, J. P. Hiuu, F.R.S., Vice-President, in the Chair. Major Srantey S. Fiower, O.B.E., F.Z.S., exhibited, and made remarks upon, some living specimens of Testudo leithit and ar example of Z'estudo ibera. Dr. P. Cuatmers Mircueni, C.B.H., F.R.S., exhibited, and made remarks upon, a photograph of the Death-mask of a young Gorilla, presented to the Society by Prof. Howard Macgregor, of Columbia University, New York. Dr. Franx M. CuapmAn gave an account of the distribution of Bird-life in the Urubamba Valley of Peru, and of his expedition down the Urubamba river from its sources in the Puna zone, at an elevation of 14,200 feet, to the Tropical zone at 3500 feet. He reviewed the existing climatic conditions in Peru, and com- mented on the faunal characteristics of the Puna, Temperate, Subtropical, and Tropical zones encountered in the descent. A comparison of the Avifauna of the forests of the humid Temperate zone (alt. 10,000—12,500 ft.) with that of the adjoining treeless, grass-grown plains and slopes of the Puna zone, showed that that of the first-named area has evidently been derived from the tropical forests at the eastern base of the Andes; that of the latter from the South Temperate zone plains of Argentina and Patagonia. The environmental change experienced in passing from the hot, tropical forests, at the foot of the mountains, to the cold temperate forests near their summits, is much greater than that existing between the plains of Patagonia and’ those of the Interandine tableland. This is obviously raflected in the degree of differentiation exhibited by the birds inhabiting these elevated forests and plains. Thus 55 per cent. of the genera of birds inhabiting the Temperate zone forests are endemic, while 646 ON SOME MARSUPIAL EMBRYOS. only 7 per cent. of the genera found on the plains of the Puna are endemic. These facts are believed to indicate that the extent of environmental change, expressed chiefly in climatic conditions, is a more important evolutionary factor than those exerted by time of isolation or extent of distance from the presumed ancestral forms. Prof. J. P. Hruz, F.R.S., V.P.Z.S., exhibited, and made remarks upon, a series of lantern-slides of some Marsupial embryos, dealing especially with the Koala (Phascolarctos) and the Wombat (Phascolomys). No. 217. ABSTRACT OF THE PROCHEDINGS OF THE ZOOLOGICAL SOCIETY OF LONDON.* May 10th, 1921. Dr. A. Smita Woopwarp, LL.D., F.R.S., Vice-President, in the Chair. Sir 8. F. Harmer, K.B.E., F.R.S., exhibited, and made re- marks upon, a mounted specimen of Rhinoceros sondaicus. Mr. THos. GerRRARD, F.Z.S., exhibited, and made remarks upon, a series of photographs of the skeleton of an old male Kivu Gorilla (Gorilla gorilla beringeri Matschie), the property of Lord Rothschild, and also a series of photographs showing the milk- dentition in the skull of a young African Elephant. On behalf of Lord Rothschild, Mr. Grerrarp exhibited the mounted specimen of the Kivu Gorilla (mounted by Messrs. Rowland Ward), and stated that it differed from the other races of Gorilla externally by its long thick pelage, intense black colouring, and large fleshy protuberance on the top of the head. Mr. D. Sera-Suirn, F.Z.S., exhibited a series of skins and photographs of some polyandrous Birds, and described their mating habits. * This Abstract is published by the Society at its offices, Zoological Gardens, Regent’s Park, N.W., on the Tuesday following the date of Meeting to which it refers. It will be issued, along with the ‘ Proceedings,’ free of extra charge, to all Fellows who subscribe to the Publications ; but it may be obtained on the day of publication at the price of Strpence, or, if desired, sent post-free for the sum of Stix Shillings per annum, payable in advance. 26 Dr. Cuas. F. Sonwrac, F.Z.S., exhibited, and made remarks upon, some abnormalities in the Carnivora. Mr. L. T. Hoesen, F.Z.S., and Mr. E. A. Spaut, exhibited, and made remarks upon, the effect of Pineal Gland administra- tion on the pigment-cells of Frog Tadpoles. My. R. I. Pocock, F.R.S., F.Z.S., communicated his paper on “The Auditory Bulla and other Cranial Characters in the Mus- telide (Martens, Badgers, etc.).” Tn the absence of the Author, Mr. G. 8S. THapar’s paper “ On the Venous System of the Lizard, Varanus bengalensis Daud.,” was taken as read. The next Meeting of the Society for Scientific Business will be held on Tuesday, May 24th, at 5.30 p.m., when the following communications will be made :— The SECRETARY. Report on the Additions to the Society’s Menagerie during the month of April, 1921. Caas. F. M.D., F.Z.8., Anatomist to the Society. “The Comparative Anatomy of the Tongues of the Mam- malia.—IV. Families 3 and 4. Cebidee and Hapalide. R. Broom, F.R.S., C.M.Z.8. On some new Genera and Species of Anomodont Reptiles from the Karroo Beds of South Africa. Rel, Pocock, HRIS:, EIZS: On the External Characters of some Species of Lutrine (Otters). C. W. Anprews, D.Sc., F.R.S., F.Z.S. Note on the Skull of Dinotheriwm gigantewm in the British Museum. j 27 The following Papers have been received :— Dr. C. F. Sonntac, F.Z.S. The Comparative Anatomy of the Koala (Phascolarctos cinereus) and the Vulpine Phalanger (77richosurus vulpeculu). R. I. Pocock, F.R.S., F.Z.S. The External Characters of the Koala (Phascolarcios) and some related Marsupials. The Publication Committee desire to call the attention of those who propose to offer Papers to the Society, to the great increase in the cost of paper and printing. This will render it necessary for the present that papers should be condensed and be limited so far as possible to the description of new results. Communications intended for the Scientific Meetings should be addressed to P. CHALMERS MITCHELL, Secretary. ZooLoGicaL Society oF Lonbon, ReEcENtT’s Park, Lonpon, N.W.8. May \7th, 1921. No. 218. ABSTRACT OF THE PROCEEDINGS OF THE ZOOLOGICAL SOCIETY OF LONDON. May 24th, 1921. Prof. E. W. MacBrips, D.Sc., LL.D., F.R.S., Vice-President, in the Chair. The Secretary read a Report on the Additions to the Society’s Menagerie during the month of April, 1921. Miss L. E. Curssman, F.E.S., exhibited, and made remarks upon, a series of lantern-slides illustrating the habits in the larval stage of the Hydropsychide (Snare-building Caddis Worms). Major Sranuzy 8S. Frower, F.Z.S., exhibited examples of three species of Egyptian Jerboas—Jaculus jaculus, Jaculus orientalis, and Scarturus tetradactylus, the latter being the first living specimens ever brought to England; and made remarks upon the distribution of the three species in North Afriea. Dr. C. W. Anprews, F.R.S., F.Z.S., gave a réswmé of his paper ‘‘ On the Skull of Dinotherium gigantewm in the British Museum.” * This Abstract is published by the Society at its offices, Zoological Gardens, Regent's Park, N.W., on the Tuesday following the date of Meeting to which it refers. It will be issued, along with the ‘ Proceedings,’ free of extra charge, to all Fellows who subscribe to the Publications ; but it may be obtained on the day of publication at the price of Sixpence, or, if desired, sent post-free for the sum of Sex Shillings per annum, payable in advance. 30 Dr. Cuas. F. Sonntac, F.Z.8., gave a réswmé of his papers (1) ‘Lhe Comparative Anatomy of the Tongues of the Mam- malia, Families 3 and 4, Cebide and Hapalide,” and (2) ‘‘ Some Points in the Anatomy of the Tongues of the Lemureidea.” Tn the absence of the Author, Prof. R. Broom, F.R.S.,C.M.Z.8., his paper, ‘‘On some new Genera and Species of Anomodont Reptiles from the Karroo Beds of South Africa,” was taken as read. Mr. R. I. Pocock, F.R.S., F.Z.8., gave a résumé of his paper ‘““On the External Characters of some Species of Lutrine (Otters).” The next Meeting of the Society for Scientific Business will be held on Tuesday, June 7th, at 5.30 p.m., when the following communications will be made :— Dr. Frank M. CHAPMAN. The Distribution of Bird-life in the Urubamba Valley of Peru. (Illustrated by lantern-slides. ) S. Mauuik, F.Z.S. New Indian Drilid Beetles. IP gots do 126 deliei. Iolaus, Wale As Exhibition of some Marsupial Embryos, especially the Koala and the Wombat. R. L. Pocock, F.R.S., F.Z.8. The External Characters of the Koala (Phascolarctos) and some related Marsupials. Dr. C. F. Sonntaa, F.Z.S. The Comparative Anatomy of the Koala (Phascolarctos cinereus) and the Vulpine Phalanger (Zrichosurus vulpecula). 31 The Publication Committee desire to call the attention of those who propose to offer Papers to the Society, to the great increase in the cost of paper and printing. This will render it necessary for the present that papers should be condensed and be limited so far as possible to the description of new results. Communications intended for the Scientific Meetings should be addressed to P. CHALMERS MITCHELL, Secretary. ZooLoGicaL Society oF LonDoN, ReceEnt’s Park, Lonpon, N.W. 8. May 31st, 1921. tee aN i aie aeons mil No. 219. ABSTRACT OF THE PROCEEDINGS OF THE ZOOLOGICAL SOCIETY OF LONDON.* June 7th, 1921 Prof. J. P. Hin, F.R.S., Vice-President, in the Chair. Major Stantry S. Frower, O.B.E., '.Z.8., exhibited, and made remarks upon, some living specimens of Testudo leithit and an example of Testudo ibera. Dr. P. CHatmers Mircueto, C.B.E., F.R.S., exhibited, and made remarks upon, a photograph of the Death-mask of a young Gorilla, presented to the Society by Prof. Howard Macgregor, of Columbia University, New York. Dr. Frank M.CHApPMAN gave an account of the distribution of Bird-life in the Urubamba Valley of Peru, and illustrated his remarks with a remarkable series of lantern-slides. In the absence of the Author, Mr. S8. Mauvtix’s paper on “ New Indian Drilid Beetles ” was taken as read. * This Abstract is published by the Society at its offices, Zoological Gardens, Regent's Park, N.W., on the Tuesday following the date of Meeting to which it refers. It will be issued, along with the ‘ Proceedings,’ free of extra charge, to all Fellows who subscribe to the Publications; but it may be obtained on the day of publication at the price of Sixpence, or, if desired, sent post-free for the sum of Six Shillings per annum, payable in advance. 34 Prof. J. P. Hiut, F.RS., V.P.Z.S., exhibited, and made remarks upon, a series of lantern-slides of some Marsupial Embryos, dealing especially with the Koala (Phascolarctos) and the Wombat (Phascolomys). Mr. R. I. Pocock, F.R.S., F.Z.S., gave a résumé of his paper on ‘The External Characters of the Koala (Phascolarctos) and some related Marsupials.” Owing to the lateness of the hour, Dr. C. F. Sonntae’s paper on “The Comparative Anatomy of the Koala (Phascolarctos) and the Vulpine Phalanger (Zrichosurus vulpecula) ” and Mr. C. Tarr Reean’s paper on “The Cichlid Fishes of Lake Nyassa” were taken as read. The next Meeting of the Society for Scientific Business will be held on Tuesday, October 18th, 1921, at 5.30 P.M. A notice stating the Agenda for the Meeting will be cireulated early in October. The following Paper has been received :— R. I. Pocock, F.R.S., F.Z.8. The External Characters and Classification of the Mustelide. Communications intended for the Scientific Meetings should be addressed to P. CHALMERS MITCHELL, Secretary. ZooLoGiIcAL Sociery oF Lonpon, -Recent’s Park, Lonpon, N.W. 8. June 14th, 1921. 23. 24, 25. 26. 27. 28. 29. 30. dl. 32. 33. 34, PAPERS. The Geographical Distribution of Orthopterous Insects in the Caucasus and in Meuiera Asia. by bP. UYAROV,; 2 CMlnp.); oe de cctinie viele o's ess Had ce bites re’s The Auditory Bulla and other Cranial Characters in the Mustelide. By R. I. Pococs, E-.R.S., F.Z.8. (Text-figures 14-18.) ....cecececsseeseoves taiahe) ct otatel sichniet ole. aie On the Venous System of the Lizard Varanus bengalensis (Daud.). By Gopinp Sincu amma, Vise, (Rext-fpuras lA.) sl es oe cs hace 6 eicciesie sa bin's/s\e sie) Ore siete selena s Abnormalities in the Common Frog (Rana temporaria). By J. H. Luoyp, M.Sc., HZ.8.- (Lext-figure 1.):. 5/0... 05. tet eteceare BStcigy oGc osname Ulaos cobs an Comparative Anatomy of the Tongues of the Mammalia.—IV. Families 3 and 4. Cebide and Hapalide. By C. F. Sonnraa, M.D., F.Z.8S., Anatomist to the Society. (Bext=fimuresia i Oas)\ ee iictele sisie's ceisjoisisiee ja) oh ew cies's cles ees ORO staresclevone Stars ya tecs - Note on the Skull of Dinotherium gigantewm in the British Museum, By OC. W. Anprews, D.Sc., F.R.S., F.Z.S. (Text-figures 1-4) .......... cece et ee eee eeee On the External Characters of some Species of Lutrinz (Otters). By R. I. Pococs, F.R.S., F.Z.S. (Text-figures 19-21.) .....0.. ees. eee eee Sal oi wi Sraptote ce cummhereeterale ae 5 The Oomparative Anatomy of the Koala (Phascolarctos cinereus) and Vulpine Phalanger (Trichosurus vulpecula). By Ouas. F. Sonntag, M.D.,F.Z.8. (Plates V., Vik; Ae ee ENGI) Uke ols manasa elena Cisvels ister stone) a elmtete/n ie eter nfal srosees socstorscere New Indian Drilid Beetles. By S. Mauuix, F.Z.S. (Text-figures 1-5.).........2- ‘ On some Abnormalities in the Carnivora. By Cuas. F. Sonntac, M. D.; E.Z.8. (Text-figures 64-65. ) eeeteesese seve Ce @eeereece OF ereoe 8 ons The External Characters of the Koala (Phascolarctos). By R. I. Pococx, F.R.S., EZS. (Text-figures 22-26.) ee seee @eonvere eereseee ee ee ee oe @oertes et eeerseeoe eo The Nestling Feathers of the Mallard, with Observations on the Composition, Origin, and History of Feathers. By J. Cossar Ewart, M.D., F.R.S., F.Z.S. (Plates I.-X. ; Text-figures 1-14.) eee rvee sees Sotelo erelscaansieveyaiose, err eeseeresesesse es sears 000080 Page 447 473 487 493 497 535- 547 579 587 - 609 LIST OF PLATES. 1921, Parr III. (pp. 447-646). Page Cuas. F. Sonntag: Pls. V.-VI. Comparative Anatomy of the Koala (Phascolarctos) 0. /'s ses sie cae es) MOLT J. Cossan Ewart: Pls. I.-X. | Nestling Feathers of the Mallard .. 609 NOTICE. The ‘ Proceedings’ for the year are issued in four parts, paged consecutively, s0 that the complete reference is now P. Z. 8.1921, p.... The Distribution is usually as follows :— : Part I, issued in March. Sey aed an ays June. ob. i September. AREA IATA December, ‘ Proceedings,’ 1921, Part II. (pp. 187-446), was published on July 8th, 1921. The Abstracts of the ‘ Proceedings,’ Nos. 217-219, are contained in this Part. The dates of Publication of ‘ Proceedings’ 1830-1858 will be found in the ‘Proceedings’ for 1893, page 436. The dates of Publication of ‘'T ransactions,’ 1833-1869, will be found in the -« Proceedings’ for 1913, page 814. PROCEEDINGS OF THE GENERAL MEETINGS FOR SCIENTIFIC BUSINESS OF AEE = ZOOLOGICAL SOCIETY ~ OF LONDON, 1921. PART. LV. conrainine Paces 647 To 887, with 6 PiatTEs aND 113 Texr-riGuREs. ey lal Uist; easel Slit Y & JANUARY 192 Ns, 4tlong) Muse PRINTED FOR THE SOCIETY, SOLD AT ITS HOUSE IN REGENT’S PARK. MESSRS. LONGMANS, GREHN, AND CoO., PATERNOSTER ROW. [Price Twelve Shillings. ] LIST. OF CONEEANES: 1921, Part IV. (pp. 647-887). EXHIBITIONS AND NOTICES. The Szcrerary. Report on Additions to the Society’s Menagerie during the months of May, June, July, August, and September, 1921......-+.. 1.0... eee ee eee eee ees Mr. M. A. C. Hiyron. Exhibition of an embryo African Elephant ..-.......-...---- Mr. E. G. Boutrncrr, F.Z.S. Exhibition of photographs of a Common Grass Snake swallLowimlertey Viletiere rele seals tetris ie PNG COM OE BE OA GOOCH 0020 AC Mr. F. Marrin Duncan, F.Z.S. Exhibition of cinematograph films of Marine Animals .. Mr. D, Serm-Smiru, F.Z.S. Exhibition of photographs of the Sacred Ibis nesting in the Society’s Gardens ................ Prof, G. Exzior Smiru, F.R.S., F.Z.S. An Account of the habits of Zurszus, illustrated by Tantern-slides 2220 ics7se) 0G hare, vee wl eee tie ei eae eet ae ee oer eee My. Epear P. Cuancn, M.A., M.B.O.U. Exhibition of cinematograph films illustrating the laying-habits of the Cuckoo (Cuculus canorus) and the life of the young Cuckoo , ‘The Secretary. Keport on Additions to the Society's Menagerie during the month of Oleoloere MeN Oss boso scat so aneobauuer EHide UM SON Akaa Ta aISCE RO OAS Ao Bu ac Mr. A. &. La Sovir, C.M.Z.S, Exhibition of photographs Cae the early Ilfe- LAVORO Ol IAU ON (OOM rolesn Gh OA ald op o wan oO FobGdG haan noah oon cat ones Page 886 886 886 886 886 886 887 Dr. A. Smiri Woopwanrp, F.R.S., V.P.Z.S. Exhibition of a Human Skull and other — TeMaims tromebrolcony Ell sNorbidhe hesitant eieiceiae eters oie ena ea Contents continued on page 3 of Wrapper. eT Pa ‘ ) ZOOLOGICAL SOCIETY OF LONDON. Tars Society was founded in 1826 by Sir Sramrorp Rarruzs, Mr. J. Sasinz, Mr. N. A. Vigors, and other eminent Naturalists, for the advancement of Zoology and Animal Physiology, and for the introduction of new and curious subjects of the Animal Kingdom, and was incorporated by Royal Charter in 1829. Patron. HIS MAJESTY THE KING. COUNCIL. yo sti HIS GRACH THE DUKE OF BEDFORD, K.G., F.B.S., President. t OO D2 AuFrep H. Cocks, Hse., M.A. Cot. diy Hiner 1 ¥enk AifON, Ceoncns Drummonn, Ksa., E. ree ae i, A ee yey Treasurer. SAFO AF OR| Neo) Vice-President: AurFrep Ezra, Ksa., O.B.E. 2 2 P. CHatmers Mrrcuetr, C2B.n, MEAy DSc. E.RAS., Secretary. Tuer Hart oF Onstow, O.B.E. Ksq., LL.D., THe Rient Hon. rae Viscount Grey, K.G., P.C. Sir Sipnny F. Harmer, K.B.E., M.A., Sc.D., F.RS., Vice- President. Pror. Jamuus P. Hirt, D.Sc., F.R.S., Vice-President. Witiiam Huntsman, Ese. Mason tHE Lorp AtAsSTAIR Rosset Innes-Ker, D.S.O. Pror.ErnestW. MacBrips,D.Sc., LL.D., F.R.S., Vice-President. Masor AtBert Pam, O.B.E. Taz Lorp Quennporoven. H.G. Tae Doxezor Rurtanp, K.G. Masor RicHarp S. Taytor. A. Truvor-Bartys, Hse., M.A. Awnrnony H. Winertetp, Ese., Vice-President. A. Smrra Woopwarp, Esa., LL.D., F.B.S., Vice- President. 2 The Society consists of Fellows, and Honorary, Foreign, and Corresponding Members, elected according to the By-Laws. It garries out the objects of its foundation by means of its collection of living animals, by its Library, and by its Scientific Publications. The Office of the Society, Regent’s Park, N.W.8, where all com- munications should be sent, addressed to “The Secretary,” is open from Ten till Five, except on Saturdays, when it closes at OnE P.M. The Library, under the superintendence of Mr. F. Martin Duncan, F.Z.S., F.R.M.S. is open daily (except Sunday) from Ten a.m. till Five p.m.; on Saturdays, Ten a.m. till One p.m. The Library is closed from Good Friday to Easter Monday, and upon all other Bank Holidays. It is also closed annually for cleaning purposes during the whole month of September. The Meetings of the Society for General Business are held in the Meeting Room at the Society’s Office on the third Wednesday of the month at 4.30 p.m. except in September and October. The Meetings for Scientific Business are held in the Meeting Room at the Society’s Office-fortnightly on Tuesdays, except in July, August, September, and December and January, at halfpast Five o’clock pP.n. The Anniversary Meeting is held on the 29th of April, or the nearest convenient day, at Four p.m.. The Society’s Gardens are open daily from Nine o’clock until Sunset. Mr. R. I. Pocock, F.R.S., F.L.S., is the resident Super- intendent and Curator of Mammals, Mr. D. Seth-Smith is Curator of Birds and Inspector of Works, Mr: E. G. Boulenger is Curator of Reptiles, Miss L. K.Cheesman, F.E.S., is Curator of Insects. Appli- cations for anatomical material or facilities for work in the Prosectorium should be addressed to the Director of the Society’s Prosectorium. TERMS FOR THE ADMISSION OF FELLOWS. Frriows pay an Admission Fee of £5, and an Annual Contri- bution of £3, due on the 1st of January, and payable in advance, or a Composition of £45 in lieu thereof; the whole payment, including the Admission Fee, being £50. No person can become a Frriow until the Admission Fee and first Annual Subscription have been paid, or the annual payments have been compounded for. Frttows elected in November and December are not liable for the Subscription for the vear in which they are elected. PRIVILEGES OF FELLOWS. Frtrows have Personal Admission to the Gardens upon signing their names in the book at the entrance gate, and may introduce Two Companions daily. The Wire or Hussanp of a FELLow can exercise these privileges in the absence of the Fellow. Until further notice, Fennows will receive 40 undated Green Cards, available on any Sunday or week-day up to the end of February of the year following the year of issue, and 20 White * Cards available on any week-day up to the same date. Twenty of the Green Cards may be exchanged for a book containing two Orders for each Sunday in the year. Twenty White Cards may be- exchanged for a book of dated Week-day Orders, each Order available for any day during the week except Sunday. Special children’s tickets are no longer issued, but the Green and White Cards are perforated, and each half is valid for a Child under twelve years of age. It is particularly requested that Fellows will sign every ticket before it goes out of their possession. Unsigned tickets are not valid. Frttows are not allowed to pass in friends on their written order or on presentation of their visiting cards. Frttows have the privilege of receiving the Society’s ordinary Publications issued during the year upon payment of the additional Subscription of One Guinea. This Subscription is due upon the 1st of January, and must be paid before the day of the Anniversary Meeting, after which the privilege lapses. Fetiows are likewise entitled to purchase these Publications at 25 per cent. less than the price charged to the public. A further reduction of 25 per cent. is also made upon all purchases of Publications issued prior to 1881, if above the value of Five Pounds. Frttows also have the privilege of subscribing to the Annual Volume of ‘The Zoological Record,’ which gives a list of the Works and Publications relating to Zoology in each year, for the sum of One Pound Ten Shillings. Separate divisions of volumes 39 onwards can also be supplied. Full particulars of these publications can be had on application to the Secretary. Fettows may obtain a TransreraBLe Lyory Ticker admitting two persons, available throughout the whole period of Fellowship, on payment of Ten Pounds in one sum. A second similar ticket may be obtained on payment of a further sum of Twenty Pounds. 4 Any Frtiow who intends to be absent from the United Kingdom during the space of at least one year, may, upon giving to the Secretary notice in writing, have his or her name placed upon the “dormant list,” and will then be called upon to pay an annual subscription of £1 only during such absence, or in the event of returning to the United Kingdom before June 30th in any year to pay the balance of the ordinary subscription. After three years a Dormant Fellow must make a further application to be retained on that list. Any Fetiow, having paid all fees due to the Society, is at liberty to withdraw his or her name upon giving notice in writing to the Secretary. Ladies or Gentlemen wishing to become Fellows of the Society are requested to communicate with “The Secretary.” P. CHALMERS MITCHELL, Secretary. Regent’s Park, London, N.W. 8. January, 1922. MEETINGS OF THE ZOOLOGICAL SOCIETY OF LONDON FOR SCIENTIFIC BUSINESS. —$—$$_ 1922, TUESDAY, FEBRUARY ...... 7 and 21. * IVEAS CH uae vey ee 7 and 21, e SAND ese ee 4 and 25. x MAvig 3, gue yen. 9 and 23. ShUNES pute eeee- He 13 The Chair will be taken at half-past Five o'clock precisely, ZAVOLOGICAL SOCIETY OF LONDON. LIST OF PUBLICATIONS. Tae scientific publications of the Zoological Society of London are of two kinds—“ Proceedings,” published in an octavo form, and ‘“ Transactions,” in quarto. According to the present arrangements, the “‘ Proceedings” contain not only notices of all business transacted at the scien- tific meetings, but also all the papers read at such meetings and recommended to be published in the “ Proceedings ” by the Committee of Publication. A large number of coloured plates and engravings are issued in the volumes of the “ Proceedings,” to illustrate the new or otherwise remark- able species of animals described therein. Amongst such illustrations, figures of the new or rare species acquired in a living state for the Society’s Gardens are often given. The “ Proceedings” for each year are issued in four parts, paged consecutively, during the months of March, June, September, and December. From January 1901 they have been issued as two half-yearly volumes, indexed separately. An “ Abstract of the Proceedings” is published by the Society on the Tuesday following the date of the Scientific Meeting to whichit refers. It is issued along with the “ Pro- ceedings,” free of extra charge, to all Fellows who subscribe to the Publications, but it may be obtained on the day of publi- cation at the price of Sixpence, or, if desired, sent post free for the sum of Six Shillings per annum, payable in advance. The “‘ Transactions” contain such of the communications made to the Scientific Meetings of the Society as, on account of the nature of the plates required to illustrate them, are better adapted for publication in the quarto form. They are issued at irregular intervals. Fellows and Corresponding Members, upon payment of a Subscription of One Guinea before the day of the Anni- versary Meeting, are entitled to receive the Society’s Publications for the year. They are likewise entitled to purchase the Publications of the Society at 25 per cent. less than the price charged to the Public. A further reduction of 25 per cent. is made upon purchases of Publications issued prior to 1881, if they exceed the value of Five Pounds. Fellows also have the privilege of subscribing to the Zoological Record for a sum of ‘wo Pounds Ten Shillings (which includes cost of delivery), payable on the Ist of July in each year; but this privilege is forfeited unless the subscription be paid before the 1st of December following. The following is a complete list of the publications of the Society already issued. TRANSACTIONS OF THE ZOOLOGICA SOCIETY OF LONDON. 4to. 20 vols. and Index. Price to Price to the Fellows. Public. Vols. I-IV. (out of print). Vol. V., containing 67 Plates... (1862-66) .... 5 4 3... 6190 NGI 5 92 eas (USE) nse. I © pon 1B O © ae VIS F 73 i) fete ULSCO=72) 5 VRE OME Orr amlch alta) on dll log i 82 ere Wiese) en US). ey sons 12 IN © Se rs! De os 99 con (UMeYENTO! -oss IAS IL GB. , 16 2 0 a Xe ss 95 Hoe (Cis o oan 102 O. 8, iy 7 Indéx,¢ViOls Ml SNe mes pate eee (@is33=79)) ren Ole Cue 1G © Vol. XI., containing 97 Plates .. (1880-85) .... 912 0. 12 16 0 Pen. lise - 65 oo (Che) so5, & & © ¢ 4 © Pll, 3 62 p oo (EMIS) 5.5. 6 8 SB, 811 0 ere AVS 47 my oo GUS) acon © & ©; EoD © to: Ge - 52 py os (ESO) 4.5 S16 6 . 714 0 be SOROVATES make py ee (UIOISIGOD) .. & 8 ©. (ol sy VD Pate all 42 CSeSI05)" 5, 6 1S 6. “ies © ay SION ES oy ds yy oo (Of 5, 4b 1) ©. s oF 80 Bx eNeLEXG. Bema: » oo (CSCSIO) ., 10 4.0. ee ee) pho tGENON shi) Oe cee CLAIM) oN 9B oak ls 7 Oo » XXIJ.—Part 1. (7 Plates & 12 Text-figs.) (mm: WONG) aos cccs0cnce 2 6 i=) oo i=) i=) : now abie to offer for sale, at the reduced price of £30, sets of Vols. V._XVI. inclusive, and In consequence of a re-arrangement of the stock of the ‘Transactions,’ the Society ‘| Heeparate papers, of which a list can be supplied, at about one-fourth their published price. PROCEEDINGS OF THE COMMITTEE OF SCIENCE AND CORRESPONDENCE OF T™E ZOOLOGICAL SOCIETY OF LONDON. 8vo. 2 vols. (Letterpress only). Payt I. 1880-31. 1 vol. 8vo., out of print. ee Lielisa2: 13. 1p easier ee AS Odi tec) NOS. Price to Price to the Fellows. Publie. PROCEEDINGS OF THE ZOOLOGICAL SOCIETY OF LONDON. First SERIES. Parts I—XV. (1833-1847). 8vo. 15 vols. (Letterpress only.) Price to Fellows : 4s. 6d. each part ; to the Public, 6s. Index 1830-1847. Price to Fellows: 4s. 6d.: to the Public, 6s. Parts I., VII.—-IX., XI., XIV., XV., out of print. SECOND SERIES. Parts XVI.-XXVIII. (1848-1860). 8vo. 13 vols. (Letterpress only.) Price to Fellows: 4s. 6d. each part ; to the Public, 6s. Index 1848-1860. Price to Feliows: 4s. 6d.; to the Public, 6s. The Parts of this series containing Coloured Plates are out of print. PROCEEDINGS OF THE SCIENTIFIC MEETINGS OF THE ZOOLOGICAL SOCIETY OF LONDON. 8vo. 40 vols. and 4 Indices. Letterpress only. With Plates uncoloured. With Plates coloured. Price to Price to the Price to Price to the Price to Price to the Fellows. Public. Fellows. Public. Fellows. Public. SOW Ass Cds ys GSk Gast ae 9s. al oie meee Gam, Sh choo CNH MSGQM EASE GUS. 08s (G8. ak a ha! 9s. Pano Sane ote Bas, BS ssn, 4456. 1863 ASVE GU at reican OSM ad achat: Qs. verses Beaten bes. 9d 45s. 1864 AGG Sh. OSES eles 9s. Ses eR eae 30s. 9d. .... 45s. 1865 Acme Usenet a OSes Gaake 9s. ee LON een ea ks B38, Mh soce 25% 1866 dle Ges saoo Stoo coos 9s. mA Rake etait Ba BYE cca 45S TNF ea ok orchicis eE at t a Ae erpe ne 9s. 12s.*. S38, Mh casa ZS, MSG Mone oa Sine og okie canind nce aie 93 Dee ce am Boe, Be soo, 45S US GO Me ree ete eset atoucn ee Neat 9s. ee al Ke We eae ier BBS, Wh econ 4K IS OMe earch haste Bees G8 ie Ran Lar 3t 9s. Sern Oe en, Eyre! BOS, 9A. seen 40s: Inidexe 1SGIZUS(O, wns onl is ith sace 6s. BES lta ese says staan sy Ae eanaay tise ee Qs. Ete e nD eure cn) la Bee Beh nono 4s, TSU WA ha tic boas midis on a deeeteey Sia 9s See ilia sate eee ag SB Bb booo CT MSO AR Crone Ra PLL a 9s Doha eet Sak 8 sooo 4h58 TE eer bale ea cP INTEL CORE PA tc 9s ml arti ter aes 36s A8s.+ HLS fed apemcgee Deeside a oven enor omen eave eae 9s Pe LE See er ee 36s 48s ISH A Ce Rta eRe ee ineser cs cucic one 9s Det aie te vee 36s 48s.t ICY een PRC eMC CRT RE TT Ono eo: ae Or eO 9s TDG ser ayn iobene 36s 48s MSSM eis lcc eae certiel crauae nant uewets 9s Ig Pare aes Bae = 36s 48s 1S Sain Giemcesicea woes big .at'o c Dae pate ci, L8H aie tres 36s. soos, Ge TESYS{0) Sa i eR Pn hrc Aerts fc 3 9s. eee a ID pve 0 Sere 36s, ajson Chery, lindexate/ 11880" cnr cma Ace Odean an 6s. WLAMSNSOO, Wer syalenansys coco BRON coggsocancnacned 36s. soon Zeeks. Index, 1881-1890 ............ Aste OG thas 6s. TROISNOOO, Wan iolnes cooo IBEGN coscconccuusucce 36s. fous atch: Index, 1891-1900 ............ Ass Gde pire O8: * No perfect copies in stoek. Tt Out of print. PROCEEDINGS or tot GENERAL MEETINGS ror SCIENTIFIC BUSINESS or tur ZOOLOGICAL SOCIETY OF LONDON. - Svo. 32 vols. and Index. Price to Price to the Fellows. Public. TOM TOU, rena WOMMIMES, ance cb ocse cc ae o3e Hach 18s. .. 245 Indes; GOL T9LOT ee roe a ys 2 se scl ciel to ace ea As Oi er, 4s 408 MOWD vols Te ese ashi taas aha Sieh ene ea ah ere eee SS pees DAs. Pt blame eter csc a Beene en bi © laa tei each aa eee NSS ints oetees 24s, J Nar Millet apearnan cet ie oar ualonthiad Ginlons 43.0 6:6 n/5°4 2 JUSOR eau ie. 24s ; Pere Cgeae renee A Umeha oO dist bold cenicg\e'ojo.0.4 WSS. ei 24s OT igen Enmore en Hise MARE au Golo yg Gw'Uhied ooo 'UNG> 6 3 TSS ud 24s et Umer aoe oan oor ep eR ae Caub t Se ean ine 24s nO ee as ere nee se Me ae aR Dias cits cian oun ainda pS op 'eho'o oc Sse Rear 24s, Mis ae Nas cuckeystay we yehSy Aahene hoist er Re ec ao eee Rae TS 8.pccgeeae 24s DS) US Macatee We Let ei ak a Pee Nearer arts Rhea Aco n6.0 ousd ots 18s. 245 Peet WP A a then URC Re Mei mR Mito Pia hin pute we pa cl hcict 9 oo 18s. 24s TRS IG heat pencal WMeouanen Nt me a damce Pee Ouran nua SER ced 0 feort ry GNo Giallyo/a aco AChR tee 3 248 cic Hal Cl Gea Nie oh tke aeeeeetinaceo t,o eMamobi Gn a G:triptov e900 Sk eiercakosine 24s INCRE ator eh er uranetbnen sree meena iaes unre coi eal oo sto"o bit 6.8 0.03 DHSS PMRS cn ihipee 24s $hort (| Pa a Emre eco ate Coker tbs tacd'a WA c0"'n We ea: 9s. 12s TUG) eh saat fad I rea ieereerain rime she eieeeniens Ho lncas Llano ths ogo Ost e ee 12s 3 Fie A ene fone tat Serr sted Ree OV or cy BGi0-Gra! Sip (oa h Git OB es ase 12s GUUS Menge Seiten Opener tM mene Pini eR GSA icha caus odie te Soc POs er 20S aa el Cee ome Gemma h mina ming hen cahiom lotta! 4 ole Lbs sae 20s DOT GF GAME etcetera ncchtauites wiser oc ori teets Mbs SMEAR RY LU ot oR er OSes sca ee 20s Bi 4 LA ee anne APR eee eC IW 28. a ialhe city 6 RSs Pete ae 24s. US ANTS ea NG Liters a anc came cA ome Eee a Gad, Va ots 5g G43 TS: Venere 24s, se uate bl Ui tre tanta meee at ancey Mera het tL Oo Wmnrs sect ch Borah si eeees 24s LISTS OF THE ANIMALS IN THE SOCIETY'S GARDENS. List of the Vertebrated Animals now or lately Living in the Gardens of the Zoological Society of London. (Highth Kdition.) 8vo. 1883. Cloth, 4s. 6d. List of the Vertebrated Animals now or lately Living in the Gardens of the Zoological Society of London. (Ninth Edition.) 8vo. 1896. Cloth, 6s. CATALOGUE OF THE LIBRARY of the Zoological Society of London (Fifth Mdition.) 8vo. 1902. Cloth, 6s. THE OFFICIAL ILLUSTRATED GARDEN GUIDE—19th Kdition (Revised)—with (1) a Railway and Street Map; (2) a Plan of the Grounds; (3) a short description of some of the principa animals in the Coliection, with 52 Photographic Illustrations and Index. Price ls. in Stiff Paper Cover, postage 2d. THE HOUSE-FLY CAMPAIGN. 3rd Edn. Illustrated. 68vo. 1916. 3d. PRACTICAL ADVICE ON THE FLY QUESTION. 8yvo. 1915. Dae ZOOLOGICAL RECORD.—Vol. 56, containing literature relating chiefly to the year 1919, was published in November 1921. Vol. 57, for the year 1920, is being prepared as usual. P. CHALMERS MITCHELL, Secretary. Regent’s Park, London, N.W. 8. January, 1922. These publications may be obtained at the Socrery’s Orricu or through any bookseller. ON ANOMODONT REPTILES FROM SOUTH AFRICA. 647 PAPERS. 30. On some new Genera and Species of Anomodont Reptiles from the Karroo Beds of South Africa. By R. Broom, FRS., CM-Z:8: [Received April 6, 1921: Read May 24, 1921. Jy (Text-figures 28-45.) INDEX. Page Page Dicynodon sollasi, sp. V.... ...... +45 648 | Eosimops newtoni, gen. et sp.n.... 663 7 andrewst, Sp. 0.......... 650 | Palemydops platysoma, gen. et a OSUOTNI) SP. Ne) Sine oe 6al | SPELT Cees ace ed Nur es ea 665 3 watsoni, sp.n. ......... 653 | Hmydopsis tr Ge ise Broo) us CLURGUS Se Neier 654 gen. n. as 666 py woodwardi, sp.N. ...... 655 = longus, sp. bans danasaene 667 A ictinops, SP. 0. ......... 656 | a Sciuroides, sp. n....... 668 macrorhynchus, sp.n.. 657 Emydops parvus, sp... ....... 669 Bainia tigriceps (Owen), gen.n.... 659 EHmyduranus platyops, gen. et sp.1 n. 670 » peavott,sp.n. ............... 659 | Hndothiodon crassus, sp. N.......... 672 LAUGH TONDASPS Heese ene OGL: The genus Dicynodon was established by Owen in 1845 for the remarkable South African fossil reptiles discovered by Bain and referred to by him as Bidentals. As the result of the labours of Owen, supplemented by that of later workers, nearly every detail in the structure of the skeleton of Dicynodon is now as fully known as that of the Dog. It was a primitive reptile with a skeleton built essentially on the same plan as that of the mammal, but with certain remarkable specializations of the skull, especially in the complete loss of incisor teeth, the presence of a large permanently growing canine in the male, and in the fact of the front of the upper and lower jaws having formed a horny beak as in the tortoise or bird. While there remains little to add to our knowledge of the skeleton, there is still much to be done ere we have “anything like a satisfactory knowledge of the large number of species which must be referred to the genus. Dicynodon first occurs in South Africa at the base of the Parevasaurus zone, which may be regarded as Middle Permian in age, and it extends through all the zones to the top of the Permian, and forms which are very closely allied, if not indeed referable to Dicynodon, are met with to thé top of the Trias. A form closely allied to Dicynodon oceurs in the Dwina of Russia, and Placerias of the Trias of North America is, if not Dicynodon also, at least very closely allied. Oudenodon, originally described as a distinct genus differing from Dicynodon in the absence of the tusk, we now know to be the female. So long as all the known species of Oudenodon differed from the known species of Dicynodon in other characters besides the absence of the tusk, it was felt wiser to keep Oudenodon Proc. Zoot. Soc.—1921, No. XLIV. A4 648 MR. R. BROOM ON ANOMODONT REPTILES as a distinct genus. In 1909, in summing up the position at that time, I stated that the evidence seemed ‘to be getting stronger in favour of Oudenodon being but the female Dicynodon.” By 1911 the discovery of tusked Tanid tuskless specimens of Dicynodon bolorhinus in one locality removed all doubt, and since then we have discovered tusked and tuskless specimens of quite a number of species. We do not yet know whether the females of all species are tuskless, and there is considerable reason to believe they are not. And it is also possible that even the males of some species at present referred to Dicynodon may be tuskless. The genus Lystrosaurus is certainly tusked in both sexes, and as all known specimens of Cistecephalus ave without tusks, presumably both sexes are tuskless. All the known specimens—about a dozen—of Kannemeyeria are tusked. Also every known specimen of Dicynodon tigriceps or allied species 1s tusked, while the few known specimens referred to Hocyclops are tuskless. Species where both sexes are tusked or both tuskless will bave to be placed in distinct genera or at least subgenera. When Owen worked at Dicynodon and Oudenodon he made nearly every skull the type of a distinct species, and there were doubtless many who considered that he was over liberal in the making of new species. Lydekker in 1890 reduced Owen’s 20 species to 11 good and 3 doubtful species, but there is now little doubt, with our much greater knowledge of the genus, that practically all Owen’s types are good. In my opinion, out of the 20 specific names only two can be regarded as synonymous— D. pardiceps and D. rectidens, both of which I refer to D. leoniceps Ow. It is hardly surprising that a genus which survived for probably over 2,000,000 years should have over 50 species already known, and probably very many more to be discovered, when we consider how many species of such reptiles as Zestwdo or Lygosoma are at present alive. Until a few years ago nearly every good specimen discovered seemed to belong to a new species. Now the corner seems to be turned, and the majority of Dicynodon skulls now picked up can be veferred to species already known. In the present paper I wish to des a iis a number of new forms T have discovered in the past three years. 2 DIcYNODON SOLLAS!, sp. n. (Text-figs. 28, 29.) This new species I have much pleasure in naming after Prof. W. J. Sollas, who with his daughter has added so consider- ably to our fuller knowledge of the Puan of the Dicynodon skull*. It is founded on a series of over 20 good skulls and * In connection with the work on Dicynodon by Prof. and Miss Sollas, I should like to point out that the skull examined by them in 1913 cannot possibly be Dicynodon leoniceps‘Owen. As the specimen is now, of course, completely gone, and as the postorbitals have manifestly been partly weathered away before the specimen was ground down, it will never be possible to determine the species with absolute certainty. In my ‘opinion the specimen was Dicynodon feliceps Owen. FROM THE KARROO BEDS OF SOUTH AFRICA. 649 many imperfect ones discovered near Biesjespoort on a horizon which I believe to be at the very top of the Hndothiodon zone, and it appears to extend a short way into the Cistecephalus zone. Though many skulls, both males and females, are known, I take as the type one of the best preserved, which I believe to be an adult male. Itis about 5 inches in length and a little less than 4 inches in width. The beak is about the same length as the orbit and moderately sharp in front. The nostrils are fairly large, and behind them the maxillaries meet the nasals, and the septomaxillaries do not appear on the side of the face. Immediately behind the plane of the nostrils is a thickening of the bones which forms a round button-like low boss in the middle line. Above and behind the canine is a thickened elongated ridge. Text-figure 28. Text-figure 29. Text-fig. 28.—Skull of Dicynodon sollasi Broom. Reduced. Text-fig. 29.—Preparietal region in Dicynodon sollasi Broom. Nat. size. The interorbital region is flat and narrow. The orbits look upwards and outwards. The parietal region is only very slightly wider than the inter- — orbital, and the large postorbitals largely overlap the parietals and almost meet in the middle line. The preparietal region is not well displayed in the type skull, but is well seen in a number of others. The pineal foramen is relatively small, and is situated a few millimetres behind the plane of the back of the postorbital arch. The preparietal is small, and is flanked by fairly large anterior processes of the parietals. The postfrontals are nearly obscured by the parietals and postorbitals, but form a distinct though small part of the orbital margins. 44* 650 MR. R. BROOM ON ANOMODONT REPTILES The following are the principal measurements of the type skull :—— Greatest antero-posterior length ...... 128 mm. Width across squamosals ' ............... SB) tees Intertemporal width eres. ee eeeeeeeee DD eae. nterorbibaliiwidichy ieee ree tere eee ee Dal pees Width across maxillary bosses ......... D2 Wes Width between mostrils 24 27-ecee- «css HRS Pope oe Width between canine roots ...:........ 32 29 Dicynodon sollast bears considerable resemblance to D. feliceps, D. ictidops, and D. testudiceps. From D. feliceps it differs in being a much more delicately built skui] and with a very much feebler tusk, the tusk in J. sollasi being about half the thickness of that in D. feliceps. In D. testudiceps the postorbitals meet each other in the middle line over the parietals; in D. sollast the postorbitals do not overlap the parietals so very greatly. D. ictidops comes fiom a horizon which is probably 500 feet lower than that which yields JD. sollasi; so that there is a strong presumption in favour of the two being distinct species, though they are undoubtedly closely allied. The preparietal in D. sollasi is smaller than in JD. ictidops, and the general contours of the bosses differ considerably in the two species. DICYNODON ANDREWS, sp. n. (Text-fig. 30.) This species is founded on a nearly perfect skull from the same horizon and nearly the same spot at Biesjespoort that yielded the type of D. sollasi. In addition to the type specimen I have two others which I refer to the same species—one from a locality between Murraysburg and Biesjespoort, and the other from a locality 3 miles east of Biesjespoort. As this latter specimen was found in near association with a specimen of Cistecephalus sp., 1b gives us the horizon as the base of the Cistecephalus zone. The species is a very strikingly new type, which may ultimately have to be placed in a new subgenus. In general shape the species resembles typical Dicynodons. The snout is unusually short, the preorbital region being only about one-fifth the length of the skull. Above and behind each nostril is a well-marked bony boss, with a fairly broad flat suleus between. Hach of the two specimens which has the mawxille preserved is tuskless. The orbits are large, with supraorbital thickening on the frontal bones. The frontal region is shghtly concave behind, and in front a low median elevation leaves each frontal bone slightly concave. The parietal region is broad, and unlike the typical species of Dicynodon, the pavietals are not overlapped by the postorbitals. The large size of the parietals makes a distinct approach to the broad-headed Dicynodonts of the type of Dieynodon tigriceps Owen, and for which I am proposing a new genus. In fact, D. andrewsi may be the ancestor of D. tigriceps. FROM THE KARROO BEDS OF SOUTH AFRICA, 651 The preparietal is well developed, and the relations of the bones in this region are very similar to the typical Dicynodon arrangement. There is a large postfrontal. Text-figure 30. Parietal region of Dicynodon andrewsi Broom. The following are the principal measurements of the type :— Greatest antero-posterior length... about 260 mm. iBasalelemetine | es aa. dees see Rea KOE a8 Greatest: widthieess ja cce aes. coke ao Ours. Wmbteronloubal waddlay 2 ahs eke. aoe Aes intertemiporaliwidthapes...8 ss iynses seas sos AGH. 5. I have much pleasure in naming the species after Dr. C. W. Andrews, F.R.8., of the British Museum. DiIcyNODON OSBORNI, sp.n. (Text-fig. 31.) This very strikingly new species is from Wapatsberg, near New Bethesda Road, and is from beds which are undoubtedly in the Lystrosaurus zone. In the seme horizon are found not only Lystresaurus, but the Therocephalian recently described by me as Moschorhinus kitchingi. When Moschorhinus was described the exact locality was unknown to me, and when later I dis- covered it I made a special expedition to the spot to try and discover the back portion of the skull. Though I did not succeed in getting the rest of the type, I got another imperfect skull, two specimens of Lystrosaurus, and the type skull which I am now describing. The present skull is not only of much interest as a new type, but is of great importance as being one of the very few Dicynodon 852 _MR. R. BROOM ON ANOMODONT REPTILES skulls known from the Lystrosaurus zone, and the only known large one. The skull lacks the lower jaw, and has lost by weathering the front of the snout, the right jugal arch and part of the left, but otherwise is fairly complete. Unfortunately, the matrix is hard, and though most of the bony surface has been displayed, it is almost impossible to be sure of the sutures. The whole skull is broad and flat. When perfect it probably measured about 163 inches in length, and it must have been fairly wide. Text-figure 31. Skull of Dicynodon osborni Broom. Greatly reduced. The snout is relatively long, and apparently resembled in some respects that of Kannemeyeria, and as in that genus, there are a pair of very powerful tusks. The nasal region has been broad and rounded, without any very striking bony bosses. The frontal region is wide and moderately flat. Low but well- developed bosses are above the anterior half of each supraorbital margin. The intertemporal region is relatively narrow, and a marked median crest is formed by the flattened parietal overlapped by the large postorbitals. The preparietal region has not been satisfactorily revealed. The pineal foramen is evidently small, and situated well back between the approaching postorbitals. Whether a preparietal | FROM THE KARROO BEDS OF SOUTH AFRICA. 653 bone is present or not the evidence does not show. There is certainly no preparietal boss such as occurs in the species of Hocyclops. The orbits look almost directly upwards. The postorbital arches are long and well developed, The squamosals must have been extremely broad, judging from the portions preserved. The following are the principal measurements of the type :— Greatest antero-posterior length... about 460 mm. Greatest) wath. ere oe Ba 40 ules Distance between tusks .................0085 "SO. 5 Mnteronbitall wid tay oe ces es <5 NGOr ar. Jnitertemamearalltwrahbhy Yeates is cect ot eens 45 Basallene tii sce desde c sess about 410 ,, Ihave much pleasure in naming this new type after Prof. H. F. Osborn. Text-figure 32. SESS asses SSS eS SS y= Top of parietal region of skull of Dicynodon watsoni Broom. The degree to which the postorbitals overlap the parietals is unknown, but most probably it is not greater than is indicated in the figure. DicYNoDON WATSONI, sp. n. (Text-fig. 32.) This new species is founded on a very fragmentary skull, of which little more than the frontal and parietal regions are pre- served, found on the mountain to the east of New Bethesda, about 800 feet above the horizon of the village and probably in the Lystroswurus zone. It is a near ally of D. osborni though quite a distinct species, and important in giving details of the structure of the preparietal region. 654 MR. R. BROOM ON ANOMODONT REPTILES The frontal region is broad and slightly convex, and the parietal region is in the same plane as the general frontal plane. There are no supraorbital bosses as in JD. osborni, the whole frontal surface being unusually smooth and slightly convex. The postorbitals are large and form a considerable part of the upper cranial surface. Posteriorly they clasp the parietals, and with them form the parietal crest. Curiously enough, each post- orbital has had its posterior process broken off near the plane of the pineal foramen, probably as the result of the bite of some carnivorous enemy ; but there is little doubt that they extended backwards, though not so far as in typical Dicynodons, as the grooves for their lodgement are preserved. Probably, however, they did not quite extend to meet the squamosals. The preparietal is moderately large and wide, and forms the margin of anterior third of the pineal foramen. The postfrontal, though fairly large, is almost entirely hidden on the surface by the frontals and the postorbitals. I have much pleasure in naming the species after Mr. D. M.S. Watson. The following are the principal measurements :— DIcYNODON cuRTUS, Sp. n. (Text-fig. 33.) The type of this new species is a complete but distorted skull found by me at Biesjespoort. The horizon is probably about Text-figure 33. Preparietal region of Dicynodon curtus Broom. Wat. size. 100 feet above that in which the type of D. sollasi was obtained, but it is quite possible that D. sollasi may have survived to be contemporaneous with this other type. in any case, though this type is similar in size to D. sollasi, it is very strikingly different in nearly every detail. ewe FROM THE KARROO BEDS OF SOUTH AFRICA. 655 The skull is short, broad, and deep, and about 4 inches in length. The snout is bent down and the caniniform processes are under the orbits. There is no great thickening of the nasals. The orbital region is narrow and the parietal region moderately wide. The pineal foramen is large and rounded. The preparietal is small and narrow, and the postfrontals very largely developed. The parietals are fairly well exposed between the postorbitals, but are extremely short. The postorbitals in the temporal region are broad and overhang the temporal fosse. The nearest ally to this species, which I propose to call Dicynodon curtus, is Dicynodon dubius Owen, from which it differs in the unusual shortness of the parietals and in other proportions. The following are the principal measurements :— Greatest antero-posterior length... about 110 mm. Greatest breadth Pi Lae aN eine Ae | ao CO PasaliMlemathe sn waaeunree..cowerccocseueene tee ts LOOT ys raterogboubaltwidishite tects. ncn. Woon ec ce thcedenis 20). 55 itertemporal width 2 kee-043.4).. cates: PY alae Between the caniniform processes about 22 . Text-figure 34. o \ ‘ Preparietal and its relations in Dicynodon woodwardi Broom. Nat. size. DiIcYNODON WoopWARDI, sp.n. (Text-fig. 34.) The skull which forms the type of this new species was sent to me from near New Bethesda by Mr. C. Kitching, and I am not quite sure of its horizon. Pretty certainly it is from some part of the Cistecephalus zone. Though the whole of the base of the skull is preserved, much 656 MR. R. BROOM ON ANOMODONT REPTILES of the upper surface is unfortunately lost, including both post- orbital arches with the whole of the postorbital bones, practically the whole of the parietal region, and both squamosals. Still, what is preserved shows that it is a strikingly new type, and I have much pleasure in calling it after Dr. Smith Woodward, of the British Museum. The skull is moderately small, having measured about 6 inches in length. The snout is long and narrow, the preorbital portion of the skull being nearly as ‘long as the postorbital. The nasal region is imperfect, but there were probably no great bony thickenings. The maxille have each a round tusk, the head of which lies under the orbit, and the tusks are relatively farther back from the front of the beak than in most species. The frontals are long and moderately narrow. Behind them lies a most remarkably broad preparietal quite unlike that of any other known species. By its sides are narrow posterior processes of the frontals. The postfrontals lie outside the frontal processes. The postorbitals are lost, but from the con- dition of the posterior parts of the squamosal it is manifest that the temporal region must have been comparatively narrow. The following are the principal measurements :— xreatest antero-posterior length... about 160 mm. Greatest breadth .......... . perhaps about 110 ,, ‘Basallplemot ly Me i See eect cee it scnn i 4 145, dm berorlomucllewiachulia naasce a crear css ae SHO es Distance betweenstuslss ye sc ee ee, DO nee DicyNnopon ictinops, sp.n. (Text-fig. 35.) This new species is formed on the nearly perfect skeleton of a small animal found by me about 30 miles from Biesjespoort on the Murraysburg Road. While it is quite possible that the form is not quite pully grown, it cannot be the young of any known species. The head is about 3} inches in length, and the whole animal from snout to the tip of the tail about 12 21 inches. The snout is smooth and rounded, without any bony thickening either on the nasals or in the supraorbital regions, except a very slight thickening on the nasal just where it overhangs the nostril. The septomaxillary does not appear to show on the face. ‘Che lacrimal and prefrontals are small. In the maxille is a small tusk. The frontals are large, and a transverse section in the orbital region would show a convex upper surface, while a section of the posterior end would have a concave upper surface. The post- frontals are small, and only a small part appears on the surface. The preparietal is rather small, its anterior end being on the plane of the postorbital arch. The pineal foramen is situated far back. The parietals are small. The postorbitals are very large, FROM THE KARROO BEDS OF SOUTH AFRICA. 657 and their posterior processes are kept apart farther than in most Dicynodon species. On some future occasion the skeleton will probably be described in detail. For the most part the bones are surrounded by very hard calcareous masses, and are very difficult to clear. Text-figure 39. Top of skull of Dicynodon ictinops Broom. The following are the principal skull-measurements :— Greatest antero-posterior length ...... eos ees araucnys Greaveshmoreagiuht. .. nese thes ie S e 68 ais hater orgies leawaalitslne yey ea eee ee Tolar Initextemiporale widths. tunel ee VO S, Basa lenethia aa aan Anus, atte about SOP tts, Distance between canines ................5. NNO Nes DiIcYNODON MACRORHYNCHUS, sp. n. (Text-fig. 36.) This new species is founded on two small skulls discovered by me at New Bethesda. For a time I considered they might be very miniature specimens of Dicynodon platyceps, the common New Bethesda species, but further careful examination shows that it is necessary to regatd them as belonging to a distinct species. The best preserved specimen is taken as the type. The skull when complete was considerably less than 3 inches \ 658 MR. R. BROOM ON ANOMODONT REPTILES in length, and if an adult it represents one of the smallest known species of Dicynodon. It is chiefly remarkable for the great length of the orbital and preorbital regions as compared with the postorbital. In the nasal region there is a large median thickening re- sembling that in Dicynodon sollasi. The frontal region is narrow and the orbits large. The postorbital arch is slender and the postorbital bones are quite unusual in shape. There is no trace to be seen, at least on the upper surface, of any post- frontal, and the postorbital takes its place, having an anterior process which forms a considerable part of the supraorbital margin. ‘The posterior process is unusually short, much of the Text-figure 36. Top of skull of Dicynodon macrorhynchus Broom. temporal wall being formed by the parietal. The pineal foramen is of moderate size, and is situated as shown in the figure given. The preparietal is long and narrow. The posterior portion and much of the palate are crushed. The following are the principal measurements :— Greatest length ......... probably about 70 mm. lankerormitallwadith: 2.0. cd0 sae eee CSO. ie Iinterremporal width... -. eee eeeeeeeee 8 a From snout to front of pineal foramen... 47 us Dicynodon macrorhynchus is allied to D. sollasi, D. testudiceps, and D. ietidops, but being from the middle of the Cistecephalus zone, must be from a horizon many hundreds of feet higher than that of any of these others. seein FROM THE KARROO BEDS OF SOUTH AFRICA. 659 BArNIA TIGRICEPS (Owen), gen. n. One of the first species of Dicynodon to be described by Owen was Dicynodon tigriceps. This is represented in the British Museum by a very fine skull, said to have come from ‘ Gonzia River, Kaffraria.” In general appearance it differs very con- siderably from the typical Dicynodow as represented by D. lacer- ticeps. The skull is massive and extremely broad and flat, and the parietal region differs in that the parietals are well developed and not covered by the postorbitals. The genus Dicynodon is so large that it would be convenient if we could lop off a few species and put them into a subgenus, even if the differentiating charac- ters did not seem of very fundamental importance. But there is another character that is of great importance. J know of nine skulls which belong either to Dicynodon tigriceps or closely-allied species, and all have tusks. It is therefore very highly probable that the female of Dicynodon tigriceps was tusked, and uot like the female Dicynodon—tuskless. And if this be so, Dicynedon tigriceps must be placed in a distinct genus. It is, of course, impossible to prove that the female was tusked, and even if 50 tusked skulls were found there would still be a doubt. As it is, I think the probability sufficiently great to make for it a new genus, bainia, after the father of South African geology, Andrus Geddes Bain-—one of the most gifted geological geniuses the world has seen. Even if’a tuskless femaie should turn up, Bainia must still stand as a subgenus. BAINIA PEAVOTI, sp.n. (Text-fig. 37.) This new species is founded on a very fine skull discovered by me at Biesjespoort. The skull is well preserved and not much crushed. One zygomatic arch is lost, also part of the base and the lower jaw. Though very similar in size to Bainia tigriceps, it differs in many important features; and as it is manifestly a new species, I have much pleasure in naming it after the late Henry G. Peavot, who for some years was the Zoological Society’s Librarian and Clerk of Publications, and whose care and kindly help all workers so fully appreciated. It is a delight to be able to add a stone to the cairn of a fellow-worker and friend who laid down his life on the battlefield. The skull is of special value in that all the sutures can be made out with perfect distinctness. In general shape it agrees more closely with Bb. tigriceps than\with B. laticeps, but differs in the shape of the zygomatic arch and in many other details which will be mentioned. ‘The snout and the whole front half of the skull is much flatter than in any other known Anomodont. The premaxilla, which is nearly complete, is relatively small and unusually flat. The nasals are short but very broad, and have each a well- developed boss which overhangs the nostril. The suture with the frontals and prefrontals is nearly transverse. 660 MR. R. BROOM ON ANOMODONT REPTILES The septomaxilla is not satisfactorily preserved, but is relatively small. ~ The maxilla is a very remarkable bone, quite unlike that in any other known Dicynodont. The bone of the left side has lost the subnasal portion, but is otherwise nearly complete, and it is so remarkably shallow that one would fancy the palatal half had been removed; yet if part of the bone has been removed, it must have been done during the animal’s lifetime. There is no trace of a tusk, yet there are the remains of the socket of a large tusk. Text-figure 37. Sor re == rs Skull of Bainia peavoti Broom. A not improbable explanation of the peculiarity is that the animal is aged and has lost its tusks, and that in consequence the palatal portions of the bone have become absorbed, as is seen in aged human and other mammalian jaws. The prefrontal is a relatively small trianguar bone which has a small and not very prominent boss above the front of the orbit. The lacrimal is small. FROM THE KARROO BEDS OF SOUTH AFRICA. 661 The jugal is large. It forms most of the lower border of the orbit, most of the postorbital arch, and a considerable part of the zygomatic arch. The frontal is unusually large and wide. Its shape will be best understood from the figure given. It has apparently a large articulation with the postorbital, but a thin little strip of the postfrontal is really wedged in between them. ‘The frontal passes back to the plane of the posterior end of the pineal foramen. The postfrontal scarcely appears on the surface, but though nearly hidden by the frontal and postorbital, it is really a bone of appreciable size underneath. The postorbital is a large and powerful bone. It forms most of the strong postorbital arch and the whole of the upper border of the temporal fossa, and pretty certainly articulates with the squamosal. The preparietal is unusually large. It passes far forwards between the frontals, and posteriorly nearly surrounds the pineal foramen. The parietals appear to be ankylosed into a fairly large bone lying mainly behind the pineal foramen. It forms a moderately large. somewhat concave intertemporal region. Anteriorly it sends forward on each side a slender process between tie frontal and postorbital, and posteriorly a long process between the postorbital and interparietal. The interparietal is a little broader than the middle part of the parietal. Laterally it articulates with the large tabular. The occiput is fairly well preserved, and was probably much like that of Bainia tigriceps, though some degree of crushing of the squamosals gives it a rather different appearance. The occipital condyle is much shorter in the present species. The squamosals are, as in all Dicynodonts, large and powerful. The shape of the zygomatic portion can be seen in the figure to differ appreciably from that in B. tigriceps. The following are some of the principal measurements :— Greatest length (oblique) ............... 500 mm. Greatestioresdtl 2 yo. 0 see. douse ev esewsen 480 ,, Beisallb lemon iatiiecus cits Sis: ily cy eal ce AGO) ee Width across nasal bosses ............... DOS, Interorbutal wadth 27225 eee ene T3307. Intertemporal width across parietals... 50 ,, Intertemporal width across post- OL Italy eee eas AA te I alk 86 BAINIA HAUGHTONI, sp.n. (Text-fig. 38.) In Haughton’s paper on the Anomodonts in the South African Museum, recently published, he figures and describes a skull (8S. A. M. No. 3328) which he believes to be a young 662 MR. R. BROOM ON ANOMODONT REPTILES I have examined this skull, specimen of Dicynodon laticeps. and while it has all the appearance of a young skull, it cannot, in my opinion, be the skull of a young D. laticeps, as it differs in having a very large preparietal and very large postfrontals. Nor can it be the young of either Bainia tigriceps or D. peavoti. And as it differs from all these three species, but manifestly belongs to this group, it is necessary to make it the type of a distinct species which I have much pleasure in naming after Mr. 8S. H. Haughton, who is doing excellent work in the vast South African field. Text-figure 38. \ ~ dD ZAILINNS EES. IV Po. Parietal, preparietal, and frontal regions of Bainia haughtoni Broom. Mr. Haughton, in addition to describing the specimen, has given a figure of the preparietal region. While his figure is essentially correct, a camera lucida drawing I have made of this region gives, I think, a little more accurately the position of the sutures. The following are some of the principal measurements :— imberorbitale width. .6-cceo see eee 46 mm. Intertemporal width ..............0...0+- Boas fda sla oae alae ee Ree ae Across the nasals FROM THE KARROO BEDS OF SOUTH AFRICA. 663 EosImMoPs NEWTONI, gen. et sp.n. (Text-fig. 39.) This genus and species is founded on a skull which I dis- covered at Victoria West, and which is of special importance as throwing some light on the age of the rocks which yielded the unique fossil forms Galechirus scholtzi, Heleosuchus scholtzi, Arnognathus parvidens, and Heleophilus acutws. When these fossils were discovered by Mr. Scholtz no other fossil had been found within 40 or 50 miles, and there was much doubt as to the age of the deposit. When I described the forms I thought it very likely that they were some of the land representatives of the Lystrosaurus zone. Shortly afterwards Mr. A. L. du Toit, of the Geological Survey, reaching Victoria West from the north Text-figure 39. Skull of Hosimops newtoni Broom. The apparent asymmetry is partly due to crushing, but largely owing to different degrees of the surface of the bones having been flaked off. through a very unfossiliferous region, came to the conclusion that the beds at Victoria West were probably of the Pareia- saurus zone. For many years I have taken Mr. du Toit’s opinion, but working from the South I have been compelled to differ. Biesjespoort, which lies about 20 miles south of Victoria West, is very rich in fossils, and is certainly at the base of the Cistecephalus zone. For miles the strata are almost perfectly horizontal, and it is hard to helieve that the beds at Victoria West can be much lower than those at Biesjespoort. Unfortu- nately, the intermediate beds so far as examined are almost Proc. Zoou. Soc.—1921, No. XLV. A5 664 . Mi. R. BROOM ON ANOMODONT REPTILES devoid of fossils till we come to the little rich patch at Victoria West. Soon this locality will be for ever drowned in a oe dam which is being made above the township. A small Dicy nodont skull was some years ago discovered in the deposit, which I referred to Pristerodon aches Huxley, and took this as confirmatory evidence of the deposit being in the Pareiasaurus-zone; but we now know that very Se forms belonging to the genus I am ‘calling Hmydopsis-oceur in the Oistecephalus zone. The new Dicynodont skull which I have discovered is nearly perfect, but being in a hard matrix cannot be very satisfactorily developed: ‘It is’ about 5 inches long and 4 broad, and is chiefly remarkable from having the preorbital region very short and deflected and the parietal region very broad. The premaxilla is extremely small and short. The nostrils are large and directed forwards. The nasals,. like the premaxilla, are also unusually small. The septomaxilla does not show on the facial surface. The maxilla though fairly well developed is very short. It has a rather small slightly flattened tusk. The prefrontal and lacrimal are both small. The frontals are long and narrow, and form only a small part of the orbital margin. Between them lies an exceptionally large and long preparietal. The postfrontals are also very large. The post- orbitals, on the other hand, have very short posterior processes which do not meet the squamosals, The parietals are well developed and broad. The pineal foramen is unusually small. The squamosal does not seem to present any unusual features. The occiput has not been cleaned of matrix, The species in the folding down of the snout recalls Lystro- saurus, though the narrow interorbital region gives the upper surface quite a different appearance. In Dicynodon moschops Broom we have a very similar folding down of the snout, and at first sight one might be led to think that this skull was the male of D. moschops. When, however, we look at the structure we find very great differences. In D. moschops we find the septo- maxillary large and forming part of the face. Here it does not appear on the face. Further, in D, moschops, though the parietals are also broad the preparietal is small, and there are no post- frontals showing on the upper surface. The species of Bainia also show some affinities with the present type, but they all differ in having the frontal region broader than the parietal. Still we may conclude that this Victoria West type has aflinities with D. moschops on the one hand and with the species of Bainia on the other. As all the known species of Bainia occur in the Oistecephalus zone, as does also D. moschops, we may, I think, conclude that the Victoria West animal is either from the Oistecephalus zone or from the zone immediately below—viz., the Endothiodon zone. In any case it is pretty certain that the Victoria West deposit cannot belong to the Pare tasaurus zone. FROM THE KARROO BEDS OF SOUTH AFRICA. 665 It is certainly very remarkable that, while we know the faunas of the Hndothiodon and Cistecephalus zones fairly well, nowhere else have we ever come across any of the forms that occur at Victoria West. Though the present Dicynodont is fairly new to the typical Dicynodons, I think it well to make it the type of a new genus, which may be called Hosimops, characterized by the very short deflected snout and the broad .and unusual condition of the parietal region. The specific name has been given in honour of Mr. E. T. Newton, F.R.S., whose work I have always greatly admired. The following are the principal measurements :—- Greatest antero-posterior length... about 140 mm. Greatest) ORE AGU hier i a1 mee Tidy, LOOW fe Janibesor bitterly, oo) ase ent eevee i BOie ee linterteniporalawadtiny (amu eek vec SOU PALEMYDOPS PLATYSOMA, gen. et sp. n. (Tl ext-fig. 40. »§ i § This new genus and species is founded on a beautiful skull and part of the skeleton of a small Dicynodont found by me at Text-figure 40. Skull of Palemydops platysoma Broom. Biesjespoort. ‘The head is barely 3 inches long, and one might consider the possibility of its being a young Dicynodon. But this cannot well be the case, as the bones are well ossified and the tusk apparently fully formed. Further, the structure of the top of the skull is unlike that in any known species of Dicynodon. 45* 666 MR. R. BROOM ON ANOMODONT REPTILES In fact, the top of the skull differs so greatly from that of typical Dicynodons, that I feel justified in making the specimen the type of a new genus, Palemydops. The skull is flat and broad. The snout is very short. There are no indications of any thickenings of the nasal bones. The maxilla is short and has a well-developed round tusk. The sutures of the snout elements cannot with certainty be made out. ‘ne frontals are moderately broad, and on the back part of each is a long deep, probably glandular pit. The postfrontals are large, and form a considerable part of the orbital margin. The postorbitals are very large, and form the whole of the inner wall of the temporal fossa. The preparietal is relatively small and narrow. The pineal foramen is small, and lies much behind the plane of the post- orbital arches. The parietals are large. The following are the principai measurements of the skull :-— Greatest antero-posterior length ...... 75 mm. GREASE eye lie ohisb cite uo Aatctea oat nace Ri aakias Oily was erntencoxglorivall whit a o2 ee oc ae 5. Sele oye Intertemporal Wath cock... cder ees AD) ee EMYDOPSIS TRIGONICEPS (Broom), gen. n. (Text-fig. 41 A.) In 1904 I described a small Anomodont skull as a species of Oudenodon and called it Oudenodon trigoniceps. With the much Text-figure 41. A. Preparietal region of Hmydopsis trigoniceps Broom. B. Preparietal region of Hmydopsis sciuroides Broom. greater knowledge which we now have it became pretty manifest that it could not be a species of Oudenodon, or rather of Dicynodon, the parietal region being quite different from the Dicynodon type» The specimen is now in the Albany Museum, and Mr. J. Hewitt, the Curator, kindly sent it to me on Joan for further examination. I discovered on breaking into the maxilla that there are three or four slender molar teeth. These are long FROM THE KARROO BEDS OF SOUTH AFRICA, 667 and rounded, and with small serrations on the posterior side of the upper portions of the crowns. Manifestly the genus is allied _to Pristerodon and to Hmydops, and also to Dieluwrodon, but it seems wisest to make a new genus for those forms with three or four slender posteriorly serrated teeth. Pristerodon has a larger number of molars—6, 7, or 8—and they are robust teeth. Emydops has a few teeth which are apparently not serrated. Diclurodon has a number of molars which have serrations both in front and behind. Emudopsis trigoniceps (Broom) may be taken as the type species of the new genus. Emypopsis Loneus, sp. n. (Text-fig. 42.) This species is founded on a collection of eight good and a number of imperfect skulls found by me at Biesjespoort. ‘They occur in exactly the same horizon as Diceynodon sollast. As most of the skulls are about 22 inches in length, and no one is more than 22 inches, we may safely conclude that they are adult skulls, Text-figure 42. Skulls of Hmydopsis longus Broom. A. Type skull. B. Skull of a topotype illustrating the variability of the preparietal. The snout is very appreciably shorter than the antero-posterior diameter of the orbit. It is above rounded and moderately smooth, though the nasal bones are uniformly thickened. The septomaxillary does not seem to be seen on the face, though it is impossible to be quite certain. 668 MR. R. BROOM ON ANOMODONT REPTILES The frontals are long and narrow, and nearly shut out from the orbital margins by the preefrontals and large postfrontals. They pass far backwards by the sides of the preparietal. The frontals are moderately flat. The preparietal differs in different specimens considerably, but the appearance of the postfrontals, postorbitals, and parietals is — very constant. I give figures of the region in two specimens to illustrate the variation. The parietals are large and broad and flanked by the post- orbitals. In no specimen is there a tusk, but in both upper and lower jaws are a few slender posteriorly serrated molars. The number appears to vary with age. In a beautifully preserved lower Jaw there are three teeth. The first is large and is on the point of being shed, the replacing tooth being seen below it. The second tooth is also well developed. he third tooth is very small. A few years ago Haughton and I described a small skull from Dunedin, which we provisionally referred to Hmydops and named Emydops piatyceps. There can be little doubt that this is very closely allied to the present species, and should be transferred to the genus Hmydopsis. The only striking difference between Hmydopsis platyceps and ~ Emydopsis longus is that in the latter the posterior part of the skull is much longer and broader, and while #. platyceps has slender tusks, #7. longus has none. The following are the principal skull measurements (in milli- metres) of three specimens :— Type. B. C. Greatest length ............ 64 68 (2 Greatest width ...... about 48 about 54 about 52 Interorbital width ......... 12 11 12 Intertemporal width ...... 18 18 20 Basaldlencthy sce. ave... 59 about 62 — EMYDOPSIS SCIUROIDES, sp. n. (T'ext-fig. 41 B.) The beautiful little skull which forms the type of this new species was found by me at New Bethesda, the famous locality which has already yielded so many interesting types. Without seriously injuring the skull, it will not be possible to say with certainty if the specimen belongs to Hmydopsis, but the general resemblance of the skull and lower jaw to those of species of Hmydopsis is so close as to render it highly probable that the specimen belongs to a species of this genus. The next striking characteristics of the skull are the large size of the orbit and the relative shortness of the parietal region. The snout is short but relatively broad. A tiny corner of the septomaxillary appears on the face. The maxilla is deep, and in this type skull at least, tuskless. The prefrontal and lacrimal are small. FROM THE KARROO BEDS OF SOUTH AFRICA. 669 The frontal is very long and narrow, and remarkable in that the postorbital portion is wider than the orbital. The preparietal is almost unique among Anomodonts in being very small and situated entirely in front of the pineal foramen. A similar condition is only known to occur in Emydopsis arctatus (Owen). The postfrontal is either entirely absent or does not appear on the surface. The postorbital is very large,and extending forward into the region usually occupied by the postfrontal, forms a con- siderable part of the orbital margin. The parietal is moderately wide but short, the part behind the pineal foramen being extremely short. ‘The pineal foramen is very small. In the lower jaw the dentary resembles that of Hmydopsis longus, but the post-dentary portion of the jaw is relatively much shorter. The following are the principal measurements of the skuil :— Greatest lemothy 0. 2.2. ..6665tekees so about 60 mm. Greatest width ......... doubtfuily about 50 ,, Toterorbitalwidtbr i) a.28).2 ono dae OSes Wnitertemporal width .0 3.015 ae. get maces Teli 45. Between caniniform processes ............ Llores 5, Basa lemau hie hoes ceacca ts one about 48 ,, Skull of Hmydops parvus Broom. The skull is slightly crushed, but could not well have been longer than indicated by the dotted line. Emypors PARVUS, sp. n. (Text-fig. 43.) Until the discovery of the present specimen, Hmydops minor was the smaliest known Anomodont, but this new find gives us a type which has certainly a shorter skull and possibly smaller in other respects. This new species is founded on a skull discovered by me at Bruintjeshoogte in the same stratum as yielded the type skull of Ictidosuchus longiceps Broom. The skull is 14 inches long and rather less in breadth, and the whole animal was probably as small as a newly-born kitten. 670 MR. R. BROOM ON ANOMODONT REPTILES The snout is very short, but the structure cannot be made out owing to crushing. The orbits are small. The frontal region is relatively narrow, and the temporal region in front nearly twice as wide. Owing to a finely crackled condition of the bones, sutures are very difficult to make out, but the limits of the preparietal can be clearly seen, The pineal foramen is small and situated well back. The general shape of the skull will be best indicated by the drawing given. There is evidence of at least one small unserrated molar, The following are the principal measurements of the skull :— Greabestjlemetin, 5... .cceherons see cay about 38 mm. Greatestuatea dtl, <5. .assh eee ee oy nee hatererbitalaid tly. oft casas eines SD) i, Mnibertemnyonall wad tho 5 ees eee Tay ae EMYDURANUS PLATYOPS, gen. et sp.n. (Text-fig. 44.) The new genus and species is founded on a small skull dis- covered by me near Biesjespoort, and is one of the most interesting Anomodonts known. The specimen was found in the deposit which has yielded the very numerous skulls of Dicynodon sollasi, and before being developed it was looked upon as another of the many duplicates. This was unfortunate, for a beautiful shoulder- girdle lying against the skull was developed out and a portion of the skull sacrificed before it was noticed that the skull is really a very remarkable new type. We have still preserved the greater part of the skull, minus the arches, but with the contact between the occiput and the parietal region lost. The skull is that of a small Anomodont with a broad, flat head and a wide palate which has on each side three or four teeth. The premaxilla is broad and shallow. As in typical Anomo- donts, it forms the greater part of the bony palate. There is a well-marked median ridge. Its relation to the palatines and maxillze will be best understood by the figure. The maxilla is well developed, but is remarkable from the fact that there is very little of a caniniform process. Some distance inside of the alveolar margin are three or four teeth in a row. On the right side the matrix has been left supporting the crowns. The first is a small sharp-pointed unserrated tooth; the second a much broader flattened tooth, also without serrations; the third probably similar to the second; and the fourth a smaller tooth imperfectly preserved. On the left side the matrix, and with it much of the teeth, has been removed. The first tooth is seen to have its root near the suture between the premaxilla and maxilla, and though it doubtless is fixed in the maxilla, it passes through the premaxilla. The others are in the posterior part of the maxilla. The palatine is well developed and has an unusually large FROM THE KARROO BEDS OF SOUTH: AFRICA. 671 palatal portion, which is rugose and has probably had a horny protection. The posterior part of the palate is not preserved. The nostrils are situated well forward and separated by the median process of the premaxilla, which passes well back. The nasals are large. Each has a well-marked boss behind and above the nostril. The septomaxille cannot be made out. The frontal is long and narrow, and passes unusuaily far behind the plane of the postorbital arch. The postfrontal is small and narrow, but forms a considerable part of the orbital margin. Text-figure 44. A. Top of skull of Emyduranus platyops Broom. B. Palate of Hmyduranus platyops Broom. C. First three molar teeth of same, enlarged. The preparietal is remarkable in being situated quite in front of the pineal foramen. It islozenge-shaped, the two posterior sides articulating with the parietals and the anterior with the frontals. The postorbitals are large. The occiput, so far as can be seen, presents no unusual feature. The following are the principal measurements :— Greatest length............. eae about 90 mm. Greatest Widhhtee. "aqme hac Be HDS Nt Intévorbital width: ee ae Gis once Intertemporal width ..................... ZO Between the caniniform processes ...... PS eget Front of snout to frontof pineal foramen 62 672 _ MR. R. BROOM ON ANOMODONT REPTILES: This little Anomodont, to which I have given the name Einyduranus platyops, shows some affinities with Cryptocynodon simus Seeley, and also with Prodicynodon pearstunensis Broom, but must be regarded as the representative of a very distinct genus. ENDOTHIODON CRASSUS, Sp. n. (Text-fig. 45.) This new species is founded on a beautiful skull discovered by me at Dunedin, Beaufort West district. The skull has lost the lower jaw, and has the parietal region badly weathered and the squamosal portion of the left temporal arch unperfect, but other- wise it is almost complete. With the exception of Hndothiodon whaitsi it is the largest Hndothiodon skull known, and it is little inferior to that other, though very different-in shape. Unlike Hndothiodon uniseries Owen and Endothiodon whaitsi Broom, it is very broad and flat, and the parietal region is only slightly elevated above the frontal plane. In general shape and proportions the species which it most resembles is Hndothiodon paucidens Broom, but that it is very different from this species will be at once evident on comparing the figures. The premaxilla when viewed from the front is like a thick inverted V, the deep notch being for the point of the mandible, and the two points are the caniniform processes, which here are formed by the premaxilla and not, as in the tuskless Dicynodonts, by the maxille. A small knob is situated in front of each nostril. The superior median process passes between the nasals, but is short. The nasals are very broad, and the bones, like all the bones of the top of the skull, very thick. They are as broad as long, and the whole preorbital portion of the skull is thus much broader than long—a condition which differentiates this species of Hndo- thiodon from ail others. The prefrontals are of moderate size, and so thickened that they may be looked upon as forming preorbital bosses. As in all typical Endothiodons, there are two longitudinal grooves passing from the front of the frontals down the middle of the nasals to the upper corners of the nostrils. The septomaxilla seems to be small and within the nostril, but the evidence is not satisfactory as to whether it also appears on the face. The lacrimal is smaller than the prefrontal, but its limits cannot be clearly made out. The maxilla is a large bone. As in all species of Hndothiodon, it has a long ridge which forms an alveolar border and at a considerable distance further in a row of teeth. Most of the teeth appear to be lost, but when complete there was probably a single row of eight teeth which measured 74 mm. The first tooth of the series is round in section and has a diameter of 7 mm.; the last, also round, has a diameter of 85 mm. The maxilla forms most of the lateral part of the face, and it sends FROM THE KARROO BEDS OF SOUTH AFRICA. 673 a long and powerful process outwards and slightly backwards below the jugal to strengthen the jugal arch. The jugal is, as in other Endothiodonts, more powerful than in the Dicynodonts, and has a large rounded descending knob. Tt sends a long process backwards inside the squamosal. Text-figure 45. Skull of Hndothiodon crassus Broom. Greatly reduced. The quadrato-jugal and quadrate (QJ and Qu) of the right side are displaced : those of the left side lost. The frontals are large, but the exact limits of the bones cannot be made out except in front as most of the orbital margins have been lost. . The preparietal is large and forms the front of the pineal foramen wall. The parietals are narrow but deep. The squamosals are powerful. The zygomatic portion is more slender than in Dicynodonts, but the occipital portion is greatly developed. 674 ON ANOMODONT REPTILES FROM SOUTH AFRICA. The quadrate and quadrato-jugal are lost from the left side, but preserved in a slightly displaced condition on the right. As in Dicynodon, they are apparently ankylosed, and their displace- ment from the squamosal and paroccipital adds further evidence to observations I have made in Dicynodon, that while the quadrate is fixed to the quadrato-jugal, the quadrato-jugal has a very considerable degree of movement on the squamosal, and thus relatively to the skull the quadrate is perhaps as freely movable as in the Lizard, in Anomodonts, and also in Cynodonts. The palate is similar to that in other Endothiodonts, and does not differ greatly from the Dicynodon type. The periotic processes do not descend to the same degree as in Dicynodon, but are clasped by the basisphenoids in a similar manner, and have each a large foramen ovale for the end of the stapes. Hndothiodon differs from Dicynodon in having a large median basisphenoid boss. The occipital condyle is flat as in other Hndothiodon species. The following are the principal measurements of the skull :— Greatest antero-posterior length...... 502 mm. Greatest breadth across squamosals.... 400. ,, Greatest breadth across jugals........ 430 ,, Breadthacrossmasalay Woe os et 180, Breadth between nostrils............ 60 WK 9 VIN idl SANTO SUA) SOO) ONT el \/ il ‘duit yang ‘URL 38 'Tep Usedy'p ‘hal INAV BRE WG GIL IS) Zara - i ‘ Z| : i ‘dur yyy TUE Nel INNS) BUSl GGL 1S) 4 Gl 8 WIGLOVOINN SS IMO) SUELO) Ul el ete! ‘UTI 1.8 Tap weary * i ‘dur yyy ‘YUL ® Tap used Pp CEES AML Wel IND) BOSE IGE (SZ wal SeiemecemmAt a nites 4 ; x ¢ ee Sindee Apes i eta Ie), Avro), MUS) 21, 1eUe GE ANIN| 2 IRL IW, Huth imp. J.Green del. et lith. aI LONG IGNeS(OMOUS) JE UMENEIN (GNSS IEOUMNUnS, 2) ink, “IWMVatesIRdOy sy ION WaleeAzr, 4 i 4, uy & t ; " s 2 , a ‘S/, SAHdIONOT SINOWUHDOTAVH 4‘ WSSVAN VHVOONOTNV'I ‘dur yqny “Yd 18 [ep uaeay'p IX Wel INDO) Bal Te (S ss Z cal wy ‘SNHONAHHOTIdS SINOYHOO Td du yany UA ual INS) BRET LG IES) el é Wat af (4 SANIVHLHAOHOVN SINOWHOOHANVHE T ‘Yat 38 'Tep uae4p' lL ON THE CICHLID FISHES: OF LAKE NYASSA. 675 36. The Cichlid Fishes of Lake Nyassa. By C. Tarr Recay, M.A., F.R.S., F.Z.8., Keeper of Zoology, British Museum (Natural History). [Received June 7, 1921: Read June 7, 1921. | (Plates I.-VI.* ; Text-figures 1-30.) The Fishes of Nyassa have been somewhat neglected in com- parison with those of Tanganyika and Victoria. hey were first described by Giinther (P. Z. 8. 1864) from specimens collected by Sw John Kirk, and nearly thirty years had passed before the same author again described fishes from the lake sent by Sir Harry Johnston (P.Z.S. 1893). Later a collection made by Captain E. L. Rhoades was described by Boulenger (Ann. & Mag. N. H. (8) 11. 1908). In Boulenger’s ‘ Catalogue of African Fresh-water Fishes,’ 111. (1915), 38 species of Cichlide are recorded from Nyassa; one of these, Petrochromis nyasse, may now be removed from the list. The supposed occurrence of the specialized Tanganyika genus -Petrochromis in Nyassa was dificult to explain; re-examination of the type of P. nyassew leads me to regard it as identical with the more recently described P. fasciolatus, and I have no doubt that the locality assigned to it was an error on the part of the collector. The loss of this species from the Nyassa list is made good by the re-establishment of Ginther’s Chromis subocularis, placed by Boulenger in the synonymy of C. johnstoni, so that the number of valid species of Cichlid hitherto described from Nyassa is 38. The present revision is based on an examination of the speci- mens in the collection of the British Museum (Natural History), including the types of all the species described by Giinther and by Boulenger, but principally on the study of a very fine collection made and presented to the Museum by Mr. Rodney C. Wood. As a result, the number of species is more than doubled, 46 being described below as new to science. Of the 84 species all but 5 (3 Tilapia, 1 Astatotilapia, 1 Serranochromis) are endemic, and the proportion of endemic genera is high, 11 out of 15, but more than half the species belong to the widely distributed genus Haplochromis. ‘The majority of the Nyassa genera are quite distinct from any found elsewhere : for example, Lhanyphochromis, which may be supposed to occupy the same place in Nyassa that Bathybates does in Tanganyika, is very different from Bathybates. There are, however, a few remarkable examples of convergent evolution in Nyassa and Tanganyikat. The fish described below as Fseudotropheus tropheops bears a great superficial resemblance to Zropheus, and has the same peculiar dentition ; another new type, dulonocara, has deep channels with large openings in the frontal, nasal, orbital, preopercular, and mandibulary bones, exactly as in Vrematocara. * For explanation of the Plates see p. 727. + For the Tanganyika genera see Regan, Ann. & Mag. N. H. (9) v. 1920, p. 33. 676 ; MR. C, TATE REGAN ON THE Synopsis of the Genera. T. Scales cywloid or very finely denticulate. A. Pharyngeal apophysis formed by parasphenoid on Teeth in narrow or moderately broad bands, outer bicuspid, inner tricuspid . ‘sate yee Male tathl lecaboles eatets ED ANee LLC ROTC Cl Teeth very small, in very “broad bands ...:.... i.e. 2. Corematodus. Teeth in a few series, slender, with expanded crowns, the outer obliquely tr uncated . EAE RN ee heat aoe eLbemaunnlapea. B. Pharyngeal apophysis eet by Deo in middle and prootics at sides. Teeth moderate, outer mostly bicuspid, inner tricuspid....., 4. Otophargne. Meet very Shouts NICU Spee es peck deter ssehie-e seek) see seenenan DEmCe LON OLN aunoa Gl. II. Scales usually distinctly denticulate. Pharyngeal apophysis formed by para- sphenoid in middle and hasioccipital at sides. A. Bones of head with small mucitferous canals with small pores. 1. Premaxillaries without anterior beak-like expansion. a. ae of membrane of spinous dorsal produced into Jappets. . Teeth conical, or outer bicuspid and inner tricuspid. Teeth of iste series in upper jaw first decreasing and then increasing in size antero-posteriorly ; band of teeth in lower jaw crescentic ...... 6. Astatotilapia, Upper jaw with a rather broad ‘band of small ‘cuspidate teeth anteriorly and a series of enlarged conical teeth on each side; lower jaw with a band of teeth with incurved lateral edges, from which a single series runs back on each side ...... sistestersye, @ Eseudotropheus. Teeth conical ; lower jaw w ith 4 anterior canines ............ 8. Cynotilapia. Teeth conical; no canines; 7 to 9 series of scales on cheek. 9. Serranochromis. Outer teeth decreasing in size posteriorly; no distinct canines; bands of teeth aa 2 to 6 series of scales on cheek ...... : 10. Haplochromis: Teeth very small and slender, forn ming narrow v bands which are interrupted at the sv mphyses. Para une Ae A eis erh: mops. g. Outer teeth very broad, compressed, with or without a pair of small lateral cusps ....... .... 12. Docimodus. b. Edge of membrane of spinous dorsal 1 running evenly between the tips of the spines ..... Mauna. \Cyntocanas . Praemaxillaries with an anterior beak-like expansion. 14, Rhamphochromis. B. Frontals, nasals, orbitals, preopercular and mandibulary bones deepl : a v, ply excavated, with large openings ..........-....0eesee ee 15. Aulonocara. 1. Trnapra A. Smith, 1840 (type 7’. sparmannt A. Smith). Teeth small or moderate, in several series, the outer bicuspid, the inner tricuspid (sometimes conical in large fish), Seales eycloid or feebly denticulate. Pharyngeal apophysis formed by pavasphenoid only. Africa; numerous species. The four species found in Nyassa may be distinguished follows :— I. Three anal spines. A. Caudal scaled only at the base. Maxillary extending to below eye .............c0ceeccc eee eeseee eee s eee 1. mossambica. Maxillary not extending to below eye ..............ceceeeeeseesses. 2. melanopleura. B. Caudal covered with small scales ............ 6.ce i.e. 8. syuamipinnis. JT. Hourjanallspines: feiss acdc a WANN as era asta ee Ne ERG TEC ROL ce CICHLID FISHES OF LAKE NYASSA. 677 1, TILAPIA MOSsAMBICA Peters, 1852. Bouleng. Cat. Afr. Fish, iii. p, 154, fig. 101. Kast Africa, 2. ‘TILAPIA MELANOPLEURA A Dum., 1859. Bouleng. Cat. Afr. Fish. iii. p. 190, fig. 123. West Africa, Congo, and Zambesi. 3. TILAPIA SQUAMIPINNIS Gtnth., 1864. Bouleng, Cat. Afr. Fish. iii. p. 183, fig. 118. Nyassa. 4, TILAPIA SHIRANA Bouleng., 1896. Bouleng. Cat. Afr. Fish, iii. p. 151, fig. 98. Nyassa and Portuguese H. Africa. 2. Corematopus Bouleng., 1896. Jaws with very broad bands of very small teeth, the outer with expanded, compressed, and obliquely truncated crowns, the inner mostiy pointed, Scales cycloid., A single species from Nyassa, CoREMATODUS SHIRANUS Bouleng., 1896. Corematodus shiranus Bouleng. Cat. Afr. Fish. iii, p. 494, fig. 342. Depth of body 22 in length, length of head 3. Snout with convex profile, shorter than postorbital part of head. Diameter of eye equal to preeorbital depth, 5 in length of head: interorbital width 23. Jaws equal anteriorly; maxillary extending to below eye. 4 series of scales on cheek. 12 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 34 scales in a longi- tudinal series. Dorsal XVI 11; last spine 2 length of head. Anal IIIT 9: third spine 4 length of head. Caudal scaly. Caudal peduncle 13 as long as deep. Body with seven narrow blackish cross-bars as in Tilapia squamipinnis, the first downwards from origin of dorsal, the last two on caudal peduncle. The type, 200 mm. long. 3. Hemirinaria Bouleng., 1902. Teeth in 3 to 6 series, small, with slender shaft and compressed and expanded crown, those of outer series obliquely truncated. Scales feebly denticulate. A single species from Nyassa. HEMITILAPIA OXYRHYNCHUS Bouleng., 1902. Henatilapia ovyrhynchus Bouleng. Cat. Afr. Fish, iii. p. 489, fig. 339. 678 MR. C. TATE REGAN ON THE Depth of body 22 to 23 in the length, length of head 3 to 33. Snout with straight or concave profile, from as long as to 13 diameter of eye, which is 33 to 53 in length of head, in adult equal to preorbital depth; interorbital width 33 to 4 in head. Jaws equal; maxillary not extending to below eye. 3 or 4 series of scales on cheek. 12 or 13 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 33 to 35 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVI 10-11; last spine from less than #2 to 3 length of head. Anal ITI 9-10; third spine from less than + to more than # head. Pectoral as long as or a little shorter than head. Caudal scaly, truncate or slightly emarginate. Caudal peduncle longer than deep. A dark spot on operculum, usually two on lateral line below spinous and soft dorsal respectively and another at base of caudal. Dorsal and caudal spotted with orange ; anal with long spots between the rays; males with dorsal, anal, and pelvic fins darker, pale-edged. Seven specimens, 90 to 190mm. long. (Voore, khoades, Wood). 4. OvopHaRYNX Regan, 1920 (type Tilapia auromarginata Bouleng.). As Tilapia, but the prootic forms part of the facet for articu- lation of the upper pharyngeal on each side. Nyassa ; two species. 1. OroPpHARYNX AUROMARGINATUS Bouleng., 1908. Tilapia auromarginata Bouleng. Cat. Afr. Fish. 11. p. 180, fig. 115. e Depth of body 23 to 2% in length, length of head 34 to 33. Snout decurved, as long as or a little shorter than postorbital part of head. Diameter of eye equal to depth of preorbital, 32 to 4 in length of head; interorbital width 3 to 33. Jaws equal anteriorly ; maxillary not extending to below eye; teeth in 3 to 5 series, inner tricuspid or conical; 60 to 75 in outer series of upper jaw, the anterior bicuspid, the posterior conical. 4 series of scales on cheek. 15 to 18 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. Scales cycloid, 33 or 34 in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVII-XVIII 10-11; last spine 2 length of head or a little more. Anal IJI 9-10; third spine stronger than dorsals, } head or a little more. Pectoral as long as head, not reaching anal. Caudal densely scaled, emarginate. Caudal peduncle 1} as long as deep. Bluish; dorsal and anal dark blue with yellow edge; dorsal with series of red spots. Two specimens, 205 and 240 mm. in length, and a skeleton. 2 CICHLID FISHES OF LAKE NYASSA. 679 2. OTOPHARYNX SELENURUS, sp.n. (Text-fig. 1.) Depth of body 24 in length, length of head 33. Snout with straight profile, as long as postorbital part of head. Diameter of eye nearly equal to depth of preorbital or cheek, 4 to 44 in length of head, interorbital width 4, Jaws equal anteriorly; maxillary not extending to below eye; teeth in 3 or 4 series, 50 to 55 bicuspid teeth in outer series of upper jaw. 3 or 4 series of scales on cheek. 12 gill-rakers on lower part of anterior arch. 35 scales in a longitudinal series, 6 from origin oH dorsal to iateral line. Dorsal XVI-XVIT 11-12; last spine length of head, longest soft rays nearly as long as head. Ian IIL 9; third spine stronger and shorter than last dorsal. Pectoral as long as head, reaching anal. Caudal sealy, deeply emarginate. Caudal peduncle 13 as long as deep. Bluish grey, uniform or with traces of darker cross-bars. Text-figure 1. tol Otopharynx selenurus. Two specimens, 135 and 155 mm. in total length (Wood). A smaller example, 90mm. long, is more slender (depth 3 1n length), but, except for juvenile characters, such as the larger eye (33 in head), i is very similar to the two deseribed ; ; colour silvery, ith traces of several cross-bars; an oblong dark spot on lateral line below middle of spinous dorsal, a band along lower lateral line. 5. CHILOTILAPIA Bouleng., 1908. Jaws with an outer series of stout, little compressed, teeth, and 3 or 4 imner series of smaller subconical teeth, which are stronger at the sides of the premaxillaries than in front. Scales cycloid or feebly denticulate. Nyassa ; a single species, Proc, Zoou. Soc.—1921, No. XLVI. 46 680 MR. C. TATE REGAN ON THE CHILOTILAPIA RHOADESII Bouleng., 1908. Chilotilapia, rhoadesti Bouleng. Cat. Afr. Fish. ii. p. 499, fig. 366. Depth of body 21 to 23 1n the length, length of head 33. Snout short, declivous ; onsen of eye 4 to 42 in length of head, preorbital depth 4, interorbital width 3. ie aws equal anteriorly ; mouth wide, with short lateral cleft. 3 or 4 series of scales on cheek. 11 eill- rakers on lower part of anterior arch. Middle pharyngeal “teeth moderately strong, subconical. 32 to 34 scales in a longitudinal series. Dorsal XV—XVI 10; last spine 2 to 4 length of head. Anal III 9-10. Pectoral a little longer. than head, reaching anal. Caudal densely scaled, emarginate. Caudal peduncle a little longer than deep. Silvery or bluish; an opercular spot ; sometimes a broad dark band on each side of back and another on middle of side; soft dorsal with series of spots. The type (220mm.) and two specimens of 180 and 200mm. (Wood). 6. AsraToTILAPIA Pellegr., 1904 (type Labrus desfontainesti Lacep.). Near Haplochromis, but posterior teeth of outer series of upper jaw increasing in size backwards. Teeth in 3 to 5 series, cuspidate or conical, those of outer series of upper jaw sometimes bicuspid anteriorly, conical posteriorly ; band of teeth in lower jaw crescentic. Middle teeth of lower pharyngeal somewhat ‘enlarged. Dorsal XIII-XVIT 8-11. Anal III 7-11. Scales 26 to 36. Africa. This genus includes three species placed by Boulerger in 7tlapia, namely, 7. sywnnertoni, 7’. burton, and 7. calliptera, and four included by him in Haplochromis, H. strigigena, H. moffati, H. desfontainesti, and H. moerwensis. These are all closely related, agreeing in the form and size of the mouth, the rather short pectorals, comparatively short caudal peduncle (as long as deep), and the rounded caudal. ASTATOTILAPIA CALLIPTERA Giinth., 1893. Tilapia calliptera Bouleng. Cat. Afr. Fish. iii. p. 222, fig. 145. Depth of body 23 to 24 in length, length of head 23 to 31. Snout with straight pr ofile, shorter aha postorbital part a head. Diameter of eye 2 to 43 in length of head, in adult seauge ly greater than depth of preerbital; interor bital width 33 to 43 in length of head. Jaws equal anteriorly ; maxillary extending to below anterior edge of eye; teeth in 3 to 5 series, 40 to a0 fe outer series of upper jaw, the anterior bicuspid, the last 2 or 3 (young) or 8 or 9 (adult), conical, increasing in size posteriorly. 3 to 5 series of scales on cheek. 8 to 10 gill-rakers on lower part CICHLID FISHES OF LAKE NYASSA. 68] of anterior arch. Middle pharyngeal teeth rather strong, conical in adult. 29 to 33 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XIV—X VI 8-11; last spine 2 to 4 length of head. Anal III 7-9; third spine stronger and usually shorter than last dorsal. Pectoral shorter than head, not reaching anal. Caudal rounded. Caudal peduncle as long as deep. An opercular spot and a dark bar from eye to end of maxillary; body with or without dark cross-bavs and a dark lateral band; dorsal and caudal sometimes spotted; males with lower fins blackish, the anal with 2 to 6 rounded orange spots. Nyassa and Zambezi. Numerous specimens, 65 to 140 mm. in total length. 7. PsEUDOTROPHEUS, gen. n. (type Chromis willtamst Giinth.). Jaws with several series of teeth anteriorly, the outer bicuspid, the inner small and tricuspid, forming rather broad curved or transverse bands; upper jaw with a series of conical teeth on each side posteriorly, more or less sharply differentiated, some or all larger than the last bicuspid teeth; lower jaw short and broad, with the lateral margin of the band of teeth incurved, and with a series of teeth on each side behind the band. Dorsal X VI-XTX 8-10. Anal III 7-9. Scales denticulate. Nyassa; five species. Text-figure 2. a. Dentition of Pseudotropheus tropheops. ib: tk Tropheus moorii. Synopsis of Species. ‘I. Jaws equal anteriorly. D. XVI-XVII 9. A. III 8. 4 series of scales on cheek........, 1. williamsi. D. XVIIL9. A.I11 9, 5 or 6 series of scales on cheek.... ... 2. zebra. Il. Lower jaw shorter than upper; mouth rounded. D. XVIII 9. A. IL 8. Depth 3im length ........................ 3, novemfasciatus. DaX Desa As LIL 78" Wepthess maene thems iene | Ln auracus. III. Mouth subterminal, transverse ; snout very convex. JD) DRONA RIC TDs en NED GAS ier SRE oo re eo Meee eee uae elas aR 9 éropheops. 46* 682 MR. C, TATE REGAN ON THE 1. PspuporRoPHEUS WILLIAMss Giinth., 1893. Chromis williamsi Ginth. P. Z.S8. 1893, p. 624, pl. lvi. fig. C. Tilapia livingstonit Bouleng. P.Z.8. 1899, p. 134, pl. x. fig. 2; Cat. Afr. Fish. 11. p. 243, fig. 162. Tilapia williamsi Bouleng. Cat. Afr. Fish. ii. p. 225, fig. 147. Depth of body equal to length of head, 3 in length of fish. Upper profile of head convex ; snout as long ag or a little longer than diameter of eye, which is 34 to 4 in length of head, nearly equal to or greater than the interorbital width, depth oe cheek, or pr sorivuanl: Mouth terminal, rounded ; maxillary extending to below anterior edge of eye; teeth in 5 or 6 series, inner small, tricuspid; 40 to 52 teeth in outer series of upper jaw, the last 3 or 4 (young) or 6 to 9 (adult) on each side conical, enlarged and sharply differentiated from the rest, which are bicuspid. 4 series of scales on cheek. 9 or 10 gill-rakers on lower part of anterior arch. Pharyngeal teeth small, slender. 53 scales in a longitudinal series, 6 from origin of dorsal to Lewera! line. Dorsal XVI-XVII 9; last spine a little less than 5 length of head. suo BEE sey ‘third spine a little shorter than dorsal. Pectoral ? length of neo not reaching anal. Caudal truncate 2, densely scaled in basal half. Caudal “peduncle as long as or a little longer than deep. Body with 6 dark cross-bars, or bars represented by a series of spots above lateral line and another on middle of side; an opercular spot and a dark bar or spot on base of caudal; two round white (? orange) spots on anal fin (pro- bably absent in @). Two specimens, types of the species and of 7. livingstoni, 105 and 75 mm. long. 2. PsrUDOTROPHEUS ZEBRA Bouleng., 1899. Tilapia zebra Bouleng. Cat. Afr. Fish. ii. p. 244, fig. 163 (1915). Depth of body 24 in length, length of head 34. Snout de- curved, a little iomec than dnamnetien noe eye, which is 4 in length of head, a little gr eater than depth of preorbital, equal to depth of cheek; interorbital width 3 in length of head. Mouth ter- minal, rounded, rather wide; jaws equal anteriorly : maxillary extending to vertical from anterior edge of eye; teeth in 4 or 5 series, 56 in outer series of upper jaw, the last 8 or 9 conical. 5 or 6 series of scales on cheek. 12 gill-rakers on lower part of anterior arch. 31 scales in a longitudinal series, 8 from origin of dorsal to lateral line. Dorsal X VIII 9; last spine nearly 3 length of head. Anal III 9; third spine a little shorter than last dorsal. Pectoral nearly as long as head, not reaching anal. Caudal probably truncate, densely scaled at base. Candal peduncle a little deeper than long. A dark bar between eyes, a second from opercular spot to occiput ; 6 vertical bars on body, the first from origin of dorsal to base of pectoral, the sixth from soft dorsal to anal: ; three round pale (? orange) spots on anal in male. The type, 105 mm. in total length. CICHLID FISHES OF LAKE NYASSA. 685 9 3. PSEUDOTROPHEUS NOVEMFASCIATUS, Sp. Nl. Depth of body 3 in length, length of head 33. Snout with convex profile, as long as diameter of eye, which is 33 in length of head, equal to interorbital width, greater than depth of preorbital or cheek.. Mouth rounded, lower jaw shorter than upper; maxiilary reaching vertical from anterior edge of eye; teeth in 5 or 6 series, about 40 in outer series of upper jaw, anteriorly bicuspid, the last 6 to 8 on each side conical, some enlarged. 5 series of scales on cheek. 10 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 32 scales in a longitudinal series, 6 from origin of dorsal to lateral line. Dorsal XVIII 9; last spine 4 leneth of head. Anal ITI 8; third spine as long as and stronger ‘than last dorsal. Pectoral shorter than head, not reaching anal. Caudal truncate. Caudal peduncle as long as deep. Body crossed by 9 dark vertical bars, the first 6 corresponding to the 6 of P. zebra, the last 2 on caudal peduncle; end of snout and lower jaw blackish; an opercular spot and a spot on base of caudal; a faint dark band on lateral line and another on middle of side; spinous dorsal with a dusky intramarginal band; soft dorsal and caudal with series of spots. A single specimen, 65 mm. in total length (Wood). 4, PSEUDOTROPHEUS AURATUS Bouleng., 1897. Tilapia curate Bouleng. Cat. Afr. Fish. ii. p. 246, fig. 164. Depth of body 3% in length, length of head 33. Snout with convex profile, a little longer than diameter of eye, which is 4 in length of head, a little greater than interorbital width, depth of preorbital, or cheek. Mouth rounded ; lower jaw shorter than upper; maxillary nearly reaching ver tical from anterior edge of eye; edn in 5 or 6 series, 45 in outer series of upper jaw, the last 4 or 5 conical, enlarged. 4 series of scales on cheek. 9 or 10 gill-rakers on lower part of anterior arch. 33 scales in a longitudinal series. Dorsal X1X 8; last spine 2 length oH head. Anal III 7; third spine as long as last dorsal. Pectoral 2 2 length of head, not reaching anal. Tanda densely scaled, eons: Caudal peduncle a little longer than deep. Lips blackish ; two black bands between eyes; a black stripe from eye, ending in a spot on base of caudal; a second near edge of back, a third on aorsal fin. The type, 75 mm. in total length. 5. PshUDOLTROPHEUS TROPHEOPS, sp. n. (Text-fig. 3.) Depth of body 2% to 3 in length, length of head 33. Upper profile of head very convex; snout as long as diameter of eye. which is 34 in length of head, ‘slightly exceeds preeorbital depth and equals depth of cheek ; interorbital region convex, its width 22 to 22 in length of head. Mouth abe minal, Lneteee ; Jaws with 8 series of small cuspidate teeth ; a series of 6 to 8 well-differentiated conical teeth on each side of see ndleegn 4 series of scales on 684 MR. C. TATE REGAN ON TH cheek. 9 gill-rakers on lower part of anterior arch. Pharyngeal teeth small, slender. 33 scales in a longitudinal series, 6 from fen of dorsal to lateral line. Dorsal XVII 10; last spine 1 length of head. Anal I! 8; third spine stronger and as long as or a little shorter than last dorsal. Pectoral a little shorter than head, not reaching anal. Caudal sealy, truncate. Caudal peduncle as long as or a little longer than deep. Traces of dark eross-bars on body bearing a series of darker spots above lateral line and another on andes of side ; an opercular spot and a spot on base of eaudal ; dorsal wid a blackish intramarginal band. Two specimens, 116 and 122 mm. in total length (Wood). Text-figure 3. , Pseudotropheus tropheops. +. 8. CYNOWILAPIA, gen. n. Near Pseudotropheus, but teeth conical, in a few series, outer large, in lower jaw forming distinct canines anteriorly. Nyassa ; ; a single species. CYNOTILAPIA AFRA Gtinth., 1893. Paratilapia ee Bouleng. Cat. Afr. Fish. 111. p. 325, fig. 218. Depth of body 2? to 3 in length, length of head 3 to 3}. Snout decurved, as Heine as or a little longer than diameter of eye, which is 3% to 32 in length of head, greater than preorbital depth ; intevorbital width 34 in length of head. Jaws equal anteriorly ; maxillary not quite reaching vertical. from anterior edge of eye; teeth conical, trisevial, outer large, 28 to 32 in upper jaw ; lower jaw with four strong anterior canines. 3 or 4 series of scales in cheek. 14 or 15 gill. rakers on lower part of anterior arch, 33 scales in a Taneuimdiungl series, 6 from origin of dorsal CICHLID FISHES OF LAKE NYASSA. 685 to lateral line. Dorsal XVII 9; last spine 2 to 4 length of head. Anal ITT 8; third spine stronger and as long as or a little shorter than last ove Pectoral Shor ter than head, not reaching anal. Caudal? Caudal peduncle as long as or longer than deep. The types, 85 and 95 mm. in total length. 9. SERRANOCHROMIS Regan, 1920. As Haplochromis, but third vertebra without inferior apophyses, fourth with a very small pair. Mouth large; teeth conical ; cheek deep, with 7 to 9 series of scales. A single species. SERRANOCHROMIS THUMBERGI Casteln., 1861. Paratilapia thumbergui Bouleng. Cat. Afr. Fish. ui. p. 328, fig. 220. Nyassa and Zambezi; Katanga and L. Bangwelu; L. Ngami; Angola. 10. Harptocuromis Hilgend., 1888 (type H. obliquidens Hilgend.). Céenochromis Pfeff., 1893 (pectoralis). Champsochromis Bouleng., 1915 (ceruleus). An outer series of bicuspid or conical teeth, decreasing in size posteriorly, and one or more inner series of smaller bicuspid), or Text-figure 4. oy) By RA ” ei ie Hit Hoe LH wea tina sO at ty Mis b a. Lower pharyngeal of Haplochromis tetrastigma. HOS aes, Bs H. placodon. conical teeth. 2 to 6 series of scales on cheek. Scales usually distinctly denticulate. Pharyngeal apophysis formed by para- sphenoid in middle and basioccipital at sides. Third vertebra with inferior apophyses, which meet below. 52 species of this genus occur in Nyassa, all of them endemic. In the great number and diversity of species of Haplochromis 686 MR. C. TATE REGAN ON THE Nyassa resembles Victoria and differs notably from Tanganyika. In the Nyassa species the caudal fin is truncate or emarginate, and appears to be always nearly completely covered with small scales in the adult fish; this feature, the prevalence of a few distinctive types of acloraion. and the absence of evident relationship to species found elsewhere lead to the conclusion that the Nyassa species are a natural group and may perhaps have evolved in the lake from a single ancestral form. This conciusion is fortified by the study of such skeletons as are available. The differences in the pharyngeal dentition between closely related species are sometimes very striking; the best examples of this are H. tetrastigma and A. alaannler, and H. similis and H. kirkii; these cases make it evident that it is not desirable to regard the development of large, round, blunt pharyngeal teeth as warranting generic separation. Synopsis of the Nyassa Species. I. Snout not longer than postorbital part of head. A. Pectoral fin with series of spots on the rays. Lower jaw projecting .... wecesecseeee * L. Levengstonrs. Jaws equal; depth of body 2 2 to 3in length . ceseteeeee 2. polystigma. Jaws equal; depth 22 in length. titdesscettectseenseseenese | Oo NACULIMaNUS. B. Pectoral fin entnapalane. 1. Preemaxillary pedicels not extending beyond anterior edge of orbits. a. Body with a series of dark spots on back alternating with a series on or above middle of side, or uniting with them to form irregular cross- bars. Jaws equal; outer teeth mostly bicuspid. Pharyngeal teeth small; spots very large . . 4, simulans. Middle ‘pharyngeal teeth enlarged, blunt ; veaudal: peduncle 15 to 1? as long as deep .. . 5. subocularis. Middle pharyngeal | teeth enlarged, plunt ; ae peduncle a little longer than deep ........ beboonsne, Gh OPDBIGDND b. Body with 6 dark cross- sai “Thome: jaw projecting; outer teeth bicuspid. Pharyngeal teeth small .... ... 7%. johnstoni. Lower pharyngeal with a eroup of enlar reed plunt teeth ... 8 sexfasciatus. ce. Body with a dark lateral stripe, slightly curved anteriorly, from head to caudal tin. Outer teeth bicuspid. o#. Lower jaw a little projecting; middle pharyngeal teeth stout and lolunraaeeeeeee shbsachou . el BOR B00 8. Jaws equal; pharyngeal teath eral. Caudal truncate or slightly emarginate .... ca LOS sunales: Caudal rather deeply emarginate; 45 to 60 tecth in outer series ..... Perce alulomeucceps: Caudal rather ‘deeply emarginate ; “94. to 30 teeth in outer SELICS ieee othe weeveeeee.. 12. maicrostoma. d. Body le a iene “inal panel 3 on rea acridine forward as a series of spots. ‘Teeth conical. Lower jaw projecting. 13. wrotenia. e. Body with a straight dark band or stripe from head to caudal fin. Teeth conical. Lower jaw projecting. 2. Mouth little oblique, below level of eye. Caudal truncate or very slightly emarginate .................. 14. fuscoteniatus. Caudal distinctly emarginate .......... lveeeedteciea)) bi holotenia. 3. Mouth very oblique, pao sone ¢ ona level with eye. Depthyot body 2eito,2= invlens th eyes. weresee weed etesseeeet 16. strigatus. Depth of body Ain dength 20.00... 6.0... ..scecececneerstenesecveeee) Le Gumidraus. ” CICHLID FISHES OF LAKE NYASSA. 687 Jf. An opercular spot and a blackish spot on or under lateral line below middle of spinous dorsal ; often a third spot below soft dorsal and a fourth at base of caudal f. . 8 to 13 gill-rakers on lower part of anterior arch (? Al. zntermedius). * Outer teeth bicuspid; lower jaw not projecting. Maxillary extending to below eye . f 18. awritus. Maxillary not extending to below eye; . pharyngeal teeth small . 19. tetrastigma. Maxillary not “extending ‘to below. eye; pharyngeal teeth large, rounded, blintey eke tae: secseueeeee 20. placodon. ** Outer teeth oouetel Jerse ‘ee puteetine (? A. intermedius). Snout a little shorter than postorbital part of head; last dorsal spine nearly + leneth of head . e 21. intermedius. Snout as long as postorbital part of head ; last dorsal spine 3, interorbital width + Jength Os head... Saree 22. modestus. Snout as long as postorbital part of head ; last dorsal spine 3 to 2, interorbital width 2 or i length Olheadeaaeeeeee 23. woodi. 8. 17 to 28 gill- Pate on lower part of anterior arch. 17 to 21 gill-rakers ; 15 or 16 dorsal spines .................. 24. chrysonotus. 23 to 28 ill- vakers : 17 or 18 dorsal spines ....... wen. 25. quadrimaculatus. g. A dark band from nape or origin of car cl to base of caudal (sometimes faint or absent in large specimens). a, Lower pharyngeal with a group of large, rounded, blunt teeth. Jaws equal; caudal peduncle 14 to 1} as long as deep ...... 26. spherodon. Jaws equal; caudal peduncle 13 as long as deep ............ 27. ericotenia. Lower jaw projecting ........ ....... 28. lateristriga. 8. Lower phar eee daw a ihe 2 maddie series a little enlarged. Eye 3 to 4 in head (in specimens of 70 to 110 mm.) ......... 29. plagiotenia. Eye 3 to 33 in head (in specimens of 120 to 170 mm.) ... 30. melanotenia. y. Pharyngeal teeth small; outer teeth of jaws forming a close-set sevies, usually bicuspid im young, some or all conical in adult. * Lower jaw a little shorter than upper... 31. guentheri. ** Jaws equal; 18 or 19 dorsal spines ...... 32. melanonotus. *k* Jaws equal or lower a little projecting; 15 to 17 dorsal spines. + Maxillary extending to vertical from anterior edge of eye. 8 or 9 gill-vakers on lower part of anterior arch ............ 33. brevis. 12 eill-rakers ; 35 or 36 scales; caudal peduncle 15 as long as deep ... Leese aaa M20LOLENU. 11 to 13 gill- rakers ; 37 to 39 ‘scales ; ‘caudal peduncle 13 WONZpasHlOnptasuel ee phar ass Meeks cae adianadin theese seas onli 35. lepturus. ++ Maxillary not extending to below eye. Scales 35 to 37; last dorsal spine} to 3 head ............... 36. rhoadesii. Scales 33 or 34; last dorsal spine quite 2 head ............... 37. atriteniatus. 0. Pharyngeal teeth slender; outer teeth of jaws conical, rather strong, and spaced. * Third anal spine as long as last dorsal; pelvic fins shorter than ea deVeer hats thane ci euebanhea den cs tution wep 38. spilorhynchus. **< Third anal spine shorter than last dorsal; pelvie fins as long as head. Maxillary not quite reaching vertical from anterior aie of eye . ee. 39. longipes. Maxillary not ‘nearly. reaching vertical: from: anteric ior edge need risa : See On ccemulouss e . Coloration Five ery, AG tmes | Ww Hen ree dark ecross-bars. o#. Maxillary extending to below anterior edge of eye. Lower pharyngeal with a group of enlarged teeth with spherical crowns ..... .... 41. macrochir. Posterior teeth of 2 middle series of lower “pharynge al somewhat pnlaveed SNA te Re EE eee ones WALA er gUTrOsSOmeds tI imclude H. modestus herve, but its coloration is unknown; the type, as preserved, is uniformly brownish. 688 MR. C. TATE REGAN ON THE 8. Maxillary not extending to below eye. Lower jaw a little the shorter; 12 or 13 gill-rakers on lower part of anterior arch.. 43. macrophthalmus. Jaws equal anteriorly ; 11 ‘oill- yakers on “Tower part ‘of anterior arch .. BbuemAGhe dnetncaBaresdacsabentonaal | CNN IEDKCOSCIRS, Jaws equal; 15 or 16 gill- yakers. 0... vse. 45. tnornatus. Lower jaw a little the ‘shorter ; 16 or 17 gill: rakers ......... 46. micrentodon. 2. Premaxillary pedicels extending to between middle of orbits. 47. eucinostomus. II. Snout longer than postorbital part of head. A. Lower jaw projecting ; maxillary not extending to below eye. Teeth cuspidate; caudal emarginate ..................-0.0:.... 48. preorbitalis. Teeth conical; caudal truncate ..:....0.slo cesses 49: compressiceps. B. Jaws equal anteriorly ; ee not ae ae to below eye. 15 to 17 gill-rakers on lower part of anterior arch ......... 50. macrorhynchus. 19 or 20 gill-rakers on lower part of anterior arch ......... dl. rostratus. C. Jaws equal anteriorly ; maxillary extending to below eye. 52. macrostoma. 1. HapPLocHRoMis Livinestonit Giinth., 1893. Haplochromis livingstonti (part.) Bouleng. Cat. Afr. Fish. 111 p. 286, fig. 194. Depth of body 22 ‘to 3 in length, length of head 2? to 3. Snout with straight profile, 14 to 2 diameter of eye, which is 43 to 54.in length of head, interorbital width 33 to 4, depth of pree- onbrinl 8 3s to 4. Lower jaw projecting; maxillary not extending to below eye; teeth in 3 or 4 series, outer bicuspid and inner tricuspid in young, all conical in adult, 50 to 60 in outer series of upper jaw. 4 or 5 series of scales on cheek. 11 or 12 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 33 to 39 scales in a longitudinal series, 6 or 7 Ou eee of dorsal to lateral line. Dorsal XVI 10- ld last spine 3 to ? length of head. Anal It 9-10; third spine 3 3 to 2 head. poe al a little shorter than head, vege himne origin otf anal. Caudal truneate or slightly emarginate. Caudal peduncle longer than Weep. Silvery (2) or bluish grey (d ); a dark bar from eye to end of maxillary ; large opercular spot; body marbled with blackish, the spots hee bands constantly nearly as in the figure of the type; pectoral with series of small spots on the rays; dorsal and anal, in male, blackish with pale edge. The type, 120 mm. long, and three specimens of 200 to 230 mm. (Wood). 2. HAPLOCHROMIS POLYSTIGMA, sp.n. (PI. I.) Haplochromis livingstonii (part.) Bouleng. Cat. Afr. Fish. 111. p. 286. Depth of body 2% to 3 in length, lene of head about 3. Snout with straight progies 13 to 24 diameter of eye, which is 4 to 6 in lene of head, interorbital width 31 to 4, depth of pre- orbital 37 2 to 43. Jaws equal anteriorly ; maxillary not extend- ing to below eye; teeth in 3 or 4 series In upper jaw, 2 or 3 in lower, outer bicuspid and inner tricuspid in young, all conical CICHLID FISHES OF LAKE NYASSA. 689 in adult, 50 to 65 in outer series of upper jaw. 4 or 5 series of scales on cheek. 10 to 13 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 32 to 34 scales in a longitudinal series, 6 from origin of dorsal to lateral line. Dorsal XVI 10- une last spine + ‘to 2 2 length of head. Anal IIT 9-10; third spine usually a little "shorter than last dorsal. Pectoral as long as or a little shorter than head, nearly or quite reaching anal. Caudal truneate or slightly emarginate. Caudal peduncle longer than deep. Body with large irregular brown spots tending to run together to form 3 longitudinal bands; head, body, ‘and vertical fins covered with numerous small dark spots ; pectoral with series of spots on the rays. Six specimens (Wood), 120-210 mm. in total length; two others (Rhoades, Cunningéon) also belong to this species. 3. HAPLOCHROMIS MACULIMANUS, Sp. D. Paratilapia modesta (part.) Bouleng. Cat. Afr. Fish. 11. p. 326. Depth of body 22 in length, length of head nearly 3. Snout with straight profile, shorter than postorbital part of head. Diameter of eye 5 in length of head, a little less than preorbital depth, # depth of cheek; interorbital width 4 in head. Jaws equal anteriorly; maxillary not extending to below eye ; teeth conical, in 4 series, about 80 in outer series of upper jaw. 5 series of scales on cheek. 11 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 33 scales in a longi- tudinal series, 7 from origin of dorsal to lateral line. Dorsal XVI Ti Amaliih: 10; seed spine 3 head, a little shorter than last dorsal. Caudal peduncle a litéle longer than deep. Traces of dark cross-bars; pectoral with several transverse series of small dark spots. A single specimen, 190 mm. long (J/oore). 4, HAPLOCHROMIS SIMULANS, Sp. n. (Text-fig. 5.) Depth of body 2§ in the length, length of head 5 to 3l. Snout with straight or eanree profile, 17 to ‘12 diameter of eye, which is 33 to 42 in length of head, eatlsoer bital width 4, depth of pre- orbital 4 to 43. Tee equal anteriorly ; : maxillary not extending to below eye; teeth in 4 to 6 series, outer bicuspid, or posterior ly conical in adults, 60 to 75 in outer series of upper Jaw. 3 or 4 series of scales on cheek. 10 gill-rakers on lower part of anterior arch. .Pharyngeal teeth small. 32 to 34 scales in a longitudinal series, 6 or 7 from origin of dorsal to lateral line. Dorsal XVI 105 duly last spine 3 to more than 2 length of head. Anal III 9; third spine as long as last dorsal. ” Pectoral as long as or a little shorter than head, nearly or quite reaching anal. Caudal tiun- eate or slightly emarginate. Caudal peduncle as long as or a little longer than deep. Silvery or yellowish ; a dark bar below eye; an opercular spot: 4 large dark spots on back alternating 690 MR. C. TATE REGAN ON THE with 4 very large, vertically expanded spots on side, the spots sometimes confluent to form broad irregular cross-bauds ; dorsal and anal dusky with yellow margin, or anal pale yellow; caudal dusky above, yellow below. Four specimens, 130 to 190 mm. in total length (Wood). Text-fi gure 5. Nie Haplochromis simulans. This species bears a great resemblance and is probably closely related to Cyrtocara venusta, but differs in the structure of the spinous dorsal fin. 5. HAPLOCHROMIS sUBOCULARIS Giinth., 1893. IE An Sig Os MOPAILG Tolle hie. aie 18>. Tilapia johnstoni (part.) Bouleng. Cat. Af. Fish. 111. p. 249. Depth of body 3 to 32 in length, length of head 3} to 32. Snout with straight or slightly convex profile, a little longer than diameter of eye, which is 34 to 3? in length of head ; inter- orbital width 32 to 4 in head, depth of preorbital 44. Jaws equal anteriorly ; maxillary not extending to below eye; teeth m 4 or 5 series, outer bicuspid, 40 to 50 in outer series of upper jaw. 3-series of scales on cheek. 10 gill-rakers on lower part of anterior arch. Middle pharyngeal teeth enlarged, rounded, blunt. 32 or 33 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XV 11-12; last spine 2 to 4 length of head. Anal III 8; third spine 3 to 2? head. Pectoral as long as head, nearly or quite reaching anal. Caudal truncate or slightly emarginate. Caudal peduncle 13 to 12 as long as deep. Two dark bars across nape; 4 dark spots at base of dorsal alternating with a series above middle of side; 2 large spots on caudal peduncle; dorsal and caudal with series of spots. Two specimens, 120 and 140 mm. in total length. aN CICHLID FISHES OF LAKE NYASSA. 691 6. HAPLOCHROMIS ORNATUS, sp. n. (Text-fig. 6.) Depth of body equal to length of head, 2? in length of fish. Snout with slightly convex profile, as long as postorbital part of head. Diameter of eye 14 depth of preorbital, greater than depth of cheek, 34 in length of head: interorbital width 4 in length of head. Jaws equal anteriorly; lips thick; maxillary not extending to below eye; teeth in 3 series in upper jaw, 4 in the lower ; 52 bicuspid teeth in outer series of upper jaw. 3 series of scales on cheek. 12 gill-rakers on lower part of anterior arch. Middie pharyngeal teeth enlarged, rounded, blunt. 31 scales in a longitudinal series, 6 from origin of dorsal to Jateral line. Dorsal XVI 10; spines rather strong, last a little more than 2 length of head. Anal III 9; third spine stronger than dorsals, 'Text-figure 6. Haplochromis ornatus. 3. 2 head. Pectoral nearly as long as head, reaching anal. Caudal feebly emarginate. Caudal peduncle a little longer than deep. Silvery; an opercular spot and a spot above eye; 5 dark spots at base of dorsal alternating with a series above middle of side; 2 dark bars across upper 3 of caudal peduncle ; dorsal and caudal with series of spots. A single specimen, 140 mm. in total length (JVood). 7. HAPLOCHROMIS JOHNSTONI Giinth., 1893. Tilapia johnstoni (part.) Bouleng. Cat. Afr, Fish. ui. p. 249, fig. 167. ; Depth of body 22 in length, length of head 3. Snout with straight profile, as long as postorbital part of head. Diameter of eye 4 in length of head, interorbital width 43, depth of pre- orbital 44. Lower jaw projecting; maxillary not extending to below eye; teeth in 3 or 4 series, outer bicuspid, 50 in outer series of upper jaw, 3 series of scales on cheek. 11 gill-rakers 692, MR. C. TATE REGAN ON THE on lower part of anterior arch. Pharyngeal teeth small, com- pressed. 31 scales in a longitudinal series, 5 from origin of dorsal to lateral line. Dorsal XVI 10; last spine a little less than 4 length of head. Anal III 9; third spine stronger than dorsals, nearly 5 2 head. Pectoral + head, reaching origin of anal. Caudal truncate ov shghtly emarginate, Caudal peduncle 13 as long as deep. Silvery; a dark bar from eye to angle of mouth; 6 dark bars on body, the first downwards from nape, the fifth from end of dorsal to behind anal; series of spots on dorsal and caudal. The type, 120 mm. in total length. 8. HaPLOCHROMIS SEXFASCIATUS, sp. n. (Text-fig. 7.) Depth of body 2% to 24 in the length, length of head 3. Upper profile of Bend a little concave ; 3 snout 1A to 2 diameter of eye, which is 4 to 5 in length of Threrudh interorbital width : x to 43, preorbital depth 4 to ree Lower jaw projecting ; eee not Text-figure 7. telo Haplochromis sexfasciatus. extending to below eye ; teeth in 3 to 5 series, 40 to 60 bicuspid teeth in outer series of upper jaw. 3 or 4 series of scales on cheek. 11 or 12 gill-rakers on lower part of anterior arch. Lower pharyngeal with a group of moderately enlarged blunt rounded teeth in the middle posteriorly. 32 scales in a longi- tudinal series, 6 from origin of dorsal to lateral line. Dorsal XV 11; last spine 2 to nec wly 4 Length of head. Anal III 8— 95 third spine stronger than igh dor al, 4 3 to2 head. Pectoral +to % length of head, reaching origin of jameld Caudal truncate or shia ney emarginate. Caudal peduncle 14 as long as deep. Body with 6 blackish cross-bars; dorsal and camel! with series of spots. Two specimens, 100 and 170 mm. in total length (Wood). Lae) CICHLID FISHES OF LAKE NYASSA. 69 9. HAPLOCHROMIS KIRKIT Giinth., 1893. Tilapia kirkii (oar ) Bouleng. Cat. Afr. Fish, iii. p. 251, fig. 169. Depth of body 23 to 22 in length, length of head about 3. Snout with straight ne 1; to if ‘Uaneie: of eye, which is 32 to 43 in length oF head, in adult equal to preorbital depth; hiner arbital width 34 to 44 in length of head. Lower jaw a little projecting ; maxillary extending to between nostril and eye; teeth in 4 or 5 well- -separated series, inner tricuspid, outer bicuspid ; 40 to 45 in outer series of upper jaw. 3 or 4 series of seales on cheek. 11 or 12 gill-rakers on lower part of anterior arch. Middle pharyngeal teeth stout and blunt. 31 to 33 scales in a longitudinal series, 5 or 6 from origin oF dorsal to lateral line. Dorsal XV 10- iM 3 last spine #2 to 3 length of head. Anal [II 8-10; third spine as long as or a ae shorter than last dorsal. Pector A as long as or a little shorter than head, reaching origin of anal. Caudal truncate or shghtly emarginate. Caudal peduncle 13 to 13 as long as deep. A dark stripe from operculum to caudal. Usually a stripe or a series of spots above upper lateral line and another at base of dorsal: ; dorsal and caudal with series of spots; anal with ocelli in males. Seven specimens, 100-160 mm. long (Johnston, Rendall, Wood). Text-figure 8 Haplochromis similis. 2. 10. HaAPLocHRoMIS sIMILIS, sp.n. (Text-fig. 8.) Depth of body 24 to 2! in length, length of head 3 to 3L. Snout with straight or slightly convex profile, as long as or longer than abana of eye, which is 3} to 32 in length of head, oreater than preorbital depth ; “pte Meal Cait 31 to 3% in 694 MR. C. TATE REGAN ON THE length of head. Jaws equal anteriorly ; maxillary ending between nostril and eye; teeth in 4 to 6 series, outer bicuspid, inner tricuspid, 40 to 52 in outer series of upper jaw. 3 or 4 series of scales on cheek. 11 to 13 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 32 or 33 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVI-— XVII 9-10; last spine 2 to 4 length of head. Anal IIT 9-10; third spine a little shorter than last dorsal. Pectoral as long as or a little shorter than head, nearly or quite reaching anal. Caudal truncate or slightly emarginate. Caudal peduncle 1z to 14 aslong as deep. Coloration as in H. kirku. Five specimens, 120-150 mm. long (Wood). 11, HaPLoCcHROMIS BREVICEPS, sp.n. (Text-fig. 9.) Depth of body 3 in the length, length of head 32. Snout a little shorter or longer than diameter of eye, which is 3 to 33 in length of head, greater than depth of preeorbitai or cheek; inter- orbital width-4 in length of head. Jaws equal anteriorly; maxillary not extending to below eye; teeth in 2 or 3 series; Text-figure 9. Haplochromis breviceps. 3. 45 to 60 bicuspid teeth in outer series of upper jaw; 2 series of scales on cheek. 14 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 33 to 35 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVI-X VIL 11; last spine 2 length of head, Anal II] 9; third spine stronger than dorsals, 3 or a little more than + head. Pectoral as long as head, not reaching anal. Caudal emarginate. Caudal peduncle 13 as long as deep. Silvery; a narrow dusky band from opercular spot to base of caudal, another along upper lateral line, a third at base of dorsal. Dorsal and caudal with series of orange spots; male with ocelli in anal. Two specimens, 85 and 135 mm. in total length (Wood). CICHLID FISHES OF LAKE NYASSA. 695 12. HAPLocHROMIS MICROSTOMA, sp.n. (Text-fig. 10.) Tilapia kirkit (part.) Bouleng. Cat. Afr. Fish. i. p. 251. Depth of body 3 in length, length of head 34 to 33. Snout straight, as long as or longer than diameter of eye, which is 33 to 4 in length of. head, a little or considerably greater than pre- orbital depth ; interorbital width 4 in length of head. Mouth small; jaws equal anteriorly; maxillary ending a little behind nostril; teeth in 4 well-separated series, inner tricuspid, outer bicuspid, 24 to 30 in outer series of upper jaw. 3 or 4 series of seales on cheek. 11 gill-rakers on lower part of anterior arch. Text-figure 10. cS) . Haplochromis microstoma. Pharyngeal teeth small. 33 or 34 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVIT 11; last spine 2 to 4 length of head. Anal III 9; third spine a little shorter than last dorsal. Pectoral as long as head, not reaching anal. Caudal deeply emarginate. Caudal peduncle 13 as long as deep. Silvery; a dark lateral band from operculum to base of caudal ; two series of small dark spots, one at base of dorsal, the other above lateral line; dorsal spotted. Two specimens, 80 mm. (Wood) and 115 mm. (Rhoades) in length. 13. HAPLOCHROMIS UROTHNIA, sp.n. (Text-fig. 11.) Depth of body 23 to 3 in length, length of head 24to 3. Snout with straight upper profile, in adult as long as postorbital part of head. Diameter of eye 4 to 42 in length of head, equal to or a little less than preorbital depth, less than depth of cheek ; interorbital width 3? to 4in length of head. Lower jaw a little projecting ; maxillary not quite reaching vertical from anterior edge of eye; teeth conical in 2 or 3 series, 35 to 45 in outer Proc. Zoou. Soc.—1921, No. XLVII. 47 696 MR. C. TATE REGAN ON THE series of upper jaw. 3 or 4 series of scales on cheek. 11 or 12 gill-rakers on lower part of anterior arch. Pharyngeal teeth slender. 33 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVI 9-10; last spine about 1 length of head. Anal IIT 8-9; third spine stronger and shorter than last dorsal. Pectoral + head, nearly or quite reaching anal. Text-figure 11. Haplochromis wrotenia. BP Caudal truncate or feebly emarginate. Caudal peduncle 1+ to 12 as long as deep. A series of about 8 dark spots near dorsal profile, a second above lateral line, a third on middle of side posteriorly confluent to form a band; dorsal with series of spots. Three specimens, 170 to 200 mm. in total length (IVood). 14. HAPLOCHROMIS FUSCOTENIATUS, sp. n. (Text-fig. 12.) Depth of body 22 in length, length of head 23. Head 23 as long as broad. Snout a little concave in front of eye, thence straight, nearly twice as long as diameter of eye, which is 5 in length of head, interorbital width 5, preorbital depth 4. Lower jaw a little projecting; maxillary not quite reaching vertical from anterior edge of eye; teeth conical, in 3 series, 60 in outer series of upper jaw. Cheek: with 4 series of scales, depth % diameter of eye. 12 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 34 scales in a longitudinal series, 6 from origin of dorsal to lateral line. Dorsal XVI 11; spines strongly increasing to last, which is 3 length of head. Anal ITI 10; third spine stronger and a little shorter than last dorsal. Pectoral + head, nearly reaching anal. Caudal truncate. Caudal peduncle 1} as long as deep. Sides yellowish; a dusky band on middle of side, a second backwards from base of pectoral ; a series of dusky spots at base of dorsal, connected with a second Col CICHLID FISHES OF LAKE NYASSA. 697 ’ series which form an interrupted band on and above lateral line. Dorsal with series of spots. Anal with some pale (? red) round spots. A single specimen, 175 mm. in total length (IVood). a } 4 * | Text-figure 12. Haplochromis fuscoteniatus. wie 15. HaPLocCHROMIS HOLOTANTA, sp. n. Paratilapia dimidiata (part.) Bouleng. Cat. Afr. Fish. iii. p. 860, fig. 244. Depth of body 33; in length, length of head 8. Snout de- eurved, shorter than postorbital part of head. Diameter of eye 43 in length of head, equal to interorbital width or preorbital | depth, less than depth of cheek. Lower jaw a little projecting ; maxillary not extending to below eye; teeth mostly conical, in 4 series in upper jaw and 3 in lower, 60 in outer series of upper jaw. 4 series of scales on cheek. 1 gill-rakers on lower part of anterior arch. Pharyngeal teeth slender. 36 scales in a longi- tudinal series, 5 from origin of dorsal to lateral line. Dorsal ROVE 12); last spine = 2 head. Anal III 10; third Spine stronger and nearly as long as last dorsal. Pectoral 2 > length of head. Caudal emarginate. Cundal peduncle 13 as lene as deep. A dark lateral band from eye to base of caudal fin; soft dorsal with series of spots. A single specimen (J/oore), 190 mm. in total length. 16. HAPLOCHROMIS STRIGATUS, Sp. n. (Text-fig. 13.) Depth of body 22 to 2£in length, length of head 24 to 3. Snout with straight or slightly « concave p1' ofile, as long as postor bital part of head. Diameter of eye 4 to 4} in length of head, equal to depth of preorbital or interorbital width, less than depth of cheek. aT* 698 MR. C. TATE REGAN ON THE Mouth oblique; lower jaw projecting; maxillary not extending to below eye; teeth conical, triserial, 55 to 60 in outer series of upper jaw. 3 or 4 series of seales on cheek. 12 or 13 gill- rakers on lower part of anterior arch. Pharyngeal teeth small. 33 or 34 scales in a longitudinal series, 6 from origin of dorsal to lateral line. Dorsal XV-XVI 11; last spine 2, longest soft rays 2 to ? length of head. Anal III 10-11; third spine stronger and shorter than last dorsal. Pectoral nearly as long as head, reaching anal. Caudal truncate. Caudal peduncle 13 as long as deep. Silvery; a blackish stripe, half as broad as a scale, from operculum to base of caudal. Dorsal with series of orange spots; other fins yellowish (@ ) or pelvics and anal blackish with red margin (< ), Three specimens, 170 to 185 mm. in total length (Wood). Text-figure 13. Haplochromis strigatus. ¥. 17. HapPLocHROMIS DIMIDIATUS Giinth., 1864. Depth of body 4 in length, length of head 31. Snout with straight upper profile, a little shorter than postorbital part of head. Diameter of eye 5 in length of head, equal to preorbital depth, slightly exceeding interorbital width. Mouth oblique; lower jaw projecting; maxillary ending far in front of eye; teeth conical, in 3 series in upper jaw, 2 in lower; 44 in outer series of upper jaw; anterior teeth of second series in upper jaw enlarged. 3 or 4 series of scales on cheek. 11 gill- rakers on lower part of anterior arch. Pharyngeal teeth small. 35 scales in a longitudinal series, 5 from origin of dorsal to lateral line. Dorsal XVI 11; last spine 3 length of head. Anal III J1; third spine as long as and stronger than last dorsal. Pectoral 2 length of head. Caudal slightly emarginate. ese ee CICHLID FISHES OF LAKE NYASSA. 699 Caudal peduncle twice as long as deep. Silvery; an opercular spot; a blackish lateral stripe, half as broad as a scale, ending in a spot on base of caudal fin. A specimen of 165 mm. (Wood), described above, has been compared with the type, a skin 220 mm. long. 18. Hapnociromis Auritus, sp.n. (Text-fig. 14.) Depth of body 24 in length, length of head 3. Snout convex, a little shorter than diameter of eye, which is twice preorbital depth, 3 in length of head ; interorbital width 43 in head. Jaws equal anteriorly ; maxillary extending to below eye; teeth in 2 series; 66 bicuspid teeth in outer series of upper jaw. 3 series of scales on cheek. 13 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 32 scales in a longi- tudinal series, 5 from origin of dorsal to lateral line. Dorsal Text-figure 14. SS ee a < Haplochromis auritus. Nat. size. XV 10; last spine a little less than 3 length of head. Anal IIT 9; third spine stronger than and nearly as long as last dorsal. Pectoral as long as head, reaching anal. Caudal rather deeply emarginate. Caudal peduncle a little longer than deep. Silvery ; a blackish opercular spot; a dark spot on lateral line below spinous dorsal; series of spots on soft dorsal and caudal. A single specimen, 80 mm. in total length (Wood). 19. HAPLOCHROMIS TETRASTIGMA Ginth., 1893. Tilapia tetrastigma Bouleng. Cat. Afr. Fish. 111, p, 250, fig. 168. Depth of body 24 to 24 in length, length of head about 3. Snout with straight or convex. profile, nearly as long as post- orbital part of head. Diameter of eye 34 in length of head, interorbital width 4, preorbital depth 4. Jaws equal anteriorly; maxillary not extending to below eye; teeth in 3 or 4 series, outer 700 MR. C. TATE REGAN ON THE bicuspid, about 60 in outer series of upper jaw. 3 or 4 series of scales on cheek. 10 or 11 gill-rakers on lower part of anterior arch. Lower pharyngeal a small weak plate bearing small compressed bicuspid teeth. 30 to 32 scales in a longitudinal series, 5 from origin of dorsal to lateral line. Dorsal XIV— XV 10-11; last spine 2 to 4 length of head. Anal III 8-9; third spine a little shorter than last dorsal. Pectoral as long as or a little shorter than head, reaching origin of anal or a little beyond. Caudal slightly emarginate. Caudal peduncle 13 to 1d as long as deep. Silvery; four blackish spots, the first on operculum, the second on upper lateral line below spinous dorsal, the third on side below soft dorsal, the fourth. at base of caudal. Three of the types, 105 to 110 mm. in total length. 20. HAPLOCHROMIS PLACODON, sp.n. (Text-fig. 15.) Depth of body 23 to 25 in length, length of head about 3. Snout with straight or convex profile, shorter than postorbital part of head. Diameter of eye equal to or greater than depth Text-figure 15. lies Haplochromis placodon. of preorbital, 83 to 4} in length of head, interorbital width 32 to 33. Jaws equal anteriorly, lower 2 length of head; maxillary not extending to below eye; teeth in 4 or 5 series, outer bicuspid, 50 to 55 in outer series of upper jaw. 3 or 4 series of seales on cheek, 8 or 9 gill-rakers on lower part of anterior arch. Lower pharyngeal a large and strong plate bearing large rounded flat teeth. 31 to 33 scales in a longitudinal series, + or 5 from origin of dorsal to lateral line. Dorsa! XVI 10-11; —— CICHLID FISHES OF LAKE NYASSA. 701 last spine 2 to 2 length of head. Anal III 8-9; third spine } to 2 head. Pectoral as long as head, extending nearly to middle of anal. Caudal emarginate. Caudal peduncle 12 to 13 as long as deep. Silvery; a blackish spot on operculum; two large blackish spots on upper lateral line, expanding upwards towards spinous and soft dorsal; a blackish spot at base of caudal. Dorsal and caudal with series of orange spots; males with a pale edge to dorsal and lower fins dusky, the anal with ocelli. Five specimens, 130 to 180 mm. in total length (Wood). 21, HaPLocHROMIS INTERMEDIUS Ginth., 1864. Paratilapia intermedia (part.) Bouleng. Cat. Afr. Fish. iu. p- 363. Depth of body 24 in length, length of head 3. Snout with straight profile, shorter than postorbital part of head. Diameter of orbit nearly equal to depth of preorbital, which is 43 in length of head. Jaws equal anteriorly or lower perhaps slightly projecting; lower jaw # length of head; premaxillary pedicels not reaching orbits; maxillary not extending to below eye; teeth small, conical, in 2 or 3 series, about 70 in outer series of upper jaw. 4 series of scales on cheek. 34 scales in a longi- tudinal series, 5 from origin of dorsal to lateral line. Dorsal XVI 11; spines strongly Increasing to last, which is nearly 3 length of head. Anal III 9; last spine 4+ head. Pectoral as long as head, reaching anal. Caudal rather deeply emarginate. Caudal peduncle a little longer than deep. Silvery; traces of two dark spots on upper lateral line; dorsal with series of spots and a pale edge ; anal with several spots. The type, a skin, 185 mm. long. Boulenger’s figure (Cat. Afr. Fish. iii. fig. 247) represents the body and fins correctly, but the head is inaccurate, the preorbital being too narrow; also the four black spots depicted are taken from a specimen of H. quadrimaculatus. H. intermedius is well distinguished from H. quadrimaculatus by the deeper preorbital, deeper cheek with 4 series of scales, shorter premaxillary pedicels, the number of dorsal spines (16 instead of 17 or 18), and the form of the spinous dorsal fin. It shows much closer agreement with H. placcdon, but has the teeth in fewer series, smaller, and conical instead of cuspidate ; also the lower jaw is notably longer. 22. HApLOcHROMIS MoDESTUS Giinth., 1893. Paratilapia modesta (part.) Bouleng..Cat. Afr. Fish. iii. p. 326, fig. 219. Depth of body 3 in length, length of head 23. Snout with straight profile, as long as postorbital part of head. Diameter of eye 44 in length of head, a little greater than przorbital depth, equal to depth of cheek; interorbital width 5 im length: 702 MR. C. TATE REGAN ON THE of head. Lower jaw projecting; end of maxillary not far short of vertical from anterior edge of eye; teeth conical, in 3 series, 70 in outer series of upper jaw. 4 series of scales on cheek. 12 gill-rakers on lower part of anterior arch. Pharyn- geal teeth slender. 32 scales in a longitudinal series, 5 from origin of dorsal to laterai line. Dorsal XVI 11; last spine 3 length of head. Anal III 10; third spine stronger and slightly shorter than last dorsal. Pectoral? Caudal ? (perhaps truncate). Caudal peduncle as long as deep. Colour ? A single specimen, 150 mm. in total length. 23. HapPLocHROMIS Woopt, sp.n. (PI. IT.) Paratilapia modesta (part.) Bouleng. Cat. Afr. Fish. 1i1. p. 326. Depth of body 23 to 3 in length, length of head 23 to 23. Snout with straight profile, as long as postorbital part of head. Diameter of eye 34 to 4 in length of head, equal to or greater than depth of preorbital or cheek ; interorbital width 6 to 7 in head. Lower jaw projecting ; maxillary not extending to below eye; teeth conical, in 3 or 4 series, 55 to 70 in outer series of upper Jaw. 3 or 4 series of scales on cheek. 11 or 12 gill- rakers on lower part of anterior arch. Pharyngeal teeth slender. 32 or 33 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XV-—XVI 9-10; last spine + to 2, longest soft rays 4 to # length of head. Anal III 9-10; third spine stronger than and about as long as last dorsal. Pectoral a little shorter than head, reaching anal. Caudal truncate or slightly emarginate, angles sometimes rounded. Caudal peduncle as long as or a little longer than deep. Silvery; about 10 faint dark cross-bars; an opercular spot; a series of small dark spots near dorsal profile; a large dark spot on side below spinous dorsal, a second below soft dorsal, a third at base of caudal. Dorsal and caudal with series of orange spots; in males dorsal with blackish intramarginal band and orange edge; pelvics and anal blackish, latter with orange spots. Six specimens, 130 to 210 mm. long (Wood); two in poor condition (fendall) also belong to this species. 24, HAPLOCHROMIS CHRYSONOTUS Bouleng., 1908. Paratilapia chrysonota Bouleng. Cat. Afr. Fish. il. p. 362, fig. 246. Depth of body 23 to 3 in length, length of head 3 to 33. Snout with straight profile, as long as or a little shorter than diameter of eye, which is about twice depth of przeorbital, 22 to 3} in length of head, interorbital width about 34. Jaws equal anteriorly ; premaxillary pedicels extending to between anterior edges of orbits, } length of head; maxillary not extending to below eye; teeth small, conical, in 3 or 4 series. 2 or 3 series of scales on cheek. 18 to 21 gill-rakers on lower part of anterior CICHLID FISHES OF LAKE NYASSA. 793 arch, Pharyngeal teeth small. 31 to 34 scales in a longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal X V— XVI 9-11; last spine 2 to 4 length of head. Anal III 9-11; third spine about 2 head. Pectoral as long as head, reaching anal. Caudal scaly, slightly emarginate. Caudal peduncle as long as or a little longer than deep. Silvery with four blackish spots. one on operculum, the second on and under lateral line below spinous dorsal, the third below soft dorsal, the fourth at base of caudal; males usually darker, often with yellow kack. Several specimens, 100 to 130 mm. long. 2). HAPLOCHROMIS QUADRIMACULATUS, Sp. n. Paratilapia intermedia (part.) Bouleng. Cat. Afr, Fish. 111. p- 363. Depth of body 21 to 2% in length, length of head 3 to 33. Snout with straight profile, shorter ham postorbital part of head. Diameter of eye 4 to 42% in length of head, depth of preeorbital 5 to 54, interorbital width 33. Jaws equal or lower slightly projecting ; Peery pedicels extending to between anterior edges of orbits, 2 2 length of head; maxillary not extending to below eye; teeth sei conical, in 2 or 3 series. 2 or 3 series of scales on cheek. 24 to 27 gill-rakers on lower part of anterior arch. Pharyngeal teeth small. 34 to 36 scales ina longitudinal series, 5 or 6 from origin of dorsal to lateral line. Dorsal XVII-X VITE 11-12; spines canes or slightly increasing from seventh or gil, last 2 to 4 length of head. Anal IIT 10-12; third spine 1 to 2 head. Pectoral as long as or shorter than head, not or barely reaching anal. Caudal scaly, emarginate. Caudal peduncle a little longer than deep. A blackish spot on operculum, a second under lateral line below spinous dorsal, usually a third between lateral lines and a fourth at base of caudal. Seven specimens (Rhoades, Whyte), 170 to 200 mm. long. 26. HapLOCcHROMIS SPH@RODON, sp.n. (Text-fig. 16.) 953. Tilapia later “istriga (part.) Bouleng. Cat. Afr. Fish. iti. p. SFA! Depth of body 25 to 3 in length, length of head 38 to 33. Snout with straight profile, as long as or a little shorter than diameter of eye, which is 3 to 34 in length of head, interorbital width 4, preorbital depth 5. Jaws equal anteriorly; maxillary not extending to below eye; teeth in 4 to 6 series, 40 to 60 bicuspid teeth in outer series of upper jaw. 2 to 4 series of scales on cheek. 9 or 10 gill-rakers on lower part of anterior arch. Middle pharyngeal teeth large, with spherical crowns. 31 or 32 scales in a longitudinal series, 5 from origin of dorsal to lateral line. Dorsal XV-XVI 10-12; last spine about 2 length of head. Anal III 8-9; third spine usually a little shorter than last dorsal. Pectoral about as long as head, reaching _— 704 MR. C. TATE REGAN ON THE anal. Caudal slightly emarginate. Caudal peduncle 1+ to 13 as long as deep. An opercular spot; a blackish band from nape to caudal. Five specimens, 80 to 125 mm. long (Wood, Rhoades), and a skeleton. ‘ Text-figure 16. Haplochromis spherodon. ol ° 27. HAPLOCHROMIS ERICOTHENIA, sp.n. (Text-fig. 17.) Depth of body 31 to 34 in length, length of head 3 to 33. Snout with straight profile, nearly as long as or shorter than diameter of eye, which is 2? to 3 in length of head, interorbital Text-figure 17. Haplochromis ericotenia. Nat. size. width 41, preorbital depth 54 to 6. Jaws equal anteriorly ; maxillary not extending to below eye; teeth in 4 series, 35 bicuspid teeth in outer series of upper jaw. 3 series of scales on cheek. 10 gill-rakers on lower part of anterior arch. Middle posterior pharyngeal teeth enlarged and obtuse, probably spherical —_ —— CICHLID FISHES OF LAKE NYASSA. 705 in adult. 31 scales in a longitudinal series, 7 from origin of dorsal to lateral line. Dorsal XV-XVI11; last spine 2 or a little more than 2 length of head. Anal IIT 9-10; third spine as long as last dorsal. Pectoral + head, not reaching anal. Caudal slightly emarginate. Caudal peduncle 13 as long as deep. Silvery ; an opercular spot; 8 faint dark ecross-bars on body; a blackish band, more or less broken up into a series of spots on the bars, from nape to middle of base of caudal. Two specimens, 65 and 72 mm. in total length (Wood). 28. HapLOCHROMIS LATERISTRIGA Giinth., 1864, (Text-fig. 18.) Chromis lateristriga (part.) Giinth. P. Z.8. 1864, p. 312. Tilapia lateristriga (part.) Bouleng. Cat. Afr. Fish. iii. p. 253. Tilapia lethrinus (part.) Bouleng. t. c. p. 254. Depth of body 2? to 3 in length, length of head 23. Snout with straight upper profile, as long as postorbital part of head. Diameter of eye 4 to 5 in length of head, interorbital width 34 to 43, depth of preorbital 33 to 4. Lower jaw a little projecting ; Text-figure 18. LUN WR) ZO NINN: Haplochromis lateristriga. ¥- maxillary exposed distally, ending below nostril; teeth in 3 or 4 series, inner small; 40 to 55 bicuspid teeth in outer series of upper jaw. Cheek with 3 to 5 series of scales. 153 gill-rakers on lower part of anterior arch. ) ry h » A (y : a. OI Os S se c Gi © © C) | a © c (0) ( WEE) SSD 4. J. The vallate papille of the Lemuroidea: A-D. Lemur fulvus; EH. L. rufifrons ; H-H. L. macaco; 1. L. mongoz; Ja-Jce. plan and elevation of papillee of L. rubriventer ; Jd. fungiform papille of Lemar. The conical papille (text-fig. 68 A) have the usual arrange- ments in clusters and rows, and their points run in the usual directions. Those on the oral part of the dorsum are small, but those on the pharyngeal part are big and cylindrical] or scale-like. These two groups are separated by a line of demarcation concave forwards, lying a little in front of the vallate papillary area. The points of the basal papiile are straight or recurved. This arrangement is also present in Hapalemur, according to Beddard (1), and Chiromys, but is absent in Microcebus, the Loriside, Galagide, and Z'arsius. The lateral organs (text-fig. 68 B-E) consist of lamineze and sulci on the dorsum alone (L. macaco) or on the lateral borders and inferior surface. The central lamine are larger than the lateral ones, and the inner border of each organ is convex towards _ OF THE TONGUES OF THE MAMMALIA. 745 the lateral vallate papille as in the Simiide. The relative position of these papille to the organs varies in different species, lying level with the central or posterior lamine. In Microcebus the lateral organs are absent, and this condition, together with the number and arrangement of the vallate papillee and state of the conical papille, links the tongue to those of the Loriside and Galagide, and distinguishes it from those of Zemw. The numbers of laminz and sulci frequently differ in the two organs of the same tongue, and the secondary sulci may be well marked or indistinct. Text-figure 68. /s ( A @ NN @ AT tana alin, Seo Se? A. Conical papille of Lemur; B-E. lateral organs of ZL. eatta (B), L. fulvus (C), I. macaco (D), L. mongoz (E); F. cross sections of the tongue of Chiromys in the anterior (a) and middle (b) thirds: (s. sublingua, 1.1. lytta uf tongue, ls. lytta of sublingua). The ventral papillary zone is narrow and has many conical but few fungiform papille, and the ventral mesial sulcus may contain a mesial] crest; it is absent altogether in some examples of each species, but present in others. The sublingua (text-fig. 69) is triangular or lyrate, and the apex is divided into a small group of slender denticles. The edges are serrated in L, catia, L. macaco, and L. varius, and entire in L. fulvus and L. fulvus rufifrons. The median crest is well developed, but the lateral ventral crests vary in size. Between the base of the sublingua and the frenal lamelle is a deep sulcus which is sometimes V-shaped. The frenum consists of two lax portions separated by the sub- lingua. The first extends from the floor of the mouth to the under surface of the sublingua, and the second runs from the upper surface of the sublingua to the mesial ventral sulcus of 5()* 50 146 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY the tongue. The frenal lamelle, which are sometimes complex, have the following characters :-— 1. LZ. catta: lamelle are two blunt rounded processes. 2. L. coronatus: \amelle form a parallel-sided flap. 3. L. macaco: triangular with rounded apex and lobulated sides. A. L, varius: two broad flaps with rounded anterior borders and sharply cut mesial borders. 5. L. mongoz: lamelle form a parallel-sided flap. The tongues of the species of Lemur resemble those of the Chimpanzee and Gibbons in the following respects :— 1. Several vallate papillae form a Y. 2, The conical papille on the pharyngeal part of the tongue are much larger than those on the oral part. 3. The lateral organs are situated at the posterior parts of the lateral borders and are convex towards the lateral vailate papille. The tongues differ from those of the Orang-Outan and Man, which have no large conical papille on the pharyngeal part of the tongtie, and the vallate papille form aV. ‘These also have the Apical Gland of Nuhn. They differ from those of the Gorilla and Orang-Outan, whose lateral organs form ladder-like patterns on the dorsum. Genus Microcrsus. Smirn’s Dwarr Lemur (JZ. murinus). Three vallate papille form a triangle with the base anterior ; the basal papille are small and circular, but the apical papilla is large, round, and granular. The conical papille imerease in size from the apex of the tongue to the epiglottis, and those on the base are not dispro- portionately large as in Lemur and Hapalemur. Lateral organs are absent. The sublingua is a flat horny plate with a row of denticles on its anterior border. One can see, therefore, that there is no resemblance in any of the essential points between the tongues of Lemur and Micro- cebus, but the latter resemble those of some of the Lorises, Galagos, and Pottos, described below, Family INDRISID&. Genus [yprRis. Flower (4) described the tongue of J. brevicaudatus as follows:—“The tongue is long and narrow, with an obtuse point at the apex, being less truncated than in Lemur. It is had, OF THE TONGUES OF THE MAMMALIA. TAG 24 inches long, and 3 inch broad at its widest point. The dorsal surface is covered with a very close-set velvety pile of small papille, apparently all of one kind. There are two small circum- vallate papille, 2; inch apart, just in front of the attachments of the palato-glossal folds.” The sublingue resemble those of the Lemuride. Text-figure 69. I. ue K. L, The sublingua of Lemur catta (A.B), L. fulvus (C), L. rufifrons (D), L. macaco (HE); Nycticebus tardigradus (E.G); Loris gracilis (H); Perodicticus potto (1); Hemigalago demidoffi (J); Galago crassicaudata (K); Tarsius spec- trum (L). Family Lorisip &. yenus NYCTICEBUS. THE Stow Loris (WV. tardigradus). Three preserved specimens were examined, and their lateral borders were seen to have no lateral organs. The apex, sulci, and bare area in front of the epiglottis were similar to those of Lemur. Three vallate papille (text-fig. 66 B) form an equilateral tri- angle with the base behind. All are circular, prominent, and 748 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY finely granular; the fosse are well marked and the vallums appear as clear zones. They are conical on elevation, the bases of the cones projecting beyond the fosse. In two specimens the papillae are of equal size, but in the third (Mus. R.C.S8.) the apical papilla is smaller than the basal ones. The fungiform papille are not numerous, and form a dorsal bounding zone on which they have the usual arrangement in rows aud clusters, but the apical cluster is small; and they are few in number on the ventral papillary zone. The conical papille have the usval arrangement in clusters and rows ; they increase in size from apex to epiglottis, and there is no sharp transition between small conical papille on the oral and large papille on the pharyngeal part of the dorsum as in Lemur and Hapalemur. Lateral organs are absent. The sublingua differed in two of my specimens. In the first it is tongue-shaped with a broad base whose angles are rounded, and its area is half that of the inferior surface of the tongue. The denticles on the centre of the anterior border are large, but the more lateral ones are smaller. The central part of the sub- lingua is thick and strong, but the lateral parts are thin. On the ventral surface there is only one strong median ridge, and the upper surface has a strong ridge which is received into the median ventral suleus of the tongue. The frenum is long and lax, so the sublingua is freely movable. In my second specimen the sublingua has no rounded basal angles, the apical denticles are more closely packed, there is only a faint ridge on the upper surface, and the whole organ is more fixed to the under surface of the tongue, The frenal lamelle of the first specimen are broad, with bluntly rounded apices, but they are narrow and pointed in the second example. Supplementary lamellz were present in Pocock’s spe- cimen (15). The median ventral sulcus has no crest, but receives the ridge on the upper surface of the sublingua. Genus Loris. Tur SuenDER Loris (Loris gracilis). The conical tongue narrows rapidly from base to apex, and its proportions are small. Three vallate papille form an isosceles triangle with the apex behind. All are circular, furrowed, and granular, and are sur- rounded by prominent vallums. None are overlapped by conical papille. The fungiform papille differ from those of Vycticebus tardi- gradus in that they extend right across the dorsum. They have the usual arrangement in rows and clusters, but the transverse OF THE TONGUES OF THE MAMMALIA. 749 rows extend far forwards at the expense of the apical cluster. Those at the posterior part of the oral division of the dorsum are large, hemispherical, polished, and close together. Few are pre- sent on the inferior surface of the tongue. The conical papille resemble those of Vycticebus tardigradus in their mode of enlargement from before backwards, and differ from those of Lema and Hapalemur. Most of the papille are cylindrical, and are surrounded by zones of interpapillary dorsum. Lateral organs are absent. The following account of the sublingua (text-fig. 69) 1s supple- mentary to those of Tiedemann (18), Otto, Vrolik (16), and Gegenbaur (5). It is leaf-shaped and w rapped round the convex inferior surface of the tongue, so 1t 1s concavo-convex on cross section. The edges are finely crenated and the apex is sharp. The denticles are smaller, less numerous, and more closely packed than those of WVycticebus tardigradus, but the characters of the central and lateral parts are similar in both cases. The mesial dorsal crest is not so pronounced as that of some examples of WVycticebus, but is equal to that of others. Both species have equally mobile sublingue, but the freedom is less than that of Lemur. The frenal lamelle are two small, rounded processes whose edges extend postero-laterally for a considerable distance. No accessory lamelle are present. The median ventral sulcus resembles that of Vycticebus tardi- gradus. Genus PERODICTICUS. BosmAn’s Porro (Perodicticus potto). The tengue, which narrows slightly from base to apex, has no apical notch and no mesial dorsal sulcus, but possesses a few small, narrow transverse sulci in the posterior part of the oral division of the dorsum. The circumvallate papille.—Van der Hoeven (6) described and figured three large papille in a triangle with the apex behind. In the specimen in the Museum of the Royal College of Surgeons there are three large excavated papille forming a triangle. In my specimen there are three in a triangle, but the apical papilla is divided into two large elongated parts enclosed in the one fossa. The fungiform and conical papilla do not differ materially from those of Vycticebus tardigradus. Lateral organs are absent. The sublingua is tongue-shaped and has nine apical denticles as in P, ibeanus. Its strong median ventral crest is bifurcated posteriorly, and the dorsal ridge is larger and sharper than that of any other species which I examined. The edges are devoid of 750 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY serrations and the degree of mobility is equal to that which exists in Vycticebus and Loris (text-fig. 69), Nussbaum (11) pointed out that the sublingua has a rod-like core, which he termed the ‘“lyssa of the sublingua.”” It has a connective-tissue sheath and consists of fat, cartilage, and connective tissue. It is connected above to the lyssa of the tongue which contains fat, connective tissue, and muscle fibres. The frenal lamelle are two conical processes whose crenated edges extend postero-laterally almost as far as the palato-glossal folds. The median ventral suleus has no crest, but receives the prominent median dorsal ridge on the sublingua. Family GALAGID&. Genus GALAGO. THe BusHy-TAILED GaALaco (G. crassicaudata). Mus. R.C.S. The tongue was preserved, so no measurements were made. Three vallate papille are arranged in the form of an isosceles triangle with long sides. They are large, oval, and excavated, the fosse are patulous, and the vallums appear as clear zones. The basal pair are opposite the attachments of the palato-glossal folds. Flower (4) also recorded three papille in the form of a triangle. ; The fungiform papille are small, rounded, and arranged in the usual manner. The conical papille increase in size from before-backwards, and there is no sharp line of contrast between those on the oral and those on the pharyngeal parts of the tongue as in Lemur and Hapalemur. Flower states that they are thick on the base of the tongue. They have denticulated points directed backwards. The lateral organs are absent, the mesial ventral sulcus has no crest, and the ventral papillary zone has no fungiform papille. The sublingua (text-fig. 69), which is flat, horny, and tongue- shaped, has a rounded anterior border bearing sixteen squat sharp-pointed denticles. It covers rather more than half of the width of the under surface of the tongue and half of the length of the free part between the apex and frenum. ‘The sides are parallel and. entire. Flower’s specimen had upturned edges and a downwardly-directed apex which adapted its form to that of the under surface of the tongue. Genus HEmIcaLaco. Demiporr’s Hremicataco (H. demadoffi). The small conical tongues of both my specimens have three vallate papille arranged in the form of a triangle with the apex OF THE TONGUES OF THE MAMMALIA. (oil directed backwards ; the basai papille are small and circular, and the apical papilla is large, cireular, and granular, with a patulous fossa. The fungiform papille are not numerous, but stand up prominently. They have the usual arrangement, but the apical cluster is small. They form a well-marked row of closely-set elements on the posterior third of each lateral border of the tongue. None are present on the inferior surface of the tongue. The conical papille give the surface of the tongue a finely- granular appearance. ‘They increase in size trom before back- wards, and there is no sharp contrast between the papille on the oral and pharyngeal parts of the tongue as in Lemur and Hapalemur. When the tongue is dried they stand up promi- nently and reach a higher level than the summits of the fungiform papille. They form a narrow ventral papillary zone. Lateral organs are absent. The sublingua (text-fig. 69), which is flat, horny, and tongue- shaped, has a rounded, denticulated, comb-like anterior border, and nearly parallel, entire edges. It has a well-marked median crest, but no lateral ridges, and it covers a larger area of the under surface of the tongue than that of the Galago. The frenal lamelle are small and pointed, and the mesial ventral suleus does not contain a crest. Pocock (15) described nine apical denticles on the sublingua of Galago monteire. Family CurroMYID&. Genus CHIROMYs. Tur Ayr-Ave (C. madagascariensis). The following description is supplementary to the existing accounts by Owen (18) and Peters (14) :— Measurements.— Total length 4:6 cm.; oral part 3:1 cm.; pharyngeal part 5 em.; width between the lingual attachments of the palato-glossal folds 1:6 em. The apex is thick and rounded, but has no notch, and the massive lateral borders are devoid of lateral organs. Mesial dorsal and mesial ventral sulci ave absent. Three circumvallate papille are arranged in the form of a triangle with the apex behind. All are circular and granular, and surrounded by nodulated vallums. Owen described two papille, about two lines apart, lying about an inch and a half behind the apex of the tongue. Peters recorded three papille arranged in a triangle, and figured nodulated vallums. The fungiform papille are not numerous. They extend right across the dorsum, but are scanty in the middle line. They have the usual arrangement, but the apical dorsal cluster is small. One large papilla bisects the base of the vallate papillary triangle. 152 DR. GC. F. SONNTAG ON THE COMPARATIVE ANATOMY None are present on the inferior surface of the tongue. They are smooth or granular, and are hemispherical or pedunculated. The conical papille have the usual arrangement. They are closely packed and exhibit considerable mutual compression. Some are low and cylindrical, but others stand up prominently. The papille on the oral part of the tongue are smaller than those on the base, and a line of demarcation, concave forwards, separates the two groups. The tongue resembles those of Lemur and Hapalemur in this respect, and differs from those of the Lorises, Pottos, and Galages. There is also an area devoid of papille in Paierh of the epidlottis ; this is bisected by the median glosso-epiglottic fold. The sublingua has been described by Owen (13), Gegenbaur (5), and Pocock (15), but several details are omitted, or not sufficiently emphasised, in their accounts. It is a flat horny plate with entire lateral margins and rounded anterior border with a central projection, whose point marks the place where the strong, denticulated median ventral ridge bends downwards in a hook. This crest increases in depth from behind-forwards, and its hook hes 9 mm. posterior to the apex of the tongue. It was shown in the descriptions of the sublingue of the Lorises that there is a thick central strip and two thin lateral parts. The same parts are present in Chiromys, but differ from those of the Lorises in that the central part is adherent to the under surface of the tongue, but the lateral parts are free; in the Lorises, on the other hand, both parts are free. Gegenbaur and Pocock both describe a free lateral margin, but do not give any idea of the extent of the free part. Jn my specimen the total width of the sublingua is 1-2 em.; the central firm adherent part is *6 em. wide, and each free lateral part is °3 em. wide. A probe can be passed for a considerable distance between the under surface of the tongue and the sublingua. No denticles project from the anterior border. Pocock (15) deseribed the two small glandular (?) pockets on each side of the frenum, and pointed out that the frenal lamelle are narrow. Gegenbaur (5) described the horny nature of the sublingua, but did not show how the thickness varies in different parts. Although one thinks that the central parts are thicker than the lateral parts, when the tongue is entire, one sees that the reverse is the case when sections ave made. The apparent thickness of the central part is due to a downward projection of the lingual muscles to which the central part of the sublingua is closely applied (text-fig. 68 F). The Lytta —The tongue possesses a strong median ridge on the under surface of the sublingua, which Owen termed the “lytta.” He did not describe sections of the tongue, for a well- developed lytta is present close to the sublingua. Gegenbaur made the same omission, but Nussbaum (11) recorded its OF THE TONGUES OF THE MAMMALIA. 753 presence. It is united to the upper surface of the sublingua in the median part of the latter. ‘lhe condition resembles that of Perodicticus, in which there are lingual and sublingual lytte. In the latter species both lytte are internal. The position of the lytta in different parts of the tongue is seen in text-fig. 68 F. Family TaRrsiip 4. Genus TARSIUS. Tue Tarsier (7’. spectrum). The classical work of Burmeister (2) and the subsequent papers by Gegenbaur (5) and Pocock (15) have dealt with most points, so the. work of any observer is now comparatively limited. In the specimen in the Museum of the Royal College of Surgeons the shape of the tongue and the characters of the vallate, fungiform, and conical papille are similar to those described by Burmeister. That author does not mention that the lateral organs are absent. There is no great contrast between the conical papille on the oral part of the tongue and those on the base. The sublingua is soft and defined from the under surface of the tongue by a groove only ; it is the most adherent sublingua among the Prosimie. The median ventral rod does not termi- nate in a knob as in Pocock’s specimen, but its free anterior extremity is rounded (text-fig. 69), and is slightly turned downwards as in Chiromys; as the specimen was preserved, I am unable to say whether this curving of the ridge was produced by the alcohol. The frenal lamella are two conical tapering processes, and differ from the conditions recorded by Burmeister and Pocock. Summary and Conclusions. 1. The tongues of the species of Lemur resemble those of the Chimpanzee and Gibbons in the Y-formation of their vallate papille, their convex lateral organs, and the contrast between the large conical papille on the pharyngeal part of the tongue and call ones on the oral part. They differ from the tongues of Man and the Orang-Outan in that the latter have no large conical papille on the base of the tongue, and their vallate papille form aV. They differ from the tongues of the Gorilla and Orang-Outan, whose lateral organs form ladder. like patterns on ale dorsum of the tongue. The tongue of JLicr Daehue differs in the essential features a those of the species of Lemur, and resembles those of the Lorises, Galagos, and Pottos. 3. The tongues of the Loriside and Galagide differ from those of Lemur in so many features that they are important for “purposes of classification. Adding them to other characters 754 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY deseribed by Pocock, Forsyth Major, and others, we obtain the following comparison :— Lemur. Loriside and Galagide. i 1. Vallate papille form a Y. . Vallate papille form a tri- angle. 2. Lateral organs convex. 2. Lateral organs absent. Conical papille small on the 3. Conical papillae increase oral part of the tongue gradually from the apex and large on the base. of the tongue to the epi- glottis. 4, Sublingua triangular or 4. Sublingua tongue-shaped, lyrate, with apical den- with. denticles along its ticles. anterior border. 5. Urethra opens above tip of 5. Clitoris traversed by urethra. clitoris. 6. Ectotympanic annular or 6. Ectotympanic external to horseshoe-shaped and in- bulla, of which it forms side bulla. the outer wall. I have not had the opportunity of examining the tongues of the Indriside, but published accounts and illustrations allly them to those of Lemar. 4. The tongue of Chiromys has its own characteristic sub- lingua, which has a larger area free from the tongue than existing accounts lead one to expect. It has no lateral organs, but its basal conical papille are large. It has two or three vallate papille. Its characters resemble those of Zemzz on the one hand and those of the Loriside and Galagide on the other. 5. The tongue of Zarsius spectrum has the most adherent sublingua. It stands by itself in the consistence and size of the sublingua and the characters of its vallate papille. 6. The frenal lamelle are frequently very complex, and supplementary lameile are present in Vycticebus tardigradus. 7. The foramen cecum and Apical Gland of Nuhn are absent. 8. The lytta is present in several forms, and is frequently accompanied by a central axis, or lytta, of the sublingua. In the other Primates this is absent. Bibliography. 1. Bepparp, F. E.—‘‘ On some Points in the Structure of Hapalemur griseus.” P.Z.S. 1884, p. 391. Burmeister, H.—‘ Beitriige zur niheren Kenntnis der Gattung Zarsius.” Berlin, 1846. Campren, F. A. W. van.—-Verhand. der Konig. Akad. van Wetenschappen, Amsterdam, 1859. Firowrer, W. H.—‘‘ Lectures on the Anatomy of the Organs of Digestion of the Mammalia.” Med. Times and Gazette, 1872 ee O00) Se lo aon 10. abt: 12. 13. 14. 15. 16. 17. 18. 19. 20. lo | OF THE TONGUES OF THE MAMMALIA. (3: . GucENBAur, U.—‘* Uber die Unterzunge des Menschen und der Siiugethiere.” Morph. Jahrb. 1884, p. 428. . Horven, J. van der.— Bijdrage tot de Kenntnis van den Potto van Bosman.” Amsterdam, 1851. Houxtey, T. H.—‘“On the Angwantibo (Arctocebus cala- barensis).” Proe. Zool. Soc. 1864, p. 314. . Kinema, P. H.—‘ Eenige Vergelijkend-Ontleedkundige Aanteekeningen over den Otolicnus Peli.” Leyden, 1855. Mecxet, J. I'.—System der vergleichenden Anatomie. Teil 4. Halle, 1829. Mitncun, F.—‘ Die Topographie der Papillen der Zunge des Menschen und der Siiugethiere.’” Morphol. Arb. 1896. Nusspaum, J.—Anat. Anz. pp. 551-561, 1896; and pp. 345- 358, 1897. OppreL, A.—Lehrb. der vergl. mikr. Anat. vol. iii, pp. 251— 270 and 391-393. . Owen, R.—‘* Monograph on the Aye-Aye.” Trans. Zool. Soc. London, vol. v. p. 33, 1863. Perers, W.—‘ Uber die Saugethiergattung Chiromys.” Abh. der Konig]. Akad. der Wiss. zu Berlin, 1865. Pocock, R. I.—*‘ On the External Characters of the Lemurs and of Tarsius.” P.Z.S. 1918, pp. 19-53. Scureper, J. L. C., vAN DER Koxnk, and W. VroiimK.— Bijdragen tot de Dierkunde, uitgegeven door het Konin- klijk Zoologisch Genootschap “ Natura Artis Magistra.”’ Kerste Deel. Amsterdam, 1848-54. Sonntac, C. F.—Papers on the Tongues of the Primates. P.Z.S. 1921, pp. 1-29, 277-322, 497-524. TrepEMANN, F’.—Arch. f. Anat. u. Physiol. vol. v. pp. 348-356, STO: Tuckerman, F.—Journal of Morphology, vol. vii. pp. 69-94, 1891. Sonnrae, C. F.—P. Z.8. 1921, pp. 547-577. ROUT AN ie ee at BN: ny) Be sel ; HE ANATOMY OF THE TONGUES OF THE MAMMALIA. 757 39. The Comparative Anatomy of the Tongues of the Mam- malia.—VI. Summary and Classification of the Tongues of the Primates. By Cuaries F. Sonntag, -M.D., H.Z.S., Anatomist to the Seciety. [Received August 8, 1921: Read November 8, 1921.] INTRODUCTION. In the five preceding papers of this series I described the structure of the tongues of the Primates and gave a short account of the phy siolog gy of the lingual glands. I also drew attention to a number of points bearing on “the phylogeny of the tongue. In the present paper I have summarised these observa- tions and pointed out how the lingual characters support systems of classification which have been based on external and skeletal characters. SUMMARY. Pigmentation.—Most of the pigmented forms are included in the Cebide. The colour, which is yellow, green, brown or bluish- black, especially the latter, is uniformly distributed all over the dorsum, or the pigmented an colourless parts may form definite patterns ; in Ateles grisescens, for example, there is a white cross on a brown background, The vailate and fungiform papille may be pigmented (e. g. in Hylobates lar) or colourless, but the lateral organs and central parts of the inferior surface of the tongue are always white. If several examples of each pigmented species are examined it will be seen that the colour varies considerably, so is of no value for purposes of classification. Most specimens of Cercopithecus patas, for example, have reddish- yellow tongues, but some tongues are colourless, and the fungi- form papille of Cercopithecus tantalus ave yellow or colourless. The bluish-black colour persists longest in preserving fluids. Form and Proportions.—Most tongues are conical, but a few are oval, spatulate or rectangular; and shape is of no value for comparative purposes. In most of the Primates the tongue is long and comparatively narrow, but in Anthropopithecus troglodytes and Gorilla gorilla the tongue of the young animal is relatively wider than that of the adult. I did not, however, see such variations in Cerco- pithecus patas or Macacus sinicus, of which I examined very young and adult examples. Cunningham showed that the tongue of Simia satyrus resembles that of Homo most closely in the relative proportions of length and width. The Apex is round, truncated or pointed, and may or may not have a notch, but the latter is usually absent from the fresh tongue, It is comparatively smooth, or roughened and tubercu- 758 DR. CG. F. SONNTAG ON TILE COMPARATIVE ANATOMY lated by sinall conical and fungiform papille. The relative quantities of fungiform papille on and immediately behind the apex vary; in the Simiide and Cercopithecide they are numerous and thickly clustered, but in the other families they are few and discrete. In some species of Zemur the apex exhibits a number of sharp- pointed processes prolonged forwards from ridges on the inferior surface of the tongue, but these vary even in different examples of each species. Sulci and Ridges.—Few fresh tongues have median dorsal sulci on the oral part of the dorsum, but many preserved specimens do; and I observed a well-marked median sulcus on the pharyngeal part in Anthropopithecus troglodytes alone. The most pronounced mesial dorsal sulcus which I observed occurred in Mystax ursulus. Median dorsal ridges are present in some Lemuroidea and Hapalidee. Wide, shallow transverse sulci separating low wide transverse ridges are present on the fresh tongues of Lemur catia and Hapale jacchus; and fine narrow sulci are seen in Gorilla gorilla and Perodicticus potto. Some of the fine sulci and ridges remind one of the patterns of fissures and ridges on the finger-tips. The median ventral sulcus is present in most tongues, and is never an artefact induced by preserving fluids as is the dorsal one in bottled specimens. It is narrow and deep, or wide and shallow, and it frequently opens posteriorly into a triangular fossa which recieves the upper end of the frenum. In some of the Lemuroidea it receives the median dorsal crest of the sublingua, and it recesses the crest on the dorsal surface of the frenal lamella of Cebus fatuellus. In Gorilla gorilla, some species of Hylobates and some of the Lemuroidea it contains a fixed crest; and it has been stated that the crest is a remnant of the sublingua, but the presence of both these structures in some Lemurs would seem to disprove that theory.’ . Lateral Borders.—The edges of the tongue are sharp, or full and rounded, and increase in thickness from before backwards in most species. Those of Chiromys madagascariensis are more massive in proportion to the size of the tongue than those of any other species of the Primates. Of the Pithecoids Simia satyrus and Anthropopithecus have the most massive lateral borders. In Tarsius spectrum, Microcebus, Chiromys, and all Loriside and Galagide the lateral borders are devoid of lateral organs and, as I hope to show in a future paper, this has an important bearing on phylogeny. In Gorilla gorilla and Simia satyrus only small parts of the lateral organs are found on the lateral borders, but in all other Primates the greater part is found there. ‘he conical and fungiform papille on the lateral borders are arranged in vertical rows and the points of the former are directed backwards. OF THE TONGUES OF THE MAMMALIA. 159 Distribution of the Papille.—In all Primates except some specimens of Simia satyrus and Symphatangus syndactylus, some Lemuroidea and Homo, papille cover the entire dorsum, apex, lateral borders, and a bounding zone of the inferior surface. In all these species there are smooth non-papillary areas on the base of the tongue, and in the Lemurs the area is bisected by the median glosso-epiglottic fold. The ventral papillary zone varies greatly in width, and its characters ave of limited value for purposes of classification. It is wide in the Simiide, Cercopithecidee, and in Cebus, Ateles, and Lagothrix, but is narrow or absent in all other Primates. In species with a wide zone the conical and fungiform papille are numerous, but in those with a narrow zone there are few or no fungiforms. In Goriila gorilla many of the fine transverse sulci on the dorsum cut the lateral borders and run inwards on the ventral papillary zone. The Circumvallate Papille.—In my paper on the tongues of the Cercopithecide I showed that all the papillary patterns present in the family will be seen in most species if sufficient examples of each ave examined. The whole series may quickly appear or it may be necessary to examine many. I am not prepared, however, to apply this rule to any other family except the Hapalidz, as insufficient specimens have come to hand. In the following list of papillary patterns, P means two . papille forming a pair, and D.P. indicates four papille arranged in a double pair. The Y type means that there are several papillae present and does not include forms in which the four papille of a double pair are arranged in a Y. Family. Patterns of vallate papille. PELL A ee ye elke se hia DP Leia Ay A Cercopithecide ......... A Vi op ln DP Celiedee reales weeds Ne ene ey yh do abe Hlanpabiclesa My s503 3: Xe Wgenaumrcbestasi ale. Goh an NG [SS ee LORUsiGeey Henkes. sae Ne ; Gala gileagrey ise. Hecs see iNew ee @hivouiyides) 45) sgh 3) 42: TANCE OLE Damen lee tigated saeco aetsicr Three papille arranged in a line. One can see, therefore, that the Simiide and Lemuride are the only families whose tongues possess more than four papillee arranged in a Y, and it will be shown later that they differ from all Scher Primates i in other respects. The papille are round or oval on plan and conical on elevation, with the bases of the cones projecting beyond the vallums. The Fungiform Papille stretch right across the dorsum, or are absent from the centre thereby forming a dorsal bounding zone. They form a cluster behind the apex, but are arranged in rows Proc. Zoou. Soc.,— 1921, No. LI. 51 760 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY of varying degrees of obliquity behind that. The apical cluster is large in the Simiide and Cercopithecide, but in all other families the transverse rows extend far forwards at its expense. In the majority of tongues with large apical clusters there are many fungiform papille on the ventral papillary zone, It has been shown by Tuckerman that the fungiform papille of the apical cluster have many taste-buds. In some specimens of Anthropopithecus troglodytes there is a row of prominent fungiform papillae occupying the mid-dorsal line of the tongue and replacing the median dorsal sulcus. It is sometimes impossible to tell whether a papilla at the posterior part of the oral division of the dorsum. is a large fungiform or small vallate form, for fossa and vallum may ihe indistinguishable even through a strong lens. Histological examination is the only proof, The funeiior m papille may have no taste-buds or these, if present, lie on ine free upper surface of the papilla; in the vallate papille, on the other hand, the taste- buds never le on the free upper surface of the papilla, but are deep down on one or both sides of the fossa. In some tongues there are more fungiform papille than are visible to the naked eye, for some are entirely concealed by overhanging conical papille (e.g. Anthropopithecus troglodytes). The fungiform papille on the ventral zone may be thickly clustered at the apex of the tongue and scanty further back, or vice versa, and the examples, and the rows in which they are arranged are close together or discrete. From the point of view of classification the most important features are the size of the apical dorsal cluster and the presence ov absence of the fungiform papille on the ventral papillary zone. ‘Although their presence or absence in the centre of the oral part of the dorsum varies greatly, it is not a character of sufficient distinctness to be of “value for purposes of classification. The Conical Papille vary im size and arrangement in the different families, and there are three types of the formere= 1. The papille on the pharyngeal part of the tongue are small :—--Homo, Sema satyrus. 2. The papille on the oral part of the tongue are compara- tively small, but those on the pharyngeal part are large and prominent :— Gor ‘lla aorilla, Anthropopithecus troglod Wee all species of Hylobates, all species of Lemur, Chiromys madagas- cariensis, and Tarsius spectrum. ‘his type also occurs in some lower Mammalia. ‘ 3. The papille gradually increase in size from the apex of the tongue back to the epiglottis :—Cercopithecidee, Cebide, Hapalide, Loriside, and Galagide. This arrangement is also ‘present im Microcebus in which the vallate papille form a triangle, so the papille are of value for distinguishing it from Lemur. The arrangement of the papille distinguishes most of the Cebide from all other families. In the latter they form a cluster behind the apex and rows of varying degrees of obliquity behind OF THE TONGUES OF THE MAMMALIA. 761 that, but in the Cebide they are dotted irregularly all over the dorsum. The tongues with smooth non-papillary areas on the pharyngeal part of the dorsum have been enumerated above. In most Primates the points on the oral part of the dorsum look backwards or backwards and inwards, but in some specimens of Simia satyrus and Cercopithecus cethiops those on the centre of the orai part run in all directions. In the Cebidee, Hapalide, Lemuroidea, and Tarsioidea the conical papille are mostly pointed, and cylindrical and globular forms are uncommon. In the other families there is a good admixture of all types. ‘ The Lateral Organs present numerous forms and are of value for purposes of classification :— 1. Organs absent :—AMicrocebus, Chironvys, Tarsius, the Lovi- side and Galagidee. The lamine and sulci form ladder-like patterns on the dorsum of the tongue, and only their outer ends cut the lateral borders of the tongue :—Gorila gorilla and Simia satyrus. 3. The organs are convex towards the lateral vallate papille: Anthropopithecus troglodytes, Hylobates (all species), Lemur (all species ). 4. There are rows of short lamine and sulci on the lateral borders of the tongue :—Cercopithecide, . The inner borders of the organs are concave towards the lateral vallate papile :—Cebide. 6. A few faint irregular laminz and sulci are present on the lateral borders of the tongue :—Hapalide. In very few specimens did I find an equal number of laminze and sulci in the two organs of the same tongue. But one must be careful net to mistake simple folds of the mucosa at either end of the organs for lamine. Histological examination is the only true test in doubtful cases, for it reveals the presence of taste-buds in the true lamine. In a few species of Cercopithecus one may find fungiform papillee situated on the lamine of the lateral organs. The degree of protrusion of the lamin and depth of the sulci vary not only in different animals but in several examples of each. The Lingual Glands are divided into apical and basal parts, but the former is most variable. The Apical Gland of Nuhn is present only in Homo and Simia satyrus. Oppel believes that it is a piece which has become cut off from a forward prolongation of the basal glandular mass. The basal mass in the Marsupialia sends forwards two prolonga- tious of variable stoutness, and it is possible that the Apical Gland of Nubn has been cut off from one of these. If that were so it would support Gegenbamr’s view that the tongues of the Primates have evolved from those of the Marsupialia. The serous and mucous glands on the pharyngeal part of the a1 762 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY tongue are developed to an equal degree in the Primates, but the degree of development of the entire basal mass varies. The pharyngeal part of the tongue possesses a variable degree of development of lymphoid nodules, and a variable number of orifices of pits and the ducts of glands; and the latter are of value for distinguishing the tongues of the various genera of the Cercopithecide from one another. Orifices are visible in all Cercopithecide, but are absent from most Cebide, so are of classificatory importance when taken in conjunction with the characters of the lateral organs and mode of arrangement of the conical papille. The following are the Cheever of the orifices in the Cercopithecidee :— Genus Presbytes :—Orifices larger and more patulous than in any other genus, and lie in the centres of large round glandular areas. The salivary glands are enormous. Genus Cercopithecus :—Orifices well-marked, discrete and not patulous, and no white areas present around them. Genus MZacacus :—Orifices like pin points. Genus Cercocebus:—Small duct orifices present at the sides of the base of the tongue. Genus Papio ._Targe duct or Unees present at the sides of the base of the tongue. : In many of the Cebide the sour on the base form zones, with concave anterior borders, in front of the epiglottis, but 1 did _ not observe a similar condition in any other family. The Prenal Lamella varies greatly in the Primates, and appears to belong to the floor of the mouth rather than to the tongue. It varies even in several examples of the same species, so is of limited value for purposes of classification. In Homo, Simia satyrus, and some specimens of Symphatangus syndactylus it appears as two simple folds over Wharton’s Ducts ; but Mr. Pocock informs me that he observed two well- marked processes in a young Simia satyrus. IJ cannot believe, however, in conformity with niy observations on the tongues of other animals, that reduction in the lamella is a change due to advancing years. In other species the lamella appears as a tviansulat or tongue-shaped process with an entire or divided apex ana with both edges entire, serrated, or bearing small sharp points. The sharpest and most prominent points on the edges occur in Cercopithecus preussi and some specimens of Cebus fatuellus. The Orifices of Wharton’s Ducts vary in their position on the lamella. as follows :— 1. On the upper surface— Gorilla gorilla. On the apex— A nthropopithecus troglodytes. 3. On the under surface— Pithecia satanas. The apical position is the commonest form, however. Tuckerman descvibed a rich nerve plexus with peripheral nerve endings in the lamella of Ateles ater, and Gegenhaur considered that the endings were tactile in function. As this condition has Ne a oe See en ee OF 'THE TONGUES OF THE MAMMALIA. 763 not been found in any other tongue there is insufficient material in which to work out its phylogenetic significance. The Subtingua.—The various Lemuroidea are differentiated from one another by the shape, mobility, and characters of the crests and denticles as follows :— 1. Sublingua triangular or lyrate, has three ventral crests, and-is very mobile. Phe apex 1s aivatted into a brushwork of denticles :— Lemur, Hapalemam, Indriside. Sublingua a large flat plate adherent to the under surface of the tongue by its central parts; no denticles present, but a strong keel-like ridge on its under surface projects forwards in the middle line :—Chiromys. 3. Sublingua large, tongue-shaped, but not quite so free as that of Lemur. ‘Vhere is only one median ventral crest, bat’ the dorsal surface has a cress of variable prominence. This crest, the characters of the denticles, and the variations in the consistence of various parts of the sublingua ure important :—- In Microcebus the sublingua is uniformly thick, the median dorsal evest is slight, and the denticles are moderately long. In WVyeticebus and Loris the central parts of the sublingua ave thicker than the lateral parts, the median dorsal crest is small and the denticles are of moderate length ; they are discrete in the former and close in the latter, but there is no essential difference between the sublinguz of these genera. In Perodicticus the median ventral crest is bifurcated pos- teriorly, the median dorsal crest is very prominent, and the denticles are long and slender. In Mecrocebus, Galago, and Hemigalago the anterior border of the sublingua is broad, but in Loris and Vycticebus it is more or less pointed. The Plice Pimbriate of the Simiide are derived from the sub- lingua by a process of phylogenetic reduction, and I showed that the plicee of Anthropopithecus troglodytes with the intervening piece of mucosa form a soft triangular field resembling a sub- lingua; this is even more marked in the tongue of the new-born child, as deseribed and figured by Gegenbaur. The plice of Phascolurctos cinereus, however, do not beund such an svea. The tongues of the Gibbons, Cercopithecide, Cebide, and Hapalide illustrate the ultimate stage of reduction, for no traces of the sublingua or plica are present as a rule in the extra-uterine stage. In the fetal Gibbon, as shown by Deniker, there is a well-marked sublingua; and T observed two minute plicee in a young Cercopithecus patas, so it is probable that the feetuses of all Primates have sublingue. If one examines a series of human tongues at different ages, one finds that the new-born child has well- -leveloped plicee or an actual sublingua provided with taste-buds. As age advances the buds disappear and the plicee diminish in size. These taste-buds _ probably account for the more acute sense of taste in the child. Experimental methods also demonstrate that the sense of taste 764 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY elicited by applications of solutions to the centre of the oral part of the dorsum diminishes as age advances. This atrophy of structure following loss of function may have played an important part in the reduction in and ultimate loss of the primitive Mammalian tongue. The Lytta—Vwo forms are to be recognised :—The lytta of the tongue and the lvtta of the sublingua, And Gegenbaur showed that the latter, when present, appears in one of two forms. In Stenops it forms a str ong central supporting rod, but in Tarsius ib is double. In Lemur itis absent altogether. Owen described the keel on the ventral surface of the sublingua as the lytta, but sections through the tongue show a well-marked lytta inside ; it is connected in the middle to the sublingua. CLASSIFICATION. The structures which are of value for purposes of classification are :— 1. The number and arrangement of We vallate papille. 2. The arrangement of the conical papille on the oral part of the dorsum. 3. The mode in which the conical papiile increase in size from before backwards. 4, The size of the cluster of fungiform papille on the dorsum behind the apex of the tongue. 5. The characters of the lateral organs. 6. The width of the ventral papillary zone, with the number, characters, and arrangement of its papille. 7. The presence or absence and characters of the lytte of tongue and sublingua. 8. The presence or absence of the apical gland of Nuhn, and the characters of the orifices of ducts and pits on the pharyngeal part of the dorsum 9. The sublingua and plice fimbriate. These features are of specific value in the case of all the Simiide, except Hylobates, and many of the Lemuroidea, but they are of generic value ouly in all otber Primates. They lend additional weight to some schemes of classification Lased on external and skeletal characters; in some cases they help us to settle points which are not supported by a large body of very conclusive evidence. The tongue of Simia satyrus resembles that of Homo most closely. Both have rounded apices, and their form and general proportions are similar. They have no large conical papille on the base of the tongue, and their vallate papille forma V. The frenal lamelle are two small folds over Wharton’s Ducts, and their plice fimbriate are equally developed, but smaller than those of other Primates; and they are the only Primates possess- ing an Apical Gland of Nuhn. In Simia satyrus the lateral OF THE TONGUES OF THE MAMMALIA. 765 organs are better developed, and the conical papiile on the oral part of the dorsum include more pointed forms, but Homo alone of all the Primates possesses a foramen cecum. The ladder-like patterns formed by the lateral organs distin- guish Simia satyrus and Gorilla gorilla trom all other Simiide which have organs convex towards the lateral vallate papille, but the tongue of Gorilla gorilla has enormous conical papille on the pharyngeal part of the tongue, large plice fimbriate, a median ventral crest, and a large triangular frenal lamella; it has no Apical Gland of Nuhn, and its vallate papille form a V or Y. The only character which really differentiates Anthropopithecus troglodytes on the one hand from Hylobates and Symphatangus on the other is the possession of plice fimbriate, for there is a similarity between them in all other features ; and Hylobates has a well-marked, triangular, bifid lamella, whereas Symphatangus syndactylus has two small folds over Wharton’ s Ducts similar to those of Homo and Simia satyrus. Their vallate papillae form a Y. In the Cercopithecide, Cehide, and Hapalide the vallate papillae never form a Y, the conical papille on the base of the tongue are never disproportionately large, and the lateral organs are neither ladder-like nor convex towards the lateral vallate papille. The chief differences between them lie in the characters of the lateral organs, the size of the apical cluster of fungiform papille, and the mode of arrangement of the conical papille on the oral part of the dorsum. In the Cercopithecide the lateral organs form rows of lamin and sulci on the lateral borders, the apical cluster of fungiform papillae is large, and the conical papillz on the oral part of the dorsum are regularly arranged. ‘I'he genera are distinguished by the characters of the orifices of ducts and pits on the pharyngeal part of the dorsum (page 762). In the Cebide the inner borders of the lateral organs are con- cave towards the lateral vallate papille, the conical papille on the oral part of the dorsum are crowded together without any definite arrangement, and the apical cluster of fungiform papille is small. The vallate papille are two in number, and there is a well- marked median suleus on the pharyngeal part of the dorsum in Aotus, but in all other genera there are more than two vallate papille, but no basal dorsal suleus. The ventral papillary zone is wide, and has few fungiform papille in Lagothria, but it has many papille in Ateles and Cebus. In Afeles the ventral fungiform papille are most numerous round the apex, but in Cebus they are most numerous farther back. The ventral papillary zone is narrow in Pithecia, Alouwatia, Callicebus, and Cacajao. It has many fungiform papille in Pithecia but few or none in the others; and Alowatta is the only 766 DR. C. F. SONNTAG ON THE COMPARATIVE ANATOMY ‘ genus possessing large glandular orifices on the base of the tongue. Callicebus has many nodules and few orifices on the base, but Cacajao has neither. The Hapalide always have a vallate triangle and a few irregular lamine and sulci in their lateral organs. The pha- ryngeal part of the tongue occupies a relatively small part of the whole, and the ventral papillary zone is small or absent. The conical papille are regularly arranged and the apical cluster of fungiform papille is small. The lingual characters are not sufficiently distinctive to classify the genera. The Lemuroidea and Tarsioidea differ from the Simize in the possession of a sublingua whose size, consistence, mobility, denticles, and minute structure vary in different families. In the Tarsioidea the sublingua is soft, small, devoid of denticles, and only delimited at the sides from the under surface of the tongue by a narrow groove. There are no lateral organs, no large conical papille, and few apical fungiform papillee. In the Lemuroidea the sublingua is large, horny, denticulated, and is move or less movable. In Lemur and Hapalemur the sublingua is triangular or lyrate, very free, and has three ventral crests. No lytta is present, The vallate papillee form a Y, and the conical papille on the base ave disproportionately large. The lateral organs are convex towards the lateral vallate papille. In Chiromys the sublingua is tongue-shaped, adherent by its central parts, has one ventral crest which is keel-like and nodulated. It has a lytta inside the tongue. Vallate papille form a pair or triangle, but never a Y. Conical papillae on base of tongue large. No lateral organs. In Microcebus, the Loriside and Galagidz the sublingua is a flat plate with one median ventral crest and a variable dorsal crest. It varies in mobility, consistence, and denticles, and the value of these characters has already been described (page 763). The vallate papille form a triangle, but never a Y, the conical papilla on the base of the tongue are not disproportionately large, and lateral organs are absent. : Tbe special points arising from these notes on classification are :— 1. The tongue of Sima satyrus resembles that of Homo. 2. Simia satyrus and Gorilla gorilla differ from all other Simiidz in the characters of the lateral organs, but there the resemblance between these two species stops. 3. The tongue of Hylobates differs from that of Anthropopi- thecus troglodytes in having no plice. 4. The separation of Symphatangus from Hylobates. 5. The great value of the lateral organs for purposes of classification. 6. The value of the lingual glands for distinguishing between the genera of the Cercopithecide. — ee a a a ae OF THE TONGUES OF THE MAMMALIA. 767 7. The genera of the Hapalide cannot be distinguished from ene another by the characters of the tongue. 8. The tongues of Lemur and Hapalemur resemble those of the Simiide most closely. 9. It is impossible to ally the tongue of Wicrocebus with those of Lemur and Hapalemur, but it resembles those of the Loriside and Galagide. 10. The characters of the tongue at once distinguish the Lorises, Galagos, and Pottos from Lemur. 11. The tongue of WZarsius resembles neither those of the Simiide nor of Lemur, and is hard to distinguish from those of the Loriside and Galagide. It has three vallate papille in a line, but these may be elements of a triangle which has been flattened out. 12. The keel on the under surface of the sublingua is not the true lytta as described by Owen. ON SOME NEW PARASITIC MITES. 169 40. On some new Parasitic Mites. By Stanuey Hirst, F.Z.S8. (Offered for publication by permission ae the Trustees of the British Museum.) [Received September 18, 1921; Read October 18, 1921.] (Text-figures 16-43.) The present paper deals with various parasitic Acari, mostly belonging to. the family Gamaside, including eight new species of Liponyssus and a new species of Rhinonyssus, also two new genera closely allied to Rhinonyssus. A new Tarsonemid mite belonging to the genus Podapolipus is also described, and a new ‘Red Spider” (Paratetranychus trinitatis) found on grapes at Trinidad. Text-figure 16. Rhinonyssus nove-guinee, sp. u., %. Ventral view. Genus Rutyonyssus Trouessart. RHINONYSSUS NOVE-GUINES, sp.n. (Text-fig. 16.) ©. Abdomen not elongated. The spinules or hairs usually present on the venter in this genus have practically disappeared, being mostly replaced by exceedingly minute and inconspicuous circular pores. Anal plate distinct, being elongate pyriform ; it is furnished with a pair of hairs. Second and third coxe each 770 MR. STANLEY HIRST ON SOME with a pair of very stout cones or tubercles; the other segments of the legs are provided ventrally with similar stout cones and also a few spinules. Length (including capitulum) *93 mm. Host: A bird (Craspedophora magnifica), New Guinea. Pro- fessor Trouessart’s collection. ; RHINONYSSOIDES, gen. nov. Closely alhed to Rhinonyssus Trouessart, but the chelicera is — f very differently shaped, consisting of a short stout basal portion and a long thin distal style, the latter has a very minute lobe near the distal end, which is very slightly curved, being in the form of a minute claw-like denticle (text-fig. 17 6). Text-figure 17. TERZIJ Rhinonyssoides trouessarti, sp.u., 2. Ventral view. a. Palp from below. b. Chelicera. RHINONYSSOIDES TROUESSARTI, Sp. n. (Text-fig. 17.) 2. Body shaped as in Rhinonyssus, the abdomen not elongated however. Anterior hairs on venter short but quite fine. A number of hairs are situated posteriorly on the venter; they are longer than the anterior pairs of hairs and are fairly slender, the ‘ NEW PARASITIC MITES. (erRil basal half being only slightly thickened, whilst the apical half is fine. Anal plate present, being pyriform i in shape and furnished with three hairs as in Liponyssus ete. First free segment of palp apparently with a process. There is a median row of minute denticles on the ventral surface of the base of the eapitulum. Anteriorly the second coxa is produced to form a stout process or spur rather similar to that present in the same position in Liponyssus but thicker and blunter. The spinules or hairs present on the proximal segments of the legs in Rhinonyssus are replaced by curious globular structures in this species; a pair being situated on each of the first three coxe and a single globule on the fourth coxa. Trochanter of first leg with two chitinous globules ventrally, second and third trochanters each with three globules and a short pointed spinule, fourth trochanter with four globules and a short pointed spinule. Distal segments of legs furnished with short pointed spinules. Claws of first legs only slightly curved (the claws of the other legs are missing in the unique specimen). Length (with capitulum) 92 mm. Host: A bird (Sphecotheres mawillaris), Australia. Professor Trouessart’s collection. NEONYSSUS, gen. nov. Allied to ZLiponyssus but with the palpi shorter, the sternal plate longer thau wide, and the hairs on the body and legs very few in number. Differing from Lhinonyssus in the presence of two dorsal shields (an anterior and a posterior one), instead of only a single anterior shield. NEONYSSUS INTERMEDIUS, sp. n. (Text-figs. 18, 19.) 2. Two dorsal shields both fairly large, their surface reticulated and with numerous punctations, posterior shield with the hinder end rather bluntly rounded off. Sternal shield longer than wide, apparently there are no hairs on it, but there are three minute dots or circles on each side. Grenito-ventral plate unusually short. Anal plate similar to that of Liponyssus, being pear-shaped. Palpus short, being very similar to that of Rhinonyssus. Apparently there is no median groove on the ventral surface of the capitulum. With the exception of the pair on the anal plate there seem to be no spines or hairs on the ventral surface. Peritreme opening dorsally asin Rhinonyssus. Legs rather short, apparently only the distal segments are furnished with hairs, a dorsal hair on the tarsus of the third leg being longer than the others, but still it is not very long ; there is also a pair of distal spines on the ventral surface of this segment. Except for the dorsal spur on the anterior surface of the second coxa which is present but quite short (obsolete), the coxe are without hairs or spines. It is possible that some of the hairs on the body and limbs of 772 Text-figure 19. Text-figure 18. MR. STANLEY HIRST ON SOME Neonyssus intermedius, sp. n., 9. Dorsal view. a. ‘Tarsus of first leg from above. Neonyssus intermedius, sp. n., 9. Ventral view. NEW PARASITIC MITES. 773 the single specimen have been destroyed by the caustic potash used in preparing it. It seems probable, however, that this absence is natural for in general appearance the mite is very like a Rhinonyssus, and the spines may be reduced or almost suppressed in the latter genus. There is a fully-developed larva inside the specimen and it resembles the parent in having only a few hairs on the legs. The claws of this larva are well developed and strongly curved. Length (with capitulum) *53 mm. Host: The above description is based on a single specimen labelled ‘“*Sur un Oiseau de Madagascar.” Professor E. L. Trouessart’s collection. . Genus Liponyssus Kolenati. The species of this genus are found both on birds and mammals. Only the species that I have been able to examine personally are included in the key given below. A number of species are not in dur collection, and these have been left out. Two dorsal shields are present in the nymphal stages of Liponyssus, whereas there is usually a single undivided dorsal shield in the adults. In the adult females of a few species, however, there are two dorsal shields as in the nymphs: for instance in LZ. viator, sp. n., where they are shaped exactly as in the nymphs of certain other species. The adult females can always be distinguished by the presence of separate sternal and genito-ventral plates on the venter, and by the delicate fan-shaped or semi-circular genital operculum situated at the front end of the latter plate. Key to the females of the species of the genus Liponyssus present in the British Museum Collection :— ie with two shields............. Ale 2 Worsomauwithtarsmpdetshield sy. et tek aes cee: wcicete 8 aie of the two dorsal shields very small ......... 3. 2. 4 Posterior of the two dorsal shields fairly large ......... ( Anterior dorsal shield small and separated from the | minute posterior shield by a considerable space... L. serpentiwm Hirst. | Anterior dorsal shield rather long and only separated i. from the posterior one by a short space ................ I. natricis Gerv. Sternal plate quite short, the two posterior hairs on it n long. Anterior spur on second coxa short............. LZ. viator, sp. n. Sternal plate longer ae margin of sternal plate strongly chitinised . OD on Posterior margin of sternal plate weakly chitinised... { Posterior dorsal shield with only three pairs of fairly J Jone hairs S. pss eee ee ee aa slate) cen, eheropteralis, 6. nom, nov. | Posterior dorsal shield with a number of fairly long feeeap IMDS: ss vetna re mete tte are Saeed vcr eactiness¢ Lan WYASEC, SPs Ds 774 MR. STANLEY HIRST ON SOME ‘Sternal plate with strong reticulate markings. A few of the hairs at the posterior end of the body longer 7 2 than the other hairs | Sternal plate smooth. Hairs at posterior end of body not much longer than the others .. ( Genito- ventral plate truncate pedoonly nel fae nished with three pairs of hairs g ioe) Genito-ventral plate rounded off or pointed pos- Re HERO My desc06 a0 doe acd ob aoa soa voGnba as vee aaa ToOSLU eee cod ehE coo Sternal plate with a pair of well-defined circular | stigma-like markings. Genito-ventral plate with 9.4 THI PUIEE) JOHN TRS) OIE WEMDES Us cag ngscoaddoecoateasoendoce penboandio: | Sternal plate without any circular markings. Genito- {ventral plate with only a single pair of hairs......... ( Sternal plate with the posterior margm thickened...... 10. ( Sternal plate not thickened posteriorly .............. Coxe of legs furnished with distinct dentiform spurs. 11-% Coxe of legs without any spurs, except an anterior dorsal one on the second CoxaA ............cee cee cee eens Coxa of second leg with a very large curved outer 12. spur 4 Coxa of second leg without any very large outer spur. Coxa of second leg with two posterior spurs ............ 13 Coxa of second leg with only one posterior spur ...... (Second coxa with a large curved inner (anterior) | spur ventrally. Second tarsus usually ending in 14.4 two spurs, one of w nchrisiveny stout 1o 0s. sass { Second coxa without any anterior inner spur ventrally. (Only the anterior dorsal spur of the second coxa and 1s J a little outer posterior spur on the third coxa present oO. (Several other spurs present on the coxee as well ...... (Sestenon end of dorsal scutum rounded off or blunt.. 16. ‘ | Posterior end of scutum pointed ................ Second coxa with only a very short enmiedl : spur posteriorly rate lips Second coxa with a large pointed spur posteriorly ( Posterior end of eenito-ventral plate rather wide and | blunt or rounded off . 18.4 : | Posterior end of genito-ventral plate narrower and {usually pointed Case HEY Re EE Anal plate rather angular in Genes 119), a plate pear-shaped, the sides being curved......... ( Size larger. Dorsai shield elongated and rather narrow. Sternal plate comparatively long and without any | distinct markings. Peritreme rather short ......... 4 | L 20. shorter and with afew deep curved littleimpressions on each side in the anterior angles. Peritreme long. Size smaller. Dorsal shield wider. Sternal plate . L. biscutatus, sp. n. L. javensis Oudimns. LL. madagascariensis, sp. n. L. sternalis, sp. n. 10. LL. ethiopicus, sp. n. nae 12. 18. L. creightoni Hirst. 13. LL. sciurinus, sp. n. 14. LL. liberiensis Hirst. oe LL. blanchardi Trouessart. 16. L. arcuatus C. L. Koch and L. earnifexr ©. L. Koch. Wie LL. confucianus Hirst. LL. macedonicus, sp. n. 19. PA LL. flavus Kolenati. 20; L, ellipticus Kolenati. LL. granulosus Kolenati. NEW PARASITIC MITES. Tigo oles plate with three pairs of hairs..................... 22. 21. Sternal plate with two pairs of hairs ..............00... 23. Dorsal shield narrow, especially posteriorly (parasitic ; a5 OUMTAUS)) Seay oekaki. necealas . cat muh rma tes eee ea CoO may ELL STiR Dorsal shield wider (parasitic on hen)... ................. D. bursa Berlese. ( Hairs on dorsum only slightly eurved.........0........... DL. silviarum 23. Can. & Fanz. Hairs on dorsum very distinctly curved.................. 24), { Dorsal shiela rather short. Anal plate rather wide Ieee nigh Ov boweeen ce iaet creat «elk, Rae Ri Ra * 24.< | Dorsal shield more elongated. Anal plate narrower Raye Gere oe eee bse ctins cudiwe bos cehecduelcoobedaannte dl Suarounaen on OOK hbaniist L. lacertinus Berlese. LIPONYSSUS NATRICIS Gerv. Liponyssus arabicus mihi described from a female specimen found on a lizard (Agama adramitana) is apparently identical with ZL. natricis Gerv. I have examined examples of the latter found on a snake at Florence kindly sent me by Prof. Berlese. A very minute posterior shield is present on the dorsum in the adult female, and Berlese seems to have overlooked it. It is very similar in shape to that present in Z. serpentiwm mihi, but is longer and only separated from the anterior dorsal shield by a short space, whereas in the L. serpentiwm the two dorsal shields are widely separated from one another. LIPONYSSUS VIATOR, sp. n. (Text-fig. 20.) Q. There are two dorsal shields, shaped as shown in figure, the anterior one being slightly produced in the middle posteriorly, whilst the posterior one has the anterior margin rather strongly concave ; posterior dorsal shield shaped rather like a tongue, the lateral margins being at first convex and then narrowing to form the bluntly pointed posterior end. Hairs on dorsal shields short, some of the hairs on the margins of the anterior dorsal shield are a little longer, and the hairs of the terminal pair on the posterior shield are longer than the others on its surface. Lateral hairs on body fairly long and the posterior hairs on uncovered part of dorsum very much longer than those on the shields. Sternal plate very short, being in the form of a narrow transverse strip; it bears three pairs of hairs, those of the anterior pair are much shorter than the posterior ones, the latter being much longer than usual. Genito-ventral plate narrow, the posterior eud rather blunt. Anal plate pear-shaped. Peritreme reaching as far forwards as the middle of the second coxa or somewhat further forwards. A minute and inconspicuous spine or spur is present on the anterior surface of the second coxa, but it is much shorter than in most species of Liponyssus; the other cox apparently are without spurs. It is not easy to make out the denticle on the first free seginent of the palp in the specimens Proc. Zoor. Soc.—i921, No. LIT. 52 776 MR. STANLEY HIRST ON SOME from Cypselus affinis, but it is quite distinct in the example from Gecinus vaillanti, being sharply pointed. Length of body (not including capitulum), 2 °65--87 mm. Host: Several specimens from nests of Cypselus affinis, Calcutta, collected by C. Paiva (28. vi. 1909) and forwarded to me by Dr. Nelson Annandale. Also an example from Gecinus vaillanti, Tangier (May 1892), presented to the Museum by . Dr. James Waterston. Text-figure 20. Liponyssus viator, sp. u., 9. Dorsal view. a. Sternal plate. Nore.—The specimen from Gecimus vaillantt agrees closely with those from Cypselus affiis in the structural details given above, but the posterior dorsal shield is rather more narrow, and the denticle on the first free segment of the palp distinct. LIPONYSSUS CHIROPLERALIS, Nom. nov. 2. Two dorsal shields are present, their surface being finely reticulated. The shape of these shields varies somewhat in specimens from different localities; the posterior one is tongue- shaped, its hinder end being narrowed and blunt. Hairs on the NEW PARASITIC MITES. ue shields of moderate length ; there are three pairs of these fairly long hairs on the posterior shield and also three or four pairs of very minute and inconspicuous hairs at its posterior end. Hairs on softer parts of body very numerous, of moderate length and rather thick. Sternal plate trapezoidal, its posterior margin is very much thickened, beg strongly chitinised as in LZ, ethio- picus and L, nyasse ; anterior pair of hairs on it quite short, the other two pairs usually longer. Genito-ventral plate with the posterior end sharply pointed. Anal plute pear-shaped. Peri- treme only reaching as far forwards as the middle of the second coxa or even shorter. First free segment of palp with a well- developed process. Second cowa with the anterior spur well- developed ; apparently there are no other spurs or spines on the coxe. Bristles on dorsal surface of legs very numerous and rather stout. Hosts: Bats. On Pipistrellus, Ambleside, Windermere (11. ii. 1912) (J. FL. Cormack); ex Dr. J. Waterston’s coll. On Myotis myotis, Oristano, Sardinia (Dr. A. HW. Krausse); ex Hon. N. C. Rothschild’s coll. On bats, at Ain Sefra, S.W. Algeria (14. v. 1913); ex Hon. N. CU. Rothschild’s coll. Off Ctenodactylus gundi, Gorge of Tilaton, Kl Kantara; specimens collected by the Hon. L. W. Rothschild and Dr. E. Hartert. Also a specimen from a bat, Tabgka, Tiberias (P. £. Schmitz) ; ex the Hon. N. C. Rothschild’s coll. Norn.-—Canestrini, relying upon the colour of living specimens, considered this species of Liponyssus with two dorsal shields to be L. arcuatus C. L. Koch (see Prospett. Acar. Ital. i. pp. 120-121, 1885), and Berlese is also of this opinion in his Acari ete. Reprt. in Italia (Ordo Mesostigmata, p. 22, 1892). Oudemans gives the name L. arcuatus to quite a different species with an undivided scutum (see Arch. Naturg. Abt. A. Heft ix. pp. 68-75, 1913). According to this author the name of this species with two dorsal shields should be Z. masculinus C. Li. Koch—a mite found on the house-mouse (see Tijdschr. Ned. Dierk. Ver. (2) vii. p. 293, 1902). It is impossible to determine the two species said by Kolenati to have two dorsal shields, viz., Steatonyssus periblepharus and S. brachypeltis (see Sitzb. K, Akad. Wiss. Wien, xxxv. pp. 186 & 187, 1859). As it seems to be impossible to ascertain the real name of this mite, I propose the new name JL. chiropteralis for it. LIPONYSSUS NYASSA, sp. DD. 2. Two dorsal shields are present in this species, the anterior one being very wide; the posterior one is much narrower and shaped like that cf Z. cethiopicus, sp. n., and L. chiropteralis, being tongue-shaped and with the anterior margin distinctly concave. It differs from the posterior dorsal shield of Z. chiro- pteralis in having a number of moderately long hairs on its surface instead of only three pairs as in that species. Hairs on uncovered (softer) part of dorsum numerous, stiff, of moderate 52* 778 MR. STANLEY HIRST ON SOME length, and sharply pointed much as in L. chiropteralis. Sternal plate rather weakly, chitinised except the posterior margin, which is very much thickened and slightly concave as in L. chiropteralis and “. wthiopicus; there are numerous fine transverse linear markings on its surface and the three pairs of hairs are long and fine. Genito-ventral plate with the posterior end rather bluntly Text-figure 21.