etee aati nee ny fetetettet eS teatgts it tele: This 3 OY ss) 0) Hate ote at i rte state siete pte tetetaty . re fies gh, AOL aad Se Sctotatetate + Ou ree Bry < eters ?: aauarty th 4 aie i ne ¥¢, at reine 7 ae Tere alee oO i ttl? ewe tele tered TET steele? tte tpt x ery ang es pecegrrye Ch he Bs wot ve + ' Me ine oy > caked eouea yas . 8 peated eat err at 02) areca Gren yy seert FHEATES PAOM CORNELL LAB of ORNITHOLOGY LIBRARY at Sapsucker Woods oe Illustration of Bank Swallow by Louis Agassiz Fuertes LABORATORY OF OF TTHOLOGY American Museum of Natural History 77th Street and Central Park West New York os pate ee a ae Cornell University The original of this book is in the Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www. archive.org/details/cu31924022526382 BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY VOLUME XXXVI, 1917 Eprror, J. A. ALLEN THE DISTRIBUTION OF BIRD-LIFE IN COLOMBIA; A CONTRIB)- TION TO A BIOLOGICAL SURVEY OF SOUTH AMERICA BY Frank M. CHAPMAN NEW YORK PUBLISHED BY ORDER OF THE TRUSTEES 1917 FOR SALE AT THE MUSEUM AMERICAN MUSEUM OF NATURAL HISTORY. SEVENTY-SEVENTH Srremr and CentraL Park West, New Yorx Crry. Boarp oF TRUSTEES. Presivent. HENRY FAIRFIELD OSBORN. First Vice-PresipEnt. Seconp Vice-PREsIDENT. CLEVELAND H. DODGE. J. P. MORGAN, TREASURER. SECRETARY. HENRY P. DAVISON. ADRIAN ISELIN, Jr. Ex-oFFIc1o. JOHN PURROY MITCHEL, MAYOR OF THE CITY OF NEW YORK WM. A. PRENDERGAST, COMPTROLLER OF THE CITY OF NEW YORK CABOT WARD, PRESIDENT OF THE DEPARTMENT OF PARKS. ELEctTive. GEORGE F. BAKER. WALTER B. JAMES. FREDERICK F. BREWSTER. A. D. JUILLIARD. R. FULTON CUTTING. CHARLES LANIER. THOMAS DeWITT CUYLER. OGDEN MILLS. JAMES DOUGLAS. PERCY R. PYNE. HENRY C. FRICK. JOHN B. TREVOR. MADISON GRANT. FELIX M. WARBURG. ARCHER M. HUNTINGTON. ARTHUR CURTISS JAMES. ADMINISTRATIVE OFFICERS. Direcror. ASSISTANT-SECRETARY. FREDERIC A. LUCAS. GEORGE H. SHERWOOD. ASSISTANT-T REASURER. THE UNITED STATES TRUST COMPANY OF NEW YORK. ili ScIENTIFIC STAFF. Freperic A. Lucas, Sc.D., Director. GEOLOGY AND INVERTEBRATE PALHONTOLOGY. Epmunp Otis Hovey, Ph.D., Curator. Cuester A. Rerps, Ph.D., Assistant Curator. MINERALOGY. L. P. Gratacap, A.M., Curator. GerorcE F. Kunz, Honorary Curator of Gems. WOODS AND FORESTRY. Mary Cyntuta Dickerson, B.S., Curator. BaRrRINGTON Moors, A.B., M.F., Associate Curator. INVERTEBRATE ZOOLOGY. Henry E. Crampton, Ph.D., Curator. Roy W. Miner, A.B., Associate Curator. Frank E. Lutz, Ph.D., Associate Curator. L. P. Gratacap, A.M., Curator of Mollusca. A. J. Murcuumr, Assistant. Witiarp G. Van Name, Ph.D., Assistant. Frank E. Watson, B.S., Assistant. W. M. Wueeter, Ph.D., Hon. Curator of Social Insects. A. L. Treapweti, Ph.D., Hon. Curator of Annulata. .Cuartes W. Lene, B.S., Hon. Curator of Coleoptera. ICHTHYOLOGY AND HERPETOLOGY. Basyrorp Dran, Ph.D., Curator Emeritus. Joun T. Nicuoxs, A.B., Assistant Curator of Recent Fishes. Mary Cyntuta Dickerson, B.S., Associate Curator of Herpetology. MAMMALOGY AND ORNITHOLOGY. J. A. Auten, Ph.D., Curator. Frank M. Cuapman, Se.D., Curator of Ornithology. W. DeW. Mitter, Associate Curator of Ornithology, lv Scientific Staff. . Roy C. Anprews, A.M., Assistant Curator of Mammalogy. H. E. Antuony, B.S., Assistant in Mammalogy. HerBert Lane, Assistant in Mammalogy. ~ James P. Cuarin, A.B., Assistant in Ornithology. Leo E. Miuter, Assistant in Ornithology. VERTEBRATE PALAONTOLOGY. Henry Fairrietp Ossorn, LL.D., D.Sc., Curator Emeritus. W. D. Marttsew, Ph.D., Curator. Water Granger, Associate Curator of Mammals. Barnum Brown, A.B., Associate Curator of Reptiles. Wiiuram K. Grecory, Ph.D., Associate in Paleontology. Cuartes R. Eastman, Ph.D., Research Associate. ANTHROPOLOGY. Cuark Wisster, Ph.D., Curator. Puny E. Gopparp, Ph.D., Curator Ethnology. Rosert H. Lowir, Ph.D., Associate Curator. Hersert J. Spinpen, Ph.D., Assistant Curator. N. C. Netson, M.L., Assistant Curator. Cuartes W. Mean, Assistant Curator. M. D. C. Crawrorp, Research Associate in Textiles. GrorcE Birp GrinneELL, Ph.D., Research Associate in Ethnology. J. H. McGrecor, Ph.D., Research Associate in Anthropology. Louis R. Suurvan, A.M., Assistant in Physical Anthropology. Lesuz Spier, B.S., Assistant in Anthropology. ANATOMY AND PHYSIOLOGY. Ratpu W. Tower, Ph.D., Curator. _ Cuartes F. Herm, Assistant. ALESSANDRO Fassri, Research Associate in Physiology. PUBLIC HEALTH. CHARLES-Epwarp A. Winstow, M.S., M.A., Curator. T. G. Hutt, Ph.D., Assistant. ; PUBLIC EDUCATION. Grorce H. Surrwoop, A.M., Curator. G. Cuyps Fisusr, Ph.D., Associate Curator. Ann E. Tuomas, Ph.B., Assistant. BOOKS AND PUBLICATIONS. Rautpu W. Tower, Ph.D., Curator. Ipa RicHaRpson Hoop, A.B,, Assistant Librarian, CONTENTS OF VOLUME XXXVI. THE DISTRIBUTION OF BIRD-LIFE IN COLOMBIA. Hnitroduction:, fod Tasha oR t aalons Reeve nedanen eae aeeedeee sae Acknowledgéments:. .2:¢:00cotseeadenenank gy it tec Gn cag ve enone geen Gey A Review of Colombian Ornithology. ............00000 00 cece cece cece Bogots, Collectionss . cra .serses see Gadus see ea kee eee Stee oka The ‘Bogotdé’ Region. ........0....0.0.0 006. cee teen Claude Wyatt’s Explorations....................500 000000 c ue : Berlepsch on a Bucaramanga Collection................ ......0.... Wirt Robinson on the Magdalena......................00. 2 hide Rona Salmon’s Collections in Antioquia.................. 0000 c cece e ee DeLattre in Western Colombia.................00..00 cece eee, ; The Michler Expedition to the Atrato.............0.....00 0000000004 Sundry West Colombian Expeditions...................0.000000.. : Mervyn G. Palmer’s Collections............0..0.000 0660 e ees The Santa Marta Region.......... 0.0000. c ccc cece eee eee 2 The American Museum’s Expeditions in Colombia........................ Expedition No. 1: Buenaventura to the Cauca Valley: Reconnaissance, Cali to Giradot.................000. eae an adeky Buea i Beene Chey Expedition No. 2: The Popayan Region...................0..00020.. Expedition No. 3: Lower end of the Cauca Valley, the Quindio Trail, Cartago to San Juan River................. boca nate Bon Biante hee oteras Expedition No. 4: Cali to San Agustin.................. ..00. 0005. Expedition No. 5: San Agustin to the Caquetdé Region........... arses Expedition No. 6: Tumaco-Barbacoas................-....000.00005. Expedition No. 7: The Bogotdé Region...................0....00000. Expedition No. 8: The Antioquia Region................... _ Auxiliary Collections.......00 000.002 An Outline of Colombian Topography......... 0... ...00 000s eee cece ees Remarks on the Distribution of Forests.........0.0 0.0.0. e eee eee eee Notes on Colombian Climatology............00 00.0. cece cece The Life Zones of the Colombian Andes..........0.. 0.000 c cece eee ees The Tropical Zone and its Faunas.....................0000, ee ere The Colombian-Pacific Fauna........0.0.0 000.0000 cece vee cece e ves The Cauca-Magdalena Fauna........ Sih iota Abana a anane eae haere The Caribbean. Fauna: :; .0025 02) sepe nace hee a ekares ouegasbeeesee The Orinocan Fauna...............0052020 eee tegany Ses fe ELE aA Ss The Amazonian Fauna................ 6a Sacaivare oma Macha Gata ouaien a oy vill Illustrations. Pace. The Subtropical Zone and its Faunas............0.000020 00 cee eee eee 135 The West Andean Subtropical Fauna...............0 00.00 eee eee eee 145 - The East Andean Subtropical Fauna.............. beeen eee 149 The Central American Extension of the Subtropical Zone.................. 151 The Temperate Zone... 0.0... eee teeta 159 Lhe; Parana: LON: pices 94:4 B04. oa REN deo acstd Sodus spite eed dle dea, Bese go seb hed Bau 166 Tabular Synopsis Showing Zonal Distribution of Families of Colombian Birds 168 Part II Classifications 26 c0ich ori cea peace pemisue ew mele err ew A Lane ee Gea Ee . 170 Number of Species Included... 0.2.0.0... 00000 cece eee eves 179 Forms Described as New... 200.0000 cece cece cece ce eneteneeennes 181 North American Migrants. ........0.0.0.0000 000 cece ccc ccc eee e ees 183 Sequence of Localities Cited........00.0.0.0.00 000 cece cece cece nas 185 PUCLCTEMCES fs Sellstac: d aitsd daca aes waded oaeetcawd-3 od SlasAn ogee, eonimen onde gd Galak des 185 English Names............... Eee Onle rene Han & te get acted idee BE Sean Sen Sater eee caibtt 186 Color: Terms 9 ices tess aio bare ik alee nee tein pide ok Peres ated Bea Tae oberon wt 186 Distributional List of Species and Subspecies......................00.005 187 Gazeteer of Colombia Collecting Stations.............0...0 0.0 0cc cee ee 640 Bib MOB a Phys xen. ct oat cts’ eae aap yesnad Alan MOAR ES ache KING Bee we ene A 657 CEO PM OIL g.3-cta dat Sah 3 ede ot sce omae ate eA em maw mh nae Oy wane ah ne 660 DANO iS nea Ca ag earns ey ane Reem tae Oa AU eas oo AEA eases nee Octet ale a 661 PuaTEs. No. FACING PAGE T.— Map of Colombia. ....00 cece eee ns Frontispiece II.— Fig. 1. The Upper Dagua near Caldas; Fig. 2. Lower Dagua (Arthur ; BN AMGTR) eel as stats rea aa tere ali ty ss Ba Png Gah Ot ctl 2 cel y le rater 8 20 IIl.— Fig. 1. Road between Caldas and San Antonio (Arthur A. Allen); Fig. 2. Forest at San Antonio (Frank M.Chapman).................. 22 IV.— Eastern side of Western Andes from San Antonio (Frank M. Chapman) 24 V.— Fig. 1. Western slope of outer ridge of Central Andes; Fig.2. Eastern slope of same ridge as above at same altitude (Frank M. Chapman)..... " 26 VI.— Fig. 1. Cauca River near Buga; Fig. 2. Forest on the Cauta River at Rio Frio (Leo E. Miller).....00.0.00 0000.0 ce ce eeeee 26 VII.— Fig. 1. Santa Isabel from Laguneta; Fig. 2. Laguneta (Frank M. Bae a Yt one a sch Ss oe g aoa egba aden now dean ees sores clnn 8 - 28 VIII.— Fig. 1. The Quindio Trail; Fig. 2. The Quindio Trail (Frank M. : Cha pre ny > 2 x sq cree sorte ais a oeeare ed ads deen ona cae hos me Sch ak a ee 28 IX.— Fig. 1. Railroad between Honda and La Dorada; Fig. 2. Plains of Tolima (Frank M.Chapman)..............00 00.00 cceeeeeseu cece. 30 X.— Fig. 1. Crest of Western Andes, west of Popayan; Fig. 2. La Gallera, Western Andes (Leo E. Miller),........ 00. ce eee eer eves eee vanes 32 Illustrations. No. XI.— Fig. 1. Boquilla Valley from Salento; Fig. 2. Stream near pelento (Arthur AyAllen yi is ean afi tad waa Pace Miacea whi oun mavawer ner ee XJI.— Fig. 1. Chicoral Bridge; Fig. 2. Giradot, upper Magdalena River (Frank M. Chapman)..........0.0.0 000000 c cece e eens XIT.— Fig. 1. Paramo of Santa Isabel, Central Andes; Fig. 2. Paramo of Santa Isabel, Central Andes (Leo. E. Miller).....................4. XIV.— Fig. 1. Between Buenaventura and San José (Arthur A. Allen); Fig. 2. Juntas de Tamand (Leo. E. Miller)......................005- XV.— Map of Central Western Colombia.................0060 eevee eee XVI.— Fig, 1. Near the Source of the Magdalena River; Fig. 2. Los Chor- rillos, above Almaguer (Leo E. Miller)..................000 00 0c eee XVII.— Fig. 1. Coast near Carthagena (Leo E. Miller); Fig. 2. Shores of the Lower Magdalena River (Frank M.Chapman)..................-. XVITI.— Fig. 1. Central Lower Magdalena River; Fig.2. A Wood Yard in the Magdalena Forests (Frank M.Chapman).....................05. XIX.— Fig. 1. Slopes above Bogotdé; Fig. 2. The Environs of Bogotd (Frank M. Chapman).......... 000.0000 cece eects XX.— Fig. 1. Eastern Andes between Bogoté and Chipaque; Fig. 2. Chip- aque (Frank M. Chapman)............. 0.0 ccc cece cette ne XXI.— Rio Negro from Monteredondo; Fig. 2. Junction of Rio Cdqueza and Rio Negro (Frank M. Chapman)...................0020 eee eeee XXII.— Fig. 1. Country near Sta. Elena, Central Andes; Fig. 2. Western Andes near Antioquia (Leo E. Miller)................. 0.000 c eee ae XXITII.— Fig. 1. The Paramillo, Western Andes; Fig. 2. Characteristic Vegetation on the Paramillo (Leo E. Miller)........................ XXIV.— Fig. 1. Alto Bonito, Rio Sucio; Fig. 2. Rio Cauca at Puerto Valdivia (Leo E. Miller).........0 0000. ccc ccc eens XXV.— Distribution of Forests in Colombia..................000 00000 XXVI.— Life Zones and Faunas of Colombia...................0 000 eens XXVII.— Western Slope of Central Andes from La Manuelita (Frank M. (CHADIAN) 5. cAh os dansk he cect as AURA en eANels aed Ae Ae AG Rae eB inhae hG XXVITI.— Fig. 1. Farallones of Cali, Western Andes; Fig. 2. Farallones of Cali, Western Andes (Frank M. Chapman).....................00% XXIX.— Fig. 1. Cauca Valley from San Antonio; Fig. 2. Cauca Volley: from Miraflores (Frank M. Chapman)....................000 eee cues XXX.— Fig. 1. Near Villavicencio; Fig. 2. Near Villavicencio (Frank M. CHAPMAN) ciek Arca ae peatie tessa Cuca Ad Heck Ge Dance ciehen sawn ee oo XXXI.— Subtropical Forest (Frank M. Chapman).................4... oe XX XII.— Heart of the Central Andes (Frank M. Chapman)............... XXXTIT.— Rio Negro Cafion near Monteredondo (Frank M. Chapman). . XXXIV.— Fig. 1. Primeval Forest at Buena Vista; Fig. 2. Forest Interior at Buena Vista (Frank M. Chapman)................ 0000s XXXV.— Figs. 1 and 2. Characteristic Trees of Temperate Zone Forest (Frank M. Chapman)......... 0.0.00: c cece eect eee n eee e eens XXXVI.— Wax Palms (Frank M. Chapman)................... 0.6: ..008 XXXVII.— Fuertes’s Parrakeet, Hapalopsittaca fuertesi (Chapm.) (Drawn by by As Puertes: hic aay coon bees aee cece oe See Ses BE PEE RE Dee BEE Eee XXXVIII.— Bills of Toucans (Drawn from fresh specimens by L. A. Fuertes) 94 96 126 132 138 140 148 150 160 258 264 328 x List of New Names. No. Pace. XXXIX.— Miller’s Ant Pitta, Grallaria milleri Chapm.; Allen’s Ant Pitta, Grallaria alleni Chapm. (Drawn by L. A. Fuertes.)............0.2545- 396 XL.— Black-headed Finch, Atlapetes fusco-olivaceus Chapm.; Yellow-headed Finch, Allapetes flaviceps Chapm. (Drawn by L. A. Fuertes.)......... . 574 XLI.— Key Map to Colombian Collecting Stations...................005. 656 Text FIGurss. 1. Life Zones of the Colombian Andes..............00.. 0000 cece eee ee 86 2. Ideal section through the Ecuadorian Andes to show zones of weiptation 87 3. IXnown range of Sapayoa enigma, a characteristic species of the Colom- bian-Pacific Fauna of the Tropical Zone........................ 107 4. Range of Zarhynchus waglert... 0... eee 108 5. Known range of Osculatia...... 0000 lil 6. Distribution of the western races of Manacus manacus...... ......... 112 7. IJsnown range of Arremonops conirostris......... 000000 ee 113 8. Ranges of Capito squamatus and C. maculicoronatus................... 114 9. Known distribution of Micromonacha lanceolata..................0... 118 10. Known range of Thamnophilus nigriceps.......00.00.000000 06 cece 119 11. Range of Ostinops decumanus......0000 00000 120 12. Known range of Myrmeciza exsul..........0.00000 000 ee a? Bpegee 121 18. Range of Donacobius atricapillus........ 000.000. 00c cc Sieg E23 14. Range of Thraupis palmarum......00 00 eee 125 15. Distribution of the Cock-of-the-Rock.........0......00..0000 ccc cee 137 16. Distribution of Formicarius rufipectus..... 0.0.00... cee ee 147 17. Distribution of Buarremon brunneinuchus. 0.0.0.0... 000 00 cee eee 152 18. Distribution of Atlapetes gutturalis.....00 00000 eee eee 154 19. Semi-diagrammatic representation of the range of Scytalopus niger.... 162 20. Known distribution of Pyroderus scutatus.......0..0.00 cece cece. 177 21. A probable case of Hybridism. Ranges of Ramphocelus icteronotus and AS ve UOS esas 20a. ons sntal aie cee Wosrattt aoe cits ersezy ty Meaeate Ahi geet .. §©611 LIST OF SPECIES AND SUBSPECIES DESCRIBED OR RENAMED IN THIS VOLUME. ; Pace. Zenaida ruficauda antioguie Chapman................. 0.00 cc cece cece eee 207 Phethornis striigularis subrufescens Chapman........................-... 283 Helianthea celigena ferruginea Chapman............... 0000. e ccc cece 298 Vestipedes paramillo Chapman..... 0.000.000. ccc ccc cece cece ee 301 Brachygalba fulviventris caquete Chapman.......................-----2.. 338 Pittasoma harterti Chapman....... 0.000.000.0000 cece cece cece. 392 Pipra leucocilla minimus Chapman........0...00. 000. c eee eee eee eee 480 Troglodytes musculus neglectus Chapman................00. 0000000200 520 Henicorhina prostheleuca albilateralis Chapman.........................-. 524 Cyclarhis flavipectus parvus Chapman.........0.0.00 00.00 ccc ccc eee ee cuee 540 BULLETIN OF THE AMERICAN MUSEUM OF NATURAL HISTORY. VOLUME XXXVI, 1917. THE DISTRIBUTION OF BIRD-LIFE IN COLOMBIA; A CONTRIBUTION TO A BIOLOGICAL SURVEY OF SOUTH AMERICA. By Frank M. CHapman. Part J. INTRODUCTION. SYNOPSIS. Part I. Introduction Acknowledgments A Review of Colombian Ornithology The American Museum’s Expeditions in Colombia ‘ Auxiliary Collections An Outline of Colombian Topography Remarks on the Distribution of Forests Notes on Colombian Climatology The Life Zones of the Colombian Andes The Tropical Zone and its Faunas The Subtropical Zone and its Faunas The Temperate Zone The Paramo Zone Bulletin American Museum of Natural History. [Vol. XXXVI, Parr II. A Distributional List of the Birds Collected in Colombia by the American Museum’s Expeditions APPENDIX, A Gazeteer of Colombian Collecting Stations Bibliography INDEX. . ae pe 1917.] Chapman, Distribution of Bird-life in Colombia. 3 Part I. INTRODUCTION. Our knowledge of the animal life of regions remote from the older centres of learning has been acquired through essentially similar channels. The casual specimens brought back, in whole or part, as curios by early ex- plorers, missionaries, travelers and adventurers were, in some instances, followed by shipments of the pelage or plumage of those species having a commercial value. Material of this kind was generally collected by natives and was lacking in data. Later came the exploring naturalists and profes- sional collectors. When not members of an expedition designed to enter some hitherto unknown region, they at first found near the pathways of trade vast territories as yet zodlogically unknown. It was only when the faunas of the regions reached by these long existing, if little traveled, routes failed to yield further novelties, that naturalists penetrated into less accessible places which, for some reason, had not lured the prospector, trapper or trader. These purely natural history expeditions have, as a rule, gone out to discover new species. Collections were made at widely separated localities with the double object of avoiding duplication of material, and of securing forms which had not before been taken. _ As long as large areas remained unexplored it is natural that we should desire a knowledge of their animal life. But having acquired this knowledge, it is also natural that we should wish to solve the problems which arise from its possession. Thus, through the sources mentioned, we now have so complete a knowledge of South American bird-life that it is not probable further*exploration will reveal any considerable number of distinct species. Tn short, we have now reached that stage in our study of the South American ornis, when, the search for species over, we may attempt to learn something of the. habits, racial variations and geographic distribution of the between four and five thousand birds known to inhabit that country. Acting on this belief, the American Museum of Natural History in- augurated in December, 1910, an intensive zodlogical survey of South America. For the present the work of this survey is restricted to the col- lecting of birds and mammals and of information concerning them and the country they inhabit. Our ultimate object is the discovery of the geo- graphic origin of South American life, but it is understood that this major problem cannot be successfully approached until we have a far more definite knowledge of faunal areas in South America than exists at present. 4 Bulletin American Museum of Natural History. [Vol. XXXVI, This knowledge is not within our reach until we have a much larger num- ber of specimens than our museums now contain. ‘These must be collected, not at widely separated localities, but at as many stations as are necessary to represent the principal physiographic and climatic areas contained in the range of the species. Our expeditions were instructed to make as complete a collection of the birds at each station as circumstances permitted. The commoner, more widely distributed species are more apt to reflect environmental influences than rarer ones of limited distribution, and are often, therefore, of more scientific value. One unfamiliar with the problems involved might imagine that we have accumulated an unnecessarily large number of specimens.!. But I regard each specimen as standing for a concrete fact. It places beyond dispute the occurrence of its species at a definite place on a certain date. The con- dition of its sexual organs helps to determine the relation between season and period of reproduction; its external characters enable us to distinguish between individual variations of sex, age and season, and those which result from environment and mark the nascent species. The bird-life of Colombia is probably as well known as that of any other part of tropical America of similar extent, but one has only to read the ‘Review of Colombian Ornithology,’ presented beyond, to realize how wholly inadequate for the ends in view, were the existing data in regard to the distribution of birds in Colombia when we began our work there. To determine the boundaries of zones and faunas as they are manifested by birds and mammals is our first aim, and in the course of this work we trust that our study of purely local conditions will at times so closely connect cause and effect, that we may throw some light on the laws governing 1 Lest we be accused of needless sacrifice of life, it will be well to state that our collections are far from sufficient satisfactorily to settle all the questions of speciation and distribution raised by our studies. C . From the standpoint of bird protection, the number of specimens taken has produced about as much effect on Colombian bird-life as would the collecting of the same number of plants have on the Colombian flora. The results of general collecting on the avifauna of a region as a whole are always negligible. It is only when the collector’s attention is focused on a certain species that its numbers are appreciably diminished. A milliner’s agent, for example, whom I met in Mendoza, Argentina, told me that he, alone, had sent the wing and tail-quills of 16,000 Condors to Paris! All were killed in the Argentine Andes where, in consequence, the species has become comparatively rare. On the other hand, eighty years of general collecting for millinery purposes in the Bogoté region has not, so far as we could observe, seriously affected the numbers of birds inhabiting it. Our expedi- tion No. 7, in passing from the Magdalena Valley over the Eastern Andes to Villavicencio, and hence through the heart of the Bogot4 region, secured over five hundred species of land-birds in some two months’ collecting, a number which clearly indicates the richness of the avifauna. Nevertheless, from this region, as stated above, hundreds of thousands, perhaps even millions, of birds have been shipped to European dealers. In view of these facts, it is hardly necessary to add that our average of twelve specimens per species has not perceptibly reduced the bird-life of the wide area over which we worked! 1917.] Chapman, Distribution of Bird-life in Colombia. 5 the origin of species and the distribution of life. It is also hoped that the technical reports on our large collections may be acceptable to the systematic ornithologist. Colombia was selected as our first field of operations, not because we believed it to be zodlogically the least known part of South America. On the contrary, so far as birds are concerned, the trade in native-made, ‘Bogoté skins’ has doubtless resulted in a greater number of specimens being shipped from Colombia than from any other part of South America. Colombia was chosen, therefore, because of its proximity, because cir- cumstances! had already aroused our interest in its avifauna, because lying at the base of ‘the Isthmus of Panama it is also at the crux of the problem of intercontinental relationships, and because it possesses more diverse physiographic and climatic conditions, combined with a greater variety of animal life, than any other part of South America of similar extent. The intensely humid Pacific, and arid Caribbean coasts, isolated Cauca and upper Magdalena Valleys, widespreading Amazonian forests and no less extensive llanos, three distinct mountain ranges and insular mountain mass of Santa Marta, each with four zones of life, give exceptionally wide scope for the manifestation of biogeographic phenomena in Colombia. From December, 1910, to April, 1915, we have had from one to six collectors in Colombia, crossing and recrossing the mountains and travers- ing the intervening valleys in pursuance of a carefully planned survey, designed to extend from sea-level to snow-line, and from the Pacific coast to the tributaries of the Amazon and Orinoco. At the outset we were impressed by the absolute necessity of determining - the level, as it were, at which a species flows before we could hope to dis- cover whence it came and whither it is going. A study of the distributional problems presented by Colombian bird- life, based on a collection of specimens from unknown altitudes, would lead to as inaccurate and confusing results as would the study of a collection of fossils from unknown geological formations. The differences between the bird-life of the Tropical and Temperate Zones, for example, are equally important whether occasioned by latitude or altitude. No one would think of removing the labels from specimens collected on the Amazon and in Argentina and then writing of them as having all been taken at one locality. But it-would be no more improper to do this than to write of the distribution of bird-life in the Eastern Andes of Colombia on the basis of a collection of native-made ‘Bogota’ skins. As a result of our labors, we are now in possession of approximately | 1° 15,775 birds and 1600 mammals, all carefully labeled with locality and 6 Bulletin American Museum of Natural History. [Vol. XXXVI, altitude, as well as many field-notes on distribution.! To these data the writer can add information gained on two expeditions which have led him across the three ranges of the Colombian Andes, from Buenaventura on the Pacific coast to Villavicencio at the eastern base of the Eastern Andes. Not only does a field experience acquaint one with the country, and all that such personally acquired knowledge implies, but it gives one a supply of negative facts which the most extensive collections cannot fur- nish. While specimens show where a species does occur, they fail to tell where it does not occur, and the latter fact is quite as important as the former. But when one is reasonably familiar with the appearance, espe- cially in life, of the birds of a country, not only the presence but also the absence of the more common or conspicuous species is noted. The alti- tudinal ranges of those most easily observed can be determined with more or less accuracy even from horseback as one travels slowly through the mountains. Climbing to the summit of ridge after ridge, and descending to the floor of the valleys between them, species appear and disappear at certain altitudes with a regularity which enables one to predict with more or less certainty when they will be found and when lost. Satisfactory determination of our Colombian specimens, and a true conception of the limits of those faunal areas lying only partly in Colombia required field-work in contiguous regions. Richardson, was, therefore, despatched to Ecuador where he collected some 4000 specimens, while with Anthony and Ball he secured 1800 specimens in eastern Panama. The Smith collection of birds from the Santa Marta region has also been of great service for comparison with our material from other parts of Colombia. The routes followed by our eight expeditions, and the localities at which we, as well as others, have collected, are shown on the map accompanying our Gazeteer of Colombian collecting stations; while full itineraries of each expedition are given beyond. It will be observed that our work has been restricted to what may be termed Andean Colombia. We have not attempted to penetrate the Amazonian forests beyond the upper Caquetd, or to explore the Ilanos east of Villavicencio. The uniformity of environmental conditions to the eastward of these points, in connection with our knowledge of Amazonian and Orinocan bird-life, warrant the belief that we should not find eastern Colombia to possess any marked faunal characteristics not shown by ad- joining regions in the same zones. The Sierras of the upper Uaupes and upper Inrida doubtless offer habitats not afforded by the country from which they rise, but the explora- tions of Rice fail to show a higher altitude in these mountains than 2850 1Cf. Bull. A. M. N. H.. XXXI, 1912, p. 189. 1917.] Chapman, Distribution of Bird-life in Colombia. 7 feet, leaving their summits therefore, so far as known, in the Tropical Zone. The zodlogical exploration of these Sierras, is, however, greatly to be desired. We have done no collecting in the Eastern Andes north of Cundinamarca, since the papers of Wyatt and Berlepsch indicate that this region has no faunal features which are peculiar to it; but we do feel the need of exact data, particularly in regard to the distribution of forests, from the extreme northern end of this range.’ In Antioquia we have felt compelled to duplicate to some extent the work of Salmon, especially in the Tropical Zone, which, lying in the region where Pacific coast and east Andean faunal elements meet, occupies a posi- tion of much importance. The Santa Marta group affords a closely related but independent problem to the one we have attacked, and its solution may well be left in the experienced hands of Mr. M. A. Carriker, Jr., whose six years’ residence in this region has given him exceptional opportunities for the continuous study of its life. Even with the restrictions named, the territory to be examined is so large, its topography so varied, its fauna so rich, and much of it is so com- paratively inaccessible, that we have covered it only superficially. But the resources at our command, and the extent of our ultimate plan, have made it imperative that we should make only a reasonably thorough recon- naissance of this part of the field, if we would hope to advance our study of the major problems involved in other parts of South America. It was a constant source of regret to us that we were not accompanied by a botanist who might have collected and identified at least the more characteristic plants of each zone and fauna. I feel, however, that the conclusions reached, based wholly on birds, have, in some respects, a greater value than if they had been based on the combined study of birds and plants. In their present form they constitute an independent contribution to zodgeo- graphy, solely from the standpoint of ornithology. The final determination of zonal, faunal and floral boundaries, will, in my opinion, be reached by the combination of similar independent contributions from the botanist and students of other branches of the animal kingdom. Meanwhile, comparison of the results here presented with those given by Wolf (Geo- grafia y Geologia del Ecuador; see beyond) based only on plants, shows a most assuring agreement. In this connection I desire to state with emphasis that the maps and profiles accompanying this report are not assumed to possess more than semi-diagrammatic accuracy. Colombian physiography is still too im- perfectly recorded to supply a base map on which faunal data might be entered. It is, indeed, so indefinitely diversified that our entire time in Colombia might have been devoted to a single mountain range and still ’ 8 Bulletin American Museum of Natural History. [Vol. XXXVI, not have given us the information needed to map its zones and faunas with a thoroughness which would ‘begin to express all the facts and factors in- volved. We must, therefore, leave to future workers the task of filling in the details of our work in Colombia, with a hope that they will find the zonal and faunal boundaries here proposed at least fundamentally correct. ACKNOWLEDGMENTS. The American Museum gratefully acknowledges the courtesies extended to it by the Colombian Government through its representatives in this country, and its administrators at the ports of Barranquilla, Buenaventura, and Tumaco. The consideration shown us by these gentlemen, and the promptness with which our equipment and supplies have been admitted to their country has materially advanced the objects of our expeditions. To the representatives of our own country in Colombia we are also indebted for many courtesies. We have received so many favors from individuals that properly to acknowledge them would require the enumeration of the names of the scores of persons on whom, for one purpose or another, we have called for assistance. Our requirements were often so unusual, or, to us, so pressing, that to meet them was frequently not a matter of their money value, and we had then to rely upon the generosity and good-will of those on whom we were in truth dependent. In this connection we are particularly under obligation to Mr. Chas. J. Eder, of the beautiful sugar estate La Manuelita, near Palmira in the Cauca Valley. After entertaining our first expedition at his home, Mr. Eder not only placed his bungalow ‘Miraflores,’ in the mountains, and ranch ‘Guengiie’ at our disposal, but supplied us with mules to make the journey from one to the other. We.have also to thank Mr. Eder for many personal attentions which added materially to our comfort, and for letters to his agents in various parts of Colombia. Mr. Henry Eder, of the firm Eder & Co., at Cali, acted as our forwarding agent during the year or more‘of our work in the Cauca region and through his efficient service we completed our labors without the loss in transporta- tion of a single specimen or item of equipment. In Bogota, Mr. F. L. Rockwood has acted in a similar capacity for. cer- tain small collections acquired since our expedition left that region. For these we have mainly to thank Hermano Apolinar Maria, Director of the admirably arranged museum of the Instituto de la Salle. 1917,] Chapman, Distribution of Bird-life in Colombia. 9 Hermano Apolinar has presented us with numbers of specimens, and has secured for us additional specimens of species. not satisfactorily repre- sented in our own collections. Notably, Cistothorus apolinari and Asio accipitrinus bogotensis. We are also under obligations to Mr. D. C. Stapleton, Mr. Charles Miller, Dr. Hamilton Rice, Gen. Rafael Santos, Sr. Jesus Velez, and Mr. Mervyn S. Palmer. ' We should indeed be lacking a sense of appreciation if we did not express our gratitude to the people of Colombia with whom at one time or another and in a thousand nameless ways, we have come in contact. From the peon by the wayside to the owners of haciendas one and all have shown us ‘the most courteous attention. When traveling through remote, unsettled regions with a valuable outfit and often considerable sums of money, we have felt as safe (possibly safer!) than when in our own homes. When in camp or at hotels, country inns or posadas, we made no special provision for guarding our equipment and supplies; nevertheless, during the five years of our work we did not suffer the loss of a single item by theft. Indeed, on passing through a certain village where one of our party had previously worked, we were stopped by a native bringing a needle and thread which had been left behind! But especially do I desire, so far as mere words will permit, to pay a tribute to the men with whom it has been my privilege to be associated on our zodlogical explorations in Colombia: To William B. Richardson, Louis A. Fuertes, Leo E. Miller, Arthur A. Allen, George K. Cherrie, Paul G. Howes, Geoffroy O’Connell, Thomas M. Ring, and Howarth Boyle. To their untiring enthusiasm and whole-souled devotion to the Museum’s interests may be credited the most valuable collections of birds and mam- mals which have been brought from any part of South America. To Richardson, veteran among collectors in tropical America, was given the exceptionally unhealthful stations on the Pacific coast. Here he suffered from fever and from beri-beri, but with the amazing vitality which has carried him through thirty years of exposure to tropical diseases, he con- tinued work when most men would have succumbed. Miller, a novice on our first expedition, showed such resourcefulness, energy and persistence in overcoming the difficulties which are the neces- sary accompaniment of collecting in the tropics, that he was subsequently selected as one of the Museum’s representatives with the Roosevelt Brazilian Expedition. His work during the rainy season in the humid Amazonian forests of the Caquet4, where with only unskilled native assistance he secured 830 birds and mammals in 30 days is a feat in tropical collecting; while his 10 Bulletin American Museum of Natural History. [Vol. XXXVI, ascent of the Paramillo is our most difficult and noteworthy piece of actual exploration in Colombia. On this latter trip he was ably assisted by Howarth Boyle. Cherrie’s extended experience in South America made him an invaluable associate on our trip in the Bogotd region. Particularly effective as a collector he was no less efficient in dealing with those details of transporta- tion and subsistance which form so important a part of field-work in thinly settled regions. He, too, was chosen as a representative of the Museum on the Roosevelt Expedition. Allen’s admirable descriptions of the country which he explored show how well qualified he was for work of this kind, and serve to double our regret that an illness contracted in the unhealthful Chocé region, should have necessitated his return to New York just as he was approaching the most productive part of Colombia. Howes, O’Connell and Ring made up in enthusiasm what, at first, they lacked in experience, and to them we owe many specimens of birds and mammals which would not otherwise have been obtained. I am sure that no other member of our various Colombian expeditions will feel that I am giving undue praise to any one member of it when I say that the best qualities each one exhibited were all present in Fuertes. Officially the artist of the expeditions with which he was connected, he filled, in truth, whatever position seemed most to require his attention. In looking for an opportunity to help others, he rivalled Cherrie, while his unbounded enjoyment of the experiences of his associates, as well as his own, made him an ideal companion. To the fellow-workers who have rendered me assistance in the prepara- tion of this paper, I am indebted no less than to those who have aided us in the field. For the loan of specimens used in comparison, I have to thank Dr. Chas. W. Richmond, of the United States National Museum, Mr. E. W. Nelson, of the Biological Survey, Dr. Witmer Stone, of the Academy of Natural Sciences of Philadelphia, Mr. Outram Bangs, of the Museum of Comparative Zodlogy, Mr. Thomas E. Penard, of Arlington, Mass., Mr. L. A. Fuertes, of Ithaca, N. Y., Mr. W. E. Clyde Todd, of the Carnegie Museum, and Mr. Charles B. Cory, of the. Field Museum. Mr. Phanor Eder, author of the authoritative work on Colombia in the Fisher-Unwin series, has supplied me with numerous references to the literature of Colom- bian exploration and loaned me from his extensive Colombian library, a number of works not elsewhere available. Professor Isaiah Bowman, Director of the American Geographical Society, has given me access to the Colombian maps in his charge, and supplied much of the data on which the map of Colombia accompanying this paper is based. 1917.] Chapman, Distribution of Bird-life in Colombia. 11 My assistant, Mr. Waldron DeWitt Miller has given me the benefit of his advice in many knotty problems, and to Mr. David S. Ball and Mrs. Alice K. Fraser, of the Department of Birds, I am also under many obliga- tions. Mr. Ball made the preliminary identifications of the Hummingbirds, while to Mrs. Fraser has fallen the clerical labor, comparison of refer- ences, proof-reading, indexing, etc., incident to the preparation of a report of this kind. Additional assistance of a more specific nature is acknowledged in con- nection with the instance in which it has been given. A REVIEW OF COLOMBIAN ORNITHOLOGY. ‘ Bogoté’ Collections Eighty years had passed since the publication of the tenth edition of Linnzus’ ‘Systema Natura’ before naturalists began to draw on the ornithological treasures of Colombia which, after eighty years more, are still unexhausted. It was apparently in 1838 or 1839 that a French collector, resident in Bogota, began to send birds’ skins to Paris. These came to the attention of Boissoneau, Lafresnaye, Des Murs and Bourcier, who described many of them as new in the pages of the ‘Revue Zoologique’ and ‘Revue et Magazin.’ Native collectors soon learned how to prepare skins which, in increasing numbers, were sent to Paris, and, apparently as early as 1840, reached London, since Fraser described several new ‘Bogota’ birds in the Proceedings of the Zodlogical Society for .1840. So large were the shipments of birds from Bogoté that in 1855 Sclater, from whose paper we learn these facts, published in the Proceedings of the Zodlogical Society, a list of 435 species personally known to him from the Bogot& region. Many of these were species of wide distribution, others were migrants from North America, but of the remainder no less than 180 had been described from “New Grenada,” as the country was then called, chiefly from the Bogota region, and of these some seventy were first made known by Lafresnaye. In, 1857 (P. Z. S., pp. 15-20) Sclater published an addendum which added 52 species to his previous list making 487 which at that time were known from the Bogota region. Since that date hundreds of thousands, possibly even millions, of birds, collected primarily for millinery purposes, have been shipped from Bogota, in the main to London and Paris. This trade probably reached its maxi- mum about 1885, when the fashion of wearing small birds on hats was at its height, but with a change in style which created a demand for the a 1917,] Chapman, Distribution of Bird-life in Colombia. 13 Bogota and in the Llanos to the east by T. H. Wheeler, but apparently no special report has appeared on this gentleman’s labors, and I am unaware of their scope, but it seems probable that many of his specimens were collected by natives. In 1899 Dr. Witmer Stone published a report (Proc. Acad. N. S. Phila., 1899, pp. 302-313) on some 77 species collected by Dr. J. W. Detwiler, chiefly from Honda to Ibagiie, and this short paper appears to be the only one which has been issued on scientifically collected birds from the Bogota region; but even this collection evidently contains many native;made skins. There are doubtless few regions in the world where accuracy in labeling specimens is of more importance than in that area whence came the so- called ‘ Bogota’ skins. In its most restricted sense this area, extending from the Magdalena Valley on the west to the base of the Andes on the east, contains four life-zones and two distinct basal faunas. While a dataless specimen may help indicate the character of the bird-life of the region as a whole, it throws no light on faunal or.zonal limits or on geographic variation under the strikingly different environmental conditions which prevail in this part of Colombia. Not only does the absence of data, particularly of altitude, make Bogot4 skins of no value in determining the limits of zones and faunas, but in many instances it has been discovered by comparison with fresh material, that the old, native-made skins have undergone so striking a change in color that they fail to represent properly the species to which they belong, and for purposes of exact comparison they are there- fore not only worthless but misleading. I shall make no attempt to list in detail the many papers consisting merely of descriptions of new birds based on Bogoté skins. Our own brief explorations show that new species are still to be found in sight of the city of Bogota itself, and for many years there will no doubt continue to be additions to the list of recognized species to which the type-locality ‘ Bogota’ is ascribed. The Limits of the ‘Bogoté’ Region— While the known ranges of the birds contained in even the earliest Bogoté collections make it evident that the native collectors worked at comparatively great distances from the city of Bogot4, so far as I am aware no definite information of the regions visited by them has been published. The following facts were gathered from dealers and collectors in the city of Bogoté during our seventh Colombian expedition: The majority of the birds’ skins brought in by natives are collected by them within twenty-five miles of the city. Fusugasug4 to the south, Anolaima, at the border of the Subtropical and Tropical Zones, to the northwest, and Choachi and Fémeque on the eastern slope of the first 14 Bulletin American Museum of Natural History. [Vol. XXXVI. range of the Andes east of Bogot4, are localities from which many speci- mens are now received. At a greater distance, the region about Villavicencio has supplied a vast number of skins. This city is the gateway of the trail to Bogota toward which, in default of an eastern outlet, the current of trade from the Llanos flows. Villavicencio thus draws on the region east to the Casanare, and’ south and east to the rubber producing forests of Amazonian Colombia. Transportation facilities and commercial relationships, therefore, make Bogoté the market for the products of the vast region lying to the east of it, and for this reason eastern Colombia has supplied a far greater number of birds’ skins than the region west of Bogota, where transportation to the marts of the world may be secured without the passage through Bogoté required by the products of the east. Nevertheless, the demand for skins by the Bogoté dealers has brought specimens from as far north as southern Santander, from west at least as far as Ibagiie at the entrance to the Quindio trail over the Central Andes,! and as far south as the head of the Magdalena Valley at San Agustin. It was here, that in April, 1912, Leo E. Miller found a native collecting with his blow-gun about forty Hummingbirds a day for a Bogotad dealer, as above related. It is apparent, therefore, that in exploring the Andes from base to sum- mit and working both to the east and west of the Eastern Range, the Bogota collectors have pursued their calling in four life-zones and two quite unlike faunas. Nevertheless, for the past seventy-odd years, ornithologists have used these Bogoté specimens in defining the characters and distribution of birds without knowing whether they came from the Magdalena Valley or the headwaters of the Meta, from the Tropic or the Temperate Zone. Even when used in a broad sense, the locality ‘Bogot4’ has come to have a far more definite meaning than, in view of the facts above recorded, should be given it. With the wider-ranging species it is obvious that Bogoté collections may contain specimens from far separated localities, and, in default of labels, it is often impossible to distinguish between geo- graphic and individual variation. In a number of instances our collections show that birds inhabiting both the western and eastern slope of the Eastern Andes, which have been sup- posed to represent one form, belong in fact to two, while in the case of the House Wren no less than three forms occupy the area which the most recent reviser of this group believed to be occupied by one. It seems not improbable that the least-known portion of the restricted 1 The type of Chamzpetes g. goudoli came from this region. 1917.] Chapman, Distribution of Bird-life in Colombia. 15 Bogoté region is the Savanna of Bogota itself. The comparatively limited number of birds found in this area has made it an unfavorable spot for the resident collectors who have, naturally, been more attracted by the richer avifauna of forested humid regions. Doubtless for this reason some of the commonest of true Bogotd species are comparatively rare in Bogota collections. ‘ During one morning I shot the types of new races of the Least Bittern, Short-eared Owl, and Yellow-headed Blackbird at Suba, within sight of the city. At the same locality Hermano Apolinar Maria secured for us specimens of distinct forms of Cistothorus and Habrura, genera which had not been previously reported from the Bogoté region. The Coot (Fulica) of the Savanna also proves to be a well-marked, undescribed form which has doubtless escaped the attention of earlier writers because of lack of material. The forested portion of the Magdalena Valley, from La Dorada north- ward, seems to have been but little visited by the Bogoté collectors who prefer the more healthful localities in the mountains to the hot, fever- infested river bottoms. Claude Wyatt's Explorations.— Aside from the native collections and the few birds secured by Wheeler and Detwiler, we have only three other sources of information concerning the bird-life of the Eastern Andes and country at their base. In January, February and March, 1870, Mr. Claude Wyatt made an ornithological reconnaissance in Santander. He left the Magdalena River at Puerto Nacional and proceeded thence by mule through Ocafia, La Cruz and Cocuta Suraté to Bucaramanga. From this point he ascended to the Paramo of Pamplona and returned to the Magdalena near what is now Puerto Wilches. He gives an excellent description of the country traversed, and the accurate data as regards locality and altitude at which he secured specimens of the 210 species he lists, makes his paper (Ibis, 1871, p. 118 ef seg.) one of real scientific value. Berlepsch on a Bucaramanga Collection.— In 1884 Count von Berlepsch published in the Journal fiir Ornithologie (pp. 273-320) a report on a collec- tion of some 800 bird skins, representing 150 species, which was sent him from Bucaramanga. These skins. were made by natives and were without data. It is probable that they came from the country immediately sur- rounding Bucaramanga, but beyond indicating in a general way the faunal affinity of this region with that of Bogoté, the collection possesses little value for distributional problems. Wirt Robinson on the Magdalena: —In 1895, Lieutenant (now Colonel) Wirt Robinson published a list of ninety-one species collected or observed by himself and his brother on a trip from Barranquilla up the Magdalena to Honda and thence to Guaduas, distant a day’s journey on the road to 16 Bulletin American Museum of Natural History. [Vol. XXXVI, Bogoté. He returned to Barranquilla over the same route. The expedi- tion occupied but a month (June 20-July 21) and afforded neither time nor opportunity for much field work. Nevertheless, the daily record of birds observed tells us what species enter into the everyday bird-life of this part of Colombia, and the records, as far as they go, are definite. It is to be regretted that Col. Robinson did not continue his journey to El Vergel, but two hours beyond Guaduas, where he would have found first-growth forest and an interesting fauna. Salmon’s Collections in Antioquia.— Proceeding to the west, we shall find that beyond the occasional mention of specimens secured by native collectors on the ‘Quindiu’ and in ‘Antioquia’, our exact knowledge of the bird-life of central Colombia has rested solely on the collections made by T. K. Salmon in the Department of Antioquia. These were reported on by Sclater and Salvin (P. Z. S., 1879, pp. 486-550) in a list of 468 species represented by about 3500 specimens. Salmon was an Englishman in the employ of the Colombian Govern- ment and lived at Medellin. His collections were made between 1872 and 1878, chiefly at and near Medellin, but he also visited the country as far west as Frontino, Antioquia, and Concordia, and as far south as Jerico, while to the north and east he reached Remedios, in the Tropical Zone, on the headwaters of the Ité, which flows into the Magdalena. His field, therefore, extended from the eastern border of the Atrato, to the western border of thes Magdalena Valleys. The locality “Sta. Elena’’ which appears so often in Sclater and Salvin’s list, and which they were unable definitely to locate, is situated a few miles east of Medellin, on the summit of the first ridge of the Central Andes between that city and the Magdalena Valley. Salmon was the first naturalist to make anything approaching a com- plete collection of the birds of a stated area in Colombia and his work is of high importance. Where his localities are not on or near the boundaries of life-zones the data accompanying his specimens are sufficient. His records from Remedios, for example, a station wholly in the Tropical Zone (alt. 2360 ft.) and at some distance from altitudes of sufficient height to support life of the succeeding or Subtropical Zone, are of much significance and give us our first, and, until the present time, practically only knowledge of the extension of Pacific coast forms into the Magdalena Valley. From Reme- dios, for example, Sclater and Salvin record Cyphorhinus pheocephalus, Thryophilus nigricapillus, Orthogonys olivaceus, and Capito maculicoronatus. When, however, his collections were made at localities where the precipitous nature of the ground and height of the mountains produced marked changes in altitude within short distances, Salmon evidently failed to appreciate 1917,] Chapman, Distribution of Bird-life in Colombia. 17 the necessity for exactness in labeling and his data, as published, are there- fore most misleading. To illustrate: from “Sta. Elena” Sclater and Salvin record among the Wrens alone, Thryophilus nigricapillus, Thryothorus mystacalis, and Cinnicerthia wnibrunnea, species which, respectively, are characteristic of the Tropical, Subtropical, and Temperate Zones and whose occurrence at one place, therefore, would be as remarkable as the successful cultivation ‘of cacao and wheat in adjoining fields! Many similar instances could be given; thus Troglodytes solstitialis, a species of the Temperate Zone, is recorded from “ Nechi” (sic), a locality in the Tropical Zone, and this inaccuracy destroys, in a measure, the value of the paper for distributional purposes. Taken, however, with what we have learned of the zonal distribution of Colombian birds, and particularly in connection with Miller and Boyle’s work (Expedition No. 8), Salmon’s paper gives us an excellent understanding of the avifauna of Antioquia. His notes on nesting-habits and carefully made collection of nests and eggs form a noteworthy contribution to our limited knowledge of the life-histories of Colombian birds. Delattre in Western Colombia.— In western Colombia small ornithological collections were made at least as early as 1846, when Delattre and Bourcier published in the Revue Zoologique descriptions of new Hummingbirds secured by the first-named author on a journey from Buenaventura through Juntas (= Cisneros) to Cali, Popayan and Pasto. Other birds collected by Delattre were described by Lafresnaye, but the total number of specimens secured by this early French traveller does not appear to have been very large. The Michler Expedition to the Atrato— Our first real knowledge of the character of Colombia’s Pacific coast avifauna we owe to Chas. J. Wood and Wm. S. Wood, Jr., who, as naturalists of the expedition under Lieut. Michler to discover a possible route for a canal from the lower Atrato to the Pacific, made a collection representing 144 species of birds on the lower Atrato, the Truando, and Nercua Rivers. This collection was reported on by Cassin in the Proceedings of the Academy of Natural Sciences of Philadelphia for 1860 (pp. 132-144, 188-197), and his paper still remains practically our only source of information of the bird-life of this part of Colombia. Of the four new species therein described by him, Pittasoma michleri, type of a new genus, is the most noteworthy. Sundry West Colombian Collectors.— In 1894, we learn from Hellmayr (P. Z. S., 1911, p. 1084), W. F. Rosenberg visited the region east of Buena- ventura working ‘chiefly at Juntas and Cali. His birds went to the late Adolphe Boucard, who published a list of the Hummingbirds in ‘The Hum- mingbird’ (Vol. V, 1895, pp. 5-7) but the bulk of the collection was never 18 Bulletin American Museum of Natural History. [Vol. XXXVI, reported on. In 1896 and 1897, Hellmayr writes, Gustav Hopke “sent a fair series” from the same district to Count Berlepsch who described several new species in the Ornithologische Monatsberichte, Vol. V, 1897, pp. 173- 176, and in Ornis, XIV, Feb. 1907, pp. 347, 361, 365. Mr. Eugene Andre, in 1899, Hellmayr continues, “forwarded a large collection of birds from the environs of Buenaventura and western slope of the Andes above that town, to Comte de Dalmas of Paris. Unfortunately, the greater part of it was subsequently destroyed by accident, and merely a list of Trochilide, by Messrs. Simon and de Dalmas (Ornis, XI, 1901, pp. 216-224),” is the only publication it produced. In February, 1898, Walter Goodfellow and Claud Hamilton landed at Buenaventura and traveled thence to Cali whence they proceeded, wa Popa- yan, the Patia Valley, Pasto, etc., to Quito. Such collections as were made in Colombia were lost in transit, but Goodfellow’s report (Ibis, 1901, pp. 300— 319; 458-480; 699-715; 1902, pp. 59-67; 207-233) on collections subse- quently made in Ecuador, contains an interesting description of the journey through Colombia with occasional observations on the birds observed. Mervyn G. Palmer’s Collections — Prior to 1910, the most important collections of west Colombian birds, however, have been made by Mervyn G. Palmer who collected in the region between Buenaventura and Cali in 1907 and 1908 and on the Upper San Juan and its sources in the latter part of 1908 and 1909. The birds believed to be new in the first-named collections were described by Outram Bangs in the Proceedings of the Biological Society of Washington for 1908 (pp. 157-161) and 1910 (pp. 71-76), but the main collection has’ not yet been reported on. The San Juan collection, numbering some 700 specimens of 201 species, : fortunately fell into the hands of Hellmayr whose paper on this material: (P. Z. S., 1911, pp. 1084-1213), prepared with an exceptionally wide knowl- edge of South American birds, is, if not the most extensive, at least the most satisfactory treatise on the birds of any part of Colombia with which I am familiar. From June 19 to July 2, 1904, W. W. Brown, Jr., representing John E. Thayer, collected vertebrates on Gorgona Island, which lies some thirteen miles off the shore of southwestern Colombia. Birds were rare both in species and individuals, examples of only fourteen species being secured. These with two others are reported on by Thayer and Bangs (Bull. Mus. Comp. Zoél., XLVI, 1905, pp. 91-98) who describe as new Sula etesiaca, Urubitinga subtilis, Thamnophilus gorgone, Cyanerpes gigas, and Cereba gorgone. The Santa Marta Region.— The Santa Marta mountains, because their isolation and altitude promised biological results of exceptional interest, 1917] Chapman, Distribution of Bird-life in Colombia. 19 have received more attention from exploring ornithologists than any other part of Colombia. They were first visited by F. Simons who, in 1878 and 1879, worked from sea-level to as high as 17,000 feet, and on both northern and southern slopes. His collections of 182 species formed the basis of papers by Salvin and God- man in ‘The Ibis’ for 1879 (pp. 196-206) and 1880 (pp. 114-125, 169-178). Simons was followed by the well-known American collector, W. W. Brown, who, in the interests of E. A. and O. Bangs, collected during the years 1897-99, approximately 2500 specimens representing some 242 species. A series of papers based on this collection was published by Outram Bangs in the Proceedings of the Biological Society of Washington, and of the ’ New England Zodlogical Club. Before Brown had left the region Herbert Smith entered it in charge of a party which planned to make a thorough study of its fauna. A seri- ous illness and prolonged revolution so interfered with Smith’s plans that he did little work above the Subtropical Zone. His collections, numbering nearly 3000 specimens representing 304 species, were purchased by the American Museum of Natural History and were reported on by J. A. Allen in the Bulletin of the American Museum for 1900 (pp. 117-183). Dr. Allen includes in this paper references to the 84 species secured by Simons and Brown but not by Smith, bringing the total number of birds known from the Santa Marta region up to 388. Since the year 1911, M. A. Carriker, Jr., who has had prolonged experi- ence in the American tropics, has been resident in the San Lorenzo moun- tains of the Santa Marta group and in the adjoining country, where he has made large collections of birds for the Carnegie Museum. W. E. Clyde Todd has described some of the species secured, and it is to be hoped that we may have a résumé of our knowledge of the exceptionally interesting bird-life of this group of mountains in which Carriker’s field studies may be employed to map its zones and faunas. No other part of the Andes has received such long continued attention from a trained collector. It appears, therefore, that aside from the Santa Marta group, and omitting reference to ‘Bogot4’ skins as of no value in an attempt to deter- mine with exactness the boundaries of life-zones and faunal areas, our knowledge of Colombian birds rests, in the main, on Wyatt’s three months’ explorations in the Eastern Andes of Santander, Salmon’s extensive col- lections in Antioquia, the work of the Michler expedition in the lower Atrato, and of Palmer on the San Juan and Pacific slope west of Buenaven- tura. It is obvious then, in view of these facts and the extent and topo- graphic diversity of the area to be covered, that we had before us a task of some magnitude when, in Noveimber, 1910, we began our field-work in Colombia. 20 “Bulletin American Museum of Natural History. [Vol. XXXVI, THE AMERICAN MUSEUM’S EXPEDITIONS IN COLOMBIA. In planning our field-work in Colombia we experienced much difficulty in securing definite information concerning means of transportation, routes, and the character of the country we proposed to visit. Aside from the use of the railways from Buenaventura to Caldas and Puerto Colombia to Barranquilla, and La Dorada to Honda, and of steamers and launches on the San Juan, Cauca, and Magdalena rivers, our work in Colombia has of necessity been conducted solely with the aid of pack animals and porters. Such limited transportation facilities in a country where topography and climate further add to the difficulties of travel, imply a lack of intercommunication between regions which, although contiguous, are separated by high mountain ranges with but few passes. We should not therefore, have been surprised often to find it impossible to learn from the inhabitants of one district even the most salient features of what to us seemed comparatively nearby districts. For this reason it has seemed to me to be desirable to publish at some length the itinerary of each of our eight expeditions in Colombia with a general description of the routes followed and stations at which collections were made. This information is presented not only for its bearing in the present connection, but as a contribution to Colombian geography. - Miller and Richardson in the Andes west of Popayan, Miller and Allen in the Paramo of Santa Isabel and in crossing from Cartago to Névita and Popayan to San Agustin; Miller in the Caqueté region and with Boyle on the Paramillo, have visited regions about which little or nothing has been published; while the narrative of those expeditions which followed more beaten trails, may have a practical value to those who, for whatever pur- pose, follow in our footsteps. That our explorations may be extended to advantage, is beyond ques- tion, for there still exist large areas in Colombia of which we know but little or nothing. The bird-life of Amazonian Colombia, probably richer than that of any other part of the republic, is known to us only through the results of Miller’s one month’s collecting in the Caquetd region; in the Llanos .proper there has been no scientific collecting; the character of the bird-life of the northern end of the Eastern Andes we know only by inference; no collections have been made in the Goajira Peninsula, and but few speci- mens have been recorded from the arid coastal region west of the Magda- lena. ‘The great Magdalena forests are still but imperfectly explored; the Central Andes south of Antioquia have been visited only by our expe- ditions; even the ornis of the Cauca Valley, as elsewhere stated, is not Butretin A. M. N. H. VoL. XXXVI, Prate II. Fig. i. Tar Upper Daaua near CaLpas (Tropical Zone; arid portion of the Cauca-Magdalena Fauna.) Fig. 2. Ture Lower Dacua (Tropical Zone; Colombian-Pacific Fauna.) 1917] Chapman, Distribution of Bird-life in Colombia. 21 satisfactorily known. Miller and Allen in their rapid crossing from Cartago to Névita took species not found by us elsewhere, while work in Tatamé Mountain and Cerro Torra in this region, the ‘Paramo’ of Frontino to the north, and Farallones of Cali to the south, would be certain to yield valuable results. The Patia Valley with its unique tropical connection with the Pacific coast, offers an unusual problem in zodgeography, while the Pacific coast itself is ornithologically unknown from the Patia to the San Juan rivers. Particularly, would I call attention to the need of further exploration in the Chocé region, and especially in the Baudé range and mountains on the Panama frontier. Expedition No. 1. Buenaventura to the Cauca Valley; Reconnaissance, Cali to Giradot over the Quindio Pass. November 10, 1910-June 4, 1911. Personnel.— Frank M. Chapman, Louis A. Fuertes, Wm. B. Rich- ardson, Leo E. Miller. Itinerary Richardson reached Buenaventura on the Pacific -coast, alone, on November 9, and proceeded at once to Caldas (alt. 2560 ft.) distant forty miles at the end of the railway under construction from Buena- ventura to Cali. He remained at Caldas until November 24, and thence retraced his steps some fifteen miles to San José (alt. 600 ft.) collecting there from November 27 to December 18. On the last-named date he left for Cali, at the eastern foot of the Western Andes, and this large, attractive city became our base of operations for the succeeding year. Collections were made about Cali until December 31, when Richardson moved to.the mouth of the pass (alt. 6600 ft.) in the Western Andes, 3100 feet above the town, and established himself at a wayside posada surrounded by forest. This locality is known as Las Cruces, from three large crosses which mark the divide, and also as San Antonio, from a small settlement just below the pass on the trail to Cali. It was at this point that Mervyn G. Palmer made part of the important collections purchased by Mr. Bangs; and at El Tigre, a ranch about 1500 feet below the divide and to the west, Eugene Andre collected. February 26, Richardson moved from San Antonio to Las Lomitas (alt. about 5000 ft.), a ranch on the Pacific slope some five miles to the northwest, and worked there until March 7. In order to be near the coast while awaiting the arrival of the remainder of the expedition, Richardson went to Los Cisneros (alt. 900 ft), also known as Juntas, at the junction of the Dagua and Las Petitas rivers, and the head 22 Bulletin American Museum of Natural History. [Vol. XXXVI, of canoe navigation on the first-named stream, and remained there until March 21. Two days later he reached Buenaventura where he was joined by Chapman, Fuertes and Miller. The whole party now went to Cali, and after depositing there a large part of their equipment and supplies, established themselves, March 29, very comfortably in a bungalow at San Antonio, immediately below. the forest which crowns the crest of the mountain. The collections made by us here in connection with those of Palmer, are believed to contain a large proportion of the birds which occur in this locality. April 8 we returned to Cali and on the 11th reached the sugar estate of La Manuelita in the Cauca Valley, some five miles north of Palmira. Here we were the guests of Mr. Charles J. Eder until the 18th, when with pack animals supplied by Mr. Eder, we moved to his bungalow, Miraflores, situ- ated on the western slope of the Central Andes, about 3000 feet above the valley, or at an elevation of some 6100 feet. May 1, Mr. Eder sent mules for us and, after a night at La Manuelita, we crossed the valley to Cali by way of Florida and Guengiie, stopping at the last-named rarich two days and reaching Cali on May 7. Some collecting was now done in the marshes of the Cauca River near Juanchito, the port of Cali. Fuertés, secured here our first specimen of Aythya nationi, a practical rediscovery of the species previously known only from a pair taken at Lima, Peru. May 13, Fuertes and Chapman began their return journey to New York in a reconnaissance down the Cauca to Cartago, thence over the Quindio Pass to Giradot on the Magdalena, and down that river to Barran- quilla, which was reached June 4. On the same date Miller and dichatdson left on an expedition to the Andes west of Popayan. On the whole, the work of this first expedition is believed to have given us a fair idea of the avifauna of the region covered. We regret now, how- ever, that no attempt was made to reach the Farallones of Cali, the highest point in the Western Andes, between the summit of the Micai Trail, west of Popayan, and the Citaré of Antioquia. With an altitude of between 9000 and 10,000 feet, it is possible that we might have found there some species of the Temperate Zone. Our Cauca Valley collections would also have been more satisfactory if we had explored a tract of primeval forest between Cali and Florida. Description of Route and Collecting Stations — The rainfall of the Pacific slope of Colombia is phenomenal. It has been known to reach 400 inches in one year at San José (see beyond). There is no dry season on the Pacific coast and it rains almost daily in this intensely humid belt. CRUNey URopuy Is9AQ ‘UOT PeRordoayqnys) CRuney vUaTepSVTT-voneyD oy Jo uoMaod pre ‘auozZ [eordo1y) SOATZ Jo} Udo JO IoT ay 7B LOG & JO ainsy oh puy Ulajso4, oO} JO sodojs usojysom soddn puv 4so10 pUL SYUNA}-001} PosSkoUd-SSOU OTL OY} SIVAOD YOTYM JSoIoy TRoIdoayqns oq} JO AMT, JaMmoyT ay} 9}ON ‘aT@os JO WON dsouoos out “AVIPTUUNY, aud} XI OY 99 OINOLNY NVQ LY Lsanog OINOLNY NVQ GNV SVQIVO N AMLAA AVOY THL T4¥1gd ‘TAXNNN “TOA ‘HON CIN CV Nisatiog 1917,] Chapman, Distribution of Bird-life in Colombia. 23 As a natural consequence the region is, as a rule, densely forested from the very margin of the sea to the summit of the Western Andes. Buena- ventura lies at the head of the bay of the same name, some fourteen miles from the sea. The shores here are lined with mangroves, and numerous small streams and estuaries make a network of mangrove-bordered water- ways. Buenaventura to Caldas.— Shortly after leaving Buenaventura, on the railway to Caldas, one reaches higher ground and enters the true coastal forest. The trees are not of great height but the growth is luxuriant in the extreme, the floor of the forest as well as limbs of trees being covered with vegetation, making progress off trails or clearings impossible without the aid of a machete. Richardson, who collected in this coastal forest at San José and Cisneros, considered it the most difficult ground to work he had encountered in a field experience of twenty-five years in the tropics. The density of the vegetation limits one’s radius of action and makes it difficult to shoot birds as well as to find them when shot; the high degree of humid- ity prevents them from drying properly, while the abundance of mosquitoes, as well as of other insect pests, makes the region extremely trying and unhealthful. Both Richardson and his native assistant suffered severely from fevers acquired in this low coast region, the avifauna of which is still far from exhausted. The Caldas Basin.— A short distance east of Cisneros, and some 1500 feet above it, the railroad, still following the shores of the Dagua, passes through a narrow cafion worn by the river, and emerges in a surprisingly arid basin or pocket in which lies the settlement of Caldas (alt. 2560 feet). The floor of the valley, and at least lower slopes of the hills by which it is sur- rounded, are covered with short grasses with occasional stands of low cac- tus, acacia-like trees and agaves. The abrupt change in climate, indicated by the striking difference in the vegetation of Cisneros and Caldas, is evi- dently due to the presence of a ridge at the western border of the Caldas Valley of sufficient height to protect the area lying east of it from the pre- vailing western winds and, consequently, from receiving a share of the moisture they carry. A part of this moisture is given up as the air-currents strike the Pacific slope of the ridge which borders the Caldas basin on the west, with the resulting heavy rainfall of the western slope. In passing over or pouring down into the valley at Caldas, the temperature of the air is doubtless raised rather than lowered and its moisture-carrying capacity correspondingly increased. Consequently, further condensation does not | occur until the higher mountains to the east are reached, and with the increase in rainfall the forests reappear. This treeless depression or valley on the Pacific slope of the Western 24 Bulletin American Museum of Natural History. [Vol. XXXVI, Andes is therefore surrounded by forest and the character and origin of its fauna is hence of much interest. It apparently cannot be derived from the humid, heavily wooded slopes above or below it, and the height of the mountains to the east is presumably sufficient to separate it from the faunally similar Cauca Valley. Nevertheless, its bird-life has evidently been derived from that valley. When, however, one observes that owing to the aridity of the eastern slope of the Western Andes the Tropical Zone ascends niearly to the San Antonio pass, it is clear that the Tropical Zones of the Caldas and Cauca Valleys are separated only by the narrow belt of timber which crowns the San Antonio pass. Hence we have numerous Cauca Valley species occurring at Caldas but apparently not elsewhere on the Pacific slope in this section. Caldas to San Antonio.— At Caldas the trail leaves the banks of the Dagua and winds gently up the slope toward the San Antonio pass. At an altitude of 5700 feet we entered the clouds and, at the same time, the lower order of the cloud forest which characterizes the Subtropical Zone. The Caldas region now appeared as a treeless depression surrounded by forest- crowned mountains. Everywhere the tree-line was as sharply defined as in a fresh clearing’ The cloud-line coincided with es tree-line. Cloudless hilltops were bare of trees. The luxuriant forest of the Subtropical Zone continues to the summit of the ridge and as far over it as the cloud’s-cap itself. Normally, this is not more than a few hundred feet, but when ravines or barrancas slope down toward the Cauca Valley the water they carry leads the forest to a much lower level than it reaches without the encouragement of such natural ‘irrigation. These wooded barrancas are separated by grass-grown ridges of the treeless eastern slope of the Western Andes. These ridges carry a limited number of species of the Tropical Zone upward almost to the San Antonio pass, Just as the forest’s arms stretching down the barrancas carry some Subtropical Zone species well below the upper limits of the Tropical Zone. The result is an inosculation of faunas occasioned by causes which are obvious enough when seen, but which the most accurately labeled specimens would not: reveal. The crest of the range is here so narrow that the descent into the Cauca Valley begins almost where the ascent from the Pacific ends. One has to go only a few hundred feet ‘below the divide to pass from the forest into a low, scrubby growth which quickly gives way to the brown, treeless slopes leading down to the Cauca Valley. Most of our collecting in this vicinity was done in the forests, but occa- sionally work was done along its border and here certain tropical species were secured, a fact which accounts for their being recorded from a locality which in reality is in the Subtropical Zone. Buuietin A. M.N.H VoL. XXXVI, Priate IV. EASTERN SIDE OF WESTERN ANDES FROM SAN ANTONIO Note descent of forest down a drainage ravine and ascent of arid zone of the Cauca Valley up a treeless shoulder of the range. (Interdigitation of Tropical and Subtropical Zones and West Andean Fauna with arid portion of the Cauca-Magdalena Fauna.) 1917,] Chapman, Distribution of Bird-life in Colombia. 25 The Cauca Valley near Cali— The Cauca Valley from Cali to Cartago has a uniform altitude of 3500 feet and an average width of possibly twenty- five miles. The rainfall is not high, ranging from thirty-five to fifty inches, and forests apparently occur only where they receive natural subsurface irriga- tion from the mountain slopes. The Cauca River, which is navigable for small steamers from Cali to Cartago, except during very dry seasons, is bordered by marshes, bamboo thickets and savannas and, in places, by heavy forests. Approaching the mountains, on each side, dryer savannas with acacias and large tracts of grazing and cultivatable land predominate and extend to the bare, rounded foot-hills which lead upward to the lower borders of the cloud forest of the Subtropical Zone. About Cali we collected in the savannas and marshes; at La Manuelita in the pastures, cacao groves and fallow fields grown with scrub and bordered by trees. At neither place did we find first-growth forests such as exist in the vicinity of Guengiie east of Florida, where, however, circumstances shortened our stay. Miller and Allen later collected in primeval forest at Rio Frio, but I feel that more work could be done to advantage in the forests of the valley. The Central Andes above Palmira.— Our location at Miraflores (alt. 6200 ft.) on the western slope of the Central Andes above Palmira, was much like that in which we had lived at San Antonio. The comfortable bungalow which Mr. Eder so kindly placed at our disposal is situated near the junction of the Tropical and Subtropical Zones. Above us was the.lower border of the luxuriant subtropical forest; below, the bush-grown or bare hills leading to the valley. If, therefore, we went down the trail we encountered chiefly tropical forms but if we climbed upward we were soon among the birds of the subtropics. Where the change in fauna also implied change in haunt the difference between the bird-life below and above our home seemed natural. Thus Ground Doves and Seedeaters were to be expected in the open grassy country toward the valley, just as Tanagers and Trogons were to be looked for in the forests higher up the mountain side. When, however, in the belt of timber bordering the Amina River, a thousand or fifteen- hundred feet further down, one found Ostinops decumanus, a strictly tropical ° species, and in not dissimilar haunts a few hundred feet aboye the bungalow, encountered Ostinops salmoni a strictly subtropical species, one was more impressed by the influence of temperature in determining life-zones. The summit of the ridge on which Miraflores is situated has an altitude of 8000 feet, and the forest growth increases in luxuriance as one mounts toward the crest. Nowhere have I seen a greater profusion of creepers, parasitic and epiphytic growth. Tree ferns here were estimated to reach a height of fifty feet. 26 Bulletin American Museum of Natural History. [Vol. XXXVI, The crest of the ridge is narrow and about 150 feet down the eastern slope the character of the forest changed completely. Tree ferns, parasites, and epiphytes largely disappeared and trees with small leaves replaced the cecropias and other large-leaved species of the western and more humid slope. There was little undergrowth and the woods bore a general resem- blance to an open beech forest. This growth persisted to the shores of a fair-sized stream at the bot- tom of an almost V-shaped valley, 1400 feet below and west of the summit of the ridge to the west. The succeeding ridge, or east wall of the valley, is of apparently the same height as the first ridge and is densely wooded to its summit. The trail, however, did not extend beyond the bottom of the valley and we made no attempt to explore the uninhabited mountains. to the east. THE RECONNAISSANCE OVER THE QUINDIO. _ Cali to Cartago.— The journey from Juanchito, the port of Cali, to Cartago was made by steamer on the Cauca River. The river is narrow enough (averaging one hundred and fifty to two hundred yards in width) to permit one to see the details of both banks; the water was high, the cur- rent about three miles an hour. The distance in an air-line between Cali and Cartago is ninety miles, by the river 172 miles; but if the winding course of the steamer increases the length of the journey, it also adds to the charm of it. The Cauca flows on the western side of the valley, its waters occasionally washing the foothills of the Western Andes. The country through which it passes is most diversified and attractive. Broad marshes flanked by dryer savannas, bamboo forests, patches of plumed wild cane, cacao groves and stretches of plantains near the small settlements or ports of the larger towns which, like Cali, were some miles from the river, made a pleasing and varied panorama of river scenery. Later we encountered heavy, primeval, bottomland forest, such as surrounds the port of Rio Frio, selected as a locality for subsequent investigation by Miller and Allen. These forests; - however, are not to be compared in extent to those which border the Magda- lena River, for example; and are doubtless limited to areas where they re- * ceive sufficient subsurface irrigation to nourish them. Large White and Snowy Egrets, the latter much the less common, Gray-green and Night Herons, Wood Ibis, Roseate Spoonbills, Cormorants, Jacanas, Pigeons (Columba rufina) a few Ducks, including an occasional Muscovy, and Cassiques (Ostinops, decumanus) were the birds most commonly seen from the steamer, while mammals were represented ‘ CRUNRY URAPUP Iso '9UOZ [TRoIdorjqns) Ceuney ueapuy jsaA4 ‘auUO0Z [eotdomqns) “‘o.INSOdXa + ‘UOTTEJOZVA JURLIMXNT pue dGQRIOARJUN BULMOTLOJ p ATJUepLia uoHeydiooid AABOY ST aJoY} ‘A[Joysam oe SPpULM Suyrersad ay, YIMOIS oTpIseied i luloo Y 1OJ OULL Cla QOO'S “LI¥) sauotayUIyy Md SY GDGIY ANVS JO AdO1Q NUALSY A AAOMY SHQNY IuMLNAQ a0 at YW UALAGQ Ao aAdo1g NUGALSA MAM “A SLVId ‘TAXNX “IOA “HON CIN V Nisgating Buvtetin A. M,N. H., Vou. XXXVI, Prare VU. Cauca River Near Buaa Note the treeless eastern slopes of the Western Andes. A flock of white Herons is feeding on the marsh. (Tropical Zone; arid portion of Cauca-Magdalena Fauna.) Forest on THE Cauca River at Rio Frio (Tropical Zone; a humid island in the arid portion of the Cauca-Magdalena Fauna.) 1917.] Chapman, Distribution of Bird-life in Colombia. 27 by a few Capybaras on the banks, and red Howling Monkeys in the bamboos. Fresnado, the port of Cartago, like Juanchito, the port of Cali, is distant three miles from the town it serves. The intervening country, again like that at Cali, is a dry, open plain or potrero. Here small, scattered acacias are the characteristic trees, and Mockingbirds, Vermilion and Tyrant Fly- catchers (Tyrannus melancholichus), Anis, Lapwings and Milvago Hawks the characteristic birds. At Cartago, thanks to the assistance of Senor Jesus Velez, we secured riding and pack animals without delay and began our journey across the Quindio the day of our arrival. Cartago to Giradot.—— The trail which crosses the Central Andes over the Quindio Pass has been travelled for centuries. Up to the lower limits of the Temperate Zone (about 9000 ft.) the country through which it passes is more or less settled and under cultivation, and its primitive character is therefore not always obvious to one en route. However, Dr. Allen’s descrip- tion of the stations at which he and Miller subsequently collected, supply the essential details, and I give here only the generalized view which one may gain from the saddle. For the first seven or eight miles, the trail, after leaving Cartago, passes over the low ridge which lies between Cartago and: Piedra Moler on the Vieja River, one hundred feet above Cartago. The country is rather arid, and more or less covered with a scrubby growth. From the summit of the ridge a view is had of a well wooded valley which opens into the Cauca Valley, now much constricted and set with hills which mark its termina- tion as a valley and passage into the more mountainous country north of Cartago. After crossing the Vieja the trail, for the succeeding ten or twelve miles, passes through a comparatively level depression known as El Hoyo de Quindio. Itis bordered by a bushy scrub and some first-growth, with much fine bamboo, which reaches its upper limit at about 5500 feet. There is no outlook until, at the end of about ten miles, the trail gradually ascends and takes to the ridges. The depression through which we have passed is now seen behind us with the Western Andes in the distance, and on each side well-wooded valleys open. A few miles further the picturesque town of Finlandia (6400 ft.) is seen and beyond it we had our first view of the main Central Andes with the snow-fields of Santa Isabel. Finlandia was reached at 4 P. M. after eight hours’ travel by mule from Piedra Moler, a distance of about twenty-five miles with an ascent, always gradual, of about three thousand feet. We were still in the foothills, which, in softly rounded, green, grassy billows, rolled downward toward 28 Bulletin American Museum of Natural History. [Vol. XXXVI, the Cauca Valley and flowed indefinitely north and south along the base of the main range of the Central Andes, which arose impressively across a plain-like valley to the east. From this point we had superb but brief views of Santa Isabel and Tolima. We passed the night at Finlandia and the following morning crossed the valley to the east. At the end of an hour we entered the first primeval forest through which the trail had passed and from this point to the summit of the ridge which overlooks the Quindio River, with the Boquilla at its base, there is much Subtropical Zone forest. Here we saw Hypopyrrhus pyrohypogaster for the first time. After fording the Quindio River, a rapidly flowing stream, at the Boquilla (alt. 6100 ft.), the trail rises steeply through an open country to Salento, which is reached in a thirty-minute climb of 900 feet. Salento (alt. 7000 ft.), standing on a shelf at the base of the main range of the Central Andes, is the last town through which the traveller to the Magdalena Valley passes until he reaches Ibagiie at the eastern base of the chain. Although one has gained an altitude of about 3500 feet above Cartago, the grade is so low that one has done no real climbing, and the ascent of the Andes may be said to begin definitely at Salento or, to be more exact, at the Quindio River, 900 feet below Salento. In an hour after leaving Salento we felt that we were in the heart of the Andes. Below lay the Quindio Valley, carpeted with grass and with a scattered growth of tall palms fringing the stream which winds through it; above was an endless array of mountains leading up to the brown paramo and gleaming snowfields of Santa Isabel. Until we reached an altitude of 9000 feet there was little growth near the trail and Allen’s detailed’ description of the collecting station near Salento must be consulted for information in regard to the nature of the primitive vegetation at this point. At the altitude named, we reached the lower limits of the Temperate Zone and coincidentally the upper limits at which the land had been cleared for agricultural purposes. In consequence, forests now bordered or were near the trail. At first they were composed of large, open-branched trees among which fine oaks were conspicuous. As we ascended they became much lower and more finely branched, with small, close-set rigid leaves, and a profusion of white moss. This Temperate Zone forest thickly covered the mountains to the mouth of the Pass. At Laguneta (10,000 ft.) it was fully developed and the abundance of bird-life induced us to select this place as a collecting station for Miller and Allen who, three months later, made a most valuable collection there particularly noteworthy for the number of Grallarias it contained. Butvtetin A. M.N. H. Vot. XXXVI, Pratre VIL. Santa IsapeL, rrom LAGuNE Note the continuous forest. (Temperate Zone.) LAGUNETA A stopping-place on the Quindio Trail near the camp of Expedition No. 3. An ox pack-train is resting. (Temperate Zone.) Butietin A. M.N. H. Vor. XXXVI, Prater VIII. THe Quinpio TRaiIL A scene in the Central Andes between Volcancito and Rio Tochecito, showing wax palms. (Subtropical Zone; West Andean Fauna.) THe Quinpio TRAIL Rio Tochecito. Compare with preceding picture to illustrate differences between vegetation, along the trail, of ridges and intervening valleys. (Subtropical Zone.) 1917.] Chapman, Distribution of Bird-life in Colombia. 29 At 9000 feet we secured a specimen of Myioborus chrysops, the first indication of an eastern slope fauna, and the following day it was found to be abundant on the Tochecito. At Laguneta, distant only three hours by mule from Salento, the bird-life had completely changed. The Subtropical species were left behind and in their places such characteristic Temperate Zone birds as Semimerula gigas gigantodes, Atlapetes schistaceus, and Psttto- spiza rieffert were seen commonly by the wayside. After crossing the Divide (11,500 ft.) the descent toward Volcancito is through a country from which the forest has been recently cut, but the evidence indicated that it had covered the mountain sides, as at a distance from the trail it still does. About 1000 feet below the summit wax palms (discovered on the Quindio ‘Trail by Humboldt and Bonpland in 1801) were, first encountered and these stately trees, in scattered groups or densely growing masses, were the most abundant aboreal form, from this point to the Toché River. They attained a height of at least 180 feet and were of especial interest to us as the home of the fine Yellow-eared Parrot (Ogonorhynchus icterotis). In places nearly every palm was occupied by a pair of these birds whose nest-holes opened just below the lowest leaves. The trail now descends by steep zig-zags to the Tochecito River (alt. ‘9000 ft.), a rushing mountain stream some ten feet in width with banks bordered by a luxuriant undergrowth and some small parasite-covered trees. Beyond these banks the mountain sides were covered with wax- palms with some bushy lower growth. Birds were not numerous. Essentially similar conditions exist to the Toché Valley: (7100 ft.) of which a most impressive view is obtained from a point on the trail, at least 2000 feet above it. To the right the eye follows the course of the beautiful foaming Rio Toché, here about eighty feet in width, the home of Torrent Ducks (Merganetta columbiana) and Ousels (Cinclus leuconotus); to the left at some distance, the floor of the valley is covered with a heavy forest growth which, unfortunately, we have not explored. Specimens of wide- ranging, plastic species taken at this point are, as might be expected, refer- able to the Magdalena Valley, rather than Cauca Valley form. ~ We had now returned to the Subtropical Zone. There is a small settle- ment on the Toché and from this point onward to Ibagiie the country border- ing the trail is, or has been, largely under cultivation. Small patches of the original forest growth were found at intervals, notably near El Eden, but the work of man near the trail and heavy clouds which often obscured all but the immediate landscape, made it difficult to gain a very clear idea of primitive conditions over this part of the trail, though distant mountain sides generally appeared to be wooded. 30 Bulletin American Museum of Natural History. [Vol. XXXVI, Ibagiie (alt. 4850 ft.), a city of several thousand people, is situated at the definite junction of the Magdalena Valley plains with the mountains. From this. point to the Magdalena River the road passes through a grass-covered, grazing country with more or less scrubby tree growth bordering the streams, but with no real forest. Highly eroded, castellated buttes, arising abruptly several hundred feet from the plain, are characteristic features of this part of the Magdalena Valley and at least as far north and east as the vicinity of La Dorada on the river, where the semiarid upper valley merges into the humid forest region. The descent is not noticeable, but at Chicoral on the Coello River and distant some thirty miles from Igabiie, we have dropped to an elevation of only 1800 feet, and Honda, on the river, is but 600 feet above the sea. The country lying between Honda and Barranquilla is described under Expedition No. 7. Expedition No. 2.— The Popayan Region. May 18, 1911-July 21, 1911. Personnel.— Leo E. Miller; W. B. Richardson. Itinerary On May 13, 1911, a few hours after Chapman and Fuertes started their homeward journey in reconnaissance over the Quindio Pass, Miller and Richardson with their pack mules left Cali for Popayan. They reached that city on the 17th, and three days were consumed there in making preparations for a trip to the Western Andes. On May 20 they left Popayan for Cerro Munchique, making their first collecting station May 22, at an ele- vation of 8325 feet on this mountain. They remained at this station until June 4, when they left for Cocal on the western slope reaching that place June 6, and working there until June 18 when they returned to Popayan for. supplies. June 24 they again left Popayan for the Western Andes working at Gallera from June 26 to July 4; La Florida July 5 to 9, and on the summit of the first ridge of the Western Andes (10,340 ft.) from July 10 to 23. At this point they found a typical Temperate Zone fauna, this being the first time this fauna has been discovered in the Western Andes. July 27 they returned to Popayan and at once left for their base at Cali. Description of Route and Collecting Stations.— “The country ‘through which we passed on the road to Jamundi is level, covered with excellent grass and given up largely to cattle ranches. Two hours beyond J amundi the country became rolling and here the lomas, or hill country, begins. At 2 P. M., May 14, we crossed the Cauca, here practically as wide as at 1 From the reports of Leo E, Miller. Ceuney euaeps¥y-eoukO Jo uomsod ple ‘auo0z peoidoay) f Ceuneg vuaepsefy-voney jo uoysod pe ‘auoz jeordoay) “aygnq pepoia “9¢ ea < : ; eu: YIM [RIOT pue enseq] Ueemjoq AoyveA wuatepseyy ssddgQ SHANA PEROT, TOMA: BUUBARS “BUSTEDABIAL: deddiy. “OUsiHeseTeUO yavuoqd VW] GNV VaNOW NaGMLaA avOoUTIV' YNITOL 40 sNivig dad v1 H UY “XT 81V1d “IANNX “TOA “HON CW CY Nisaiing 1917.] Chapman, Distribution of Bird-life in Colombia. 31 Juanchito. We passed through a heavy growth of bamboo, creepers and brush about a mile before reaching the river and heard here several howling monkeys. “After leaving this bottom-land the country again became rolling. The hills are bare except for a wiry grass. There were no cattle. “May 15, we continued to pass through a bare, rolling country and at 4:30 P. M. reached an elevation of 5900 feet. The following day the country began to look more attractive. The road generally ran along the top of a ridge and we could see for many miles. Everywhere the hills were covered with low trees and dense shrubbery. At an altitude of 6400 feet we saw Green Jays (Xanthoura), Blue Swallows (Pygochelidon), Andean White-throats (Brachyspiza) and heard Compra Pans (Grallaria ruficeps). There were also Black Thrushes (Merula gigas gigantodes)..” Popayan to Munchique— “At first the country is comparatively level with small clumps of trees and large cornfields. There were also groves of oranges, apples of rather poor quality, fair peaches and good bananas and plantains. A very little cacao and a great deal of coffee is grown. “The second day out (May 21) the country was rolling and barren except for a few clumps of trees and brush. Many Black Merulas and Green Jays were seen. After leaving Chappa, on May 22, the road became very bad, rough and steep. At an altitude of 7200 feet we entered the forest and at 4 P.M. that day made camp at 8325 feet, on the eastern slope, in a small old clearing entirely surrounded by virgin forest in which we collected.” Cerro Munchique to Cocal.— “One hour after leaving camp we reached the top of the ridge (alt. 8800 ft.). Below was a sea of clouds, over which the Pacific could be dimly distinguished. It is distant fifty miles, but owing to the numerous ravines and ridges the natives take eight days to reach the coast. There is but a narrow trail through the dense forest which here is like that on the crest of the ridge above Miraflores. “The trail is like a stairway down which one goes with much difficulty, some of the steps being six feet or more. We-camped by the Rio Tambito, a narrow, swift stream running through a 20-foot gorge, and after crossing another ridge (alt. 6900 ft.) through the forest, reached Cocal (alt. 4000 ft.)? at 4 P. M. the next day. “Cocal is a settlement of negroes who are practically savages. They live in miserable huts and wear no clothes. The mountain sides are very steep and the jungle all but impenetrable. I estimate that fully thirty 1 Indicating the arid Subtropics, to which the Temperate Zone Black Merula descends.— F. M. C. 2 The collections from Cotal include species which we have not elsewhere found at so low an eleva- tion, but Mr. Miller tells me that some of the birds labeled Cocal were taken at a higher altitude than the settlement. 32 Bulletin American Museum of Natural Hislory. [Vol. XXXVI, percent of the specimens shot were lost in the undergrowth. We found our first Cocks-of-the-Rock in the tall palms here.” Popayan to Gallera.— “For the first day the country is rough and bare. On the morning of the second day we reached virgin forest on the eastern slope of the Western Andes at an elevation of 7500 feet. After reaching the crest of the range (alt. 10,340 ft.) the road runs for about four miles along it. . From this point one may see the Pacific faintly. The vegetation is scarce, scrubby and stunted. Here we worked under difficulties, living in the small tent. The wind blew almost constantly and there were a number of severe electrical storms, during which it blew, rained and hailed with great violence. These storms are preceded by dense fog, so that it was not possible to go any distance from camp without danger of falling hundreds of feet off the trail. “The new government road runs down the west side in zig-zags. ‘Two thousand feet down we struck the heavy forest. Gallera (alt. 7000 ft.) is a camp of road laborers in the very heart of the forest. There is not one side trail. The forest is impenetrable and we found few birds.” Expedition No. 3.— Lower end of the Cauca Valley, The Quindio T: rail, Cartago to San Juan River. August 22, 1911-January 7, 1912. Personnel.— Leo E. Miller; Arthur A. Allen. Itinerary.— Miller and Allen sailed from Cali on a Cauca River steamer August 22, 1911, and arrived at Cartago on the 25th. Cartago was left on the 27th and Laguneta, just below the Quindio Pass, reached on the 28th. Here they pitched their tent at an altitude of 10,300 ft. and remained until September 11, when they retraced. their steps as far as Salento and, the fol- lowing day, began the ascent to Santa Isabel through the Boquilla Valley. They reached the Paramo, at an altitude of 12,700 feet, on September 13, and camped there unti! the 20th, then moved to a point about a thousand feet lower and collected there for three days. _ September 25 they returned to Salento and on October 3 they began a trip over the Quindio Trail, to Chicoral in the Magdalena Valley, with the object of collecting the more characteristic species at a number of localities and thus determining zonal and: faunal limits. Chicoral was reached October 6 and collections were made there until the 13th, when they began their return journey, stopping at El Eden October 17-21, Rio Toché October 23-27, and returning to Salento October 31. Collections were made in the vicinity of Salento until the 13th when they returned to Car- tago en route to Rio Frio on the Cauca. Here they worked in the heavy Buuvetin A. M.N. H. Voi. XXXVI, Puate X. Crest or WeEsTERN ANDES, WEST OF POPAyANn. Scene near camp of Expedition No. 2; alt. 10,340 ft. (Temperate Zone.) La GaLuERA, WESTERN ANDES ; Near camp of Expedition No. 2. (Subtropical Zone; West Andean Fauna.) 1917] Chapman, Distribution of Bird-life in Colombia. 33 forest until December 2, when again they went to Cartago to make prepara- tions for the crossing of the Western Andes, the most difficult journey thus far undertaken. Cartago was left December 7, and, through a misunderstanding, the journey to the San Juan Valley was made as rapidly as possible without pause for collecting. The few specimens secured en route indicate that the fauna differs in some respects from that of a section through the same chain from Cali to Buenaventura, and we cannot but regret the absence of speci- mens from this region. Juntas de Tamana was reached December 14 and left December 20; and collections were also made at Névita from December 21 to 27. During these two weeks 277 birds and 39 mammals were collected under the unfav- orable conditions of the rainy season. Doubtless this over-exertion made both men more than usually susceptible to the pernicious type of malaria which prevails in this unhealthful region. Both contracted severe attacks of fever, and on returning to Cali, January 7, were under a physician’s care for several weeks. Description of Route and Collecting Stations.— A general account of the route between Cali and Cartago and Giradot, based on the reconnaissance made by Chapman and Fuertes in May, 1911, will be found under ‘ Expedi- tion No. 1.’ The following detailed descriptions of the stations on the route at which collections were made were prepared by Allen, who also writes the report on the ascent to the Paramo of Santa Isabel and on the journey from Cartago to the San Juan region. Rio Frio. “Most of our collecting here was done in the forest on the east bank of the river except for two trips to a rather extensive marsh some- what south of the Rio Frio, and about an eighth of a mile back from the Cauca. Some collecting was also done along the northerly edge of the forest, the country becoming more open in this direction, the forest extend- ing to a much greater distance south from the port. “The forest reminds one considerably of our northern deciduous forests in which the giant oaks are replaced by ceibas and the maples by cecropias. ‘There are comparatively few ferns, orchids or epiphytes of any kind except a few “pines” (bromelias) and very little moss. The forest floor is covered with dead leaves and with little undergrowth except about clearings and more open spots in the forest where it is very dense. But it was in such places that the birds were most abundant, as we found here a greater abun- dance of individuals than anywhere else, though the number of species was perhaps more limited than in the ‘cloud’ forest. 1 That is, forest of the Subtropical Zone.— F. M. C. 34 Bulletin American Museum of Natural History. [Vol. XXXVI, “T did not attempt to explore the marsh to its full extent because of its evident treachery and the difficulty of progress through it. It was appar- - ently surrounded on all sides. by forest, although toward the south this seemed to thin out and was perhaps no more than a fringe about the edge. The marsh appeared as though it sometimes might be a lake although at this time of year (November) there was little open water. The predominant vegetation was a coarse sedge, the tangled roots of which furnished the only ‘support for one attempting to walk through it, for elsewhere, and below these roots, was bottomless black muck. About the edge of the marsh, fringing the forest, was a rank growth, almost impenetrable, of thorny bushes and growing out into the marsh were scattered bushes of marsh mallows (Hibiscus?). Where the sedges had not yet established themselves, were extensive rafts of the water hyacinth and a plant that looks like coarse lettuce. On these rafts were flocks of Jacanas and Spurwings, White Ibis, Roseate Spoonbills, and White Egrets. In the small patches of open water floated Muscovy Ducks, Cinnamon Teal, Fulvous and Black-bellied Tree Ducks. In the sedges Rails skulked and there were many Screamers. The ‘Black Marsh Hawk (Circus) skimmed low over the sedges just like our northern bird, and a few King Vultures sailed high overhead. It was a wonderful spot, I should like to have spent a month in studying it alone.” El Roble.— “El Roble, at an altitude of 7100 feet, is the last posada before descending into the valley of the Boquia. The collecting here was ‘done in two kinds of places: the comparatively level forest at the altitude given, and the forest at a considerable lower level filling the valley of one of the tributaries of the Boquia. The level forest was not much less humid ‘than that at Laguneta, with correspondingly less moss on the trees and on the forest floor. That along the stream was just as humid, if not more so ‘than at Laguneta. Along the trail and about clearings, etc. the plants and ‘birds were similar to those about Cali, but in the forest and particularly along the stream, the forms were those of the Subtropical Zone. The ‘feentury plant’ and the tree fern seemed to me to reach their greatest luxuriance here, the tree fern in the forest, the yucca in the open country, along roadsides, ete.” Salento.— “Most of our collecting near Salento was done along the ‘Boquia River at an altitude of 6500 feet. The flora and fauna of the open -ecountry of the Boquia Valley and about Salento is similar to that of El Roble with a somewhat larger amount of the Cauca Valley element such as ‘the Spanish bayonet, yuccas, plantains, a few bananas and oranges. Com- 1 To prevent confusion with a station of the same name above FusugasugA, in the Eastern Andes, specimens taken at El Roble are listed under Salento, the Salento collecting ground being nearby and in the same zone.— F. M. C.- s Ceuneg ueepuy jsaM ‘euoz reordo14qng) ‘eyMbog oy} 07Ul SuLMoy wWra4s ¥ (‘euney uvapuy JseaAA ‘9UOZ Teotdo1.qng) OLNTIVG WOU AGTIVA VITIINDOG OLNGIVEG YVAN 1X F1V1d ‘IAXXX “10A “ACN CW 'V Nigeria g 1917,] Chapman, Distribution of Bird-life in Colombia. 35 paratively little of this open-country fauna was collected and most of the birds came either from the narrow strip of woods along the river, or from Santa Rita, a small tributary with well-wooded banks, entering from the west. The Santa Rita is a beautiful dashing mountain stream, cutting through the rock in narrow gorges or pouring over huge boulders and arched over by luxuriant vegetation, rich in moss and epiphytes; perhaps the most luxuriant that we found except in parts of the Western Andes. Here lived the Cock-of-the-Rock and Pharomacrus.” Laguneta.— “The forest in which we pitched our tent at Laguneta, at an altitude of 10,300 feet, is fairly open. The vegetation includes a few “small palms, tree ferns, orchids and epiphytes of many kinds but gives one the general effect of some of our denser northern forests. The large trees, of which some are oaks, are rather sparingly branched and thinly leaved so that sunlight reaches the ground in most places. The orchids, ‘pines,’ moss and other epiphytes on the branches are responsible for as much shade as the trees themselves. The undergrowth is not dense except in the clear- ings —it being the vines and ‘climbing bamboo’ that makes the forest impenetrable. The forest floor is remarkably bare with very few herbaceous plants (due to season ?), few ferns, and no moss (on the ground). The leaves are mostly thick and heavily glutinized or covered with down and, though some are large, the average is small. The clearings resemble our northern clearings in general appearance — grown up to bushes and small trees. Here occurs a pokeberry upon which certain birds feed. The under- brush is always extremely dense and almost impenetrable without a knife. “We remained in this camp from August 30 until September 11. During this time we had very favorable weather with but little light rain. For several days, however, we had very high winds. The temperature was very uniform averaging 48° at 6: 30 A. M. and 64° at noon (the nearest to maxi- mum and minimum that we could get).” Rio Toché.— “Most of the collecting here was done along the river where there was a sparse growth of trees, but two trips were made up the river to where it was heavily forested, humid and luxuriant, resembling the banks off Santa Rita near Salento. A few birds, notably Atlapetes flaviceps, were taken from the brush covering the cleared mountainsides of the open valley not far from the trail.” El Eden.— “The country about El Eden at an altitude of 8500 to 9000 feet, seemed intermediate between the valley type, such as was found at Salento, and the cloud forests of Laguneta. This is due, I suppose, to the large amount of clearing and the comparatively small extent and isolation (?) of the forest. In the forest, birds were very scarce and in the open country bitds were also less abundant than elsewhere. We were disap- 36 Bulletin American Museum of Natural History. [Vol. XXXVI, pointed in the collecting here since forms were neither abundant nor dis- tinctive.” Chicoral.— “Chicoral is on the west side of the Magdalena Valley at an altitude of 1200 feet. The Valley at this point is very arid, even along the - streams, although this was probably emphasized by the fact that it had not rained for five months when we arrived. There were many more cacti and palmettos than in the Cauca Valley, with few epiphytes or orchids on the trees. Most of the collecting was done in the sparse woods along the river and scrubby places about the pastures and a little on the open plain which was covered with coarse dry grass. Along the river birds were very plentiful, including many migrants from North America.” Salento to the Paramo of Santa Isabel. “The Valley of the Boquia leads northeast past Salento and the trail to the Paramo follows this valley to the very headwaters. The valley, which we crossed at Boquia on our way to Laguneta, at an altitude of 6100 feet, rises gently until, where we left it, it’ was 8300 feet. It is broad and open with little vegetation, except a narrow girth along the stream and a scattered growth of magnificent palms. These continue to the head of the valley and up on the mountainside to at least 9500 feet. “Leaving the valley of the Boquia at 8300 feet, the trail leads almost due north up the mountainside at a very sharp angle. The trail is poor and in some spots practically obliterated. The lower mountainside is very bare except for the scattered palms, having been burned and cleared. Corn and wheat are growing in spots but most of the country is closely cropped by cattle. The open country continues to an altitude of 9300 feet; here the forest begins, and so far as we could observe resembles very closely that of Laguneta. The trail now becomes slightly more marked so that one has no difficulty in finding the way. At 4:20 P. M. we reached the second house above the valley, at an altitude of 10,550 feet, where we stopped for the night. The next morning we continued on our way to the Paramo. The trail leads through large clearings and patches of woods similar to that of Laguneta until the 12,000-foot mark is reached where the forest appears quite different. The trees are large, the woods more open with an abun- dance of moss even on the forest floor —less of the ‘climbing bamboo,’ which has been replaced by another species more like huge grass. The moss itself is very different in appearance being almost black. A few yellow orchids were in bloom. The appearance of this woods leads us to decide to stop and investigate it for a few days on our way back. ‘This forest continues to from 12,400-12,600 feet where a decided change takes place. The large trees disappear and smaller, more or less recumbent species, take their place, resembling large bushes. Large bushes are intermingled Burierin A. M.N. H. Vor. XXXVI, Pirate XII. CuicoraL BRIDGE Collecting ground of Expedition No. 2. (Tropical Zone; arid portion of Cauca-Magdalena Fauna.) GIRADOT, UPPER Macpautena River (Tropical Zone; arid portion of Cauca-Magdalena Fauna.) 1917] Chapman, Distribution of Bird-life in Colombia. 37 with them and sphagnum, gentians, dwarf lupines, yellow-eyed grass, a yellow sorrel, similar to ours but larger, a buttercup, a peculiar fern, and numerous composites, mostly dwarfed rosettes but one a very showy purple and yellow species, called “arnica” by the natives. The wet places, along streams or in sink holes, give one the impression again of our northern bogs only here there is very little or no sphagnum. Its place is taken by a peculiar daisy whose thick set rosettes of short stiff leaves form great hummocks over soft places. Blueberries were plentiful but bitter, woody, and inedible, except for the birds. “When we reached the Paramo, we made for the top of the ridge and looking over found a beautiful little valley suspended there in mid-air. The lower end of it was wooded at the sides but the centre and upper end was open Paramo. Down the centre rushed a sparkling mountain stream which made up in sound what it lacked in size. On either side of the valley the ridge rose from 13,500-15,000 feet and the valley itself lay at about 12,700. “We descended into the valley and pitched our tent at the edge of the woods. In this valley was done most of the collecting. We later explored up to an altitude of 15,200 feet (nearly the limit of vegetation and above the base of the snowline on the peaks) and found the vegetation practically the same and the fauna identical, except that it was much more condensed in the valleys, especially along the stream. The vegetation extended to about 15,500 feet and above that all was bare, frosted rock. At this altitude and open situation, birds were very few but without exception were identi- cal with those in the valley. The woods which cover the mountainside below, and the nature of which has already been described, extend up the sides of the valley for aboyt a half mile to an altitude of 12,600 feet. The centre of the valley is open from 12,400 feet up but the typical Paramo does not begin until 12,600 feet. “The birds of the open Paramo are comparatively few in species and not extremely abundant in numbers, especially on the ridges. They are most abundant along the stream and in the swamp which occurs mostly along the stream. “The birds of the woods bordering the Paramo were of course very different and consisted mainly of Laguneta species with a few others of like nature. ae “We remamed at this camp for just a week and collected 200 birds. The weather was very pleasant although the nights were cold, several times ice forming in the water-pail. The temperature at 6 A. M. varied from 37° to 45°, dependent upon whether it were cloudy or not, and at noon got up as high as 76°. The mornings were generally more or less clear but shortly 38 Bulletin American Museum of Natural History. [Vol. XXXVI, after noon clouds usually rolled up the valley surrounding us in fog; the ridge at the end of the valley was seldom seen in the afternoon. “From the Paramo, on September 2, we retraced our steps through the 12,000 foot woods to a vacant house at 11,500 feet, intending to remain several days and work the woods. We left on the third day, however, after making 75 skins, for we found that the fauna was not strikingly differ- ent from that of Laguneta.” Cartago to Novita.— “We left Cartago December 7, Miller having been fortunate enough to locate some oxen going to Salencio. Crossing the river the valley is more rolling than on the other side, but just as arid. Forty minutes brings one to Anserma but the trail branches off just before reaching the town and soon winds up a long ridge to an altitude of 6800 feet, drops across a valley to 5800 feet, and then rises again to 7800 feet. This country is much like that across the valley on the Quindio Trail, and the birds, so far as observed, likewise. There is little sign of humidity until 6800 feet is reached where moss on the trees and an abundance of ferns bespeak of the ‘cloud zone.’ There was little or no bamboo except along one stream low-down (3800 ft.) and no apparent change in the birds until this upper zone (above 6800 ft.) was reached. This humid forest reached its greatest development on the western slope of the ridge although it was very distinct on the eastern slope down to 6800 feet. On the western slope it extended down to perhaps 6000 feet. (I have not this altitude exactly.) On the top it is extremely rank and luxuriant — far more than we had met before, although perhaps not more so than at San Antonio. Birds were very scarce, as usual in this humid forest; the trail descends very steeply from this forest through more or less cleared country to Salencio, at an alti- tude of 5,500 feet. Here we staid two days while Miller engaged the peons for the rest of the trip. In the river valley (5000 ft.) below is a heavy growth of bamboo. Between this and humid forest (at a little distance from: town) is a good forest growth of an intermediate nature. “Leaving Salencio the trail follows up the river Bueltas, a small, dashing mountain stream — much like the Santa Rita near Salento — with luxuriant woods or mossy cliffs on either side, much moss, epiphytes, ete. It is a typical Cock-of-the-Rock stream, and we had not waded many miles over its slippery rocks or on its mossy logs before I got a nice male of Rupicola sanguinolenta. After leaving Salencio we knew scarcel dry moment till we reached Juntas.. The first night at an altitude of feet, it rained all night, and we had neither tent nor blankets, for the peon carrying them deserted us without our knowing it. “The trail follows up the Bueltas to its very headwaters (alt. 6,600 ft.) where it leads sharply up the mountain through deep crevices or gulleys, Bouruetin A. M.N. H. Vou. X CXVI, Prats XII. Paramo or Santa Isapev, CenTRAL ANDES Collecting ground of Expedition No. 2. (Paramo Zone.) Paramo or Santa Isapet, CenTRAL ANDES Near camp site of Expedition No. 2; alt, 12.500 ft. (Paramo Zone.) 1917] Chapman, Distribution of Bird-life in Colombia. 39 in places grown over above by the luxuriant vegetation forming tunnels, until an altitude of 7,400 feet is reached. The forest now is very luxuriant, similar to that on the top of the first ridge. Here was an assemblage of birds mostly new to me. I should like to have been able to work it thor- oughly, but of course we could not stop even for a day on account of the scarcity of food. “This extremely humid forest extends down the western slope of this second ridge until 4000 feet is reached where the change, which is coiplete at 3,500 feet, begins. The moss and epiphytes gradually disappear, the forest floor becomes dryer, the bamboo appears and becomes abundant, and the forest takes on much of the character of the Rio Frio vegetation. The change in the birds is likewise very noticeable. This forest continues down and across the valley and up the east side of the third ridge to practi- cally the same altitude, it being only on its very top that the humid zone prevails. The trail coming down the west slope of the second ridge follows the ridge which separates the Ingaré from the Avita, which flow together at El Puente to form the Tamandé. El Puente is a collection of some half dozen bamboo houses filled with shiftless, long-legged negroes where one can get but a few expensive plantains by way of supplies. The fauna of the valley contains a large percentage of the coast forms, birds which we saw for the first time, but found very common in the Chocé proper. “The top of the last ridge compares favorably with the similar altitude of the second ridge, and is not nearly so humid as the higher altitude, though strikingly different from the bamboo zone below. The west slope of this last ridge is different from any I have before described. There is but little bamboo or none. It is more humid than that zone and yet there is no moss and comparatively few epiphytes. The coast fauna, I believe, there extends nearly to the top —at least to about 3,500 feet. The forest at Juntas, Névita and Noanam4 seemed practically the same in nature as this western slope and though we took different birds at each place I presume it was due to the short time spent in each.” Juntas de Tamané:— “ Altitude 400 feet. Except for the clearing in which the small village is located, the entire country is covered with a rich, humid, steaming forest of large trees and comparatively little undergrowth except that formed by the giant vines hanging down from the branches and occasional patches of fern. There are many epiphytes and but little moss, reminding one of Rio Frio, only much more humid. Birds were abundant along the edge of the clearing and along the forest trails but, as elsewhere, scarce in the deep woods.” Névita:— “Altitude (150 feet). Although Ndévita has the reputation of being one of the wettest spots in Colombia, the forest seemed less humid 40 Bulletin American Museum of Natural History. [Vol. XXXVI, than that at Juntas de Tamané and not very different from that at Rio Frio. The clearing in which the town lies is of much greater extent than at Juntas de Taman, and as a result more of the open country birds such as the little Black and White Finches, Blue Tanagers, etc. were found. In the forest birds were most abundant about small clearings or plantain patches which filled these small clearings, especially about flowering trees. “We left Névita in a canoe hoping to make good connections with the steamer at Noanamd which however did not appear. Fortunately Mr. D. C. Stapleton was passing up the river in his launch which was to return in a few days, and he offered to ship us back to Buenaventura, an invi- tation we gladly accepted. “The country seems about the same along the San Juan until one gets to sea-level where there is a great increase in the number of species of palms, and from the little we could see from the launch, the forest appears much denser and more luxuriant.” Expedition No. 4.—Cali to San Agustin. February 27- April 7, 1912. Personnel. L. E. Miller and A. A. Allen. Itinerary.— Proceeding to Popayan over the route followed by Miller and Richardson in May, 1911 (See Expedition No. 2), Miller and Allen, accompanied by J. T. Lloyd, left Popayan on foot February 27, 1912, and traveled southward to La Sierra (Feb. 29-March 4) and Almaguer (March 9-18). At the last-named point they turned to the east to cross to the Magdalena Valley, stopping at Valle de las Pappas (March 22-April 4), and reached San Agustin April 7, after a difficult and trying journey. Allen suffered much from a recurrence of fever acquired in the Chocé and shortly after arriving at San Agustin his condition became so serious that he was obliged to go to Bogoté for treatment and subsequently was invalided home. Description of Route and Collecting Stations.— The following notes are supplied by Dr. Allen: Popayan to San Agustin.— “Leaving Popayan (Feb. 27) the country continues very similar to that to the north of the city ranging, from 4700 to 6800 feet in altitude and sparsely covered with vegetation except in the immediate vicinity of the rivers. (The haze or the fog was always so dense that observations of distant ranges or peaks was impossible so that the notes must of necessity be restricted to the country in the immediate vicin- ity of the trail). “The fauna and flora likewise. continues practically the same, being similar to that of the open country just below El Roble which I have called Butretin A. M,N. H. Vor. XXXVI, Plate XIV. Between BUENAVENTURA AND San JOSE (Tropical Zone; Colombian-Pacific Fauna.) Junras pe TAMANA Typical Chocé Country (Tropical Zone; Colombian-Pacific Fauna.) 1917,] Chapman, Distribution of Bird-life in Colombia. 41 ‘Transition.’ It_is probably the lower edge of the transition, however, as along the streams the large bamboo is prevalent and several of the bamboo zone birds noted. Eight miles (approximately) to the south the town of Timbio is reached, situated on a small riversof the same name. We are now on the headwaters of the Patia and the streams lie in very deep valleys in places with almost perpendicular sides, averaging a thousand feet in depth. Were it not for these valleys the country would be fairly level, but the steep descents and ascents in crossing these streams makes the travelling slow and difficult. A day and a half journey from Timbio lies the town of La Sierra situated on a saddle-back ridge just before the trail drops into the cafion of the Patia proper. Just before coming to the town small groves of rather open forest occur and here we stopped for three days to collect, securing sixty-six birds and ten mammals. To the east a trail follows the ridge up the sides of Sotar4, which is not visible from the trail but which shows considerable forest growth, commencing at what I should judge to be between 8000 and 9000 feet. We camped at 6800 feet; having been told there was no water higher up nor pasturage for the mules. The forest in which we collected at La Sierra was comparatively dry and open, and very limited in extent. Birds were scarce both as to numbers and species, and but a very few new to our former collections were taken. Here was seen the only Condor of the expedition, and it was flying rather high over the ridge toward Sotaré. “Leaving La Sierra (where very few supplies can be secured) the trail is very poor and probably nearly impassable in the wet season, descending steeply to the Patia which here flows at 4700 feet, a rushing, rocky torrent similar to the Toché in size, but crossed by a strong brick bridge which leads one to believe the trail has degenerated. Along the river is a jungle of low trees, but elsewhere the country is covered by coarse grasses and sedges with no higher vegetation. A steep and then gradual ascent brings one to the town of San Miguel, a row of some forty houses straddling the ridge. It is supposed to be but four hours from La Sierra, but our pack mules re- quired a full day. Here we learned for the first time the truth concerning the trail from Almaguer to San Agustin, and the one that continued on to Pasto, and had to alter our plans accordingly. The trail next soon strikes into the valley of the Rio La Vega which is, I believe, another tributary of the Patia and quite similar to it in its precipitous sides. The trail follows a niche in its side for the rest of the day until the town of La Vega is reached at an altitude of 7500 feet. Hereabouts are greater signs of industry than noted elsewhere in most parts of the country, the precipitous mountain sides being covered with corn-fields or wheat, and neatly marked off with beautiful hedges; no forest as yet, however, and the fauna still “ transition.” “Leaving La Vega, the trail continues up the river for about thirty 42 Bulletin American Museum of Natural History. [Vol. XXXVI, minutes and then cuts steeply up the mountain-side, winding much until the very top of the ridge is reached at an altitude of 10,350 feet. The trail then follows this ridge for about two miles and then descends on the other side two hours’ distant to Almaguer. This ridge extends approximately northeast and southwest, and shows evidences of timbering along the lower line of forest growth. At present, the whole crest of the ridge is covered with most luxuriant forest but it extends downward for but a few hundred feet (altitude). From evidences on the south side of the ridge, I should judge that this forest may have originally extended down as low as 9500 feet, this lower stretch now being covered with high bushes of the “ole- ander”’ type, and with occasional trees of size. The forest itself is a most luxuriant one of the ‘cloud [= Temperate] zone’ type, being much more luxuriant and mossy than that at Laguneta at Santa Isabel. Here we col- lected nine days — March 9-18. Although the rainy season was not sup- posed to have set in, it rained every day and the forest was always draped with fog. The trail along the ridge has been recently widened which, to- gether with several side trails, made excellent collecting grounds. As usual; however, birds were scarce, and a considerable number of species were found with nests in the process of construction, a few with eggs, and a few with young on the wing; and the majority of all birds with enlarged repro- ductive organs. “The present lower limit of this forest is about 10,000 feet, the upper limit under 11,000, for in following up a ridge which leads off at an angle from the one of the trail, open places with stunted trees and numerous paramo species of shrub and herbaceous plant were encountered as low as 10,600 feet, although the ridge did not extend high enough for real paramo. These open areas were similar, Miller stated, to the crest of the Andes on which he and Richardson had collected west of Popayan. The flora and fauna of this moss forest was very similar to that at Laguneta and Santa Isabel, comparatively few species new to our former collections being taken. “March 18, we broke camp to start for San Agustin. One long day’s travel, or a day and a half, as we had to travel, brings one to the town of San Sebastian. The trail from Almaguer crosses the ridge to the north- east at 9600 feet, which is below the present lower edge of the moss forest, and then descends steeply into the Valley of the Caquiona at 7700 feet. The trail then follows down this valley for about an hour and crosses another ridge into the valley of the San Sebastian, at the head of which is the town of San Sebastian (alt. 7600 ft.). It is a small town of some fifty or sixty houses, where the necessities can be secured on market day; that is, bread, meat, rice, beans and sugar,:but at other times it is rather devoid of life. The fauna of these last two valleys appeared similar to that of La Sierra, BULLETIN D Y .X LLETIN A. M.N.H, VoL XXXVI, Prare XV Zardgoz Reet “Ae clPtedra 5 lier 1S) N By B.San mi Ln, La Torres eet (re Manuela, f Se) au Sway an ahr HS en enna wen Scale of Miles ° 10 20 eat ERAT eleva ULLAL evation in feet 76° Map or CENTRAL WESTERN COLOMBIA (Drawn by F. Miiller under the direction of Frank M, Chapman.) Dotted red line indicates routes of the Museum expeditions. — Collections were made at localities underlined in red. 1917, Chapman, Distribution of Bird-life in Colombia. 43 transition, with innumerable Black Merulas and the common Song Sparrow (Brachyspiza). The trail next crosses two ridges at 8000 and 9000 feet respectively, the nature of the. country and paramo remaining practically the same. It then ascends the third and last ridge very steeply to the Pass at 10,500 feet, and then descends more gradually into the Valle de Pappas at 9900 feet. This last ridge is covered with the moss forest similar to the one above Almaguer; the forest extending down to 9600 feet with its flora and fauna the same so far as observed. “From the Pass, the valley appears perfectly flat, with patches of forest and open meadow through which endlessly winds a fair-sized stream. The valley is perhaps a mile and a half wide, and the mountains all about it so far as could be seen through the clouds covered with dense forest. Above this forest again on all the higher peaks and ridges was another area of strict Paramo covering their tops, at this time now covered with snow. The stream was called by the Indians the “Cosiacu”’ and said to be the head- waters of the Caquetd. “Descending into the valley, the vegetation is found to be similar to that of the Paramo of Santa Isabel, although here at an altitude of but 9900-11,000 feet, long sedges with numerous similar herbaceous plants and bushes, and numerous “ frailejones” were scattered about; at intervals occur small clumps of forest similar to that at the edge of the Paramo of Santa Isabel. The trail, where it has been repaired with brown soil and guide logs, is very good, but in other places. where composed of black muck, the natural soil, it is almost impassable for the mules. Here we staid ten days, finding quarters in one room of a finca, to which we had been recommended by its owner in Almaguer. “The fauna of this valley, while containing very few new forms, is very interesting. As contrasted with Almaguer, where the birds were just commencing to nest, here nidification was about completed for most species. Trees had ceased flowering, and most of the Hummingbirds had disappeared. “We left the Valle April 3, on the trail for San Agustin. The trail leads practically northeast upward steeply in places, and very rocky, until the top of the Paramo is reached at 12,300 feet. It was extremely rainy and foggy so that we could not see far, but it was very noticeable that there was no sharp line to tree growth as at Santa Isabel. One looks down into narrow valley covered with Paramo vegetation, while all about the moun- tains are heavily forested in places probably up to 13,000 feet; but even on these wooded slopes the forest is not continuous, but here and there occur patches of the Paramo vegetation scattered about rather miscellaneously — their presence perhaps determined by the nature of the soil rather than the altitude alone. Most of these ‘Paramo Valleys’ appear to me to be the 44 Bulletin American Museum of Natural History. (Vol. XXXVI, basins of ancient lakes which have, in some way, broken through their barriers and left behind them only those broad flat-bottomed beds of loose muck, which have gradually been covered over with the growth of coarse sedges, except where the small stream still meanders. “Just below timber-line the forest is extremely dense with a great deal of moss, caladiums, etc., and with a tree of the banyan type quite prevalent; the fringe of stunted trees is restricted. The trail continues along the ridge for a short distance fringed by low growth, and then begins a steady descent. At. 11,000 feet a fair-sized mountain stream is crossed, and then the trail follows approximately down its valley, extremely rocky and stony in parts, and ever very wet with a stream flowing down it. When a level stretch is reached, it is generally very marshy, making progress difficult. In places great cliffs rise perpendicularly for hundreds of feet at either side of the valley, and waterfalls tumble uninterrupted from the top to the river below — at least’ a thousand feet. These walls could be seen, however, only at intervals when the fog parted for an instant; at other times one could not see fifty feet in advance. Thus the trail descends to Santa Marta at an altitude of 9000 feet. Santa Marta is a rather large but unfinished building used as a general posada by all the Indian packers. It is situated in a beautiful amphitheatre of perhaps a half-mile in diameter, whose perpen- dicular walls are pierced only by the ingress and egress of the stream (and trail). The river even here is a swollen torrent and called the Magdalena by the Indians; all about is the luxuriant moss forest. It would make an ideal collecting spot. “A long day’s trip over a trail which is comparable only with that . between Cartago and Névita, brings one to Los Monos which is nothing but a small lean-to situated at the edge of a small clearing. Three hours further, ascending and descending, brings one to Pefiaseca, a niche. in a perpendicular cliff under-cut so as to be perfectly dry, and no shelter of any kind has been erected or is necessary. A few hundred feet below, almost straight down, rushes the Magdalena, here a mad torrent. The altitude is but 7000 feet, but the moss forest extends uninterrupted down its course and covers its sides,— a wonderful country! I was sorry not to be in a better position to appreciate it. This country between Santa Marta and Pefiaseca was the most inviting of the whole trip, and the trail the worst. From Pefiaseca to San Agustin, two days, the trail is much better though not good. Leaving the Magdalena at Pefiaseca the trail winds up the opposite ridge until an altitude of 7800 feet is reached, and then commences a gradual descent. All of this is strangely enough covered with a luxuriant moss forest, though less so than that across the valley at the same altitude. It extends down to about 7000 feet where a decided change is noticeable, (ouo0z a),e10dw9,L) Couog 9} vr1odtu9,L) *499J 00O'OL JO OpNI}V UY 7B Ysa1OJ JURLINXNE$ ATTRUOTdaox Gq ‘seddeg St] op a8 A 2} UT aua0g UANOVNIY AAOWY ‘SOTTINHOH,) So'T YUAAIY VNAIVGOV]Y AHL 40 GOUDOG AHL UVaAN a TAX TLVIG “TIAXXX “TOA ‘HON CIN V Nisaiiag 1917,.] Chapman, Distribution of Bird-life in Colombia. 45 and though still heavily wooded is more second growth and much less moss. The trail descends again to the river at 6500 feet and follows it to 6100 feet; it then ascends another ridge rather steeply up to 6800 feet (Las Cham- bas — stop for night) and descends again to 6100 feet, and the river which it follows for but a short distance rising abruptly and then gradually until 7000 feet is reached and a long gradual descent begun to San Agustin. This latter country is much more open than that first reached at 7000 feet, but there is abundant evidence of deforestation until the last descent is commenced to San Agustin, which seems naturally more arid and less for- ested except along tributary streams which are forested even down below the altitude of San Agustin (6000 ft.). None of these are in the near vicin- ity of the town, however, the country being semi-arid and more or less like that about Cali. Here I was laid up completely and unable to do any col- lecting whatever. Here we met Sefior Nieto of the Bogoté engineers and discovered that our barometer was reading 900 feet too high, so that the altitude of San Agustin should be 5000 feet. When this error commenced I do not know. “The country from San Agustin to Neiva and thence to Giradot in general, is very similar to that about the headwaters of the Cauca, being semiarid (more so than the Cauca) except along the streams, where con- siderable coffee and cacao is raised.” Expedition No. 5.— San Agustin to the Caqueté Region. April 7 - Sept. 1, 1912. Personnel.— L. E. Miller. Itinerary.— Illness having compelled Allen to leave the country, Miller carried out the plans of the original expedition, assisted only by natives. From April 7 to 25, and again on May 19 to 21, he worked in the vicinity of San Agustin going far enough from the city to reach the virgin forest. It was during this period that he discovered a nesting colony of the Cock-of- the-Rock. April 27 to May 5 he was resident at La Palma, and from May 7 to 19 at La Candela, both in the forest respectively south and west of San Agustin. Returning to San Agustin May 20, preparations were made for the trip over the new government trail to the Caqueté Region. The Eastern Andes were crossed at Andalucia (7000 feet) and some col- lecting was done on both eastern and western slopes (May 30-June 20). Florencia was reached June 24, and collections made there until July 6, while at La Morelia the work was pushed vigorously from July 8 to July 26. Mr. Miller is, so far as I am aware, the first ornithologist to enter Amazonian 46 Bulletin American Museum of Natural History. [Vol. XXXVI, Colombia, and his collections from Florencia and La Morelia add many species to the known avifauna of Colombia. Description of Collecting Stations The following notes are supplied by Mr. Miller: La Palma.— “La ‘Palma is the name given a place about one day south of San Agustin, on the same trail we came on from the Cauca. There are a number of clearings and large fields of corn; also some pastureland. Around these extends the virgin forest; altitude 5500 feet. The place is not far from the junction of the Magdalena and Mulales. “The forest is very dense. There are many palms; much large timber and a great deal of moss. Ferns, also, are abundant. The country is mountainside, cut by numerous ravines. Paths there are none, and it was invariably necessary to follow along a small rivulet or ravine. Birds in general were scarce.” La Candela. “A small Indian ranchito, a day west of San Agustin (with pack animals) bears this name. There is a comparatively small clearing, surrounded on all sides by giant forest. The altitude is 6500 feet. The trail for first half of the way leads through the open country that surrounds San Agustin, and then through the forest, and is fair, but narrow so that the pack mules pass with difficulty. “The lower growth of the forest consists almost exclusively of palms which reach a height of perhaps thirty feet. The trees are immense, being the thickest, tallest and straightest I have seen in Colombia, and include giant ‘cedars.’ The forest floor is littered with dried palm leaves, but open and easy to traverse. Birds are not abundant, but more plentiful than at La Palma. “The climate was cool and delightful with but little rain. There is very little moss in the forest.” . Andalucia.— “Before attempting to cross the Eastern Andes, it was thought necessary to make a short survey of the western slope. It had been absolutely impossible to get any reliable information as to the road, etc. to Florencia, and the only way to learn the conditions was to work near the trail and find out from the travellers who chanced along that way. A trip was therefore undertaken to the top of the range, one day from Guada- loupe, the place being called Andalucia, alt. 7000 feet. The altitude of Guadaloupe is 2500 feet. “ Andalucia is a single, very large, boarded house, owned by the Govern- ment, situated on a narrow ridge with a large clearing on each side. At least during this season (May-June) the weather was most severe; fog, strong wind, almost continuous rain and very cold, almost recalling condi- tions on a paramo. Also, the forest was dense, and the vast number of 1917,] Chapman, Distribution of Bird-life in Colombia. 47 fallen trunks and branches rendered the greater part of it impenetrable. Birds were scarce, but small mammals were plentiful. After three days a camp was established further down on the western slope where conditions were more favorable. nS It was our intention to remain only a week at most, but a series of mis- sions in Guadaloupe attracted everybody for miles around including the peons, so I was left alone and of course could not leave until their return, two weeks later. “On the western slope, the great forest extends down to 3,500 feet, and in the cafions and ravines, down to 3000 feet. -The lower part is comparatively open, with numerous ferns and palms, and as one reaches higher altitude there is much moss, many ‘pines,’ parasites, creepers, etc., similar to forests of the same altitude previously described. Birds are not plentiful, and maromals scarce. “Up to 3500 feet there are large areas of wild cane and bamboo along the streams, and in these the most successful trapping was done.”’ The Magdalena Valley. to the Caquetaé Region.— “ About an hour’s ride from Altamira, over a nicely constructed gravel road, takes one to the town of Guadaloupe. Just before reaching the town it is necessary to cross the Rio Saraza, which at this season (June to August) was a swift, muddy stream over a hundred yards wide. The town has a population of about one thousand, numerous small stores, weekly market, etc. All around are thickets of bamboo and wild cane; the altitude is 2450 feet. This zone extends up to nearly 3500 feet. “The new government road follows closely along a small ‘quebrada,’ the name of which I could not ascertain, but it is probably the Imaya or Matayna, and a full day’s travel with packs takes one to Andalucia, altitude 7000 feet, which is practically the top of the range. The forest up to this point has been described elsewhere, and is continuous over the eastern side until Sucre, altitude 2800 feet, is reached, the second night. The only difference here is that one finds less moss and epiphytes. Sucre is a large board house constructed by the government, and contains the telegraph office, the present end of the line which is being constructed by the govern- ment to Florencia. “From Sucre one continues gradually downward until shortly after noon, to an altitude of 1500 feet. Then there is a sharp hill about 500 feet high, called Llegua Gorda. This was the only bad part of the trail, and, ‘on account of the deep mud, two of the three pack mules had to be unloaded and the packs carried to the top. About two hours beyond we camped at a small hut called La Recluta where there is a large clearing, fine pasture, corn, yuccas and other farm products. It is only a half day to Florencia 48 Bulletin American Museum of Natural History. [Vol. XXXVI, from here, over a fairly good road, level, with the exception of a small rise or two which, however, are of no consequence. “Florencia is a small town, with a few hundred inhabitants, but growing rapidly. The altitude is 675 feet. The whole Department of the Caqueta contains but two thousand souls not including Indians, according to the alcalde of Florencia. Provisions can be had at Florencia, but prices of everything but meat and corn are very high. “Our first work was done a short distance above the town, at an eleva- tion of 1000 feet, at the ranch of one Don Blas. The clearing was the largest I had seen in this locality, there being fodder, plantains, cacao and corn. In this open country birds were abundant. The surrounding forest was comparatively open, and not far away. From the elevated position one has a good view of the Caqueta country, a perfect ocean of forest stretch- ing out ahead as far as the eye can see, which on clear days is many miles. The sight is most impressive. There is not a single rise visible and the forest is of uniform height. “The forest is comparatively open, that is, free from dense undergrowth. The trees are tall and there are a few tree ferns, many climbing lilies and also many palms. There is not much moss and along the streams there is much bamboo and also wild cane, often mixed with dense clumps of creepers, tall grass and thorny bushes. In places there are small clumps, perhaps a few acres in extent, of dense low trees resembling cecropias and called “estrojo.”” Streams and rivers are numerous and one is at once impressed with their large size and depth. Also, while swift, they are so silent that one may be near a large river and not know of its presence until at the very edge. “Clouds hang low, often descending to the ground, especially in the early morning and late night, causing a dense fog. We happened to strike the country in the height of the rainy season, but there were frequently intervals of three bright days with not a drop of rain. On other days the showers, which were heavy, were confined to early morning, the afternoon, after 4 P. M., and night. It rarely rained all day long. About 4 P. M. a cool wind invariably sprang up. At noon the heat was rather intense but not nearly so great in the Magdalena Valley (as I later discovered) below Neiva. The nights were cold so that two blankets were none too many. The ex- pedition was without a thermometer so no observations as to temperature could be made. It is said that during the dry season (December, Janu- ary and February) the heat is terrific and there is much fever owing to the clouds of mosquitoes that emerge from the pools left by the receding water. “La Morelia is two days’ southeast from Florencia, between the Bodo- quera and Pescado. It seems as if the elevation should be greater than 1917:] Chapman, Distribution of Bird-life in Colombia. 49 Florencia, but the aneroid registered 600 feet. The trail is bad and all but impassable to mules. There is practically no difference in the forest, but probably there are more streams. There is a ridge of low hills near by, to the east, not over 500 feet higher than the surrounding country, and also heavily forested. Some of the larger birds taken here are said to have been common around Florencia some years ago, but to have retreated with the approach of civilization.” Expedition No. 6.—Tumaco-Barbacoas. J uly 26-Oct. 18, 1912. Personnel.— W. B. Richardson. Itinerary.— Richardson reached Tumaco by steamer from Panama, July 26. He left there July 30, arriving at Barbacoas August 3. In this unhealthful locality he worked until September 10, when an attack of beri- beri forced him to seek a higher altitude and he continued up the trail toward Pasto to Ricaurte, at an elevation of about 4500 feet. He remained at Ricaurte until September 30, and then returned to Barbacoas, stopping on the way down, as he had on the way up, at Buenavista on the Pasto Trail. Barbacoas was left about October 8, and Tumaco reached October 13. From this point Richardson sailed for Esmeraldas, Ecuador, and for the following year collected in that country. Description of Route. and Collecting Stations — The following informa- tion is taken from Richardson’s letters and reports: “The island of Tumaco is dry, sunny, and sandy with only stunted vegetation; and, on one side, mangroves. There are only a few common birds there. On reaching the mainland at Sala Honda, at the mouth of the Patia, everything changes and the next one hundred miles is through a dense swamp of flooded forests. It is inhabited only by negroes who live on the river banks and cultivate patches of rice and plantains and cut wood for the steamer. Their huts are built of bamboo on poles five to eleven feet above the ground, and they last only a few years. When abandoned their thatched roofs are soon converted by nature into veritable ‘roof’ gardens; a mass of vines and parasites, ferns, mosses, and even corn and bananas growing on top of them until they cave in. “ After four days by steamer and canoe, I managed to reach Barbacoas. The surrounding country is much like that which exists between Buena- ventura and Cisneros on the road to Cali, thick, heavy forest and impene- trable jungle all matted together with vines and undergrowth. “Nothing is cultivated but plantains. The only paths through the forest lead to gold washings. For that reason I did much collecting from a canoe. 50 Bulletin American Museum of Natural History. [Vol. XXXVI, “T arrived at Barbacoas (a town of 4000 negroes and 50 whites) in what was supposed to be the dry season, but it rained about twice a day during my stay. Nine months of the year it is said to pour, and even zinc roofs corrode. The climate is like a Turkish bath. “The journey from Barbacoas to Ricaurte is about a 75-mile gradual climb over a good road. The intervening country is very broken and uncul- tivated and is inhabited only along the road, over which hundreds of Indians and mules laden with freight from Pasto and the interior pass daily. “Ricaurte, with an elevation which I believe to be between 4000-4500 feet,!-is on the upper edge of the forested zone which extends to this point from the coast. The country above Ricaurte is open, bare and grassy and through it one may pass over steep ascents to the paramo. “At Ricaurte it is dryer, the forest is less luxuriant and dense than further down, but the country is very broken except along the gradually winding road. I found there many of the birds of San Antonio, Munchique, and Miraflores. ~ “At Buenavista (alt. 1200 ft.) between Barbacoas and Ricaurte it rained steadily for six days. Only once did I get a glimpse of the snow cap of Mt. Cumbal in Ecuador.” Expedition No. 7.— The Bogoté Region. January 19- April 16, 1913. ey Personnel.— Frank M. Chapman, George K. Cherrie, Louis A. Fuertes, Paul G. Howes, Geoffroy O’Connell, Thomas M. Ring. Ttinerary.— On January 19, 1913, we sailed from Barranquilla up the Magdalena, and by taking advantage of stops for fire-wood and cargo, collected 300 birds during our twelve-day voyage to La Dorada, the port of Honda. At Honda (alt. 600 ft.) we remained from February 2 to 9, collecting in the immediate vicinity of the city, at the hacienda El Triunfo, a few miles to the north, and at the beautifully situated posada, El Consuelo (alt. 3300 ft.) distant four hours’ ride on the mule trail to Bogoté. February 10, we left Honda on mules over this trail for Bogoté. No collecting was done en route during the three-day journey to Facatativé, where a train was taken to Bogoté, but our familiarity in life with many of the more common species observed, permitted us to make notes on their altitudinal distribution as we rode slowly through their respective zones. 1 Mr. Richardson’s barometer not having reached him he was unable to learn accurately the eleva- tion of Ricaurte. His collections, however, show that it is in the Subtropical Zone.— F. M. C. Butretin A. M.N. H. Vou. XXXVI, Puate XVII. Coast NEAR CARTHAGENA (Tropical Zone; Caribbean Fauna.) Suores oF THE Lower MacpaLena RIVER The neighboring savannas support many cattle. (Tropical Zone; Caribbean Fauna.) 1917.] Chapman, Distribution of Bird-life in Colombia. 51 We remained in Bogoté from February 12 to 20 buying mules and sup- plies for our proposed journey to Villavicencio, distant some ninety miles by trail, at the eastern base of the Andes, and during this period collected on the Savanna in the vicinity of the city. In order to simplify the problem of transportation and to avoid over- taxing the limited resources of wayside posadas, our party was divided into two sections for the journey to Villavicencio. This plan necessitated the use of only three saddle- and three pack-mules. At the end of a day’s journey of twenty to twenty-five miles, the first section stopped. After resting a day the mules were sent back for the second section. On its arrival the first section advanced another day’s journey. Relayed in this manner we collected to a limited extent en route, at Chipaque (alt. 9000 ft.), Quetame, (alt. 4600 ft.) and Monteredondo (alt. 4500 ft.). The first party reached Buena Vista (alt. 4500 ft.), on the summit of the last ridge of the Andes (the first ridge above Villavicencio) February 28, the second, March 2. March 5, the first party left for Villavicencio, where it was joined by the second party March 10. The first party remained at Villavicencio until March 15, while the second party returned to Buena Vista March 13. In all, therefore, we had nearly two weeks’ collecting at Buena Vista and Villavicencio. March 16, we all left Buena Vista for Bogoté. Additional mules were hired to avoid delay and the journey was accomplished in the regulation pack-train time of three days. March 19, we left Bogota for Fusugasugé, distant about 35 miles to the south at the upper border of the Tropical Zone of the Magdalena Valley. Collections were made in the vicinity of Fusugasuga (alt. 5464 ft.) and at Aguadita (alt. about 6500 ft.) March 25 to 31, at El Roble (alt. 8100 ft.) in the Subtropical Zone, April 1 to 4, and at El Pifion (alt. 9600 ft.) in the Temperate Zone April 1 to 5. April 5 to 9 was devoted to packing, and on April 10 we left Bogota on our homeward journey, during which no birds were collected. Barranquilla to Honda via the Magdalena River.— The Caribbean coast of Colombia, both because of a low and irregular rainfall and the character of the soil is comparatively arid. Acacias, cacti and other xerophytic forms are the prevailing types of vegetation. This region, however, is so remote from Bogota that, so far as I am aware, none of the birds which, in Colombia, are restricted to it are found in Bogota collections. It is not until one has passed Calamar and reached the vicinity of Banco, about 150 miles from the mouth of the Magdalena, that the humid, forested region is reached. More favorable soil and increased rainfall, doubtless following condensation attributable to the proximity of the Eastern Andes, 52 Bulletin American Museum of Natural History. [Vol. XXXVI, presumably are responsible for the change from the open, scrubby vegetation of the arid coastal area to the luxuriant forests which now almost continu- ously line both shores of the river. Away from the border of the river, however, at least on its eastern bank, the arid zone continues as far as Puerto Nacional whence, according to Wyatt (Ibis, 1871, p. 117) “the first few miles” of the road to Ocafia runs through “small savannas, tracts of open grassy country sprinkled with a few stunted trees, or through woods.” In the more northern part of this humid region tributary streams may make their contribution to the muddy waters of the Magdalena through marshy or low-lying land, but farther up the river the banks are higher and the shores of entering streams are forested. The humid zone of the floor of the Magdalena continues with no diminu- tion in the luxuriance of the vegetation as far up the river as La Dorada, about 600 miles from its mouth. Between this place and Honda a marked change occurs. Strongly eroded buttes with castellated outlines appear, the soil is thinner and less fertile, and although the rainfall is not so low as at Barranquilla (Mr. Miller, the manager of the railway between La Dorada and Ambalema, tells me that at Mariquita, a few miles east of Honda, it has ranged in a few years observation from 85 to 100 inches annually) the vegetation suggests that of an arid or semi-arid region. The heavy forests are replaced by a more stunted growth and there are large tracts of open country devoted to grazing. This condition apparently prevails to the head of the Magdalena Valley. é Honda to Bogoté.— In a region which has been inhabited by white man for as many years as that lying between Honda and Giradot, and between these towns and the plateau of Bogoté, it is often difficult to determine just what changes man has wrought in the character of the country. At present, however, in following either the mule trail from Honda or the railroad from Giradot, one sees but little forest growth between the Magdalena river and the Savanna of Bogota. In the upper Magdalena Valley proper, the absence of heavy forest, as has been remarked, is doubtless due to the character of the soil, but on the mountain slopes the first-growth timber has no doubt disappeared in many places before the agriculturist. Remains of this forest were discovered between El Consuelo and El Alto de Sargento on the first ridge -of the Andes east of Honda, where at an altitude of some 4000 feet, we found such characteristic species of the Tropical Zone as Formicarius analis and Myrmelastes immaculatus. At El Vergel (alt. 5500 ft.), on the summit of the second ridge, or that lying east of Guaduas, there is a small area of apparently primeval forest in which oaks, some 75 feet in height, were prominent and the presence here of Xanthoura yncas galeata, Brachyspiza capensis, Melanerpes flavi- Buuietin A. M.N. H. Vor. XXXVI, Prare XVIIT. CentraL Lower MaapaLtena River The country is heavily forested. (Tropical Zone; Cauca-Magdalena Fauna.) sr A Woop Yarp In THE MaGpaLeNna Forests Many birds were collected at such localities when the steamer stopped for fuel. (Tropical Zone; Cauca-Magdalena Fauna.) 1917] Chapman, Distribution of Bird-life in Colombia. 53 gula and Grallaria ruficapilla indicated that we had reached the second or Subtropical Zone. Beyond this place the country is again more or less thickly settled and coffee plantations have replaced the forest which probably once existed here. Certain birds were abundant but the conditions were obviously unnatural and far from satisfactory to one who would study animal life in’ an undisturbed environment. Possibly owing to the absence of the luxuri- ant forest which usually occurs at an altitude of from 6000 to 9000 feet, such birds of the Temperate Zone as Semimerula gigas and Sturnella magna meridionalis were encountered as low as 6000 feet. At an altitude of 7300 feet, on the line of the railroad from Giradot to Facatativé, one passes through a broad belt of superb first-growth forest, such as doubtless once occupied the slopes now given to agriculture on the Honda trail, but at no other place was primeval forest observed from the railway. Aside, therefore, from the few days at El Consuelo and obser- vations made from mule-back on the road to Facatativaé, we did no work in the country lying between Honda and the Bogoté Savanna. In the country above Fusugasugé, however, to be presently described, primeval conditions were found and representative collections made of the birdlife of the Subtropical and Temperate Zones of the western slope of the Eastern Andes. | The Bogoté Savanna.— Bogota, a locality to which so many species of birds have been attributed, has, as a matter of fact, a comparatively restricted avifauna. Situated at an elevation of 8600 feet, near the southern end of the great Savanna which is so striking and unusual a feature of Colombian Andean topography, and at the western base of the chain which encloses the Savanna at the east, it is in the arid portion of the Temperate Zone. The word arid, as used here, does not necessarily imply sterility, but indi- cates the existence of conditions which prevent forest growth in a zone where, under favorable circumstances, such growth should occur. For example, at the altitude of Bogoté on the trail from that city to Fusugasuga beyond Cibaté, luxuriant forest growth is found and, in consequence, the upper limit of the Subtropical Zone here reaches upward to somewhat over 9000 feet, or nearly, if not quite, to the divide at El Pifion. This forest is obviously due to the heavy rainfall which prevails at that point, just as on the Savanna of Bogoté the lack of forest is possibly attribut- able to insufficient rainfall! However this may be, practically the only tree we saw on the Savanna between Facatativé and Cibaté, is the intro- 1 The rainfall at BogotA is given by Petre (‘The Republic of Colombia,’ London, 1906) as 42 inches for the six months’ wet season. 54 Bulletin American Museum of Natural History. [Vol. XXXVI, duced eucalyptus, and the existence here of a peculiar race (O. a. peregrina) of such a plains-loving species as Otocoris alpestris implies that the region is naturally treeless.’ ; So far as we observed, every available square foot of the Savanna is used for pasturage or agricultural purposes, chiefly the growing of wheat and corn, the grazing area, however, prevailing. The Bogoté River, as it winds through the Savanna, in places widens into small pools the banks of which are bordered by reeds and cat-tails. In the rainy season depressions of from a few square yards to others of several hundred acres or more become lagoons, and it is in these restricted localities that the resident, as well as winter visitant water-birds of the Savanna are found. To the North American ornithologist the bird-life of the Savanna holds so many familiar forms that it was difficult for us to realize that we were .. within 300 miles of the Equator. By ascending the mountains Guadalupe or Mont Serraté one may reach ‘the Paramo Zone, at an elevation of between 11,000 and 12,000 feet. Our plans to visit this zone with its restricted avifauna, did not, however, mature, and we touched this upper life-zone only at its lower border on the higher parts of the trail between Bogota4 and Chipaque. , Bogoté to Villavicencio.— The trail from Bogota to Villavicencio and the Llanos of eastern Colombia leads directly over that ridge of the Eastern Andes at the western foot of which the city lies. A few squares south of the Central Plaza one turns eastward and the ascent begins before the city limits are reached: The country is rolling rather than precipitous, and for a considerable distance the trail leads over comparatively level country. The actual divide is situated at the extreme eastern edge of the ridge, some ten miles from Bogoté, where from the mouth of the pass, at an altitude of approximately 10,700 feet, one looks down the extremely steep eastern slope to the valley of Chipaque two thousand feet below. The average height of that part of the ridge traversed by the trail is about 10,000 feet, and nowhere does it rise higher than 10,300 until the pass is approached. In limited areas well-developed Temperate Zone forest exists, but the country for the greater part is covered with a bushy scrub, or with low ferns. Both to the north and south cones or spurs of the ridge rise at a sharp angle to as much as 2000 feet above the trail. On the slopes with a northerly exposure, timber- line extended to approximately 11,000-11,500 feet. On southerly slopes it was about 500 feet lower and under these conditions, frailejons, one of the most characteristic paramo plants, grew abundantly almost down to the level.of the trail. The pass, using this term in the broad sense to cover the higher parts of the trail between Bogota and a point where the descent to Chipaque begins, lies in the Temperate Zone and though it is frequently Bouuuetin A. M.N. H. Vor. XXXVI, Prate XIX. SLOPES ABOVE Bocota Junction of Temperate and Paramo Zones. Tur Environs or Bocora View of the Savanna from the beginning of the trail to Villavicencio. (Temperate Zone.) 1917] Chapman, Distribution of Bird-life in Colombia. 55 referred to as paramo, it barely reaches the level at which true paramo be- gins. The prevailing winds are easterly and the clouds which have passed over the lower ridges to the east here give up their moisture, producing a climate marked by chilling winds with driving mist and rain. The upper portions of the eastern slope above Chipaque are covered with dense, but rather poorly developed Temperate Zone forest, the last forest-growth through which the trail passes until one reaches the eastern- most ridge in the range. : As one descends to Chipaque the forest, doubtless partly because of natural conditions and partly because of the demands of agriculture, gradu- ally disappears. We were favored in securing quarters at an hacienda about a mile west of the town of Chipaque and some five hundred feet above it. The avifauna here was characteristic of that of the Temperate Zone, .a single specimen of Gallinago nobilis and one of Cistothorus equatorialis taken at approximately ten thousand feet, indicated our proximity to the Paramo Zone above, while a single specimen of Henicorhina leucophrys guttata taken in the dense undergrowth bordering a deep ravine cut by the C4queza, illustrated the tendency of a lower zone fauna to penetrate the zone above along the protected banks of barrancas. The absence of forests between the summit of the range above Chipaque and the most eastern ridges of the range, or approximately between 9500 and 4500 feet altitude, was a great disappointment to us, since it prevented us from securing a collection in the Subtropical Zone forests of the range. This is the most serious gap in our field work in the Bogoté region. Such forest doubtless exists in other parts of the range, but in the region traversed by us it was restricted to the summits of the higher mountains and ridges where, under the circumstances, it was not accessible. That its exploration would yield most interesting results for comparison with those obtained near Fusugasugé is indicated by the discovery of a new form of Ostinops sincipitalis, taken by Ring with much difficulty from a forest-crowned summit rising to 1000 feet above Monteredondo, and by the striking differ- ence found to exist between the Jays of the genus Xanthoura inhabiting opposite slopes of the range. At Quetame (alt. 4800 ft.), our next station, the trail continues to pass - through an arid, treeless region with some tree-growth along the water- courses in ravines or lateral barrancas, and occasionally a crown of forest on some rounded crest high enough to receive moisture. The first evidences of Amazonian bird-life were observed just east of ‘C4queza, and about thirty miles from Bogota where, at an altitude of 5500 feet, Tanagra episcopus was noted, and from this point it became increas- ingly common as we journeyed toward the Llanos. 56 Bulletin American Museum of Natural History. [Vol. XXXVI, The scarcity of arborescent vegetation at Quetame necessarily limited the numbers of species of birds found there. Altitudinally in the Sub- tropical Zone, only the scanty growth of timber along streams flowing into the Rio Negro itself furnishes a haunt for the tree-inhabiting species of this zone, but in such localities the few species secured were members of the zone in which Quetame is situated. Examples are Xanthoura cyanodorsalis, Grallaria ruficeps, and Cinclus. But mingled with these birds were outlying representatives of the Tropic Zone below; for example Planesticus ignobilis and Tanagra. episcopus, while on the grassy hillsides or along the hedge-rows such characteristic species — of the Temperate Zone as Sturnella meridionalis and Planesticus gigas found ) their lower limit. At Quetame, then, although the avifauna was in the main that of the Subtropic Zone, representatives of both the zone below (Tropic) and zone above (Temperate) met, a condition we have not found elsewhere. At Monteredondo, some five miles east of Quetame, but at about the ‘same altitude (4800 ft.), arborescent vegetation was somewhat more ‘developed and from this point it increased steadily in size and abundance. Ten miles further east the mountain slopes rising from the southern side of the Rio Negro were heavily forested from base to summit, here a matter of about 3000 feet, but the slopes on the northern side, or those having a southerly exposure, were still comparatively bare; a condition possibly due to the fact that the prevailing winds are northeast rather than southeast. As we traveled_eastward the forested areas continued to increase, the most eastern ridge of the range being covered with superbly developed primeval woods from the Llanos at their base to their crest (alt. 4500 ft.). On the western side, where the slopes reached the Rio Negro, now some 2500 feet below, the forest continued to the water’s edge. In this region we made our base at the posada of Buena Vista, situated at the side of the trail on the summit of the ridge directly above Villavicencio lying some 3000 feet below at the base of the Andes. Buena Vista was our most productive station. From the surrounding region have come many Bogoté skins. Within two hundred yards of the posada lie as finely developed tropical forests as I have ever entered; the trees are of exceptional height (averaging over one hundred feet), the forest floor is comparatively open. Numerous trails greatly facilitate the. passage of the collector and I recall with unalloyed pleasure our experiences in this delightful locality. The fauna of Buena Vista is mainly that of the Tropical Zone, with the addition of some species from the zone above. The fact, however, that the ridge reaches an altitude of only 4500 feet, and that there are no higher ridges nearby, evidently limits the number of Subtropical Zone representa- tives. Bouuuetin A. M. N. H. Vor. XXXVI, Piare XX. EasTERN ANDES BETWEEN BoGotTa AND CHIPAQUE (Temperate Zone.) CHIPAQUE View looking west. (Temperate Zone.) 1917,] Chapman, Distribution of Bird-life in Colombia. 57 As has been remarked, the eastern slope of the eastern ridge of the Andes is forested from summit to base. At the foot of the ridge which rises abruptly from the level plain at its feet, the forest of the mountain gives way to the grasses of the Llanos. The environs of Villavicencio, to which our collecting was confined, are largely under cultivation. The original llano grasses have given way to Para and Guinea grass; various crops are cultivated; trees border the byways and are found in clumps on small areas in the bottom-lands. Somewhat farther east stretches of forest occur. The differences between the bird-life of Buena Vista and that of Villavicencio are due to the char- acter of the country rather than to altitude. Doubtless had we collected in the forests at the base of the mountain, we should have found many species which we secured only at Buena Vista. Bogoté to Fusugasugé4.— Our disappointment in finding so little undis- turbed, primeval country along the trail from Honda and the railroad from Giradot to the tableland was forgotten when we discovered the admirable conditions for collecting which exist between Bogoté and Fusugasugé. As aregion to which Bogotafios repair for climatic change, Fusugasuga is often thought of as a suburb of the larger city, a conception which is fostered by the apparent proximity of the two cities as seen on maps. It is probable, however, that, excepting a few raptorial species, not two birds are common to both places. Beginning our journey at Cibaté, at the southern end of the Savanna, it is doubtful if, even in Colombia, one could encounter more pronounced faunal changes than occur in the fifteen miles lying between that place and Fusugasuga. Cibaté has the same altitude as Bogoté and, except to the south, the surrounding country, and doubtless its bird-life, are essentially similar; but within a few minutes after leaving the railway station, which marks the terminus of the twenty-miles of track constituting the Ferrocaril del Sur, one begins to climb the low hills which form the southern rim of the Savanna. Where not under cultivation, the ground is covered with a scrubby second growth which, within three miles, is gradually replaced by the low, stunted, moss-covered forest of the humid Temperate Zone. The ascent is now barely perceptible but it continues to the posada of El Pifion, at the extreme southern edge of the ridge dividing the Savanna of Bogota from the slopes below it. At this point the trail drops abruptly into the almost gorge-like valley which leads to Fusugasug4. On each side, and separated by less than a mile, stand the precipitous walls which flank the valley. Heated air from the broad, radiating surfaces of the semi-arid | Magdalena plains below is _ drawn ‘up ‘this slit in the mountains and at El Pifion meets the cooler 58 Bulletin American Museum of Natural History. (Vol. XXXVI, atmosphere of the tableland. Condensation ensues and in consequence the rainfall here, and in the valley immediately below, is doubtless unusually high. Evidently for this reason, the forest of the Temperate Zone at El Pifion is more luxuriant than we have found it elsewhere, while the floor and where circumstances permitted, the sides of the valley below were clothed with heavy, subtropical forest broken only by artificial clearings until one reached a point some 1500 feet above Fusugasugé. From this point down- ward the forest has been cleared and replaced by the coffee plantations which surround the town of Fusugasugé. Our nearest collecting station to El Pifion was El Roble, a posada 1000 feet lower down on the trail to Fusugasugé. Nowhere in Colombia have we found so great a faunal change in so short a distance as that which occurs between these two points. Indeed one has to go only a few hundred feet below El Pifion to pass from the Temperate, completely into the Subtropical Zone. So steep is the trail that one seems to be descending a flight of stairs. Within a dozen steps the rolling ground of the dividing ridge is lost to view, and one is at once protected from the chill winds of the tableland. Very quickly a striking change is observed in the vegetation as the larger, more open-limbed, liane-draped trees of the Subtropical Zone replace the smaller, thickly branched, moss-covered ones of the Temperate Zone. About 1200 feet below. El Roble, we collected at a way-side posada known as Aguadita. The valley is here somewhat wider, but the heavy subtropi- cal forest, essentially like that found at El Roble, and broken only by occa- sional clearings, still prevails. A short distance below Aguadita the primeval forest ends and the coffee plantations begin and continue to and beyond Fusugasugé. While climati- cally in the Subtropical Zone, the clearing away of the original forest-growth has permitted a number of species characteristic of the semi-arid Tropical - Zone of the Magdalena Valley to extend their range up the mountain slope. Examples are Mimiis, Tanagra cana and T. palmarum.*’ Our party was stationed only a day or two at Fusugasug4, collections being made from this -point in the forests 1500 feet above the city. Expedition No. 8.—The Antioquia Region. November, 1914— March 26, 1916. Personnel.— Leo E. Miller; Howarth Boyle. Itinerary— Miller and Boyle reached Medellin via the Magdalena River to Puerto Berrio, November 11, 1914. After establishing their base ‘in this city they proceeded at once to Sta. Elena, one of Salmon’s most ‘important collecting stations,.on the summit of the first ridge of the Central Buiietin A. M.N. H. Vor. XXXVI, Pirate XXI. <3 Rio Necro rrom Montrereponno Upper limits of arborescent vegetation in the Rio Negro Valley; Bogota-Villavivencio Trail at the right. (Tropical Zone; Orinocan Fauna.) JuNncTIon oF Rio Caqueza ano Rio Necro A scene near Caqueza. Bogota-Villavicencio Trail leaves the floor of the valley at the right. (Fusion of Temperate, Subtropical, and Tropical Zones.) 1917.) Chapman, Distribution of Bird-life in Colombia. 59 Andes east of Medellin. Here they worked from December 15 to 23, and then proceeded twelve miles further east to Barro Blanco remaining there from November 25 to 29. Returning to Medellin they collected for three more days (Dec. 1-4) at Sta. Elena and arrived at their base December 5. In order to determine to what extent the fauna of the lower Atrato has reached the heavily forested portions of the lower Cauca region, the expedi- tion left Medellin December 9 and reached Puerto Valdivia at the head of navigation on the Cauca five days later. Collections were made here until December 26, when the party retraced its steps to La Frijolera, some 5000 feet higher, working there from December 29 to January 4. About 600 birds were secured on this lower Cauca trip. These, in connection with Salmon’s records from Remedios, Miller and Boyle’s later work at Malena near Puerto Berrio, and some 300 specimens collected chiefly by the Bogota expedition along the lower Magdalena, doubtless give a fair indication of the extent to which Pacific coast forms have entered this region. Return- ing to Medellin preparations were made for the ascent of the zodlogi- cally unknown Paramillo, at the northern end of the Western Andes, the most important and most difficult piece of exploration planned for this expedition. : Medellin was left January 14, and Peque reached on the 19th. Here the mules were replaced by Indian porters, and after four days, which evi- dently made exceptional demands on the strength and perseverance of the explorers, the Paramo was reached January 23. Between this date and February 1, over 150 birds were secured. Several were new to science and the collection as a whole very clearly shows the character of this, the highest point found by us in the Western Andes. Most of the species belong to the upper Temperate rather than the Paramo or Alpine Zone; such char- acteristic Paramo species as Phrygilus and Upucerthia were apparently wanting. On the other hand, a number of species were taken which had been previously found on the crest of the Andes west of Popayan, the only other point at which we have discovered a Temperate Zone in the Western Andes. Chief among these is an excellent series of Diglossa gloriosissima, hitherto known only from the specimens secured by Richardson and Miller at an elevation of 10,300 feet west of Popayan. A fine series was also taken of Diglossa brunneiventris, previously found in Colombia only by Salmon at Sta. Elena, showing that the Colombian form is not separable from the Peruvian race. The final work planned for this most successful expedition was on the Atrato slopes of the Western Andes. From the Paramillo, Miller and Boyle returned February 9 to Buritica, left this place on the 9th and reached Dabeiba, their first station, on the 11th. Collections were made here from 60 Bulletin American Museum of Natural History. [Vol. XXXVI, February 12 to 14, when they moved ten miles further down the river to Alto Bonito where, between February 16 and 23, they secured 255 birds. A two-day stop (Dec. 25-26) was made at Dabeiba in returning to Medellin. Only two stations now remained to be visited; Malena near Puerto Berrio, selected by Miller as a favorable point on his way into Medellin, and La Playa, near Barranquilla. One hundred birds were taken at the first-named place March 9-11; and one hundred and five at the latter, March 23-26. Description of Route and Collecting Stations The following notes are supplied by Mr. Miller: La Playa:—“At La Playa, a few miles from Barranquilla, one enters a typical stretch of the Arid Coastal Zone. The country is level or gently undulating, sandy, and covered with a sparse growth of cacti and thorny shrubs. Toward the river vast shallow salt lakes, swamps and mangrove thickets abound. “‘T was greatly surprised to find that the dry, sand-dune country con- tained an abundant and varied bird-life, while the green mangrove jungles were practically uninhabited. In the lagoons, Pelicans, Black Jacanas, Herons, Sandpipers and Anhingas are very abundant, and Terns occasion- ally pay them a visit. “Tn the dry brush, large Wrens, Synallazis, Pigmy Owls, Mockingbirds, Thamnophilus, Ground Doves, Sycalis, Orioles and Honey Creepers (sugar birds) form the characteristic avifauna; and occasionally one meets Buccos, Piculets, Woodpeckers and Herons, Parrots and Parrakeets. Many of the bushes are loaded with bulky nests; but they are well protected by the thorny branches, in spite of the fact that they are most conspicuous. One visit was made at the end of the breeding season, though some species were still with eggs and small young (March 23-26.) ” Malena:—“ The first stop on the railway line beyond Puerto Berrio is the village called Malena, and as the locality appeared to offer ideal facili- ties for collecting, the expedition moved to that point March 10th and spent four days collecting in the surrounding forest. “Malena contains perhaps forty huts, and is situated in the heart of the immense Magdalena forest. The forest is high consisting mainly of giant ceibas, with comparatively little undergrowth; the tagua palm is abundant. There is also an abundance of bamboo. “ Bird-life fairly teems along the edge of the forest, but in the woods there is little life except clouds of mosquitoes. “During our visit there was but little rain and everything was dry. Birds seemed to adhere to the banks of the small streams and, with few ex- ceptions, did not band together in flocks.” 1917,] Chapman, Distribution of Bird-life in Colombia. 61 Puerto Berrio to Medellin:—‘ Coming up from Puerto Berrio, the heavy Magdalena forest extends up until only about 1500 feet, although patches of it ascend much higher, to the top of the lower foothills several thousand feet high; beyond the forest belt begins open brush and scattered forest country. Wild cane and bamboo jungles grow in profusion up to Cisneros, alt. 3700 feet, which is the end of the eastern stretch of railroad. Beyond Cisneros rocks appear — sandstone and granite, in boulders and outcrop- ping ledges. By the time the highest point is reached (carriage road) at La Quiebra, altitude 5425 feet, the hills are practically bare. Coming down from La Quiebra toward Medellin the country is the same. The railroad again begins at Botero and follows the Medellin River to the city; the valley of this river is covered with a tall, slender willow growth which I have not seen in any other place in Colombia. The trees are like some seen about Popayan where a few have been planted along a driveway; but here they form a pretty compact forest.” Santa Elena and Barro Blanco:—‘ Santa Elena, is an interesting place with an elevation of 9000 feet, which is the top of the ridge, although there are a few peaks near which reach to 9500 feet. The temperature ranged between 40° and 62°, and the vegetation reminds me much of the crest of the Andes west of Popayan. There are, however, a few patches of forest, but most of the country is grassy and brush-covered. “Continuing toward Rio Negro (a large town) from Sta. Elena, a good trail goes gently downward into an immense, practically level plateau from 7000 to 8000 feet high. All of this is pretty well settled, devoid of forest, and ‘cultivated; corn fields cover practically the whole region, with an occasional pasture and small area of brush. Beyond the plateau the peaks again rise about a thousand feet, with open or brush-covered sides and tops fringed with low forest. Barro Blanco (our camp) was at the foot of this hemming- in ridge. Although the altitude is 7200 feet, the heat at mid-day is intense. ‘Tall fan palms grow, though in small numbers, and there are clumps of feath- ery bamboo. The fauna is quite different from that at Sta. Elena although some forms such as Planesticus gigas and Brachyspiza are abundant. It does not seem possible that this part of the Central Range was ever forested, that is, as heavily as the jungle we found at San Antonio, above Cali, although there are evidences that vast tracts have been cleared. All indica- tions are that the mountainsides were to a large extent brush-covered with bushes, ferns, and climbing bamboo up to fifteen feet high, with clumps of low, rather open forest on the peaks. The soil for the main part is clay and rocky. December, January and February, June, August and Septem- ber are the dry months; March, April and May, July, October, and No- vember are the ‘winter’ months, during which it rains. Our experience 62 Bulletin American Museum of Natural History. [Vol. XXXVI, is that there are pretty heavy showers in the late afternoon and rarely a rainstorm at night; the mornings were bright and cloudy, with low-hanging clouds until shortly after sunrise. Occasional gusts of clouds blew in during the mornings, but lasted a few minutes only. The prevailing wind was from the east.” Medellin to Puerto Valdivia:—‘ From Medellin we took the train to Barbosa, an hour and a half away; then we took mules. The trail at first goes up very abruptly from 4625 feet to 8100 feet, which point we reached at noon, three hours after starting. The country is barren of forest, although there are a few small patches of brush. The high plateau is rough and broken with many granite boulders strewn about. Santa Rosa, 9200 feet, seven leagues from Barbosa is a town of a few hundred houses, situated in almost desert country; there are numerous mines in the vicinity, and many diggings and tunnels are visible from the trail. This dry, desert country continues for about two leagues beyond Santa Rosa; then small patches.of open, rather stunted forest begin and continue for three leagues, interspersed with llanos; this forest has little undergrowth, but the trunks and branches are covered with short yellowish moss; apparently there is not much rainfall. Woodpeckers (Melanerpes) abound in this semi- forested zone. Now follow two mere leagues of almost barren country until the town of Yarumal, 7000 feet, is reached. Yarumel is a good-sized town, nearly as large as Santa Rosa. It rests on a steep hillside, so steep, in fact, that it is difficult to walk on the streets. “About a league beyond Yarumal magnificent first-growth forest begins, and continues with minor interruptions only until Valdivia, six leagues away. This forest reminds me much of that at San Antonio, above Cali, and there is doubtless an abundance of rain; small torrents are also numer- ous. The altitude of Valdivia is 4200 feet. In the immediate vicinity of the town the forest has been cut away, but a mile beyond it again starts and continues down to the Cauca River. This lowland forest is as tall or taller, but has less moss, ete. than the high country forest. The distance from Valdivia to Puerto Valdivia is one and a half leagues. We made the trip from Medellin to the port in four and a half days. “The Cauca, at Puerto Valdivia (alt. 360 ft.), flows directly between the Western and Central Ranges, without any valley whatever. The mountains slope up sharply right from the water’s edge on both sides and are heavily forested except for a few small clearings where corn and cacoa grow, but the clearings are too few.and far between to amount to anything. “The climate was hot, the temperature often reaching 85°, but a daily breeze in the afternoon, blowing up the Cauca, cooled the atmosphere con- siderably. We had comparatively little rain. The rainy months are April, Buuuetin A. M.N. H. Vout. XXXVI, Pirate XXII. CountTRY NEAR Sta. Evena, CENTRAL ANDES The original forest has largely disappeared. (Fusion of Subtropical and Temperate Zones.) WestTeRN ANDES NEAR ANTIOQUIA (Tropical Zone; arid portion of the Cauca-Magdalena Fauna.) 1917,] Chapman, Distribution of Bird-life in Colomiia. . 63 June, July, August, October and November, with violent wind and hail storms in July. The bird-life is abundant and interesting. “On the return trip we stopped at a place called Frijolera (alt. 5000 ft.) just this side of the town of Valdivia, principally to collect mammals for which the country looked promising; here we took many more fine birds, : and a good many mammals of numerous species.” Exploration of the Paramillo— “The name Paramillo is applied to that lofty spur of the Andes jutting out of the western range slightly below lati- tude 7°. To explore this section the expedition left its base at Medellin on January 14, 1915, with equipment sufficient for about three weeks’ actual field-work. “The very good trail strikes toward the northwest, ascending the moun- tainside rapidly, so that four hours after starting we had reached the top of the range. A great cleft forms a natural pass 8750 feet high, and saves a climb of at least an additional 1000 feet. The slope on the other (western) side is more gentle. * “We were immediately impressed with the barren nature of the country for, with the exception of a few patches of low brush and the clumps of withered grass, no vegetation was to be seen; and an occasional glimpse of the Cauca River far below suggested the picture of a broad yellow ribbon lying upon a brown, rocky plain. “That night we reached a small town called San Geronimo, elevation 3200 feet. Near the town small patches of ground are irrigated with water brought from mountain brooks and distributed through a network of arti- ficial ditches; in these spots rice, corn and pasturage grow but rather scantily on account of the rocky nature of the soil. “Next morning we were on, the road before six; a few hours later, on crossing the top of a small ridge, we came suddenly upon the town of Sope- trén completely hidden in a fertile little valley filled with palms, mangoes, and other beautiful trees; the cluster of some hundreds of neat white houses with red tile roofs, the well-kept streets, and the multitude of birds flutter- ing among the deep green foliage rendered Sopetrdn quite the most attractive town of its size I have seen in Tropical America. At noon we reached the Cauca and crossed that sluggish, muddy stream on a well-built suspension bridge probably 800 feet long. Gravel: banks flank the sides of the river, and bare sandy islands divide its waters; the elevation at this point is approximately 2000 feet. One league beyond the Cauca lies the town of Antioquia, altitude 2600 feet. The valley of the Cauca is here five to ten miles wide, rolling; and supports no vegetation except occasional clumps of mimosas and cacti which rather add to its desert-like appearance. The high ranges of the Western and Central Andes hem it in like huge walls of pink clay and sandstone. 64 Bulletin American Museum of Natural History. [Vol. XXXVI, “January 16, we reached Buriticé. Immediately after leaving Antio- quia, a mere ledge of a trail begins the ascent of the Coast Range, and while the safety of the two pack animals caused a good deal of anxiety, it was nevertheless a relief to escape from the intolerable heat of the low country. The altitude of Buritied is 6200 feet. “On account of the jaded condition of the animals, we spent the morn- ing of January 17, at Buriticdé; we took advantage of this time to divide the equipment, leaving such material as we expected to use on our subsequent visit to the.Rio Sucio. Leaving at noon, we reached a small settlement known as Tabocal, altitude 5400 feet, at 5 P. M. We could now no longer see the Cauca, our view having been shut off by a ridge of mountains sev- eral thousand feet in height which rises out of the valley between the-ridge we were on and the river. A slight change was perceptible in the character of the country; extensive areas covered with low brush dotted the other- wise barren landscape, though far apart; and on the extreme tops a both ranges a thin fringe of greeil could be distinctly seen. “Beyond Tabocal the country is extremely broken, there being frequent rises and descents of 2000 feet; and several separate mountains, not con- nected with the main ranges, stand here and there like huge monuments, rising from a basal elevation of 3000 feet to 8000 or 9000 feet, which natur- ally magnifies their already impressive proportions. “Late in the afternoon of the 18th, we reached an altitude of 8000 feet and entered a fine strip of forest, the first we had seen since leaving Medellin; this is the beginning of the forested zone, which examination showed to be at an equal height:on both the Central and Western Ranges, and to con- tinue to the tops, which appear to rise to an altitude of 9000 feet or more. The night was spent at an Indian hut called La:Meseta, altitude 7900 feet, just below the forest belt, and situated in the midst of an extensive strip of maize. “Peque, the end of the journey by mule, was reached at noon on the 19th. After leaving La Meseta the trail goes down abruptly; the town has an altitude of only 5000 feet. ““Peque boasts of about fifty decaying mud huts and its population is mostly of Indian descent, including some pure-blooded Indians; one of the latter, Julian David, received us most cordially and rendered us every pos- sible assistance in securing the porters for the ascent of the Paramillo. “Some of the country surrounding Peque once doubtless bore a light forest growth, with heavier forest in the ravines; but by far the’ greater part is naturally bare or covered with a dense growth of brush. I was told that at: the time of the Spanish Invasion, 40,000 Indians inhabited this region; and as there are several mountain streams supplying an abundance Buiietin A. M.N. H. Vor. XXXVI, Plate XXIII. Tue ParaMILLO, WESTERN ANDES Camp of Expedition No. 8 at left; alt. 12,500 ft. (Paramo Zone.) CHARACTERISTIC VEGETATION ON THE PARAMILLO (Paramo Zone.) 1917] Chapman, Distribution of Bird-life in Colombia. 65 of fresh water and the soil responds fairly well to cultivation, there seems to be no reasion why it should not have supported an extensive population. - “The forest zone which, as stated before, begins at La Meseta, at 8000 feet, gradually extends its limits downward as we go farther north, until at Peque it reached as low as 5000 feet in the deeper and well-watered ravines; and as previously reported, at Puerto Valdivia it reaches the very edge of the Cauca. “We secured four half-breed porters to carry the equipment; and as there was no trail to the Paramillo, a fifth man was secured to go in advance and clear an opening with his machete. “On the 21st we started at 6 A. M., following a short trail that led to a lonely hut known as El Madeiro; this three hours’ walk took us through country covered with large areas of tall brush, blackberry briars and guavas, with occasional patches of forest, some of which had recently been burnt. Arriving at El Madeiro (8000 ft.) we plunged into the magnificent forest, going in a due westerly direction; it was our plan to follow along the top of an undulating ridge, which one of the men said was the shortest and easiest route. “ At first the forest was fairly penetrable, but soon it assumed the charac- ter of the well-known San Antonio (above Cali) jungle, being composed of a solid wall of moss, ferns, creepers and epiphytes which burdened every tree-trunk and branch. “On account of the long climb, we made camp at 3 P. M., at an altitude of 10,000 feet, having ascended 5000 feet in eight hours actual marching. Water was obtained in a ravine over 1000 feet lower down on one side of the ridge, and I may here add that this was the only water we had until reaching the Paramillo, so that we went nearly two whole days without drinking. “The second day’s march we had hoped would be over a gentler slope; but it was soon discovered that our ridge was composed of a succession of knolls rising from 500 to 1000 feet above the main level, and the forest grew denser constantly. We had to cut practically every foot of the way. In places we actually walked over the top of the masses of vegetation; the branches were a solid tangle of creepers, climbing bamboo, bromelias and mosses, and formed spongy aérial bridges; more often it was easier to bur- row through, and frequently ‘tunnels’ many yards long were cut through which the carriers ctawled on hands and knees. The tops of some of the hills were void of trees, their place being taken by a dense growth of grass-like bamboo, wild oleander, thick-leaved shrubs, and thickets of a tall, coarse grass with leaves eight feet tall and six inches wide. We camped this night 11,350 feet up; the men eagerly cut down clumps of 66 Bulletin American Museum of Natural History. [Vol. XXXVI, bromelias hoping to obtain water, but all the leaves contained were a few drops of liquid mud, utterly unfit for use. Although we travelled steadily for ten hours, I doubt if we covered more than three miles. ; “ A few hours after starting on the morning of January 23, we emerged suddenly from the dark forest; instead of the tall, overburdened trees, there were extensive areas of bushes, evergreens, stunted pines, and plants with thick, round, rubbery leaves, interspersed with clumps of tall, rank ferns. Beyond:stretched the bleak; wind-swept surface of the Paramo. “The Paramillo region is composed of a series of sharply inclined peaks, the highest of which attains an elevation of 13,000 feet, interspersed with ravines and deep fissures. The surface consists mainly of dark sandstone which in many places has been shattered so that a thin litter of the particles cover the basal rock; occasionally a thin vein of white quartz crops out, especially where, as often occurs, the strata stand in a perpendicular posi- tion. Water there is none; we discovered but one small trickling brooklet; but at the bottom of one of the crevices several pot-holes were found, each containing several hundred gallons, and apparently remaining perpetually full. “Each night the temperature fell to 28°, and ice formed in our pails half- an-inch thick; in the morning the ground was white with frost. On the evenings of January 28, 29 and 30, there fell very slight, short showers of rain. The prevailing wind came from the west. “The vegetation is of a typical paramo character, consisting of low clumps of ‘frailejones,’ blueberry bushes and tough grass. In the ravines there are thick bushes and stunted trees, all heavily moss-covered. “The vast expanse of level grazing country, and the broad marshes and sphagnous areas so common to Sta. Isabel and El Valle de las Pappas were conspicuously absent. “Naturally, country of this character is not very well adapted to the support of an extensive fauna. Birds were extremely scarce, and strange to say, exceedingly wary. “On all sides, excepting a break toward the west, the Paramillo is sur- rounded by ridges, some reaching an elevation of 12,000 feet, the tops of which are covered with dense forest, so that it stands like a mountainous, brown ‘island amid the sea of green. The Paramo of Frontino could be seen about twenty or twenty-five miles to the southwest, in a separate ridge, not directly connected with the Paramillo.” Buiritica to the Rio Sucio.— “ After completing the work on the Paramillo, we returned to Buritica, which place was reached February 7, and on the 9th we began the trip to the Atrato drainage. “Crossing the Western Range was comparatively easy, as the trail is excellent; we reached the top four hours after leaving Buritica, the altitude Buiuetin A. M. N. H. Vor. XXXVI, Puate XXIV. ALTO Bonito, Rio Sucio (Tropical Zone; Colombian-Pacific Fauna.) Rio Cauca at Puerto VaALpivIA View looking north, near the southern limit of forest on the lower Cauca. The Central Andes arise from the right hand bank, the Western Andes from the left. At no other place do these ranges so closely approach each other. (Tropical Zone; Cauca-Magdalena Fauna.) 1917,] Chapman, Distribution of Bird-life in Colombia. 67 of the pass being 8000 feet. Then the trail winds steadily downward, adhering closely to the sides of the Rio Cafiasgordas (Rio Sucio). We spent the first night at the town of Cafiasgordas, altitude 5000 feet, a settlement of Perhaps 5000 people, and next day reached a house called Orobajo, altitude 3225 feet. The river which begins as a small spring just this side of the summit of the range, here attains a width of two hundred feet and is a Taging torrent, the muddy water leaping and dashing over a bed strewn with huge rounded boulders of granite. Along the banks grow strips of dense bamboo, wild cane and brush interspersed with small patches of culti- vated ground; these strips of verdure do not exceed a few hundred yards in width, and beyond that the country is bare or covered with short, thin grass. This was a surprise to me as I had expected to find the whole west slope heavily forested. “We reached Dabeiba late on the 11th; as we reached the summit of the last little knoll, a beautiful panorama was spread before us; a perfectly level valley several miles long and a mile wide, covered with light green vegetation lay at our feet; in the center stood a cluster of forty or fifty white huts — the town of Dabeiba. Here and there a white area contrast- ing strongly with the green, marked the location of a cotton field; and through the center of the valley flowed the Sucio, now swollen to a rapid, muddy stream a few hundred feet wide. The sides of this valley are hemmed in by successive bluffs of sandstone, rising one above the other and at some distance apart, and I could never quite convince myself that this region had not once been covered by a good-sized lake. “‘Dabeiba marks the beginning of the coastal forest zone; the change from open country to forest is not gradual, but sharply marked. After col- lecting in this locality three days we accepted the invitation of an acquaint- ance and moved to his house, called Alto Bonito, ten miles down the river from Dabeiba. At the latter place the altitude is 2000 ft.,! and Alto Bonito is 1500 ft. Primeval forest covers all the surrounding country and the abundance of bromelias, ferns, and parasites indicates an abundant rainfall, although there is little underbrush. “Right days were spent at Alto Bonito, and a great many specimens secured that were new to us; but a large percentage were identical with those collected at Puerto Valdivia. “The work at Alto Bonito provided the last link in the chain of facts regarding the forested areas of northwestern Colombia, together with facts concerning the extension of the mountain ranges. ““The Western Cordillera terminates in the Cerro Aguila, just below 1] am convinced that our aneroid registers at least 500 ft. too high; but I have given its readings throughout, except at Puerto Valdivia and at La Playa (Barranquilla). 68 Bulletin American Museum of Natural History. [Vol. XXXVI, 9°, on the Golfo de Urab4, altitude less than 1000 feet. Gradually, north of the Paramillo, the range becomes lower and lower; in lat. 73°, the highest peak, called Alto Esmeralda, does not exceed 4000 feet; and the Abibe, a few roiles further north, attains an altitude of only 3600 feet. Beyond this the range is described as being mere hills. “A new road has just been completed from Turbo on the Gulf of Uruba (Darien) to Monteria, on the Rio Sint. This cuts right across the country about which we know the least, and I was interested to learn two facts:— the highest point in the road is 800 feet, and, every mile of the distance was cut through heavy, primeval forest. This, it would seem, provides for an easy passage for Atrato forms to the Cauca, at least to such forms (forest) which can ascend up to 800 feet; and this the collections from Puerto Val- divia and Alto Bonito should prove. However, I believe that forms which adhere strictly to the coast country, that is to the forest growing in the perpetual rain-belt, would be barred to a great extent from entering the Cauca, for, from all I can learn, the Cerro Aguila extends to the very coast notwithstanding that it is not so marked on our map.! “* After having entered the forests of the Cauca, there should be no ob- stacle to prevent birds entering the Magdalena forests, via the mouth of the Cauca and the San Jorge which furnish a natural connecting link.”’ AUXILIARY COLLECTIONS. Satisfactory determination of the specimens secured by our expedition in Colombia has been greatly facilitated by comparison with collections already existing in the Museum, or with others which have been recently made primarily for use in this connection. Under the first head should be mentioned the Lawrence Collection with its numerous types, and, particu- larly, the H. H. Smith Santa Marta collection. This contains some 3000 specimens of about 300 species collected chiefly in the Tropical and Sub- tropical Zones of the Santa Marta mountains. Under the second head, a collection made in 1912-1913, in Ecuador by W. B. Richardson, is deserving of first place. This contains some 4000 specimens including much topotypical material. Second place is accorded to a collection of some 1800 specimens made in eastern Panama by Rich- ardson, H. E. Anthony and D.S. Ball. These collections have been of the utmost value in identification and in distributional problems. 1This information was given to me by Sr. Cspinos, Director of the School of Mines in Medel- lin; and also by Ernesto White who was building the road I wrote you about.— L. E. M. 1917.J° Chapman, Distribution of Bird-life in Colombia. 69 When our own collections lacked the specimens needed to answer the question at issue, an appeal to allied institutions or to collectors has rarely failed to produce them. ‘The Batty Collection.— In 1898, the American Museum purchased from the late J. H. Batty 290 bird skins which had been collected by him in the region about Cali. The specimens are fully labeled, but I have been able definitely to locate only a few of the places at which they were taken, and with some exceptions they are therefore not listed in this paper. It should be added that the collection contains only one or two species not obtained by our Museum expeditions. It was through this collection and informa- tion secured in many long talks with Mr. Batty that I was first attracted to the Cauca Valley. The Mrs. Kerr Collections — In 1908 the American Museum purchased from an American woman, Mrs. Elizabeth L. Kerr, one hundred and ninety- four bird skins which she had collected in Colombia west of Honda, in the Magdalena Valley, and on the eastern slopes of the Central Andes up to an altitude of 3000 feet. Later Mrs. Kerr was commissioned to collect specimens in the Atrato Valley, and the two hundred skins secured by her and listed under the locali- ties she visited are the only ones we have secured from this region, except those taken by Miller and Boyle at Dabeiba and Alto Bonito. The Hermano Apolinar Maria Collections— Through the cordial rela- tions established by our Expedition No. 7 with Hermano Apolinar Maria, Director of the Instituto de la Salle of Bogotd, we have since received from him a number of small lots of birds from the Bogot4. Some have been col- lected at our request, some have been sent for identification, some in ex- change, others as a donation. They have included a large proportion of rare and interesting species. Brother Apolinar’s codperation has been especially valuable in securing specimens of species but poorly or not at all represented in the collections made by our Expedition No. 7. The Gonzales Collections — Manuel Gonzales, a native of Bogota, while employed as a general helper by our Expedition No. 7, was taught to prepare bird skins. A collecting outfit was left with him and he subsequently sent us some six hundred specimens, chiefly from the region about: Bogota, but including also seventy-nine from Barrigon at the head of navigation on the Meta. The localities visited by Gonzales, with the number of speci- mens collected at each are listed in the Gazeteer. 70 Bulletin American Museum of Natural History. [Vol. XXXVI, An OUTLINE OF COLOMBIAN TOPOGRAPHY. Detailed descriptions of Colombian topography, so far as we are familiar with it, are given im the itineraries of our various expeditions and in the characterization of the zones and faunas proposed. The physical, and zodgeography of the country is also presented in the accompanying maps. But at the risk of some repetition, it is proposed to present here an outline of Colombian physiography emphasizing those features which are of special importance in the study of the distribution of its life. It should be observed that in Ecuador the Andean system is, faunally, composed of but one range with, consequently, but one Pacific and one Atlantic slope, and one Temperate Zone bordered on each side by the Pa- ramo Zone islands of the higher peaks. But shortly after crossing the Colombian boundary this great range branches into three clearly defined ranges, each one of which is separated from its neighbor by a valley which descends to the Tropical Zone. The Magdalena Valley, lying between the Eastern and Centra] Andes, is never less than thirty miles in width and, in its lower part, much wider. The Cauca Valley, lying between the Central and Western Andes, from somewhat north of Popayan to north of Cartago has a width of twenty to thirty miles, but in Antioquia it is contracted to the width of the Cauca River from the eastern and western banks of which the Central and Western Andes respectively arise. Except in this region, therefore, the three ranges of the Andes in Colom- bia nowhere approach one another, and at no place do their upper life- zones — Subtropical, Temperate, and Paramo— come into contact with the corresponding zone of the neighboring range. It should further be noted that all three ranges terminate in the Tropical Zone, the Western and Central in Colombia, the Eastern in Venezuela. . It follows, therefore, that their three upper zones end, as it were, in the air; that is, at their most northern part, they have no such connection, through gradual descent of zonal boundaries with increasing south latitude, as, for example, exists between the Temperate Zone in Colombia at from 9500 to 12,000 feet, and the Temperate Zone in Argentina at sea-level. The Western Andes have no peaks reaching to snow-line, and we know of only four points at which they enter the Temperate Zone, one of which, the Paramillo, is near the northern end of the range. There appear to be no passes below 4900 feet (Cresta de Gallo, 4924 ft.), the average elevation is approximately 7000 feet, and the summit of the range is therefore usually in the Subtropical Zone. The Central Andes have a number of snow-crowned peaks; Paramo 1917.] : Chapman, Distribution of Bird-life in Colombia. 71 Zone islands are not infrequent, I know of no pass below 10,000 feet, the average height of the range may be said to be about 11,000 feet, and its summit is therefore largely in the Temperate Zone. The Eastern Andes also possesses several snow peaks and numerous Paramo Zone islands. So far as we have learned, at only one point, until one approaches the northern extremity in Colombia, do they fall below the Temperate Zone, the pass at Andalucia between the upper Magdalena Valley and the Caqueté region having an altitude, as determined by Miller, of only , 7000 feet. In addition to these main branches of the Andean system, all of which are connected at their base north of the Ecuadorian boundary, Colombia possesses three other mountainous areas; the Baudo-Panama, what may be called the Amazonian, and the Santa Martan. The Baudo mountains, lying west of the upper Atrato form the true Pacific Coast Range. They are said to attain an altitude of 5500 feet, making’ their summit subtropical. In discussing the northward extension of the Subtropical Zone into Central America, evidence is presented which is believed to indicate that this range once possessed a greater altitude connecting it with the mountains of the Panama boundary at the north, and Western Andes at the south, at which time it formed a fourth Colombian branch of the Andean system on which the Subtropical Zone was carried into eastern Panama. Little is known about the mountains lying east of the Eastern Andes on upper Amazonian drainage (as before remarked), but I can find no evidence of their having an altitude of over 3000 feet, and if this be true, they do not reach above the Tropical Zone. Hamilton Rice! writes that the Sierra Chiribiquete “may be a counterfort thrown out from the Suma Paz, and is a chain of crag-like peaks and hog-backs rising to an altitude of over 2800 feet.” He doubts the existence of the Tunahi or Padavida range, shown by Codazzi. The zodlogical evidence supports the geological belief that the Santa Marta mountains are of independent formation and have had no connection with the Andes. As such, the life of this group above the Tropical Zone, is insular and the study of the geographical origin of its forms is a clearly circumscribed problem, supplemental to that presented by the life of the main Andean chain. Aside from these smaller mountain groups, it is obvious that the exten- sion, almost the entire length of Colombia, of three distinct, high mountain ranges, effectively cuts up the Tropical Zone through which they pass into several sections each of which is more or less segregated from the other. 1 Geog. Journ., 1914, p. 144. 72 Bulletin American Museum of Natural History. [Vol. XXXVI, Thus, the great region lying east of the Eastern Andes, the western ex- tension of the Amazonian and Orinocan basins, is separated by this range from the Magdalena Valley. This valley, in turn, is walled about by the Eastern and Central Andes and, so far as tropical life is concerned, is accessible only at its northern end. The Cauca Valley is similarly isolated and is open only at the north, while the Pacific coast region is shut off at the south by the deserts of south- ern Ecuador and Peru, and at the east by the entire Andean system. Like the Magdalena and Cauca Valleys it, too, is apparently to be entered only at the north. Tropical Colombia, therefore, may broadly be spoken of as consisting of the western portion of the Amazon-Orinoco Basins and three cul de sacs which debouch on its northern coast. To what extent existing faunal conditions are dependent on existing topography, and to what extent they have been brought about by what may be called pre-Andean topography, remains for us to determine. REMARKS ON THE DISTRIBUTION OF FORESTS. The detailed information which we gathered in relation to the distribu- tion of forests in Colombia will be found in the descriptions of the routes traversed by our expeditions, as well as in the Gazeteer. We covered, however, so comparatively little ground, and conditions change so abruptly, that data are lacking for anything but the most generalized statements in regard to the extent of the forest areas of Colombia as a whole. The pres- ence or absence of forests, however, has so important a bearing on the boundaries of faunal areas, that from the zodlogist’s point of view, even generalized statements are of value. The Forests of the Tropical Zone.— The Tropical Zone possesses five large areas of heavy, humid forest. Named in order of their importance they are (1) the Amazonian; (2), the Pacific coast; (8), the lower Cauca-Mag- dalena; (4) the Maracaibo Basin, only the western portion of which enters Colombia, and (5) the Santa Martan. The Amazonian forest region occupies all that part of Colombia lying east of the Andes and south of the Rio Guaviare. Its northern boundary, therefore, lies‘just north of the divide between Amazonian and Orinocan drainage. It forms, in fact, the northwestern corner of the vast forests of upper Amazonia which, southward, reach to Bolivia. Miller (Expedition No. 5), writing from an elevated position near Florencia says “one has a good view of the Caqueté country,.a_perfect ocean - of forest stretching out ahead as far as the eye can see, which, on clear days, Burt. A. M. N. H. Vor. XXXVI, Prate XXV Ss 2 A N ® |" ae ‘ 1 B GULF OF RK Santa Marta , r fi Lea i A 5 Ls, C pes 3 § Carthagen: © By hes <4 mn ¢ enaventur. Bona Island? Gor; Arauen| ou a i ( = ( Mountain Forest Tropical. ZONE Forest DISTRIBUTION OF (-~] Unrorestep AREAS LT} Lranos FORESTS IN COLOMBIA 1917.] Chapman, Distribution of Bird-life in Colombia. 73 is many miles. The sight is most impressive. There is not a single rise visible and the forest is of uniform height.” Dr. Hamilton Rice writing from a point west of the Sierra Chiribiquete (Lat. 1° 10’ 16”; Long. 72° 12’ 34” and therefore slightly south of east from Miller’s station) says:' “From this high land the Chiribiquete was seen to extend southeast as far as the eye reached, the rest of the country being undulating and forest-covered. ...Occasionally the forest gave place to a dense growth of bush (rastrojo), a wild tangled thicket, difficult to get through, on a bottom of black, boggy mud, and especially hard on the carriers.” The same writer (i. c., p. 144) also refers to the densely forested plains of the Inirida and Uaupes, while his description of the transition from the Llanos north of the Guaviare to the forested region south of it (I. c. p. 145) I quote in full: “Tn passing from north to south across such a stretch of country as that between San Martin and the Caqueté district, one may note differences of Amazonian vegetation characteristic of each different level of land. First there are the grassy savannas or campos with their knolls, glades, thickets, and scattered scrub; then the vegetation of the lowlands or rebalsa edging the rivers and inundated in the wet or winter season; next the forests of the low plains or monte bajo, which when seen from above appear more evenly topped and lighter than the woods on land above the highwater mark (monte alto), which are known as Virgin or Primeval forest. On closer inspection the trees of the low plain are seen to be lower and more scattered than those of the high land, without any great abundance of palms or lianas, but with a profusion of ferns. In the Virgin forest the trees are densely packed and high, from which emerge solitary individuals still more lofty, overtopping even the highest palms, and from whose massive masts are spread diverse forms of crowns and summits, dome, pyramidal, and cande- labra, the whole interwoven by an intricate meshwork of lianas and vines. The vegetation of the rebalsa near the river bank is often low and bushlike, but gradually increases in height, the further it is from the bank, until, at the point to which the highest floods reach, it almost rivals the trees of the monte alto in height.” The Pacific coast forest extends from northwestern Ecuador northward to eastern Panama, and from the shore-line eastward to the forests of the Subtropical Zone, or to the summits of outlying ridges. Under the condi- tion first-named the entire Pacific slope of the Western Andes from sea to summit is covered with unbroken forest, such, for example, as Allen and Miller (Expedition No. 3) found on their section from Cartago to Névita. 1 Geog. Journ. August, 1914, p. 150. 74, Bulletin American Museum of Natural History. [Vol. XXXVI, Under the second-named condition, the continuity of the forest is broken by such an arid pocket as we found at Caldas, or Miller and Boyle encount-- ered above Dabeiba. At the northeast, on the right side of the lower Atrato, the Pacific coast forest apparently connects with that of the Cauca-Magdalena, and this is the only connection of which we definitely know between any of the forested areas of the Tropical Zone. Mr. Douglas Fyfe, an American engineer, sit- uated in northern Colombia, writes me of the country at the border of the Colombian-Pacific forests and the Caribbean Savannas: “The Sint is‘a wide, sluggish stream meandering through a broad plain of very deep allu- vium. Along its banks are situated beautiful grass-covered cattle ranges. The river is carrying a vast amount of sediment and gradually setting up a large area at its mouth forming deltas and reed swamps. Numerous water- fowl seem to inhabit these swamps the year around. The country lying east of the Sind for about twenty miles is low-lying and under water part of the year; cienagas in fact. Beyond this lie the savannas. The country west of the Sind is entirely forested to the Pacific coast, the forest beginning at the Atlantic sea shore and extending inland without a break to the interior.” The differences between the forest of the lowland and that of the foot- hills are well described by Allen in the narrative of his journey with Miller just alluded to. In connection with the change in gradient, character of the soil, etc., we have here two widely varying types of environment which doubtless account for the fact that some species appear to be confined to the bottomlands. The Cauca-Magdalena forest is contained chiefly in the Department of Antioquia. Miller’s exploration and inquiries show that it extends from the northeastern border of the Pacific coast forest northward to the Carib- bean coast and thence eastward to the shore of the Rio Sinu, when it meets the western border of the marshes which pass into the Caribbean savannas. Thence it extends southward up the lower Cauca, and doubtless also the Neché, to about Lat. 7°, and up the Magdalena Valley to La Dorada on the Magdalena River, where, on the valley floor, it terminates abruptly and is succeeded by the Savannas of the upper Magdalena. Along the foothills of both the Eastern'and Western Andes, the forest extends south of La Dorada. Northward, in the Magdalena Valley, it is bounded on the west by the San. Jorge and on the east by the foothills of the Eastern Andes, while its northern limit lies near Banco. Eastward of this point, it may occupy the foothills of the Sierra de Motillones and Sierra del Valle de Upar, at the northern end of the Eastern Andes, but I have no definite information in regard to this region. Whether the Cauca-Magdalena forest is connected with that of the Maracaibo Basin, remains therefore undetermined. 1917.] Chapman, Distribution of Bird-life in Colombia. 75 The presence of heavy tropical forest in the southern part of the Mara- caibo basin has been made known to me by W. H. Osgood and Ned Dearborn, both of whom have visited this region in the interests of the Field Museum. As above remarked, J do not know whether this forest is connected with that of the lower Magdalena Valley by a belt of foothill forest, such, for example, as we found above Villavicencio; but at the east it appears to be bordered by the arid coastal region and Venezuelan Andes. It seems, therefore, to be an isolated area; nevertheless, it lies at the door through which many species have entered the Cauca-Magdalena fauna. Its con- nections to the eastward are of much importance, but I have no information concerning thein. Of the Santa Marta forest, M. A. Carriker, Jr., our authority on that region, writes me that on the Carribbean or northern side, from about Cabo de San Juan de Guia, to a point known as Camarones, the forest of the Sierra Nevada extends right to the coast. He adds: ‘Whether there is any forest connection between the Sierra Nevada and the Cerros Negros (Eastern Andes) on the watershed between the Rio Hacha and Rio Cesar, I do not know; most likely there is some such connection, although I am not of the opinion that it is heavy forest. This watershed is very low.” The forest west of the Sierra Nevada, Carriker writes, “extends west- ward to the shores of the Cienaga Grande. ...The region between Cienaga Grande and the Magdalena is swampy and contains many sluggish water- ways and is inundated frequently... .” Unforested Tropical Areas.— That part of the Tropical Zone in Colombia not covered by heavy forest growth may be grouped under four heads: (1), the Llanos; (2), the Caribbean; (3), the upper Magdalena and (4) the upper Cauca or Cauca Valley proper. Aside from these major divisions there are semi-arid pockets like the upper Dagua on the western slope of the Western Andes, bare foothills such as exist above Dabeiba, or brush- grown valleys like that of the lower Rio Negro, but these and similar local variations do not affect the general truth of our classification. The Llanos occupy that part of Colombia lying east of the Andes and north of the Rio Guaviare. Of them Rice (J. ¢., p. 139) writes: “The Llanos, extending from the Cordillera to the Orinoco river, and from the Arauca to the Guaviare, are covered with denise, tall grasses, from which here and. there rise groups of palms and bushes and belts of trees. They are well watered by innumerable streams, varying in volume and size from immense rivers to rivulets, which in winter season overflow the lower lands.” Of the Llanos of San Martin the same author writes: “To the east they are broken by swales, swamps, and eminences of mesa and scarp formation which push the Humadea river northeast and deflect the Ariari southeast. The high Llanos, at the foot of the forested hills are usually rocky, and semi- ie 76 Bulletin American Museum of Natural History. [Vol. XXXVI, arid, covered with wild dense jungle growth,-and so deeply ravined as to-be impassable. These merge into the low Llanos, over which are scattered Savannas, immense meadows of fine succulent grasses, dotted with high bushes, clusters of palms, and thickets of other trees. During the wet or winter season, much of the land is subject to inundation, and consequent fertilization, by the swollen rivers.” Hettner,! writing of the country lying at the eastern and western bases of the Eastern Andes in the Bogota region, describes the Llanos and accounts for their lack of forests as follows: “The two lowlands which take in the western and eastern foot of the Cordilleras, bear altogether different plants; the lowland of the Magdalena River below Honda is covered with a thick, primeval forest; the eastern low-lands, the so-called Llanos, are on the other hand, wide grassy plains, which are only interrupted by a line of forests on the banks of rivers. This difference of plant growth has a relation with the rainfall, for the lowland of the middle part of the Magdalena River has two rainy and two dry seasons, which are, however, of short duration; in the Llanos, on the other hand, the one dry season shrinks together to a few weeks, while that of the other is extended to from five to six months. With so long a dry season near the equator, the forest growth is out of the question. If we should travel southward in the Llanos and reach the territory of the equatorial rains, we should find here likewise a luxuriant forest, and vice versa, at the lower part of the Magdalena River, somewhat north of 8 degrees, where the tropical rains begin, the forest is gradually crowded out by grassy plains, interspersed with single trees, or in other words, by savannas.” The Caribbean forestless region corresponds to the Caribbean faunal area. It is a semi-arid or arid region in which the absence of forest-growth is presumably due to insufficient or irregular rainfall with long periods of drought. The coast region itself, from the Rio Sint to the Goajira Peninsula, except for the mangrove-bordered lagoons, and the section of forest-covered shore north of the Sierra Nevada of Santa Marta, may be described as arid. Cacti, thorn-bearing bushes and other xerophilos forms comprise its char- acteristic vegetation. (Farther inland we pass to the semi-arid savannas, a grazing country with scattered mimosas and acacias and frequently marshes. Carriker writes: “The semi-arid coast belt begins a few miles south of Cienaga (town on the Santa Marta Railway) and extends around the coast 1‘Die Kordillere von Bogota,’ Petermann’s Mittheilungen, Erganzungsband, 22, p. 76. I am indebted to Dr. Chester A. Reeds for this and following abstracts from Hettner’s valuable paper. 1917.] Chapman, Distribution of Bird-life in Colombia. 77 to somewhere about Cabo de San Juan de Guia, where the forest of the Sierra Nevada (as above quoted) extends right down to the sea all along the coast to a point known as Camarones, from which point to Rio Hacha is a region of scrub and cactus, along the coast, but opening out into savanna and scattered woodland toward the south, and continuing on around the Sierra Nevada to the region of the Valle de Upar and Rio Cesar Valley... The whole of the Goajira Peninsula, from Rio Hacha east, is an arid region of cacti and thorny scrub.” , Hettner’s reference to the savannas of the lower Magdalena, “some- what north of 8 degrees or just north of the region of tropical rains” has been given in connection with his description of the Llanos quoted on a preceding page. The two remaining unforested areas, the upper Magdalena and upper Cauca Valleys, are described both in our itineraries and in connection with the characterization of the Cauca-Magdalena fauna of which they form arid sections. — Hettner (I. c., p. 79) referring to the upper Magdalena region writes: “ Another interruption of the forest was probably formed by the strata of sand and gravel, whose porous soil, destitute of nutriments, is not suffi- * cient for the forest; the mesa of Fusugasugé, for instance, probably always consisted of thorny underbrush and grass. Similar, although still more adapted to the drought, because it lies in a warmer climate, are the strata of sand and gravel of Medina or the tuff plains on the Magdalena River above Honda.” The same author (I. c., p. 80) describes the vegetation of this and similar semi-arid localities in this part of Colombia as follows: “Tn many places, the bushy vegetation shows plainly its adaptation to drought, and this is generally the case in the vicinity of the lighter mimosa forest, which probably took in the place of this brush originally. It is small-leaved and thorny and generally shows the acacia form; the agaves with their sword-shaped, sharp-edged, fleshy leaves, and the most varied forms of cacti, whose juicy trunks give them the ability to overcome drought, are numerous and often form impenetrable thickets, while at different spots, they stand singly, and leave the bare earth between them. In such places the land almost bears the appearance of a desert, and only the donkey finds a meagre meal. This vegetation, adapted to the drought, occurs in almost all the altitudes, on the banks of the Magdalena River above Honda, as well as in the valley of the Rio Bogoté between Tocaima and Jiradot, or in the vicinity of Cicuta, and at Neiva, or at Soacha in the southern part of the plateau of Bogoté, but it is, however, more extensive and more devel- oped in the lower, warmer parts than in the cooler altitudes.” wa 78 Bulletin American Museum of Natural History. [Vol. XXXVI, The Cauca semi-arid region, begins slightly south of the head of navi- gation in the Cauca river, in Antioquia and extends up that much con- stricted region lying between the Western and Central Andes, through which the Cauca flows, to the Cauca Valley. The topographic isolation of this valley is therefore faunally increased by the aridity of the region which lies between it and the humid lower Cauca-Magdalena region. The Cauca Valley, as elsewhere remarked, is far from arid. Neverthe- less, its forests are largely restricted to the banks of streams and low-lying areas where a natural subsurface irrigation gives them the water the insuffi- cient rainfall does not directly provide. It must, however, be remembered that the Cauca region has been settled for many years, and that deforestation to create grazing for cattle has long been ‘practiced. It is probable, therefore, that the forested area was formerly more extensive than it is at present. Mountain Forests— Under the general head of Mountain Forests we may include the forests of the Subtropical and Temperate Zones. Both may be classed as cloud forests, the lower limit of the first being determined by the altitude of condensation, the upper limit of the second by that of the temperature at which tree-growth ends. Depending for their continuity on topography, or the relation between slope exposure and prevailing air currents, on the relative altitude of con- tiguous ridges and other comparatively local conditions, a detailed survey would be required to make anything approaching an accurate map, showing the distribution of mountain forests in Colombia. It may, however, be said in a general way that Subtropical forests, or those lying between approx- imately 5000 and 9500 feet, because of their lower altitude, are more or less continuous on one slope or both, of all three ranges of the Colombian Andes. Temperate Zone forest, both because of the higher altitude required, and also because of the lower rainfall which usually prevails on this zone, is less continuous than that of the Subtropical Zone. Unforested Mountain Areas.— As indicated by the statements just made, the forestless tracts above the tropics are found largely in the Temperate Zone, where on the Bogoté Savanna, or in that region lying south of Popa- yan, for example, there are wide areas which are treeless. What is believed to be an approximation to existing facts is expressed, semi-diagrammatically, in the accompanying map in which the yellow color employed represents not only the arid coast region, the savannas, and plains of the Tropical Zone, but also treeless, grass-covered slopes and paramos of the higher summits. 1917.] Chapman, Distribution of Bird-life in Colombia. 79 NotEs ON COLOMBIAN CLIMATOLOGY. Temperature:— The temperature of any given locality in Colombia shows so little fluctuation throughout the year that the seasons are marked, not by increase of cold or heat, but by rainfall. To demonstrate the narrow and regular path travelled yearly by the thermometer in Colombia, and as a contribution to the unfortunately limited amount of published meteoro- logical data from that country, I append a summarized record of the tem- perature for the year 1907 at the sugar estate of La Manuelita in the Cauca Valley near Palmira. These records were made by Mr. Chas. J. Eder and are selected as an average from a series covering a period of ten years. It will be observed that there is only a difference of six degrees in the average weekly temperature for the entire year. Temperature for the year 1907 at La Manuelita, Col. Week ending Highest Lowest Average for week. Jan. 4 85 66 744 11 86 63 743. 18 86 63 75 25 84 65 73 Feb. 1 86 64 7A 8 86 67 73 15 86 64 73 22 85 64 73 March 1 86 62 72 8 86 64 75¥ 15 86 64 75 22 88 66 76 29 85 64 743 April 5 86 64 755 12 87 66 75 19 86 66 76 26 86 67 763 May 3 86 65 75 10 85 67 75 17 85 64 734 24 86 66 73 31 83 63 74 June 7 85 62 74 14 86 64 75s 21 86 66 75 80 Bulletin American Museum of Natural History. [Vol. XXXVI, Week ending Highest Lowest Average for week. July 5 86 63 742 12 86 64 74h 19 85 65 74 26 89 59 743 August 2 87 65 76 9 91? 66 veh 16 90? 61: 76 23 90? 66 78 30 80 64 773 Sept. 6 89 65 76 13 91 64 764 20 89 64 763 27 90 65 763 Oct. 4 88 67 76 11 86 64 763 18 89 65 77 25 88 64 753 Nov. 1 87 66 75 8 87 66 76 15 87 68 774 22 86 66 75 29 86 66 76 Dec. 6 86 65 76 13 85 65 75 20 87 64 76 27 90 66 774 The preceding observations were made in the Tropical Zone, but an equal stability in the yearly range of temperature is shown at localities in the Subtropical as well as Temperate Zones. Thus, Regel states that the lowest monthly average recorded at Bogota is 57° (July, 1880), while the highest, 61°, occurred in the same month of the succeeding year. This surprisingly small annual range in the temperature of any one locality doubtless accounts for the comparatively small variation in the limits of life-zones, the boundaries of which are primarily isothermal. It is, therefore, of importance for us to know to what extent temperature is affected by altitude and I insert here the following pertinent observations from Robert Blake White’s ‘Notes on the Central Provinces of Colombia’ (Proc. R. G. S., V, 1883, pp. 263, 264). “From a numerous series of observations of the mean temperatures at different altitudes in the cordilleras, collected from a great many observers, I have formed a table of mean temperatures corresponding to a series of altitudes from sea-level up to 16,400 feet in height, which will be found very generally applicable over the whole of the Colombian territory. These 1917.] Chapman, Distribution of Bird-life in Colombia. 81 mean temperatures are derived from observations made on distinct systems, but as a rule the temperature of the earth, in a part sheltered from the sun and rain, at a depth of 30 inches from the surface of the ground, will repre- sent in these latitudes the mean temperature of the locality. In tropical regions, where vegetation is not exposed to great variations of temperature, the most important point to which the agriculturist should look is the mean temperature, if he would judge correctly of the climate of any locality. “Table of Mean Temperatures in the U. 8. of Colombia, between 2° and 6° N. Lat., compiled from observations by Humboldt, Caldas, Boussingalt, Mosquera, Reiss, Stubel, and White. Height above Mean Height above Mean Sea-level. Temperature. Sea-level. Temperature. Feet. Fahr. Feet. Fahr. 0 324° 9,020 55.4° 820 80.4 9,840 53.6 1,640 78.4 10,660 50.9 2,460 76.3 11,480 48.2 3,280 74.3 12,300 45.5 4,100 71.2 13,120 42.8 4,920 68.0 13,940 40.1 5,740 65.3 14,760 37.4 6,560 62.6 15,580 32.0 7,380 59.9 16,400 30.2 8,200 57.2 “The mean in the greater altitudes varies somewhat according to the greater or less extent of snow-covered mountains, and in the lesser altitudes the temperature is affected by the open or inclosed character of the valleys and by the presence or absence of vegetation. Generally, however, it will be found that the above means are sufficiently near the truth to be of practical utility.” Rainfall: — As before remarked, the seasons in Colombia are charac- terized by dry and wet periods rather than by variations in temperature. It is, therefore, to be regretted that data in regard to rainfall are even less satisfactory than those relating to temperature. Furthermore, variations in rainfall, both at the same locality and at neighboring stations, are so great that the statistics available merely demonstrate the need for additional observations before even the most conservative generalizations can be made on this branch of Colombian meteorology. Two types of the seasonal distribution of rain are commonly recognized in Colombia. In one, a wet season of six months duration is followed by a dry season of equal length. In the other, wet seasons each of three months duration are separated by dry seasons of equal length. 82 Bulletin American Museum of Natural History. . (Vol. XXXVI, Under the first-named condition, rain usually falls from May or June to November or December, and the season is termed “invierno”’ or winter; while the months from November or December to May or June are dry and the season is known as “verano” or summer. North of latitude 8° the seasons are characterized by one dry and one rainy period; south of this latitude two rainy and two dry seasons are the rule. The comparatively arid Caribbean Fauna possesses therefore but one rainy season, while the humid Cauca-Magdalena Fauna has two, annually. There is, however, much irregularity both north and south of latitude 8°, while the amount of variation in annual precipitation at stations separated by only a few miles may exceed 300 inches! For example, at San José, thirty-seven kilometers from Buenaventura, the observers of the Pacific Railway recorded a deposit in 1912 of 400.88 inches, while during the same year Caldas, distant 45 kilometers from San Jose, received only 54.46 inches, a difference of 346.42 inches. The topographic conditions responsible for this remarkable variation will be found described under the description of the route followed by our expedition No. 1. Ocean currents, comparative temperatures of the air over land and sea, prevailing direction of the wind in relation to mountain slopes, relative height of ranges in the same chain of mountains are among the more im- portant local causes affecting rainfall in Colombia. The subject is inti- mately related to the distribution of life and particularly to the breaking up of zones into faunal areas, but, as already remarked, few exact data exist and I therefore merely present those I have obtained through the courtesy of others, without further comment. Record of Rainfall at Pato Mines on the Rio Neché, Antioquia, from August, 1918, to July, 1914. 1913. August September October November December 1914. January February March April May June July Pee erp e OROCO HO Nonon NPR WOORMMON WANWA | 143.6 1917.] Chapman, Distribution of Bird-life in Colombia. Record of Rainfall at certain Stations on the Pacific R. R. Contributed by R. Alvarez Salas. Stations. Buenaventura San José Km. Caldas Km. Palmar Km. Lomitas Km. Yumbo Km. 1910. 1911. 1912. 1913. 83 323,96 248,66 265,10 234,90 262,86 37 270,00 277,37 400,88 296,10 82 57,08 46,66 54,56 50,11 31,09 98 34,06 25,41 109 33,34 158 37,73 ——- Annual Rainfall at La Manuelita, Cauca Valley, 1900-1910. Year. 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 Amount. 37.97 45.21 33.80 56.38 87.74 83.79 39.96 47.80 54.94 55.13 48 .50 Days on which Rain Fell. 123 136 148 131 141 113 126 144 177 172 160 Average Monthly Rainfall at La Manuelita, Cauca Valley, 1900-1910. Month. January February March April May June July August September October November December Amount. 3.19 3.10 4.16 6.09 5.37 2.92 1.50 1.49 2.98 5.81 4.78 3.20 + 44.59 84 Bulletin American Museum of Natural History. [Vol. XXXVI, THE LIFE ZONES OF THE COLOMBIAN ANDES. The ascent of a lofty mountain on which, faunally, several hundred feet of altitude may represent several hundred miles of latitude, is a profoundly impressive and fascinating experience for, the zodgeographer. One is amazed by the distinctness of the life-zones encountered and is led to speculate on the origin of their strongly characterized floras and faunas. That the alti- tudinal distribution of plants should conform closely to belts or zones, the limits of which are determined primarily by temperature, is not surprising; but that such mobile creatures as birds should be confined within certain more or less definite boundaries by these invisible barriers is a convincing evidence of their potency as well as of the sensitiveness of the organisms on which they act. It does not follow that every species will occur in only one zone, it appar- ently being a general law that wide latitudinal or faunal distribution implies also great altitudinal or zonal range. Examples may be found among birds as widely apart in relationships and habits as the Turkey Vulture and House Wren. Both range at sea-level from the South Temperate to the North Temperate Zone; both are found from the tropical lowlands to the temper- ate uplands of the Colombian Andes. These birds, and a few others like them, are, however, marked exceptions and while a species may advance a small distance beyond its true zone, a surprisingly large number of species are found in only one zone. The zones themselves are not, of course, more sharply defined than the ranges of the species which characterize them. No one can stand at the foot of a snow-crowned mountain in the tropics without at once realizing that temperature, as it is influenced by altitude, is obviously the dominant factor in producing the floras and faunas en- countered between base and summit. Where humidity, and in certain instances, character of the soil, add their influence, the boundary lines between life zones are often very sharply defined. One may pass, for example, from the upper border of the arid tropics on the eastern slope of the Western Andes at San Antonio into the dense forests of the humid subtropics on their crest in less than two minutes, and experience a complete change in bird-life. But even where temperature alone is active, and there is no marked difference in rainfall, the forest being continuous, an alti- tudinal difference of 1000 feet may bring one into an essentially new avi- fauna. Such a phenomenon we observed when traveling from El Pifion (alt. 9600 ft.) to just above El Roble (alt. 8600 ft.), on the trail between Butt. A. M. N. H. Vor. XXNVI, PratE XXVI 6 A N fy x i 1 bi GULF OF h\ g ENEZUEL x i bine — Zany, Cc ase Waa Carthagen: A LAKE eee 4 eee Buenaventur: f to Gorgona Island $ phe oth eG 2 ee Ae Qu “ aN 2, iar SQ Z é 3 3 a a “Sl x ea : . \ : : Cee, ~ wy | } \| be Z X 5 a a 2 I ere eal 1 Y a eho, aa 10 78 7 \ EZ _ 74 7: 7 7 FAUNAS FAuNAS CoLomBian-PAcIFIC Sustrorrcat { [iggg:«“West Anvean \ [7] Cauca-Macpaena 2 ZONE \ Gag East Anpean TROPICAL 5 ee Zoxe | CO) Carippean TEMPERATE ZONE AND FAUNA [5 Ortnocan Paramo GHB Zone And Fauna (CZ Amazonian LIFE ZONES AND FAUNAS IN COLOMBIA 1 The dotted area is the arid portion of this Fauna. 1917.) Chapman, Distribution of Bird-life in Colombia. 85 Bogota and Fusugasugé. The first-named locality has a highly developed Temperate Zone life; while at the second, the fauna of the Subtropical Zone is equally well represented. A study of the bird-life of the Colombian Andes, shows, therefore, that it is distributed in four zones, and since the lower zone lies wholly within the tropics it follows that the remaining zones are all altitudinal. While I have been tempted to use names for them which seemed especially descrip- tive locally, it has been deemed far more desirable to accept existing terms which are generally applicable. These zones with their altitudinal bounda- ries are as follows: Tropical Zone sea-level to 4500-6000 ft. Subtropical Zone 4500-6000 ft. to 9000-9500 ft. Temperate Zone 9000-9500 ft. to 11,000-13,000 ft. Paramo Zone 11,000-13,000 ft. to snow-line (15000 ft.). These divisions correspond to the ‘Tierra Caliente,’ ‘Tierra Templada,’ ‘Tierra Fria,’ and ‘Paramo’ of other authors, but the altitudes here given are higher than those based on temperature alone. Basing the limits of his divisions upon an apparently purely arbitrary assignment of isotherms to zonal boundaries, Hettner! places the upper limits of the Tierra Caliente [= Tropical Zone] at 1000 metres; the Tierra Templada [= Subtropical Zone] between 1100 and 2000 metres; the Tierra Fria [= Temperate Zone] between 2100 and 3000 metres, and the Paramo between 3100 and 4000 metres. It will be observed that the limits of only the upper zone conform to those determined on the distribution of bird-life. It is reassuring, therefore, to find a much closer agreement between the zonal boundaries here given, based on the distribution of birds, and those based upon the distribution of vegetation presented by Wolf,? who, as a result of his studies of the flora of Ecuador, gives for both the Western and Eastern Andean slope four zones of life as follows: Tropical sea-level to 1600 metres. Subtropical 1600 to 3000 metres. Subandine [= Temperate] 3000 to 3400 metres. Andine or Paramo 3400 to 4600 metres. The limits of the two lower zones, for which Wolf employs names I had independently adopted, are essentially the same as those IJ here give for Colombia. The third, which Wolf calls the Subandine, but for which a continental-wide view of the subject suggests the name of Temperate Zone 1 Kordillere von Bogoté, p. 70. 2 Geographia y Geologia de Ecuador, p. 435 et. seq. “salar UL ayBog yeyaoztI0 FL “yoo UF aeog peorps A “moug fenjedieq “¢ ‘qu0Z oyelodulay, “¢ ‘auo0Z yesidonqng °% ‘quoz ouremeg “F -ouog yeaidory, “T ‘afeuleip ULoOULIQ 0} S809 OBIOVg VY} WOIJ sapuy UeIquIojo— oy} Jo sodues sory} 043 YsnoIg} e]yosd oNeuoreIserp-TwES VY ‘SopUY UBIGNIO[OD oy} Jo souoZ-9yrT “T “SW 7“SAoN sofuadlAeriA kaite ‘ eupjep sey 1 ' fy- N aTIeA by eysiaeuan|g a i caer shes cenen ine aeteo ae eae : Be See \ ear @ fexanbe> @ : 7 @ ac aaoe nbediyo\ SETAE S: siQQNV NUBLSIM en ¢ 2 anbeduyy S706 oa ypoLoy / ¢ el e ‘ ejaunBe ssid ssfed orpuin}, £ V leAesTeques : + SAQNV |INNALSVS ere 5 SAQ\INV WVaLNaS en B 0002 oe Oze Yoe Crd ELE 35% Cad Fee BOG Bor oa On 27) wer 70 96 49 ey ze cn : 86 1917, Chapman, Distribution of Bird-life in Colombia. 87 as more appropriate, is somewhat less extensive altitudinally than in Colom- bia, and the fourth, or Paramo, is correspondingly larger. Local conditions, including the much greater superficial area of the Paramo Zone in Ecuador, and the isolation of the temperate interandine region from Subtropical influences, are no doubt responsible for these differences. Wolf’s zones, as well as the comparatively simple topography of the Andean system in Ecuador, are shown in the accompanying diagram from his standard work (I. c., p. 441). The extent, general, and ornithological characters of the zones herein proposed are presented in detail beyond, but here I offer several general considerations in regard to Andean zonal life as a whole. Any attempt to explain existing conditions must be preceded by an effort to picture to ourselves the effect on the fauna of a tropical region of the uplift in it of a mountain system to snow-line. If at some point in the heart of the humid tropics, let us say upper Amazonia, progressive cooling 000 m oe BIE EEN Eee IS LANNE RENCE COS eee te L Oe SECRET : 3 . Fig. 2. Ideal section through the Ecuadorian Andes to show zones of vegetation. 2. Llanos. 3. Tropical and Subtropical Forests. 4. Interandean Region [= Temperale Zone]. 5. Andean Region [= Paramo Zone]. 6. Perpetual Snow. (From Wolf, 1892.) should eventually produce a snow-covered area surrounded by successive, concentric, climatic belts leading gradually to the surrounding tropics, we should have no more striking climatic change than has been brought about by the elevation of the Andes. Geologists, I believe, are agreed that this great mountain system is of Tertiary origin, and that there have been pronounced uplifts as late as the Pleistocene. Perhaps, therefore, we are warranted in assuming that the range had not acquired sufficient elevation to become an effective barrier ’ to the distribution of tropical life prior to the latter half of the Tertiary. However this may be, one’s imagination is stimulated by an attempt to follow the course of events as a gradually increasing elevation, with its subsequent changes, brought into existence new habitable areas of the earth’s surface with strikingly different climates from that of the base out of which they had arisen. 88 Bulletin American Museum of Natural History. [Vol. XXXVI, Where such an uplift created a mountain system as continuous as the Andes now are, these new areas were doubtless populated by latitudinal extension of range from regions having similar climates, and by altitudinal extension as the pressure of life from immediately contiguous regions below forced species upward, the more adaptable of which survived. Although having the shorter journey, the change of environment would be greater for those species coming from another zone in the same latitude than for those coming from the same zone in a perhaps distant latitude. We should, therefore, expect to find greater variation in what may be called zonal representatives than in altitudinal representatives. A study of the existing fauna supports these theories of the origin of zonal life and the degree of variation it presents. Thus the birds of the Subtropical Zone have been almost wholly derived from the zone below; those of the Temperate Zone came in part from the Subtropical, in part from the same zone at sea-level, while nearly all those of the Paramo Zone have come from the sea-level equivalent of this zone in southern South America. It follows, then, that the birds of every zone above the tropics have been derived from a lower level. There are some exceptions to this rule but they do not affect the general truth of the statement. In comparative varia- bility the fauna of the Subtropics differs more from the ancestral stock in the tropics than do the altitudinal forms of the Temperate and Paramo Zones from their distant sea-level derivatives of the South Temperate Zone, with which indeed they are often specifically identical. Hence it follows that uniformity of life increases with altitude, while as a corollary, the number of species decreases; uniformity of environment being apparently the underlying cause. The sometimes marked difference in the character of alluvial bottom- lands and slopes arising from them, even when both are wooded, exerts a strong influence on the range of some species of the Tropical Zone. Certain terrestrial birds, like Pittasoma, for example, are confined to the muddy shores of slow-flowing streams. Others, like Opisthocomus, do not leave the growth along the borders of such streams. Still others frequent the floor of the lowland forest. Such restrictions of range, however, appear to me to be of habitat rather than of zone, and do not, in my opinion, require’a subdivision of the Tropical Zone. We obtained no evidence of altitudinal migration among Colombian birds, though it is probable that Hummingbirds range up and down moun- tain sides in search of certain flowers. We cannot of course expect to find conclusive evidence of the geographic origin of all the species of a given zone. Possibly the ancestral forms and 1917.] Chapman, Distribution of Bird-life in Colombia. 89 point of origin of only those it has acquired most recently may be determina- ble, while the derivation of the earlier arrivals may forever be unknown to us. Consequently we have a large group of species whose history is lost to us and which, therefore, we can only assume have entered their zone under the influences which are still active, and the cumulative effects of which they exhibit. Aside from creating areas where, under the influence of a new environ- ment, evolution has evidently proceeded at a highly accelerated pace, the topographic changes incident to the elevation of the Andes have profoundly affected the distribution of life in the Tropical Zone. A comparison of the bird-life of the Pacific coast of Colombia and north- ern Ecuador with that of the Tropical Zone at the eastern base of the Andes in southeastern Colombia and eastern Ecuador, induces the belief that we have here, in part, a pre-Andean fauna, the Pacific portion of which has been cut off from that of upper Amazonia by the Andean uplift. The specific identity of many birds common to both areas is evidence that but little change has taken place in their surroundings since their ranges were dis- connected, and in such cases evolution has, so to speak, been at a standstill. But the elevation of the intervening territory to snow-line has brought into play most of the environmental influences one finds between the equator and the poles, and where in an unchanged basal zone species remained as constant as their habitat, in the new region they sprang forward in an evolutionary race. The evidence on which this theory of the Amazonian origin of Pacific coast life is based is presented in detail beyond. The strongly marked characteristics of the Colombian Pacific Fauna, however, indicate that even in the Tropical Zone evolutionary influences. have been active since the isolation of the Pacific coast region. The bird-life of the Cauca Valley and upper Magdalena Valley appears to have been acquired under existing topographic conditions. The fact that the forests of the Pacific coast compare in Juxuriance with those of upper Amazonia, while forests are of small extent in the Cauca Valley and are wanting in the upper Magdalena Valley, may in part explain the marked difference in the bird-life of these valleys and that of the Pacific coast. Heavy forest, however, exists in the lower Cauca-Magdalena region, the bird-life of which has evidently been acquired in part from the Pacific coast, in part from east of the Andes, suggesting that this region was not above sea-level prior to the Andean uplift. These, however, are faunal rather than zonal problems, and they will be discussed more fully in the succeeding pages. Our studies of the faunal effects of the appearance of the Andes must not be restricted to those changes wrought by the uplift of this system, but 90 Bulletin American Museum of Natural History. [Vol. XXXVI, must also take into consideration the subsequent modifying factors of glaciation, subsidence and erosion. In ‘Die Kordillere von Bogota’ Dr. Alfred Hettner! gives the results of his studies of glaciation in the Eastern Andes showing that at their maxi- mum development glaciers which are now retreating were, in at least one instance, 460 meters lower than at the time his observations were made. He writes (p. 74): “During the last years and decades the boundaries of the snow and the glaciers have probably retreated to a large extent. All the people who served me as guides here agreed to the same opinion; and that the retreat must have been quite marked, may be gathered from the words of a peddling Indian tradesman, who expressed his lively ill-will for the visits of strangers in the mountains, for he thought they were to blame for the disappearance of snow. Upon the snow-covered mountain of San Paulin, the smooth nature of the strata on the western side for several hundred meters below the present snow line clearly shows that it was until recently covered with snow. We have previously heard that from the foot of the Sugar-loaf Glacier to a distance of 700 to 800 m. and a difference in height of 80 m., a great amount of end moraines are found, upon which no plant growth has yet been able to settle, which must, therefore, be of a very recent origin. About contemporaneous with the retreat of the snow-line of the Alps and that of many mountains, the snow-line of the Colombian Andes also re- treated.” On the succeeding page (75) he adds: “In more ancient times the glaciers must have been much larger, the climate consequently cooler, for I noticed on the slope of the valley, to a distance of at least four or five km. from the present foot of the glacier, about 460 m. lower, a row of plain end moraines grown over, and it is possible that they extend down further. “Even if the observation that is taken in general delineations and in text-books does not possess any proof, the fact of the existence of an ice age in the Andes of Colombia, and supposedly also of Venezuela, may be considered as certain. Regarding the appearance of two ice ages, no intelligence has yet been gathered from the equatorial Andes up to the present time.” ' It seems evident, therefore, in view of the different climatic conditions which must have prevailed in the Andes during the period of maximum glaciation, that the existing zonal boundaries are post-glacial. Without 1 Petermann’s Mittheilungen Erganzungsband 22, pp. 74, 75. I have to thank Dr. Chester A. Reeds for these references and the accompanying translation. 1917.] Chapman, Distribution of Bird-lfe in Colombia. 91 a more detailed knowledge of the Andean Ice Age, and of the variation in altitude with latitude, of Andean life-zones, we can only speculate on the extent to which the zones were affected by the rise and fall of a glacial period. We may with some hesitation suggest that so pronounced a boreal type as Otocoris is a recent bit of glacial diift stranded on the Savanna of Bogota. But on the other hand, with far more confidence, we may believe that the undoubted northward extension of South Temperate Zone species along the Andes, with increasing altitude, to the mountain crests of Colombia, has been coincident with the retreat of the glaciers; and the often close rela- tion of these birds with their sea-level, ancestral form supports this view. Whether or not this be true there can be no question of the southern origin of most of the species of the Paramo and Temperate Zones. The trend of life in the Tropical Zone is less susceptible of determination. One cannot say that life does not radiate from an equatorial center and flow north, and south, to the limits of the Tropical Zone; though the north- ward current in America is now not only stronger but reaches farther. In the Subtropical Zone with its extremities reaching into Mexico, the evidence also indicates a current setting toward the north. If, however, this northward bound current is of post-glacial origin, it apparently follows that the former subtropical bridge, which carried the numerous subtropical species now found in Costa Rica to that country from Colombia (as suggested beyond), has disappeared since the Glacial Period. It is in this connection that we find our best illustration of the biogeo- graphic effects of the two other modifying factors,— subsidence and erosion. The retreat of the glaciers to higher altitudes with the resulting upward extension of life-zone boundaries, is accountable for the formation of Paramo Zone islands separated by Temperate Zone areas. Again Temperate Zone islands have apparently been formed by erosion of the mountain crests which at one time connected them. This appears to have happened in the Western Andes where the close relation now existing between the life of the Temperate Zone of the Andes west of Popayan and that of the Paramillo at the northern end of the range, suggests the former continuity of the Temper- ate Zone on the crest of that range. Apparent proof of subsidence, doubtless accompanied by erosion, is found in what I have termed the ‘Panama fault’ in the Subtropical Zone which, after terminating at the northern end of the Western Andes, reap- pears again on the crest of the higher mountains of eastern Panama and of western Panama, though in the intervening areas it is widely separated by the Tropical Zone. The evidence on which this theory of the former con- tinuity of the Subtropical Zone from Colombia to Costa Rica rests, is pre-~ sented beyond. 92 Bulletin American Museum of Natural History. [Vol. XXXVI, The whole question of the origin of Andean life-zones is, from the orni- thological point of view at least, a new one, and I cannot at this stage of our knowledge hope to do much more than open it for discussion. The fundamental factors in zone formation, and the conclusions reached from our studies in Colombia, may be stated as follows: Factors. Rise of the Andes Past Glaciation Erosion Subsidence Temperature Present ; Humidity ; } Environmental factors. Soil Conclusions. 1. Existing faunal conditions in the Tropical Zone are, in part pre- Andean, in part post-Andean. The humid Pacific coast, for example, contains many species which appear to have occupied this region prior to the Andean uplift. The life of the Cauca Valley, on the other hand, seems to be of post-Andean origin, its analysis supporting the geological evidence that this valley was occupied by a lake until post-tertiary time. 2. The Subtropical Zone nowhere reaches sea-level. Its life has been derived from the Tropical Zone. Because of its greater age and because altitudinal extension of range may imply greater environmental change than occurs in latitudinal extension, its life varies more widely from that of the ancestral area than does the life of the remaining two zones. 3. When a Colombian Subtropical Zone form differs from its Ecuadorian ' representative, there is usually one race in the East Andean Fauna and another in the West Andean Fauna. In some instances, however, the Colombian form is alike in all three branches of the Andes, though these three arms of its range are separated by the intervening tropical areas, indicating that the same characters have been developed in the individuals of each mountain chain by parallelism. | - 4. Receding glaciers, erosion, and subsidence have produced zonal islands and zonal ‘faults.’ The distributional evidence on which, for example, the Panama ‘fault’ is shown to have occurred indicates that the Andes of Colombia and the higher mountains of western Panama and Costa 1917.] Chapman, Distribution of Bird-life in Colombia. 93 Rica were, until comparatively recent geologic times, connected by a range having an altitude of not less than five thousand feet. 5. The Temperate Zone reaches sea-level in the South Temperate Zone. Its life is derived in part by zonal, in part by latitudinal extension and is more recent than that of the Subtropical Zone. 6. The Paramo Zone reaches sea-level in southern South America. Its life is derived by altitudinal extension and is more recent than that of the Temperate Zone. 7. The present trend of the distribution of life is northward. Few boreal species have entered Colombia in recent geologic times. 8. With rare exceptions (e. g., Brachyspiza capensis peruviana) no species extends its range from an upper to a lower zone. 9. Wide latitudinal range usually implies wide altitudinal range. 10. Uniformity of life increases with altitude. v Tue TropicaL ZoNE AND ITS FAUNAS. The Tropical Zone in Colombia occupies all that part of the country lying approximately below an altitude of 5000 feet. In some few places it does not extend much above 4500 feet; in others it reaches to about 6000 feet. Its limits are determined primarily by temperature, but they are further dependent upon humidity, as humidity itself affects temperature through radiation. On the heavily forested Pacific slope of the Western Andes the lack of favorable radiating surfaces is conducive to a lower temperature than is found at a corresponding level on the barren eastern slopes of the same range. In consequence, the upper margin of the Tropical Zone is at least 1000 feet higher on the eastern than on the western side of these mountains. Unfortunately no exact data on temperature are available in this con- nection, but the influence of radiation was observed in a marked manner in the succeeding or Subtropical Zone at the San Antonio Pass (alt. 6800 ft.) where the road from Buenaventura to Cali crosses the Western Andes. With a regularity which has given it the name of the “San Antonio Wind,” at two o’clock each day a strong, cold, westerly wind, usually with a driving mist, sweeps over the crest of the range. This phenomenon is apparently attributable to decreasing barometric pressure following radiation from the comparatively open floor of the Cauca Valley, when, already urged by the prevailing westerly wind, air from the coast rushes into the area of lower pressure and is condensed as it reaches the higher parts of the range. 94 Bulletin American Museum of Natural History. [Vol. XXXVI, Where zonal boundaries are determined not only by altitude (= tem- perature) but also by humidity, they are more clearly defined than when attributable to temperature alone. As with faunas, suitability of haunt or habitat here plays an important part. The upper limits of the Tropical Zone in the Cauca Valley, for example, are arid and meet the lower borders of the Subtropical Zone at the lower limit of condensation. In journeying from west to east one therefore passes from the dripping, cloud-wrapped forests of the western slopes and crest of the Western Andes to. the arid, treeless eastern slope of the range. The forest-inhabiting birds of the Subtropical Zone stop as suddenly as the forest itself and they are succeeded by certain species of the arid Tropical Zone which find a suitable haunt on these treeless slopes, whence they have extended their range upward from the dry savannas of the Cauca Valley below. Descending to and crossing this valley, we ascend the bare foothills of the western slopes of the Central Andes only to reverse the experience, as at the cloud line (about 6500 ft.) one leaves the arid Tropical Zone and enters the forests of the Subtropical Zone. On both sides of the valley, however, an extremely interesting interdigita- tion of zonal boundaries is observed as the arid Tropical Zone climbs up the barren ridges or crests-of the spurs of the foothills, while the forests of the Subtropical Zone seem to flow down the drainage areas or arroyos be- tween them. Under such conditions Tropical Zone species are found at higher altitudes than Subtropical Zone species on the same mountain slopes, and the importance of a personal knowledge of the local factors is obvious. -The altitude to which the Tropical Zone ascends is determined, there- fore, primarily by temperature, but, as with faunal boundaries, humidity may exert an important influence not only as it increases the temperature but provides an environment better adapted to the wants of certain species of the arid Tropical Zone than to those of the Subtropical Zone. Our experience at Buena Vista, in the Eastern Andes, made it apparent that the altitude of the upper margin of the Tropical Zone may also, to some extent, depend on the altitude of the range or ridge concerned. Here a heavily forested spur rises from the Llanos, with only gallery forests, to an altitude of 4500 feet. Its direct physical connections are therefore with the Tropical Zone, and in spite of the favorable environment and the alti- tude, very few Subtropical Zone species were secured. -- Accessibility may therefore be a factor in fixing zonal boundaries. It is accessibility which chiefly distinguishes the Tropical Zone from zones above it. ‘Life may enter it wherever it comes in contact with areas lying below an altitude of approximately 5000 feet; and a glance at an orographic on qoId 9} JO IYSL ol} Xa ot} Tau OUT] JSOIOJ BY} 9B ST‘ TON UOMIpedxg| Jo UOMs B‘solOpRIlyY “AoT[VA ay} oAOGe YdJ QOO'S JO 9pNINTe UL OF SpUIISR OLY OUOZ [RITdOIL Pv oy, ‘SopUY UslOJSO A, 9} UL SUOTJIPUOD ALTIUS JO [IeJOpP B AO] AT OVI OSTR aag “WMOYS APART OF PLI dot} OYP AN svorw ouoZ [eRotdory, ple JO JUoIse PU SAUTARI OFLUTeIP dT]? UMOP YSaIOJ aTOZ [RoIdoayqng JO WWaosep 9IL VLITHONY JT VY] NOUd SAGNY IVULN ) 40 FdIOTg NugLseT AQ IIAXX 38£V1q ‘IAXXX “TOA ‘HON CIN CV Nisading 1917.] Chapman, Distribution of Bird-life in Colombia. 95 map will at once show how much more of the earth’s surface lies below than above this level. Within its latitudinal limits the Tropical Zone may therefore be spoken of as a sea of life in which the upper zones are mere islands. The comparison fails, however, when one examines the conditions under which life exists in the Tropical Zone, for instead of finding that uniformity of aspect which characterizes the sea, we find a diversity of environment far beyond that shown by any of the upper life zones. Shore-line, marsh, savanna, llano, plain, and forest afford homes for a correspondingly wide variety of forms, and, in connection with the extent of the area, go far to account for the richness of its life. Returning again to our simile of the sea, when as in Colombia, the upper zone islands assume the rank of peninsulas or are numerous enough to be likened to archipelagos, more or less land-locked bays are formed which, chiefly through their isolation, become centers for the development of new types. All these characteristics of the Tropical Zone, as compared with those of the zones above, are present in Colombia, and an attempt to define its faunal areas results in the recognition of no less than five more or less clearly defined Faunas, as follows: 1. The Colombian-Pacific. The Cauca-Magdalena, including both humid and arid sections. The Caribbean. The Orinocan. . The Amazonian. Of these the first, or Colombian-Pacific, is the most important since in connection with a high degree of humidity, not equalled elsewhere on the tropical Pacific coast, or indeed in the western hemisphere, it combines an isolation which has made it the principal local area of adaptive radiation in Colombia.! To: a limited extent the Cauca-Magdalena region, both in its humid lower Cauca and upper Cauca portions, has produced some new forms, and the Caribbean Fauna, with its extension on the Venezuela coast, appears to have been the birth-place of some of the species which are confined to it. ‘Orinoecan’ and ‘Amazonian’ are terms provisionally applied to the respectively semi-arid and intensely humid portions of the Orinocan and Amazonian basins, which find their western borders at the base of the East- ern Andes. They are merely small parts of much larger faunas and possess no distinctive features of their own. With these general remarks on the St 1 Cf. Osborn, The Age of Mammals, 1910, p. 22. 96 Bulletin American Museum of Natural History. [Vol. XXXVI, Tropical Zone in Colombia as a whole, we may attempt to describe and define its faunas. The range of temperature throughout the zone is so small that this factor has not to be considered in determining faunal boundaries. Humid- ity, character of the soil, and ease of access are the agents which have been active in faunal development in Colombia. Of these the first is by far the most important. The arid upper Dagua basin, on the western slope of the Western Andes, is surrounded by the humid forests of the Colombian- Pacific Fauna and Subtropical Zone. Nevertheless, its life was derived through the Cauca Valley from east of the Andes, suitability of environ- ment as it is controlled by rainfall, here proving far more potent than ease of access from immediately contiguous regions, which do not possess species adapted to, an arid habitat. Birds of the Tropical Zone. Family Tinamide. Colinus cristatus parvicristatus ! Tinamus major ruficeps Odomtephorus eens marmoratus “ “ ; param « « hoe cali Rhynchortyx cinctus australis Cryparus arlepeant Family Columbide e soul soui Columba speciosa . “ caquetee © rufina ee “ cauce « goodsoni « “ modestus « subvinacea berlepschi € adspersus yapura - plumbea propinqua De variegatus salvini Zenaida auriculata « . kerrie “« — ruficauda robidsoni Famalne \Ovaaie Chrsmepelia pasventien a Crax alector . se nana, “ panamensis “ minuta eleodes “ alberti rufipennis rufipennis Penelope ortoni “ « caucee : jacqdagu Claravis pretiosa livida Ortalis columbiana columbiana ! Leptotila verreauxi verreauxi be & caucee } 6 rufaxilla dubusi k “ guttata m « pallidipectus Fe « garrula se plumbeiceps Pipile cumanensis i pallida a < « eo Family Odontophoride. Osculatia ee Colinus cristatus decoratus Oreopeleia veraguensis Q « ~ leucotis+ . montana + 1 Ranging upward to the Subtropical Zone. Buvuietin A. M. N. H. Vor. XXXVI, Puatre XXVIII. FaRALLONES OF Cai, WESTERN ANDES View across the Cauca Valley from La Manuelita, near Palmira, distant about twenty miles. FARALLONES OF CALI, WESTERN ANDES View up the Cali River, from the town of Cali. The summit of this part of the Western Andes is unexplored. It may reach the Temperate Zone. 1917,] Chapman, Distribution of Bird-life in Colombia. Family Opisthocomide Opisthocomus hoazin Family Rallide Pardirallus nigricans nigricans Aramides cajanea cajanea « wolfi (vide Hellmayr) Amaurolimnas concolor guatemalensis Anurolimnas castanciceps ¢ hauxwelli Porzana flaviventris Creciscus senops = albigularis Neocrex columbianus Gallinula chloropus pauxilla Ionornis martinicus Family Heliornithide Heliornis fulica Family Podicipedide Podilymbus podiceps 2 Colymbus dominicus brachyrhynchus Family Laride Phaétusa chloropoda Rhynchops nigra cinerascens Family Charadriide Belonopterus cayennensis 1 AXgialitis collaris Himantopus mexicanus Gallinago brasiliensis Family Parride Jacana spinosa «“ ~ melanopygia nigra Family Edicnemide Burhinus bistriatus “ Family Zurypygide Eurypyga major Family Psophiide Psophia napensis Family Ibidide Theristicus caudatus Harpiprion cayennensis Phimosus berlepschi Family Plataleide Ajaia ajaja , Family Ciconiide Jabiru mycteria Family Ardeide Ardea cocoi Herodias egretta Egretta candidissima Florida cerulea, Hydranassa tricolor tricolor Agamia agami Nycticorax nycticorax nevius * Cochlearius cochlearius # Pilherodias pileatus Butorides striata Tigrisoma lineatum & salmoni Ixobrychus erythromelas Family Palamedeide Palamedea cornuta Chauna chavaria Family Anatide Cairina moschata Dendrocygna discolor & bicolor # Querquedula cyanoptera Marila nationi Nomonyx dominicus Family Phalacrocoracide Phalacrocorax vigua vigua Family Anhingide Anhinga anhinga Family Cathartide Gypagus papa Catharista urubu 2 Cathartes aura aura * 97 1 Ranging upward to the Subtropical Zone. 2 Ranging upward to the Temperate Zone. v 98 Bulletin American Museum of Natural History. Family Falconide Polyborus cheriway ! Tbycter americanus Milvago chimachima! Circus cinereus + « — buffoni } Micrastur guerilla interstes ! Parabuteo unicinctus harrisi 1 Accipiter superciliosus 4 ventralis 8 bicolor Tachytriorchis albicaudatus exiguus Asturina nitida Rupornis magnirostris magnirostris ! « . ruficauda ! Busarellus nigricollis Urubitinga urubitinga ee schistacea, plumbea Leucopternis semiplumbea, « plumbea (vide Hellmayr) Lophotriorchis isidorii ? Spizaétus ornatus ie tyrannus Herpetotheres cachinnans cachinnans : . fulvescens “ Rostrhamphus sociabilis Leptodon uncinatus oe palliatus Harpagus bidentatus Ictinia plumbea ! Falco fusco-ceerulescens 1 « rufigularis ! Cerchneis sparveria cauce 2 « “ intermedia ? Family Bubonide Otus choliba “ watsoni Lophostrix cristatus stricklandi Ciccaba virgata virgata “ ~ nigrolineata Glaucidium brasilianum brasilianum brasilianum phaleenoides Tyto perlata subsp. [Vol. XXXVI, Family Psittacide Ara ararauna “ macao “ chloroptera “ militaris militaris “ severa Aratinga wagleri 1 eruginosa zruginosa, Pyrrhura melanura pacifica Psittacula conspicillata conspicillata a - caucee “ spengeli sclateri Brotogeris jugularis © devillei Amazona inornata = amazonica ochrocephala ochrocephala « f panamensis salvini Pionus menstruus Eucinetus pulchra Pyrilia pyrilia ? Family Alcedinide Megaceryle torquata torquata ! Chloroceryle amazona « americana americana « inda Family Momotide Urospatha martii martii s “« semirufa Electron platyrhynchus platyrhynchus “ “ minor ss i pyrrholzmus Momotus subrufescens subrufescens mn “ reconditus momota ignobilis Hylomanes momotula obscurus “ Family Caprimulgide Nyctibius longicaudatus Chordeiles acutipennis acutipennis es texensis 1 Ranging upward to the Subtropical Zone. 2 Ranging upward to the Temperate Zone. 1917.] Chapman, Distribution Uropsalis lyra, Hydropsalis climacocerca Nyctidromus albicollis albicollis ! Stenopsis cayennensis cayennensis $ & monticola Antrostomus rosenbergi Family Cypselide Streptoprocne zonaris albicincta + -Cheetura spinicauda fumosa £ cinereiventris sclateri “ # occidentalis Cypseloides brunneitorques brunneitor- ques ? Family Trochilide. Androdon equatorialis Threnetes cervinicauda « ruckeri fraseri Glaucis hirsuta affinis «“ eenea Pheethornis yaruqui sancti-johannis “ fraterculus moorei “ hispidus oseryi oe anthophilus griseogularis striigularis striigularis os nee subrufescens Eutoxeres condamini « aquila aquila “ “ salvini heterura Campylopterus equatorialis Florisuga mellivora mellivora Agyrtria viridissima subsp. “ fluviatilis Polyerata amabilis « rosenbergi Lepidopyga goudoti s celina Saucerottea saucerottei - viridigaster Amizilis tzacatl tzacatl a “ — jucunda Hylocharis grayi ! & humboldti “ « of Bird-life in Colombia. 99 Damophila julie julie Chlorostilbon gibsoni . heberlini melanorhynchus ! s poortmani poortmani Thalurania fannyi fannyi é nigrofasciata Chalybura buffoni buffoni ! « ceruleogaster . urochrysa Colibri delphinz Anthracothorax nigricollis nigricollis Chrysolampis elatus Boissoneaua jardini Heliothrix barroti Anthoscenus superba stewarti Calliphlox mitchelli Popelairia conversi Family Trogonide Pharomacrus pavoninus Trogonurus strigilatus strigilatus . strigilatus chionurus « curucui cupreicauda a bolivianus Chrysotrogon caligatus columbianus s ramonianus Curucujus melanurus melanurus s e macrourus massena australis Family Cuculide Coccyzus melacoryphus Piaya cayana columbiana ? “ — mesura } nigricrissa 1 “ rutila rutila « « gracilis Neomorphus salvini Tapera nevia + Crotophaga ani 2 se major Family Capitonide Capito aurovirens “ ~ maculicoronatus rubrilateralis «“ “ 1 Ranging upward to the Subtropical Zone. 2 Ranging upward to the Temperate Zone. 100 Bulletin American Museum of Natural History. Capito maculicoronatus pirrensis « squamatus hypoleucus quinticolor auratus auratus Family Ramphastide Ramphastos piscivorus brevicarinatus “ swainsoni ! ambiguus abbreviatus cuvieri culminatus citreolamus Pteroglossus pluricinctus - castanotis castanotis a torquatus nuchalis sanguineus flavirostris flavirostris s humboldti Selinidera reinwardti a oe spectabilis Family Galbulide Galbula ruficauda ruficauda is . pallens melanogenia “ tombacea tombacea « albirostris chaleocephala Brachygalba salmoni (vide Scl. & Salv.) a fulviventris fulviventris e x caquetze Galbaleyrhynchus leucotis a“ rig “ “ Family Bucconide Bucco capensis Notharcus pectoralis S hyperrhynchus leucocrissus o tectus subtectus Argicus macrodactylus Nystactes noaname Nystalus radiatus Hypnelus ruficollis ruficollis Malacoptila fusca & mystacalis & panamensis poliopis € panamensis Micromonacha lanceolata [Vol. XXXVI, Nonoula frontalis Monasa flavirostris « morphceus peruana pallescens pallescens « sclateri “ ~ nigrifrons Family Picide Chloronerpes xanthochlorus « litee Chrysoptilus punctigula punctipectus at . guttatus f « striatigularis . i. ujhelyii Melanerpes cruentatus i pucherani pucherani rubricapillus rubricapillus Veniliornis fidelis e ruficeps hematostigma & kirki cecilii Celeus loricatus loricatus “« mentalis Campephilus rubricollis a melanoleucus . malherbi Cniparchus hematogaster splendens Ceophlceus lineatus mesorhynchus Picumnus cinnamomeus as squamulatus squamulatus olivaceus olivaceus 1 £ * harterti granadensis granadensis “ a antioquensis Family Conopophagide Conopophaga aurita ss castaneiceps castaneiceps 1 os i chocoensis Family Formicariide Cymbilaimus lineatus lineatus « « fasciatus Taraba unduliger « transandeana transandeana # € granadensis Thamnophilus nigriceps s punctatus punctatus 1 Ranging upward to the Subtropical Zone. 1917.] Thamnophilus punctatus atrinucha canadensis pulchellus “ doliatus doliatus - radiatus albicans £ tenuipunctatus multistriatus 1 Thamnistes equatorialis . anabatinus intermedius Clytoctantes alixi Dysithamnus puncticeps puncticeps “ “ flemmingi leucostictus capitalis capitalis ardesiacus ardesiacus Thamnomanes glaucus Myrmotherula pygmea e surinamensis pacifica Myrmopagis fulviventris « hematonota ornata ornata axillaris melena « “ « ~ albigula € schisticolor schisticolor ! i interior i cinereiventris pallida Herpsilochmus rufomarginatus frater Microrhopias grisea intermedia s boucardi consobrina Ramphocenus melanurus trinitatis e rufiventris griseodorsalis Microbates cinereiventris cinereiventris «“ “ magdalenz « collaris Cercomacra sclateri $ tyrannina tyrannina! ‘ nigricans € berlepschi Anoplops bicolor zequatorialis “ « dague e « ~ bicolor Myrmeciza melanoceps . maculifer maculifer “ “ cassini « lemosticta nigricauda & longipes boucardi se «“ panamensis Myrmelastes immaculatus immaculatus Chapman, Distribution of Bird-life in Colombia. 101 Myrmelastes immaculatus berlepschi ! Gymnocichla nudiceps sancte-marte Dichrozona cinctus Hypocnemis cantator peruviana ft hypoxantha, Hylophylax lepidonota « nevia theresre « « ~ neevioides Myrmoborus leucophrys leucophrys i myiotherinus elegans Phenostictus macleannani macleannani Rhopoterpe torquata torquata Formicarius colma nigrifrons . analis connectens « nigricapillus destructus ~ analis saturatus Chamezxza brevicauda columbiana £ & nobilis? Pittasoma harterti £ rosenbergi y, michleri (vide Cassin) Grallaria brevicauda minor & modesta “ guatimalensis chocoensis Hylopezus dives barbacor € “~ fulviventris « perspicillata periopthalmica : ss perspicillata Family Dendrocolaptide Furnarius agnatus Synallaxis mcesta meesta e « ~ obscura es albescens albigularis ¢ pudica pudica “ « ~ nigrifumosa, a « caucee & gujanensis columbianus & cinnamomea fuscifrons « candzi candsi . rutilans caquetensis Xenerpestes minlosi (vide Berlepsch) Hyloctistes subulatus subulatus “ is assimilis Automolus melanorhynchus a ochrolemus turdinus dorsalis 1 Ranging upward to the Subtropical Zone. 102 Automolus pallidigularis pallidigularis € infuscatus infuscatus s nigricauda saturatus i cinnamomeigula Philydor rufipileatus consobrinus « pyrrhodes ¢ ruficaudatus Ancistrops strigilatus Xenops genibarhis littoralis Sclerurus brunneus £ albigularis albigularis mexicanus obscurior Glyphorhynchus cuneatus subsp. i. castelnaudi pectoralis Dendrocincla lafresnayei lafresnayei . . phzochroa zeiphorhyrcls guttata guttatoides zequatorialis equatorialis lachrymosus lachrymo- sus lachrymosus alarum nanus nanus insignis Dendroplex picus picus a “ picirostris Picolaptes albolineatus Campylorhamphus trochilirostris procur- voides trochilirostris zuelensis thoracicus pusillus + Dendrocolaptes validus validus ! sd sancti-thome sancti- thom sancti-thome radiolatus a“ “ «“ e vene- “ “ Family Tyrannide Ochthornis littoralis Fluvicola pica Arundinicola leucocephala Copurus colonus fuscicapillus “ ~ leuconotus Machetornis rixosus flavigularis Platytriccus albogularis ! Placostomus coronatus Bulletin American Museum of Natural History. [Vol. XXXVI, Craspedoprion xquinoctialis a pacificus Rhynchocyclus sulphurescens asemus & si assimilis f i exortivus « marginatus marginatus « viridiceps ~ flaviventris aurulentus klagesi Todirostrum cinereum cinereum 1 e sclateri e nigriceps a latirostre schistaceiceps superciliare Euscarthmus striaticollis zosterops “ septentrionalis Lophotriccus spicifer . squamecrista minor Orchilus atricapillus Atalotriccus pilaris pilaris Hapalocercus meloryphus Serpophaga cinerea cana ! Inezia caudata intermedia Mionectes olivaceus hederaceus « olivaceus pallidus Hipnamoryhe oleaginea oleaginea 3 parca Leptopogon superciliaris poliocephalus s amaurocephalus Capsiempis flaveola leucophrys Pheomyias murina incompta Camptostoma pusillum pusillum “ « napeum « cauce Microtriccus brunneicapillus brunnei- capillus Tyrannulus elatus reguloides Tyranniscus chrysops chrysops ! Elenia flavogaster flavogaster 1 « gigas parvirostris Myiopagis viridicata accola “pallens “ Legatus albicollis Sublegatus glaber Myiozetetes cayennensis cayennensis = similis columbianus 1 Ranging upward to the Subtropical Zone. 1917.] Chapman, Distribution of Bird-life in Colombia. Myiozetetes granadensis « similis connivens Pitangus sulphuratus rufipennis a « caucensis sulphuratus subsp. - i lictor Sirystes albocinereus Myiodynastes maculatus nobilis Megarhynchus pitangua Myiodynastes luteiventris “ Onychorhynchus coronatus castelnaudi Cnipodectes subbrunneus (vide Hell- mayr) minor Myiohuie barbatus barbatus § atricaudus \ a villosus sulphureipygius aureatus fasciatus fasciatus Terenotriccus erythrurus fulvigularis Myiotriccus ornatus stellatus a phcenicurus Pyrocephalus rubinus rubinus « “ ~ heterurus Empidochanes cabanisi Mitrephanes berlepschi eminulus Sayornis nigricans cineracea Myiochanes brachytarsus Myiarchus tyrannulus tyrannulus fortirostris (ferox) panamensis (ferox) venezuelensis apicalis « tuberculifer tuberculifer “ “ nigriceps “ Tyrannus niveigularis melancholicus satrapa 1 Muscivora tyrannus Family Pipride Pipra erythrocephala erythrocephala berlepschi “ mentalis minor velutina coronata Cirrhopipra filicauda Macheropterus striolatus “ «“« Allocopterus deliciosus Chloropys holochlora holochlora € litee Chiroxiphia pareola napensis Corapipo leucorrhoa “ altera altera Manacus manacus abditivus ¢ interior s « bangsi < « ~ flaveolus vittellinus vitellinus 7 . milleri Scotothorus turdinus rosenbergi i “ stenorhynchus Sapoyoa enigma Family Cotingide Tityra cayana “ ~ semifasciata semifasciata i ¢ columbiana . ee esmeralde “ ~ buckleyi “« a. albitorques (vide Hellmayr). Platypsaris homochrous homochrous ff minor Pachyrhamphus cinnamomeus . magdalenze castaneus saturatus polychropterus niger atricapillus “ Lathria cinerea “, unirufa castaneotincta Lipaugus simplex ew holerythrus holerythrus “ ee rosenhergi Attila citreopygus citreopygus Euchlornis jucunda Cotinga nattereri Carpodectes hopkei Querula purpurata Cephalopterus ornatus Family Hirundinide Tridoprocne albiventris Progne chalybea chalybea Pheoprogne tapera immaculata Atticora fasciata 1 Ranging upward to the Temperate Zone. 103 104 Neochelidon tibialis sieledopion® ruficollis ruficollis « — equalis! uropygialis + Family Sylviide Polioptila livida plumbeiceps ‘ “« daguee schistaceigula Family Troglodytide Heleodytes minor bicolor zonatus brevirostris nuchalis nuchalis turdinus hypostictus albobrunneus harterti Thrgephilne leucotis galbraithi galbraithi albipectus bogotensis rufalbus cumanensis leucopogon nigricapillus schotti Pheugopedius fasciato-ventris fasciato- ventris hypospodius Troglodytes musculus striatulus £ neglectus “ « « « Henicorhina inornata € leucosticta prostheleuca albilateralis Leucolepis salvini q phzocephalus pheocephalus Microcerculus marginatus marginatus « . occidentalis squamulatus antioquensis ri ~ & Family Mimide Mimus gilvus tolimensis ! a “ columbianus Donacobius atricapillus albovittata Rhodinocichla rosea rosea, Family Turdide Planesticus pheopygus ¢ gymnopthalmus tristis daguee : ignobilis ignobilis Bulletin American Museum of Natural History. Planesticus ignobilis goodfellowi ! « « — debilis + - obsoletus columbianus € albiventer ephippialis “Family Vireonide Vireosylva flavoviridis flavoviridis as chivi cauce Pachysylvia semibrunnea * . flavipes flavipes minor Cyelathis flavipectus canticus parvus “ Family Mniotiltide Compsothlypis pitiayumi elegans } « pacifica Dendroica petechia zquatorialis Geothlypis semiflava € sequinoctialis Basileuterus bivittatus chlorophrys if auricapillus olivascens rufifrons mesochrysus fulvicauda semicervinus i s fulvicauda a“ Family Fringillide Cyanocompsa concreta cyanescens cyanea caucse Oryzoborus angolensis < crassirostris crassirostris funereus Sporophila grisea grisea 1 minuta minuta ! castaneiventris aurita aurita . “ — opthalmica muralle & gutturalis + Tiaris olivacea pusilla + “ bicolor omissa Volatinia jacarini splendens Pitylus grossus Saltator maximus « ~ olivascens cerulescens azare striatipectus striatipectus ! “ “ “ “ “ 1 Ranging upward to the Subtropical Zone. [Vol. XXXVI, 1917] Astragalinus columbianus 1 Sicalis flaveola « arvensis minor Ammodramus savannarum cauce Myiospiza manimbe columbiana Md cherriei “ aurifrons Arremenops conirostris conirostris € € inexpectata « te chrysoma Emberizoides sphenurus Paroaria gularis Arremon aurantiirostris erythrorhynchus i. € occidentalis 1 ff axillaris Family Cerebide Ceereba luteola luteola «mexicana columbiana 1 mexicana cauce Dacnis cayana cayana . “ ~ ecerebicolor « angelica egregia egregia venusta fuliginata leucogenys Cyanerpes cyaneus pacificus « cerulea microrhyncha Chlorophanes spiza exsul # “ cerulescens ! Family Procniatide Tersina viridis occidentalis 1 Family Tanagride Tanagra cyanocephala cyanocephala ! « aurea pileata « xanthogastra chocoensis ! i e brevirostris 1 “ concinna saturata a olivacea humilis “ fulvicrissa purpurascens « omissa & crassirostris crassirostris melanura « ehrysopasta Chapman, Distribution of Bird-life in Colombia. Tangara, chilensis « ~~ schranki johannze xanthogastra florida auriceps vitriolina ! lavinia lavinia palmeri mexicana boliviana inornata, larvata fanny Buthraupis rothschildi Thraupis episcopus leucoptera ceelestis ccelestis cana cana glaucocolpa palmarum melanoptera Ramphocelus nigrogularis : dimidiatus dimidiatus carbo carbo & “ unicolor flammigerus chrysonotus icteronotus Chlorothraupis olivacea £ stolzmani Pheenicothraupis gutturalis Heterospingus xanthopygius Tachyphonus rufus ! ¢ luctuosus « surinamus surinamus i delattri Eucometis cristata cristata Mitrospingus cassini Erythrothlypis salmoni Hemithraupis peruana ee guira guirina Cissopis leveriana minor Schistochlamys atra Family: Icteride Zarhynchus wagleri wagleri Gymnostinops guatimozinus “ yuracares Ostinops decumanus ba angustifrons Cacicus cela, 1 Ranging upward to the Subtropical Zone. 106 Bulletin American Museum of Natural History. [Vol. XXXVI, Cacicus vitellinus Leistes militaris “ heemorhous affinis Icterus auricapillus “ uropygialis pacificus « — mesomelas salvini Amblycercus solitarius « — hondee s holosericeus flavirostris ? « — xanthornus xanthornus Cassidix oryzivora violea * Megaquiscalus major assimilis Molothrus bonariensis atronitens . : “ “ cabanisi Family Corvide « & eequatorialis Cyanocorax affinis affinis Agelaius icterocephalus icterocephalus s violaceus , THE FAUNAS OF THE TROPICAL ZONE. The Colombian-Pacific Fauna.— Th: fauna for which I propose this name is one of the most circumscribed and sharply defined, and possibly the most strongly characterized of any fauna of tropical South America. Certainly no other area of similar extent in the Tropical Zone has so many birds which are peculiar to it. In brief, this fauna occupies the entire humid Pacific coast region of tropical South America. This includes the whole extent of the Colombian coast and the more northern portions of the coast region of Ecuador. Its southern limits are marked by the northern limit of what may be termed the Equatorial Arid Fauna. The boundaries of these faunas remain to be determined. Roughly speaking they may be found in the Province of Manavi, where the Arid Fauna finds its northern limit near Bahia Caraque. The humid Colombian-. Pacific here recedes from the coast and extends south of Bahia Caraque in the region of heavier rainfall near the base of the Andes. Northward, the Colombian-Pacific Fauna contributes an important element to the composite life of the lower Cauca-Magdalena district, which it enters through the forested region at the end of the Western Andes; north- westward it continues into the Tuyra region of eastern Panama, where it is also associated with Cauca-Magdalena species, which, like Ostinops decu- manus and Donacobius atricapillus albivitta, are of purely eastern origin. Many Central American species appear to have been derived from this small but important area, and its influence may be traced even into southern Mexico; but as a fauna, we may perhaps set its northern boundaries in the Tuyra River system of eastern Panama. Climatically, the Colombian-Pacific Fauna is distinguished by an excep- tionally heavy rainfall. Exact meteorological data are lacking, but, as else- 1 Ranging upward to Subtropical Zone. 2 Ranging upward to the Temperate Zone. 1917,] Chapman, Distribution of Bird-life in Colombia. 107 where stated, the precipitation at San José, east of Buenaventura, on the Colombian coast, reached 400.88 inches in the year 1912, an amount doubt- less not equalled in any other part of the Western Hemisphere. There is no marked dry season in those portions of this region with which we are familiar and, as might be expected, it is everywhere covered with luxuriant tropical forests. 107 Fig. 3. Known Range of Sapoyoa enigma, a Characteristic Species of the Colombian-Pacific Fauna of the Tropical Zone. ‘a The arid pockets, like the upper Dagua basin, which lie between the coastal forests and those of the Subtropical Zone, have nothing in common with the Colombian Pacific Fauna, their life evidently having been derived from the dryer country lying to the east of the Western Andes. Some 150 species and subspecies are now known from Colombia which are largely or wholly restricted to the Colombian-Pacifie Fauna. Others, 108 Bulletin American Museum of Natural History. [Vol. XXXVI, known as yet only from Ecuador, doubtless occur in it; for it must always be remembered that not only are there portions of this fauna which have never been visited by an ornithologist, but also that.in no other part of tropical America is the collector confronted by more unfavorable conditions than those which prevail in the humid coastal district of Colombia. Of the 150 species and subspecies which are known to characterize this ‘B0° 7o° 60° Sewer Fig. 4. Range of Zarhynchus wagleri. northward to Mexico. A Tropical Zone species of the Pacific Coast which ranges fauna in Colombia, less than fifty can be classed as representative races of widely distributed species, leaving therefore approximately one hundred, or two-thirds the known characteristic forms, as autocthonous. This remarkably large proportion for a continental area of such com- paratively limited size indicates that the environment is unusual, the isola- tion effective, and possibly also that the region was formerly more extensive. 1917.]- Chapman, Distribution of Bird-life in Colombia. 109 It is not probable that the environment differs appreciably from the densely forested region lying at the eastern base of the Andes; but that this humid Pacific coastal region is isolated from other regions of essentially similar character is evident. To the south, the arid coasts of Peru and northern Chile extend far below the limits of the Tropical Zone. To the east the Andes form an evidently impassable bulwark three zones high, the upper two of which are largely or wholly devoid of forest growth. To the west, lies the Pacific Ocean and, consequently, under existing topographic conditions, these luxuriant coastal forests of northern Ecuador and western Colombia..ean have received their life only from the north. That some of their forms have entered it from this direction is obvious; but they are to be found among the fifty races which represent widely distributed and chiefly Amazonian species. Examples are Cymbilaimus lineatus fasciatus, Myrmotherula surinamensis pacifica, Tityra semifasciata esmeralde, Stelgidopteryx r. uropygialis, Basileuterus fulvicauda, Arremon e. occidentalis, and other birds which have what may be called a completed distribution, that is, occupy all the territory in which they might be expected to occur. Once having reached the forested, lower Cauca-Magdalena district, in which they are all represented, there is nothing to prevent these species from ranging southward to western Ecuador through the Colombian coastal forests. We cannot, however, regard this Cauca-Magdalena district as the gate- way into the Colombian-Pacific Fauna for those west coast species which are unknown in the Cauca-Magdalena Fauna. Many of these, as has been said, advance westward and northward into Central America and it may be suggested that the autocthonous Colombian-Pacific forms have originated in Central America, since access from other areas seems to be impossible under existing conditions. Possibly some of them may have been so derived, but the fact that a large part of them are unknown north of eastern Panama prohibits a, belief in their northern origin. Even those species which like Neomorphus salvini and Selinidera spectabilis are found as far north as Nicaragua, assuredly cannot be considered to have entered the Colombian-Pacific Fauna from the north. Both belong to Amazonian groups, and both, in South America, are known west of the Andes only in the humid coast region; neither having been recorded from the lower Cauca-Magdalena district or from western Venezuela. Other species fall into this same class, that is, they are repre- sented in the Colombian-Pacific Fauna and also east of the Andes, but at the north their ranges are apparently not connected. A list of certain species 110 Bulletin American Museum of Natural History. [Vol. XXXVI, in this interesting group is given at the end of this section. In their distri- bution I believe that we have a clue to the origin of many species of the humid Ecuador-Colombian littoral whose presence in that region may not be otherwise accounted for. Briefly, these forms appear to have been derived from upper Amazonia before the Andes had acquired a sufficient altitude effectually to separate, as they do now, the Tropical Zones at their eastern and western bases. A strong affinity exists also between the fresh-water fishes of these two regions. Scharf remarks: “The fresh-water fish fauna of the Pacific slopes of southern Ecuador still exhibits such affinity to that of the Amazon, that the Ecuador mountains could only have had a slight elevation until com- paratively recent geological times.” 4 Henn, in confirmation of these statements, writes that “the fishes of the Pacific slope are in general widely distributed Amazonian types; none of them would cause surprise if taken at Manaos.”’ ? Wolf states that the flora of humid western Ecuador is essentially like that of Panama and the Chocé region of western Colombia, and adds many species are identical with or belong to the same genus as those found on the eastern slopes of the Andes. (Geographia y Geologia del Ecuador p. 439.) Having in mind the possibility of the Amazonian origin of the Pacific humid fauna, W. B. Richardson in 1913, after his explorations for the American Museum on the Ecuador coast, made, at our request, a section across the Andes from Santa Rosa, south of Guayaquil, through Zaruma and Loja to Zamora in the Tropical Zone at the eastern base of the Andes. It is proposed to report on his collections and notes in a subsequent paper, but in this connection it may be said that the results of his journey strongly suggest, as the topography of the region indicates, that this section was one of the latest to be closed to the passage of Tropical Zone forms from one side of the Andes to the other. In journeying from Loja to Zamora, Richardson crossed the intervening mountains, which here attain an altitude of 11,500 feet; but the Rio Zamora, rising in the Loja Valley breaks through these mountains at a much lower elevation, below of course, that of Loja, which is given by Richardson as 7260 feet. This theory of the transandean origin of the Pacific humid fauna affords a satisfactory explanation for the presence in western Ecuador and south- western Colombia of a number of common species which are also represented in eastern Panama, or the Cauca-Magdalena district, but are unknown on 1 Distribution and Origin of Life in America, 1912, p. 360. 2 Arthur Henn, Science, N. S., XL., 1914, p. 603. 1917,] Chapman, Distribution of Bird-life in Colombia. 111 90° 80° 70° 60° Fig. 5. Known Range of Osculatia. Osculatia sapphirina occurs in the Tropical Zone at the eastern base of the Andes (probably north to the latitude of the Guaviare River). Osculatia purpurata inhabits the Colombian-Pacific Fatima of the Tropical Zone at the western base of the Andes, their ranges apparently being separated by the Andes. 112 Bulletin American Museum of Natural History. [Vol. XXXVI, the west coast of Colombia north of the Patia River. Reaching western Ecuador, before the Loja region had attained its present altitude, they have also advanced westward around the northern end of the Andes in Colombia, but have not as yet completed their distribution either by ranging north- ward from Ecuador or southward from northern Colombia. 30° 80° 70° 60° Fig. 6. Distribution of the western races of Manacus manacus. Illustrating the apparent absence . on the Pacific coast north of the Patia River of a common, widely distributed Amazonian species which is found in western Ecuador and northern Colombia. M, -M, he 4 1. 2. M purus. 3. Manacus manacus interior. AA, fl, i} M;, r bangsi. 4, Manacus manacus abditivus. 5. J 6. ' 7 Me lanochlamys. @ Mil ie 1917.] Chapman, Distribution of Bird-life in Colombia. 113 Doubtless the absence of some of these species from western Colombia may be apparent rather than real, but such common, conspicuous species as Manacus manacus, and Arremenops conirostris are not likely to escape the collector, if present. Possibly the heavy forests of the Colombian- Pacific may have prevented the entrance of Arremenops into this region, but this explanation cannot be offered to account for the absence of Curu- 90° 70° ~ G0" °° Pe, 2 12d . on eo “BRITISH oo ve @ T HONDURAS 6, 90° 70° 6o Fig. 7. Known range of Arremonops conirostris. Illustrating the apparent absence on the Colombia-Pacific coast, north of the Patia River, of a species represented in Panama and northern Colombia by the same or an allied species. 1. Arremenops conirosiris chrysoma. 2. Arremenops conirostris inexpectata. 8. Arremenops conirostris richmondt. 4, Arremenops conirostris conirostris. cujus melanurus, while the theory of transandean origin will explain why. the west Ecuador form of this species should agree with that of Amazonia (C. melanurus melanurus) rather than with that of the Cauca-Magdalena district (C. m. macrourus). 114 Bulletin American Museum of Natural History. [Vol. XXXVI, Even within the restricted limits of the Colombian-Pacific Fauna and under the uniform conditions which prevail in it, considerable racial varia- tion has occurred. Note, for example, the distribution of the Capito- squamatus-maculicoronatus group as mapped herewith. Forms from the Atrato River not infrequently differ more or less from those from the San “ —" BRITISH “tut NDURAS 60° Fig. 8. Ranges of Capito squamatus and C. maculicoronatus — to illustrate the breaking up of a characteristic Colombian-Pacific Fauna group into a distinct species south of the Patia River and three races north of it. 1. Capito squamatus. 2. Capito maculicoronatus rubrilateralis. 3. Capito maculicoronatus pirrensis. 4. Capito maculi tus li t Juan River and southward, and others living south of the Patia differ from those found north. This river indeed appears to form the northern limit of a number of species, but in view of the lack of knowledge of the coast region lying between the Patia and Buenaventura, it is not well to be positive in this connection. 1917,] Chapman, Distribution of Bird-life in Colombia. 115 List of Species and Subspecies which Characterize the Colombian-Pacific Fauna. Tinamus major latifrons 1 Crypturus kerriz © berlepschi soul modestus Crax panamensis Penelope ortoni Odontophorus parambe Rhynchortyx cinctus australis Columba goodsoni Columba subvinacea berlepschi Leptotila plumbeiceps e pallida a cassini 2 Osculatia purpurata Oreopeleia veraguensis Aramides wolfi Creciscus albigularis Leucopternis plumbea « semiplumbea Herpetotheres cachinnans fulvescens Pyrrhura melanurus pacifica + Eucinetus pulchra Electron platyrhynchus platyrhynchus Antrostomus rosenbergi ! Androdon equatorialis Threnetes ruckeri fraseri Glaucis enea Pheethornis yaruqui sancti-johannis i striigularis subrufescens Eutoxeres aquila salvini ef “ ~heterura Polyerata rosenbergi i amabilis Amazilis tzacatl jucunda Hylocharis humboldti Thalurania fannyi fannyi Chalybura urochrysa Boissoneaua jardini Heliothrix barroti Trogonurus strigilatus chionurus Curucujus massena australis . Neomorphus salvini “ Capito maculicoronatus rubrilateralis “ ~ squamatus 1 “ — quinticolor Ramphastos swainsoni « ambiguus abbreviatus Pteroglossus sanguineus Selinidera spectabilis Galbula melanogenia Notharcus pectoralis Nystactes noanamz Nystalus radiatus Malacoptila poliopis poliopis * ff « panamensis ? Micromonacha lanceolata + Monasa pallescens pallescens ? Chloronerpes lite Melanerpes pucherani pucherani Celeus loricatus loricatus 3 Celeus loricatus mentalis * Cniparchus hematogaster splendens Picumnus olivaceus harterti 1 Cymbilaimus lineatus fasciatus Thamnistes anabatinus intermedius Dysithamnus puncticeps puncticeps ? i i. flemmingi ! Myrmotherula surinamensis pacifica Myrmopagis fulviventris € axillaris albigula Microrhopias boucardi consobrina Microbates cinereiventris cinereiventris Cercomacra berlepschi Anoplops bicolor bicolor ? «“ « daguze 4 sequatorialis ! Myrmelastes immaculatus berlepschi Phenostictus macleannani macleannani Myrmeciza maculifer maculifer * i Me cassini 2 f lemosticta nigricauda Hylophylax nevioides Formicarius nigricapillus destructus Pittasoma harterti ! “ “a 1 Known only from south of the Patia River. 2 Known only from north of the San Juan region. 3 Known only from south of the San Juan region. 4San Juan River Region. 116 Pittasoma rosenbergi * « michleri 2 Grallaria guatimalensis chocoensis “ perspicillata perspicillata * ee periopthalmica ! Hylopezus dives barbacoz ! Synallaxis pudica nigrofumosa Automolus nigricauda saturatus * Hyloctistes subulatus assimilis Xiphorhynchus equatorialis equatorialis ee lachrymosus lachrymosus Campylorhamphus thoracicus ! Craspedoprion pacificus Rhynchocyclus marginatus marginatus Todirostrum sclateri 1 Lophotriccus squameecrista minor 3 Cnipodectes subbrunneus Myiobius villosus “3 sulphureipygius aureatus Myiobius lite Terenotriccus erythrurus fulvigularis Mitrephanes berlepschi eminulus Tyrannus niveigularis ! Pipra mentalis minor Allocopterus deliciosus 4 Chloropipo holochlora litz Corapipo altera altera Manacus manacus bangsi ! # vitellinus vitellinus Scotothorus turdinus rosenbergi Sapoyoa enigma Tityra semifasciata esmeraldz } Lathria unirufa castaneotincta Lipaugus holerythrus holerythrus ? € * rosenbergi ? Euchlornis jucunda ! Cotinga nattereri Carpodectes hopkei Neochelidon tibialis Bulletin American Museum of Natural History. [Vol. XXXVI, Stelgidopteryx ruficollis uropygialis Poliptila livida dague « ~ schistaceigula Heleodytes albobrunneus harterti Thryophilus nigricapillus schotti fe leucopogon Henicorhina inornata Leucolepis pheocephalus phzocephalus Microcerculus marginatus occidentalis . squamulatus antioquensis Planesticus tristis dague Pachysylvia minor Dendroica petechia equatorialis 1 Basileuterus fulvicauda semicervinus £ bivittatus chlorophrys ! Sporophila aurita aurita ? ‘ “ — ophthalmica ? Arremonops conirostris chrysoma + Arremon aurantiirostris occidentalis Dacnis cayana ccerebicolor Cyanerpes cyaneus pacificus Tanagra xanthogastra chocoensis Tanagra saturata Tangara florida auriceps « — johannee « ~ palmeri & lavinia lavinia Buthraupis rothschildi Ramphocelus icteronotus Chlorothraupis olivacea f stolzmani ! Heterospingus xanthopygius Tachyphonus delattrii Mitrospingus cassini Erythrothlypis salmoni Zarhynchus wagleri wagleri Cacicus uropygialis pacificus Molothrus bonariensis zequatorialis ' 1 Known only from south of the Patia River. 2 Known only from north of the San Juan region. 3 Known only from south of the San Juan region. 4San Juan River Region. 1917.] Chapman, Distribution of Bird-life in Colombia. 117 List or Species on REPRESENTATIVE Forms Founp IN THE Tropical ZoNE oF SoUTHEASTERN COLOMBIA AND EasTERN Ecuapor AND WESTERN CoLomBia anp NortTawesTeRN Ecuapor, THE RANGES OF Waicu, SEPARATED BY THE ANDES, ARE NoT KNowN To BE CoNNECTED.! Eastern or Amazonian Leptotila rufaxilla pallidipectus Osculatia sapphirina Pyrrhura melanura melanura Electron platyrhynchus pyrrholemus Curucujus melanurus melanurus Selinidera reinwardti Micromonacha lanceolata Thamnistes sequatorialis Microrhopias quixensis Microbates collaris Hylopezus dives fulviventris Hyloctistes subulatus subulatus Dendrocolaptes sancti-thome radiolatus Myiotriccus phoenicurus Hapalocercus meloryphus ? Cnipodectes subbrunneus minor Chloropipo holochlora holochlora Manacus manacus interior Cephalopterus ornatus Leucolepis salvini Sporophila aurita muralle Arremonops conirostris conirostris Tanagra schranki Cacicus cela Western or Pacific Leptotila rufaxilla dubusi Osculatia purpurata Pyrrbura melanura pacifica Electron platyrhynchus platyrhynchus Curucujus melanurus melanurus Selinidera spectabilis Micromonacha lanceolata Thamnistes anabatinus intermedius Microrhopias boucardi consobrina Microbates cinereiventris cinereiventris Hylopezus dives barbacoze Hyloctistes subulatus assimilis Dendrocolaptes sancti-thome thom Myiotriccus ornatus stellatus Hapalocercus meloryphus Cnipodectes subbrunneus subbrunneus Chloropipo holochlora, lite Manacus manacus bangsi « « melanochlamys Cephalopterus penduliger Leucolepis phzeocephalus phzeocephalus Sporophila aurita ophthalmica Arremonops conirostris chrysoma Tanagra florida auriceps Cacicus flavicrissus sancti- The Cauca-Magdalena Fauna.— The faunal area to which I would apply the name Cauca-Magdalena embraces that part of the Tropical Zone which is drained by the Cauca and Magdalena River systems from their source northward to the arid coastal region, or Caribbean Fauna. It is divisible into arid and humid sections. The former embraces the entire Cauca Valley and extends northward into Antioquia nearly to the upper limits of navigation on the lower Cauca, and,also the upper Magda- __ Re. 1In a future paper on the distribution of bird-life in Ecuador it is proposed to treat more fully of the origin of the avifauna of the Tropical Zone of the Pacific Coast. In this connection I present only some of the more striking instances of Tropical species common to both the Pacific Coast and eastern Ecuador and Colombia, whose range appears to be separated by the Andes. 2S, W. Ecuador, Magdalena Valley, and Amazonia, but unknown on Pacific-Colombian coast. 118 Bulletin American Musewm of Natural History. [Vol. XXXVI, lena Valley from its head northward to the upper limits of heavy forest growth near La Dorada. The latter includes the lower Cauca-Magdalena forests which cover the bottomlands and lower slopes of the mountains from La Dorada northward to the vicinity of Banco. To the northeast the country bordering the Rio Cesar is of the open, savanna type, and belongs to the Caribbean Fauna, but it is possible that a belt of Tropical Zone 90° Fig. 9. Known distribution of Mier ha I lata, a species of the Tropical Zone which is found at both the eastern and western bases of the Andes. forest skirts the Eastern Andes and crosses its northern end to connect with the forests of the southern Maracaibo district... This connection, however, is not known by me to exist. In any event, it is not probable that the association of species forming the Cauca-Magdalena Fauna extends far into the Cesar Valley. To the west, the Cauca-Magdalena Fauna reaches at least to the Tuyra 1 Compare Simons’s map of the Goajira Peninsula (Proc. R. G. S., 1885) where a considerable area at the northeastern end of the Eastern Andes bears the name “Montes de Oca (woods).” 1917,] Chapman, Distribution of Bird-life in Colombia. 119 district of eastern Panama, but from the lower Atrato Valley westward it merges so completely with the Colombian-Pacific Fauna that any attempt to map their respective boundary lines in this region must be purely arbi- trary. : The Cauca-Magdalena Fauna possesses comparatively few species pecu- liar to itself, its life consisting chiefly of forms received from the Ama- zonian region on the east and Colombian-Pacific Fauna on the west. The “BRITISH J] ! HONDURAS 107 Fig. 10. Known range of Thamnophilus nigriceps Scl. A species of the humid Colombian-Pacific Fauna of the Tropical Zone. result is a composite group to which neither of the above names ‘could properly be applied. The region, however, is more than a meeting ground for species originating elsewhere. It contains a number of such strongly. marked indigenous forms as Capito hypoleucus, Xenerpestes minlosi, and Gymnostinops guatamozinus, and for this reason, as well as for convenience in descriptive zodgeography, it is deserving of faunal rank. 120 Bulletin American Museum of Natural History. [Vol. XXXVI, 70° 0° go? “BRITISH ! HQNOURAS 9a° 80° Fig. 11. Range of Ostinops decumanus. A Tropical Zone, Amazonian species which enters Colombia from the east, extends southward up the Magdalena and Cauca Valleys and westward to western Panama, but is unknown on the Pacific Coast of Colombia except on the lower Atrato. Dotted area — General South American range. Black area — Range west of the Andes. a“ 1917.) Chapman, Distribution of Bird-life in Colombia. 121 The humid portion of the Cauca-Magdalena Fauna lies chiefly in the Department of Antioquia. It is largely covered with heavy, primeval forest, and is connected with the forests of the Colombian-Pacifie Fauna by the forested area at the northern end of the Western Andes. This continuous forest growth has made the lower Cauca-Magdalena district easy of access to Pacific coast forms, which have entered it in such large 90° 70°, Gi 90° Fig. 12. Known range of Myrmeciza exsul. A species which enters the Cauca-Magdalena Fauna from the west. numbers that it might well be considered a part of the Colombian-Pacific Fauna, had not invasion from the east given the Amazonian region even greater claims upon it. When, however, one compares the narrow strip of country lying to the west with the vast area lying to the east, the proportion of western to east- 122 Bulletin American Museum of Natural History. [Vol. XXXVI, ern forms is surprisingly large. It is, however, to be especially noted that whereas many of the Amazonian forms have crossed the Cauca-Magdalena district and entered Panama and even Central America, no distinctly Pacific coast species appears to have gone east over the route at the northern end of the Eastern Andes by which Amazonian speciés have traveled west- ward. This fact might indicate that in spite of the proximity of the district whence it has been derived, the west coast element is of more recent origin, but a comparison of the changes which have occurred in both groups since their establishment in the humid Cauca-Magdalena Fauna does not con- firm this theory, species of western origin showing as much racial variation _ as those from the east. The abrupt cessation of forest growth on the floor of the Magdalena Valley at La Dorada marks the southern limit of the range in that valley of the forest-inhabiting species which characterize the humid Cauca-Magda- lena Fauna. Tropical Zone forest extends at least as far south as the lati- _ tude of Honda, on the slopes of the mountains, carrying with it such forest- loving species as Myrmelastes immaculatus and Formicarius analis saturatus, but beyond this the avifauna of the Tropical Zone of the upper Magdalena Valley is composed of species which frequent plains, thickets and low scrubby woods. Examples are Colinus cristata leucotis, Ortalis columbiana columbiana, Psittacula conspicillata conspicillata, Thamnophilus radiatus albicans, Myrmeciza l. boucardi, Arremenops conirostris conirostris and A. c. inexpectata, Thraupis c. cana, Thraupis palmarum melanoptera, etc. With but few exceptions all the Tropical Zone species inhabiting the upper Magdalena Valley have evidently entered it from the north passing the forested area lying between Banco and La Dorada. Its life, therefore, resembles that of the arid lower Magdalena or Caribbean Fauna, rather than that of the Tropical Zone at the eastern base of the Eastern Andes in the same latitude. Not only have these mountains proved a barrier to extension of range directly over them, but of equal, or in view of the low altitude of the Anda- lucia Pass (7000 ft.) of possibly greater importance, is the lack of forests in the upper Magdalena which would afford a favorable home for the spe- cies inhabiting the densely wooded region at the eastern base of the moun- tains. In several instances, however, notably with species not so strictly con- fined by zonal boundaries as is customary, it is evident that forms of the extreme upper Magdalena Valley have entered it from the east over the mountains. The known examples are Piaya cayana mesura (upper Mag- dalena specimens agreeing with those of the eastern slope of the Andes in- stead of with those of the vicinity of Honda), Conopophaga castaneiceps 1917.) Chapman, Distribution of Bird-life in Colombia. 123 90° Fig. 138. Range of Donacobius atricapillus. An Amazonian species which enters Colombia from the east, ds the Magdal Valley to at least Honda, crosses the lower Cauca-Magdalena district to the lower Atrato and Tuyra district in eastern Panama, but is unknown in the Cauca Valley. Dotted area — general South American Range. Black area — Colombia range. 124 Bulletin: American Museum of Natural History. [Vol. XXXVI, castaneiceps (found.in the forests on the west slopes below Andalucia), Myiotriccus 0. phenicurus, Tanagra chilensis, and Schistochlamys atra. , It should be added that our work has been done about the borders of this upper Magdalena district. Of the fauna of the floor of the valley, I — feel that we have still much to learn. The second southward extending arm of the Cauca-Magdalena Fauna enters the Cauca Valley from Antioquia and reaches almost to Popayan. It seems highly inaccurate to speak of so fertile a district as the Cauca Valley as arid, nevertheless, in the light of our present knowledge, it must be ranked as an arid branch of the Cauca-Magdalena Fauna. The marshes and bayous of the Cauca River support a variety of aquatic and palustrine species unknown to the dryer upper Magdalena, but aside from this difference there is a marked similarity in the Tropical Zone life of each. There is more forest growth in the Cauca than in the upper Magdalena Valley, in spite of widespread deforestation. Localities like those visited by Allen and Miller at Rio Frio, and the country through which we passed about Guengiie, seem well-adapted to the needs of forest-haunting species; nevertheless, we have thus far failed to find in the Cauca Valley a single representative of the families Momotidae, Trogonide, Galbulide or Buc- conide, and but one species of Ramphastide, the widely distributed Aulaco- rhynchus hematopygius. Possibly the comparatively limited amount of forest-growth may account for the apparent absence of those species of these groups which inhabit the lower Cauca region, and might therefore be expected to occur in the Cauca Valley. But it is evident that its isolation, and the fact that the Tropical Zone enters it at the north where it is separated from the forests of the lower Cauca by long stretches of treeless, truly arid country, are all factors which must be taken into consideration in accounting for its apparently limited life. That this life is actually limited I believe to be a fact, but I also believe that further collecting in the forests of the valley will result in the discovery of species which have not thus far been taken there. In spite, therefore, of the physical differences between the upper Mag- dalena and Cauca Valleys, their land-bird life is much the same. In both instances it has been derived indirectly from east of the Andes by a current which appears to have flowed northward around the end of the Eastern Andes, and thence southward up to the heads of the valleys. The upper Magdalena, being far more accessible geographically, and having a narrower belt of humid tropical forest at its mouth, has received the greatest number of species. The following common birds for example, of the upper Magdalena are as yet unknown from the Cauca Valley: Broto- o\3 : BUENOS AIR “100° 90! go* 60° Fig. 14. Range of Thraupis palmarum. A widely distributed South American Tropical species which enters Colombia from the east; ranges southward to the head of the Magdalena and westward through the lower Cauca-Magdalena district to the Pacific coast, north to Costa Rica, and as far south at least as Buenaventura, but is unknown from the Cauca valley. Dotted area — General South American range. Black area — Range in Colombia and adjoining territory. 125 126 Bulletin American Museum of Natural History. [Vol. XXXVI, geris jugularis, Galbula ruficauda, Myrmeciza longipes, boucardi, Furnarius agnatus, Atalatriccus p. pilaris, Hapalocercus meloryphus, Manacus manacus flaveolus, Heleodytes m. bicolor, H..brevirostris, Planesticus a. ephippialis, Saltator maximus, Arremenops conirostris, Thraupis palmarum melanoptera, Cyanocorax affinis, etc. To this list others might be added. Doubtless further field work in the Cauca Valley will result in the discovery of some of them, but it seems clear that there will still remain a goodly number whose absence can be accounted for only by the isolation of the valley. I dwell on the fact for, it seems clearly to indicate that the life of the Cauca was acquired under existing topo- graphic conditions. Notwithstanding its nearness to the Pacific coast and the comparatively low altitude of the Western Andes, few species have entered the Cauca Valley from the Pacific coast. This cannot be wholly due to the differences in the physical characteristics of these two districts, since there are many species of the Pacific coast which would find a congenial home in the Cauca Valley, but which have evidently been prevented from entering it by the intervening mountains. In the appended list of nineteen species common to the Pacific coast and Cauca Valley, at least thirteen are of East Andean origin, and are more likely to have entered the Cauca Valley from the lower Cauca-Magdalena district than from the Pacific coast. Marila nationi is a duck of unknown origin, Leptotila plumbeiceps occurs as far north as Central America and is doubtless found in the lower Cauca-Magdalena district, where Pachy- rhamphus dorsalis is also found, leaving therefore only two species, Tanagra saturata and Manacus vitellinus vitellinus which with any certainty can be said to have entered the Cauca Valley from the Pacific coast. The life of the Cauca Valley has been therefore derived, with surpris- ingly few exceptions, from that part of South America lying east of the Andes, and has but slight affinity with that of the Pacific-Colombian Fauna. But if the Pacific coast has given but little to the life of the Cauca Valley, the valley has made noteworthy contributions to the arid upper Dagua Basin lying in thé Tropical Zone on the western slope of the Western Andes. Of thirty-three species taken by Richardson at Caldas on the upper Dagua, sixteen are common to the Pacific coast and the Cauca Valley, thirteen are known from the Cauca Valley but not from the Pacific coast, while only one, Sayornis n. cineracea, is recorded from the Pacific coast but not from the Cauca Valley, where, however, it may occur. This case clearly illustrates the necessity of giving due consideration to suitability of environment in any attempt to solve distributional problems. The Pacific-Colombian Faunal area, an intensely humid district, cannot be expected to’ enter Butietin A. M.N. H. Vor. XXXVI, Prare XXIX. Cauca VALLEY FROM SAN ANTONIO The Cauca Valley lies below and the Central Andes arise behind the clouds. Trail to Cali in the foreground. Taken from the lower border of bushy vegetation near the San Antonio Pass. (Junction of Tropical and Subtropical Zones.) re MORE Sintec Cauca VAaLLey rrom MIRAFLORES Note the level valley floor. The Western Andes appear in the background. (Junction of Tropical and Subtropical Zones. ) 1917.] Chapman, Distribution of Bird-life in Colombia. 127 largely into the Cauca Valley, a comparatively arid district; but the tropical life of the valley, on the other hand, has crossed the Subtropical Zone of the Western Andes and entered a favorable environment on its western slope. It should be noted, however, that on the treeless lower slopes of the eastern side of the Western Andes, the Tropical Zone extends to a greater altitude than it does on the humid western slopes. At San Antonio and Cresta de Gallo, for example, it practically reaches the divide from which one can almost see the Dagua basin, the upper margin of which lies not more than 1200 feet below. Only a small barrier, therefore, prevents the passage of species from the Cauca Valley to the upper Dagua Valley. On the western slopes of the Western Andes, the humid Tropical Zone does not attain so great an altitude as does the arid Tropical Zone on its eastern side, and the Subtropical Zone to be crossed is correspondingly wider. In spite of its isolation from other regions possessing similar character- istics, the Cauca Valley has given rise to but few geographical forms, and this fact in connection with its apparently limited life suggests that the existing fauna has been acquired at a comparatively recent date. It had occurred to me that possibly the floor of the Cauca Valley is an ancient lake-bed but with no geological evidence to support this theory, I had hesitated to advance it, but on re-reading Robert Blake White’s ‘ Notes on the Central Provinces of Colombia’ (Proc. R. G. §., V, 1883, p. 250) after the preceding observations had been written, I find this exceedingly inter- esting statement: “Directly to the eastward of this group [Supia and Tad6 Moros] of igneous rocks lies the great volcanic centre of Herveo, Tolima and Santa Isabel, and there, can be no doubt that the valley of the upper Cauca was for some time in the post-tertiary period converted into a lake, owing to the upheaval of the flanks of the volcanoes mentioned. However, their action also produced a fracture parallel to the opposing western cordillera, and the waters of the Cauca at last worked their way northwards and now run through one of the grandest ravines imaginable.” Here then, we have an apparently satisfactory explanation of the charac- ter of the Cauca Valley fauna, which appears indeed to be of post-Andean origin. List of Species and Subspecies which Characterize the Humid Cauca-Magdalena Fauna. Crax alberti Capito hypoleucus Amazona salvini Ramphastos citreolemus Pyrilia pyrilia Brachygalba salmoni Momotus subrufescens subrufescens Campephilus malherbi “ reconditus Thamnophilus nigriceps 128 ?Clytoctantes alixi Xenerpestes minlosi Gymnocichla nudipes sanctee-martz Todirostrum nigriceps Corapipo leucorrhoa Pachyrhamphus magdalene Heleodytes zonatus brevirostris Thryophilus leucotis Bulletin American Museum of Natural History. [Vol. XXXVI, Pheugopedius fasciato-ventris fasciato- ventris Tanagra concinna Tangara inornata Pheenicothraupis gutturalis Gymnostinops guatimozinus Cacicus vittellinus Tropical Zone Species which enter the Humid Cauca-Magdalena Fauna from the West. Crypturus ‘‘boucardi” (vide Scl. & Salv.) Leptotila cassini Oreopeleia veraguensis Creciscus albigularis Leucopternis semiplumbea Electron platyrhynchus minor Cheetura spinicauda fumosa Androdon equatorialis Polyerata amabilis Heliothrix barroti Anthracothorax nigricollis nigricollis Trogonurus strigilatus chionurus Capito maculicoronatus rubrilateralis Ramphastos swainsoni Notharcus pectoralis Nystalus radiatus Malacoptila panamensis panamensis Monasa pallescens sclateri Melanerpes pucherani pucherani Celeus loricatus mentalis Cniparchus hematogaster splendens Dysithamnus puncticeps puncticeps Myrmopagis fulviventris Microrhopias boucardi consobrina Microbates cinereiventris magdalene Myrmelastes immaculatus immaculatus Phenostictus macleannani macleannani Myrmeciza maculifer cassini Myrmeciza lemosticta nigricauda “ longipes panamensis Hylophylax nevia nevioides Hylopezus perspicillata perspicillata Xiphorhynchus lachrymosus alarum Dendrocolaptes sancti-thome sancti- thome Cnipodectes subbrunneus ?Rhynchocyclus marginatus marginatus Pipra velutina Manacus vitellinus milleri Lathria unirufa castaneotincta Lipaugus holerythrus holerythrus Cotinga nattereri Neochelidon tibialis Polioptila schistaceigula Thryophilus nigricapillus schotti Leucolepis phzocephalus pheocephalus Oryzoborus funereus Dacnis venusta fuliginata Dacnis egregia egregia Tanagra saturata Tangara larvata fanny Ramphocelus icteronotus Chlorothraupis olivacea Heterospingus xanthopygius Tachyphouus delatri Mitrospingus cassini Erythrothlypis salmoni Zarhynchus wagleri wagleri Cacicus uropygialis pacificus Cyanocorax affinis affinis Tropical Zone Species which enter the Humid Cauca-Magdalena Fauna from the East. Odontophorus guianensis marmoratus Aramides cajanea cajanea Phaétusa chloropoda Rhynchops nigra cinerascens Jacana nigra Phimosus berlepschi Jabiru mycteria Agamia agami 1Qnly species which are unknown in western Ecuador, and whose eastern origin is therefore undoubted, are here included. 1917.] Pilherodias pileatus Ixobrychus erythronotus Chauna chavaria Aratinga wagleri Psittacula conspicillata conspicillata Brotogeris jugularis Amazona amazonica Amazona ochrocephala panamensis Uropsalis lyra Glaucis hirsuta affinis Lepidopyga ccelina Chrysolampis elatus Pteroglossus torquatus nuchalis Galbula ruficauda ruficauda ?Jacamerops grandis (= anna?) Nonnula frontalis Chrysoptilus punctigula striatigularis Melanerpes rubricapillus rubricapillus Ceophleeus lineatus mesorhynchus Conopophaga castaneiceps Myrmeciza longipes boucardi Formicarius analis saturatus Synallaxis albescens albigularis ?Automolus pallidigularis pallidigularis Xiphorhynchus nanus nanus Campylorhamphus trochilirostris vene- zuelensis Fluvicola pica ?Euscarthmus septentrionalis Atalotriccus pilaris pilaris Inezia caudata intermedia Pipramorpha oleagina parca Phzomyias murina incompta ?Microtriccus brunneicapillus brunnei- capillus Chapman, Distribution of Bird-life in Colombia. 129 Myiozetetes similis columbianus Pitangus sulphuratus rufipennis « lictor ; Myiodynastes maculatus nobilis Myiobius fasciatus fasciatus Myiarchus ferox panamensis Tyrannus melancholichus satrapa Pipra erythrocephala erythrocephala Machaeropterus striolatus Manacus manacus abditivus + Tridoprocne albiventris Progne chalybea chalybea ?Mimus gilvus columbianus Donacobius atricapillus albovittatus Planesticus ignobilis ignobilis ‘ " albiventer ephippialis Cyclarhis flavipectus canticus Geothlypis equatorialis Pachysylvia flavipes flavipes Sporophila grisea grisea “ minuta minuta Tiaris olivacea pusilla “ bicolor omissa Saltator striatipectus striatipectus Arremonops conirostris conirostris * Emberizoides sphenurus Tanagra olivacea humilis & crassirostris crassirostris Thraupis palmarum melanoptera Ramphocelus dimidiatus dimidiatus Eucometis cristata cristata Hemithraupis guira guirina Schistochlamys atra Ostinops decumanus Icterus xanthornus xanthornus List of Species and Subspecies Known only from the Cauca Valley. Ortalis columbianus cauce Chamepelia rufipennis cauce Psittacula conspicillata cauce * Stenopsis cayennensis monticola Synallaxis pudica cauce Camptostoma caucze Pitangus sulphuratus caucee Planesticus ignobilis goodfellowi * Vireosylva chivi cauce Cyanocompsa cyanea cauce * Ammodramus savannarum cauce Ccereba, mexicana cauce Ramphocelus fammigerus 1 Represented in western Ecuador by M. m. melanochlamys and in southwestern Colombia by M. m. bangsi, but unknown on Pacific coast north of Patia River. 2 Represented in western Ecuador by A. c. chrysoma, but unknown on Pacific coast north of Patia River. 3 Occurs also in the arid upper Dagua Valley. 130 Bulletin American Museum of Natural History. [Vol. XXXVI, List of Species and Subspecies of the Colombian-Pacific Fauna which have entered the Cauca Valley. 4 Leptotila plumbeiceps ! Marila nationi ? Micrastur guerilla interstes * Rupornis magnirostris ruficauda * Piaya rutila gracilis * Piaya cayana nigricrissa 3 Ceophlceus lineatus mesorhynchus * Taraba transandeana transandeana * Myrmopagis schisticolor schisticolor * Xenops genibarbis littoralis 5 Rhynchocyclus sulphurescens asemus 3 Myiopagis viridicata accola * Myiobius barbatus atricaudus * Myiarchus tuberculifer nigriceps ° Manacus vitellinus vitellinus + Tanagra xanthogastra chocoensis ? Tanagra saturata 1 Ramphocelus flammigerus ! ‘List of Species and Subspecies Collected in the arid Upper Dagua Basin on the western slope of the Western Andes, showing how large a proportion of them has been derived from the Cauca Valley. Colinus cristatus leucotis ° Cheemepelia passerina nana ° Leptotila plumbeiceps ® Belonopterus cayennensis ® Psittacula conspicillata cauce ® Saucerottea saucerottei ® Hylocharis grayi * Chlorostilbon melanorhynchus ° Tapera nevia ® Crotophaga ani § Synallaxis albescens albigularis ® Todirostrum cinereum cinereum ° Myiopagis viridicata accola ° Myiobius fasciatus fasciatus Pyrocephalus rubineus heterurus ° Sayornis nigricans cineracea ” Myiarchus apicalis 5 Muscivora tyrannus © Troglodytes musculus striatulus ° Mimus gilvus tolimensis Planesticus ignobilis > Compsothlypis pitiayumi elegans § Geothlypis semiflava ° Cyanocompsa cyanea cauce > Sporophila grisea grisea ® é minuta minuta 5 gutturalis ° Tiaris olivacea pusilla ® Saltator striatipectus striatipectus ° Tanagra cyanocephala cyanocephala « saturata ® Tangara vitriolina > Ramphocelus dimidiatus dimidiatus § £ chrysonotus Tachyphonus luctuosus ® Molothrus bonariensis cabanisi ® “ The Caribbean Fauna.— The Caribbean coast of Colombia (except for a small section of the base of the central part of the Santa Marta group) 1 Unknown east of the Andes. 2 Known, elsewhere only from the vicinity of Lima, Peru. 3 Represented east of the Andes. 4 This species appears to be the only Tropical Zone species of the Pacific coast which has entered the Cauca Valley over the Western Andes. A representative, but strongly marked form, Manacus vitellinus milleri, occurs on the lower Cauca River at Puerto Valdivia, on the route by which this species might have been expected to enter the Cauca Valley. 5 Recorded from the Cauca Valley but not from the Pacific coast. 6 Recorded from Pacific coast and Cauca Valley. . 7 Recorded from Pacific coast but not from Cauca Valley. 1917,] Chapman, Distribution of Bird-life in Colombia. 131 from the mouth of the Sinu River to the end of the Goajira Peninsula is arid. Farther east the arid coastal strip extends into Venezuela but I have not at hand exact data from which to determine its eastern limits. In Colombia, in addition to the coastal district, this arid area occupies the valleys lying between the Santa Marta groups and the Eastern Andes and extends southward up the Magdalena Valley to the northern limits of the forested, humid Cauca-Magdalena Fauna at approximately the junction of the rivers from which this fauna takes its name. In those parts of this region with which we are familiar, the rainfall is said to be small and irregular. In consequence there are no forests, the Open Savannas supporting a scanty growth of acacias, mimosas, occasional cacti and other xerophytic forms. In the vicinity of rivers, marshes and bayous afford a home for numer- ous aquatic and palustrine forms, and near the coast there are vast expanses of red mangroves bordering the bodies of tidal water. These might indeed be set aside as constituting a small but distinct faunal area. We have, however, done no collecting in them and I am unable to treat of their fauna, but as a rule, aside from water birds, such regions contain few char- acteristic species. In addition to such widely distributed scrub and savanna-inhabiting species as Pyrocephalus r. heterurus, Muscivora tyrannus, Saltator olivascens, Thraupis cana cana, etc., all of which appear to be of eastern origin, this arid district possesses enough forms of its own to warrant, in my opinion, its being distinguished as a distinct faunal area, for which the name Caribbean seems appropriate. Some of the characteristic species of the Caribbean Fauna have crossed the forests of the Magdalena and reached the arid upper Magdalena Valley. Examples are Colinus cristatus, Brotogeris jugularis, Furnarius agnatus, Heleodytes m. bicolor and H. brevirostris, but such distinctive species as Psittacula spengeli, Picumnus cinnamomeus, and Synallaxis candeit do not appear to be known beyond the confines of the Caribbean Fauna. List of Species and Subspecies which Characterize the Caribbean Fauna. Ortalis garrula + ?Brotogeris jugularis Colinus cristatus decoratus Galbula ruficauda pallens Cheemepelia passerina albivitta Hypnelus ruficollis ruficollis. Chlorostilbon heberlini Chloronerpes xanthochlorus Aratinga xruginosa #ruginosa Chrysoptilus punctigula ujhelyi Psittacula spengeli Picumnus cinnamomeus 1 Recorded only from the Santa Marta district. 132 Bulletin American Museum of Natural History. [Vol. XXXVI, Thamnophilus canadensis pulchellus Heleodytes nuchalis nuchalis Microrhopias grisea intermedia "y zonatus brevirostris Furnarius agnatus Troglodytes musculus atopus Synallaxis cinnamomea fuscifrons Mimus gilvus columbianus e candei candei Thraupis glaucocolpa Empidochanes cabanisi Ccereba luteola luteola Heleodytes minor bicolor The Orinocan Fauna. — In our work at Buena Vista and Villavicencio, as well as at Florencia and La Morelia, we merely touched the western margin of a major faunal region which reaches to the eastern border of the continent. To define its minor divisions, even were data at hand, would take us far beyond the limits of our subject. It will, however, answer our present purpose to apply, at least provisionally, the term Orinocan Fauna to that. part of this region with which we are concerned and restrict our comparison of its life to that of the contiguous areas here under review. As has been stated elsewhere (see Expedition No. 7), the llanos or plains of the Meta, at Villavicencio, come directly to the base of the Andes. They bear no large forested areas, but the banks of streams are sometimes wooded, and where the streams overflow there are usually patches of forest growth. But the slopes of the outermost ridge of the Andes are heavily forested from base to summit. Our collecting about Villavicencio, was done in the open fields and in the strips of woodland. At Buena Vista, some 3000 feet higher, we col- lected only in the forests and about their borders. Nevertheless there was a surprising similarity in the arboreal tropical bird-life of these apparently quite different localities. Thus, of one hundred and fifty species (chiefly Passeres), forty-eight were found only at Buena Vista, fifty-eight only at Villavicencio, while fifty were common to both places. The Amazonian element is apparently quite as strongly shown at Vil- lavicencio as at Buena Vista; twenty-three Amazonian forms collected by Miller at La Morelia and Florencia, being also taken about Villavicencio, and eighteen at Buena Vista. It seems obvious, therefore, that the life of these two localities as it is represented in our collections, may be treated collectively. 7 Compared with that of the Amazonian Fauna of southeastern Colombia, it contains a much smaller number of pure Amazonian forms, and much larger number of wide-ranging species characteristic of the arid and semi- arid portions of northern South America. There are also several so-called Guianan species which have not as yet been recorded from upper Amazonia. Indicating doubtless a lower humidity, several species common to both the Florencia and Villavicencio districts are represented at the last-named Butrerin A. M.N. H. Vor. XXXVI, Piatra XXX. Near VILLAVICENCIO Junction of Llanos with Andes. (Tropical Zone; Orinocan Fauna.) Near VILLAVICENCIO Exit of Rio Guatequia from the Andes. (Tropical Zone; Orinocan Fauna.) 1917.| Chapman, Distribution of Bird-life in Colombia. 133 locality by paler forms. Examples are: Crypturus soui soui, Leptotila rufaxilla pallidipectus, Brachygalba fulviventris fulviventris and Synallaxis masta mesta. List of Species and Subspecies found at Villavicencio and Buena Vista which have not been Recorded from the Amazonian Fauna of Colombia or from eastern Ecuador and which, therefore, Characterize the westward extension of the Orinocan Fauna. Crypturus soui soui Todirostrum superciliare superciliare Crax alector Leptopogon superciliaris poliocephalus Colinus cristatus parvicristatus ie amaurocephalus Columba rufina Pheomyias murina incomta Leptotila rufaxilla pallidipectus Myiozetetes granadensis Stenopsis cayennensis cayennensis Pachyrhamphus cinnamomeus Brachygalba fulviventris fulviventris Thryophilus albipectus bogotensis Chelidoptera tenebrosa e rufalbus cumanensis Veniliornis fidelis Pheugopedius hypospodius Thamnophilus doliatus doliatus Troglodytes musculus neglectus es tenuipunctatus Pachysylvia flavipes flavipes Ramphocenus melanurus trinitatis Geothlypis sequinoctialis Cercomacra tyrannina tyrannina Oryzoborus angolensis Grallaria modesta a crassirostris crassirostris Synallaxis mcesta moesta Sporophila grisea grisea & gujanensis columbianus. Myiospiza cherriei Automolus turdinus Arremonops conirostris conirostris Selerurus albigularis albigularis Arremon axillaris Glyphorhynchus cuneatus Ceereba luteola luteola Dendroplex picus picus Tanagra aurea pileata Picolaptes albolineatus Tangara vitriolina Campylorhamphus trochilirostris vene- Thraupis episcopus leucoptera zuelensis Ramphocelus carbo unicolor Machetornis rixosa flavigularis Icterus xanthornus xanthornus Amazonian Fauna.— When we enter that vast territory lying east of the Andean system, we leave behind us the more distinctive features of the Colombian fauna. We stand now, as it were, on the shores of a great ocean of life which stretches far beyond the boundaries of Colombia. No adequate analysis of its affinities can be based on the study of a restricted part of it. The problem is as wide as the combined Amazonian and Orinocan basins. It should, therefore, be understood that in applying the term Amazonian Fauna to that portion of tropical Colombia included in the Amazonian drainage system, it is not intended to imply that we have here a definite faunal area, but that the faunal affinities of this southeastern section of the republic are with that wide-spreading region to which the name Amazonia is commonly, if somewhat vaguely, applied. 134 Bulletin American Museum of Natural History. [Vol. XXXVI, So far as I am aware, no attempt to map the faunal areas of Amazonia has as yet been made, but it is evident that in spite of its enormous extent the comparatively uniform climatic and topographic conditions which prevail throughout its forested portions have produced correspondingly uniform faunal characteristics. Such small, sedentary species as, for example, Dichrozona and Rhopoterpe range, unchanged from the base of the Andes through two thousand miles of forest to the lower Amazon, and the local differences in flora occasioned _by the wide variations between low and high water on the larger streams are often more pronounced than those existing in districts at opposite borders of the region. The distinctive feature of Amazonia is its forests, so well described by Wallace, Bates, Spruce and others. As elsewhere remarked, the northern limit of this Amazonian forest coincides approximately with the Guaviare River, beyond which lie the Llanos, but a heavily forested belt extends much farther north along the lower slopes of the Andes. According to Rice, as already quoted, the Sierra Chiribiquete lying south of the Guaviare, reaches an altitude of 2850 feet. Judging by our work at Buena Vista (alt. 4500 ft.), on the outermost spur of the Eastern Andes above Villa- vicencio, this is not sufficient altitude to produce a marked change in fauna. It remains, however, to be discovered to what extent the apparent isolation of these mountains and the possibly different environmental conditions they may offer, has modified the forms inhabiting them. As might be expected, there is a close resemblance between the bird-life of Amazonian Colombia, as it is revealed by Miller’s work at La Morelia and Florencia, and that of eastern Ecuador as that has been made known by the so-called ‘Napo’ specimens. It goes without saying that notwithstanding the large collections secured ‘by him in a limited time, Miller’s month at La Morelia and Florencia enabled him to get only enough material to show the faunal features of the region, and their Amazonian character is indicated by the appended list of, distributionally, the more significant species. List of the More Characteristic Amazonian Species collected at Florencia and La Morelia, in Southeastern Colombia. Penelope jacqtagu Creciscus enops Ortalis guttata Psophia napensis Pipile cumanensis Otus watsoni Opisthocomus hoazin Psittacula sclateri Anurolimnas castaneiceps Electron platyrhynchus pyrrholemus hauxwelli Hydropsalis climacocerca 1917,] Pheethornis fraterculus moorei Eutoxeres condamini Campylopterus obscurus equatorialis Agyrtrina fluviatilis Pharomacrus pavoninus Trogonurus bolivianus Chrysotrogon ramonianus Curucujus melanurus melanurus Capito aurovirens “auratus auratus Ramphastos cuvieri Pteroglossus pluricinctus . castanotis castanotis flavirostris flavirostris « humboldti Selinidera reinwardti Galbula tombacea tombacea Bucco capensis Argicus macrodactylus Malacoptila fusca Monasa flavirostris “ morphceus peruana nigrifrons Veniliornis ruficeps hematostigma Campephilus melanoleucus Dysithamnus capitalis capitalis ¢ ardesiacus ardesiacus Cercomacra sclateri Rhopoterpe torquata Dichrozona cincta Myrmeciza melanoceps “ « Chapman, Distribution of Bird-life in Colombia. 135 Hypocnemis cantator peruviana # hypoxantha Hylophylax lepidonota as neevia theresse Synallaxis rutila caquetensis Automolus infuscatus infuscatus Philydor ruficaudatus Ancistrops strigilatus Sclerurus brunneus Xiphorhynchus guttatoides insignis Dendrocolaptes sancti-thome radiolatus Ochthornis littoralis Todirostrum latirostris . Lophotriccus spicifer Pipra coronata Cirrhopipra fulicauda Macheropterus striolatus Chiroxiphia pareola napensis Lathria cinerea Lipaugus simplex Cephalopterus ornatus Myospiza aurifrons Paroaria gularis Tanagra chilensis e schranki f xanthogastra Ramphocelus nigrogularis Cissopis liveriana minor Gymnostinops yucares THE SUBTROPICAL ZONE AND ITS FAUNAS. The Subtropical Zone lies approximately between the altitudes of 5000 and 9000 feet, or from the average upper limits of the Tropical Zone to the lower limits of the Temperate Zone. Its inferior boundary is consequently as variable’ as the superior boundary of the Tropical Zone, and hence may vary from 4500 to 6500 feet in accordance with the conditions mentioned in outlining that zone. The altitude at which it meets the Temperate Zone is also governed by humidity. Apparently, however, it never exceeds 9500 feet, but in the absence of the forest which distinguishes the Subtropical Zone it descends to the level at which forest is encountered. Should the forest be entirely wanting, the zone, so far as birds are concerned, is also missing or but suggested by the occurrence of the few scrub-haunting species like Xanthoura, which are found in it. With them will be associated b 136 Bulletin American Museum of Natural History. [Vol. XXXVI, both Tropical Zone and even Temperate Zone species whose habits fit them to life in more or less open, bushy places. Such an association was found at Quetame (alt. 4500 ft.) in the Eastern Andes, where in the scanty growth of low trees bordering the Rio Negro and its tributaries we found Xanthoura yneas cyanodorsalis, Schistochlamys atra and Thraupis episcopus leucoptera, while in the immediately adjoining fields was Sturnella magna meridionalis, and in bordering hedgerows, Semimerula gigas gigas and Brachyspiza capensis peruviana. Usually, however, the Subtropical Zone is as clearly defined as the condi- tions to which it owes its characteristic features. It is pre-éminently a zone of forests, the product of the heavy rainfall and high degree of humidity prevailing at the altitude in which the Subtropical Zone is found. In the field, we termed it the ‘Cloud Zone,’ so closely does its lower border coincide with the height to which clouds descend on the mountain- sides. This term, however, may be also applied to the two upper zones, Temperate and Paramo, though cloud forests exist only in the Subtropical and Temperate Zones, the temperature of the Paramo Zone being evidently too low to permit of forest growth. Furthermore, the rainfall decreases as the altitude increases. The forests of the Subtropical Zone, particularly on windward slopes, present a luxuriance of growth not equalled even in the Tropical Zone. The lower zone produces nobler, taller trees (we saw nothing in the subtropics to approach the ceibas of the basal zone), but in profusion of undergrowth, of parasites and epiphytes which thrive in this region of clouds, the Sub- tropical Zone excels. It is the zone in which we found tree ferns attaining their maximum height of approximately fifty feet, in which a climbing bamboo grows in impenetrable tangles, in which orchids, bromelias and plants of similar habit occupy every available point of vantage, clustering thickly on the limbs and even trunks of trees; while every spot not occupied by some other form of plant-life, is cushioned with moss. From each leaf and limb water is constantly dripping, the bromelias are usually full to overflowing, the moss is like a saturated sponge. Even when, at intervals, the sun penetrates the clouds, the falling drops suggest a shower. In view of the altitude attained by the Subtropical Zone, far higher mountains are required to act as effective barriers to its extension across the ranges on the slopes of which it lies. This fact, in connection with the exceptional continuity of the Subtropical Zone forests, gives to the life of this zone a uniformity which, when one considers its length and the distance which its arms are sometimes separated, is surprising. Latitudinally, the Subtropical Zone extends from central Venezuela and Mexico at the north southward through Colombia to western Ecuador 1917] Chapman, Distribution of Bird-life in Colombia. 137 90° 80° Fig. 15. Distribution of the Cock-of-the Rock. A Subtropical Zone species represented in the West Andean Subtropical Fauna by one form (Rupicola peruviana sanguinolenta), and in the Central and Eastern Andes by a closely related but distinct species (R. peruviana aurea), the ranges of which are separated by the Temperate Zone of the higher intervening mountains. , 138 Bulletin American Museum ef Natural History. [Vol. XXXVI, on the Pacific slope of the Andes, and, on the Atlantic slope, to eastern Bolivia, where Miller and Boyle found its southern end between the Yungas of Cochabamba and the mountains above Santa Cruz. Throughout approximately 2500 miles of its length, that is, from Vene- zuela to Bolivia, the bird-life of this zone is so remarkably uniform in charac- ter that, as with the Amazonian and Orinocan elements in the Colombia fauna, one cannot arbitrarily segregate any part of it and study it satisfac- torily to the exclusion of the rest. Our field work is now chiefly directed toward the acquisition of adequate collections of ‘the birds and mammals of this exceptionally interesting zone of life, and until our work is completed, it will be impossible to speak of the zone as a whole or ofits altitudinal boundaries outside of Colombia. A study of the northward extension of the Subtropical Zone into Costa Rica reveals what may be termed a ‘zoological fault’ in Panama, while a compari- son of the Costa Rican representatives of Colombian species with those recorded from the Santa Marta group gives some significant results which will be presented later. Data are unfortunately lacking for a satisfactory comparison of the Subtropical Zone bird-life of the Andes with that of the mountains of south- ern Venezuela, the Guianas, and southeastern Brazil. The birds of the Subtropical Zone, as might be expected, are almost exclusively: forest-dwellers; the Green Jay (Xanthoura) is found about forest-borders and is one of the few Subtropical species inhabiting semi- arid places. The Dipper (Cinclus) and Torrent Duck (Merganetta), while confined to mountain streams, do not require that the shores be forested. But exceptions of this kind are rare. Tanagers are the most numerous in species as well as in individuals, the family Tanagride being the only one which, in Colombia, has more species in the Subtropical Zone than in the Tropical Zone. The Thrushes, while far less numerous in species, have almost as large a proportionate representation. Guans, Trogons, Capitos, Toucans, Dendrocolaptids, Cotingas and Wrens are all characteristic of the Subtropical Zone and, in the Colombian Andes, have about half as many species. in it as in the Tropical Zone. The Flycatchers are about one-half as numerous in the subtropics as in the tropics. Families of forest- inhabiting Tropical Zone birds which have a comparatively poor representa- tion in the Subtropical Zone! are the Pigeons, Parrots, Woodpeckers and Orioles, of which there are about four times as many species in the tropics as in the subtropics, and Formicarians, of which we found only seventeen 1It should be understood that these statements refer only to the results obtained by us in the Colombian Andes. Butvretin A. M.N. H. Vou. XXXVI, Puate XXXI. SUBTROPICAL FOREST Characteristic scene in the forest at San Antonio, summit of the Western Andes. Note the profusion of parasitic growth. (Subtropical Zone; West Andean Fauna.) 1917,] Chapman, Distribution of Bird-life in Colombia. 139 species in the subtropics as compared with eighty-two in the tropics. Many species of this last-named group, it is true, are scrub haunters, nevertheless, their abundance in heavy tropical forests is shown by the fact that Miller secured twenty-four species in the Amazonian Fauna in a month’s collecting. As might be expected, few true Finches inhabit the Subtropical Zone, but the Tanager-Finches of the genus Atlapetes are almost restricted to it. The Bucconide are represented by the only one of the eighteen tropical species which ranges upward to the subtropics; the Motmots by but one, while the Galbulidz appear to be wholly absent. Satisfactory data are wanting for a study of the origin of the bird-life of the Subtropical Zone. Our work in Colombia merely touched a portion of the vast area in which field studies and carefully labeled collections must be made before one can treat of the zone as a whole, and, as before stated, its life is too uniform to permit of conclusions being based on the study of a part. It appears, however, that so far as birds are concerned, the Tropical Zone differs from the Temperate and Paramo Zones in two important respects — one of which is the corollary of the other. First, the Subtropical Zone, latitudinally, does not extend beyond the limits of the Tropical Zone with which, when altitude permits, from Bolivia to Mexico, it is practically coterminous. Second, the Subtropical Zone, as a faunal area, does not descend to sea-level. Consequently it follows that the Subtropical Zone is always an altitudinal zone, and it also follows that its life, as a whole, was derived from the tropics. a To what extent the altitude of the Subtropical Zone is affected by lati- tude, I am not as yet prepared to say. Brachyspiza capensis peruviana, a species of the Subtropical and Temperate Zones, is found at sea-level on the Island of Curacao. It descends the Rio Negro east of Bogoté to Quetame (alt. 4500 ft.) but was wanting at Buena Vista. It, however, occurs at Caicara on the Orinoco. This species is really most characteristic of the Temperate Zone, but is also common in the arid subtropics. Its further descent to the Tropical Zone forms, therefore, an exception to the rule that the life of any zone is derived from a lower level. Assuming that species found in all three ranges must have had a com- mon point of origin south of the latitude where these ranges leave the Ecuadorian Andes, it is interesting to ascertain the results following their isolation. In most instances when there is appreciable racial variation, two races develop, one of which is found in the East Andean Fauna, the other in the West Andean Fauna. Where only one race is evolved it is generally found in the East Andean Fauna, while the West Andean form resembles that inhabiting western Ecuador. In some few instances the 140 Bulletin American Museum of Natural History. [Vol. XXXVI, form of each range is alike, all consequently differing from the Ecuadorian form in the same manner; indicating, therefore, that their characters have arisen independently and hence by parallelism. Examples are Ocreatus underwoodi underwoodi and Boissoneaua flavescens flavescens. Usually, however, specific cases of variation in altitudinal distribution show a tropical species ascending to the subtropics rather than the reverse. Thus, Leptotila verreauxi verreauxt is a form of the Tropical Zone in the Magdalena Valley and eastward, but L. v. occidentalis in the Cauca region is found in the Subtropical Zone. Colwmba subvinacea berlepschi of the tropics also grades into C. s. bogotensis of the subtropics. Lophotriccus squamecrista squamecrista inhabits the Subtropical Zone in all three ranges of the Colombian Andes, but in southwestern Colombia and western Ecua- dor’ L. s. minor is found at sea-level. Thryophilus nigricapillus connectens is a subtropical form of the tropical T. n. schotti; Cacicus uropygialis uro- pygialis is a subtropical form of C. n. pacificus. These are all intergrading forms and as such evidently illustrate how subtropical species are derived from tropical species. The inosculation of the upper limits of the Tropical Zone with the lower limits of the Subtropical Zone make it difficult to undér- stand how, under existing conditions, these intergrading, representative forms could become specifically separated. When, however, we attempt to explain the origin of the numerous wholly distinct species and genera now restricted to the Subtropical Zone we must take into consideration the profound climatic changes caused by elevation of the Andes, and by subsequent periods of glaciation which have produced wide fluctuations in zonal levels. We must also consider points of origin and subsequent zonal dispersion followed by complete geographic segregation from the parent form. For example, the basal, or tropical ancestor of Rupicola peruviana is possibly Rupicola rupicola; but the former has extended its range through the Subtropical Zone to Bolivia, while the latter is restricted to the Guianan region. Again, Pyroderus scutatus granadensis is now found in the Tropical Zone of the Eastern Andes of Colombia, but P. s. scutatus, the probable parent form, is found only in southeastern Brazil and eastern Paraguay. ‘But until we have a far more detailed knowledge of the geological history of the Andes and especially of the extent to which these mountains have been glaciated, we shall not be in a position to discuss satisfactorily the origin of its Subtropical Zone life. Meanwhile, as an ornithologist, I present further data in regard to its Colombian elements. The sharply defined topography of the Andean system in Colombia gives an equally clear definition to the zones of subtropical life which lie on its slopes. In all three Andean ranges, north of Popayan, they are widely separated below by Tropical Zone slopes and valleys, except at the Buuvetin A. M.N. H. Vou. XXXVI, Puate XXXII. Heart or top Crnrray ANDES View of the Rio Toché from above El Pie de San Juan. Torrent Ducks and Dippers were common on this stream. (Subtropical Zone; East Andean Fauna.) 1917] Chapman, Distribution of Bird-life in Colombia. 141 northern ends of the Western and Central Andes. In the Central Andes the Subtropical Zone of the western slope is prevented from coming in contact with the same zone on the eastern slope by the Temperate Zone until the northern end of the range is reached in Antioquia. In the Eastern Andes the zones of eastern and western slopes doubtless also meet where decreasing altitude near the northern end of the range permits them to cross the divide, and Miller’s work in the southern part of this range shows that the pass at Andalucia (alt. 7000 ft.) is in the heart of the Subtropical Zone which here, as in the greater part of the Western Andes, occupies both slopes and the crest of the range. Notwithstanding these, chiefly terminal, connections it is evident that the three branches of the Subtropical Zone in Colombia are sufficiently isolated from one another to become centers of local, adaptive radiation. The life of the Subtropical Zone as a whole, however, is remarkably uni- form, more than half of its characteristic species being distributed throughout its greater part. Thus of the 230 distinctively Subtropical Zone species found by us in Colombia, 121 are present either as unchanged or intergrading forms in all three Andean ranges. The remaining 109 species may be distributionally classified as follows: Peculiar to the Western Andes 31 species a“ “ “ Central “ 9 “ “ “ « Eastern a“ pA “ Common to the Western and Central Andes ! 14 ¢ is « « Eastern and Central Andes ? 33 - The facts expressed by this analysis ‘appear to require the recognition of at least two subdivisions of the Subtropical Zone of the Colombian Andes which I suggest be known as: 1. The West Andean Subtropical Fauna. 2. The East Andean Subtropical Fauna. Before treating of these minor divisions of the Subtropical Zone it seems desirable to give a list of the species we collected in it. Birds of the Subtropical Zone. Family Tinamide Family Cracide Tinamus tao Penelope cristata Nothocercus bonapartei & perspicax f intercedens (vide Hellmayr) Aburria aburri 1 Chiefly the western slope. 2 Chiefly the eastern slope. 142 Bulletin American Museum of Natural History. Chamepetes goudoti goudoti + Family Odontophoride Odontophorus hyperythrus « strophium Family Columbide Columba albilinea albilinea * aS subvinacea bogotensis Leptotila verreauxi occidentalis Oreopeleia montana bourcieri " linearis linearis Family Anatide Merganetta columbiana Family Bubonide Glaucidium jardini Family Psittactde Ognorhynchus icterotis Pyrrhura calliptera “. souancei Amazona mercenaria Pionus chalcopterus Hapalopsittaca amazonina Family Momotide Momotus equatorialis equatorialis Family Caprimulgide Lurocalis rufiventris Stenopsis ruficervix 1 Family Trochilide Doryfera ludovicie ludoviciz Pheethornis guyi emiliz “ syrmatophorus syrmato- phorus “ Campylopterus falcatus Agyrtria viridiceps Uranomitra francize Thalurania colombica colombica i fannyi verticeps Colibri cyanotus “ — iolata Simonula berlepschi syrmatophorus columbianus Phaiolaima rubinoides rubinoides en ‘ eequatorialis Heliodoxa leadbeateri Helianthea torquata ! S coeligena columbiana « — ferruginea Lafresnayea lafresnayei ! Ensifera ensifera ensifera ! Boissoneaua flavescens ! Vestipedes aurelize aureliz “ caucensis Ocreatus underwoodi underwoodi Urosticte benjamini benjamini Adelomyia melanogenys melanogenys “ « cervina ! Heliangelus exortis 1 Cyanolesbia kingi kingi . mocoa mocoa emma “ coelestes Schistes geoffroyi « ~ albogularis Cheetocercus mulsanti a heliodor Klais guimeti Family Trogonide Pharomacrus antisiensis - auriceps Trogon personatus Trogonurus collaris Family Capitonide Eubucco richardsoni granadensis “_ bourcieri bourcieri fs £ occidentalis Semnornis ramphastinus Family Ramphastide Ramphastos ambiguus ambiguus Andigena nigrirostris nigrirostris ¢ e spilorhynchus occidentalis Aulacorhynchus albivitta albivitta & “« — pheolemus griseigularis hematopygius & “ “ “ 1 Ranging upward to the Temperate Zone. [Vol. XXXVI, 1917.] Family Bucconide Malacoptila mystacalis Family Picide Chloronerpes rubiginosus gularis “ s buenavista Melanerpes flavigula Veniliornis oleaginus fumigatus & s aureus ¢ dignus Campephilus pollens 1 Family Hylactide Scytalopus micropterus micropterus Family Conopophagide Conopophaga castaneiceps castaneiceps Family Formicartide Thamnophilus unicolor multistriatus Dysithamnus semicinereus semicinereus “ extremus ? Myrmopagis schisticolor schisticolor oe ss interior Drymophila caudata caudata Terenura callinota Pyriglena picea Formicarius rufipectus carrikeri Chamezza turdina Grallaria alleni we hypoleuca a ruficapilla ruficapilla Grallaricula costaricensis - nana “ cucullata Family Dendrocolaptide Lochmias sororia Synallaxis azaree media! . pudica pudica a unirufa ! Siptornis antisiensis “ erythrops griseigularis # striaticollis Pseudocolaptes boissonneauti boisson- neauti Automolus ignobilis Chapman, Distribution of Bird-life in Colombia. 143 Automolus holostictus Philydor montanus striaticollis Thripadectes flammulatus virgaticeps sclateri Xenicopsis subalaris subalaris “ . mentalis Sclerurus mexicanus obscurior Margarornis perlata + “7 stellata Premnornis guttata Xenops rutilus heterurus Premnoplex brunnescens brunnescens Dendrocinela tyrannina tyrannina Xiphorhynchus triangularis Xiphocolaptes promeropirhynchus ! Picolaptes lacrymiger lacrymiger 1 : warscewiczi Campylorhamphus pucherani Family Tyrannide Ochthceca cinnamomeiventris sa gratiosa + Mecocerculus poecilocercus Platytriccus flavigularis Craspedoprion fulvipectus Peecilotriccus ruficeps ruficeps . “ — rufigene Euscarthmus granadensis ! Pseudotriccus annectens $ pelzeni pelzeni Ceenotriecus ruficeps ruficeps + Lophotriccus squamecrista squameecrista Hapalocercus acutipennis Pogonotriccus peecilotis e opthalmicus Oreotriccus plumbeiceps Mionectes striaticollis poliocephalus Leptopogon erythrops Phyllomyias griseiceps griseiceps ff cauce Acrochordopus zeledoni Tyranniscus nigricapillus nigricapillus } « cinereiceps ¢ Elenia chiriquensis chiriquensis Elenia pudica brachyptera « “ pudica Conopias cinchoneti 1 Ranging upward to the Temperate Zone. 2 Found also in the Cauca Valley. 144 Bulletin American Museum of Natural History. Myiodynastes chrysocephalus minor Hirundinea sclateri Myiobius flavicans villosus cinnamomeus pyrrhopterus pulcher pulcher " « ~ bellus Myiotriccus ornatus ornatus Empidochanes peecilurus Myiochanes ardosiacus ardosiacus Myiarchus cephalotes Family Pipride Pipra leucocilla minimus « a“ coracina isidorei isidorei Chloropipo flavicapilla Piprites tschudi Masius chrysopterus “ ~ corunulatus Family Cotingide Pachyrhamphus versicolor e dorsalis Lathria fuscocinerea fuscocinerea “ ~ eryptolopha Attila brasiliensis parambze Rupicola peruviana aurea 4 “ sanguinolenta “ _ Stictornis cinctus Euchlornis riefferi riefferi “ “ ~ occidentalis lubomirski aureipectus (vide Scl.& Salv.) Heliochera rufaxilla Pyroderus scutatus granadensis a occidentalis Cephalopterus penduliger «“ Family Hirundinide Pygochelidon cyanoleuca Family Troglodytide Cinnicerthia olivascens ! Odontorhynchus branicki Thryophilus nigricapillus connectens Pheugopedius spadix e mystacalis mystacalis [Vol. XXXVI, Pheugopedius mystacalis amaurogaster « sclateri Troglodytes solstitialis pallidipectus ! Henicorhina prostheleuca eucharis leucophrys guttata © = brunneiceps Leucolepis dichrous Family Cinclide Cinclus leuconotus Family Turdide Myiadestes ralloides venezuelensis Planesticus serranus i fuscobrunneus leucops if cauce Catharus birchalli - phzopleurus € dryas Family Vireonide “ Vireosylva josephe josephe Pachysylvia semibrunnea Cyclarhis nigrirostris Family Mniotiltide Myioborus verticalis verticalis Basileuterus cinereicollis cabanisi tristriatus tristriatus coronatus + Family Catamblyrhynchide Catamblyrhynchus diadema ? Family Fringillide Sporophila luctuosa Saltator atripennis atripennis " caniceps Spinus xanthogaster Brachyspiza capensis peruviana * Pseudochloris citrina antioquize Lysurus castaneiceps Atlapetes flaviceps eS fusco-olivaceus gutturalis gutturalis latinuchus latinuchus “ “« 1 Ranging upward to the Temperate Zone. 1917.) Chapman, Distribution of Bird-life in Colombia. 145 Atlapetes latinuchus elzoprorus io albifrenatus Bs crassus « semirufus Buarremon brunneinuchus " atricapillus Family Cerebide Diglossa sittoides similis ~ albilateralis 1 personata ! cryptorhis Diglossopis cerulescens czrulescens Conirostrum albifrons y atrocyaneum “ “ Family Tanagride Chlorophonia pretrei Tanagra xanthogastra brevirostris Chlorochrysa, calliparsea bourcieri e pheenicotis nitidissima Pipreidea melanota venezuelensis Procnopis vassori ! Tangara guttata tolime « « ~ bogotensis “ “« — rufigula aurulenta aurulenta si Be occidentalis «“ ~ icterocephala «“ gyroloides gyroloides o bs catharine a ¢ bangsi « — nigroviridis nigroviridis « cyaneicollis ceruleocephala a ‘3 granadensis & ruficervix ruficervix as labradorides’ melanotis parzudaki Tangara venusta i atricapilla Iridosornis porphyrocephala Buthraupis cucullata cucullata + . edwardsi s melanochlamys = aureocincta Compsocoma somptuosa victorini « od antioquiz - as cyanoptera . notabilis Dubusia temiata 1 Sporathraupis cyanocephala auticrissa, Calochetes coccineus Piranga testacea testacea « ~~ rubriceps ! Phoenicothraupis cristata Tachyphonus rufus Creurgops verticalis Chlorospingus albitempora nigriceps : flavipectus e canigularis . flavigularis flavigularis . semifuscus Hemispingus frontalis oleagineus € atropileus ! # melanotis Oreothraupis arremonops Family Icteride Ostinops salmoni “ alfredi sincipitalis “ « ~ neglectus Cacicus uropygialis uropygialis Icterus giraudi Hypopyrrhus pyrohypogaster Family Corvide Xanthoura yneas galeatus ay “ — eyanodorsalis THE FAUNAS OF THE SUBTROPICAL ZONE, The West Andean Subtropical Fauna.— The West Andean Subtropical Fauna occupies that part of the Subtropical Zone which extends along the Western Andes from southern Ecuador, or at the northern end of the arid 1 Ranging upward to the Temperate Zone. 146 Bulletin American Museum of Natural History. [Vol. XXXVI, Peruvian region, northward through Colombia. To it should doubtless be added the subtropical western slope of the Central Andes which, while far from possessing all the forms that characterize. the West Andean Subtropical Fauna, has a closer affinity with that fauna than it has with the East Andean Subtropical Fauna. This is indicated by the presence of such distinctive West Andean species as Geotrygon bourciert, Formicarius rufipectus carrikert, Chlorochrysa nitidissima and Ostinops salmont. Furthermore, with species which have representative races in the West Andean Fauna and the East Andean Fauna, the form of the western slope of the Central Andes usually agrees with that of the Western Andes, that of the eastern slope with that of the Eastern Andes. Taking the Central Andes, as a whole, however, the East Andean Subtropical element is much stronger than that received from the Western Andes, a fact obviously attributable to existing topography and to the humid connection at the head of the Magdalena Valley. On the Pacific slopes of the Andes this faunal belt stretches continuously from its southern end to northern Colombia. Whether it exists on the summit of the Baudo range unfortunately is not known. It reappears in dilute form on the crests of the higher mountains of eastern Panama and Costa Rica, and its influence extends even to southern Mexico. Its appar- ent absence between the higher portions of eastern and western Panama has already been mentioned and will be referred to in detail later. While apparently always present on the western slope of the Western Andes, it is developed on the eastern slope of this range only above an eleva- tion of 6500 feet; the altitude of condensation, as explained in writing of the Tropical Zone, being higher on the eastern than on the western slope of this range. Its forests stretch, apparently without a break, along the western slopes of the Central Andes above the Cauca Valley, are wanting in southern Antioquia, but reappear in the more northern part of that department. Here the Western and Central Andes are separated only by the Cauca River from opposite banks of which they respectively arise. At this point the subtropical forests of these ranges are within a short distance of one another. Doubtless for this reason forms elsewhere restricted to one range may in some few instances here be found in both. Further south, these ranges are separated by the increasingly wide Cauca Valley until one reaches the ‘ knot’ of Popayan, but although this attains the altitude of the Subtropical Zone it is lacking in the heavy forests which characterize it and the West Andean Subtropical Fauna is, therefore, not connected here with its Central Andean branch, a fact which presumably accounts for the comparatively small number of West Andean forms found in the Central Andes. The distinguishing characteristics of the West Andean Subtropical 1917] Chapman, Distribution of Bird-life in Colombia. 147 Fauna are those of its zone. It cannot be said to have arid sections. When the humidity falls below the point required to produce the forests which’ Subtropical Zone species require, the zone practically disappears and its place is taken by an upward extension of the Tropical Zone and a downward extension of the Temperate Zone. Aside from representative forms of species of general Subtropical distri- 90° 80° = 70° 60° Fig. 16. Distribution of Formicarius rufipectus. A species of the West Andean Subtropical Fauna which occurs in eastern Panama, western Panama and Costa Rica but is unknown in the inter- vening Tropical Zone. It is represented in the Subtropical Zone of eastern Ecuador by F. thoracicus. 1. Formicarius rufipectus rufipectus. 2. F.r. carrikeri. bution in Colombia, we have taken thirty-one species peculiar to the West Andean Fauna, but only twenty-two peculiar to the East Andean Fauna. Nevertheless, as we have seen, the West Andean Fauna occupies a compara- tively restricted, isolated area, and at the south, whence it seems evident subtropical life was derived, it is entirely cut off from corresponding areas. . 148 Bulletin American Museum of Natural History. [Vol. XXXVI, ‘The East Andean Fauna, however, is but the more northern portion of a ‘belt of forest which extends southward to central Bolivia. Expressed in miles, the West Andean Fauna measures from north to south about 850 miles, the East. Andean, with its subdivisions, nearly 2500 miles. Something more than isolation is required to explain the presence in the Western Andes of so large a number of species in proportion to its area. The Subtropical Zone of the Central Andes, with its narrow basal connection with the East Andean Fauna and long peninsula-like projection, is almost insular in its isolation; but it has few indigenous species. If this fact is attributable to its inaccessibility, one may reply that the subtropics of the Western Andes are even less accessible. It therefore seems reasonable for us to believe that the Subtropical Zone of the Western Andes, as well as the Tropical Zone at its western base, received its life in part from what is now upper Amazonia, before the Andes were sufficiently elevated to act as an effective barrier between the Subtropical Zones lying on their eastern and western slopes. Since, however, this was obviously at a-later date than that.at which the Tropical Zones of the eastern and western slopes were separated, there is a closer relation between the life of the upper than between that of the lower zones. List of Species and Subspecies which Characterize the West Andean Subtropical Fauna. ' Nothocercus intercedens Veniliornis oleaginus aureus 1 Penelope perspicax !' Formicarius rufipectus carrikeri + Geotrygon bourcieri ! Grallaricula costaricensis Leptotila verreauxi occidentalis + Synallaxis azare media * Pheethornis syrmatophorus syrmato- Siptornis erythrops griseigularis phorus Automolus ignobilis ro Agyrtria viridiceps Xenicopsis subalaris subalaris 2 Thalurania fannyi verticeps Margarornis stellata Phaiolaima rubinoides sequatorialis Picolaptes warscewiczi Helianthea cceligena ferruginea Campylorhynchus pucherani Vestipedes aureliee caucensis Peecilotriccus ruficeps rufigene # Adelomyia melanogenys cervina. Pseudotriccus annectens Cyanolesbia emme Elenia pudica brachyptera ! ga coelestes. Myiarchus cephalotes Schistes albogularis Masius corunulatus Eubucco bourcieri occidentalis + Attila brasiliensis parambe: Semnornis rhamphastinus Rupicola peruviana sanguinolenta Andigena, nigrirostris occidentalis, Euchlornis riefferi occidentalis — Aulacorhynchus albivittatus pheolemus Pyroderus scutatus occidentalis Chloronerpes rubiginosus gularis ? Thryophilus nigricapillus connectens 1 Found also on the western slope of the Central Andes. 2 Found also on both slopes of the Central Andes. Bouuietin A. M.N.H. Vov. XXXVI, Prats XXXII. Rio Negro Canon NEAR MONTEREDONDO A scene near the western limit of arborescent vegetation on the Rio Negro. Trail at the left. (Junction of Tropical and Subtropical Zones.) 1917.) Chapman, Distribution of Bird-life in Colombia. 149 Pheugopedius spadix Tangara gyroloides bangsi mystacalis mystacalis ? Tridosornis porphyrocephala 2 Henicorhina prostheleuca eucharis Buthraupis cucullata cucullata Leucolepis dichrous 2 5 edwardsi Planesticus fuscobrunneus ! a melanochlamys Saltator atripennis atripennis a aureocincta Lysurus castaneiceps Compsocoma somptuosa cyanoptera Atlapetes latinuchus latinuchus e notabilis . crassus Pheenicothraupis cristata Diglossa cryptorhis Chlorospingus flavigularis marginatus Chlorochrysa phcenicotis « semifuscus nitidissima 1 Oreothraupis arremonops Tangara rufigula Ostinops salmoni 2 af aurulenta occidentalis 1 Cyanolyca pulchra « icterocephalus The East Andean Subtropical Fauna.— The Subtropical Zone in the Eastern Andes, like the Tropical Zone at their eastern base, is merely a part of a much larger region. Our work in Colombia, therefore, can be con- sidered merely as a contribution to the general subject. Even with this limitation it must be confessed that our explorations covered so small a part of the range that we are sadly lacking in detailed information concern- ing its altitude, the distribution of its forests, and its bird-life. Miller’s section across the Eastern Andes from the upper Magdalena Valley to the Caquetaé region (see Expedition No. 5) showed, as has been elsewhere stated, that the Andalucia Pass has an altitude of only 7000 feet, and that both slopes of the range are here forested, the western down to an altitude of 3000 feet, the eastern continuously. In other words, at this point, the Subtropical Zone occupies both eastern and western slopes as well as the crest of the range. It is at this point, and possibly also further south, that the subtropical life of the Amazonian slopes of the Andes enters the upper Magdalena and thus gains access to the Central Andes. The Subtropical Zone evidently extends nearly to the northern end of the Eastern Andes in Colombia, where Cerro Pintado has an elevation of about 8600 feet. Here it is separated from the Santa Marta group by the valley of the Rio Cesar. The zodlogical as well as geological evidence indicates that there has been no connection between these mountains, and the Santa Martan portion of the Subtropical Zone should doubtless rank as a faunal area. It is also probable that the Venezuelan branch of the Subtropical Zone is deserving of recognition as a distinct faunal area, though it has close relations with the East Andean Fauna of Colombia. 1 Found also on the western slope of the Central Andes. 2 Found also on both slopes of the Central Andes. 150 Bulletin American Museum of Natural History. [Vol. XXXVI, As stated in outlining the boundaries of the West Andean Fauna, the subtropical portion of the Central Andes, as a whole, is more closely related to that of the Eastern, than to. that of the Western Andes, but the West Andean element on its western slopes is sufficiently strong to make it seem desirable to place that slope in the West Andean Fauna, while the eastern slope may unquestionably be placed in the East Andean Fauna. In the latitude of Bogotd, the Eastern Andes have a width of about one hundred miles, and it is not improbable that the subtropical eastern slopes may have forms either not found on the western slopes or ‘representative of them. Ostinops alfredi neglectus and O. a. sincipitalis, and Xanthoura yncas cyanodorsalis and X. y. galeatus, are evidently representative races of this kind, the first named of each species being found on the eastern, the second, on the western subtropical slope of the range. As stated in the itinerary of our expedition (No. 7) to the Bogotd region, our route on the eastern slope of the range between Bogoté and Villavicencio did not take us into the humid subtropics and we are not therefore in a position to compare the subtropical life of both sides of the range. The East Andean Fauna of Colombia has but few species which are restricted to it; its practical physical identity with those portions of the Subtropical Zone to the north and south prevent that isolation which renders cumulative the effects of environment on an organism. While none of the species in the appended list are found in the Western Andes, most of them range beyond the Colombian portion of the Subtropical Zone. List of Species and Subspecies which Characterize the East Andean Subtropical Fauna. Nothocercus bonapartei ! Vestipedes aureliz qurelise Penelope cristata ! Adelomyia melanogenys melanogenys Odontophorus strophium Cyanolesbia kingi kingi Geotrygon linearis linearis ! mocoa mocoa Ognorhynchus icterotis * Schistes geoffroyi Pyrrhura calliptera Eubucco richardsoni granadensis as souancei 2 “ ~ bourcieri bourcieri 1 Hapalopsittaca amazonina Ramphastos ambiguus ambiguus ! Lurocalis rufiventris Andigena nigrirostris nigrirostris Pheethornis syrmatophorus columbianus Aulacorhynchus albivitta albivitta } Phaiolaima rubinoides rubinoides Chloronerpes rubiginosus buenaviste Heliodoxa leadbeateri Veniliornis oleaginus fumigatus Helianthea cceligena columbiana Terenura callinota Lafresnayea lafresnayi Pyriglena picea ! 1 Found also on the eastern slopes of the Central Andes. 2 Found also on both slopes of the Central Andes. Butietin A. M. N. H. Vor. XXXVI, Prater XXXIV. PrimevaL Forest ar BuENA Vista Photographed with a 14-inch lens at a distance of about five hundred yards. The eastern slope of the Eastern Andes is here heavily forested. (Tropical Zone; Orinocan Fauna.) Forest [Interior at Buena Vista A detail of the preceding picture. (Tropical Zone; Orinocan Fauna.) 1917.] Chamezza, turdina 2 Grallaria hypoleuca ! Grallaricula nana 2 Lochmias sororia 2 Siptornis antisiensis ! a striaticollis 1 Xenicopsis subalaris mentalis 1 Platytriccus flavigularis ! Pseudotriccus pelzelni pelzelni Hapalocercus acutipennis 2 Leptopogon erythrops Phyllomyias griseiceps griseiceps Acrochordopus zeledoni Elenia pudica pudica ? Conopias cinchoneti ? Hirundinea sclateri Pipra leucocilla coracina “ isidorei isidorei Piprites tschudi 2 Masius chrysopterus ! Lathria fuscocinerea fuscocinerea # « ~ eryptolopha Rupicola peruviana aurea ? Euchlornis riefferi riefferi ? “ aureipectus 2 Pyroderus scutatus granadensis ! Pheugopedius sclateri 2 & mystacalis amaurogaster Chapman, Distribution of Bird-life in Colombia. 151 Planesticus serranus ne leucops ! Catharus birchalli « dryas ! Basileuterus cinereicollis cinereicollis Sporophila luctuosa 2 Saltator atripennis caniceps Atlapetes albifrenatus . semirufus Chlorochrysa calliparza bourcieri } Pipridea melanota venezuelensis Tangara guttata bogotensis Tangara aurulenta aurulenta ! « gyroloides catharine “ ¢. ceruleocephala, ! s melanotis 1 « parzudaki 3 Compsocoma somptuosa victorini + Calochetes coccineus Chlorospingus albitempora nigriceps ” - flavipectus as flavigularis flavigularis Hemispingus melanotis £ superciliaris superciliaris Ostinops alfredi sincipitalis = « ~ neglectus Xanthoura yneas cyanodorsalis Tue CENTRAL AMERICAN EXTENSION OF THE SUBTROPICAL ZONE AND THE Panama ‘Fautt,’ The range of several Subtropical Zone species extends as far north as Mexico. Conspicuous among them is the Towhee-Tanager, Buarremon brun- neinuchus which is found from southeastern Peru to the mountains of the State of Vera Cruz, a distance of over 2500 miles, doubtless a more extensive distribution than is shown by any other subtropical species. Aflapetes gutturalis, a common Colombian subtropical bird, is found as far north as Guatemala, but, as a rule, South American subtropical species do not go further north than Costa Rica. Here, apparently, judging from the data supplied by Mr. Carriker’s valuable work, they sometimes descend to lower 1 Found also on the western slope of the Central Andes. 2 Found also on both slopes of the Central Andes. Bulletin American Museum of Natural History. [Vol. XXXVI, 90° sf B Eg ‘MA 4 Pa 4 yee, | r° i eer ee M CO dvénrsn er, J] | HONDURAS, ‘Le Rea ic: : "ot are a e Set 9 Evty ed < “Be € o ) 10" RAT VT egal 2d 90° = Fig. 17. Distribution of Buarremon brunneinuchus. A common species of the Subtropical Zone which ranges from southeastern Peru to Mexico but is unknown in the area between eastern and western 3 Panama. 1917.] Chapman, Distribution of Bird-life in Colombia. 153 levels than we have found them in Colombia, possibly in response to local conditions peculiar to that miniature transcontinental republic. The close relation existing between Colombia and Costa Rican subtropical bird-life is most striking. In many cases, the same species is common to both countries. In others, slightly differentiated races of the same species occur, and in others still, unquestionably representative, but now distinct, non- intergrading species are found. A list is appended of the species which may properly be considered as belonging to these classes. It shows that some sixty-odd species of Colombian subtropical birds are present or represented in Costa Rica, and usually also in western Panama. About twenty-three of these are found on the subtropical crests of the mountains of eastern Panama, but between these localities these species are not known to occur. An orographic map shows that the Subtropical Zone of the northern end of the Western Andes is separated from the subtropical crests of the mountains on the Colombian-Panama boundary by a tropical area of approximately seventy-five miles in width; while the Subtropical Zone of eastern Panama is separated from the same zone in western Panama by not less than three hundred and fifty miles. Doubtless some of the species in the following list will be found in the intervening area. Thus far, however, not one of the species included in the appended table has been recorded from between eastern and western Panama and more than two-thirds of them are unknown from between the northern end of the Western Andes and western Panama. In other words, there is an apparent hiatus in their range of somewhat over four hundred miles. This statement is based not alone on published data, but on the examination of numerous specimens, including those contained in Goldman’s fine collection from the Canal Zone and adjoining territory and eastern Panama which, through the courtesy of the Biological Survey, I have been permitted to see. Goldman reached the subtropical Zone on Mt. Pirri' and found there most of the subtropical species listed under Eastern Panama in the subjoined table. None of these, however, was taken elsewhere, though in his work in and near the Zone he collected on Cerro Azul at an altitude of 3000 feet. Anthony and Ball, of the American Museum Panama Expedition of 1915, discovered a number of subtropical species (including the distinct Scytalopus panamensts) on the crest of Mt. Tacarcuna, at an elevation of about 4500 feet. But Richardson’s extensive collections from the Tropical Zone of eastern Panama, as might be expected, contained none of the species which characterize the higher, subtropical altitudes. If then these sib- tropical species are not found in the tropics immediately below the zones in which they occur, it is of course not to be expected that they will occur in the tropics elsewhere. Consequently, the absence of subtropical altitudes 154 Bulletin American Museum of Natural History. [Vol. XXXVI, between those of eastern and western Panama is in itself evidence of the absence of subtropical species in this area. If the facts then are as stated, it remains for us to explain what we may term this Panama ‘fault’ in the Colombian-Costa Rica subtropical stratum of life. To one familiar with the influences governing the distribu- tion of birds, the occurrence of so large a number of species, including many et 77 BRITISH 1 HQNOURAS . Fig. 18. Distribution of Allapetes gutturalis. A common species of the Subtropical Zone in Colombia (A. g. gutturalis) and from western Panama to Guat la (A. g. br ) but which is not known in the intervening area. of sedentary habit, at two such widely separated localities, is conclusive proof that the localities themselves were connected. The fact that they are joined by the comparatively low land between them has no bearing on the case. The barriers which confine subtropical species to their zone are so effective that these birds could no more cross the Tropical Zone separating the areas they inhabit, than they could an intervening sea. Some of these 1917.] Chapman, Distribution of Bird-life in Colombia. ' 155 species, indeed, are so eminently terrestrial that they rarely fly more than a few yards, and a continuous flight of several hundred miles would for them be impossible. It is true that birds populate remote oceanic islands, but we do not find among island forms such sedentary species as Formicarius rufipectus, Gral- laricula flavirostris, and Siptornis erythrops, etc., which in but slightly differ- entiated races are common to the Subtropical Zones of Colombia and Costa Rica. Nor do we find species establishing themselves in regions which are already occupied. Early arrivals on oceanic islands encounter no opposi- tion, but the mainland offers no such favorable opportunities for settlement. The available space is taken and the emigration of even a score of species from one mainland home to settle in another mainland home, at a distance of several hundred miles, is an unknown phenomenon in the distribution of bird-life. In this connection it is quite to the point for us to compare the Subtropi- cal Zone bird-life of the Eastern Andes with that of the Santa Marta moun- tains. The subtropical portions of these mountains are separated by not more than forty miles; nevertheless, of the one hundred and ninety-eight species received by us from the Subtropical Zone of the Eastern Andes only fifty-odd have been recorded as present or represented in the Santa Marta group. On the other hand, as we have seen, some sixty species of Colombian subtropical birds are present or represented in Costa Rica, though here their ranges are separated by four hundred miles, or ten times as great a distance as that lying between the northern end of the Eastern Andes and the subtropical Santa Marta slopes. Geologists are, however, agreed that the Santa Marta mass is of independent origin and has never been connected with the Andean system. Zodgeographically, its life zones “above the Tropical Zone are, therefore, as much islands as though they stood out in the Caribbean Sea. The absence from these zones of many common, widely distributed species is doubtless to be accounted for by the fact that they have never been connected with the corresponding zones in the Andean system. If, therefore, so small a proportion of East Andean subtropical species have crossed the forty miles lying between that range and the subtropical slopes of Santa Marta, we certainly cannot account for the presence in a continental area, distant four hundred miles, of a larger proportion of West Andean subtropical species under the assumption that they have reached Costa Rica fortuitously or by emigration. To my mind, the existence respectively in Colombia, Costa Rica and eastern Panama of these two ends and a fragment of the Subtropical Zone, is conclusive proof of a former physical connection of the areas concerned, 156 Bulletin American Museum of Natural Hsitory. [Vol. XXXVI, at which time this zone stretched’ more or less continuously from north- western Colombia through Panama to Costa Rica. If this be true, it follows that the mountain system of Panama must have had an elevation of not less than five thousand feet. The ornithologi- cal evidence also indicates that this range could not have had a greater elevation than nine thousand feet. This statement anticipates a treatment of the Temperate Zone, which, as will be shown, is present at the northern end of the Western Andes, above an altitude of 9000 feet. The higher mountains of western Panama and Costa Rica also exceed this altitude but there is no such close relation between the bird-life of the Colombian and Costa Rican Temperate Zones as exists between that of their Subtropical Zones. The Temperate Zone avifauna of Costa Rica, so far as one can trace its origin, appears to have been derived largely from the north, while that of Colombia has been derived chiefly from the south. We apparently are therefore justified in assuming that there has been no actual connection between the Temperate Zones of these two regions and this, in turn, implies that the intervening mountains have not exceeded an elevation of 9000 feet. ie Subsidence, of which there is abundant geological evidence, and erosion are the factors which have doubtless brought the mountain system of Panama below the subtropical level, except at its eastern and’ western extremities. In the effort to picture to ourselves this region as it existed when, as we believe, the Subtropical Zone extended from Colombia to Costa Rica, we may try also to imagine the fate of those individuals of representa- tive subtropical species which inhabited that part of the zone which sub- sequently disappeared. As their home gradually contracted in extent, only two lines of retreat were open to them. They could go downward into the Tropical Zone, or they could go horizontally toward that part of their zone which had not sunk into the zone below. Species of the Tropical Zone extend their range into the Subtropical Zone, but I recall only one instance of the reverse occurring. Not only are upper zone forms held by those environmental bonds which determine their zone, but the causes which prevent an island form from taking root on the mainland doubtless also prohibit a bird from extending its range to a zone below the one it occupies. In both instances the ground is occupied. So we find none of the species which we believe to have occupied the former Panama Subtropical Zone in the Tropical Zone of that country. Successful retreat toward either end of the disappearing zone would be even more difficult than to the zone below, for here the resulting over- crowding would bring them into competition with species of similar habits. 1917.] Chapman, Distribution of Bird-life in Colombia. 157 I conclude, therefore, that those individuals of subtropical species which I believe inhabited the Subtropical Zone of Panama went out of existence with their zone. The student of living species, unlike the paleontologist, has no means of determining geologic time. When the Panama ‘fault’ occurred cannot therefore be determined from zodlogical evidence alone. The absolute identity of many of the birds inhabiting the two widely separated ends of the zone implies that they have undergone no change since their ranges were disconnected. But neither degree of variation nor stability afford a measure of time. Still one may believe that under the influence of isolation the more plastic species would show some differentiation from one another and the fact of the continued close resemblance of forms, which elsewhere vary geographically, indicates that this ‘fault’ in the subtropical stratum took place at a comparatively recent period. The facts in the case suggest that the subsidence which has occurred in Panama, and made parts of its southern coasts the islands of the Gulf of Panama, has also involved the littoral of Colombia. The trend of the Western Andes and the existence of the Atrato valley, make it improbable that this range was connected with the range on the Colombian-Panama boundary. If this be true, we may ask how so many subtropical species could cross from the Western Andes to eastern Panama, and so few go from the Eastern Andes to the much nearer Santa Marta group. The Baudé, or true coast range, through a more southern connection with the Western Andes might, however, have formed the bridge between the main Andean system and the mountains of eastern Panama. We should then have had four, instead of three ranges of the Andes in Colombia. The evidence in support of this theory is far from conclusive, but includes the apparent necessity of a larger tropical area than now exists at the Pacific base of the Andes for the development of the Colombian-Pacific Fauna, and the strong probability, as shown by its fauna, that Gorgona Island was once a part of the mainland. , Our attempts to reach the summit of the Baudé range have unfortunately failed; but such collections as have been made there by Mrs. Kerr appar- ently show that some Central America species rare, or not found by us in the Western Andes, were common in the coast range. Our data, however, are far from satisfactory, and further exploration in this range would, in my belief, result in the discovery of facts of much significance to the zodgeographer. 158 Bulletin American Museum of Natural History. [Vol. XXXVI, Species of the Subtropical Zone * of Colombia with their Representatives in this Zone in Eastern Panama, Chiriqui and Costa Rica. Colombia Nothocercus intercedens Columba albilinea albilinea Claravis mondetoura ? Geotrygon linearis linearis ? Glaucidium jardini ? Bolborhynchus ferrugineifrons ” Doryfera ludovice ludovice Eutoxeres aquila salvini Saucerottia sophie saucerrottei £ cyanifrons Chlorostilbon gibsoni Colibri cyanotus Heliodoxa jacula jamesoni 4 Calliphlox mitchelli Lophornis delattrei Pharomacrus auriceps Trogon collaris Capito bourcieri occidentalis Semnornis rhamphastinus Aulacorhynchus albivittus pheolemus Chloronerpes rubiginosus gularis Melanerpes flavigula Veniliornis oleaginus aureus Scytalopus micropterus micropterus Dysithamnus semicinereus a puncticeps puncticeps Formicarius rufipectus carrikeri Grallaricula flavirostris costaricensis Siptornis erythrops griseigularis Pseudocolaptes boissoneauti Thripadectes virgaticeps sclateri Lochmias sororia 3 Philydor panerythrus ? « montanus striaticollis Xenicopsis subalaris subalaris Sclerurus albigularis albigularis * Margarornis squamifera, Eastern Panama LTT E. a. salvini H. j. henryi L. delattrei P. auriceps T. collaris subsp. C. b. salvini A. ceruleigularis cognatus | V. o. aureus 8. panamensis D. mentalis suf- fusus F, r. carrikeri G. f. brevis S. e. griseigularis L. sororia X. s. subalaris Chiriqui-Costa Rica N. frantzi C. a. crissalis C. mondetoura G. chiriquensis G. jardini B. lineolus D. veraguensis E. a. salvini 8. s. sophize C. e. alfaroana C. caniveti salvini C. cyanotus H. j. henryi C. bryante L. helene P. moccino costaricensis T. puella C. b. salvini Tetraganops frantzi A. c. ceruleigularis C. r. uropygialis M. formicivorus striati- pectus V. o. sanguinolentus 8. argentifrons D. m. septentrionalis D. p. puncticeps F. r. rufipectus G. f. costaricensis 8. e. rufigenis P. lawrencei T. rufo-brunneus P. panerythrus P. variegaticeps X. s. lineatus 8. canigularis M. rubiginosa 1 With exceptions as marked all occur in West Andean Subtropical Fauna. 2 Eastern Andes, 3 Central Andes. 4Ecuador; unknown as yet from Colombia. 1917.] Chapman, Distribution of Bird-life in Colombia. . 159 Colombia Eastern Panama Chiriqui-Costa Rica Premnoplex brunnescens brunnescens __P. b. brunnei- P. b. brunneicauda cauda, Sittasomus zequatorialis 2 — 8. sylvioides levis Picolaptes lacrymiger lacrymiger = P. affinis neglectus Cephalopterus penduliger —— C. glabracollis Pachyrhamphus versicolor versicolor — P. v. costaricensis Acrochordopus zeledoni 3 — A. zeledoni Pseudotriccus pelzelni ! P. pelzelni Myiochanes ardosiacus Elenia pudica pudica M. lugubris E. frantzi frantzi Lophotriccus squamecrista squaame- _ L. s. minor L. s. minor crista Myiadestes ralloides venezuelensis M. coloratus M. melanops Catharus birchalli “ C. m. costaricensis fuscater fuscater =— C. f. hellmayri Cinelus leuconotus C. ardesiacus? Henicorhina leucophrys guttata H. 1. guttata H. 1. collina Troglodytes solstitialis T. festinus T. ochraceus Vireosylva josephe josephe V. j. costaricensis Myioborus verticalis verticalis M. v. verticalis M. aurantiacus Compsothlypis pitiayumi C. p. speciosa Sturnella magna meridionalis 1 8. m. alticola Chlorospingus albitempora nigriceps C. novicius novicius Piranga leucoptera ardens P. 1. latifasciata “ testacea, “ testacea Buthraupis melanochlamys B. ceruleigularis Euphonia cyanocephala * E. elegantissima, Saltator olivascens S. grandis Buarremon brunneinuchus Atlapetes gutturalis gutturalis Lysurus castaneiceps Brachyspiza capensis peruviana Spinus xanthogaster B. brunneinuchus A. g. brunnescens L. crassirostris B. ¢. peruviana S. xanthogaster ATELY THE TEMPERATE ZONE. The Temperate Zone lies between the upper limit of the Subtropical Zone and the lower limit of the Paramo Zone, or, approximately, between the altitudes of 9000 and 12,000 feet. In humid regions its upper boun- daries coincide with timber-line. Where the Subtropical Zone is arid and treeless, certain Temperate Zone species may descend to the upper borders of the Tropical Zone. Where a 1 Eastern Andes. 2 Central Andes. 160 Bulletin American Museum of Natural History. [Vol. XXXVI, mountain peak or range does not enter far into the Paramo Zone, the Temperate Zone may reach a higher than average level. This is especially true if the zone is forested; but in the absence of forests the Paramo Zone encroaches upon the upper border of the Temperate Zone which may not then exceed 11,000 feet. In Colombia, except when interrupted by the Paramo Zone, the Tem- perate Zone occupies the crest and both slopes of the ranges on which it occurs. There is, therefore, no such difference in the life of its eastern and western slopes as is found, for example, in those of the Subtropical Zone of the Central Andes. In the comparatively low Western Andes, the Temperate Zone north of Popayan is present only at a few disconnected localities. Richardson and Miller found it west of Popayan, and Miller and Boyle discovered it on the Paramillo at the northern end of the chain. At both places Dzglossa gloriosissima was common, but the species has yet to be taken elsewhere. Doubtless the Temperate Zone is also present on the “Paramo’’ (so- called) of Frontino, and on the other higher peaks of the northern part of the Western Andes. In the Central Andes it is probably continuous as far north as Sta. Elena, east of Medellin, and, except for the subtropical break at Andalucia, the Temperate Zone appears to occupy most of the summits of the Eastern Andes, though I am unable to state its northern limits in Colombia. In Venezuela this zone reaches the vicinity of Merida. In the Santa Marta group, if one may judge from Dr. Allen’s summary of our knowledge of its bird-life, the Temperate Zone holds comparatively few representative species. Of seventy-three species which I list as characteristic of the Temperate | Zone in the Eastern Andes, only § seven are specifically, and only eleven are generically represented in this zone in the Santa Marta moun- tains. Further field-work will doubtless add to the list of Santa Martan Temperate Zone species, and perhaps explain certain anomalies in distribu- tion contained in Allen’s paper. For example, Buarremon assimilis, which we have found commonly in the Temperate Zone of all three ranges but never below, is recorded from Bonda (2 adults, 2 juv. in nestling plumage), a locality near sea-level. Again, Myospiza manimbe, which we have never found above the Tropical Zone, is recorded on the authority of Bangs from the Paramo of Macotama (alt. 11,000-15,000 ft.). Southward, the Temperate Zone increases greatly in area in the inter- andine valleys of Ecuador, and on the tablelands of Peru and Bolivia. In Argentina and Chile it descends to sea-level at a latitude not yet determined. Unlike the Subtropical Zone, the fauna of which is almost wanting in arid, treeless regions, the Temperate Zone has strongly marked humid_and Buuuetin A. M.N. H. Vor. XXXVI, Plate XXXV. CHARACTERISTIC TREES OF TEMPERATE ZONE ForEs?. (Photographed near Laguneta, Central Andes.) 1917.] Chapman, Distribution of Bird-life in Colombia. 161 _arid divisions. The first is characterized by a dense low forest of thickly branched, sturdy trees on which grow numerous epiphytes and parasites, including a great variety of mosses. Allen’s description of the Laguneta region (Expedition No. 3) gives an excellent idea of Temperate Zone forest. The arid portions of this-zone include bush-grown or treeless slopes, and the Sar Savanna of Bogota, with its exceptionally favorable haunts for plains- nd marsh- loving Species. 7 interesting Sbnaion of species obviously derived from lower zone forms of the same latitude and from forms inhabiting the same zone at a different Thus, Zenaida ruficauda ruficauda, Z. r. antioquie, Troglodytes musculus columbe, and Agelaius icterocephalus bogotensis are racial, intergrading representatives of Tropical forms of the same latitude. While, though now specifically distinct, Penelope montagni, Trogonurus assimilis, Andigena hypoglaucus appear respectively to be zonal representatives of P. cristata, T. personatus, and A. nigrirostris, all of which are found in the adjoining lower zone. ' Of the second group, or true Temperate Zone species which have ex- tended their range to the higher parts of the Andes from a latitude where this range reaches sea-level, Porphyriops melanops bogotensis is but a slightly differentiated form of P. m. melanops of Paraguay, Scytalopus niger is found unchanged at sea-level in Chile, and Catamenia analis schistaceifrons is an intergrading form of C. analis analoides which is found on the coast of Peru. One of the most interesting results of our study of zonal life is the dis- covery that two forms of Streptoprocne zonaris inhabit the Andes; one, S. z. albicincta, occurring in the Tropical and Subtropical Zones; the other, S. z. altissima, in the Temperate Zone. Although these closely related forms, when feeding, are doubtless associated, we have found no intergrades be- tween them. The latter indeed, more closely resembles S. z. zonaris of southern Brazil than it does S. z. albicincta, and it is not improbable that it has been derived from zonaris through extension of range with increasing altitude in the Andes, rather than from albicincta. The latter, however, is also a racial representative of zonaris. Hence apparently two forms having a common ancestor with which both intergrade, meet as species. We must look not only to more southern, but to more northern latitudes if we would discover the ancestral type from which certain species of the Temperate. Zone in Colombia were derived. Possibly no more convincing proof of the ‘northern origin of a Colombian Temperate Zone race could be asked for than is furnished by Otocoris alpestris peregrina, a common species of the Bogoté Savanna, to which it appears to be restricted. es Fig. 19. Semi-diagrammatic representation of the Range of Scylalopus niger, a Temperate Zone species. which ranges from the sea-level in Chile to 10,000 feet in the Colombian Andes, without showing racial variation. Its range is not known to be as continuous as the diagram indicates. 162 1917,] Chapman, Distribution of Bird-life in Colombia. 163 Fulica americana columbiana, Ixobrychus exilis bogotensis and Asio flammeus bogotensis are also Savanna forms of presumably northern origin which have reached Colombia under climatic conditions no longer existing, and are now associated there with species of southern origin, which have apparently arrived during prevailing conditions. Characteristic forms of the Temperate Zone whose origin is not now determinable, are species of the genus Grallaria, Ochtheca, Diglossa and Buthraupis, while the Hummingbirds, of which we found some sixteen species, are more numerous than the members of any other family. The more uniform climatic conditions of higher altitudes, as well as of higher latitudes, tends to create corresponding uniformity in their life. Unlike the zones below it, the Temperate Zone of Colombia cannot be divided into smaller faunal areas. Even when Temperate Zone islands of the same range are as widely separated as are those of the Western Andes, there is striking similarity in their bird-life. Hence we conclude that as with the Subtropical Zone of Colombia and Costa Rica, they were at one time connected, and owe their present isolation to erosion in the interven- ‘ing area. This belief is strengthened when we compare the life of what we believe to be a true “oceanic” Temperate Zone island in the Santa Marta group, with that of the same zone in the Eastern Andes, and find how few Andean species have crossed to the Santa Martan Zone. In the Central and Eastern Andes the Temperate Zone is too continuous to permit of isolation with subsequent differentiation. In defining the boundaries of the Subtropical Zone we have seen that when the Andean system of Ecuador and southern Colombia develops into three distinct ranges, the Subtropical Zone of the Pacific slope is continued northward in the Western Andes, that of the Amazonian slope in the East- ern Andes, while the Central Andes, having an indirect connection with both Western and Eastern ranges, has received a certain amount of life from each, but has little of its own. When, however, we examine the topographical relations of the Temperate Zone, we find that the Central Andes carries a direct northward extension of the great Ecuadorian interandine temperate region, and as such it has some species, particularly at its southern end, unknown elsewhere in Colom- bia. Examples are Angaretes p. equatorialis, Conirostrum fraseri, and Urothraupis stolzemant. the: he study of peers is forcibly illustrated i in the Colombian Temper- ate Zoné by the bird-life of the the Bogota Savanna. pparently no other part of this zone in Colombia possesses the physical characteristics of of that area. Though evidently ‘much modified by cultiva- “Le 164 Bulletin American Museum of Natural History. [Vol. XXXVI, tion it still may be described as a flat, treeless plain suitable for the wants of such plains-loving species as Horned Larks and Pipits, dotted with sloughs and reedy marshes, ideal haunts for Coots, Rails, and Bitterns. Nowhere else in the Temperate Zone of Colombia do these and certain other species find the conditions they require. The Savanna, therefore, constitutes the entire range of such forms as Rallus semiplumbeus, Porphy- riops melanops bogotensis, Fulica americana columbiana, Ixobrychus exilis bogotensis, Asio flammeus bogotensis, Habrura pectoralis bogotensis, and Otocoris alpestris peregrina, when it follows that, lacking this favorable locality, none of these species would be represented in Colombia. Various other species, generally wide-ranging tropical and subtropical forms, are found more or less regularly in the Savanna, but are unknown elsewhere in the Colombian Temperate Zone. Examples are Egretta candi- dissima, Florida cerulea, Gallinula galeata, all of which have been found in the Savanna by Hermano Apolinar Maria, whom I have to thank for this information. So small a part of the Andean Temperate Zone is contained in Colombia that what has been written here can be considered only as a contribution toward the solution of a very large problem. Birds of the Temperate Zone. Family Tinamide Nothocercus julius Family Cracide. Penelope montagni Family Columbide Zenaida ruficauda ruficauda “ as antioquix Family Rallide Rallus semiplumbeus Porphyriops melanops bogotensis Fulica americana columbiana Family Bubonide Asio flammeus bogotensis “ stygius Family Psittacide Pionus seniloides seniloides Hapalopsittaca fuertsi Family Cypselide Streptoprocne zonaris altissima Family Trochilide Helianthea helianthea id bonapartei £ lutetia lutetia Lafresnayea saiil saiil Agleactis cupripennis cupripennis Vestipedes vestitus vestitus . “« smaragdinipectus mosquera derbyi longirostris Metallura tyrianthina tyrianthina Chalcostigma herrani Ramphomicrom heteropogon micro- rhynchus Opisthoprora euryptera Psalidoprymna victoria victoriz © gouldi gouldi Family Trogonide Trogonurus assimilis “ “ Family Ramphastide Andigena hypoglauca 1917.] Family Picide Hypoxanthus rivoli rivoli “ brevirostris Veniliornis nigriceps equifasciatus Family Hylactide Scytalopus niger * griseicollis infasciatus Myornis senilis Acropternis orthonyx ri Family Formicariide Chamzza mollissima Grallaria squamigera . ruficeps rufocinerea monticolor milleri Oreopezus rufula rufula “ “ “& Family Dendrocolaptide Schizceaca fuliginosa Synallaxis a. elegantior “ gularis gularis , subpudica Thripadectes flammulatus “ Family Tyrannide Ochthodieta fumigatus Ochtheeca cenanthoides fumicolor brunneifrons frontalis lessoni Mecocerculus leucophrys setophagoides # stictopterus uropygialis Aneretes parulus equatorialis . agilis Family Cotingide Euchlornis arcuata Heliochera rubricristata a“ «“ « Family Hirundinide Orochelidon murina Family Troglodytide Cinnicerthia unirufa ie unibrunnea Cistothorus apolinari Troglodytes musculus columbe Chapman, Distribution of Bird-life in Colombia. 165 Family Turdide Semimerula gigas gigas © “ gigantodes Family Mniotiltide Myioborus ornatus chrysops Myiothlypis nigrocristatus Basileuterus luteiviridis . richardsoni Family Motacillide Anthus bogotensis Family Alaudide Otocoris alpestris peregrina Family Fringillide Pheucticus uropygialis uropygialis Catamenia inornata minor f analoides schistaceifrons homochroa Spinus spinescens “ ~ nigricauda Spodiornis jardini Atlapetes schistaceus “ pallidinuchus pallidinuchus € « papallacte “ Buarremon assimilis Family Cerebide Diglossa gloriosissima < brunneiventris lafresnayei aterrima Conirostrum sitticolor a rufum fraseri 9 a“ e Family Tanagride Tridosornis dubusia dubusia 6 “ — ignicapillus ceruleoventris Peecilothraupis lunulata lunulata e palpebrosa palpebrosa < « olivaceiceps « “ Buthraupis cucullata gigas e eximia chloronota ad «“ — eximia Sericossypha albocristata Hemispingus atropileus 166 Bulletin American Museum of Natural History. [Vol. XXXVI, Hemispingus superciliaris nigrifrons Agelaius icterocephalus bogotensis Pseudospingus verticalis - Sturnella magna meridionalis Urothraupis stolzmanni Microgleus subalaris Psittospiza riefferi niefferi Family Corvide Family Icteride Cyanolyca armillata armillata Cacicus leucorhamphus « : quindiuna Tue Paramo ZONE. The name ‘Paramo’ is locally applied to any treeless region lying above 10,000 feet. Thus, the road from Bogota to Chipaque is commonly said to pass over the Paramo of Boqueron whereas, faunally, it nowhere extends above the Temperate Zone. The true Paramo Zone extends from the upper limit of trees to the lower limit of snow. On Santa Isabel, in the Central Andes, Allen and Miller found this zone between the altitudes of 12,500 and 15,200 feet, but where the upper border of the Temperate Zone is arid and lacking in forest the paramo appears to reach a lower level. Thus, on the range east of ‘Bogota, the mullein-like ‘frailejon’ so characteristic of the Paramo Zone, grows abundantly at 11,000 feet, and some plants of this species are found even lower. . Where, however, on peaks which do not rise to snow-line, humid condi- tions prevail, the tree-line may reach 13,000 feet and the lower level of the Paramo Zone be correspondingly higher. Miller and Boyle reached this altitude on the Paramillo, the highest point in the Western Andes, but found there but few species which may be considered as representative of this zone, including Orodynastes striaticollis striaticollis. We have elsewhere found this bird only on the Paramo of Santa Isabel, but I observe that Bangs records it from an altitude of 8000 feet in the Santa Marta group, and it possibly may not be a true Paramo Zone form. Conditions on the Paramillo seem favorable for the presence of species which are common on Santa Isabel, and their absence can only be attribut- able to the isolation of this peak and may be considered to indicate that. at no time has it been connected with areas having the life of the Paramo Zone. This zone in its full development is therefore, apparently absent in the Western Andes. In the Central Andes it occurs in at least twelve places between the Paramos of Las Pappas and Santa Isabel. Both these localities were visited by our expeditions (Nos. 3 and 4) and Allen’s careful description of the country traversed should be read in this connection. 1917,] Chapman, Distribution of Bird-life in Colombia. 167 In that part of Colombia lying south of Popayan, where the Andean system retains much the same character it possesses in Ecuador, there are no less than twenty mountains which rise above the lower level of the Paramo Zone. , The Eastern Andes possesses some twenty mountain summits of suffi- cient altitude to support a Paramo Zone fauna, but few of them have been explored. This zone also exists in the Andes of Merida, Venezuela, and in the Santa Marta group.! The proportion of paramo species in the last-named range is higher than that of any of the three lower zones, a fact which possibly is due to the open nature of the haunts of paramo birds and their consequent exposure to storms which may transport them considerable distances. If we except so cosmopolitan a genus as Gallinago, the species of the Paramo Zone of Colombia are all of southern origin. All the genera repre- sented reach sea-level in the south Temperate Zone and most of them are absent from the Tropical Zone. Cinclodes, Upucerthia and Muscisaxicola are admirable examples of South Temperate Zone genera which, with increasing altitude, have extended their range northward to the very limits of the Paramo Zone. Even the Condor, a sea-level bird of Patagonia, makes what we think of as his true home on the summits of the Northern Andes, where the factors which determine zonal boundaries keep him to his true level quite as effectively as they do a diminutive Marsh Wren. Like that of the Temperate Zone, the life of the Paramo Zone in Colombia re- quires no faunal subdivisions. Allen and Miller’s work on Santa Isabel shows that the Central Andes, as the topography of the region indicates, is the main northward extension of the Andean system. Muscisaricola columbiana and Upucerthia excelsior columbiana, both representing genera hitherto unknown in Colombia, were found by them in numbers. Doubtless additional work in the Paramo Zone of the Central Andes would reveal the presence of other southern forms. Birds of the Paramo Zone. Family Charadriide Family Cathartide Gallinago nobilis Sarcorhamphus gryphus “ jamesoni Family Trochilide Family Anatide Pterophanes temmincki Nettion andium Vestipedes paramillo 1 The occurrence of Cinclodes in the Paramo Zone of the Santa Marta group and of the Andes near Merida, Venezuela, is surprising. No other species of this genus is known from nearer than Ecuador. Possibly the genus will still be discovered in the Colombian Andes. 168 Bulletin American Museum of Natural History. [Vol. XXXVI, Metallura williami Siptornis flammulata quindiana O abelli xypogon stiibelli Family Tyrannide Family Hylactide Scytalopus canus " sylvestris Orodynastes striaticollis striaticollis Muscisaxicola alpina columbiana Family Formicariide Family Troglodytide Upucerthia excelsior columbiana Cistothorus equatorialis Family Dendrocolaptide Family Fringillide Leptasthenura andicola Phrygilus unicolor grandis Siptornis fammulata multostriata . “ geospizopsis Tabular Synopsis by Families of Zonal Distribution of Colombian Birds Collected by the American Museum’s Expeditions. Zones Sub- Tem- Families Tropical | tropical perate Paramo 1 141} PIMA MAIO vs 2s c2 se oe dae eet Gah nea Rake fe CTACIDB ssh. c.e ieee be oe wd meee ee MRSA Odontophoride......... 0.0.0.0. eee Colum bidiees & since seins coins ag 'y eee see Opisthocomide......... 0.6.00. cee ee eee Pallide a iss asenis e-cas whune-swahea'e aw duns vented Podicipedide: » . 2.4 dace. e0noeg eeu eune ves Heliornithide.. 2... 0... kee eee eee Tyaridatts. ee 2:cie nex Bes vee Noms te Kae es Charadriide sc. 4.22 cee cis wns eee ota sews i 22g (6 Laer ee ea ae Eurypygide.......: Sinha ania Hideaway ae aanees Cidicnemide.. 2.1.1... eee eee ee PSOphiidee ns. soe. ns shal od Rawle aoa waa ONhas Tbididits. ssa. ne oar Orie aes ewe ees oan s on > wy 2+11 3 11 CLCONM Es pease. Ghee ee area pane ga a Rae RRS Arceidte 5:54 2462 sa nankle's on cthty aaeet aki ne Se Me Ao Palamedeide................... beak eSO Ne Amhin Gide, os siceaiie se bards svt nee suena wa dares Cathartide... 0.0.0... ccc ccc cee eee BR i) ee (OH RaAN|®ArR AMER OPRNE pa ENnaon wo 22 21 1 1 Ranging upward from the Subtropical Zone. ? Ranging upward from the Tropical Zone. . 1917,] Chapman, Distribution of Bird-life in Colombia. 169 Zones Sub- Tem- Families Tropical | tropical perate Paramo Balconidee scsi. 0a tues eee none Seas 33, 11? 1+5?2 Bub ON da sisi nce 5 sic agaee de wierd dhe ede ne wens 8 1 2412 PSIttAGd B's. 2.4 se 4.ca peso isine ee Be din wee aoe RAS 22 6+1? 2 Alcedinids' jess saeerex needa ene. 4 Momotide sxe sede os deteac dice ie sce licesecacnd 9 1 Caprimulgide.............00. 0. cece ee 9 2+1? 1 Oo 5 22 MeO CHUA coc oe has ceasoaeaetsce eens pees 48 37+32) 16+61! 4 Trogonide:. <3... gos ce ves va eres ep enee en ane 10 4 1 CUCU Seeks a eacbo ws did Vek pee os eae ae 9 42 Capitonide sc:55-4 2 wien aateas Abe canes ene 7 4 Ramphastide.... 0.0.0... cece eee eee 17 7+1?2 1+11 Galbulide. 0... eee eee ee 8 H 3300E(s10) 11 (oF: ar a ne ve a 19 12 PICU soecatee stausrdea cd oss a weal ae Sage bscrased SOES 25 7+1 3+11 Conopophagide..............00.ce eee ee 3 12 Hylactide........ siidsld pag pteady wad fea as 1 5 2 Formicariide......... 0.0.0. e cece cece eee 82 17+5? 7 1 Dendrocolaptide....................-00. 48 27+32) 5451 3 SEY PANN hex feiss ¥ did 258 soe dade Sie iad Beaubien 90 38+5?) 10+51 2 PUDTVG SS 62 iat deestidoirdaar sci eand dala aaa deed 22 7 Cotingide .......... Biscay A Sits d nays ditties 24 15 2 Hirundinide................ Tata senaader a ca 8 142? 1 OYlVid Bc aceak aa cccanay Head oa aen’s owe eee 3 Troglodytide......... 00. e cece eee eee 23 12 4421 1 Cinch seisecaauch evade sieeve eee helen 1 Maid i560: er aaisa 4 cates band aauans 4 1? dL 6 Fs eer re oe 8 8+32 2 WIPCONID2:. 0.5 ce vie SAK BA OVA Mare tance wads 7 3+12 5 Mnhiotiltide.. 0.0.00... eee ee eee 10 5+12 5 Motacillide... 0.00... 0... cece eee eee 1 AlaUdidssi2 .iaucs qui heae ox meee Xe oes mone 1 Catamblyrhynchide.................00005 ‘1 1 Prin uli ee s.2:s ves coceataucia s gee ath aaneea ecoen 35 17+72 | 11411 2 (Coprebidass ss sists dc sacgs eva val ote dale ei haley aus 13 74+22} 8421 Procniatide.. 0.00. ccc cece awe euee 1 12 Tanagridie... i000 ccs once ea cee eee aoe ve 51 51+72 | 15+51 TCberidas ss Seis oaks cunts arate n west eal ns ses 22 6412] 4411 Cornvidee js. es eee y HOS FHS ees Ea 2 2 2 1 Ranging upward from the Subtropical Zone. 2 Ranging upward from the Tropical Zone. Part II. A DISTRIBUTIONAL LIST OF THE BIRDS COLLECTED IN COLOMBIA BY THE AMERICAN MUSEUM’S EXPEDITION. Classification.— It is greatly to be regretted that no one system of classification is accepted as authoritative by writers on South American birds. Everyone who has experienced the annoyance of referring to faunal papers, the writers of no two of which may have adopted the same system of classification, and which as authors’ ‘separates,’ are usually without an index, should admit that convenience of reference is here of first importance. The writer has seen too many systems of classification accepted and rejected to have much faith in the stability of any now, in greater or less measure, current. So far as he personally is concerned it is immaterial which one of half a dozen now in use be followed, but it is material that we use that one consistently. ij If we except Sclater and Salvin’s ‘Nomenclator Avium Neotropicalium’ (1873), which included the birds of Mexico and Central America as well as those of South America, only one list of South American birds, as such, has ever been published. This, Brabourne and Chubb’s ‘Birds of South America,’ is not only as authoritative in the present,state of our knowl- edge, as, we can perhaps expect such a general work to be, but it conforms tothe Classification of Sharpe. From the standpoints of both scientific excellence and expediency it. seems therefore eminently desirable to accept the classification of this work, and I have adopted it in the present paper. The numbers in parentheses preceding each name in the systematic portion of this paper are those of Brabourne and Chubb’s work. Nomenclature-— The nomenclature in Brabourne and Chubb’s list is binomial, the authors, having deferred an expression of opinion of the subspecific relations of the forms listed until these forms were treated at length in succeeding volumes of their work, a plan, which, owing to the unfortunate death of the senior author, will now never be realized. It is, of course, out of the question to use binomial nomenclature in the present paper, and the trinomials employed follow current usage, or express the author’s views as they have been formed through a study of the material at hand and under the requirements stated below. ~ Aside from this necessary change from binomialism to trinomialism, I have followed the nomenclature of Brabourne and Chubb’s list, except 170 1917,] Chapman, Distribution of Bird-life in Colombia. 171 in a limited number of cases where additional material has led to other conclusions. From the standpoint of nomenclature, pure and simple, I have made no attempt to revise the names they present. Treatment of Genera.—I do not approve of the present-day excessive multiplication of genera. I believe that we should treat what we term genera much as we treat species, and when the variations from a given generic type do not result in actual segregation, but simply mark the con- necting stages, then such variations should be considered of subgeneric value. We all exhibit a tendency to forget that a genus is, in a large measure, an artificial creation, and that the characters on which it is based are ill- defined, unstandardized, and arbitrarily employed. In the hands of the systematist whose talent, often highly developed for analysis, leads him to magnify the importance of minor characters, classification becomes an end rather than a means. Accepting the doctrine of evolution he nevertheless seems determined to prove the theory of special creation. It is his business to assort, arrange and pigeon-hole certain facts as these facts are repre- sented by specimens. The necessity for drawing up diagnoses, keys and descriptions for the identification of these specimens leads him to search for differences rather than resemblances. To these differences he gives names, and to these names we apparently cannot avoid attributing a signifi- cance they are often far from possessing. As a result, nomenclature over- shadows classification and facts are obscured or wholly disguised by names. There is unquestionably. urgent need for a thorough generic revision of many groups of South American birds, but the reviser should not feel com- pelled to found a new genus on every species showing a departure from the set of artificial characters he has assigned to the so-called type. Further- more, in order to determine whether the differences observed are of generic or subgeneric value, his revision should be based on all, not a part, of the species of the group concerned. It was first intended in the preparation of the following list of species collected by us in Colombia, to use currently accepted generic terms, and when recent authorities differed to attempt to reach an independent de- cision based on original investigation. I soon found, however, that the instances in which authorities differ are so numerous, that a proper con- sideration of the points at issue would require both far more time and material than were available, and I was forced to abandon this plan. I have, consequently, followed sometimes one author, sometimes another, and the results here presented I frankly confess to be both inconsistent and unsatisfactory. I feel that there is no hope for uniformity in the treatment of this question of genera, until systematists reach some agreement in 172 Bulletin American Museum of Natural History. [Vol. XXXVI, regard to what constitutes a generic character. At present we are without such a standard. The genus of one author is the subgenus of another, and is not recognized at all by a third. In many instances, therefore, a genus becomes merely a personal expression of opinion concerning the taxonomic value of certain admitted characters. The same species may be referred to a dozen or more different genera by as many writers, all of whom may agree on the details of structure and relationships involved, and disagree only on questions of nomenclature. If the ultimate object of systematic zodlogy were classification based only on analysis, we might be warranted in carrying dissection and descrip- tion to their utmost limit and applying to the results as many names as the most minute differences discovered seemed to require. It is, however, commonly agreed among biologists that the primary object of systematic zodlogy is to provide a nomenclature which can be used with some degree of precision, and which shall be based not wholly upon analysis but to a degree upon synthesis as well. =, So far as species and their geographical races are concerned, the trinomial system of nomenclature permits the systematist to recognize but slightly differentiated forms by name without obscuring their more essential rela- tionships. He may not express lines of descent, even if they are known, he may indeed name first the most recent offshoot of a certain stock, but in the naming of a subspecies he does not disguise its group relationships. In our treatment of species, intergradation, known or probable (see beyond under Treatment.of Subspecies) is the test which determines sub- specific status. But in our treatment of genera, this test is largely ignored. We have, it is true, subgenera, but so far as practical nomenclature is concerned we are trinomial with species and binomial with genera. In consequence, many of our generic terms are just as false, just as misleading, just as far from conveying an idea of actual relationships, as though we were to use a binomial for every subspecies. They are, indeed, more misleading since in the latter case the name employed would indicate at least generic relationship, while in the former all suggestion of relationship may be lost. I am aware that the two cases are not wholly comparable, and that strictly to apply the test of intergradation to generic groups, would, in some cases, place in the same genus species, which in the light of our current understanding of what constitutes a genus, could not be considered as generically ‘related; and thereby lead to nomenclatural results as undesirable as those based on excessive analysis. But I am also aware of how erroneous an impression may be given by unduly emphasizing differences which are obviously of less importance than resemblances, and then, chiefly for con- 1917.] Chapman, Distribution of Bird-life in Colombia. 173 venience in classification, labeling them with a name to which’ we can no more help attaching the attributes of an entity, than we can avoid think- ing of political boundaries as physical facts. Two species, for example, may be ninety-five percent alike, and five percent unlike, in their so-called generic characters. The five percent of unlikeness is‘made the basis of a generic division, a new name is given and the ninety-five percent of resemblance is thereby effectually concealed by the combined results of analysis and nomenclature. I will illustrate with the following examples: The American Quails to which the name Bob-white is commonly applied, are by most authors placed in two genera, Colinus (or its nomenclatural equivalent Ortyx) and Eupsychortyx. The ‘generic’ differences between Colinus virginianus of the United States and Eupsychortyx cristatus of northern South America, structurally as well as geographically the most widely separated species of the group, are found in the feathers of the head and wings; ‘Eupsychortyx’ having the central crown-feathers elongated and forming a well-defined crest, while the first (outer) primary is slightly shorter than the eighth, rather than longer, as in Colinus. There is also a well-marked difference in pattern of coloration, though both are obviously quails. Current standards in ornithology would accept these differences as of generic value, and if the two species mentioned were the only ones concerned, the appropriateness of the generic distinctions mentioned would not be questioned. But in Central America, from Yucatan to Costa Rica, or in other words, from the southern limit of the range of true Colinus in Guate- mala to the northern limit of the range of Eupsychortyx in western Panama, there are found several species in which a crest is evident, in which the wing-formula is sometimes that of the northern bird, at others that of the southern, and which also approach the type of coloration shown by the South American species. In short, so far as the characters mentioned are concerned, these Central American species are intermediate or connecting forms. Practical evidence of the truth of this statement may be found in their treatment by various recent authorities, some of whom refer them to one ‘genus’ some to the other. In spite of their superficial unlikeness, it is apparent, therefore, that a line generically separating North American Bob-whites from South American Bob-whites cannot be drawn, and hence it follows that’ a nomen- clature which recognizes generic distinction between them, conceals the biologically significant fact of their group relationship. Evidence of the closeness of this relationship not found in the birds’ skins is supplied by their notes. The calls of all the species have not as 174 Bulletin American Museum of Natural History. [Vol. XXXVI, yet been recorded, but I can affirm from personal experience that the call of Colinus nigrogularis of Yucatan and of Colinus cristatus leucotis of Colombia is essentially the familiar “ Bob-white ” of Colinus virginianus. The voice of the southern bird lacks the volume of that of the northern one, but the notes and their peculiar quality are the same in all three. In view of all these facts, I feel that the actual relationships of the species of this group are more nearly expressed by referring them all to the genus Colinus. To illustrate further what I feel. to be the evils of unwarranted generic separation, there is found in southern South America (Bolivia, southern Brazil, Argentina and Chile) a group composed of those species of small black and white woodpeckers, the close relationship of which to the Downy Woodpecker is obvious, and which have generally been considered as con- generic with that species. They all have the back barred as in Dryobates scalaris, of the southwestern United States and Mexico, and the more western species (lignarius) has the underparts. heavily streaked. Since, however, in this respect it differs from the south Brazilian species cancellgtus, more than does that species from scalaris, this cannot be considered a charac- ter of generic importance. The southern birds, it is true, have all, instead of only the lateral tail-feathers barred as in our northern species; but although. I believe that pattern of coloration is often a much better generic character than differences of degree in the shape of bill, relative length of wing, etc., no one, I think, would claim that this difference in the pattern of tail-marking would warrant generic separation. The only ‘structural’ differences said to exist between the North American and South American species of these Woodpeckers, are to be found in the relative length of the primaries and secondaries, which results in giving the southern birds a more ‘rounded’ wing. On the basis of this character it is, therefore, proposed to segregate them in the genus Dyctiopicus Bonap., a proceeding which would conceal what I believe to be the generic relationships of the Downy Woodpeckers of South America and North America, a relationship of high importance zodgeographically since in the territory lying between Bolivia and western Panama no form of this group is known to occur.