LIST OF FIGURES IN TEXT FIGURE Memoir No. 1 page 1 . Shores of the Atrato 11 2. The Cauca River near Cali (Juancita) 12 3. Narrows of the Rio Magdalena below Girardot 13 4. Tambo, Atrato basin 14 5. The Alagdalena at Girardot 15 6. Upper Rimac Valley 21 7. Map of Central Peru from Callao to Rio Ucayale 23 8. Pectoral spines of Hexaneniatichthys simonsi (Starks) 29 9. Hexanematichthys henni Eigenmann 30 10. Cetopsorhamdia nasus Eigenmann and Fisher 36 11. Rhamdia wagneri {Gunther) 38 12. Rhamdia cinerascens (Gunther) 39 13. Nannorhanidia nemacheir Eigenmann and Fisher 41 14. Teeth of upper jaw and palate of Pseudoplatystoma fasciaturn (L.) 46 15. Astrohlepus trifasciatus Eigenmann 52 16. Hemiancistrus landoni Eigenmann 72 17. Pseudancistrus pediculatus Rigewmaim 79 18. Pseudancistrus carnegiei Pig'enmaim 79 19. Chw.tostoinus fischeri i^temdachner 83 20. Ancistrus centrolepis Regan 87 21. Sketch-map of the upiier drainage area of the Rio Meta 220 Memoir No. 2 1. Nyctotingis osborni gen. et sp. nov 363 2. a. Leptobyrsa splendida Drake, sp. nov 374 b. Leptobyrsa elegantula Drake, sp. nov 374 Memoir No. 3 1. Photograph of left side of skull of Diplodocus. (C. M. No. 11,161) 385 2. Anterior view of the teeth of Diplodocus. (C. M. No. 11,161) 388 3. Posterior view of teeth of Diplodocus. (C. M. No. 11,161) 388 4. Jaws of Diplodocus seizing a mussel 390 vii LIST OF PLATES PLATE I. IT III. IV. V. VI. VII. VIII. IX. X. XL XII. XIII. XIV. XV. XVI. XVII. XVIII. XIX. XX. XXI. XXII. XXIII. XXIV. XXV. XXVI . XXVII. XXVIII. XXIX. XXX. XXXI. Memoir No. 1 Xyliphius, Bunocephalus, Cetopsorhamdia, Chasmpcranus. Pseudopimelodus, Microglanis. Hemicetopsis, Perugia, Pseudoplatystoma. Cetopsorhamdia, Imparginis, Nannorhaindia. Trachycorystes fisheri. Astroblepus. Astroblepus. Corydoras, Hoplosternum, Cheiridodus. Hemiancistrus. Hemiancistrus , Pseudancistrus. CochUodon, Lasiancistrus . Pseudancistrus, Choetosto^nus, Ancistrus. Chcetostomus, Ancistrus. Loricaria. Loricaria. Sturisoma. Curimatus. Curimatus. Parodon, Apareiodon, Prochilodus, Saccodon, Characidium. Prochilodus, Copeina, Leporinus. Pijrrhulina, Lebiasina, Piabucina. Brycon. Brycon. Hyphessobrycon, Astyanax, Thoracocharax. Charax, Roeboides. Gilbertolus, Acestrocephalus, Ctenolucius. Alollienisia, Rivulus, Gambusia, Pseudopoecilia, Poeciliopsis. Stolephorus, Joturus, Stellifer, Sicydium. Hemieleotris , Gobius, Awaous, Thalassophnjne. ypquidens. JEquidens, Cichlasoma . IX X MEMOIRS OF THE CARNEGIE MUSEUM PLATE XXXII. XXXIII. XXXIV. XXXV. XXXVI. XXXVII. XXXVIII. XXXIX. XL. XLI. XLII. XLIII. XLIV. XLV. XLVI. XLVII. XLVIII. XLIX. L. LI. LII. LIII. LIV. LV. LVI. LVII. Cichlasoma, Geophagus. ^quidens, Cichlasoma. Plecostomus, Moenkhausia, Pomadasys, Sternarchus. Sternarchus, Sternopygus. Map of Northwestern South America, Section One. Map of Northwestern South America, Section Two. Map of Northwestern South America, Section Three. Memoir No. 2 Neotropical Tingitidce. Memoir No. 3 Fig. 1. 'View of left side of skull of Diplodocus, (C. M., No. 11,161) Fig. 2. Skull and cervical vertebrae of Diplodocus, (C. M., No. 3,452) Fig. 1. Superior view of outer surface of skull of Diplodocus, (C. M. No. 11,161). Fig. 2. Palatal view of skull of Diplodocus, (C. M., No. 11,161). Fig. 1. Photograph of palatal view of skull of Diplodocus, (C. M. No. 11,161). Fig. 2. Photograph of palatal view of skull of Diplodocus, (C. XL, No. 11,161). Skull of Diplodocus, (C. XI., No. 11,255). Memoir No. 4 Dolichorhinus longiceps Douglass. Skull and lower jaw (No. 11,071 C. XL, Cat. Vert. Foss). Dolichorhinus. Skulls, Nos. 11,072, 3,117, 11,080, C. M., Cat. Vert Foss. Dolichorhinus. Skull, N^o. 11,081, C. M., Cat. Vert. Foss. Dolichorhinus longiceps Douglass. Position of skeletons in quari-y Dolichorhinus. Bones of fore limb. Dolichorhinus longiceps Douglass. Bones of fore foot. Dolichorhinus longiceps Douglass. Sacrum. Dolichorhinus longiceps Douglass. Sacrum and caudals. Dolichorhinus longiceps Douglass. Bones of hind limb. Dolichorhinus longiceps Douglass. Bones of hind foot. Dolichorhinus longiceps Douglass. Views of manus and pes. Articulated skeleton of Dolichorhinus longiceps Douglass. Restored Skeleton of Dolichorhinus longiceps Douglass. Restoration of Dolichorhinus longiceps Douglass. Publications of the Carnegie Museum, Serial No. 112 MEMOIES OF THE CAENEGIE MUSEUM VOL. IX NO, 1 W. J. HOLLAND, Editor THE FISHES OP WESTERN [SOUTH AMERICA, PART I The Feesh-Water Fishes op Northwestern South America, Including Colombia, Panama, AND THE Pacific Slopes of Ecuador and Peru, Together with an Appendix upon the Fishes of the Eio Meta in Colombia BY C. H. EIGENMANN PITTSBURGH Published by the Authority of the Board of Trustees of the CARNEGIE INSTITUTE October, 1922 u JA r lo 'i'A'f •'/?!! iVUi Corrigendum p. 5, 3rd line from bottom, for “Covitagina” read Cartagena. MEMOIES OP THE CARNEGIE MUSEUM. VoL. IX No. 1. THE FISHES OF WESTERN SOUTH AMERICA, PART I. THE FRESH-WATER FISHES OF NORTHWESTERN SOUTH AMERICA, INCLUDING COLOMBIA, PANAMA, AND THE PACIFIC SLOPES OF ECUADOR AND PERU, TOGETHER WITH AN APPENDIX UPON THE FISHES OF THE RIO META IN COLOMBIA.* By C. H. Eigenmann. (Plates I-XXXVIII.) INTRODUCTORY. The present memoir deals with the fresh-water fishes of the area between the Panama Canal and the southern boundary of Peru, between the Pacific Ocean and the continental divide from Crucero Alto in Peru to northern Ecuador, thence along the crest dividing the Amazon, Orinoco, and Maracaibo Basins on the east from the Magdalena Basin to the west. It thus includes the fishes of all of the streams on the Pacific slopes of Peru, Ecuador, Colombia, and southern Panama, as well as of the rivers of Colombia flowing north into the Caribbean Sea and the Chagres River of Panama, which also empties into the Caribbean. The fishes of Chile will be dealt with elsewhere. I planned an expedition to Colombia as early as 1891. The trip was aban- doned, however, and other work engrossed my attention. In 1904 when the United States took over the Panama Canal, I thought the time opportune to arouse interest in a general biological survey of Panama. I wrote first of all to the United States National Museum and received in reply a communication dated April 15, * Contribution from the Zoological Laboratory of Indiana University, No. 172. 1 2 MEMOIRS OF THE CARNEGIE MUSEUM. 1904, and signed by Dr. Richard Rathbun, Assistant Secretary. He said in substance that “the Smithsonian Institution is . . . endeavoring to arrange . . . for a complete natural history investigation, which would, . . . include- the fresh- water fishes. . . . The fullest opportunity will be given for collecting in all the important groups.” In Science, N.S., XXII, July 7, 1905, pp. 18-20, and in the Popular Science Monthly, June, 1906, pp. 515-530, and more fully in “Reports of the Princeton University Expeditions to Patagonia,” III, 1910, pp. 275-511, 1 published estimates of the importance of a biological survey of Panama and the bearing of the distri- bution of the fishes in this area upon the general problem of the distribution of fishes in South America. At the meeting of the American Microscopical Society at Sandusky, Ohio, in August, 1905, resolutions were passed on my initiative, urging a biological survey of Panama. The following reasons, as given, were published in the proceedings of the meeting: “Panama is a point of strategic importance in the study of the distribution of fresh-water organisms in South and Middle America. It is certain that the Pacific slope fresh-water fauna of South and Middle America was derived from the Atlantic slope fauna. The Isthmus of Panama is one of the possible routes of migration. The Panama Canal, when completed, will destroy natural barriers and cause the faunas of the two slopes to mingle to a great extent. It will thus permanently obliterate the natural and primitive conditions, and it is highly desirable that a biological survey of this region be made before the completion of the canal.” At the meeting of the International Zoological Congress at Boston in 1907 the Section on Zoogeographical Distribution passed similar resolutions and submitted them to the general session. Mr. Alexander Agassiz, the presiding officer, mis* understanding the resolutions and believing that they called for the duplication of work upon the oceanic faunas already in large part done under his direction, remarked when the resolutions were introduced, that he hoped the Congress would “have nothing to do with such a foolish proposition.” I therefore sent a communication to the Council through Professor Henry Fairfield Osborn, giving an abstract of the report on which the resolutions were based, accompanied by distributional maps and explaining in detail the purpose of the resolutions. The Council, of which Mr. Agassiz was President, thereupon approved the resolutions, and they were passed unanimously by the general session. At the meeting of the American Association for the Advancement of Science EIGENMANN: pishes of northwestern south AMERICA. 3 in December, 1907, a resolution was adopted urging the President and Congress to make provision for a biological survey of the Panama Canal Zone. The per- manent Secretary was instructed to send copies of the resolution to the President, the Vice President, the Speaker of the House, and the Secretary of the Smithsonian Institution. Similar resolutions were passed by the Ohio Academy of Sciences, the Indiana Academy of Sciences, and the American Society of Naturalists. In Science, ioT December 16, 1910 (N.S., Vol. XXXII, pp. 855-856) appeared an article, from which the following extracts have been taken : “A biological survey of the Panama Canal Zone is about to be undertaken under the direction of the Smithsonian Institution. . . . “Secretary Walcott considered these appeals [resolutions mentioned above] and under his direction a meeting of representatives of the National Museum, the Bureau of Fisheries, and the Biological Survey, the Bureau of Entomology, and the Bureau of Plant Industry of the Department of Agriculture was held. Their decisions confirmed the desirability of such a survey and in consequence of their opinions he prepared the . . . memorandum which was submitted to President Taft. “President Taft fully approved the plan for a biological survey and suggested that such arrangements be made with the Secretary of. War, the Secretary of Agriculture., and the Secretary of Commerce and Labor as would enable him to have their active cooperation in this important work. The arrangements are now ,in an advanced state, and field parties will be sent to the isthmus at an early date.” SOURCES OF THE COLLECTIONS AND ACKNOWLEDGMENTS. With other naturalists sent to investigate various branches of zoology, the late Dr. S. E. Meek, from the Field Museum of Natural History of Chicago, and Mr. S. F. Hildebrand, representing the U. S. Bureau of Fisheries, spent the winter of 1910-11 in and about the Panama Canal Zone studying the fishes. Later they spent a second season in Panama, exploring among other places the southern part of the Tuyra Basin. Their report, a splendid volume, has appeared as “The Fishes of the Fresh Waters of Panama” in the publications of the Field Museum of Natural History (Pub. 191, 1916, Zool. Series, Vol. 10, No. 15, pp. 1-374, 26 plates). In preparing this report, Mr. Hildebrand availed himself of the collections and laboratories of Indiana University during the winter of 1915-1916. Dupli- cates of the Meek and Hildebrand Collection are in the Indiana University Mu- seum, and the Carnegie Museum received some of them through an exchange with the U. S. National Museum. 4 MEMOIRS OF THE CARNEGIE MUSEUM. To carry the work southward to the San Juan and Atrato Rivers I arrived at Cartagena, Colombia, on January 3, 1912. I was accompanied by Mr. Arthur Bierhaus, who, however, returned after our first month in the field. The full account of this expedition appears elsewhere in this volume under the caption, ^‘The Colombian Reconnaissance.’’ Manuel Gonzales of Bogota became my chief assistant from Bogota to Buenaventura and Cartagena. He later collected for me within a radius of a few days’ travel of Bogota, especially to the eastward of Bogota in the head-waters of the Rio Meta. Brother Apolinar Maria, Director of the ‘'Institute de la Salle” of Bogota has sent me various collections between 1914 and 1918, especially from the "llanos,” or jilains, at the eastern base of the Andes, the streams of which form a part of the Orinoco system by way of the Meta River. These collections, together with those of Gonzales mentioned above, are reported upon in Appendix I of this paper. After my return from Colombia Mr. Carl G. Fisher and Mr. Hugh McK. Landon, both of Indianapolis, provided the means to enable Mr. Charles E. Wilson and Mr. Arthur W. Henn, at the time students in Indiana University, to visit western Colombia in 1913. The Carnegie Museum received the first series of fishes from this expedition and granted my request that an account of this new material, appropriately illustrated, should be incorporated in my report upon the fishes of Colombia. Mr. Landon and Indiana University later provided the means to enable Mr. Henn to continue his trip to southern Ecuador and along the interandean plateaus , of Ecuador. The first series of this expedition is in the collections of Indiana University, the second series is in the Carnegie Museum. An account of this trip occurs elsewhere under the title, " The Landon Expedition to Colombia and Ecuador.” I have had the cooperation of the American Association for the Advancement of Science, the Bache Fund of the National Academy of Sciences, the University of Illinois, and lastly of Mr. Will G. Irwin of Columbus, Indiana, who helped carry the survey as far south as Puerto Montt in Chile during the Irwin Expedition. Collections from this area have been lent to me by the Field Museum of Natural History, the National Museum, and the Museum of Comparative Zoology. I have also had the privilege of examining some of the collections of the Academy of Natural Sciences of Philadelphia and of Leland Stanford, Jr., University. I have examined over ninety-five per cent of the entire recorded fauna. I failed to see, or failed to recognize, slightly less than five per cent of the total fauna. The drawings were made by Mr. W. S. Atkinson of Stanford University and EIGENMANN: fishes of northwestern south AMERICA. 5 Dr. Clarence H. Kennedy, now of Ohio State University. The photographs of the fishes are the joint product of Dr. W. A. Cogshall, Miss Maud Siebenthal, and myself. Indiana University relieved me of teaching duties to enable me to collect the material and to prepare this memoir. Physical Features.^ The Pacific slope of South America, four thousand Hiiles long and rarely over one hundred miles wide, resembles a veritable shoe-string in shape. Conditions in this area vary from extremely wet to extremely dry, from wet tropical to dry and wet temperate, and from heat to cold, as one goes south from Panama to Cape Horn. The rainfall in the Canal Zone exceeds two hundred inches per annum. In Buenaventura it is said to be between two hundred and fifty and four hundred inches per year. This condition prevails to the Rio Esmeraldas in Ecuador (Veatch, “Quito to Bogota,” p. 163). South of the Rio Esmeraldas the country becomes more and more arid. On the coasts of Peru and of Chile south to Copiapo the rainfall is negligible; it does not average one inch per annum. In Peru all of the water for agriculture is derived from the rivers descending from the mountains, and in a portion of Chile, between the Loa and Copiapo Rivers, even this source fails. In Serena, central Chile, the annual amount has ranged from about two to eight and one-half inches per annum (observations made between 1869 and 1910); and in Santiago between four and thirty-one inches, the latter, an extreme record, in one of the years between 1873 and 1910. In Concepcion the rainfall has been between 26.6 and 40 inches during the period from 1876 to 1910; in Valdivia between seventy-three and one hundred and forty-three inches from 1872 to 1910; and at Puerto Montt between seventy-one and one hundred and twenty- eight inches. The amount of rainfall also varies very greatly with the altitude at any cross- section. Behind the coastal range there are local dry areas, even in the wet regions of Colombia. The upper Dagua river runs through such a “rain shadow” between Caldas and Cisnero, and the upper Cauca runs in the shadow of the western Cordil- lera and is comparatively arid. Throughout Peru, Ecuador, and Colombia there are two main chains of the Andes, the Maritime or Western Cordillera, extending from near Gevitagina in ^ This account of the physical features of the country is in large part a reproduction of what ap- peared in the Indiana University Studies, No. 45, pp. 5-11 (1920). 6 MEMOIRS OF THE CARNEGIE MUSEUM. Colombia to Cape Horn, and east of these the older of the two, called Blanca (or white) in Peru, Oriental in Ecuador, and Central in Colombia. Peru. The physical features of western Peru are very simple. The crests of the Western Cordillera always form the divide between the Titicaca or Atlantic drainage and the Pacific drainage. The crest has an elevation of over 14,000 feet everywhere, except in northern Peru. At Huarmarca, inland from Paita, a dip in the crest has an elevation of only 6,700 feet. Enoch (“Peru,” 1910, p. 11), says: “The traveller who enters the interior of Peru from the Pacific Coast must invariably cross the Andes at an altitude of 14,000 feet or more, for the passes of the main Cordillera all reach this elevation. There is one exception in the northerly part of the country towards the frontier of Ecuador, where a low gap exists in the Andes of some 6,700 feet elevation; but this is the only exception in thousands of miles of continuous mountain chain.” In Peru the Pacific slope' is drained by a large number of rivers rising in the western Andes. After a comparatively short and very swift course they either empty into the ocean, or are lost in the sands near the coast, or are more or less exhausted in irrigation projects. Only one of the rivers has a north and south trend for any considerable distance. This is the Rio Santa in central Peru, which in its upper course flows between two chains of the Western Cordillera. All of the rivers have a very great seasonal fluctuation. The Peruvian Bureau of Engineers has made observations extending through a number of years. Ac- cording to their observations the crest of the flow is usually reached in March and the minimum about August. The minimum is reached earlier in the south than in the north. The period from January to June is usually unfavorable for ichthyo- • logical exploration. Their results, as given in Boletin No. 84, Cuerpo de Ingenieros de Alinas de Peru, Lima, 1917, are generalized in the following table: The stretches between successive rivers on the Pacific slope of Peru are in most cases extremely dr}^ deserts, or mountain masses into which the rivers have cut deep gorges. The Vitor River in southern Peru, for instance, rises in an upland meadow (over 14,000 feet), flows through a region of volcanic ash, and has in its middle course a valley (Vitor Valle) about a mile wide, cultivated in vines, figs, small fruits, and grain. It then falls to a lower level near the coast, where there is another valley. Looking from tbe hills about Yura near Arequipa toward the ocean, the land is a billowy mass of arid, sand-drifted mountains and plains, with nothing green visible anywhere.^ 2 At Arequipa there is one species of Pygidmm, At Tiabaya a species of “p^J® rey” (Basilichthys) EIGENMANN: fishes op northwestern south AMERICA. 7 Discharge of Various Rivers of Peru. (Cubic meters per second, always at the same date in the months indicated.) Name of River. Maximum. Month. Minimum. Month. Santa Cub. M. 1,250 Feb.-March Cub. M. 37.0 June-Oct. Chira 1,250 Jan.-March 5.4 Sept.-Nov. Tumbez 750 Jan.-April 10.0 Sept.-Nov. Piura 480 Feb.-March 0.0 Sept.-Feb. Chicama 320 U a 1.5 June-Nov. Pativilca 280 u 10.7 May-Oct. Jequetepeque 250 March 1.3 June-Nov. Canete 240 Feb.-April 13.2 June-Oct. Lambayeque 220 Jan.-April 1.4 ({ Huaura 190 Feb.-April 9.0 (( U Zana 170 Mar.-April 0.7 . June-Dee. Moche 170 Feb.-April 0.3 June-Nov. Pisco 140 Feb.-March 0.8 July-Oct. Ica 140 a li 0.0 May-June Rimac no ii 5.9 June-Sept. Chancay 100 Feb.-April 1.0 June-Oct. Chincha 90 Feb.-March 0.1 June-Nov. Mala 80 Feb.-March 0.2 (( (( La Leche 80 Feb.-May 0.8 U (C Viru 70 Feb.-March 0.0 June-Dee. The Rimac has a somewhat different course. With its tributaries it rises in small glacial lakes having elevations of from 15,000 to 16,000 feet, and inhabited, in part at least, by Orestias. There is then a descent of a few thousand feet with very swift water, which is not suitable for fishes. Within this belt the streams are clear in the morning, but in the afternoon the melting of frozen ground rolls down thin mud, in which nothing can live.^ Between Rio Blanco and Lima, distant in a straight line less than fifty miles, the river has a fall of over nine thousand feet. At Chosica it has an annual fluc- tuation between a minimum of ten cubic meters per second in September and a maximum of one hundred and fifteen cubic meters per second in March. The Jequetepeque in northern Peru, with a total length of about seventy-five miles and a more gentle slope than the Rimac, has a minimum flow of about five is added. At Vitor Valle the Pygidium could not be found and the “peje rey” shares the river with a giant shrimp. I could not get to the lower course of this river. ^ Even at Rio Blanco at about 10,000 feet elevation we secured nothing. At Matucana (7,500 feet), we secured a few small specimens of Pygidium, Basilichthys, Lebiasina, and Bryconamericus, the four species constituting the entire fauna of the river. At Chosica (1,900 feet) and at Lima we secured the same four species in abundance. 8 MEMOIRS OF THE CARNEGIE MUSEUM. cubic meters per second in September and a maximum of two hundred and twenty in March.^ The Piura River at Piura is reduced during the dry season to a few stagnant pools, in which the fishes become crowded together. Some of them starve, but others succeed in living through the dry season. In southern Peru the interandean region is occupied by Lake Titicaca. In northern Peru, this region is drained by longitudinal rivers, which turn eastward and eventually reach the Atlantic. As stated above, throughout the whole of Peru and northward to the Tumbez River the divide between the Pacific and Atlantic drainage follows the crest of the Western Cordillera. Ecuador. In Ecuador the crests of the two main chains of the Cordilleras are but a few miles apart. They are joined here and there by cross-ridges, formed in part of old lava-flows, which divide the area between them into a series of highland parks, having elevations of from six thousand to ten thousand feet. Some of the parks drain into the Pacific, others into the Atlantic. The continental divide thus lies along the crest of the eastern chain from Popayan in southern Colombia to as far as Cotopaxi in northern Ecuador. It then shifts westward to the crest of the Western Cordillera, then again to the Eastern Cordillera, again to the Western, once more to the Eastern, finally shifting to the western crests, where it remains through all of Peru to southern Chile. It may be questioned whether the northern parks of Ecuador are drained into the Pacific, because the heavy rainfall has enabled the Patia and the tributaries of the Esmeraldas to cut back through the Western Cordillera and thus to annex the interandean streams, or whether the present trend of these interandean rivers is not due to the late formation of the Cordillera of Bogota, which in southern Colombia and northern Ecuador is piled up against the Cordillera Oriental. In the center and south of Ecuador, others of the interandean parks are tapped by Pacific slope rivers, such as the Tumbez, Rompida, Canar, Chanchan, and Chimbo. The Rio Patia in southern Colombia rises near Popayan, flows between the Eastern and Western Cordilleras southwestward to about ninety miles north of the Equator, where it breaks through the Western Cordillera and flows north- westward to empty into the Pacific near Tumaco. A large southern tributary, the Guaitara, rises between the two Cordilleras about forty-five miles north of the ^ In addition to the fishes of the Riinac river, the following species occur here: Mquidens coeruleo- punctatus, Pimelodella yuncensis, Brycon atricaudatus, Philypnus maculatus, and Astroblepus rosei. Basil- ichthys semotilus is not found so far north. The first and third are outriggers of the Guayas fauna. EIGENMANN: fishes of northwestern south AMERICA. 9 Equator and flows between them until it joins the Patia, where the latter bends from a southwestern to a northwestern course. The Rio Mira, with a length of about a hundred miles, flows northwestward, emptying into the Pacific at the northern border of Ecuador. The Esmeraldas, with a general trend nearly parallel to that of the Mira, drains the parks about Quito and empties into the Pacific approximately sixty miles southwest of the mouth of the Mira at 1° North. The rivers emptying directly into the Pacific between the Esmeraldas and the Guayas are all small, the largest of them, the Rio de Chone and the Rio de Portoviejo, are less than forty miles long, measured from source to mouth. South of Portoviejo, the country is dry and the rivers are still shorter. In the area between Cuenca and the coast the streams of the Atlantic slope rise within about thirty-five miles of the Pacific coast. According to Wolf and to Sievers, the Coastal Cordilleras between Esmeraldas and Guayaquil reach a height in places of two thousand three hundred feet. In the north about Esmeraldas and Manabi they are of late Tertiary and Quaternary origin. Southward about Portoviejo they consist of older formations. The youngest land of Ecuador lies between the Coastal Cordilleras and the Western Cordillera. Even as late as Quaternary times, the present Guayas Basin was a gulf reaching from Machala to the base of the Cordilleras. This gulf has been largely filled by sediments, thus forming the present Guayas Basin. The Cretaceous mountains of the Coastal Cordilleras reach a height of two thousand three hundred feet, while the Tertiary portions reach a height of from six hundred to one thousand feet. The rolling land of the Quaternary has an elevation of from sixty to two hundred and fifty feet. Between the Coastal Cordilleras andThe Western Cordil- lera there are a number of characteristically lowland streams with a north and south trend. The southern part of the former gulf is drained through the Vinces, Caracol, Chimbo, and Barranca Alta Rivers into the southward flowing Guayas, which is an extension of the Rio Vinces. Paralleling the Vinces, the Daule drains the area west of it to within about thirty miles of the coast. Colombia. In Colombia conditions become complicated. The western Andes of Ecuador are continued through the whole of Colombia to Cartagena. The eastern Andes, as the Cordillera Central, are also continued through the whole of Colombia to Santa Marta, but are cut in two by a great fault occupied by the valley of the lower Cauca and the lower Magdalena. 10 MEMOIRS OP THE CARNEGIE MUSEUM. It has been questioned whether the Sierra of Santa Marta really belongs to the Eastern Andes. If not, the Central Cordillera ends south of the junction of the Cauca and the Magdalena. The two chains coalesce near Medellin in central Colombia. South of Popayan the Patia valley between the two old chains of the Cordilleras is drained into the Pacific. North of Popayan it is drained by the Rio Cauca, which starts in a high interandean plateau about Popayan^ and flows to Cartago, where it begins a turbulent course through the knot of the Western and Central Cordilleras to Caceres, from which place it flows more gently to the Magdalena at a point where in former times it probably emptied into a bay similar to the present Lake Maracaibo. The complications in Colombia are due to the formation of two younger chains of Cordilleras. One of these is the Cordillera Oriental of Colombia or the Cordillera of Bogota. These Eastern Andes and the plains of Bogota have been studied by Hettner (^‘Die Kordilleren von Bogota,’^ Petermann’s Mittheilungen, Erganzhft. No. 104, 1892). He finds that the Cordillera of Bogota begins between latitude 1° and 2° North, as a range of low hills joined to the Eastern Cordillera of Ecuador. These hills are cut through by the tributaries of the Amazon flowing from the Eastern Cordillera. They gain in height at 2° and are no longer crossed by streams. The upper Magdalena has cut into these Cordilleras lengthwise, so that it runs between two of the chains as far north as Honda. At Honda the Magdalena cuts through the westernmost chain of the Cordillera of Bogota and flows in the depression between the Central Cordillera and the Cordillera of Bogota. Towards the north the Cordillera widens and then divides into several chains separated by plains.. The westernmost of these is the Sierra de Peri j a, which extends to the Sierra Nevada de Santa Marta; the middle chain is replaced on the north by Lake Maracaibo; the eastern and largest chain becomes the Cordillera de Merida, which ends at the depression of Barquisimeto. Beyond this are the Caribbean Mountains, which form the coastal chain of Venezuela. The Cordillera of Bogota, aside from a few Quaternary deposits laid down ^ Veatch (“Quito to Bogota,” p. 139) says: “Timbio is in the drainage basin of the Patia River at an elevation of 5,900 feet, while Popayan is on a tributary of the Cauca a few miles to the north at an elevation of 5,600 feet and, as we rode along, we were naturally on the alert to determine exactly when we crossed the inter-oceanic divide, but were unable to do so. The slight rise north of Timbio is very flat-topped, and from anything we could see from the trail the Rio Timbio could quite as well have joined the Cauca as the Patia.. The intervening elevation is much less striking than the “Cuchilla de Dolores,” which lies between two tributaries of the Patia. “After a time we saw in a slight depression in the plain the white walls and red roofs and church- towers of Popayan, and not till then did we know that we had crossed the divide.” EIGENMANN: fishes of northwestern south AMERICA. 11 after the formation of the Andes^ consists probably entirely of Cretaceous sedi- mentary rocks. The youngest strata, the Guaduas layers, may be Tertiary, but this has not been determined paleontologically and is doubtful. A few rocks older than the Cretaceous are the crystalline schists with quarzite and quarzitic con- glomerate, seen near Quetame. The southern part of the Cordillera of Bogota consists of erect or slightly inclined mountainous folds, comparable to the Jura. The westernmost areas consist of the Guaduas layers and are the youngest. During the entire Cretaceous, and perhaps during part of the Tertiary, the region was submerged. The formation of the mountains probably began in the Tertiary. The fourth of the Cordilleras of Colombia is the Coastal Cordillera. This is the youngest of the great mountain chains of Colombia. Hettner (“Die Anden des westlichen Columbiens,” Petermann’s Mittheilungen, 1893, p. 129) has also studied this chain. According to his observations, the Coastal Cordillera begins at the Bay of Buenaventura and extends northward through more than three degrees to the slight depression occupied by the Truando River, which separates it from the mountains of Darien. It reaches a maximum height of eighteen hundred meters, but ordinarily does not exceed one thousand meters. Near the middle there are really two chains, between which the Baudo flows. The western slope is very steep, the distance between the crest and the ocean being in some places only a few hundred feet. According to Karsten the rocks of the Coastal Cordilleras contain, as fossils, species of shells and corals, Fig. 1. Shores of the Atrato. The tall tree has more than fifty pendent bird-nests upon it. 12 MEMOIRS OF THE CARNEGIE MUSEUM. which still live in the Pacific Ocean at the base of these mountains. It seems that the Coastal Cordilleras are of late Quaternary origin. There are no crystalline rocks. If Hettner is right about the age of the Coastal Cordilleras, then the streams at present flowing westward from the Western Cordillera such as the Calima, Cucurrupi, Jujiado, Sipi, Tamana, Condoto, and,, upper San Juan rivers of the Pacific drainage, and the Raspadura, upper Atrato, Certegui, Yurri, and Sucio rivers now of the Atlantic, flowed directly into the Pacific at no very remote period. Along with the formation of the Coastal Cordillera, there was formed the trough between them and the Western Cordillera, which is indicated at present by the Gulf of Uraba on the north and the Bay of Buenaventura on the south. Fig. 2. The Cauca River near Cali (Juancita). Rafts of bamboos. The rivers Atrato and San Juan must have been later developments: the Atrato gathering the waters of the Raspadura and the streams north of it flowing from the Western Andes and emptying them into the Caribbean; the San Juan gathering the waters of the streams between the upper San Juan and the Rio Calima flowing from the Western Cordillera and emptying them into the Pacific. It seems then that the oldest of the present rivers of Western Colombia is the Cauca. The Magdalena, which is the largest river, developed with the forma- tion of the newer Cordillera of Bogota. The youngest rivers are the San Juan EIGENMANN: fishes of northwestern south AMERICA. 13 and the Atrato. A lowering of Colombia north of Buenaventura by as little as two hundred feet would convert the valleys of the San Juan and the Atrato into two long bays, or a strait, and cause the Magdalena, the Cauca, and the Cesar Rivers to empty independently into a lake or great bay, like Lake Maracaibo, extending from Santa Marta to slightly above El Banco. The Colombian Reconnaissance. Accompanied by Mr. Arthur J. Bierhaus, I reached Cartagena on January 3d, 1912. It had been my intention to ascend the Rio Atrato, descend the Rio San Juan, go eastward over the Quindio Pass to Girardot, descend the Rio Magdalena, and sail home from Cartagena. A side trip to Bogota was planned. However, when I reached Colombia, there was an unusual drought and the route had to be reversed. There is no fresh water about Cartagena and at the earliest moment we left for Soplaviento on the Dique. Thence, I went to Calamar on the Magdalena. From Calamar the expedition went up the Magdalena River by steamer to La Dorada, collecting at various stopping places, Barbosa, El Banco, Canaletal, Puerto Wilches, Penas Blancas, and Puerto Berrio. From La Dorada the route was by rail to the upper part of the Magdalena, collections being made on the way at Honda, especially in Bernal Creek. The upper part of the Magdalena was followed to Girardot, where extensive collections were made. From Girardot the route led first over the western rim of the plain at an elevation of about 8,800 feet to Bogota on an elevated plain among the eastern Cordilleras. Collections were made on the plain near Puente la Suba, north of Bogota, and at Madrid near the Fig. 3. Narrows of the Rio Magdalena below Girardot 14 MEMOIRS OF THE CARNEGIE MUSEUM. western margin of the plains of Bogota (about 8,500 ft.). A return was made to Girardot, from which place a pack-train conveyed the expedition via Chicoral to Quatro Esquinas, Ibague, Toche, across the Quindio Pass of the Central Cor- dillera (at an elevation of 11,200 feet) to Boquilla, Piedra Moler, and Cartago near the Cauca River. The Cauca being too low for the regular steamers, the pack-train was used through the Cauca valley by way of Paila, Buga La Grande, Buga to Cali, collections being made at Paila and at Cali and in the Cauca near Cali. By still another pack-train the Western Cordillera was crossed. Near Cali the continental divide is at an elevation of 6,000 feet. After collecting at Caldas (3,722 feet) the valley of the Dagua was descended by rail, collections being made at Cisnero (1,046 feet), at Cordova (120 feet) and in the tidal waters of this river. From Buenaventura, on the Pacific coast of Colombia, a steamer was taken up the San Juan River to Puerto Negria; thence a dugout and a crew of Indians carried the expedition as far as Istmina. Collections were made at both the latter places and half-way between them. From Istmina, after a ride of two hours up a little stream and across the low continental divide (300 feet), the valley of the Atrato was entered near Tambo. By dugout the settlement of Boca de Raspadura Fig. 4. Tambo, Atrato basin. was reached. Thence the Raspadura was followed into the Quito River, which was traversed to Quibdo. Collections were made at Boca de Certegui and near the EIGENMANN: fishes of northwestern south AMERICA. 15 towH of Quibdo, at the junction of the Quito River with the Atrato. From Quibdo a specially chartered steamer was taken to Rio Sucio, where additional collections were made. From Sucio a steamer carried the expedition back to the starting point at Cartagena. The reconnaissance outlined above was made possible through the hearty cooperation of Dr. William Lowe Bryan, President of Indiana University. The Trustees of the University granted me leave of absence for the purpose of the trip. President Bryan joined me in giving a note to the Bloomington National Bank for a sum advanced toward expenses. When it was found that the expenses would exceed the original estimate, he borrowed an additional sum and cabled it to me in Colombia. The Bloomington National Bank kindly lent the money without interest. . _ . On my return from Colombia the Carnegie Museum purchased the first series and the duplicates of the collection, after the second series had been set aside for the Museum of Indiana University, and thus covered more than the entire expenses of the trip, with the understanding that it was to be my duty to prepare a report on the expedition. Everywhere along the line of travel, I met with the most courteous cooperation on the part of citizens of. Colombia and on the part of others traveling, or tempor- arily in residence. Among those who deserve special mention are: Mr. W. E. H. Diekin, Mr. Thomas Miller, Mr. Harry D. Cutbill, Dr. Felipe Zapata, and Dr, R. A. Salas of the railroads of Colombia, all of whom furnished me with free transportation. Mr. Henri Banneau, a commercial traveler from Paris, who was familiar with all the traveled parts of South America, became enthusiastic over the fishing. Under his guidance the boat-crew on the steamer up the Magdalena secured valuable material. At Honda and about Bogota he himself entered actively into the work of collecting, and between Calamar and Bogota he relieved me entirely of the vexatious handling of my baggage. Mr. L. M, Monsanto of New York kindly acted as interpreter during the earlier part of the journey. Brother Apolinar Maria at Bogota secured guides to the best fishing-places. Mr. Edward H. Mason of Cali helped in various ways both before and after I landed in Colombia. Mr. J. A. Mayolo of Buenaventura and the steamship company he represents granted me special favors; I am further indebted to Mr. Mayolo for letters of introduction and other courtesies. The complete itinerary follows. Wherever it was possible, if the steamer on the Magdalena stopped for but an hour, collections wece made: January 3d-10th, at Cartagena, in the bay. llth-lSth, at Soplaviento, in the 16 MEMOIRS OF THE CARNEGIE MUSEUM. Dique. 14th-15th, at Cartagena (in connection with the Custom House). 16th- 18th, at Calamar, in the Rio Magdalena. 19th-27th, on S. S. ‘‘Neiva” to La Dorada. Collections were made: 20th, at Barbosa; 21st, at El Banco; 22d, at Bodega Central; 23d, at Penas Blancas; 24th, at Puerto Berrio (‘’elevation 429 feet)”; 27th, at La Dorada. 28th, in Bernal Creek at Honda; 29th, at Rios Perico and Guarino, six miles below Honda. Fig. 5. The Magdalena at Girardot. February 1st, Girardot, Rio Magdalena. 2d-3d, Bogota. 4ih, Puente La Suba north of Bogota. Sth, in Zerrezuela river at Madrid. 6th, at Falls of Te- quendama. 8th-llth, at Girardot ’on the Rio Magdalena. 12th, at Chicoral. 13th, in Rio Gualandai (1,566 feet). 14th-16th, at Ibague. 16th-17th, were spent on road to Toche. 18th-19th, spent in crossing the Quindio Pass (11,200 feet) to Boquilla, Rio Quindio (5,725 feet). 20th, at Balsa. 21st, at Piedra Moler, Rio Viejo. 22d-23d, at Cartago (3,012 feet). 24th, at Saragosa. 25th, at Paila. 26th, at Buga la Grande. 27th, at Buga. 28th, at La Torre, on the Cauca. 29th, at Cali (3,312 feet). March lst-3d, Rio Cauca and Rio Cali. 4th, Caldas, Rio Dagua (3,722 feet). 5th, Cisnero, Rio Dagua and Rio Pepita (1,046 feet). 6th, Cordova, Rio Dagua (120 feet). 7th-12th, at Buenaventura. 13th, up the Rio San Juan. 14th-15th, EIGENMANN: fishes of NOKTHWESTERN south AMERICA. 17 Puerto Negria, on Rio San Juan. 16th, Island, La Cruce, Rio San Juan. 17th, Depulcito, Rio San Juan. 18th-20th, Istmina, on Rio San Juan. 21st-22d, Boca de Raspadura. 23d Managru, near Boca de Certegui. 25th-27th, Quibdo. 28th, down the Rio Atrato. 29th-30th, Rio Sucio. April 2d, Cartagena, di/i, sailed for New York. The Landon-Fisher Expedition of Indiana University to Colombia. A second expedition to Colombia was made possible by Mr. Hugh McK. Landon and Mr. Carl G. Fisher, of Indianapolis. Mr. Arthur W. Henn and Mr. Charles E. Wilson, both undergraduates in Indiana University, left in December, 1912. They landed at Tumaco, near the south-western comer of Colombia. After devoting about a month to the Telembi River, a tributary of the Patia, they separated. Mr. Wilson went to the San Juan River, collecting in the upper San Juan Basin, the Condoto River at Condoto, and in the San Juan River, Istmina, and Tado of the Pacific side, and on the Atlantic side at Tambo, Raspadura, Boco de Raspadura, Managru, Quibdo, in the Atrato between Quibdo and Rio Sucio, and especially in the Truando River emptying into the Atrato near Rio Sucio. Mr. Wilson writes: “In all places I received the most courteous treatment and kind cooperation from the citizens of Colombia and from others traveling or in temporary residence. Those who deserve special mention are: Mr. J. A. Mayolo of Tumaco, to whom I am indebted for letters of introduction throughout the Choco, for passes on his steamers on the Patia and Telembi Rivers, and for many other courtesies during my stay in Tumaco. Captain Maximilian Llorente and Mr. Andres Knudson of Barbacoas aided the expedition by securing men and canoes for fishing. Mr. Knudson kindly acted as interpreter, and Captain Llorente gave us rooms in his home during the stay in Barbacoas. Mr. Jorge Mercado and Mr. G. Otero Vazquez of Buenaventura relieved me from looking after the trans- portation of my baggage. Mr. Thomas A. Copeland of Istmina kindly acted as interpreter and cared for me from Buenaventura to Istmina. Others in Istmina, who deserve special mention, are Mr. Antonio Asprilla, who secured men and canoes for collecting; and Mr. Andres Arizala, who loaned me his servant for the work on the Condoto River. Mr. C. J. Indii and Mr. Frank Ocoassen of Pefialisa furnished men and canoes and aided with the collecting. In Quibdo Mr. Tufik Meluk was of the greatest assistance in securing men and canoes for the expedition. I am also indebted to Mr. Meluk for letters of introduction. In Rio Sucio Mr. Ed. Kromer kindly acted as interpreter and greatly aided the expedition with the 18 MEMOIRS OF THE CARNEGIE MUSEUM. work in the Truando River. In Cartagena the A. & T. Meluk Co. kindly aided the expedition.” The Landon Expedition to Colombia and Ecuador. Through the continued liberality of Mr. Hugh McK. Landon of Indianapolis Mr. Arthur W. Henn was enabled to remain in South America and spend the time between February 15, 1913, and March, 1914, in collecting fresh-water fishes in Colombia and Ecuador. The following account by Mr. Henn summarizes his travels: “ On February 15, 1913, I left Tumaco for Barbacoas. Three days were spent in securing the neces- sary pack-animals; then I left for Tuquerres, arriving after five days’ travel. Tu- querres (10,090 feet) is on the Andean plateau. After diligent collecting a single species of Astroblepus was obtained. On March 4th, accompanied by Mr. Fridolin Hiimmerles of Tuquerres as an' interpreter and two arrieros or muleteers, I went northward from Tuquerres for the basin of the Upper Patia. Our route was in general that of the geologist, A. Stiibel. March 6th was spent in Ancuya (5,000 ft.) where a single speciles of Astroblepus was collected. Crossing the Rio Guaitara, our next stop was Los Llanos de Sandona (5,000 =t ft.), where a collection was made containing three species. Passing northward through Tambo and Penol, we arrived on March 13, at a place, composed of three huts, known as Guayabillo. This is directly on the brink of the canyon of the Patia. On this and the following day we descended some three thousand feet and collected in the Patia just above the mouth of the Guaitara. The elevation here, as determined by my barometer, was about 1,500 feet; the great swiftness of the river and the huge boulders made collecting difficult, but the collection is assumed to be representative. We returned to Tuquerres by way of Pasto and continued on to Barbacoas. After two days at Barbacoas, I left on the steamer ‘Bolivar’ for the mouth of the Rio Telembi, from which point I continued up the Patia by canoe. The Rio Magui, the first large tributary of the Patia above the Telembi, was ascended to the negro village of Payan. During the return to Barbacoas we fished in all available places. These localities may be somewhat arbitrarily given as (1) the Rio Magui, between its mouth and the village of Payan; (2) the Rio Patia between the Magui and the Telembi; (3) the Rio Telembi below Barbacoas. “After returning to Tumaco and waiting more than a week for the steamer, I sailed for Buenaventura, arriving there on April 23, 1913. After waiting in Buenaventura an additional week, I left on the small river-steamer “Buenaventura” for Puerto Negria, the head of navigation on the Rio San Juan. The following EIGENMANN: fishes of northwestern south AMERICA. 19 day with two negroes I left, going down-stream in a canoe. Because of flooded conditions fishing was desultory, and most of the fishes secured were obtained by barter with the Indians. Five days were occupied in drifting down the San Juan; the nights were spent in the occasional houses of the Indians. The last large tributary, the Bio Calima, was ascended. Thorough collections of fishes were made in two places: (1) a small creek of the Rio Calima just above the junction of the Calima with the San Juan; and (2) a point probably thirty miles up the Calima, known as Boca del Guineo. Ascending the creek El Guineo to its source, a short portage was made to another creek, known as San Joaquin, where a second canoe was taken to Buenaventura. “Leaving Buenaventura almost immediately, I arrived in Guayaquil, Ecuador, on May 16, 1913. A few days later, I left with Mr. R. B. Jones, an American residing in Guayaquil, for an hacienda owned by him just south of Naranjito. While here, collections were secured in a small creek known as Estero Verdes, tributary to the Rio Chanchan, which borders the hacienda, and in a clear, deep river, the Rio Barranca Alta, some two hours on horseback south of the hacienda. Later I sailed from Guayaquil for the Province of Manabi, arriving at Bahia de Caraquez, its chief port, on June 21. Going on the short railroad to Chone, I there made collections in the river, and then returned to Calceta, from which point I continued by horse to Portoviejo, where collections were also obtained. These rivers are small, traverse a very dry region, and contain few species of fishes. The return to Guayaquil was made by steamer from Manta. “I again left Guayaquil on July 17 in a launch for Daule, a small town on the Rio Daule. Delayed a few days by fever at this point, I continued in a launch to the village of Santa Lucia and then proceeded by canoe to another village known as Colimes higher up the river. Extensive collections were made at this point. Then returning to Daule in a large canoe, I embarked in a smaller one, and, after paddling two days along a winding cut-off from the Rio Palenque, I arrived at Vinces, where thorough collecting was done. “ Returning to Guayaquil, I went over the Guayaquil and Quito R. R. to Quito, stopping off for collecting at Huigra (4,000 ft.); Riobamba, Atlantic drainage (9,020 ft.); Latacunga, Atlantic drainage (9,055 ft.) and Quito (9,375 ft.). A short trip for collecting was made to Mindo (4,108 ft.), situated in a region of tremendous rainfall on the western slope of Mt. Pichincha. Somewhat later (October 18, 1913), I went north to El Angel (10,000 ft. dz) in the Province of Carchi, the most northerly province of Ecuador. Here Astroblepus was collected. From this point a trip was taken down the valley of the Rio Chota, or Mira, to a small property 20 MEMOIES OF THE CARNEGIE MUSEUM. two leagues below the hacienda Paramba, known as “Maria Luisa,” belonging to Sehor Cesar Mena of Angel. Unfavorable conditions made this trip almost a failure, nevertheless a good collection of Astroblepus was obtained on the return from a place opposite the settlement, Guallupi (5,000 ft. ±) farther up the Rio Chota. “ Being forced to return to Quito by revolutionary developments in the province of Carchi, I remained there until February 9, 1914, when I again went back to El Angel to secure the collections, continued on over the Colombian frontier, twice changing mules, and returned over the road to Barbacoas. Traveling as rapidly as possible, though long delayed by waiting for steamers at Barbacoas and Tumaco, and by the quarantine at Panama, I did not reach New York until the very last of March, 1914. “Throughout the course of the trip I met with numerous courtesies from resi- dents of the countries traversed. Chief among these persons may be mentioned Sr. Jose A. Mayolo of Tumaco, Colombia. Mr. Mayolo provided free transporta- tion on the ‘Linea Costanera Fluvial de Vapores’ for myself and my extensive outfit and collections on several trips between Tumaco and Barbacoas; in the Choco he gave me recommendations to the authorities, his friends, and business associates; assisted in employing laborers; and performed other innumerable services. His favors coming, as they did, when I knew but little of the language and customs of the country, were well-nigh indispensable. For the ‘Sierra,’ similar services were performed for me by Mr. John W. Bidlake, an American resident of Tuquerres. In Guayaquil Mr. R. B. Jones showed me a number of courtesies and Mr. H. W. Henderson of the Guayaquil and Quito Railway Company kindly provided free transportation to Quito. Sr. Alberto Santos of Bahia did me several favors. I was a guest at several haciendas; among which may be mentioned those of Sr. Cesar P. Garzon of Mindo, of Sr. Jose Tamaya, and of Sr. Cesar Mena of Carchi. The Panama Railroad Steamship Line courte- ously provided free transportation of freight from Panama to New York and reduced fare for myself. “The expenses of this Expedition were largely provided by Mr. Hugh McK. Landon of Indianapolis. He not only provided funds for the trip as originally planned, but later gave an additional sum, which allowed considerable modification and extension of the plans and enabled the expedition to remain in the field nearly four times as long as had been originally intended.” EIGENMANN: fishes of northwestern south AMERICA. 21 The Irwin Expedition. The Irwin Expedition of Indiana University, of 1918-19, carried the survey of the rivers of the Pacific slope from northern Peru to southern Chile. It also descended im the Atlantic side into the Huallaga Basin to two thousand feet, the Chanchamayo to two thousand feet, the Urubamba to three thousand feet, and surveyed the Titicaca Basin. The main account of the Irwin Expedition will appear with the reports on these regions. The portion of the Irwin Expedition, which comes within the present territory, deals with the Chira, Piura, Rimac, and Chili Rivers, enumerating these rivers in order from north to south, but not in the order in which they were examined. All four of these rivers belong to the dry Peruvian coast. The Chira River rises in southern Ecuador in the southernmost of the interandean parks of Ecuador and empties into the ocean north of Paita. I collected at Sullana early in January, 1919. The river had been swollen by recent rains in the mountains and collecting was difficult. The watep was muddy and fishes were scarce and scattered. A species of Curimatus was obtained here; and this is the southernmost record for the genus on the Pacific slope. The rest of the fauna was the same as that farther south at Pacasmayo, Fig. 6. Upper Rimac Valley. Conditions at Piura on the Piura River, only a few miles south from Sullana, were the reverse of those at Sullana. I reached Piura at the end of the dry season. On asking where the river was, I was told that it had not yet come down. The 22 MEMOIRS OF THE CARNEGIE MUSEUM. river-bed was damp in places and shallow excavations provided water for gardens and dwellers along its shores. There were a few pools in natural depressions about the pier and alongside of some rocks. These were covered with green slime and were crowded with starving fishes. A single species each of Pimelodella, Pygidium, Brycon, Bryconamericus, Lebiasina, and JEquidens formed the entire ffeuna. Both Sullana and Piura are reached by rail from Paita. The Piura River reaches the ocean south of Paita. The next stop farther south was at Pacasmayo. Con- ditions here are similar to those at Paita. The town clusters about a little rivulet, which always carries water and is affected by the tide. Behind the town there is a considerable lagoon. The Rio Jequetepeque rises in the Andes west of Cajamarca and empties into the ocean north of Pacasmayo. Collections were made at Llallan (elevation 2,437 ft.), at Cultambo, and in the outlet of the lagoon just above and below the locks controlling its level. Llallan is at the junction of two clear, swift streams. At Cultambo I collected near the railroad bridge. The water was clear with alternating swift stretches or even miniature rapids and quiet pools. The fauna was the same as that at Piura with the addition of an Astroblepus, which could not be expected to occur in the portion of the Piura I examined. Astroblepus is probably found nearer the mountains in the Piura River. A Liza is abundant in the lagoon, but we were not fortunate enough to get one. Pacasmayo was the southernmost locality for Mquidens, Pimelodella, and Brycon. We should have made collections in the Santa River, the largest and most constant river of Peru. However, several matters prevented its examination; the regular steamers do not stop at Chimbote at its mouth and the time needed for its exploration was therefore uncertain. The Rimac River south of the Santa con- tained everything we secured at Pacasmayo except Astroblepus, Brycon, AHquidens, and Pimelodella. Somewhere between Pacasmayo and Callao they disappear. Astroblepus is known to occur in the Santa Basin, which is the southernmost locality on the Pacific Slope. One question thus left unanswered is whether Brycon, AHquidens, and Pimelodella extend as far south as the Santa, or not. It will probably be found that they do. A second question is whether the “Peje rey,” Basilichthys semotilus, is found in the Santa? It is found in the Rimac, and has not been taken* in the Jequetepeque. Again a priori considerations would answer this question in the affirmative. A question exceeding the other two in interest, because its answer is more problematical, is whether the Santa, being the largest and most permanent of the Peruvian rivers, harbors a more abundant representa- tion of the ancient fauna of the Pacific slope of Peru? The answers to these questions must be left for the future. EIGENMANN: fishes op northwestern south AMERICA. 23 the fauna enormous The Rimac River, emptying just north of Callao, has been examined at various times. During the Irwin Ex- pedition collections were made at Puenta Piedra in a shallow lagoon north of the river on the railroad to- ward Ancon. The water was about “knee-deep” and full of fishes, prin- cipally Bryconamericus peruanus and Lebiasina himaculata. Many of these were badly infected with Saprolegnia and a peculiar wavy shortening and up and down distortion of the caudal portion of the spine. Collections were made in little side-branches and rivu- lets emptying into the Rimac at Lima. Four species, Lebiasina bimaculata, Bryconamericus peruanus, Pygidium punctulatum, and Basilichthys semo- tilus formed the total fauna at Lima, Chosica, and Matucana. At Chosica (elevation 2,800 ft.) the river is ca- nalized and very swift. A pool below the dam of the, electric-light plant in the city yielded the usual fauna with “ Cube,” a native poison. Basilichthys was abundant in the pool. Chosica lies at the foot of the gorge of the Ri- mac. Matucana lies in a little widen- ing of the gorge at about 7,800 ft. elevation. The usual fauna was se- cured during a stay of several days, devoted largely to becoming accli- mated to the elevation. At Rio Blanco at an elevation of ten thousand feet we secured nothing. Undoubtedly of the Rimac, especially in its upper parts, has become affected by the masses of waste daily dumped into it by the smelter at Casapalca. X