z 
 
 o 
 
, RUS PE LA IUHQUK , 29 
 
 Assortment uombreux t-l 
 rie de livres Anglais, Al- 
 Mands.Italions.Kspagiiols 
 irtugais.Langucs dii Nord 
 
\ 
 
SCIENTIFIC RESULTS 
 
 A JOURNEY IN BRAZIL. 
 
 BY LOUIS AGASSIZ, 
 
 AND HIS TRAVELLING COMPANIONS. 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY 
 OF BRAZIL. 
 
 BY OH; FRED. HARTT, l^ c 
 
 PROFESSOR OF GEOLOGY IN CORNELL UNIVERSITY. 
 
 WITH ILLUSTRATIONS AND MAPS. 
 
 BOSTON : 
 FIELDS, OS GOOD, & CO. 
 
 LONDON : 
 TRUBNER & CO., 8 & 60, PATERNOSTER ROW. 
 
 1870. 
 
,\ 
 
fltocroft Library 
 
 To PROFESSOR LOUIS AGASSIZ, 
 
 DIRECTOR OF THE THAYER EXPEDITION. 
 
 MY DEAR SIR : 
 
 I have the honor to offer you this volume on the Physical 
 Geography and Geology of Brazil as a summary of the scientific 
 results of my explorations as an attache of the Thayer Expedition, 
 together with those of a second private expedition, r- the natural 
 outgrowth of the former, made to continue investigations which 
 I had been obliged to leave unfinished. 
 
 I take this opportunity of acknowledging my deep indebtedness 
 to you for the interest you have taken in my scientific studies, for 
 your constant wise counsel and advice, and for a thousand kind 
 nesses received at your hands. 
 
 With the highest consideration and respect, I have the honor 
 to be, 
 
 My dear Sir, 
 
 Your former pupil, 
 
 CH. FRED. HARTT. 
 
 CORNELL UNIVERSITY, ITHACA, N. Y., 
 May 30, 1870. 
 
PREFATORY NOTE. 
 
 THIS volume is the result of two journeys made by myself 
 in Brazil. The first was undertaken under the direction of 
 Professor Agassiz in connection with the Thayer Expedition 
 in the years 1865 and 1866. On this journey I studied very 
 carefully the Geology and Physical Geography of the coast 
 between Rio and Bahia, going over a very large part of the 
 ground on horseback or in canoe. As a sketch of the jour 
 ney is to be found in the " Journey in Brazil," it is not ne 
 cessary to repeat it here. My companion was Mr. Edward 
 Copeland, of Boston, one of the volunteer aids of the Expe 
 dition, and I take pleasure in acknowledging my indebted 
 ness to him for the very valuable assistance he rendered me. 
 On this journey Mr. Copeland and I made very large collec 
 tions of marine invertebrates and fishes, though we did not 
 neglect to secure other objects of natural history. I hope 
 that these collections will throw much light on the fresh 
 water and marine animals of the coast, and, as they were 
 made at frequent intervals between Bahia and Rio de Ja 
 neiro, that they will prove valuable in the study of the 
 distribution of these animals. My studies of the stone 
 and coral reefs and of the geology of Brazil proved so very 
 attractive to me that on the year following I returned to 
 Brazil, and spent my vacation, several months, in examining 
 
Vi PREFATORY NOTE. 
 
 the coast between Pernambuco and Rio, exploring more 
 particularly the vicinity of Bahia and the islands and coral 
 reefs of the Abrolhos. I was aided pecuniarily on this 
 second expedition by Mr. John Lockwood, of the Adelphi 
 Academy, Brooklyn, the New York Association for the Ad 
 vancement of Science and Art, and the Cooper Institute ; 
 but my thanks are especially due to Miss Chadeayne, the 
 principal of an excellent ladies school in Jersey City, to 
 Mr. Van Nostrand, of Newark, to my friend Major O. C, 
 James, of Bergen, without whose generous aid I should not 
 have accomplished my expedition and brought home my 
 voluminous collection, and to Mr. J. E. Mills, and Mr. R. 
 L. Dugdale, of New York. 
 
 I wish that I could adequately acknowledge the innumer 
 able kindnesses and attentions I have received from Captains 
 Tinklepaugh and Slocomb, and the officers on the splendid 
 steamer of the generous Messrs. Garrison and Allen, and 
 from a hundred kind friends in Brazil. The hospitality with 
 which I was everywhere received in Brazil, and the assist 
 ance offered me wherever I went in the prosecution of my 
 studies of the country, have made me love the land of the 
 Sabia, and it is my sincerest wish in acknowledgment of 
 so much kindness to be to some humble degree instrumental 
 in removing false impressions so current about Brazil, and 
 to make the resources of the Empire better known in Amer 
 ica. In the course of the following pages I shall have 
 frequently opportunity to acknowledge the aid not only of 
 Brazilians, but of foreigners resident in the country. 
 
 This volume was intended at first as a report to Professor 
 Agassiz as the Director of the Thayer Expedition, embra 
 cing simply the results of my explorations as a geologist 
 of that expedition, together with those of my second inde- 
 
PREFATORY NOTE. Vll 
 
 pendent journey, both reports to be published among the 
 contributions of the Thayer Expedition. During the prep 
 aration of these reports, and in consequence of the delay in 
 publication, I have had the opportunity of examining more 
 or less critically the works of the majority of the writers on 
 Brazil, and the volume, from a simple report of my own in 
 vestigations, has grown to a general work, in which I have 
 incorporated the best results of others who have written on 
 the Geology and Physical Geography of Brazil. I have to 
 acknowledge valuable contributions to this volume from 
 Messrs. J. A. Allen, Orestes H. St. John, and Thomas Ward, 
 all of whom were employed on the Thayer Expedition. I am 
 indebted to Professor Jeffries Wyman for an interesting com 
 munication with reference to a Botocudo skull, and to Pro 
 fessor Alpheus Hyatt for his valuable paper on the Creta 
 ceous Fossils of Maroim ; Professor 0. C. Marsh has kindly 
 examined and described in the Journal of Science a few 
 reptilian remains I collected at Bahia, Professor Verrill de 
 scribed the radiates of my second journey, and Mr. S. J. Smith 
 has published a valuable paper on the Crustacea of the same 
 journey. Professor Marsh s paper I have given almost en 
 tire. From Professor Verrill s I have drawn largely, and I 
 have given the general results of that of Mr. Smith. A part 
 of the chapter on the Coral Reefs appeared in the American 
 Naturalist, together with several of the wood-cuts, which 
 have been kindly lent me for this volume. I must ex 
 press my thanks to Messrs. Putnam, Packard, Morse, and 
 Hyatt, of the Peabody Academy, for valuable assistance ren 
 dered. 
 
 Professor Agassiz has generously allowed me the use of 
 books, maps, and photographs, and has assisted me in vari 
 ous ways. I regret exceedingly that his illness has pre- 
 
yiii PREFATORY NOTE. 
 
 vented him from preparing for the work the paper on the 
 fishes which he contemplated. 
 
 In the recent excellent work " On the Highlands of the 
 Brazil," by Captain Burton, the celebrated African traveller, 
 now English Consul at Santos in the Province of Sao Paulo, 
 the author uses throughout the definite article prefixed to 
 the name of the Empire. Burton says, " I do not call the 
 country Brazil,- which she does not ; nor does any other 
 nation but our own." * Captain Burton s reasoning does 
 not seem to me quite conclusive. It is true that the Portu 
 guese say " o Brazil, the Brazil" but the article is not 
 prefixed to give any particular definiteness to the name. 
 They say also a Inglaterra, a Franca, o Paraguay, &c., the 
 definite article being applied to almost all names of places. 
 So we have in French le Bresil, la France, VAngleterre. In 
 Italian the definite article is used before the name of a 
 country when the whole of it is meant, but ordinarily it is 
 not employed. In Spanish it is used much as in French. 
 
 Since the Brazilians use the article before the names of 
 other countries as well as their own, in accordance with a 
 custom followed by other Romance languages, there would 
 seem to be no better reason for saying Tlie Brazil, contrary 
 to English usage, than " TTie France" because the French 
 do. It is true that many English writers on Brazil have used 
 the article, but the majority have not, and no one but Bur 
 ton uses it to-day. " The Brazils" as that author has re 
 marked, is an anachronism occasionally seen, but only 
 proper between the years 1572 and 1576, when the country 
 was divided into two governments. It is more high-sound 
 ing than the ordinary form, and so probably remains in use. 
 In the North European languages, German, Hollandish, 
 
 * Vol. I. p. 3, note. 
 
PREFATORY NOTE. IX 
 
 Danish, or Swedish, the name is derived from the Latin form 
 Brasilia; * in Danish and German, Brasilien. In like man 
 ner we have the German Italien, Sicillen, &c. In none of 
 these languages is the article used. I have followed in this 
 work the common usage, and have omitted the article. 
 
 The use of the article before the names of places in Brazil 
 is very puzzling, and foreigners writing on the country are 
 sure to commit blunders. The names of the provinces are 
 especially difficult. Several of them are derived from rivers, 
 as Amazonas, Para, Parahyba, &c., and these take the article 
 in Portuguese, as do Ceara, AlagSas (plural), Bahia, and 
 Espirito Santo ; but those bearing the names of saints, to 
 gether with Pernambuco, Sergipe, Minas Geraes, Matto 
 Grosso, and Goyaz, do not. The same difficulty is met with 
 in the names of rivers, serras, &c., and it is impossible to 
 give a general rule to guide one in writing them. Most 
 writers seem to be impressed with the idea that Portuguese 
 is only bad Spanish, and that it will do just as well to write 
 Brazilian names in the Spanish form ; so we find some of 
 our best authors on Brazil using San Francisco for Sao 
 Francisco. The only safe way seems to be to give geogra 
 phical names exactly as used by the Brazilians themselves. 
 
 Brazil, in its climate, people, and productions of all kinds, 
 was at the time of its discovery totally different from Europe. 
 The European colonists in Brazil therefore had no names to 
 give to the things they saw about them. Such was not the 
 case in North America, where the early explorers found 
 animals and plants resembling those of Europe, and they 
 recognized the bear, the wolf, the codfish, the herring, the 
 oak, &c., &c. But in Brazil all was new, and the euphoni- 
 
 * Hakluyt used the form Brasilia, and it is to be found in the works of other 
 old English wriiers. 
 
X PEEFATORY NOTE. 
 
 ous indigenous names were adopted and incorporated, and 
 to-day the Portuguese of Brazil is full of them, and places, 
 too, bear to a very large extent their Indian names. Par and 
 wide, up and down the coast, and through the country, was 
 found distributed the great nation of the Tupfs, speaking 
 everywhere the same general language, and it is from this 
 language, now spoken over a large part of Brazil, that these 
 names were taken. These names have been written with 
 the Portuguese pronunciation, and have often been much 
 corrupted, so that their orthography varies very much. 
 Foreigners murder them fearfully. Among the writers on 
 Brazil Bates and Burton are perhaps the most accurate in 
 their use of Portuguese and Tupi names. In this work I 
 have followed the best authorities, and I have taken particu 
 lar pains to insure correctness in the geographical names ; 
 but since there is no standard of orthography for these 
 names even in Brazil, and since one is frequently obliged to 
 depend on an ignorant guide for the name of a place or 
 an object, absolute uniformity and accuracy are out of the 
 question. 
 
 In looking up the derivation of Brazilian geographical 
 names, I was led into a study of the word " Brazil," but I 
 soon found that Humboldt had preceded me, and I am able 
 to add little to what he has said on the subject. Since there 
 are afloat so many incorrect ideas concerning the deriva 
 tion of the name, it seems not unadvisable to give here in 
 brief the result of Humboldt s researches.* 
 
 According to Humboldt the name, under the various forms 
 of Bracie, Brazil, BerzilJ appears on Italian maps from 
 
 * Gc ographie du nouveau Continent, Tom. II. p. 214, to which the reader is 
 referred for the details of the discussion. 
 
 t Among the many curious old forms of the name Brazil may be mentioned 
 
PREFATORY NOTE. II 
 
 1351 to 1459, applied to one or more of the islands of the 
 Agores, and more particularly to a point of the island of 
 Terceira, which still bears the name. 
 
 For three centuries before the discovery of the route to 
 the Indies around the Cape of Good Hope there was known 
 in Europe a dye-wood called bresill, brasilly, bresilji, braxi- 
 lis or braslle, which appears to have been derived from one 
 or more East Indian species of Ccesalpina and Pterocarpus. 
 
 As to how this dye-wood came to bear the name of Brazil 
 I know not, and I fear that any attempts to derive it from 
 Sanskrit or other roots will lead to nothing satisfactory.* 
 Anghiera speaks of the occurrence of Brazil wood in Haiti, 
 which was known long before the discovery of Brazil. 
 Grinoeus speaks in 1499 of the Brazil-wood seen at Paria 
 (Payra). 
 
 Humboldt says : " In proportion as discoveries extended 
 themselves to the south of Cape Santa Augustinho, especially 
 after Pedro Alvarez Cabral, in May, 1500, had taken posses 
 sion of the Terra de Santa Cruz, the commerce in the red 
 wood of Continental America became more active. On the 
 fourth expedition of Vespucius, in which one of the ships 
 was lost on the shoals around Fernando Noronha, a cargo 
 of Brazil-wood was taken in near the Bahia. All the world 
 
 that of Presill, found in fln ancient publication, described by Humboldt in his 
 Geographic du nouveau Continent, Tome V. p. 239, entitled Cop! a der Newen 
 Zeytuncj auss Presillig Latidt. 
 
 * Sec Humboldt, op. cit. Tome II. p. 222. It is always unsafe to proceed 
 m the investigation of the etymological derivation of proper names, unless one 
 has historical evidence of some sort to guide him. 
 
 t Decade I. Liv. 4. p, 11, quoted by Humboldt. Anghiera says that there 
 were found at Haiti " sylvas immensas, quaj arbores nullas nutriebant alias 
 prsctaqnam coccineas quarum lignum mercatores Itali verzlnum, Hispani bra- 
 silum appellant." 
 
Xii PREFATORY NOTE. 
 
 knows that little by little, in the first half of the sixteenth 
 century, this same abundance of dye-wood has caused the 
 name of Terra de Santa Cruz, given by Cabral, to be changed 
 into the Terra de Brasil, changement inspir par le de"mon, 
 dit Fhistoriographe Barros, car le vil bois qui teint le drap 
 en rouge ne vant pas le sang versd pour notre salut. So 
 from the Asiatic Archipelago the name Brasil has passed 
 to a cape of the island of Terceira, and then to the southern 
 shores of the New Continent." 
 
CONTENTS. 
 
 CHAPTER I. 
 
 THE PROVINCE OP RIO DE JANEIRO. 
 
 PAGE 
 
 The Serra do Mar. The Serra da Mantiqueira, and the Pico do Itatiaiossii. 
 The Rio Parahyba do Sul and its Tributaries. Description of the Bay of 
 Rio, its Islands, Tides, &c. The Sugar-Loaf. The Corcovado. The 
 Gneiss of Rio. The Organ Mountains. Geological Observations along 
 the Cantagallo Railroad. Drift Phenomena at Rio, Tijuca, and on the Dom 
 Pedro II. Railroad. Decomposition of Gneiss in situ and its Effect on the 
 Forms of the Hills. Recent Rise of the Coast. The Coast between Rio and 
 Cape Frio, its Lakes, Salines, &c. Cape Frio. Os Buzios. Islands of 
 Santa Anna. Frade de Macahe. Campos dos Goitacazes, their Lagoons, 
 Canal, &c. Rio Parahyba. Sao Joao da Barra. Sugar Plantations. 
 City of Campos. The Rio Muriahe". Tertiary Beds. Sugar Fazendas. 
 Sao Fidelis. The Gold-Mines of Cantagallo. Geological Notes on the 
 Country between Sao Fidelis and Bom Jesus. The Rio Itabapuana. The 
 Serra d Itabapuana. The Garrafao. Barra do Itabapuana ... 1 
 
 CHAPTER II. 
 
 PROVINCE OP ESPIRITO SANTO. 
 
 Barreiras do Sirf. Itapemerim. Coast between Itape merim and Benevente. 
 
 Benevente. Guarapary ; Consolidated Beach, Corals, &c. Rio Jecii. 
 Bay of Espirito Santo. Nossa Senhora da Penha. Victoria. Decompo 
 sition of Gneiss and Formation of Boulders of Decomposition. Recent Rise 
 of the Coast. Corals, &c. of the Bay of Victoria. Rio Santa Maria. 
 German Colonies. Fisheries. Sand Plains. Tertiary Plain at Carapina. 
 
 Mestre Alvaro. Serra. Nova Almeida. Rio Reis Magos. Santa 
 Cruz. Basin of the Rio Doce. Description of the River. Guandii ; its 
 Colony and Agricultural Resources. Porto de Souza. Geology of Vicin 
 ity. Luxuriance of Vegetation on the Doce. Woods. Game. Fran- 
 cylvania. Climate of the Doce. Linhares. Lagoa Juparanaa. The 
 Future of the Doce. American Colonists. Salt Trade. Barra Secca. 
 Sea-Turtles. Consolidated Beaches and the Mode of their Formation. 
 Character of Coast between the Rivers Doce and Sao Matheos. Rio Sao 
 Matheos described. Geological Features. Fertility of its Lands. Cocoa- 
 Palms and their Distribution. City of Sao Matheos. Rio Itahunas. 
 Cliffs of Os Lencoes. Coast between Itahunas and Rio Mucury . . .56 
 
XIV CONTENTS. 
 
 CHAPTER III. 
 
 PROVINCE OP MINAS GERAES. THE MUCURY AND JEQUITINHONHA 
 
 BASINS. 
 
 The Basin of the Mucury. Porto Alegre. Description of the River below 
 Santa Clara. Luxuriance of Forest Vegetation. Santa Clara. Minas 
 Geraes a Land-locked Province. Want of Roads. The Philadelphia Road 
 and the Mucury Colonies. Difference in Topography and Soils between 
 the Tertiary and Gneiss Lands west of Santa Clara. Urucii, its Dutch Col 
 ony, Soils, Climate, &c. Philadelphia and its German Colonies. Great 
 Fertility of the Mucury Basin. Character of Country between Philadel 
 phia and the Head-waters of the Mucury. The Basin of the Jequitinhonha. 
 The Rio Pardo. General Geological Structure of the Jequitinhonha- 
 Pardo Basin. The Head- waters of the Setubal, their Geological Features 
 and Catinga Forests. Geological Excursion from the Fazenda de Santa 
 Barbara to Alto dos Bois. Difficulty of geologizing in Brazil. The Bra 
 zilian Campos. The Chapadas between Itinga and Calhao. The great 
 Calhao- Arassuahy Valley. Magnificent View over the Valley from the 
 Chapada at Agua da Nova. Calhao and the Geology of its Vicinity. De 
 scription of the Country between Calhao and Sucuriu. The Chapadas. 
 Minas Novas, its Geology, Gold-Mines, &c. Occurrence of Gold in Drift. 
 Gold-Mines of the Arraial da Chapada ; their former Richness : not yet 
 worked out. Decomposition of Clay Slates in the Minas Novas Region. 
 The Rio Arassuahy. The Rio Jequitinhonha from its Confluence with the 
 Arassuahy to the Sea described : its Geology, Vegetation, Commerce. 
 The Salto Grande 125 
 
 CHAPTER IV. 
 
 THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 
 
 The Geology of the Abrolhos. Trap-bed, Fossil Plants, &c. Land Fauna 
 and Flora; Spiders, Lizards, and Sea-Birds. The Cemetery of the Frigate- 
 Birds. The Whale and Garoupa Fisheries. Importance of these Fisheries. 
 The mythical Brazilian Reef. The Coral Reefs and Consolidated Beaches 
 confounded by Travellers and Writers. The Author s Discovery of the 
 Porto Seguran Coral Reef. Coral-building Corals found almost wholly to 
 the north of Cape Frio. The Fringing Reef of Santa Barbara ; its Struc 
 ture and Life. Corals found on the Reef. Star-fishes, Ophiurans, &c. 
 Resemblance between the Echinoderms of the Abrolhos and West Indies. 
 The ChapeirCes. The Parcel dos Abrolhos ; its Appearance ; forms a seri 
 ous Obstacle to Navigation. Safe Canal west of the Islands. The Parcel 
 dos Parades. The Recife do Lixo. Its great Extent. The Submerged 
 Border and its Coral Growth. The Coral Fauna of Brazil. The Mille- 
 pores and their Stinging Properties. The Reefs of Timbebas, Itacolumf, 
 Porto Seguro, Santa Cruz, Camamu, Bahia, Maceio, and Pernambuco. 
 The Roccas . 174 
 
CONTENTS. XV 
 
 CHAPTER V. 
 
 PROVINCE OP BAHIA. COAST SOUTH OF SAO SALVADOR. 
 
 Tne Tertiary Lands between the Rivers Mucury and Peruhype ; their Vegeta 
 tion, &c. Colonia Leopoldina and its Coffee Plantations. Villa Vicosa. 
 The Canal joining the Rivers Peruhype and Caravellas. Formation of 
 Beaches and Beach Ridges. Coast between Caravellas and Porto Seguro. 
 Monte Pascoal. Porto Seguro and its Recife, or Consolidated Beach. 
 Santa Cruz and its Reef. Coast northward to the Jequitinhonha ; the La 
 goa do Braco, Campos, &c. The Canal Po-assu and the Rib da Salsa. 
 Mangrove Swamps between the Jequitinhonha and Pardo. Cannavieiras 
 The Salt Trade of the Jequitinhonha. Description of the lower Part of 
 the Rio Pardo ; Cacao Plantations, &c. Coast northward to Ilhe os. 
 Prince Neuwied s Description of the Country between Ilhe os and Con- 
 quista, Possoes, and Cachoeira ; the Forests, Campos, Social Plants. &c. 
 Ilheos. Rio and Lagoa Itahype. Dead Coral Banks. Rio das Contas. 
 Bay of Camamii. Turva Deposits. Villa de Camamu. Coast northward 
 to the Bahia de Todos or Santos. The Bay of All Saints described. Ilha 
 Itaparica. Rio Jaguaripe and Nazareth. Rio Paraguassu. Description 
 of River below Cachoeira. The Tram-road. Sant Amaro and the Agri 
 cultural Institute 215 
 
 CHAPTER VI. 
 
 THE SAO FRANCISCO BASIN. 
 
 The Explorations of Halfeld, Liais, St. John, Allen, Ward, Burton, &c. Gen 
 eral Shape of the Basin. Its uniform Width. The Sao Francisco Valley 
 hollowed out of a Series of horizontal Beds of Limestone and Sandstone. 
 The Chapadas. The so-called " Serra" separating the Sao Francisco from 
 the Tocantins Basin an irregular Strip or Table-land of Sandstone. The 
 Serras of Araripe and Dous Irmaos. Table-topped Hills in the Valley of 
 the Sao Francisco; Outliers of the Chapadas. Doubts about the Age of the 
 Sandstones and Limestones. Limestones of the Rio das Velhas. Remains 
 of Extinct Quadrupeds in Brazil, spoken of by Cazal, Spix and Martins, &c. 
 Claussen s Discoveries in the Caves at Curvelo. Dr. Lund s exhaustive 
 Researches at Lagoa Santa. Caves described ; their Number, Extent. Sta 
 lactites and Deposits of Bones in Saltpetre Earth. Immense Quantities of 
 small Bones brought in by Owls, &c. Large Number of Fossil Animals dis 
 covered by Lund. Former Existence of Megatheria, Mylodons, Mastodons, 
 immense Armadillos and Cats, Horses, &c., in Brazil. Remains of a Race 
 of Man of high Antiquity. Reinhardt s Generalizations. The Rio de Sao 
 Francisco of the Sixteenth Class among the Rivers of the World, but third 
 in Rank in Brazil. General Description of the Stream. Its Affluents, 
 the Rios Para, Paraopeba, and Das Velhas. The Rio das Velhas alone 
 capable of being made navigable for Steamers. The Siio Francisco naviga 
 ble with but few Interruptions for two hundred and sixty-four Leagues below 
 
XVI CONTENTS. 
 
 the Rio das Velhas. Cost of removing Obstructions. Proposed Railway 
 from Joazeiro to Piranhas. Fertility of Low Lands of Sao Francisco Val 
 ley. Liais s Picture of the Campos . .274 
 
 CHAPTER VII. 
 
 THE PROVINCE OF BAHIA, INTERIOR. 
 
 Journeys of Spix and Martius, Nicolay and Lacerda, Allen, and other Explor 
 ers. Geological and Physical Features of Country between Malhada and 
 Cachoeira, described by Von Martius. Sandstones. Remains of Masto 
 dons found near Villa do Rio de Contas. Immense Copper Boulder from 
 Cachoeira. Rev. Mr. Nicolay s Report of Journey from Cachoeira to the 
 Chapada Diamantina. Occurrence of Diamonds in Sandstones. Lime 
 stones. Sterile Plains. Diamantiferous Sands of the Chapada. The 
 Diamond Mines of Sincora and Lencoes. Annual Yield of the Provinces in 
 Diamonds. Mr. Allen s Report of a Journey from Chique-Chique, via 
 Jacobina to Cachoeira. Country between Cinque -Chique and Jacobina an 
 immense Limestone Plain. The Chapada at Jacobina a detached flat- 
 topped Mass of Sandstone. Gneiss Hills. " Lake Plain," east of Jaco 
 bina. Knobs. Potholes, probably of Glacial Origin. Eastern Sandstone 
 Plain. Climate, Vegetation, &c. of Route. Difference in Topography 
 between Gneiss Regions of Bahia and the Mucury described and accounted 
 for. Former greater Extension of Forests. Von Martius s Description 
 of the Country between Cachoeira and Joazeiro. Country near Feira da 
 Conceicao. Serra do Rio Peixe. Rio Itapicuru. Want of Rain at 
 Queimados. Serra de Tiuba. Tanques and Fossil Bones near Coche 
 d Agua, Barriga Molle, and Neighborhood. Monte Santo. The Great 
 Meteorolite of Bemdego. Rock Inscriptions. Villa Nova da Rainha. 
 Joazeiro to be the Terminus of Bahia and Sao Francisco Railroad. Rio de 
 Salitre. Salt Licks. Mr. Allen s Note on the Salt of the Sao Francisco 
 Valley. Saltpetre. Geology of Country between Carunhauha and Urubu. 
 Change in Geological Structure, Climate, Vegetation, &c., below Urubu . 294 
 
 CHAPTER VIII. 
 
 PROVINCE OF BAHIA, GEOLOGY OF THE VICINITY OF SAO SALVADOR 
 AND THE B\HIA AND SAO FRANCISCO RAILROAD. 
 
 Topography of the Vicinity of Sao Salvador da Bahia. The Upper and 
 Lower Cities. The Population, &c. The Harbor. The Commerce of 
 the City and Province. The Climate, &c. The Bahia Steam Navigation 
 Company. The Bahia and Sao Francisco Railroad. The Paraguassu 
 Steam Tram-road. The Gneiss of Bahia. Decomposition. Drift De 
 posits. Consolidation of Beaches. Stone Reef at Rio Vermelho. 
 Blown Sands covering the Drift of the Hills. Mr. Allport s Description of 
 the Cretaceous Beds of Monserrate and Platafonna. Fossil Fishes, Croco- 
 
CONTENTS. XVli 
 
 diles, &c. Description of several Species of Fossil Mollusks. Cretaceous 
 Beds of Plataforma and Vicinity. Prof. Marsh s Notice of the Reptilian 
 Remains. Fossil Fishes at Agua Comprida. Gneiss at the Rio Johannes. 
 Tuboleiros and Sand Plains of Camassari. Peculiarities of the Topog 
 raphy of the Tertiary Hills. Tabatinga Clay. Sand Plains and Ta- 
 boleiros of the Imbupahy. Peat-Bog. Drift. Diamond-washings at 
 Pitanga. Cretaceous Strata at Pojuca. Piassabas. Campos of Ala- 
 goiuhas. Tertiary Hills. Character of Vegetation 333 
 
 CHAPTER IX. 
 
 THE PROVINCES OF SERGIPE AND ALAGOAS, AND THE RIVER SAO 
 FRANCISCO BELOW THE FALLS. 
 
 The Province of Sergipe. Its Division into Mattas and Agre&tes. The Rio 
 Real. L stancia; New Red Sandstone, Sugar Plantations, &c. Sand 
 Dunes. The Rio Vasabarris. The Rio Cotinguiba. Aracaju. Cre 
 taceous Beds with Inocerami at Sapucahy. Marolm. Cretaceous Lime 
 stone with Ammonites. Fossil Turtles." Sugar Plantations. Messrs. 
 Schramm and Company. The Bar of the Sao Francisco. Sand Dunes 
 of the Pontal. Character of the River below Penedo. Aracare\ Villa 
 Nova and its Cretaceous Sandstones. The City of Penedo and its Geology. 
 
 Its Commerce and Fair. Notes on the Piranha and its Habits. Propria. 
 
 Morro do Chaves and Cretaceous Fossils. Traipu. Iron Ore. Cam 
 pos, Vegetation, Cactuses, &c. Pao de Assucar. Cattle Fazendas. 
 Piranhas. County flat and covered by Boulders. The River Valley a 
 Narrow Gorge in a Gneiss Plain. The Falls of Paulo Affonso. Halfeld s 
 Description. Liais s Description. Comparison between Paulo Affonso 
 and Niagara. Mastodon Remains from near the Falls. Climate of the 
 Sao Francisco below the Falls. Steam Navigation. Character of the 
 Coast of the Province of Alagoas, South of Maceio. The Lagoas. The 
 City of Maceio and the Geology of its Vicinity. Tertiary Beds. Harbor 
 and Reefs .... . . , . 3 , 
 
 CHAPTER X. 
 
 PROVINCE OF PERNAMBUCO. 
 
 The Limits, Area, &c. of the Province. Its Topography, Geology, Climate, 
 Soils, &c. The Rivers. Productions of the Province. The City of Per- 
 nambuco or Recife Derivation of these Name. Situation of the City. 
 The Stone Reef. The Port formed by it. Shallowness of Water along this 
 part of the Coast. Pernambuco a Calling Station for Foreign Shipping. 
 The Pernambuco and Sao Francisco Railroad. Table of Heights along the 
 Line. Island of Itamaraca. Fossiliferous Limestones. Fertility of the 
 Island, Cocoa-Palm Groves, &c. Fernando de Noronha. Darwin s De 
 scription of the Geology of the Island. Its Dryuess and Sterility . . 427 
 
XViii CONTENTS. 
 
 CHAPTER XL 
 
 THE PROVINCE OF PARAHYBA DO NORTE. 
 
 Limits of the Province. The Serra or Plateau of the Cairir/s Velhos. The 
 Climate, Production*, &c. of the Province. Fertile Lands found only along 
 the Coast. The Rio Parahyba do Norte, its Navigability. The City of 
 Purahyba. The Consolidated Beach at the Mouth of the River. The 
 River and Town of Marnanguape. The Geology of the Vicinity of Para 
 hyba. Cretaceous Limestone with Fossils. Observations of Professor 
 Agassizand Mr. Williams. Mr. Williams s Observations on the Geology of 
 the Country between Parahyba and the Gold-Mines of Pianco. Mode of 
 Occurrence of the Gold. " The Tasso Brazilian Gold-Mining Company 
 (Limited)" 440 
 
 CHAPTER XII. 
 
 THE PROVINCE OF RIO GRANDE DO NORTE. 
 
 Limits of the Province, its Position, Mountain, and River Systems, &c. The 
 Rio Piranhas. Vegetation. Productions. The Carnahuba Palm and its 
 Uses. Cochineal. Cattle. Climate. Natal. Geology of the Province 451 
 
 CHAPTER XIII. 
 
 THE PROVINCE OF CEARA. 
 
 Geographical Position and Limits of the Province. The Serra da Ybiapaba. 
 
 Its Topography and Geological Structure. Serra de Araripe. River 
 Basins. General Sketch of the Geology of the Province. Climate. 
 Character of the Soil. Productions. City of Fortaleza Population of 
 the Province. Gardner s Sketch of the Geology of Ceara. Character of 
 Country in the Vicinity of Aracaty. Description of Country between 
 Aracaty and Ico. Serra de Pcreira. Villa do Ico and Vicinity. Country 
 between Ico and Crato. Gold Washings. Cnito. Serra de Araripe. 
 Villa da Barra do Jardim. Description of Fossil Fish Locality. The Fishes 
 noticed by Spix and Martins and Others, and described by Professor Agassiz. 
 
 Glacial Phenomena of Vicinity of Fortaleza spoken of by Professor Agassiz. 
 
 Mammalian Remains. Minerals. Meteorolites 456 
 
 CHAPTER XIV. 
 
 PROVINCE OF PIAUHY. 
 
 Geographical Position, Limits, &c., of the Province The Rio Parnahyba 
 and its Tributaries. Description of its Basin. General Geological Struc 
 ture and Topography of the Province. Table-topped Hills of Sandstone. 
 
CONTENTS. XIX 
 
 The Serra dos Dous IramSos and its Structure. Discussion of Gardner s 
 Observations on the Geology of Piauhy and Ceara. Gardner mistaken in 
 referring the great Sandstone Sheet to the Cretaceous. Sandstones prob 
 ably Tertiary. Their great Extension over Brazil. Distribution of the 
 Cretaceous Beds in Brazil. Climate of Piauhy. The Campos Mimosos 
 and the Campos Agrestes. Peculiarities of their Vegetation. Productions 
 of the Province, Population, &c 473 
 
 CHAPTER XV. 
 
 THE PROVINCES OF MARANHAO, PARA, AND AMAZON AS. 
 
 Sandstones of the Coast of Maranhao. The Interior composed of Metamor- 
 phic Rocks. Gold-Mines of Turi and Maracassume. Climate of the Prov 
 ince. Rains. Cities of Maranhao, Caxias, &c. Pororoca or Bore at the 
 Mouth of the River Mearim. Professor Agassiz s Sketch of the Geology 
 of the Amazonian Valley. His Theory of the Mode of Deposition of the 
 Amazonian Beds. Discussion of this Question. Cretaceous Rocks in the 
 Amazonian Valley 484 
 
 CHAPTER XVI. 
 
 PROVINCES OF GOYAZ AND MATTO GROSSO. 
 
 The Geographical Position of the Province of Goyaz. The Chapada da Man- 
 gabeira. Geology of the Vicinity of Natividade. Gold- Washings of the 
 Serra da Natividade, the Arraial da Chapada, and the Arraial da Conceicao. 
 Structure of the Serra at the Town of Arrayas. The Serra Genii. 
 Subterranean Streams. Western and Southern Goyaz composed of Meta- 
 morphic Rocks. Distribution of Gneissose and Granite Rocks in Western 
 Brazil. The Montes Pyreneos and their Height. The Rio Araguaya and 
 its Navigation. Dr. Couto de Magalhaes. Ilha do Bananal. Note on 
 Piranhas. Gold, Diamonds, Iron, and Chrome Ores. Climate, Forests, 
 Population, &c. The Western Part of the Plateau of Brazil composed of 
 undisturbed Beds of Sandstone, &c. The Amazonas-Paraguay Water-shed 
 a Plain without Serras .... 496 
 
 CHAPTER XVII. 
 
 PROVINCES OF SAO PAULO, PARANA, SANTA CATIIARINA, AND RIO 
 GRANDE DO SUL. 
 
 The Serra do Mar of Sao Paulo a Plateau. Its Character. Drainage in the 
 Provinces of Sao Paulo and Parana to the Westward. Eastward-flowing 
 Streams of no Importance The Sao Paulo Railroad. Description of the 
 
XX CONTENTS. 
 
 Country, along the Railway between Santos and Sao Paulo by Major 0. C. 
 James. Geology of Vicinity of Sao Paulo. Mawe s Description of the 
 Gold-Mines of Jaragua, and the Method of extracting the Gold. Country 
 westward to Campinas. Iron-Mines at Ypanema. Serra Arassoiava or 
 Guaraooiava. Climate, Products, &c. of the Province of Sao Paulo. 
 General Topographical Features of the Province of Parana, its Climate, 
 Productions, &c. Matte or Paraguayan Tea. Tea-Planting in Brazil. 
 Rivers. Colonies. Paranagua. Coal Basin on the Rio Tuberao in the 
 Province of Santa Catharina. General Description of the Physical Features 
 of the Province and that of Rio Grande do Sul. History of the Coal-Mines 
 of Brazil Observations of Perigot, Bouleich, Avt 5 -Lallemant, Plant, &c. 
 Description of the Coal-Fields of the River Jaguarao. Engineer McGinty s 
 Report on the Candiota Coal. Coal Basin on Rio Sao Sepe. Basin near 
 Sao Jeronymo , 505 
 
 CHAPTER XVIII. 
 
 THE GOLD-MINES OP BRAZIL. 
 
 Geological Distribution of Gold in Brazil. Gold in Gneiss, at Jaragua, 
 Cantagallo, Pianco, and elsewhere. The richest Deposits found in Veins 
 traversing Clay Slates. Character of Auriferous Quartz. Granular 
 Quartz, or Caco. Gold when associated with Sulphides rarely visible. 
 The auriferous Iron Ore, Jacutinga. Gold-Mines of Sao Joao d El-Rei. 
 The Morro Velho Mine, Mode of Occurrence of the Gold, Method of Extrac 
 tion, Yield, &c. The Gongo Soco Mine. The Rossa Grande Gold-Mining 
 Company. Mines at Morro de Santa Anna, Congonhas do Campo, Sao Vi 
 cente, Cata Branca. The Gold-Mines of Brazil not yet fairly developed . 632 
 
 CHAPTER XIX. 
 
 RESUME OP THE GEOLOGY OF BRAZIL. 
 
 Eozoic Rocks and their Distribution in Brazil. Absence of Limestones. 
 The Silurian Age in Brazil. The auriferous Clay-Slates of Minas probably 
 Lower Silurian. Note on the Silurian of the Andes. The Distribution of 
 Marine Animals in the Palaeozoic. The Devonian Age in Brazil and South 
 America. The Carboniferous of Brazil and Bolivia. The New Red Sand 
 stone. The Cretaceous, its Distribution in Brazil and South America. 
 Several distinct Periods represented. Tertiary Rocks. Drift. The Gla 
 cial Phenomena of Patagonia. Tapanhoaeanga. The Drift of Rio and 
 of the Region of Decomposition. The Drift of the Dry Region of Bahia, 
 Sergipe, and Alagoas. Examination into the Merits of the various Theories 
 proposed to account for the Formation of the Brazilian Drift. The Theory 
 of Subaerial Decomposition. Wave Action during a Subsidence. Wave 
 Action during Elevation. All these Theories unsatisfactory. The Glacial 
 Hypothesis the most reasonable 547 
 
CONTENTS. XXI 
 
 APPENDIX. 
 
 ON THE BOTOCUDOS. 
 
 Origin of the Name Botocudo. Stature. Physical Form and Characteristics. 
 Manner of Wearing Hair. Lip and Ear Ornaments. Professor W yman s 
 Description of Skull of Botocudo from Sao Matheos. Comparison with 
 other described Botocudo Skulls. Color of Botocudo. Manner of Painting 
 the Body. Dislike to being clothed. Bows and Arrows described. 
 Gerber s Enumeration of the Tribes. Von Tschudi s Description of the 
 Tribes. Ranches and Huts. Food. Mode of procuring Fire. Manu 
 factures. Marriage Customs. The Botocudos cruel Husbands. Facility 
 with which Wounds heal. Treatment of Children. Religious Ideas. 
 Belief in the Bad Spirit, Janchon. No Belief in a Supreme God. Burial 
 Customs. War Customs. Cannibalism. Dance. The Botocudos fast 
 disappearing. Botocudo Character. Geographical Distribution of the 
 Botocudos. Peculiarities of their Language, Pronunciation, Grammatical 
 Structure, &c. Botocudo Vocabularies 577 
 
LIST OF ILLUSTRATIONS. 
 
 COAL-BEDS ON Rio CANDIOTA, Rio GRANDE DO SUL . . . Frontispiece. 
 
 MAP OF PART OF PROVINCE OF Rio DE JANEIRO Page 5 
 
 THE SUGAR-LOAF, CORCOVADO, AND GAVIA, FROM SAO DOMINGO . . 10 
 
 SECTION AT PRAIA GRANDE 13 
 
 THE ORGANS 16 
 
 SECTION ALONG THE LINE OF THE CANTAGALLO RAILROAD . . . .22 
 
 THE CASCATINHA AT TIJUCA 27 
 
 SUCTION OF DRIFT AT BENNETT S, TIJUCA 28 
 
 BOULDERS AT TIJUCA 30 
 
 IDEAL SECTION OF DRIFT-COVERED GNEISS HILL 31 
 
 DIAGRAM TO SHOW CONTRAST IN MOULDING BETWEEN -DRIFT-COVERED 
 
 AND BARE SURFACES 34 
 
 COAST JUST EAST OF Rio 37 
 
 CAPE FRIO 39 
 
 ONE OF THE ISLANDS OF SANTA ANNA 42 
 
 FRADE DE MACAHE" 43 
 
 LOOKING UP THE Rio PARAHYBA FROM ABOVE CAMPOS .... 49 
 
 THE PEDRA LISA 64 
 
 SERRA OF ITAPEMEKIM, SEEN FROM THE SEA 58 
 
 COAST SOUTH OF GUARAPARY 61 
 
 BAY OF ESPRITO SANTO, LOOKING UP TO THE PAO DE ASSUCAR, AND CITY 
 
 OF VICTORIA 66 
 
 MONTE JUTUQUARA AND GNEISS HILLS NEAR VICTORIA .... 67 
 
 F()RTALEZA DE PERITININGA 68 
 
 BOULDER OF DECOMPOSITION, VICTORIA 69 
 
 DECOMPOSING SURFACE, ILHO DO Boi 70 
 
 NOSSA SENHORA DA PENHA, VICTORIA 70 
 
 ANCIENT SEA-LEVEL ON PAO D ASSUCAR, VICTORIA . . . . .7V 
 
 ARRANGEMENT OF LITTORAL FAUNA AT VILLA VELHA .... 73 
 
 MORRO DE MESTRE ALVARO 80 
 
 COAST BETWEEN VICTORIA AND Rio DOCE 85 
 
 Rio DOCE AT PORTO DE SOUZA 92 
 
 VIEW ON Rio DOCE 94 
 
 LOOKING UP THE DOCE FROM NEAR LINHARES -97 
 
LIST OF ILLUSTRATIONS. XX111 
 
 LAGOA JUPARANAA, LOOKING TOWARDS THE OUTLET .... 102 
 
 LAGOA DO Aviso 103 
 
 FAZENDA OF CAMPO REDONDO, SAO MATHEOS 120 
 
 SANTA RITA RIVER VALLEY 143 
 
 CORRUGATED QUARTZ VEINS AT MINAS NOVAS 158 
 
 ISLAND OF SANTA BARBARA DOS ABROLHOS 175 
 
 THE ISLANDS OF THE ABROLHOS FROM THE SOUTH .... 176 
 
 RECIFE DO Lixo 202 
 
 SECTION ACROSS BORDER OF Lixo REEF 210 
 
 PROFILE SKETCH OF RIVER VALLEY NEAR PORTO SEGURO . . . 225 
 
 MONTE PASCOAL FROM THE SEA 226 
 
 SECTION ACROSS STONE REEF AT PORTO SEGURO 229 
 
 BAYS OF SANTA CRUZ AND CABRAL 233 
 
 VIEW OF THE COAST SOUTH OF THE Rio DE CONTAS .... 260 
 
 MAP OF CAMAMU BAY 262 
 
 MAP OF BAHIA AND VICINITY 267 
 
 GROUND VIEW OF THE LAPA VERMELHA 282 
 
 SECTION OF Rio SAO FRANCISCO VALLEY 310 
 
 VIEW OF VALLEY WALLS NEAR JACOBINA 313 
 
 KNOBS NEAR THE SfiRRA DA TERRA DURA 315 
 
 SECTIONS OF SEVERAL CUTTINGS ON THE BAHIA AND SAO FRANCISCO 
 
 RAILROAD 355-367 
 
 TERTIARY HILLS NEAR POJUCA TUNNEL 371 
 
 THE TABOLEIROS NEAR ALAGOINHAS 376 
 
 SAND-DUNES AT THE MOUTH OF THE Rio REAL 380 
 
 BAR OF Rio COTINGUIBA 382 
 
 MAROIM 384 
 
 VILLA NOVA AND PENEDO 397 
 
 TRAIPU" FROM NEAR MARCACAO 407 
 
 LOOKING DOWN THE RIVER FROM PAO D ASSUCAR 410 
 
 CATTLE FAZENDA AND GARDEN NEAR PAO D ASSUCAR . . . 411 
 
 VIEW NEAR ALLEGRIA 412 
 
 VlEW LOOKING UP TOWARDS PlRANHAS 413 
 
 CACHOEIRA DE PAULO AFFONSO 415 
 
 PERNAMBUCO 434 
 
 ISLAND OF FERNANDO DE NORONHA 438 
 
 MOUTH OF Rio PARAHYBA DO NORTE 443 
 
 SKETCH-MAP OF THE Rio PARAHYBA DO NORTE 444 
 
 MOUTH OF Rio GRANDE DO NORTE 455 
 
 CUTTING ON THE SAO PAULO RAILWAY, SHOWING DRIFT LYING ON 
 
 DECOMPOSED ROCK 608 
 
 P>OTOCUDO MAN AND WOMAN 581 
 
 SKULL OF BOTOCUDO 585 
 
 4 
 
GEOLOGY AND PHYSICAL GEOGRAPHY OF 
 BRAZIL. 
 
 CHAPTER I. 
 
 THE PROVINCE OF RIO DE JANEIRO. 
 
 The Scrra do Mar. The Serra da Mantiqueira, and the Pico do Itatiaiossii. 
 The Rio Parahyba do Sul and its Tributaries. Description of the Bay of 
 Rio, its Islands, Tides, &c. The Sugar-Loaf. The Corcovado. The 
 Gneiss of Eio. The Organ Mountains. Geological Observations along 
 the Cantagallo Railroad. Drift Phenomena at Ilio, Tijuca, and on the 
 Dom Pedro II. Railroad. Decomposition of Gneiss in situ and its Effect on 
 the Forms of the Hills. Recent Rise of the Coast. The Coast between 
 Rio and Cape Frio, its Lakes, Salines, &c. Cape Frio. Os Btizios. Isl 
 ands of Santa Anna. Frade de Macahe. Campos dos Goitacazcs, their La 
 goons, Canal, &c. Rio Parahyba. Sao Joao da Barra. Sugar Plantations. 
 
 City of Campos. The Rio Muriahe. Tertiary Beds. Sugar Fazcndas. 
 
 Sao Fidelis. The Gold Mines of Cantagallo. Geological Notes on the 
 Country between Sao Fidelis and Bom Jesus. The Rio Itabapuana. The 
 Serra d ltabapuana. The Garrafao. Barra do Itabapuana. 
 
 THE province of Rio de Janeiro is almost entirely com 
 posed of gneiss, and this gneiss region is mountainous and 
 high. The Serra do Mar, skirting the coast of the province 
 of Sao Paulo, enters the southwest corner of the province of 
 Rio, and, composed of a great number of parallel* ridges, 
 often much broken, traverses it from one end to the other. 
 These mountains form the edge of the great Brazilian 
 plateau, which consists along its eastern border of a 
 broad band of gneiss. The course of the Serra do Mar is 
 approximately east-northeast, so that, as the coast of the 
 
2 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 province of Rio de Janeiro runs eastward from Sao Paulo 
 to Cape Frio, the Serra do Mar, on entering the province 
 of Rio, trends gradually away from the coast, passing by 
 the head of the Bay of Rio. The gneiss plateau is bor 
 dered in the eastern part of the province by low plains, 
 tertiary and recent. The minor ridges of the grand Serra 
 do Mar break down abruptly on the edge of these low 
 grounds. In the western part of the province, where 
 the Serra skirts the coast, that coast is often high, bold, 
 very irregular in outline, and bordered by numerous rocky 
 islands. South of the Serra do Mar, at Rio, lie several 
 isolated mountains and ranges of hills, really belonging 
 to the same great mountain system, but separated from 
 the plateau by low plains similar to those north of Cape 
 Frio. 
 
 To the northwest of the Serra do Mar, and separated 
 from it in part by the valley of the river Parahyba do Sul, 
 is another great mountain range called the Serra da Man- 
 tiqueira, which is also composed of gneiss, and belongs to 
 the same system of upheaval as the Serra do Mar. This 
 range separates itself from the coast range near the city of 
 Sao Paulo, and, lying inside the Serra do Mar, skirts the 
 coast to a much greater distance to the north than the 
 latter does. The ridge properly called the Serra da Man- 
 tiqueira accompanies the northern part of the province of 
 Rio for a few miles, when the province boundary line 
 leaves il and runs off to the eastward. 
 
 At a distance of four or five miles from the northwest 
 corner of the province, in the Serra da Mantiqueira, is the 
 Pico do Itatiaiossu, which appears to be the highest point 
 in Brazil, and, according to the Revista Trimensal do 
 Institute Historico e G-eographico Brasileiro, has an al- 
 
THE PROVINCE OF RIO DE JANEIRO. 3 
 
 titude of about 10,300 feet.* It is said to be volcanic in 
 structure, and two craters are reported to exist on it, to 
 gether with sulphur springs and sulphur deposits.! I have 
 never seen the Itatiaiossu, but I have the strongest doubt 
 as to its being a volcano. Snow occasionally falls on this 
 mountain during the winter, and is said to remain some 
 times for several days. There can be no reasonable doubt 
 about this being the highest peak in Brazil. Itacolumi, 
 so long famed, is, according to Burton, only about 6,400 
 feet in height. According to Eschwege, this last peak is 
 5,720 feet ; Gerber says 1,112 metres (3,650 feet) ; while 
 the highest point cited by Gerber is the Alto da Serra da 
 Piedade, in the Municipio de Sabar4, which Liais makes 
 only 1,783 metres (5,853 feet). Gardner estimates the 
 height of the Organ Mountains at from 7,500 to 7,800 
 feet. It is interesting to observe, as Burton has remarked, 
 that the summit line is not in the interior in Brazil, but 
 close to the coast. Almost precisely south of the Itati 
 aiossu, in the province of Sao Paulo, is a high point in 
 the Serra do Mar, which gives rise to two rivers, one, a 
 small stream, the Rio Pirahy, flowing east and then north 
 east between two ridges of the Serra do Mar ; the other, 
 the Rio Parahyba do Sul, under the name of Pirahytinga, 
 which flows southwestward and westward for about eighty 
 miles, when it escapes northward around the end of the 
 ridge which has so far formed its barrier on the north, and, 
 doubling upon itself, flows thence, with a general east-north 
 east direction, behind the Serra do Mar, traversing the whole 
 
 * Dr. Candido Mendes de Almeida, in his Atlas do Imperio do Brazil, 
 makes the height only 2,994 metres, or 9,829 feet. 
 
 t Burton, Explorations, &c., Vol. I. p. 61. He does not here speak from 
 personal observation. 
 
4 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 province of Rio de Janeiro, until, some forty-five miles from 
 the sea, the mountains break down on the edge of the pla 
 teau, when it reaches the Campos dos Goitacazes, and 
 empties into the sea after a course of 102 Brazilian leagues 
 (Gerber), or 408 miles. 
 
 According to Gerber, the altitude of the river above the 
 sea, at the mouth of the river Parahybuna, is 272 metres. 
 The stream is so much broken up by rapids above Sao Fide- 
 lis that steam navigation ends there, and in the rest of its 
 course it gives passage only to canoes and large montarias 
 propelled by the setting-pole or towing-rope. 
 
 The principal affluents of the Parahyba are the Rio 
 Pirahy, already mentioned, which affords steam naviga 
 tion for a short distance above its mouth ; the Rio Preto, 
 which rises in the Itatiaiossu, and is a much larger stream, 
 falling into the Parahyba north of Rio ; the Pirapitinga, 
 which rises in the Serra of the Pardo ; the Piabanha, which 
 descends from the Scrra do Mar from near Petropolis, 
 and enters the Rio Parahyba at Entre Rios, a little above 
 the Parahybuna; (this latter descends from the Man- 
 tiqueira, on the opposite side of the basin, in the same 
 meridian with the Piabanha) ; the Pomba, which rises in 
 the Serra da Mantiqueira, a few miles east of Barbacena ; 
 and the Muriab.6, which takes its source in the same range, 
 a few leagues to the east. The gneiss region of Rio, where 
 uncultivated, bears a most vigorous virgin forest growth, 
 and its soils are particularly favorable for the cultivation of 
 coffee ; the great valley of the Parahyba, above Sao Fidelis, 
 and the valleys of its affluents, are largely occupied by coffee 
 plantations. The same is true of the gneiss regions of the 
 north, whose topography and soils are, over large districts, 
 favorable for coffee planting. 
 
THE PROVINCE OF RIO DE JANEIRO. 
 
6 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 With these words of introduction, let us now look at the 
 general points of interest to the geologist and physical geog 
 rapher in the vicinity of Rio, and then examine the province 
 to the east. 
 
 The Bay of Rio is a basin penetrating inland about twenty 
 miles. It is only about a mile wide at the mouth, or Barra, 
 but widens immediately. The shores on each side run in 
 and out with deep, sweeping curves, making beautiful bays ; 
 and, three miles inside, passing the city of Nictherohy on 
 the east and Rio de Janeiro on the west, the bay widens 
 rapidly, and then, with the same irregular shore line, attains 
 its greatest width about six miles from the mouth, when it 
 contracts and runs off inland toward the northeast. It con 
 tains several islands ; those near the city, as the Ilha das 
 Cobras and Ilha Enxada, are of gneiss, like that of the adja 
 cent hills. 
 
 The Ilha do Governador, a large island which I have ex 
 amined on the eastern side, is composed partially of gneiss ; 
 but I strongly suspect that it is largely made up of the 
 same tertiary clays which I have found on the Cantagallo 
 Railroad, near Porto das Caixas. It seems too level to 
 consist of gneiss alone. This island possesses fertile soils, 
 but the ants are a terrible plague. Burton says that there 
 are kjoekkenmoeddings upon it. Ancient shell-heaps have 
 been reported from Brazil by Dr. Henry Naegeli of Rio, 
 and they occur at various places along the coast. At Santos 
 there are some very extensive ones, and St. Hilaire speaks 
 of numerous shell-heaps on the coast of the province of 
 Espirito Santo near Santa Cruz. 
 
 The smaller islands arc gneiss, though, as in Paquetd, we 
 find isolated gneiss masses united by stretches of sands con 
 taining recent shells. Spix and Martius have called attention 
 
THE PROVINCE OF RIO DE JANEIRO. 7 
 
 to the great number of palms growing on the islands in the 
 Bay of Rio, owing to the dampness and heat of the climate. 
 The bay is nowhere very deep ; along the shores it is very 
 shallow ; and Mouchez s map shows the soundings pro 
 gressing quite uniformly from all sides, from one to four 
 teen or eighteen metres. Near the middle of the bay, off 
 Rio, we find the greatest depth, thirty-one metres. 
 
 Near the mouth of the bay the shores are rocky and sandy ; 
 the water is clear, and the bottom is composed of sand and 
 shells ; but around the whole head of the bay, where the 
 shores are low, and a host of little rivers bring down great 
 quantities of silt, the basin is bordered by extensive man 
 grove swamps, and the bottom is shallow and muddy. 
 
 In the turbid and somewhat brackish waters of the head 
 of the bay oysters of an immense size flourish, often grow 
 ing attached to the roots of the mangroves. Fish are very 
 abundant, and the curraes, or weirs, for taking them are 
 conspicuous objects in the shore scenery. The waters of 
 the interior of the bay are exceedingly clear and bright, and 
 off the islands of Paqueta and Governador a deposit of 
 shells, with a calcareous mud, is in process of accumulation. 
 Almost all the shells are small, and consist chiefly of spe 
 cies of Area, Venus, Murex, Cardium, Dentalium, <tc. I have 
 looked diligently for corals, but the only Madreporians I 
 have seen in the Bay of Rio consist of a couple of species 
 of Astrangia* found growing on the shells, and contributing 
 really nothing to the deposits there accumulating. The tide 
 in the Bay of Rio, as near as I can learn, rises from three to 
 five feet,f but is very uncertain, as has been remarked by 
 
 * Professor Agassiz tells me that a species of Porites has been collected by 
 Dr. Naegcli. 
 
 t On the map of Rio, by M. Barral, it is set down as three feet. Spix and 
 Martius are in error in making it fourteen or fifteen feet. 
 
8 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Spix and Martius.* I have seen the tide stand at the same 
 level for a whole day at the docks in Rio. On one occasion 
 I visited Paqueta, and during a day and a half the tide 
 was high ; but, shortly after my return to Rio, it fell to an 
 extraordinarily low level, and remained so for many hours. 
 This appears to be owing to the fact that the bay opens 
 to the southward, while the mouth is very contracted, so that 
 the waters may be piled up on the coast by a southerly 
 wind in such a way as to prevent the tide from flowing 
 out, while other winds may depress the level of the waters 
 for as long a period. 
 
 The entrance to the harbor is very bold on both sides, 
 and is sentinelled by steep gneiss hills, rounded or coni 
 cal, wooded or bare, presenting lichen-blacked precipices, 
 whose faces are smoothed and rounded in a most remarkable 
 manner. On the eastern side some of these hills are more 
 than a thousand feet in height. They are clustered closely 
 together, and stretch off in a sea of hills along the bay for a 
 few miles on the eastern side, while along the coast, eastward 
 toward Cape Frio, they form a line of irregular mountains 
 of much grandeur and beauty. On the west side of the 
 entrance to the bay is the conical peak of the Pao de Assu- 
 
 * Spix and Martius, Reise, Vol. I. p. 95. " The interior basin of Rio de 
 Janeiro has its tide as well as the ocean. At new and full moon, high water, 
 which rises fourteen or fifteen feet, sets in at thirty minutes past four ; the ebb 
 sometimes continues a whole day without intermission, at which time the current 
 is the strongest on the west side of the bay. On the other hand, when the flood 
 begins, a whirling current is remarked on the east side. The flood continues a 
 shorter time than the ebb, and usually runs at the rate of three or four sea 
 miles per hour. This strong flood has more than once led the captains of the 
 ships into error, and caused them to cast anchor too close to the shore." 
 
 The same authors remark that the saltness of the water of the bay is rather 
 less than that of the ocean. This is to be expected, from the large drain of 
 fresh water into it. 
 
THE PROVINCE OF RIO DE JANEIRO. 9 
 
 car (Pot de Beurre of the early French colonists), with its 
 smooth, precipitous sides. This celebrated rock belongs to 
 a short range of hills that runs westward to the LagQa das 
 Freitas, and which is separated from the Corcovado, and the 
 hills of which the latter forms a part, by the valley of Bota- 
 fogo. The Sugar-loaf is said to be over a thousand feet in 
 height.* The other hills are much lower, and are rounded, 
 with precipitous sides facing the valley. They are composed 
 of heavy beds of gneiss, which is very homogeneous, coarse 
 grained, and often very porphyritic in structure, with 
 large crystals of flesh-colored orthoclase feldspar. These 
 beds have very nearly the same strike as the trend of the 
 hills they compose, and dip southward at a moderate angle, 
 as may be observed in the cliffs along the pass leading from 
 Botafogo to the Praia de Copocabana. The Corcovado is 
 a sharp, angular peak, which separates itself from the great 
 mass of hills lying back of Rio, rising by a long, narrow 
 incline on the northwestern side to a point only a few rods 
 in area on top. On three sides it drops off in a splendid 
 sheer precipice, which on the south is several hundred feet 
 high. Below this precipice the mountain presents to 
 ward the south a very steep wooded slope. A similar 
 slope, not a talus, runs below the cliffs on the opposite 
 side of the valley, and may be observed in very many 
 hills of the gneiss region. In the Corcovado we find 
 rocks similar to those of the Sugar-loaf, but varying con 
 siderably in character, being generally, near the top, well 
 laminated, sometimes with large crystals of feldspar, and 
 not infrequently full of garnets. The dip varies somewhat, 
 arid I find in my field-book two notes of a slight southerly 
 
 * Burmeister says 1,212 feet. (Reise nach Brasilien, p. 57.) Dr. Almeida 
 makes it 373 metres, or about 1,426 feet. 
 1* 
 
10 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 THE SUGAR-LOAF, COKCOVADO AND GAVIA, FROM SAO DOMINGO. 
 
 dip near the top, while an observation taken some distance 
 below the reservoir on the aqueduct gave a marked north 
 ward dip. The general dip of the gneiss in the peak, as well 
 as in the hills east, is certainly northward, and one sees the 
 same dip in certain ledges of rock in the harbor, which lie 
 in the line of strike of the beds of the Corcovado. The val 
 ley which separates the hills of the Corcovado from those of 
 the Sugar-loaf is therefore an anteclinal valley. Professor 
 Agassiz has independently made the same observations, and 
 has come to the same conclusions.* 
 
 Three miles southwest of the Corcovado are the Tres Irmaos, 
 a group of conical peaks very interesting to the student of 
 topography, and a couple of miles west of these is the Gavia, 
 an isolated, tower-like, flat-topped mountain, said to be 3,000 
 feet high. Within the irregular ring formed by these rnoun- 
 
 * Pissis has called attention to the resemblance borne between the fine 
 grained gneiss, with garnets overlying the porphyritic variety forming the base 
 of the Corcovado, and that of the rocks at Copocabana. (Mem. de Vlnstit. 
 de France, Tom. X. pp. 362, 363.) 
 
THE PROVINCE OF RIO DE JANEIRO. 11 
 
 tains is the beautiful Lagoa dc Freitas, a sheet of water held 
 in a basin among the hills, and shut out from the sea like 
 the lagoas so common along the coast east and west of Rio 
 by the throwing up of a sand-beach across its mouth. 
 Travellers all speak of the romantic beauty of this spot, and 
 it is worthy of their praise ; for though clothed in the warm 
 verdure of the tropics, it is really Swiss-like in the character 
 of its scenery. If the geologist has any soul, any love for 
 the beautiful, there is no scene which, with all his cold anal 
 ysis of topographical and geological elements, is more likely 
 to impress him as an artist s work. I know of no view 
 which has affected me so much not only as a scientific 
 observer, but as a man as that of the vicinity of Rio 
 from the top of the Corcovado. There are a thousand sub 
 jects for observation and study ; and, with all, there comes 
 over one a feeling akin to, but infinitely more deep and 
 impressive than, that which one experiences when in some 
 old cathedral he sits down to study the sublime creation of 
 one of the old masters. He who can lean over the parapet 
 that crowns the Corcovado, and look down more than 2,000 
 feetf on the temple of palms of the Botanical Garden, and on 
 the silent Lagoa de Freitas, " another sky," in whose blue 
 depths sail soft fleecy clouds, who can gaze on the proud 
 encircling peaks, green with an everlasting spring, and shiv 
 ering with silvery reflections from the Cecropias, who 
 can look out over the island and sail dotted sea, and the 
 surges creeping up on the long, curving sea-beaches, and 
 then over the bay, with the city fringing widely its sweep 
 ing curves, the sea of hills beyond, the majestic Serra dos 
 Orgaos heaving its great back, in the exquisite blue distance, 
 far above the level line of the clouds, its great minarets 
 
 * 2,179 feet is the exact height. Burmeister says 2,164. 
 
12 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 sharply defined against the purple ether, and can intelli 
 gently take into consideration all the geological, climatic, 
 and other natural laws which have determined the elements 
 of beauty and usefulness in the scene, and not have his whole 
 soul moved within him in homage to the Artist whose hand 
 has moulded continents, carved out their lineaments, spread 
 over them their mantle of vegetation, and peopled them with 
 living forms, has not gone beyond the alphabet and gram 
 mar of his science, and has no idea of the literature of 
 Nature. 
 
 The Corcovado is only one peak of a mass of hills which 
 occupies a large area west of Rio. It is united by a pass 
 1,000 feet in height to another mass lying northward of the 
 last, and which culminates in a sharp conical peak called 
 Tijuca, the latter being about 3,447 feet in height. The 
 Tijuca range is connected with a group of hills which ex 
 tends several leagues to the west. "Westward of the Gavia 
 is a stretch of plain only a few feet in elevation above the 
 sea, and broken up by very numerous lakes, some of which 
 are of considerable size. Between the hills which I have 
 just described and the Serra do Mar there is a wide extent 
 of low country, in some parts perfectly flat and very low, 
 in others somewhat diversified with hills. I have seen 
 this country only from the railway-train, but I would sug 
 gest the more than probability that a considerable part of it 
 may be made up of beds of stratified clays, like those of the 
 Cantagallo Railroad, presently to be described. The Serra do 
 Mar, as well as the whole range, including the Organ and 
 Cantagallo Mountains, as has been already stated, is com 
 posed of gneiss. 
 
 At Rio the gneiss varies very much in texture. As a gen 
 eral thing it occurs in very thick homogeneous beds, varying 
 
THE PROVINCE OF RIO DE JANEIRO. 13 
 
 from an exceedingly coarse porphyritic kind containing large 
 crystals of black mica, and crystals of pink feldspar several 
 inches long, to a fine, even-grained, compact, light-gray 
 variety. Sometimes it is very distinctly laminated, fissile, 
 flaggy, or schistose. Garnets are very common in it. It 
 is largely quarried for building purposes in all directions, 
 and the finer-grained kinds are much used for paving, not 
 only in the streets of Rio, but in other towns on the 
 coast. The numerous quarries afford very excellent sec 
 tions. The unbroken compactness and the little jointing 
 of rock, even on the surface, strike the observer as re 
 markable. Many of the hills are monoliths. The dip 
 and strike are, owing to the homogeneity of the rock, gen 
 erally very indistinct. Immense granite veins, with the 
 minerals composing them often very coarsely crystallized, 
 traverse the rock, together with veins of milky quartz and 
 occasional dykes of diorite or greenstone. Faults are numer 
 ous. I copy from my note-book the following section -made 
 from the barracks north of Praia Grande on the eastern side 
 of the harbor and extending southward to the city : 
 
 SECTION AT PRAIA GRANDE. 
 
 At the barracks, the gneiss, a, is dark gray with a fine 
 lamination. The crystals of feldspar are very small, and 
 there are a great many red garnets. Strike N. 40 E. 
 Dip 30 S. 
 
 About 800 feet southward of the barracks is a large quarry, 
 in a bed of very compact gneiss, d, showing scarcely a trace of 
 stratification. It is composed of feldspar in large crystals, 
 
14 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 mica, finely crystallized, and very little quartz. Garnets are 
 abundant, and the rock is traversed by veins, a number of 
 which I have noted as dipping steeply southward. In one 
 of these veins, which was four inches in width, the sides 
 consisted almost wholly of large, coarse crystals of a light, 
 flesh-colored feldspar with a little green mica, while the 
 middle portion was made up of clear, glassy quartz of a 
 light reddish color, with an occasional large crystal of black 
 mica. 
 
 Above these beds are others, in which I have noted that 
 the crystals of feldspar lie with their longer sides parallel to 
 the plane of stratification. In a large vein here we find the 
 same arrangement of the materials as in that above de 
 scribed. 
 
 /. Gneiss with very large crystals of feldspar. 
 
 g. Thin beds of compact mica-slate, or extremely fine 
 grained and distinctly bedded gneiss. Strike N. 35 E. 
 Dip 35 S. 
 
 Here the section fails on reaching the city. In a hill im 
 mediately east of the church of Sao Joao the gneiss is 
 principally composed of large feldspar crystals. Strike N. 
 80 E. Dip 35 S. In the point between Praia Grande and 
 Sao Domingo the rock is like that last seen, the strike 
 being N. 80 E., and the dip southward. Just south of Fort 
 Sao Domingo there are seen veins of iron ore in the gneiss ; 
 the dip is about 45 S., and the strike about N. 80 E. 
 
 I have carefully examined the cuttings on the Dom Pedro 
 II. Railroad, from Belem, at the southern base of the Serra 
 do Mar, to Ypiranga, in the valley of the Parahyba do Sul, 
 and I have studied them in detail, on foot, from the mouth 
 of the great tunnel, which pierces the crest of the Serra, 
 to Ypiranga. The whole ridge is composed of gneiss, which 
 
THE PROVINCE OF RIO DE JANEIRO. 15 
 
 varies very much in character, as a general rule being dark 
 gray in color, well laminated, often fine-grained and schis 
 tose. The strike varies, according to my observations, from 
 N. 45 E. to N. 80 E., and the mean of thirty-four observa 
 tions of strike taken along the road from the southern en 
 trance of the great tunnel to the Barra do Pirahy, would 
 give N. 62 E.* The dip is almost invariably northward 
 from Belem to the Parahyba, so that the Serra is here a 
 monoclinal ridge, but it is very probable that the same strata 
 are repeated. There seems to have been much dislocation of 
 the strata, and faults are common. In some cases the beds 
 are much plicated, though, as above remarked, the general 
 dip is remarkably uniform. Quartz and granite veins are 
 very numerous, and trap dykes are not uncommon. At 
 Ypiranga is a thin bed of crystalline limestone, which is ex 
 posed near the railway, and is more or less quarried for 
 burning into lime. 
 
 The Serra dos Orgaos and the Serra da Estrella, so fre 
 quently described, is a heavy ridge lying at the head of the 
 Bay of Rio, and belonging to the Serra do Mar. It is a mag 
 nificent mountain pile, which, to the east, runs up into a 
 series of picturesque aiguilles. I have never myself visited 
 the Organs, and I quote Professor Agassiz s remarks on the 
 structure of this ridge. He says : f " The chain is formed by 
 the sharp folding up of the strata, sometimes quite vertically, 
 in other instances with a slope more or less steep, but always 
 rather sudden. To one standing on the hill to the east of 
 Theresopolis, the whole range presents itself in perfect profile ; 
 the axis, on either side of which dip the almost vertical beds 
 
 * Of eight reliable observations of strike, taken in the vicinity of Rio, the mean 
 is N. 55 30 E. ; but these observations vary from N. 10 E. to N. 80 E. 
 t Journey in Brazil, p. 491. 
 
GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 THE ORGANS. 
 
 of metamorphic rocks composing the chain, occupies about 
 the centre of the range. 
 
 " To the north, though very steeply inclined, the beds arc 
 not so vertical as in the southern prolongation of the range. 
 The consequence of this difference is the formation of more 
 massive and less disconnected summits on the north side, 
 while on the south side, where the strata are nearly or quite 
 vertical, the harder set of beds alone have remained stand 
 ing, the softer intervening beds having been gradually disin 
 tegrated. By this process have been formed those strange 
 peaks which appear from a distance like a row of organ- 
 
THE PROVINCE OF RIO DE JANEIRO. 17 
 
 pipes, and have suggested the name by which the chain is 
 known. They consist of vertical beds, isolated from the 
 general mass in consequence of the disappearance of contig 
 uous strata. The aspect of these mountains from Rio is 
 much the same as from Theresopolis, only that from the 
 two points of view one being to the northeast, the other 
 to the southwest of the range their summits present them 
 selves in reverse order. When seen in complete profile, 
 their slender appearance is most striking. Viewed from the 
 side, the broad surface of the strata, though equally steep, 
 exhibits, a triangular form rather than that of vertical col 
 umns. It is strange that the height of the Organ Mountain 
 peaks, so conspicuous a feature in the landscape of Rio, 
 should not have been accurately measured. The only pre 
 cise indication I have been able to find is recorded by Liais, 
 who gives 7,000 feet as the maximum height observed by 
 him." * 
 
 " These abrupt peaks frequently surround closed basins, 
 very symmetrical in shape, but without any outlet. On 
 account of this singular formation, the glacial phenomena 
 which abound in the Organ Mountains are of a peculiar 
 character." f 
 
 Pissis | has described the Organs in very much the same 
 
 * Gardner makes them 7,500 to 7,800 feet. 
 
 t Similar aiguilles are common elsewhere on the coast in the gneiss belt, and 
 present the same features as are described by Professor Agassiz. I may cite 
 here, as examples, the Trade de Macahe , the Pedra Lisa and the Garrafao, near 
 Itabapuana, the peaks of Itapemerim, &c., all of which I shall describe here 
 after. 
 
 { " Le leptinite s y trouve fortement redresse et presente des lames colossales 
 qui, vues de Eio de Janeiro dans le sens de leur e paisseur, apparaissent comme 
 de hautes et minces aiguilles place es les unes a cote des autres et ressem- 
 blant assez a des tuyaux d orgues." Mm. de I Inst. de France, Tom. X. 
 p. 360. 
 
 B 
 
18 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 terms as Agassiz. The same author calls attention to the 
 great steepness of the southern slope of the Organ Moun 
 tains, and speaks of the little streams which, precipitat 
 ing themselves over it, are visible, as I can testify, from 
 Rio, a distance of more than thirty miles, appearing 
 like silvery lines drawn down the blue flank of the Serra. 
 The needles of the Organs are usually represented as seen 
 from their thinnest side, which would give one the idea that 
 they were tower or chimney-like masses. Professor Agassiz 
 has figured one of the most noted of the peaks as seen in 
 this way. Through his kindness I am enabled to present a 
 side view of the same peak, from a photograph in his posses 
 sion. The views of the Organs in Gardner s Travels, and in 
 the majority of works on Brazil, are very poor. 
 
 The gneiss of the vicinity of Rio and of the Serra do Mar 
 is remarkably unproductive in useful minerals. Indeed, I 
 do not know of any mineral deposits of economic value in 
 the region, except gold, which occurs in these gneisses in 
 Sao Paulo, at Cantagallo, and elsewhere, but not very abun 
 dantly. The almost entire absence of limestone is remark 
 able. I have nowhere seen any trace of graphite. 
 
 The Rio Macacu is one of the little streams entering the 
 Bay of Rio near its northeastern extremity. The whole coun 
 try bordering the bay, as above stated, is one great man 
 grove swamp, and from the mouth of the river to the Porto 
 da Villa Nova this is the character of its banks. At this 
 place the soil consists, like that of the rest of the swamp, of 
 a soft, dark-blue mud, or clay, containing a little decomposed 
 mica in silvery flakes. This mud at Villa Nova is twenty- 
 four feet deep, and is underlaid by sand and gravel contain 
 ing recent shells ; but farther back from the river, as might 
 be expected, it is not so thick, being only about six feet. 
 
THE PROVINCE OF RIO DE JANEIRO. 19 
 
 The general surface of the country here is perfectly flat, 
 and only about one foot above water level at spring-tides. 
 The river * is one hundred feet wide at the Porto, but it is 
 very deep, and its banks are bordered with rush-like plants, 
 the quina, a plant with a triangular stem, and the periperi, 
 or Brazilian papyrus, so much used for mats, while the 
 swamps are covered by mangroves, tapibuia trees, and 
 a dense growth of plants, such as love the salt marshes. 
 
 Leaving the river at Porto da Villa Nova, and follow 
 ing along the extension of the line of the Cantagallo Rail 
 way, one soon leaves the swamps, and, rising a few 
 feet, finds himself on a plain of coarse white sand, in 
 which are seen exposed in excavations an abundance of 
 recent shells, like those which lie on the beaches along 
 the bay, Venus flexuosa being especially common. These 
 plains are sparsely wooded, and support a vegetation 
 quite different from what we have observed in the swamps. f 
 Conspicuous among the trees, or rather large shrubs, is 
 the pitangueira (Eugenia), noted for its refreshingly acid, 
 red fruit. Bromeliaceous plants are common, together 
 with cactuses, &c. Where, however, a soil has accumu- 
 
 * Burmeister remarks upon the coffee-brown color of the water of this river. 
 The vicinity is very unhealthy, intermittent fevers of a typhoid character pre 
 vailing. 
 
 t Spix and Martius describe a similar sand-plain on the road from Rio to 
 the Imperial Fazenda of Santa Cruz : " On the way hither we remarked 
 a stretch of ground composed of coarse dry granite sand. The low but 
 very pleasing wood covering it resembled, in its shining green, stiff foliage, 
 our laurel woods, but, as a token of the tropical climate, it was charac 
 terized by the multiplicity of the flower-forms." The same authors men 
 tion the following plants as occurring on this ground : " Schinus Aroeira, 
 terebinthifolia Raddi ; Pohlana (Langsdorffia Leandr.) instrumentaria Mart.; 
 Spixia fieteranthera Leandr. ; Dyrsonima nitidissima Humb. ; Sapium illcifolium 
 W. ; Ahodea Physiphora Mart. ; Petroea racemosa Nees ; Solena grand! folia ; 
 Seriarue, Paulinice, sp., &c." (Spix and Martius, Reise, Vol. I. p. 181.) 
 
20 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 lated, the vegetation assumes the dignity of a forest growth. 
 Presently we come upon some low hillocks, very rounded in 
 their outline, which stand like islands in the sand-plain. 
 Some of these are cut through by the railway, and we 
 see that they are isolated and denuded outliers of a for 
 mation underlying the sands. Rising above the general 
 level of the sands, they are bathed by them round about, 
 like islands. Some of these hillocks are composed of a bed 
 of white or reddish arenaceous clay, obscurely stratified, 
 like a kaolin mixed with sand, with an occasional quartz 
 pebble, and irregularly tinged by yellow or red ferric oxide. 
 Over the evenly rounded surface of this clay is spread a 
 thin sheet of quartz pebbles, generally well rounded, fol 
 lowing all the sinuosities of the surface on which it Jies, 
 though the bed varies much in thickness. Over this pebble 
 sheet is a concentric coat several feet thick, of a perfectly 
 structureless, arenaceous clay, consisting of decomposed 
 feldspar and fragments of quartz, deeply colored by ferric 
 oxide, and resembling the unwashed and unassorted pro 
 duct which would result from the mechanical trituration of 
 decomposed gneiss, with a mixture of the clays just de 
 scribed. All the hillocks were covered by the same ma 
 terial. At Porto das Caixas we rise by a steep incline 
 some thirty feet, more or less (I have no note of the exact 
 height), to a level plain of large extent. Cuttings at 
 the railroad depot show that it is composed of a horizontal 
 deposit of the same tinted sandy clays we found occupying 
 the centre of the hillocks just described, and that it is over 
 laid by a thick bed of structureless red clay, separated from 
 the underlying deposit by a layer of quartz pebbles. Taking 
 the train, we go westward over this plain some ten miles.* 
 
 * Being unfurnished with a reliable map, I can only give distances ap 
 proximately. 
 
THE PROVINCE OF RIO DE JANEIRO. 21 
 
 Part of it is dry and sparsely wooded, but there are large 
 areas occupied by swamps, in which grows a more or 
 less luxuriant swamp vegetation, a dismal scene, present 
 ing sickly-looking trees loaded with orchids, ferns, and 
 other parasites, and draped heavily with a species of Til- 
 landsia. By and by we reach the end of a spur from the 
 Serra do Morro Queimado, and, skirting the western base of 
 the hills, gradually ascend, passing from the plain to the 
 gneiss valley of the Rio Macacu, and, at the present ter 
 minus of the railroad at Cachoeiras, find ourselves in a 
 narrow valley, among gneiss hills, at the foot of the Serra 
 do Morro Queimado, an eastward continuation of the Serra 
 dos Orgaos. The gneiss here is of the same general 
 character as that at Rio. An observation by the river 
 side, near the residence of Mr. Williams, the superin 
 tendent of the railroad, gave the strike N. 70 E. The 
 dip was vertical. Garnets were very abundant. 
 
 On examining the soil, whether high up on the slopes, in 
 the hills, or elsewhere in the numerous cuttings of the roads, 
 &c., the rock is seen to be covered by the same red clay 
 that we have observed forming the surface-coating of the 
 plain at Porto das Caixas ; and only a superficial examina 
 tion of the intervening country is necessary to show that 
 the same red clay covers, not only the clay hillocks of the 
 railway extension and the plains described, but the hills also, 
 descending everywhere to the level of the sand-plain. In 
 addition, it is to be seen frequently underlaid by an irregu 
 lar sheet of quartz pebbles, and it very often contains an 
 gular fragments of quartz and sometimes masses of gneiss, 
 though these last are rare. It is absolutely structureless, 
 and, much as it may resemble the stratified clays above de 
 scribed, or soft, decomposed gneiss, it needs only a very short 
 
22 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 experience to enable one to distinguish from a hand specimen 
 whether it be one or the other. Generally this structure 
 less clay is deep red ; but on the surface it is often more 
 or less yellowish. The following ideal section illustrates 
 the geology of the country between the Bay of Rio and the 
 Serra do Morro Queimado, along the route over which we 
 have travelled. 
 
 w 
 
 SECTION ALONG THE LINE OF THE CANTAGALLO RAILROAD. 
 
 a. Gneiss, Eozoic. 
 
 b. Stratified clays, Tertiary. (Drift, Agassiz) 
 
 c. Drift. 
 
 d. Raised beaches with recent shells. 
 
 e. Modern mud deposit. 
 f. Bay of Rio. 
 
 This section tells a very clear story. Late in the tertiary 
 period, long after the hills of Rio were upheaved, and when 
 the country stood at a slightly lower level, the stratified 
 clays which I have described were deposited over the 
 whole area of the basin of the Bay of Rio, and the adjoin 
 ing flat country. These were afterward upheaved, most 
 probably, as I shall attempt to show hereafter, to a much 
 greater height than at present, and subjected to denudation 
 by water and glacial action. Owing to the shape of the basin 
 and the great number of streams flowing into it, as well as 
 to the fact that it must have been the focal point toward 
 which the glaciers of the encircling hills converged, it is not 
 to be wondered at that, along this part of the coast, the clays 
 have been so completely denuded and swept away, leaving 
 only small patches fringing the shores in certain somewhat 
 
THE PROVINCE OF RIO DE JANEIRO. 23 
 
 sheltered positions. It seems an argument in favor of the 
 prevalence of glacial action in the region, that the denudation 
 is so complete, else we should have expected to find the sheet 
 largely cut through by valleys, leaving more numerous outly 
 ing masses. As to the age of these clays, more hereafter. I 
 have referred them to the Tertiary, though Professor Agassiz 
 is inclined to regard them as drift. The superficial red 
 clay deposit I believe, with Professor Agassiz, to be drift. 
 The stratified sands were deposited in shallow water when 
 the sea stood only a few feet higher than at present, and 
 they have been elevated by a recent rise of the coast, a 
 rise which I believe to be still going on. The recent muds, 
 now accumulating from sediment brought down by the 
 streams, form a marshy fringe outside the raised beaches. 
 Having introduced the reader to the general geological 
 and topographical features of the vicinity of Rio de Janeiro, 
 let us now examine somewhat closely the drift phenomena 
 observable there. In the following remarks I have pur 
 posely confined myself to the results of my own personal 
 observations made during extended excursions over the 
 country near Rio, and a detailed examination of every 
 cutting on the Dom Pedro II. Railroad, from the Great 
 Tunnel to Ypiranga, on the results of which survey I have 
 made a long report to Professor Agassiz. In connection 
 with this subject we will take into consideration some 
 of the topographical elements so remarkable in the hills 
 of Rio and the Serra do Mar. If we examine the 
 gneiss hills at Rio de Janeiro and the vicinity, we 
 find that they are invariably covered, where the slopes 
 are not too steep, by the same coat of red soil which 
 we have observed on the Cantagallo Railroad. This 
 may vary more or less in the coarseness or fineness of 
 
24 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 its ingredients, but it invariably presents everywhere the 
 same general character of a sheet of structureless, unar- 
 ranged material, composed of ground-up gneiss, perfectly 
 devoid of stratification, and always of a deep red color pass 
 ing into yellow near the surface, especially where the 
 material is sandy and light. There is rarely any humus, 
 because the decay of vegetable matter is too rapid to allow 
 of its accumulating as a soil, as in northern countries. This 
 clay sheet varies in thickness from a few feet to one hun 
 dred ; sometimes it is stiff and bakes very hard, in other 
 cases it is more sandy and light. Usually it is quite free 
 from admixture with boulders, but sometimes angular frag 
 ments of quartz of considerable size occur in it, together with 
 rounded or angular masses of gneiss or diorite. The latter 
 rocks are almost always in a decomposed state, and, except 
 to the experienced eye, are recognizable only in fresh cut 
 tings. Under this clay one sometimes finds, as in the cut 
 tings at Tijuca, a thin layer of quartz pebbles like that seen 
 on the Cantagallo Railroad, but this is not always present. 
 The surface of the gneiss on which the drift rests is always 
 moutonnee and remarkably evenly rounded down, and the 
 sheet of quartz pebbles lies immediately upon it, following 
 all its curves ; but the pebble sheet may be wanting over 
 large areas, or vary very suddenly and irregularly in thick 
 ness. The gneiss in situ is almost invariably decomposed 
 beneath the drift to a depth varying from a few inches to 
 one hundred feet. The feldspar has been converted into 
 clay, the mica has parted with its iron, &c., but the altered 
 crystals of the gneiss still occupy their relative position 
 with reference to one another. The planes of stratification 
 are well marked, and the veins of quartz, though cracked 
 up, remain in place. 
 
THE PROVINCE OF EIO DE JANEIRO. 25 
 
 This extraordinary decomposition of the Brazilian gneiss 
 and other rocks has long attracted attention, and Darwin 
 has described it very accurately in his Geological Observa 
 tions. He was of the opinion that it had taken place under 
 the sea before the present valleys had been excavated.* Pis- 
 sis also described it ; but it appears to me that he has greatly 
 over-estimated the depth to which the softening of the rocks 
 has extended, when he says that, in the gneiss region between 
 Sao Fidelis and the Serra dos Orgaos, the gneiss has been 
 decomposed to a depth, in some places, of 300 metres ! ! f 
 
 This decomposition results, in my opinion, from the ac 
 tion of the warm rain-water soaking through the rock, and 
 carrying with it carbonic acid, derived not only from the 
 air, but from the vegetation decaying upon the soil, together 
 with organic acids, nitrate of ammonia, &c. I believe that 
 the remarkable decomposition of the rocks in Brazil has 
 taken place only in regions anciently, or at present, covered 
 by forest. Heusser and Claraz have suggested that it is 
 aided by nitric acid. They say : " It is, without doubt, de 
 termined by the violence and frequency of the tropical rains, 
 and by the dissolving action of water, which increases with 
 the temperature. It is necessary to observe, moreover, that 
 this water contains some nitric acid, on account of the thun 
 der-storms which follow each other with great regularity 
 during many months of the year." | 
 
 Professor Agassiz has discussed this subject at some 
 
 * Darwin, Geological Observations, p. 144. 
 
 t Pissis, Mem. de I Inst. de France, Tom. X. p. 358. It is well to remember 
 tbat, before the glacial origin of the clays overlying the decomposed rock was 
 pointed out by Professor Agassiz, the thickness of these clays was included 
 in the estimate of the depth to which the decomposition had taken place. 
 
 J MM. Ch. Heusser and G. Claraz, Ann. des Mines, 5 m< Serie, Me m., Tom. 
 XVII. p. 291. 
 
 VOL. i. 2 
 
26 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 length in the Journey in Brazil, his opinion being that the 
 softening of the rock is due to the action of warm rain-water. 
 It has been objected to this theory that the stone employed 
 in Brazil for building purposes endures remarkably well, 
 showing, after the lapse of centuries, very little change. 
 This argument, in my opinion, is of very little weight, for 
 a smooth, naked surface from which the water runs off 
 rapidly, and which is, the greater part of the time, dry, is 
 placed under conditions very different from those of the 
 gneiss overspread by a thick coating of wet drift paste, 
 and constantly soaked with water. 
 
 Brazil is not the only country in which the rocks have 
 softened to a great depth. The same phenomenon has been 
 observed in the Southern States of the Union and in India.* 
 I have seen gneiss decomposed to a depth of several feet in 
 the vicinity of New York. 
 
 When the gneiss is fine-grained, homogeneous, and not 
 distinctly stratified, it is often difficult to distinguish the 
 rock decomposed in situ from the drift ; but the line of 
 demarcation between the two, even when not marked by 
 the pebble sheet, is usually easily distinguishable even in 
 old cuttings. The slope of a railway cutting is apt to gully 
 out along the line of junction of the drift and the underly 
 ing decomposed rock. This line is invariably gently undu 
 lating, and one never sees jagged edges of strata or angular 
 masses projecting upward into the drift. 
 
 Professor Agassiz has spoken of the valley of Tijuca below 
 Bennett s as a locality where the drift is very beautifully 
 exhibited. The mountain mass of Tijuca is separated from 
 the mass of the Corcovado group by the pass of Boa Vista, 
 
 * Dr. Benza says that in the Ncelgherries granite is sometimes decomposed 
 to a depth of forty feet. (Madras Journal of Literature, &c., Oct. 1836, p. 246.) 
 
THE PROVINCE OF RIO DE JANEIRO. 27 
 
 which is about 1,000 feet above the sea. Eastward runs 
 the valley of Andarahy downward toward the city, while 
 to the westward one descends a most romantic valley to 
 a great alluvial plain, in a sort of bay or amphitheatre 
 among the hills. This valley, occupied by a tumbling 
 mountain stream, descends very rapidly, ending abruptly 
 below the Cascate Grande at some height above the plain. 
 
 THE CASCATINHA AT TIJUCA. 
 
 Minor valleys from Tijuca and the mountains to the 
 south descend and join this valley. At Bennett s the drift 
 clay is full of boulders of quartz, gneiss, and greenstone. 
 If we ascend the brook which flows through Mr. Bennett s 
 
28 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 fazenda for a few rods, we shall find that it has cut its bed 
 through the general clay sheet which everywhere covers 
 the hills, and it is perfectly easy to see that this sheet is in 
 no place a deposit thrown down by the brook. This loose 
 material consists of a brownish or reddish earth without the 
 slightest signs of stratification, in which are buried boulders 
 of gneiss, usually rounded and of many qualities, together 
 with rounded masses of quartz. 
 
 SECTION OF DRIFT AT BENNETT S, TIJUCA. 
 
 At c is a very large boulder of homogeneous, unlaminated, 
 fine-grained gneiss, from around which the drift has been 
 washed away, and this rests on a mass, in situ, of gneiss 
 coarsely crystallized (porphyritic) and well laminated. At 
 e we see the gneiss in situ, much decomposed, and laid bare 
 in a cutting. 
 
 The solid gneiss beds are well seen in the bed of the 
 brook, and the rock is generally quite coarse and porphyritic 
 in structure. I find a note of a strike N. 55 W., which, from 
 the generally uniform northeastward strike seen elsewhere, 
 might be suspected to be an incorrect observation ; but at 
 the bridge, a little farther down the brook, I found the 
 strike to be N. 10 "W. If we follow down the valley we 
 shall find the soil full of boulders, and some of these are 
 many feet in diameter. 
 
 As we descend the valley still further these boulders are 
 seen lying bare, not only in the brook where the water has 
 washed away the loose material, but on the hillsides. I 
 
THE PROVINCE OF RIO DE JANEIRO. 29 
 
 think that no geologist familiar with drift phenomena, who 
 should suddenly find himself in this valley, would have even 
 the slightest suspicion of their being anything else than the 
 most ample testimony of the former prevalence of glacial 
 action over the region ; yet, in the beginning of his drift 
 studies in Brazil, he is almost sure to commit some 
 gross blunders, for it is not even a general rule that the 
 loose boulders found on the surface are erratics. On my 
 first visit to Tijuca, very soon after my arrival in Brazil, and 
 after Professor Agassiz had announced the discovery of 
 drift at Rio, I was struck with the appearance of some 
 trap masses on a hillside near Bennett s, which looked re 
 markably like erratics ; but a close study of them satisfied 
 me that they resulted from the surface decomposition of 
 a great trap dyke. Not descending far enough into the 
 valley, and satisfying myself that a great proportion of the 
 gneiss masses that I examined at the time were not erratics, 
 I came most decidedly to the conclusion that the surface 
 deposits of Rio were not drift, but were in some way due to 
 the decomposition of the rock, as had been heretofore sup 
 posed. I desire to record here the fact, that I began my 
 studies of the Brazilian drift with a conviction that Pro 
 fessor Agassiz was wrong, and I feel much gratified that 
 my independent observations have so fully confirmed the 
 results of his own. If one descends the valley towards the 
 Cascate Grande, he will see that the valley is heaped with a 
 confusion of immense boulders tumbled one upon another ; 
 masses of greenstone, weighing hundreds of tons, piled up 
 with those of gneiss of all qualities. Where these are bare 
 they are always rounded, as is seen in the engraving, but I 
 believe that this is referable, to a very large extent, to a 
 concentric and even decomposition of the surface ; but 
 
GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 BOULDEIiS AT TIJUCA. 
 
 there is no resisting the conclusion that we have here a 
 morainic deposit from a glacier which anciently occupied the 
 valley. The above woodcut is from a stereograph published 
 by Leuzinger at Rio, and represents the boulder masses above 
 the Cascate Grande. Descending to the plain below the 
 Cascate, w r hich is also seen in the woodcut, one traces the 
 drift clays and boulders quite down to the plain, when they 
 end abruptly, and the flat lands are seen to be of alluvial 
 origin resting on sea sands, of the same age as the sands of 
 Paqueta and the Cantagallo Railroad extension. One can 
 not find in the plain, nor did I ever see anywhere on the 
 beach sands, either drift clays or boulders. 
 
THE PROVINCE OF RIO DE JANEIRO. 
 
 31 
 
 Returning to the Dom Pedro II. Railroad, we may trace 
 the drift sheet everywhere from Belem to the Barra do 
 Pirahy, over the whole Serra do Mar ; and one may here 
 study its structure in the most detailed way. From Bclem 
 to the Parahyba River the same red clay entirely covers 
 the surface, lying even on very high slopes. Nowhere is 
 there the slightest sign of stratification, and it is sharply 
 defined from the alluvial deposits of the river. The same 
 pebble sheet is seen almost everywhere, though in cutting 
 after cutting it may sometimes be wanting. Boulders are 
 rare, and are almost invariably so decomposed as to be 
 seen only in . fresh cuttings. The rock on which the drift 
 rests is always smoothly and evenly rounded down. The 
 following diagram will illustrate the structure of one of 
 the gneiss hillocks of the Serra do Mar. 
 
 IDEAL, SECTION OF DRIFT-COVERED GNEISS HILL. 
 
 a a. Drift clay. 
 
 ff. Angular fragments of quartz. 
 
 b b. Pebble sheet lying on rounded surface of gneiss. 
 
 d d. Gneiss in gitu, but decomposed. 
 
 e e. Gneiss undecomposed. 
 
 g ff. Quartz and granite veins traversing both solid and decomposed gneiss. 
 
 Among the first elements in the Brazilian mountain 
 scenery which attract the attention of the observer on 
 
32 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 approaching the coast near Rio are the bare rock slopes 
 which, instead of forming ragged precipices, as in northern 
 latitudes, are most remarkably smooth, and devoid of irregu 
 larities ; and these surfaces, where the rock is homogeneous 
 in texture, usually have a high slope, and frequently de 
 scend and dip beneath the sea level, the sea washing over 
 their even faces.* Sometimes the top of a hill is bare, 
 evenly moulded, and round, or a rounded boss projects 
 from the side of a hill. These smooth cliifs and rock-slopes 
 are heavily striped above the sea by broad lines drawn down 
 their faces, blackened by a sort of cryptogamic growth ; 
 so that these slopes, of a rich, rather purplish, black tint, 
 show very strangely, especially when set into the mantle of 
 verdure which covers the hills. Of this character are the 
 bare mass of the Sugar-loaf, the precipices of the Corcovado, 
 and the steep slopes of hundreds of hills in the vicinity. 
 Approaching nearer to one of these slopes, we find that its 
 surface is often scored by a system of little rain-courses, 
 and is covered by cactuses and bromeliaceous plants. 
 
 The Pedra Bonita is a bare mass of rock opposite the 
 Gavia, but not so high. It is partly surrounded by almost 
 vertical slopes, and on top, over a large area, is bare and 
 rounded off. Here, though one may observe the same 
 thing elsewhere, the way in which this rounding and 
 smoothing down of the rock is produced may be studied, 
 and one is soon forced to believe, that, whatever the gla 
 ciers may have contributed to the shaping of the topo- 
 
 * These evenly rounded, wave-washed rock slopes are very interesting, and 
 have already been called attention to by Darwin (Geological Observations, p. 
 144). Such slopes may be seen not only on the shores of the quiet bays, but 
 exposed to the full wash of the Atlantic waves, as at the mouth of the Bay of 
 Rio and elsewhere. In describing, farther on, the Bay of Espirito Santo, I shall 
 show how these wave-washed slopes originate. 
 
THE PROVINCE OF RIO DE JANEIRO. 33 
 
 graphical features of the country, the general moulding of 
 the hills has been due primarily to subaerial denudation. 
 The gneiss of the Pedra Bonita is decomposing. Where the 
 rock is level the decomposed feldspar, <fec. has been washed 
 away by the rain, and we have the rock covered by a thin 
 coating of loose quartz grains, which I hardly need say are 
 angular. Where there is a decided slope, the loose sand is 
 washed or blown away. The rock itself is much softened 
 on the surface, and, to a considerable depth, the feldspar 
 has been more or less altered. This semi-decomposed layer 
 forms a concentric coat over the whole rock. Sometimes 
 this is continuous and unbroken, but if the area exposed be 
 large, we usually find that there is a tendency for it to sep 
 arate itself from the undecomposed rock below, and to 
 crack up. Sometimes this layer is only a few inches in 
 thickness, in others it may be several feet. When the 
 surface is horizontal, or nearly so, the tendency is for it 
 to break up finally into small angular pieces, which, wasting 
 away and rounding down by decomposition, cover the rock 
 with loose, boulder-like masses, or are entirely removed, 
 leaving a smooth, unencumbered surface. If the slope be 
 very steep or vertical, the decomposed mass may fall or slide 
 off. On very steep slopes, as those of the Sugar-loaf, Cor- 
 covado, Morro de Santa Theresa, and elsewhere, this sheet 
 scales off and falls, breaking up below T . The Brazilian gneiss 
 cliffs rarely have a talus of broken fragments at their base. 
 The decomposition going on evenly all over their face gives 
 only sand and clay, washed down by the rains and distrib 
 uted over the earthy slope below, and the half-decomposed 
 fragments soften and finally decompose entirely. One must 
 be very careful in examining a cutting that runs under a high 
 cliff, for the earth resulting from the decomposition of the 
 
 2* C . 
 
34 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 face of the cliff resembles more or less the drift-earth spread 
 over the surface of the ground. Oftentimes a mass of 
 half-decomposed rock separates itself from the face of the 
 cliff in a great lenticular sheet. This may crack across 
 horizontally, particularly if the plane of stratification cuts 
 the surface of the cliff in this way, and the lower half may 
 drop off, leaving an overhanging portion attached to the 
 cliff. One may observe hanging masses of this kind at 
 tached to the precipice of the Sugar-loaf and Corcovado, 
 and in innumerable other gneiss localities. If we exam 
 ine one of the rounded gneiss hills, as, for instance, one 
 of those just behind the Dom Pedro II. Hospital at Bota- 
 fogo, where part of the hill is bare and steep, looking 
 at the hill in a cross-section, we may observe that the gen 
 eral rounded curves of the drift-covered portion are quite 
 out of harmony with those of the bare portions, which are 
 usually flatter, and we may further notice that the very steep 
 slopes or precipices are usually on the side of the hill away 
 from which the strata dip. This is the case on both sides 
 of the Botafogo valley. In the following diagram I have 
 tried to represent the difference in moulding between a 
 glaciated surface and that of one of these bare cliffs. 
 
 DIAGRAM TO SHOW CONTRAST IN MOULDING BETWEEN DRIFT-COVERED AND 
 BARE SURFACES. 
 
THE PROVINCE OF EIO DE JANEIRO. 35 
 
 The outline of the drift-covered portion of the hill, a b, 
 has been unchanged since the drift ; for though the rock has 
 been protected by the drift, the gneiss has decomposed, but 
 has remained in situ. Over the face, b c, the decomposi 
 tion has gone on also ; but since the slope has not been 
 protected, it has worn constantly backwards, growing grad 
 ually more regular in outline as each new decomposed coat 
 was thrown off. If one will take the pains to examine 
 the curves resulting from the decomposition of bare sur 
 faces, and those which result from decomposition where, 
 as under a drift-sheet, the decomposed rock remains in sitti, 
 he will be convinced that the moulding of the surface of the 
 gneiss on which the drift rests is of an altogether different 
 character from that resulting from simple subaerial denu 
 dation. 
 
 The sands containing recent shells along the Bay of Rio, 
 and elsewhere, and which rise to the height of a few feet 
 above high-water mark, bespeak a recent elevation of the 
 coast, as has been observed independently by his Majesty the 
 Emperor, Dr. Capanema,* and myself; but there are other 
 proofs of the same upheaval in the holes excavated by sea- 
 urchins, which are found in the vicinity, in many places 
 many feet above high-water mark. 
 
 The islands of Maricas lie a few miles off the coast 
 
 * Dr. G. S. de Capanema has written more or less on the geology of Brazil. 
 One of his papers mentioned by Burton bears the title, " Decomposifao dos 
 Penedos no Brasil," and was published at Rio in 1866. I am very sorry that I 
 have never seen any of the papers of this geologist, who has travelled very ex 
 tensively over the Empire. I know of his work only through quotations or 
 references. From a MS. of Major Coutinho, placed in my hands by Professor 
 Agassiz, I should infer that Dr. Capanema is a disbeliever in the glacial origin 
 of the surface deposits claimed by Professor Agassiz and myself to be drift, and 
 that he rather considers them to be the result of decomposition alone. 
 
36 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 between Rio and Cape Frio. They are composed of 
 gneiss in beds of unequal hardness, and a most excellent 
 opportunity is afforded for the comparison of glaciated 
 rock surfaces and surfaces denuded by subaerial decom 
 position, or wave action. The general surface of the isl 
 ands is rounded, and is covered by drift clay. Some bare 
 rocks are evenly rounded, but along the shore, \vhere the 
 waves beat, the softer beds are worn away more deeply than 
 the hard ones, and the surface is very irregular. On the 
 rocks a species of Echinometra (E, Michelini Desor), a sea- 
 urchin with moderately long, dark-purple spines, is very com 
 mon, living in a zone extending downwards from some dis 
 tance below high-water mark. Here they are found, each in a 
 cup-shaped depression worn in the rock, and in some places 
 the rock is fairly honeycombed on the surface by their 
 " nests." Above the zone of living sea-urchins the nests 
 are found, but they are empty, and in protected localities, 
 as, for instance, in narrow openings between rocks, they 
 may be observed to extend to a height of several feet above 
 high-water level, showing that the elevation of the coast has 
 been very recent. I have observed that the nests appear 
 less and less well preserved the higher we examine them, 
 which has convinced me that the upheaval has been a 
 gradual one, and I believe that it is still in progress. 
 
 The coast between Rio and Cape Frio is, for a large part, 
 high and exceedingly picturesque. Many of the hills are 
 bare and strangely shaped. The following sketch represents 
 the coast as seen from near the island of Maricas. To the 
 left are the hills of Rio. 
 
 Long sea-beaches stretching across bays formed by the 
 hills have formed lagoons along this coast, and the low 
 grounds between the mouth of Rio and Ponta Negra 
 
THE PROVINCE OF EIO DE JANEIRO. 
 
 37 
 
 COAST JUST EAST OF EIO. 
 
 are largely occupied by swamps and lagoons. Of the latter 
 the most notable is the Lagoa de Marica, which is some six 
 miles long, salt, and separated from the sea, like the LagQa de 
 Freitas, by a sand-beach, through which the inhabitants are 
 obliged occasionally to cut, in order to give passage to the 
 waters of the lake during times of freshets. This lake, as 
 well as the Others along this coast, is extraordinarily rich 
 in fish. 
 
 A sandy shore extends from west to east from Ponta 
 Negra to Cape Frio. " It may be eight or ten leagues wide 
 from the sea to the cordilheira, being roughened in this in 
 terval by several serras, and occupied in parts by various 
 lagoas. All the flat part of this plain is useless for agricul 
 tural purposes on account of the depth of the sand, and its 
 being overflowed a part of the year." * This low country ap 
 pears to extend along the Rios Sao Joao and Una to the 
 sea north of Cape Frio. 
 
 Lagoa Saquarema is a large lake lying east of Ponta Ne 
 gra, and separated from the sea, like the Lagua Marica, by a 
 sand strip. " It is three (Portuguese) miles long, and three 
 
 * Diccionario Geografico, art. Marica. This work is a mere compilation, 
 and is full of inaccuracies. In quoting from it I have done so with much care. 
 
38 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 quarters of a league in its greatest width. It is salt, full of 
 fish, and separated from the sea by a narrow tract of sandy 
 ground. When the adjoining plains begin to be covered 
 by the floods of the streams which empty into it, the inhab 
 itants of the vicinity open an outlet to the ocean at the 
 eastern extremity, which remains an unfordable river during 
 the winter, at the end of which the surf closes it up." * 
 
 Lag6a Araruama is a narrow strip of salt water about 
 twenty miles long, and with a varying width, in some places 
 several miles, lying on the coast west of Cape Frio, and 
 separated from the sea by a narrow strip of sand thrown 
 up by the waves. Cazal says f that " it empties itself 
 from its eastern extremity by a canal twenty-six bra^as 
 large at its mouth, which is situated at a distance of a 
 league and a half north of Cape Frio. Notwithstanding 
 that a number of streams empty into it, its waters are salted 
 by the communication which they have witlf the ocean. 
 The tide makes itself felt as far as Ponta Grossa, which is 
 situated at about its middle. Thence westward the waters 
 go with the winds. It abounds in fish of various kinds. 
 In some parts it is many brac,as in depth, J in others 
 one may wade from one shore to the other." Milliet says 
 that " between the sea, the city, and the Lake Ararauma 
 are natural salines, which were prohibited by royal letters 
 on the 26th of February, 1690, and 18th of January, 1691, 
 the Portuguese government intending thereby to favor the 
 
 * Coroyrafia Brazilian, Tom. II. p. 38. This old work by Cazal, published in 
 1818, is very much more reliable than the Diccionario Geografico. 
 
 t Corografia Brazilica, Tom. II. p. 38. 
 
 | Prinz Max. zu Neu-Wied, Reise nach BrasHien, 1" Band, 85" Seite. " "Wir 
 fauden das Wasser der Lagoa von geringer Tiefe und so klar, dass wir 
 den weissen Sandboden des Grundes mit seinen Korallengewiichsen deutlich 
 wahniehmen kormten ; bci der gcringen Tiefe sassen wir oft fest." 
 
THE PROVINCE OF EIO DE JANEIRO. 39 
 
 commerce of salt of its own European possessions. Not 
 withstanding the royal decrees, Domingos da Silva Ribeiro, 
 judge-in-ordinary of this city, ordered, in 1768, that the 
 communication between Lake Araruama and the salines of 
 Mac,abamba be closed, and in the following year the above 
 salines furnished in six months 50,000 alqueires of salt." 
 
 Cape Frio is the name given to the southernmost point of 
 a high, precipitous gneiss island, situated at the angle where 
 the coast line, coming eastward from Rio, bends northward 
 toward Cape Sao Thomd. It is only about three miles long, 
 very irregular in outline, and is almost divided into two 
 parts. The northern is, according to Mouchez, 394 metres 
 in altitude. A lighthouse was erected on this point, but it 
 
 CAPE FRIO. 
 
 proved to be at too great a height, being above the level of 
 the clouds, so that it had to be abandoned. The present 
 lighthouse is situated at a lower level on the southern point. 
 The island is separated from the main-land by a narrow 
 but rather deep channel. The land opposite consists of a 
 group of gneiss hills, formerly islands, which have been 
 united together by sand beaches and sand plains, which 
 extend northwestward, joining the beach of Mac,ambam- 
 ba, or Massambamba, and running northward to the Rio 
 Itajuru. 
 
40 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Between the city of Cape Frio and Os Buzios the coast is 
 very irregular in outline, rather low, with small gneiss hills 
 along the shore, forming occasional rocky promontories 
 which are united by curving sea-beaches. I find in my 
 notes the query whether the low flat lands seen from the sea 
 may not be Tertiary. Towards Armagao the coast is bor 
 dered by low gneiss hills, which, owing to the northwest 
 (landward) dip of the rocks, and the correspondence of 
 their strike with the general trend of the coast, as well as 
 from the way in which they have been worn by the waves, 
 present to the sea perpendicular, rugged cliffs of no great 
 altitude. Along this coast are quite a number of little 
 rocky islets of gneiss. 
 
 At Os Buzios * the coast line suddenly bends off to the 
 northwestward, and runs towards Sao Joao, when it curves 
 round and sweeps off with a northeast trend to Cape Sao 
 Thome*. The gneiss shows itself on the shore of the point 
 just east of the town. It is well laminated, much plicated, 
 and has a general low dip to the northwestward. The 
 gneiss of the point on the western side of the town has very 
 much the same character. The gneiss behind the town 
 seems to be overlaid by tertiary clays, which show them 
 selves in the point west of the town ; but I had no opportu 
 nity of examining them with much care. They are so much 
 denuded that they are not easily recognizable by topograph- 
 
 * Os Buzios takes its name from its richness in shells. Among the species I 
 collected there were Cassis Madagascariensis and Cyprcca exanthema. Woodward 
 in his Manual says that no Cyprseas occur on the Brazilian coast. Cypraea 
 exanthema is not at all rare, and occurs also at Bahia, where I have found an 
 other little species in great abundance. Several species of corals occur at Os 
 Buzios ; Mil/epora akicornis is especially abundant, and the rocks are covered 
 with patches of the common Palijthoa, together with a Zoanthus with an emer 
 ald disk, common elsewhere on the coast. 
 
THE PKOYINCE OF RIO DE JANEIRO. 41 
 
 ical features alone. There is a small hill just east of the 
 village which bears a little church. It is cut away on the 
 shore by the waves, and forms a little bluff composed of 
 rounded quartz pebbles derived from the rocks of the 
 vicinity, cemented together loosely by a soft, greenish 
 clay. I have seen nothing like it elsewhere. Westward 
 of Os Buzios is quite a range of gneiss hills, lying between 
 the head- waters of the Rios Garcia and Trapiche. The 
 shore beyond the rocky point west of Os Buzios appears, for 
 some distance, to be tertiary, but there are some patches 
 of recent sands. These deposits appear to terminate on the 
 shore with a bright red cliff, very conspicuous from a 
 distance, when the tertiary bluffs recede from the coast, 
 and a sandy flat, backed by low plains, extends on to the 
 gneiss hills of the Serra de Sao Joao. Thence to Macahe 
 much of the shore is low, with stretches of tertiary clays, 
 more or less denuded, and gneiss hills. North of Macahd 
 the tertiary plains soon recede from the coast running off to 
 ward Campos on the Rio Parahyba do Sul, the land border 
 ing the coast being flat, more or less swampy, and diversified 
 by numberless shallow lagoons, some of which are of large 
 extent. Off Macahe, and distant a few miles from the 
 shore, lie the little gneiss islands of Santa Anna, known 
 to coasters as their only retreat, north of Os Buzios, when 
 storms or northeast winds prevent them from passing Cape 
 Sao Thom6. There is usually a strong current off the 
 Cape, but it is very variable. During the prevalence of 
 a long northeast blow it runs southward with such rapidity 
 that it is impossible for coasting vessels to tack against 
 it. They may succeed in beating up from the islands 
 of Santa Anna, close to the shore, until they reach the 
 Cape, but they are then swept back. I was once nearly 
 
42 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 a fortnight beating off this coast during the prevalence of a 
 northeast breeze, and I visited the islands many times. 
 They are of no very especial interest, but one may observe 
 here the sea-urchins nests, raised above high-water mark, 
 and may study some of the topographical features developed 
 by decomposition. I add a little sketch of the westernmost 
 island, as seen from the northeast, to show the smoothly 
 rounded character of its steep sides. 
 
 ONE OF THE ISLANDS OF SANTA ANNA. 
 
 The Serras, always clothed with the virgin forest, stretch 
 along, at varying distances from the coast, in a magnificent 
 range of hills, with steep slopes toward the sea, forming one 
 of the grandest panoramas of mountain scenery on the 
 coast of Brazil. Of the altitude of these hills I have no pre 
 cise information, but I should estimate some of them as at 
 least 6,500 feet. There is one very conspicuous, obelisk-like 
 peak, lying behind Macahe, but standing somewhat in 
 advance of the general range of the Serra, and called the 
 Frade de Macahe . Lieutenant Mouchcz* gives its altitude 
 on one of his charts as 1,750 metres (5,745 feet), which 
 
 * Lieutenant Mouchez has on his charts given the heights of various hills 
 along the coast, but I find no note as to whether they are the results of actual 
 measurements. 
 
THE PROVINCE OF RIO DE JANEIRO. 43 
 
 would make it almost as high as the Serra da Piedade in 
 Minas. The Serras break down on reaching the valley of 
 the Parahyba River, just below Sao Fidelis, when, in de 
 tached Serras, they recede somewhat from the coast, cross 
 the Muriahd and Itabapuana rivers, tying in with the Serras 
 of Itabapuana and Itapemerim, and forming some very 
 picturesque mountain scenery. 
 
 All the flat, sandy, and swampy land, interspersed with 
 lagoons, which borders the Rio Parahyba almost to Campos, 
 a city some miles above its mouth, as well as that which 
 stretches southward to Macahe, or thereabouts, is of very re 
 cent formation, and is composed principally of sands and silt 
 brought down by the river. Off the coast of Sao Thomd, 
 as well as for some distance northward, the water is very 
 shallow, and much discolored. These lands are bordered 
 by a long, dreary sand-beach.* The country behind is, 
 to a considerable extent, covered by shrubbery and trees, 
 but there are extensive open plains where herds of cattle 
 graze. By means of a ditch uniting the lagoons, and dig 
 nified by the name of a canal, water communication with 
 Macahe has been opened,! and a considerable trade is 
 carried on between the two places, or the settlements on the 
 route, by means of canoes. Much of this swampy ground is 
 excellent for rice, an important product of this part of 
 the country. 
 
 The lakes of this region are very numerous. They are all 
 shallow, but some are several leagues in diameter. The 
 
 * Owing to the northeast trend of the beach south of Cape Frio, and the 
 prevalence of northeast winds, the waves strike the beach obliquely, and there 
 is a tendency for the sands to move southward. There are no dunes here. 
 
 t I have understood that this canal was finished, but Pompco says that in 1864 
 it was only two thirds completed. 
 
44 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 largest is Lagoa Feia, an irregular lake some twenty miles 
 long, lying about ten miles south of Campos. Owing to its 
 great area and its very small depth, its waters are kept con 
 stantly turbid through their agitation by the winds. It re 
 ceives from the west quite a little stream, the Rio Macabu, 
 which rises near the Serra do Frade. This river is navi 
 gable for canoes for some twenty miles above the lake. 
 Another mountain stream of considerable importance rises 
 among the Serras just north of the Macabu, and uniting in 
 its course several large lakes, also enters Lagoa Feia. 
 This lake is united on all sides with a multitude of lagoons 
 of greater or less size by a perfect network of little channels, 
 so that its waters flow partly to the north into the Para- 
 hyba, while in part they escape into a system of long narrow 
 lagoons that stretch along just behind the beach ridges of 
 the shore at Cape Sao Thome, and communicate by channels 
 across the beach with the sea. One of these lagoons, which 
 passes by the name of the Rio Iguassu, is some fifteen miles 
 long. It has evidently been formed by the throwing up in 
 very recent times, probably during the prevalence of some 
 very heavy storm, of a line of sand-beach just outside of the 
 shore. Similar lagoons are found elsewhere along the 
 coast, as for instance just south of Belmonte, on the 
 Jequitinhonha, and I believe that the great line of sand- 
 beaches stretching along the coast Avas thrown up, to begin 
 with, by an extraordinarily heavy storm which prevailed 
 along the whole coast, and which, in many instances, where 
 the water was very shallow, disturbed the bottom at some 
 little distance outside of the shore line, throwing up a sand 
 barrier which, through the drifting of sands by the winds, as 
 well as by the action of the waves, has since reached its pres 
 ent dimensions. During the overflow of the Parahyba its 
 
THE PROVINCE OF RIO DE JANEIRO. 45 
 
 waters back up over much of the plain on both sides of the 
 river, and the country becomes largely submerged. At the 
 time of the enchente, or annual freshet, the inhabitants cut 
 outlets across the sea-beaches for some of the lakes in the 
 southern part of the campos. North of the Parahyba, and 
 near the city of Campos, there is another large lake called 
 Lagoa do Campello, and the country thence northward to the 
 Guaxindiba is full of lagoons and cut up by little channels. 
 
 The Parahyba empties into the sea by two mouths, distant 
 some two miles from one another, and between which is the 
 Ilha do Lima. The delta of the Parahyba projects two or 
 three miles beyond the general line of the coast. The 
 mouths of the river are obstructed by bars, over which the 
 waves at times break fearfully, and an entrance can ordina 
 rily be effected only at high tide ; yet small coasting steamers 
 and vessels do enter, and small river steamers, and some 
 times even schooners, ascend as far as Campos. At the 
 mouth on the south side is the miserable little town of Sao 
 Joao da Barra, built on a sand-bank which admits of no cul 
 tivation whatever. It contains some two thousand inhabi 
 tants, who subsist principally by fishery, shipbuilding, and 
 commerce. It owes its importance to the sole fact that ves 
 sels arc often long delayed either off the bar waiting to enter, 
 or inside waiting to go out, and this keeps up a little trade. 
 In its lower course the river is wide and shallow. Mangrove 
 swamps and low grounds, sometimes covered with bushes 
 and trees, often waving with the tasselly spikes of the 
 Uba (Gynerium parvifolium Nees), border it for a few 
 miles, but by and by the banks, which are composed 
 of the richest alluvial clay, grow higher, and thence to 
 Campos the beautiful river is bordered by immense sugar 
 plantations, and the scenery is enlivened by frequent 
 
46 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 fazendas and the fabricas with their tall smoke-stacks. 
 The plains are covered with thick beds of silt, deposited 
 by the river during the annual overflows. 
 
 The waters of the Parahyba, as well as of all the moun 
 tain streams of the province flowing through gneiss re 
 gions, are very turbid and usually of a milky brown 
 ish tinge, throwing down a copious sediment, even in dry 
 times ; but when the river is swollen by rains, the quan 
 tity of silt is very much increased. This material, whether 
 derived from the gneiss rock itself or from the drift, consists 
 chiefly of decomposed feldspar and mica, and the water of the 
 river is glistening with the minute silvery flakes of the latter 
 mineral. The soil deposited by the river is very productive 
 in sugar-cane, and the region round about Campos manu 
 factures a very large amount of sugar and rum, the former of 
 a very good quality ; and this is the principal product of 
 the plains. At Campos the country, though flat, is some 
 what higher, and one may see, from an inspection of the 
 river-bank, that the alluvial deposits are underlaid by 
 tertiary clays which have been more or less denuded. 
 
 Notwithstanding the turbidness of the water of the Para 
 hyba, it is, when the sediment has been deposited, very pot 
 able, and maybe preserved for a long time. The usual cus 
 tom is to keep the water in great earthen jars, sometimes 
 weeks or months before using it. 
 
 The city of Campos is a respectable town of about twenty 
 thousand * inhabitants, built on the right bank of the river. 
 Its trade consists principally in sugar and coffee, and it is a 
 place of extraordinary stir. The vicinity is flat and fertile, 
 
 * At least so says Pompeo in his Geography. As there has been no regular 
 census, it is impossible to give with accuracy the number of inhabitants of 
 Brazilian towns. 
 
THE PROVINCE OF RIO DE JANEIRO. 47 
 
 and largely cultivated for sugar. In the vicinity of Campos 
 the Goyabeira, or Guava-tree of the West Indies (Psidium 
 Guaiava Raddi), is very largely cultivated, and the fruit is 
 manufactured into a sweetmeat which is exported in great 
 quantity from Campos. There are extensive low tracts 
 which in the wet season in part for the whole year 
 form shallow lagoons and marshes. In these marshes, as 
 well as in the ditch, dignified by the name of canal, which 
 runs Macahe-ward, Mr. Copeland and I collected a great 
 abundance of ampullariae, planorbes, &c. The former I 
 found laying in June. The eggs were large and salmon-col 
 ored, and were attached in bunches to the grass. These 
 swamps are also rich in fish, Piabas, Acaras, Trahiras, &c. 
 
 The Rio Muriah^ is a little stream entering the Para- 
 hyba from the north a short distance above Campos, tak 
 ing its rise in the province of Minas. At its mouth it is 
 perhaps 400 or 500 feet wide. It is navigable for a few miles 
 to the first falls. The following observations were made 
 during an ichthyological excursion up the river : The lands 
 in the vicinity of the mouth of the river along the left bank 
 of the Parahyba are alluvial and flat, so far as I could 
 see. Two miles up the river, near the fazenda of the 
 Baroness of Muriatic", the ground rises somewhat above its 
 general level, is hummocky, diversified by immense ant 
 hills, and covered by a red drift soil. 
 
 Ascending a little farther, higher grounds of the same 
 character are cut through, and they are seen to be composed 
 of tertiary clays and sandstones, as may be observed in a 
 bluff near Jundia. At Pestrella the land rises to a height 
 of seventy feet, and is composed below of a thick bed of 
 coarse dark-red sandstone, extending below water-level, 
 over which are beds of whitish and red sandy clays. 
 
48 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 These hills are outliers of the tertiary beds of the coast, 
 much denuded by glacial action. A large part of the lands 
 bordering the river are low, and above the fazenda of the 
 Barao d ltabapuana they consist of beds of horizontally 
 stratified sands above, with irregularly stratified brown 
 clayey earth below. In this last were layers of a dark ma 
 terial, which appeared to be made up of leaves. I may here 
 state that it has been reported that coal, or lignite, occurs 
 in the vicinity of Campos, I have seen no signs of pala?ozoic 
 or secondary rocks anywhere in the province of Rio de 
 Janeiro, and at Campos I could not learn of the existence 
 of any such deposits. The report may have originated from 
 these recent vegetable deposits on the Muriahe. Gneiss 
 shows itself at the fazenda of Piranga. Above the fazenda 
 of Oiteiro the plains cease, and the river winds among gneiss 
 hills, some of which are of considerable altitude ; the Morro 
 do Sapateiro, distant a few miles northwest of the Santa Rita, 
 being perhaps 2,000 feet in altitude, while the Serra da 
 On^a, on the opposite side of the river, is set down by Mou- 
 chez as 1,400 metres in height. 
 
 The soils of the higher lands I examined on the Muriahe 
 were not good, but those of the lower were. The principal 
 product of the region is sugar, and there are some immense 
 fabricas on the river, as, for instance, those of Taepebas* 
 and the one belonging to Senhor Joao Caldas Vianna, Jr., 
 which were the only ones we visited. 
 
 At the warehouse of Senhor Amaral, at the head of naviga 
 tion, the country consists of gneiss, with rather low rounded 
 hills bordering the river, and higher ones on the southwest. 
 At the warehouse I found the strike N. 65 E., dip 85 W., 
 
 * This factory is worked by steam, and the molasses is separated from the 
 sugar by centrifugal motion. 
 
THE PROVINCE OF RIO DE JANEIRO. 49 
 
 and a short distance up the river at the fazenda de Santa 
 Maria das Taepebas I observed a well laminated gray and 
 fine-grained gneiss, with strike N. 60 -62 E., dip nearly 
 vertical. Garnets are abundant in the rock here. 
 
 In ascending the Parahyba from Campos to Sao Fidelis, 
 the head of steam navigation, we have before us one of the 
 finest pieces of river and mountain scenery on the coast. 
 
 LOOKING UP THE EIO PARAHYBA FROM ABOVE CAMPOS. 
 
 For about one third of the distance above Campos the 
 country bordering the river is flat. It then becomes higher 
 and more uneven, and gneiss is seen in the river-banks. 
 What appears to be the highest point of the Sapateiro 
 range is on the Parahyba River, on the left bank. It is 
 a large dome-shaped hill, with more or less bare, precipi 
 tous sides. Between it and the river are several other quite 
 prominent hills, in the first of which are quarries, from 
 which a large part of the gneiss used for building purposes 
 in Campos is obtained. On the western side of the river 
 opposite the Sapateiro range stretch westward the serras of 
 Sao Fidelis in a series of sharp peaks. The height of some 
 of the peaks must be at least 3,500 feet, probably more. 
 The sides of the mountains are all regularly rounded like 
 those of Rio. Like the Organs, some of the hills are very 
 
50 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 sharp, but they are not so prominent as to give to this land 
 scape so striking a character as the Organs do to the scenery 
 of the serra of that name . The gneiss of Sao Fidelis is similar 
 to that of Rio, and contains a large quantity of garnets, some 
 of the crystals of which are an inch and a half in diameter. 
 At a rocky place by the river-side I observed a strike of 
 N. 64 E., the dip being vertical. At Sao Fidelis navigation 
 ends. Above that point the river is full of rapids, obstruct 
 ed by rocks, and is navigable only for canoes and the like. 
 The lands of the immediate vicinity arc not largely culti 
 vated. 
 
 At Cantagallo, during the reign of the first Vice-King of 
 Brazil, gold was discovered by certain seekers and smug 
 glers of gold, garimpeiros, who, quietly taking possession of 
 the place for many years, extracted gold in secret, and it 
 was a long time before the Brazilian government discovered 
 the region whence so much gold found its way to the capital. 
 Mawe says that the rock of the locality is granite composed 
 of feldspar, hornblende, quartz, mica, sometimes holding gar 
 nets, evidently gneiss, like that of the Serra do Mar to the 
 westward. He states that the gold comes from the lowest 
 bed of cascalho, or gravel, occurring always in rounded 
 grains, and that he never saw a crystallized specimen. Gold 
 and ferric oxide were the only metallic substances found 
 here. At the time of Mawe s visit (1808) so little gold was 
 extracted, that the quinto, or royalty, paid to the govern 
 ment would scarcely suffice to pay the officers and soldiers 
 appointed to collect it. Von Tschudi says * that the gold 
 of Cantagallo came from the bed of a stream. 
 
 Another locality w r here gold was formerly washed is Santa 
 
 * Reisen durch Swd-America, Drittcr Band, 176 t<! Seite. This volume con 
 tains a lengthy description of Cantagallo and Nova Friburgo. 
 
THE PROVINCE OF EIO DE JANEIRO. 51 
 
 Rita, a place about five Brazilian leagues northeast of Can- 
 tagallo. Mawe * describes the gold as occurring in a bed 
 of cascalho,! or gravel, overlaid by earth. The layer of 
 cascalho varies in thickness from two feet to seven or eight 
 inches, and lies under a thickness of four or five feet of 
 earth. 
 
 The Cantagallo region was never very rich in gold. At 
 Santa Rita and in the vicinity Mawe found heavy deposits 
 of limestone. 
 
 From Sao Fidelis I made a horseback journey in company 
 with Mr. Copeland across the country northward by way of 
 the Yallao Grande, to Bom Jesus on the Itabapuana River, the 
 dividing line between the provinces of Rio and Espirito Santo. 
 
 For the greater part of the distance to the Rio Muriahe 
 the road led through the most dense and luxuriant virgin 
 forest. Little was to be seen of the rock, or even of the 
 soil, but I observed that the drift clay, where exposed, 
 often contained boulders of gneiss ; and masses of rock 
 of large size were sometimes seen resting upon it. On 
 
 * " Travels in the Interior of Brazil, particularly in the Gold and Diamond 
 Districts of that Country. By Authority of the Prince Regent of Portugal. By 
 John Mawe." I do not know the date of the English edition. I have a copy of 
 the American reprint, which appeared in 1816, and was published in Philadel 
 phia. It has as a frontispiece a large steel engraving, representing a number 
 of negroes at work under a long thatched shed washing for diamonds, an en 
 graving which has been copied over and over again, and is familiar to every 
 young student of geography. A German edition in my possession, entitled 
 Reisen in das Innere von Brasilien, was published in 1816 at Leipzig, while 
 Burton mentions a French work bearing the title Voyages dans I lnterieur du 
 Bre sil en 1809 et 1810, published in the same year, and which I suppose to be a 
 translation of the above " Travels," though Mawe began his explorations in 
 Brazil in September, 1807. This is a work of much interest, and contains many 
 valuable geological facts. 
 
 t The word cascalho. pronounced cascalyo, means gravel in Portuguese. In 
 Brazil the auriferous cascalho is almost invariably composed of quartz pebbles. 
 
52 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the Vallao Grande, about a league west of the river Muriahd, 
 I saw well-laminated gneiss with a strike of N. 64 E., 
 and a northward dip. The valley of the Muriahe I 
 found well cultivated where I crossed it, and furnishing 
 large crops of sugar-cane. Leaving the river, our course lay 
 over a serra, which our guide called Matuca. It is com 
 posed of gneiss, and must be over 2,000 feet high. On our 
 descent on the northern side, I observed thick beds of a kind 
 of gneiss composed almost entirely of quartz, and in the 
 drift I saw boulders of this rock mixed up with boulders of 
 common gray gneiss.* From this serra to Bom Jesus the 
 country is all gneiss, with low rounded hills, the whole 
 being covered with a most vigorous forest growth. The Rio 
 Itabapuana is a little stream comparable to the Rio Pirahy on 
 the Dom Pedro II. Railroad. Between Bom Jesus and the 
 Ribeirao do Jardim the land is rather low, and diversified 
 by rounded gneiss hills of inconsiderable elevation. The 
 river is bordered by flat, alluvial lands, often marshy, the 
 resort of great numbers of water-birds, piagocas (Parr a 
 Jacana), cranes, &c. 
 
 On the Espirito Santo side, between the Ribeirao do Jar- 
 dim and the Ribeirao Formoso, begins the Serra de Sao 
 Romao e Santa Paz, or the Serra de Itabapuana, which rises 
 abruptly from the river and, more or less broken, runs off 
 in a northeast direction to the Rio Itape merim. The hills 
 on the Rio Itabapuana are more than a thousand feet high, 
 and are composed of gneiss which dips southward at a 
 moderate angle and with its usual strike. They are very 
 precipitous on the southern side, the rocks being covered 
 by an abundant growth of cactuses, <fec. On the same side 
 of the river, in the angle between the Ribeirao Formoso and 
 
 * I was told that there was limestone in this serra, but I saw none. 
 
THE PROVINCE OF RIO DE JANEIRO. 53 
 
 the Itabapuana, is a solitary irregular conical peak called the 
 Pedra Formosa, which, standing alone in front of the amphi 
 theatre formed by the Serra of Sao Romao, and that of Santa 
 Paz, forms a fine piece of mountain scenery. Thence onward 
 to the fazenda of Sao Pedro the country is still gneiss, the 
 hills are low, covered by a most fertile red drift soil, and are 
 heavily forest-clothed. The river is rocky and swift, and just 
 below the fazenda there are some considerable rapids which 
 extend for nearly a mile, and would furnish an abundant 
 water-power. The soils of this vicinity are very good, and 
 excellent cane, coffee, and cotton are raised. The cotton 
 which I examined on the fazenda of Senhor Martinho Fr. 
 Medino was the finest I ever saw on the coast. The 
 gneiss of this vicinity, and of Porto da Limeira, lies remark 
 ably horizontally. At the rapids at the fazenda of Senhor 
 Martinho it has a northward dip of some ten degrees only. 
 At Sao Pedro and Porto da Limeira it is well laminated, but 
 it has a very irregular dip and strike, though the dip is 
 generally northward. The rock is full of granite and quartz 
 veins. About three miles south of Limeira is a remarkable 
 isolated peak called the Garrafao,* or the " demijohn" which 
 forms a very conspicuous landmark, visible from a consider 
 able distance off the coast at sea. It is precipitous on all 
 sides, and as it is long from east to west, and very narrow, 
 it presents very different aspects according to the position 
 from which it is seen. From some points of the compass it 
 appears dome-shaped, from others like an immense tower or 
 pillar rising out of the generally plain country. Mouchez 
 estimates its height at 910 metres ; I should set it down as 
 between 2,500 and 3,000 feet. 
 
 * It has almost precisely the same structure and nearly the same form as the 
 Garrafao among the " Organs." 
 
54 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Between the Rio Itabapuana and the Parahyba, eight or 
 ten miles south of the Garrafao, and in a line with the Serra da 
 Oiic;a, is a remarkable group of gneiss hills, which is visible 
 from Cape Sao Thome, a distance of at least forty miles. 
 One of these is a very sharp, conical needle, called Pedra 
 Lisa.* This needle is seen in the following; sketch. 
 
 Descending the river from Porto da Limeira, one soon 
 leaves the gneiss region, and comes upon a flat country, for 
 the most part very heavily wooded, and more or less diver 
 sified by shallow lagoons, one of which, Lagoa Feia, is quite 
 extensive. The river is very narrow and tortuous, and only 
 navigable for very small steamers. Much of the land is very 
 low, and must be frequently overflowed ; but there are some 
 considerable patches of tertiary, which are however much 
 denuded. At the fazendas of Senhores Pedro Mendes and 
 Antonio Martini these lands rise to a height of perhaps sixty 
 feet, and on the river they are seen to be composed of the 
 characteristic tertiary sandstones and clays. The little 
 village of the Barra do Itabapuana, principally inhabited by 
 
 * I have seen this remarkable peak from all sides. It always appears as a 
 needle or sharp cone. Mouchcz makes it 1,150 metres in height, and I do not 
 think he has over-estimated it. 
 
THE PROVINCE OF RIO DE JANEIRO. 55 
 
 fishermen, is built on a strip of sand on the right bank of 
 the river, near the mouth. It is separated from the shore 
 by a narrow, shallow channel, or lagoon, which runs south 
 ward from the river, parallel with the shore, and just behind 
 the beach ridge. This lagoon communicates with a marshy 
 tract covered by mangroves,* south of which red tertiary 
 sandstones are exposed on the edge of the marsh, and the 
 land rises apparently some twenty feet, forming to the south 
 ward a large patch of tertiary. Opposite the town there is 
 a large sandy island, which is separated from the beach by 
 a lagoon which stretches northward along the shore for 
 some distance. 
 
 A league or more to the south of the Barra do Itabapuana f 
 are two or three rocky points of tertiary sandstone, present 
 ing low red cliffs. The same rock is said to occur at Man- 
 guinhos. Isolated masses of this rock, covered at high tide, 
 occur off the Barra of the Itabapuana, and also at Man- 
 guinhos. 
 
 The mouth of the Itabapuana is, like the Parahyba, ob 
 structed by a sand-bar, and is entered with difficulty. The 
 water is shallow off the coast, and vessels sometimes anchor 
 outside the bar and take in cargoes of wood, <fcc. 
 
 * The mud of the mangrove swamps is very soft, being composed of the 
 finest kind of silt, and it is black and stinking with decaying matter. It is full 
 of shells, leaves, the exuviae of crabs, &c. 
 
 t Ita, in Tupy, means stone, and poam or puam an island ; and I suspect that 
 the name Itabapuana may have been given from the little rocky islands men 
 tioned below. Cazal gives the name of the river as Camapuan or Cabapuanna, 
 and says that the savages called it Reritigba. (Coroyrafia, Vol. I. p. 61.) 
 
56 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER II. 
 
 PROVINCE OF ESPIRITO SANTO. 
 
 Barreiras do Siri. Itapemerim. Coast between Itapc merim and Bcnevente. 
 
 Bencveiite. Guarapary ; Consolidated Beach, Corals, &c. Rio Jecii. 
 Bay of Espirito Santo. Nossa Senhora da Penha. Victoria. Decom 
 position of Gneiss and Formation of Boulders of Decomposition. Recent 
 Rise of the Coast. Corals, &c. of the Bay of Victoria. Rio Santa Ma 
 ria. German Colonies. Fisheries. Sand Plains. Tertiary Plain at 
 Carapina. Mestre Alvaro. Serra. Nova Almeida. Rio Reis Magos. 
 Santa Cruz. Basin of the Rio Doce. Description of the River. Guan- 
 dii : its Colony and Agricultural Resources. Porto de Souza. Geology 
 of Vicinity. Luxuriance of Vegetation on the Doce. Woods. Game. 
 
 Francylvania. Climate of the Doce. Linhares. Lagua Juparanaa. 
 
 The Future of the Doce. American Colonists. Salt Trade. Barra 
 Secca. Sea-Turtles. Consolidated Beaches and the Mode of their Forma 
 tion. Character of Coast between the Rivers Doce and Sao Matheos. Rio 
 Silo Matheos Described. Geological Features. Fertility of its Lands. 
 Cocoa-Palms and their Distribution. City of Sao Matheos. Rio Itahu- 
 nas. Cliffs of Os Le^oes. Coast between Itahiinas and Rio Mucury. 
 
 A SHORT distance northward of the Rio Itabapuana, and 
 not far from the sea-shore, is Lake Maroba, from which flows 
 the river of the same name. Between the Itabapuana and 
 this river the coast lands are low and marshy. Just south of 
 the Barra do Maroba the lands rise somewhat along the shore, 
 and tertiary red sandstone shows itself in the beach. The 
 tertiary bluffs of the Itabapuana sweep round back of the 
 lake, and come down to the shore north of the Barra, and 
 are continued thence northward in a fine range of bluffs 
 called the Barreiras do Siri, which, from the bright red 
 colors of the clays and sandstones composing them, present 
 a very picturesque appearance from the sea. These- cliffs arc 
 
PROVINCE OF ESPIRITO SANTO. 57 
 
 seventy to eighty feet in height, and the country lying back 
 of them is a wooded plain. 
 
 The lowest bed seen in the cliffs of Siri is a coarse, dark- 
 red sandstone, with indistinct stratification, and, where ex 
 posed on the beach, full of holes, presenting an appearance 
 more like that of the surface of a lava stream than anything 
 else. This mass of sandstone is penetrated by deep, perpen 
 dicular, pipe-like holes, which, in many cases, communicate 
 with one another. This sandstone rises to a height of about 
 twelve feet above low-water mark, and is overlaid by a bed, 
 about twenty feet thick, of a sandy clay, whitish and reddish,* 
 which penetrates into the cavities of the sandstone. The 
 sandstone seems to result in part from the irregular cement 
 ing of the sandy clays by oxide of iron. The clays are soft 
 and show no distinct stratification. The proportion of sand 
 varies very much, some of the clays being exceedingly fine 
 in texture, like kaolin. They are not at all plastic. The 
 color varies from a pure white to a bright red, and some 
 times the clay is variegated with curved lines of red or 
 yellow, so as to look like fancy Castile soap. Over the clay 
 is an irregular deposit of very dark-red sandstone, which is 
 well stratified, and sometimes forms lenticular masses ; and 
 over this in turn lies a bed of red clay, which I could not 
 well examine. Between the clay and the soil, which is usu 
 ally light brown, there is a layer of pebbles and iron-stone 
 nodules. A few miles below Itapemerim the tertiary lands 
 
 * Prinz Max. zu Neu "VVied, Reise nach Brasilien, Vol. I. p. 169, speaks of 
 these cliffs, and gives the following note, which 1 leave in his own words : 
 
 " Der Untersuchung des Herrn Professor Hausmann zu Gottingen zufolgc 
 gehurt dieses Fossil, welches einen Hauptbestandtheil eines grossen Theils 
 dieser Kiiste von Brasilien ausmacht, zum verharteten Steinmark, wohin man 
 auch die sachsische Wunder-Erde zahlt. Es stimmt in alien Kennzeichen mit 
 dem Steinmarke iiberein." 
 3* 
 
58 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 recede from the beach, and are broadly denuded on both 
 sides of the river. The Itapemcrim is a much larger stream 
 than the Itabapuana. It rises near the frontier of Minas 
 Geraes, west of Victoria, behind the Serra do Pombal, and 
 has a course of about eighty miles. It is shallow in its lower 
 course, and of little importance. There is an extensive 
 alluvial plain bordering the river on the south side for a few 
 miles above the town, in part belonging to the fazenda of 
 the Barao do Itapemerim ; this tract of land is very fer 
 tile, and a considerable part is cultivated for sugar. There 
 are a few fazendas farther up the river. That of Muqui, 
 belonging to the Baron, is built on a gneiss hill. 
 
 The serras approach nearer the coast in going northward, 
 and in the neighborhood of Itapemerim are very high, pre 
 senting the same topographical peculiarities as in the south. 
 
 SERRA OF ITAPEMERIM, SEEN FROM THE SEA. 
 
 About twenty miles west of the town of Itapemerim, and 
 south of the river, is a very remarkable irregular peak called 
 the Frade, while a few miles to the northeast is a group of 
 needles, which presents an exceedingly strange appearance 
 when seen from certain points of the compass, appearing 
 sometimes like the fingers of a giant hand held up against 
 the sky. Some of these needles arc at least 3,000 feet high. 
 The whole country lying behind them, even beyond the lim 
 its of the province, is very mountainous, and is composed 
 of gneiss. I was informed that specular iron occurs in the 
 serras of Itapemerim, but I had no way of verifying the 
 report. 
 
PROVINCE OF ESPIRITO SANTO. 59 
 
 In 1723 the first settlers on the Itapemerim discovered 
 gold in one of its affluents on the north, called the Rio do 
 Castello.* " A decree of the 17th September of the following 
 year, 1724, determined that the auriferous lands should 
 be divided into small portions among all those who wished 
 to employ themselves in the labor of mining, with the con 
 dition that they should subject themselves to the laws of 
 the sesmarias and to the impost on the gold." f A gold- 
 washing town was established at the confluence of the Cas 
 tello with the Itapemerim, but the hopes of a rich yield 
 proved deceptive,:}: and with the incursion of the Indians 
 the place went down. 
 
 North of the Itapemerim the coast is bordered by exten 
 sive plains of coarse white sand, like those of the Island of 
 Paqueta and of the extension of the Cantogallo Railroad. 
 These plains are quite destitute of grass, and are covered 
 sparsely by clumps of stunted trees, among which is the 
 Pitangueira (Eugenia pedunculata) , with an abundance of 
 cactuses ( Cereus~) , and bromeliaceous plants. About half 
 way to the little town of Piuma, distant some eight miles 
 north of Itapemerim, gneiss hills stretch along the coast. 
 Among these is the Morro de Agah, one of the noted land 
 marks of the coast, a sharp, saddle-shaped ridge, appear 
 ing pyramidal when seen from the north. Gneiss shows itself 
 occasionally on the beaches, and there are a few little gneiss 
 islands off the coast. The Piuma is a very small river of 
 little or no importance. 
 
 * Dice. Geog., Art. Castello. 
 
 t The Cororjrafia Brasilica, published in 1817, speaks of these mines as 
 having been abandoned on account of the incursions of the Indians. 
 
 | Von Tschudi, Reisen durch Slid America, Drittcr Band, 60 S Seite. 
 
 I observed on the shore, just south of the point of the Agah, a rock 
 exposed, which appeared to be a sandstone. It was much decomposed, and 
 I was unable to examine it very carefully. 
 
60 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 From Piuma to Benevente, which is a ride of only about 
 two hours, and cannot be more than from four to six miles, 
 the road leads over low gneiss hills, because the shore is 
 rocky and the beach has to be abandoned. The Rio Iri- 
 ritiba, or Benevente, is a small stream, navigable for 
 some eight leagues above its mouth, which is narrow, sit 
 uated between gneiss hills, and unobstructed by a bar, a 
 circumstance owing to the protection of this part of the 
 coast by a rocky point extending southward some two 
 miles east of the mouth. The water at the entrance is 
 deep, and vessels of considerable tonnage may enter at 
 any time of the tide. Just inside, the river spreads out 
 into a broad sheet of water, so that it forms one of the 
 best and most frequented harbors on the coast of Es- 
 pirito Santo. Vessels frequently find refuge in the little 
 bay behind the point east of Benevente, where they are 
 well sheltered from northeast storms. The town is a 
 small one, and situated at the mouth of the river, on 
 the northern side, at the base of a low gneiss hill.* 
 Besides ship-building, its trade consists in wood and a lit 
 tle coffee. According to Yon Tschudi, the lands lying back 
 of Benevente are very fertile, and the place possesses 
 natural advantages which might, if properly used, make 
 it of much importance. Leaving Benevente, and going 
 northward, the path crosses obliquely a projecting point, and 
 passes over and among low hills of red sandstone and clays, 
 the sandstone cropping out frequently at the base of the 
 hills. The soil covering these hills is gray, and appears to 
 be very rich. In some parts it is of a deep pinkish-red color. 
 On leaving the hills a plain of white sand is reached, like 
 
 * Gneiss, gray, well laminated. Strike N. 55 E. Dip varying from ver 
 tical to 88 N. Rock intersected by numerous granite veins. 
 
PROVINCE OF ESPIRITO SANTO. 61 
 
 that of Itapemerim. This occupies a sort of bay in the 
 tertiary lands, which soon reach the coast again and form 
 a long line of red cliffs, extending for several miles along the 
 shore, almost to the little fishing village of Miahype. This 
 village is situated on a narrow sand-bank uniting a small 
 mass of gneiss to the shore, off which a similar mass forms 
 an island. Thence northward the tertiary lands extend 
 along the shore, with narrow patches of sands in front, and 
 with occasional interruptions, to the village of Guara- 
 pary,* when the shore becomes very much indented, and 
 the distribution of the various formations are too compli 
 cated for description here. Many of the low hills of 
 the vicinity have a basis of gneiss, but are capped with 
 tertiary clays. The gneiss shows itself in a number of 
 rocky points and ledges, and at the mouth of the river 
 there are several islands. The gneiss is gray, but it is 
 very micaceous, the mica being in moderately large black 
 crystals. I find in my note-book an observation of strike 
 N. 80 E., dip 80 S. 
 
 2^_- - .-:: . _- 
 
 COAST SOUTH OF GUARAPARY. 
 
 In a little cove just south of the town is a large detached 
 mass of sandstone, lying just in front of the beach, and at 
 
 * The country lying between the coast and the Serra do Pcro Cao, distant 
 some six miles from Guarapary, is composed of low hills, some of which are 
 undoubtedly tertiary, interspersed with higher ones, as the Orobd, which are 
 gneiss. 
 
62 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 low tide washed by the waves. It consists of quartz sand 
 cemented by carbonate of lime, and the rock is exceedingly 
 hard. It is well stratified, and the layers are nearly hori 
 zontal. The rock is cut up by joints, which the sea has 
 widened, so that it is much broken up, while the washing 
 out of the calcareous cement by the sea has left the surface 
 very ragged. This mass is only part of a solidified beach 
 which has been laid bare by an encroachment of the sea. 
 At low water it furnishes to the naturalist a very rich har 
 vest of marine animals, for in the holes near its edge grow 
 masses of Siderastrcca stellata Verr., while Acanthastrceas 
 of considerable size grow attached to its sides. Favias and 
 Ag-aricias, <fec., occur in great abundance, and the rocks are 
 covered with great patches of the common brown Palythoa 
 of the coast, and of a spreading, -green-disked Zoanthus, not 
 determined. Several beautiful species of sea-anemones, 
 deeply tinted, are very common, but the species have not yet 
 been made out. At low tide there may be obtained from 
 the rocks beautiful specimens of Hymenog-orgia qucrcifolia, 
 Eunida humilis, and Plexaurella dichotoma. Sea-urchins 
 (Echinometra Michelini) are exceedingly abundant here, and 
 are used for food by the natives. They live packed securely 
 away in deep holes, not only in the sandstone of the reef, but 
 also in the gneiss of the adjacent points. Here also the 
 beautiful little crimson star-fish, Echinaster crassispina, is 
 very abundant, occurring among the sea-weeds and in little 
 pools betwixt tide-marks ; and, in the tide-pools, Ophiura 
 cincrea, together with a number of other species of the same 
 order, occurs. Holothurians, some of them a foot long, are 
 very abundant, packed into crevices, in tide-pools, and 
 under rocks ; and a pretty comatula, Antcdon Dubenil 
 Bolsche, or Braziliensis Liitk., may be seen through the 
 
PROVINCE OF ESPIRITO SANTO. 63 
 
 water covering the rocks like rosettes of brown feath 
 ers."* 
 
 This locality is rich in crustaceans, but not especially so 
 in mollusks ; indeed, the whole coast has a rather poor 
 molluscan fauna. A large octopod is common here, and is 
 hooked out from the crevices in the rocks by the inhabitants, 
 who use it for food. Guarapary is an excellent collecting- 
 ground for marine invertebrates, though not so good as Vic 
 toria. Inside the mouth of the bay, in water which is brack 
 ish and impure, occurs a slender-branching, tender, nodose 
 Halcyonoid, undetermined. The entrance to the harbor of 
 Guarapary is good and secure, and shelter for shipping is 
 furnished by the little islands, Escalvada and Raza, lying off 
 the coast ; but the marshy lands in the vicinity make the 
 place exceedingly unhealthy, and notwithstanding the lands 
 to the west of the town are good, and woods and valuable 
 balsams and fish are in abundance in the vicinity, the un- 
 healthiness of the climate has placed a ban upon its growth. 
 The river is a little one, with a course of only a few leagues, 
 and takes its rise in the Serra do Pero Cao. It is said to 
 offer navigation as far as the coast serra, and to unite in its 
 course a number of little lakes. 
 
 In going northward from Guarapary the path leads, first, 
 over a gneiss district bordering the northern side of the 
 river, and then descends to a plain of white sand, sparsely 
 covered by trees. Crossing this, low tertiary hills are 
 reached, with more or less gneiss, bare in places, espe 
 cially along the shore, and you come upon the little brook 
 Pero Cao, beyond which the shores are sandy and flat as 
 
 * Of these Radiates, as well as the other marine invertebrates of the reef, Mr. 
 Copeland and I made a considerable collection, which is in the Museum of 
 Comparative Zoology, in Cambridge, but has not yet been worked up. 
 
64 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 far as a small river called by my guide Una, but which does 
 not seem to be represented on the chart of Mouchez. This 
 little stream escapes into the sea just south of a projecting 
 point of gneiss. Between this point and the serra the coun 
 try is low. The distance of the mountains from the sea is 
 only about five miles. A point or two passed, and one 
 reaches a long sand-beach, which extends some eight miles 
 north to a low gneiss point called Ponta da Fructa. Along 
 this beach are, in some places, sand-dunes twenty to twenty- 
 five feet in height. 
 
 From the Ponta da Fructa, northward, to the Ponta dc Jecii 
 a distance of seven or eight miles stretches an almost 
 straight sand-beach, behind which are plains, sandy and 
 marshy, a perfect batrachian paradise. 
 
 The Ponta do Jecu is a gneiss hill, somewhat similar to 
 that of the Fructa, and formerly an island, but now joined 
 to the mainland by a sand-beach. There are other smaller 
 hills in the vicinity. According to a sketch in my own 
 note-book, the river enters the sea to the south of the point, 
 but Mouchez s chart shows it entering on the north side, 
 which was probably the case at the time his chart was 
 made, the mouth having been closed on the south by a 
 storm. The river Jecu rises among the serras to the west, 
 and is an insignificant stream, apparently smaller than the 
 Muriahe. It is with difficulty navigable for canoes for only 
 a short distance. Some five miles above its mouth a canal, 
 cut long ago by the Jesuits, runs off northward and commu 
 nicates with the port of Victoria, distant about five miles. 
 This was done to facilitate the transport of the products of 
 the country to Victoria, as well as to avoid the dangerous 
 passage by sea from the mouth of the river around the reefs 
 and rocky point on the south of the bay of Espirito Santo. 
 
PROVINCE OF ESPIRITO SANTO. 65 
 
 The cutting of this canal is said to have improved the health 
 of the region of the Jecti. 
 
 On this river, some thirty miles from its mouth, and 
 somewhat farther from the city of Victoria, was estab 
 lished, in the year 1847, a German colony, Santa Isabel, 
 among the gneiss hills lying east of the serra, in a region 
 healthy and fertile, and proper for the culture of coffee, 
 cotton, &c. ; but the colony has not been prosperous, 
 owing to bad management and the want of roads.* 
 
 From the mouth of the Jecu to the bay of Espirito Santo 
 the sand-beaches continue, backed by the sandy and marshy 
 plains of the Campos de Piratinanga.f 
 
 The Bay of Espirito Santo is about two and a half miles 
 wide, and irregular in shape. On the north is the Ponta 
 do Tubarao, with a rocky sandstone shore, whence sweeps 
 around westward and southward a long sand-beach, join 
 ing a rocky point, near which enters the channel of the 
 Rio da Serra. Thence southward, for a mile or more, 
 the land is high and irregular, and the shore consists of 
 sea-beaches between projecting gneiss points. We then 
 reach the entrance of a narrow, irregular channel, the 
 estuary or bay of the Rio Santa Maria, that extends 
 westward among gneiss hills. On the south side of the bay 
 is an irregular conical gneiss hill, some 700 feet high, 
 called Monte Moreno, forming a rocky point. West of this, 
 and separated only by a short sand-beach and a small stream 
 coming from the swamps to the southward, is another coni 
 cal hill, some four hundred feet high, crowned by the pictu- 
 
 * Tschiuli (Reisen, etc., Drittcr Band, 8* Seitc) gives as the mean annual 
 temperature for the locality +18 Reaumur = 70.50 Fahr. 
 
 t St. Hilaire, in speaking of the sandy plains between Jecu and Victoria, 
 says that the vegetation covering them resembles in many points that of the 
 elevated plateaux of Minas Novas. (Tome II. 2 de Partie, p. 229.) 
 
 E 
 
66 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 resque pile of the convent of Nossa Senhora da Penlia. West 
 of this, between the Morro de Nossa Senhora da Penha and 
 high gneiss hills, is the deep cove of Villa Yelha, with the 
 ancient village of the same name built on the edge of the 
 sand-plain. Westward of this cove the shore stretches to 
 the Pao de Assucar, along the southern side of the channel 
 of the Santa Maria. It consists of a number of high rocky 
 points, united by mud-flats and sand-beaches. The oppo 
 site shore of the channel is of the same general character. 
 North of Monte Moreno and of the mouth of the channel 
 are two more high gneiss islands, lying one north of the 
 other. There are, besides, many smaller ones, together 
 with a number of rocks and skerries, and the channel is 
 obstructed near the Pao de Assucar by islands and rocks. 
 The Pao de Assucar is a precipitous, irregularly conical 
 gneiss hill, 400 to 500 feet high, falling off to the north, 
 presenting to the channel a smooth, almost vertical, face. 
 Here, by the projection of a point from the northern side, 
 the channel is suddenly narrowed down to a w T idth of only 
 about 600 feet. Passing the Pao, the channel widens out 
 into a most spacious harbor, and on the northern side, in a 
 fine amphitheatre among the hills, is built the city of Vic 
 toria. This basin extends only about a couple of miles west 
 of the Pao. At its head it receives the waters of the canal 
 from the Jecu, and of the Rios Crubixaand Santa Maria. A 
 channel extends northward, and, uniting the mouths of sev 
 eral rivers, passes round the hills of Victoria and enters the 
 Bay of Espirito Santo, thus rendering them an island. This 
 island is composed of gneiss, is very high, rugged, and 
 clothed with forest. To the north and west the coun 
 try is a plain, while the shore alone, along the south side 
 of the bay and channel, is hilly. The hills of the island 
 
PROVINCE OF ESPIRITO SANTO. 
 
 67 
 
 and adjoining mainland then form an isolated group, of 
 which the main mass just behind the city must be fully 1,000 
 feet in height. The channel of the harbor is a narrow valley, 
 which, owing to its rocky sides, has been easily kept open. 
 The gneiss of the locality is very homogeneous, porphyritic, 
 and of the same general character as that of the coast 
 southward. As a general thing the hills are dome-shaped 
 and regularly rounded, as is represented in the following 
 sketch, but sometimes they are conical. In some cases one 
 or more sides, or the top, is bare and smooth, as is the case 
 with the Pao de Assucar. These bare surfaces are almost 
 
 MONTE JUTUQUARA AND GNEISS HILLS NEAR VICTORIA. 
 
 always rounded with remarkable regularity, and are never 
 jagged and angular, like our northern precipices, or the 
 cliffs on the Sao Francisco, below the falls of Paulo Affonso. 
 This is owing to the uniform decomposition of a surface 
 unbroken by joints or planes of stratification ; for many of 
 these mountains are actually formed of a single, unbroken 
 mass of gneiss. The cliffs are rarely vertical, and not un- 
 frequcntly form bare places on a mountain-side, set in a 
 framework of verdure. Such a bare slope is represented in 
 
68 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the following sketch of the Fortalcza de Peritininga, below 
 Villa Velha, Bay of Espirito Santo. They are usually 
 
 FORTALEZA DE PERITININGA. 
 
 stained by perpendicular lines, or bands, of a rich, deep, 
 inky, purplish-black color, being some minute lichen 
 growth, and covered with scattered tufts of beautiful bro- 
 meliaceous plants, orchids, cactuses, <fec., which give them 
 a very picturesque appearance. Just below the city of 
 Victoria proper there is one of these bare hillsides, which 
 forms an exceedingly attractive element in the romantic 
 scenery of the island. 
 
 Standing in an amphitheatre among the beautiful hills, 
 and in full view from the sea, forming a most valuable land 
 mark for the sailor, is a conical mountain, bearing on top a 
 tower-like mass, the face of which is excavated on the east 
 ern side by a considerable cavern, in which, it is said, in old 
 times fugitive slaves took refuge. This mountain is called 
 Jutuquara,* or Frade de Sao Leopardo, according to Mouchez. 
 Its height must be 700 feet at least, probably more. It is 
 represented in the sketch on the preceding page. The gneiss 
 
 * Prince Ncu Wied gives a wretched drawing of this mountain in his work. 
 
PROVINCE OF ESPIRITO SANTO. 69 
 
 BOULDER OF DECOMPOSITION, VICTORIA. 
 
 hills, down to a certain level, are covered, as in the south, by 
 drift clay, in which are imbedded rounded and angular frag 
 ments of quartz and gneiss. This forms a rather coarse 
 and arenaceous soil, which is not so fertile as the drift soils 
 of Rio. Decomposition obtains here as elsewhere. The 
 rocky shores and islands of the bay are lined with rounded 
 masses of gneiss, often of immense size, and which, lying 
 loosely about, have all the appearance of erratic boulders. 
 Similar masses we have already found on the shores of Pa- 
 queta and many of the other islands in Rio harbor. Some 
 times these boulders of decomposition are seen perched inse 
 curely on the tops of other rocks, as is the case near the Pao 
 de Assucar. I have seen no locality where the formation of 
 these boulders is better exemplified than here. On the hill 
 sides the surface of projecting rock-masses undergoes a sort 
 of softening, which causes it to separate from the undecom- 
 posed rock beneath, and break up into irregular fragments by 
 the formation of a system of cracks. Through these fissures 
 the water finds access. Each one of the fragments decom 
 poses all round, and the loose decomposed material being 
 washed away, these masses become rounded, separate more 
 and more from one another, and sometimes fall over, so 
 
70 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 as to lie on the surface of the soil. One must therefore 
 be exceedingly careful not to make blunders in examining 
 them. The loose rocks lying on. the side of the Morro de 
 Nossa Senhora da Penha, and carefully represented in the 
 accompanying view, are boulders of decomposition. Where 
 the rock undergoing decomposition is on the sea-shore, 
 and the action of the waves assists in the removal of the 
 decomposed material as soon as formed, the effects produced 
 may be still more striking. The easternmost extremity of 
 the Ilha do Boi, just opposite the Pcnha, is a projecting, 
 sloping mass of compact gneiss, as represented in the follow 
 ing rough sketch. The whole surface of the gneiss is soft- 
 
 DECOHFOStNQ SURFACE, ILIIA DO BOI. 
 
 ened to a depth of several feet, and has shrunk entirely 
 away from the undecomposed rock. This sheet has cracked 
 through perpendicularly to the surface, and covers the rock 
 below like a pavement. The action of the waves has, as 
 represented, removed these loose fragments from over a 
 considerable area, which is left very regularly rounded and 
 uncracked. In other localities this decomposition and denu 
 dation have gone on until only a few of the heavier blocks arc 
 
PROVINCE OF ESPIRITO SANTO. 
 
 71 
 
 left on the surface, presenting the appearance of erratics. 
 On the rocks known as the Pacotes, lying a little to the 
 southward of the bay, off the coast, several large boulders 
 of this kind are seen lying, presenting the appearance of 
 buildings. It might be objected to our theory of the glacial 
 origin of the Brazilian surface-clays and pebbles that they 
 were formed by a decomposition of this kind along the 
 shore of a slowly sinking continent. This would never pro 
 duce such a coating of clay as forms the drift of Brazil, 
 and there would certainly be associated with the deposits 
 stratified sands and gravels and silts, which are nowhere 
 to be found. Farther on we shall discuss in detail the whole 
 subject of Brazilian drift. 
 
 Along the northwestern face of the Pao de Assucar runs 
 an irregular horizontal line, as represented in the woodcut be 
 low. This line consists of a series of very shallow hollows, 
 
 AUCIENT SEA-LEVEL ON PAO D ASSUCAR, VICTORIA. 
 
 sometimes continuous, and evidently worn by wave-action 
 within comparatively recent times. This old wave-line is not 
 traceable along the whole extent of the cliff. I first ob- 
 
72 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 served it in the latter part of August, 1865. In September, 
 1867, I revisited the locality shortly after a tide of the full 
 moon, which had left a well-marked muddy line running 
 around the base of the Pao de Assucar, and which the suc 
 ceeding tides had not reached. Measuring as nearly as I 
 could from the middle of the wave-line, the mean of two 
 measurements gave me as the height of the old water- 
 level above the high-tide level of the 13th September 3.16 
 metres, or a little more than seven feet. From the old line 
 to the upper edge of the zone of oysters is 3.56 metres. 
 On the face of the cliff of the Pao, in a little cove on the 
 western side, I cut a groove with my chisel indicating the 
 height reached by the tide of the 13th September, 1867. 
 This same old water-line may be seen in several places on 
 the rocks on the opposite side of the channel below the Pao 
 de Assucar, as well as on the face of a cliff at the western 
 end of the beach at Villa Yelha, where, as nearly as I could 
 judge, it had the same height above the sea. This wave- 
 line marks a period of rest when the continent, stand 
 ing for some time at the same level, gave an opportunity 
 for the little waves of the sheltered port to excavate the 
 line. No such line marks the present sea-level, and I 
 infer, from that and other facts, that the land is at present 
 rising. 
 
 The water in the middle of the bay is very shallow, and the 
 bottom appears to be a bank that comes so near to the surface 
 that the sea sometimes breaks over it. Between the Moreno 
 and Point Tubarao the average depth is about fifteen metres. 
 The depth decreases on entering the channel between the Ilha 
 do Boi and Monte Moreno, where it is from four to nine metres. 
 As the channel contracts the depth increases, and just below 
 the Pao de Assucar it reaches twenty-four metres, while in the 
 
PROVINCE OF ESPIRITO SANTO. 73 
 
 narrowest point, at the Pao de Assucar, it is sixteen metres. 
 In front of the town the depth varies from six to ten metres, 
 and an excellent and spacious anchorage-ground is offered. 
 The water opposite the town is turbid, and the littoral fauna 
 is characterized by an abundance of oysters, covering the 
 rocks and piers and mangroves, to within a few inches of 
 high-water mark. This is a very small species, with exceed 
 ingly sharp, wavy edges, and used for food in Victoria. The 
 coves between the rocky points along the channel above the 
 bay of Villa Velha are muddy, and often lined by mangroves. 
 Going down the channel the oysters grow less numerous, 
 and give way to barnacles and mussels. On the north side 
 of the cove at Villa Velha the arrangement of the principal 
 elements of the littoral fauna is as is represented in the 
 following diagram. 
 
 a. Zone of small barnacles, breadth three to four inches. 
 
 b. Little black mussels, all reaching the same upper level, but forming irregular 
 patches, extending in some places as far as a metre below high-tide mark. 
 
 c. Large scattered barnacles, and green sea-weeds ( Ulvce). 
 
 d. Coarse brown sea-weeds, corallines, &c., and oysters. 
 
 The sea-urchins begin above d, and extend below low water mark, but their 
 untenanted nests occur up to. and beyond high-water mark. 
 
 On the beach at Villa Velha, particularly near the west 
 ern end, dead corals are thrown up in great numbers. In 
 deed, they are more abundant there than on any other South 
 American beach I have seen. The commonest species is the 
 Mussa Harttii Verrill, which is thrown up by the waves in 
 
74 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 great quantities, and evidently grows in the immediate 
 vicinity ; but owing to the muddiness of the waters of the 
 cove I could not, in the absence of a dredge, find it alive. 
 It is usually drifted ashore attached to sea-weeds. So abun 
 dant is it in the muddy beach at low tide that it is collected 
 by the inhabitants of the Villa for lime-making, and is called 
 cachimbo, or " pipe-stem." This species is almost invariably 
 incrusted with beautiful Bryozoa. On the same beach occurs 
 a species of Siderastrcsa, and a Pectinia, or Symphyllia, none 
 of which I have observed living in the bay. I have seen in the 
 sands underlying the muddy shores farther up the bay shells 
 and corals which cannot now live in the turbid and brack 
 ish water. Before the rise of the land the bay extended 
 farther on both sides. The river has since narrowed its 
 channel, and the turbid waters have driven down the coral 
 fauna nearer the mouth of the bay. I doubt whether the 
 Musses are now to be found living at Villa Velha. I believe 
 that the specimens on the shore have been long dead, and 
 are thrown up by storms. Leaving Villa Velha, and going 
 down the bay, the oysters disappear as the shore becomes 
 more exposed. Outside the harbor, and at the base of Monte 
 Moreno, is a little island called Ilha Baleeiro. The tide- 
 pools of this island and of the adjoining shore are rich 
 in corals, and at low tide one may collect Hymenog-org-ia, 
 Eunicicc, Plexaurcllce, and all the species common at Guara- 
 pary. I am not aware of the existence of any coral-banks 
 in the bay and vicinity. The Bay of Victoria would be a 
 rich ground for dredging, and so would be the banks lying 
 off the harbor. In fishing I have brought up on the hook 
 masses of nullipores, &c., loaded with life, and at the mouth 
 of the bay I captured in this way a large many-rayed star 
 fish. Victoria appears to have a reputation as a locality 
 
PROVINCE OF ESPIRITO SANTO. 75 
 
 for shells ; but the littoral mollusk fauna of Brazil is very 
 poor, and the shells of the beaches are badly broken and 
 worn.* 
 
 The Rio Crubixa enters the harbor of Victoria just above 
 the city. According to the Diccionario Geog-rafico, " it de 
 scends from the Cordilheira dos Aimores among rocks, 
 amongst which is found a certain species of coral of a 
 dark color and fragile, with which the Botocudo women 
 are accustomed to decorate their heads, necks, arms, and 
 legs." 
 
 The Rio Santa Maria is a much larger stream, which 
 rises some fifty miles, more or less, northwest of Victoria, 
 in the Serra dos Aimore s. 
 
 On the Rio Santa Maria is located the colony of Santa 
 Leopoldina, and as the history of the colony and of the 
 country where it is situated has an especial interest at this 
 
 * I find in one of the letters of the celebrated Joseph de Anchieta a state 
 ment that the manati occurred in the bay of Espirito Santo, and his description 
 seems to me of sufficient interest to warrant my giving it in his own words : 
 
 " DE BOVE MARINO, .Hac quoad rationem temporis, jam ad alia tran- 
 seamus. Piscis quidam est qucm Bovem marinum dicimus, Indi Igiiaraqua. 
 nommant frequens in oppido Spirito Sancto et aliis versus Boream habi- 
 tationibus, ubi aut nulla est, aut exigua admodum, et minor quam apud 
 nos frigores injuria: hie ingentis est magnitudinis herbis pascitur, quod 
 ipse gramina depasta scopulis, quos a2stuaria alluunt, inha;rentia in 
 dicant. Bovem mola corpore superat, cute obtegitur dura, elephanti 
 colorem refercnti ; duo vellut brachia quibus natat, habet ad pectus sub 
 quibus et ubera, ad quse proprios foetus nutrit, os bovi per omnia similis. 
 Esui est congruentissimus, ita ut ducernere nequcas, utrumque carnis, an 
 potius piscis loco habcri dcbeant ; ex cujus pinguedine, qua? cuti ex maxime 
 circa caudam inhseret, ad modo igni fit liquamen, quod jure butyro comparari 
 et baud scio an possit antecellerc ; cujus ad omnia cibaria condienda olci vice 
 usus est : Ossibus solidis, et durissimis quse, possint eboris vices gerere, totum 
 corpus est compactum." Colleccdo de noticias para a historia e geoyraphia das 
 nacoes ultramarinas. Lisboa, 1812. 
 
76 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 time, when efforts are being made to colonize the Brazil 
 ian coast, I translate the following from an account of a 
 visit to the colony by Von Tschudi.* He says: "We 
 crossed swiftly the Lameirao of the bay and steered into the 
 river Santa Maria. Its current is quite slow, and offers, con 
 sequently, no particular hindrance to the ascent. Not far from 
 its mouth it receives from its left bank the little river Cara- 
 pina, and somewhat farther west lies, on the right bank, the 
 Porto da Pedra, consisting of a couple of houses and a large 
 venda. Up to this point there is sufficient water, even for 
 
 steamers of moderate draft Following the many 
 
 windings of the river in ascending it to the north-northwest, 
 we reached, after an eight hours journey, the junction of 
 the river Mangarahyba with the Eio Santa Maria. The lo 
 cality becomes, the farther up the stream one goes, more 
 and more hilly. The banks of the river themselves, where 
 the character of the lands allows, are in part inhabited by 
 Brazilians, who have here laid out little fazendas and occupy 
 themselves principally with the raising of horses and cattle. 
 The affluents of the Santa Maria are very unimportant. 
 From the south there empty into it the Rio Curipe, the Rio 
 Tauha, the Rio Una, and several other brooks whose names 
 I have forgotten ; on the north the Rio Jacuhy, Rio Trame- 
 rim, (Jatamcrim ?) and a couple of entirely unimportant 
 little rivers near the settlements Murinho, Aruaba, and Pen- 
 diuca. A rather extensive property is that of Senhor Jose 
 de Qucimado on the left bank, several miles up the stream 
 from Porto da Pedra ; over against it several little islands 
 stand out upon the surface of the stream. Likewise on the 
 northern bank lies the hamlet of Santa Maria, distant some 
 seven to eight leagues above the mouth of the stream, from 
 
 * Von Tschudi, Reisen durch Brasilicn, Dritter Band, Cap. I. 
 
PROVINCE OF ESPIRITO SANTO. 77 
 
 which it takes its name. Where the Rio Mangarahy unites, 
 or makes barra with the Santa Maria, lies the extensive 
 fazenda of Jose* Claudio de Freitas. From this estate up the 
 river the Rio Mangarahy is navigable for boats only a short 
 distance. The navigability of the Rio Santa Maria ends at 
 about one league from the mouth of the Rio Mangarahy at 
 the Cachoeira de Jose , above which rocks make the channel 
 of the river impassable 
 
 " The Rio Santa Maria rises westward from the colony of 
 Santa Leopoldina in a mountain range, on whose western 
 slope are the sources of several of the tributaries of the Rio 
 Doce. It first becomes important after receiving the river 
 Mangarahy. This last is formed of a number of mountain 
 brooks, of which the greater part rise in the southern part 
 of the colony. It receives two larger tributaries on the 
 south, one of which, the Rio do Medio, bounds the colony on 
 the south, and the other, considerably smaller, on the east ; 
 the Brazo do Sul takes its rise beyond the colony. Its north 
 ern tributaries are very numerous, but are only worthy of the 
 name of brooks. The most considerable are the Riberao da 
 Sumaca, Corrego Isabel," &c., &c. 
 
 " The margins of the Rio Mangarahy, from its union with 
 the Rio Santa Maria to the colony, are quite thickly inhab 
 ited by an agricultural population, and much more consider 
 ably than those of the main river, because the locality here 
 repays labor better. It appears that on certain places on the 
 head-waters of the Rio Mangarahy traces of wash-gold were 
 found ; at least the names California de Dentro (in the 
 colony), and California de Fora on the southern bank of the 
 Rio do Meio, would indicate it. The colony was founded 
 in 1857 by Germans, who settled on the Santa Maria, and 
 its branch, the Riberao das Farinhas. This choice of locality 
 
78 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 proving injudicious, another spot farther south, near the Quar- 
 tel Braganza, was chosen, and settled also with Germans ; but, 
 through the worst possible management on the part of gov 
 ernment employe s, as well as the bad quality of the lands, the 
 colony became demoralized, and has been a failure." 
 
 " The territory of Santa Leopoldina is composed of quite 
 high, for the most part steep, mountains and narrow valleys, 
 rarely broader than the channel of the river which flows 
 through it. The soil consists principally of quartz sand, the 
 surface soil, usually two to three inches thick, is held to 
 gether by a network of fine roots with some humus. Only 
 in certain places, where one of the valleys widens somewhat 
 near a stream, layers of rich soil brought down from the 
 mountains are found, and here, naturally, is also the great 
 est fertility. The national custom of preparing the soil for 
 the first cultivation, by axe and fire, is, for situations like 
 those of Santa Leopoldina, the most destructive. The heavy 
 fire from the burning of the felled forests destroys partially 
 the layer of humus and organic substances, and although 
 ashes remain as nourishment for the future harvest, it is at 
 the same time deprived of quite a deep layer of soil in which 
 it can take root, and which, in addition, the moisture binds 
 together. Through the cutting down of the trees the steep 
 mountain slopes are exposed to the full influence of the 
 tropical rains, and through these the best part of the culti 
 vated fields is washed away and carried to the Rio Santa 
 Maria, which finally deposits it in the Lameirao of the Bay 
 of Victoria. It is a well-established fact that in Santa Leo 
 poldina the soil, through culture, becomes more quickly un 
 fruitful than in any other colony. The forests with which 
 the mountains of Santa Leopoldina are covered do not pre 
 sent the same majestic appearance as those of the north and 
 
PROVINCE OF ESPIRITO SANTO. 79 
 
 south of the province. They have much more the appear 
 ance of a weak second growth (Capoeiras) than of a virgin 
 forest. All those plants which, to the practised eye of the 
 Brazilian husbandman, bespeak a fruitful soil, such as the 
 Pao d Alho (Garlic-tree), Jacaranda (Rosewood), Taquara- 
 assu (Bamboo), <fcc., either do not appear, or are rep 
 resented by very feeble specimens. The cultivated plants 
 correspond in their development with the forest vegetation. 
 The corn remains low, and ordinarily produces small ears. 
 The stalks often dry up before the ears make their appear 
 ance. The roots of the mandioca are smaller in the second 
 year than they are in the other colonies, and frequently be 
 come, according to the testimony of the colonists, black and 
 useless, which fact is to be accounted for by the want of an 
 adequate thickness of humus. The black beans fail entirely. 
 Equally unfavorable is the character of the soil for the growth 
 of the coffee-tree ; in the first year, while it yet needs but lit 
 tle nourishment, it grows very favorably ; but in the second it 
 sickens, and, as a rule, gives out. Colonists who had set 
 out 1,000 to 1,200 coffee-trees, possessed at the end of the 
 second year only a couple of hundred, and so soon as these, 
 on the following year, had bloomed, and the fruit had set, 
 the leaves rolled up, fell, and the little trees, without ex 
 ception, gave out." 
 
 While in Victoria I met numbers of colonists from Santa 
 Leopoldina, and all told the same story.* There can be no 
 doubt that a lamentable mistake was committed in establish 
 ing the colony in so wretched a region. The lands of the 
 
 * There is a strong prejudice against Germans in this part of the country. 
 They are represented as idle and given to drinking, and I am very sorry to 
 say that it is fully confirmed by my acquaintance with those colonists I met in 
 Victoria. Elsewhere the Germans make good colonists. 
 
80 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 central part of the province of Espirito Santo are very poor, 
 whereas those of the north and south are very fertile ; and 
 Victoria, though possessed of a most excellent harbor, will 
 never in itself, in all probability, become a place of much 
 importance because of the want of fertility of the sur 
 rounding country, and the impracticability of making 
 it one of the ports of the province of Minas Geraes. 
 There are some good lands, as I shall have occasion to 
 show, lying to the northeast, and the as yet undeveloped 
 agricultural regions of the Rio Doce are distant only some 
 sixty miles. The cutting of a canal to unite the Doce by 
 water with Victoria has been advocated, but in my judgment 
 it is not practicable. In the event of the successful coloniza 
 tion of the Doce, a railroad to some point on that river could 
 easily be constructed, so far as the physical difficulties to be 
 encountered are to be considered. The mouth of the Doce 
 is so dangerous to enter that it will never answer as a port, 
 and the Rio Sao Matheos, lying to the north, though it may 
 be entered by small vessels and steamers, is, after all, ill 
 suited to be the port of the Doce. By making Victoria the 
 outlet of the commerce of the Doce many advantages will be 
 gained. Sao Matheos will never answer as a port for a 
 foreign trade. The products of that region, as well as of 
 the Doce, are more likely to go to Rio than elsewhere for 
 final shipment to foreign ports, and this trade is now carried 
 on in small vessels and coasting steamers. The voyage 
 from Rio to Sao Matheos, though often very short, is uncer 
 tain, owing to the prevalence of northeast winds, and the 
 passage of the point of the Doce is often difficult. By mak 
 ing Victoria the port, the voyage to Rio would be shortened 
 one hundred miles at least, and made very much more easy, 
 while Victoria, being a port admitting ships of large ton- 
 
PROVINCE OF ESPIRITO SANTO. 81 
 
 nage, might be made the centre of a direct trade with foreign 
 ports. 
 
 At present the province of Espirito Santo, though 
 possessing abundantly the sources of wealth, is one of 
 the poorest and most wretched of the Empire. The 
 water bordering the coast is very shallow, and just off 
 Victoria are the extensive banks of Victoria, which are 
 very rich in fish, especially garoupas, pargos, vermelhos 
 (species of 8erranus~), &c., and they are much resorted to by 
 the fishermen of the coast, especially from Guarapary ; but 
 such is the sloth of the fishing population that nothing is 
 done to develop this mine of wealth, and the Victorienses 
 eat codfish and European sardines when they might export 
 fish to Europe ! 
 
 I had an opportunity of fishing one day on these banks, 
 and can testify to the abundance of the fish, but I had no 
 opportunity of dredging ; occasionally, as above remarked, 
 the hooks brought up masses of nullipore rich in life, and 
 from a depth of some fifty feet I collected in this way a 
 species of Pterogorgia, apparently new. 
 
 On the northern side of the island of Victoria is, as al 
 ready remarked, a tidal channel running westward and com 
 municating with the Rio Santa Maria. This channel is seen 
 in the accompanying engraving of the Morro de Mestre Al- 
 varo. It receives and carries to the sea a part at least of 
 the water of the Santa Maria and of the river westward. It 
 is but an estuary, its waters ebbing and flowing with the tide. 
 It is very turbid, and oysters grow along its banks for a 
 long distance above the Passagem. I am not aware that 
 it is ever used for the purposes of navigation. Northward of 
 this channel are a few gneiss hills, and thence to the grand 
 mountain mass of the Morro da Serra, or Mestre Alvaro, 
 
 4* F 
 
82 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 stretches a sand-plain which extends eastward to the Ponta 
 do Tubarao and westward to the Rio Santa Maria, bounded 
 on the north by the serra and the steep slope of the edges 
 of the tertiary strata, which lie between the serra and the sea- 
 coast, as seen in the engraving. This plain consists on the 
 surface of coarse white sand without shells. The absence of 
 sea-shells from these marine deposits appears remarkable at 
 first sight, but it is no doubt largely owing to the fact that 
 they have been dissolved out by rains. Where these plains 
 have been cut through by rivers, shells are seen in the lowest 
 beds, as, for instance, on the left bank of the Itabapuana, 
 a few miles above the Barra. The plain north of Victoria 
 bears the same kind of vegetation as that which character 
 izes the plains farther south. Just previous to the last rise 
 of the land these plains were under water, and the hills 
 of the Victoria group stood as islands in the mouth of 
 a bay. 
 
 Crossing the sand-plains, a steep ascent of about fifty feet 
 at Carapina brings one to the level of the tertiary plains, 
 where one leaves the sands. These plains are covered by a 
 clayey soil varying much in fertility. Near Carapina there 
 are some lands available for cultivation, and part of the plain 
 is covered by trees. At Carapina the soil is clayey, with 
 very little sand, and of a slate-blue color ; but going east 
 ward toward the sea the soil grows drier and more sterile. 
 The trees are very scattered and coarse-barked, and when in 
 clumps are free from undergrowth. The open plain is covered 
 by a scanty growth of tall coarse grass, and is diversified by 
 great numbers of ant-hills, irregularly dome-shaped struc 
 tures of clay, often quite round, as hard as stone, and resem 
 bling boulders scattered over the plain. I presume that the 
 character of the vegetation over this region is, as elsewhere, 
 
PROVINCE OF ESPIEITO SANTO. 83 
 
 in part due to fires, which, periodically set by the inhabi 
 tants, have killed all except the most hardy trees and other 
 plants. The Rio Carahype is a little stream which rises, 
 according to Gerber s map, to the northwest of the serra ; 
 but according to my notes and observations it takes its origin 
 in the hills north of the serra, and, traversing the plain, 
 reaches the sea a few miles north of the Ponta do Tubarao. 
 It is only a respectable brook, of no service for navigation, 
 unless it be for canoes. It has cut down through the ter 
 tiary beds, and has a rather deep channel. Its valley is 
 narrow, with steep sides. Near the town of Serra the valley 
 is about one hundred feet in depth, showing that the ter 
 tiary clays, lying against a sloping bottom, probably also 
 thickening towards the hills, form beds that slope gently 
 from the hills to the sea. In some localities near Serra 
 gneiss is exposed in the bottom of the river. Near the 
 hills the surface is irregular, and the plain becomes un 
 dulating and broken, the result, I believe, of the action 
 of glaciers, as well as of streams, from the hills. The 
 soils of the tertiary lands bordering the high grounds, 
 as in the vicinity of Serra and along the Rio Reis Magos, a 
 few miles from its mouth, are quite good, and are used for 
 the culture of coffee, cotton, &c. The forests of these regions 
 are more luxuriant than is elsewhere the case on the plains. 
 Lying to the north of the river, and not far from the shore, 
 is the large and shallow Lagoa Jacune*, which during heavy 
 rains overflows and sends its waters into the Carahype, at 
 which time that river widens its channel and opens its 
 bar. 
 
 The Mcstre Alvaro, or the Morro da Serra, is an isolated, 
 irregularly pyramidal mountain-mass of gneiss standing in 
 the gneiss plain a few miles northwest of Victoria, and 
 
84 GEOLOGY AND PHYSICAL GEOGKAPHY. 
 
 presenting on all sides very similar outlines. Its height I 
 should estimate at about 3,500 feet. It stands like a 
 pyramid on the plain, majestic and alone. More pleas 
 ing and symmetrical in outline, and more isolated in its 
 position, it is more beautiful than Tijuca ; and, seen from 
 the sea, it looms up grand and blue against the tropical sky 
 and the far-off line of serras, which lie along the horizon like 
 the front of an approaching storm. It is densely covered 
 with forest, but on its slopes are extensive coffee-plantations, 
 which yield well. The Diccionario Geografico says that for 
 merly emeralds and magnetic iron were found there.* In very 
 striking contrast with the quiet, sans souci air of Victoria, with 
 its grass-grown streets, is the little village of Serra, which, 
 owing to its being situated in an agricultural region, is one 
 of the most business-like towns in the province of Espirito 
 Santo. This little town is built near the base of the hills 
 on the northeast side of the Mestre Alvaro. I was told that 
 water communication existed between it and the Bay of Vic 
 toria, but my visit to the locality was rather hasty, and this 
 may not be correct. Northward of the Mestre Alvaro 
 are a few gneiss hills stretching in a line toward the Rio 
 dos Reis Magos. West of the Mestre there is much low 
 land. The appearance of the coast from Victoria to within 
 a few miles of the Rio Doce, a distance of nearly fifty miles, 
 as seen from a point at sea about ten miles east of Riacho, 
 is given in the accompanying illustration. The main Serra 
 dos Aimore s is seen in the background, stretching along like 
 
 * Saint Hilairc ascended the Mestre Alvaro. He speaks of meeting with the 
 bamboo Tnquara-assu in the forest, at a considerable elevation above the plain, 
 and remarks that these plants require humidity and considerable elevation. 
 ( SAINT-HI LAIRE, Voyages sur le littoral du Br&il, 2 nd Partic, p. 275.) Von 
 Martius says that the bamboos flourish principally at a height of 1,800 to 
 2,000 feet above the sea. 
 
PROVINCE OF ESPIRITO SANTO. 85 
 
 a wall at a distance of fifteen to twenty-five miles, while 
 in front are the isolated groups of hills of Victoria, Mestre 
 Alvaro, Nova Almeida, &c. 
 
 The Rio Reis Magos, or Apiapitanga, is a little stream 
 which rises in the serra northwest of the Mestre and empties 
 into the sea fifteen miles north of the Bay of Espirito Santo. 
 Canoes ascend it only twenty miles. It empties into a little 
 bay, on the south side of which, situated partly on the level 
 of the tertiary plain and partly beneath the slope, is the 
 ancient village of Nova Almeida.* The tertiary beds con 
 tinue northward to the port of Aldea Velha, or Santa Cruz, 
 north of which they disappear from the coast, and give way to 
 marshes and swamps that occupy the coast thence to the Rio 
 Doce. Between Nova Almeida and Santa Cruz, as well as to 
 the north, the sandstone beds form in some parts rocky shores, 
 off which lie isolated skerries, which obstruct both bays. 
 
 The river Santa Cruz is, like the Reis Magos, a little 
 stream affording navigation for canoes only. Its principal 
 affluents are the Piriqui-assu f and Piriqui-mirim. It opens 
 into a little bay like that of Nova Almeida, and which affords 
 a harbor for small vessels. The bays of Nova Almeida and 
 Santa Cruz are noteworthy in that they open broadly to the 
 sea, and, unlike the mouths of the other rivers, are unob 
 structed by bars, a circumstance which is, perhaps, owing 
 to the fact that the rivers emptying into them bring down 
 but little sediment. 
 
 * Here was established anciently a Jesuit missionary station, with a school 
 for furnishing instruction in the Tupi language. 
 
 t Saint-Hilaire describes heaps of oyster and other shells bordering the river 
 Piriqui-assii, near Aldea Vclha, which are without doubt Kjcekkenmceddings. 
 Similar shell-heaps, or ostreiras, as they are called in Brazil, are found on the 
 coast of Sao Paulo, and on the Ilha do Governador, in the Bay of Rio. They 
 often contain human remains, pottery, &c. 
 
86 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 I have sailed several times close along the shore from Santa 
 Cruz to the Eio Doce, but I have never landed, and I am 
 unable to describe it in detail, as well as the flat lands lying 
 behind it, nor have I any trustworthy information concern 
 ing them. The shore is bordered by a sand-beach like that 
 stretching south of the Parahyba do Sul, and the lands in 
 its rear are plains diversified by swamps, shallow lagoons, 
 and channels which, having never been explored, are laid 
 down incorrectly on all maps. Much of this low ground is 
 suitable for pasturage, and, so far as my observations go, the 
 greater part is very heavily wooded. 
 
 The great Serra da Mantiqueira, separating itself from the 
 Serra do Mar near Sao Paulo, runs to the north of that 
 serra, and almost parallel with it, on the northern confines 
 of the provinces of Sao Paulo and Rio do Janeiro, into 
 Minas, passing near Barbacena, whence it continues with 
 the same trend, under various names, beyond the Rio Doce. 
 To the southeast it gives rise to a host of little rivers 
 which flow into the Parahyba do Sul, while on the north 
 eastern side the Rio Grande, one of the tributaries of the 
 Parana system, and the Rio Doce take their rise. From 
 Barbacena a cordilhcira runs northward, separating the 
 waters of the Rio Doce from those of the Rio Grande and 
 the Sao Francisco, while another line of serras, with a gen 
 eral northeast course, leaving the Serra do Espinha^o near 
 Diamantina, forms a water-shed between the rivers Doce 
 and Jcquitinhonha. The basin of the Doce forms an un 
 equal-sided quadrilateral figure, whose sides measure very 
 nearly as follows : the northeastern side 120, the southeast 
 ern 230, the western 190, and the northern 90 miles, which 
 would make the area drained by the river very much larger 
 than that of the Parahyba do Sul. Gerber estimates the 
 
PROVINCE OF ESPIR1TO SANTO. 87 
 
 area embraced within the basin of the Doce at 2,300 square 
 leagues. The Rio Doce, under the name of Chopoto, takes 
 its rise only a few miles from Barbacena, in the angle formed 
 by the union of the Serra da Mantiqueira and the Serra do 
 Espinhac,o. From this point the waters flow westward into 
 the Rio das Mortes, a tributary of the Parana system, east 
 ward into the Pomba, one of the branches of the Parahyba, 
 and northward into the Doce. Its course is at first about 
 north-northeast up to latitude 19, when it bends abruptly 
 around the Serra das Ibiturunas, and then flows in a south 
 east direction to the confines of the province, when, after 
 passing a series of rapids called the Escadinhas, it reaches 
 lower grounds, and is thereafter navigable to its mouth. 
 
 This river is of so much importance that I add a transla 
 tion of a description of it from the Diccionario Geografico, 
 Art. Doce. 
 
 " The ancient geographers considered the Ribcirao do 
 Carmo as the principal origin of the Rio Doce, while others 
 assert that it rises near Sabara, in the Ribciro Santa Bar 
 bara, which empties into the Piracicaba. But if one means 
 by the source of a river that point which is at the greatest 
 distance from its mouth, we ought to place that of the Rio 
 Doce at twelve leagues * to the east of the Villa of Barba 
 cena, in the spot where the Rio Chopoto begins. This river 
 runs about twenty leagues in a northward direction, receiv 
 ing on its left bank the Rio das Pirangas, besides various 
 streams from both sides. It only offers easy navigation for 
 canoes, when, having watered the settlement of Santa Anna 
 do Deserto, it inclines toward the northwest, being aug 
 mented by the stream Turvo on the right and the Rio 
 
 * According to Gerbcr s map, the source of the Eio Doce (Chopoto ) lies at 
 about five leagues east of Barbacena. 
 
88 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Gualacho * on the left. The waters of this river make it 
 change its course a little towards the east, and both united 
 precipitate themselves over the fall called Inferno ; below 
 this fall the Rio Doce takes the name it bears, and flows 
 gently on, receiving on the right the little Rio da Casca, 
 and on the left the Piracicaba, and, six leagues farther on, 
 it passes over reefs blackened by time, whence the name 
 Escuro given to this fall. Three leagues lower down, on its 
 left, are the mouths of the rivers Santa Antonio and Corren- 
 tes, at a distance of eight leagues from one another. f Below 
 the last of these river-mouths is the Cachoeira Bagauriz, 
 where a pointed rock divides the waters of the Doce, which 
 subdivide again and again before uniting anew in a species 
 of basin, formed, as it appears, by several islets. This basin 
 extends for a distance of two leagues, and, because of the 
 current, much dexterity in the government of a canoe is re 
 quired to reach it. At the end of this series of islets the Rio 
 Doce takes a more quiet course, and, the mouth of the Sac,u- 
 hipequeno passed, it becomes once more turbulent, and is 
 successively impeded by the little Cachoeira d llha-Braba, 
 with that of Fiqueira, much more dangerous, in the Serra 
 Beteruna, where it is necessary to transport the canoes by 
 land a distance of fifteen brayas, and with that of the Re- 
 bojo-do-Capim ; five leagues farther down stream the Rio 
 Sa^uhi-grande comes to swell it on the left with its tribute 
 of waters, after having watered the comarca of Serro Frio. 
 Passing this tributary, the Rio Doce receives from different 
 quarters an innumerable number of limpid brooks, and 
 
 * According to Eschwcge, quoted by Gcrbcr, the Barra do Gualacho is 341 
 metres above the sea-level. 
 
 t On Gerber s map the mouth of the Correntes is only two and a half leagues 
 below that of the Santa Antonio. 
 
PROVINCE OF ESPIEITO SANTO. 89 
 
 makes many turns before arriving at the fall denomi 
 nated Cachoeirinha, over which canoes pass without being 
 unloaded ; still farther down it receives on its right the Ribe- 
 irao Laranjeira, and a little beyond, to the right, the Cuiate, 
 which empties into it, when the main river becomes majestic 
 for the distance of two leagues, below which various reefs, 
 with some falls of little importance, and whirlpools make 
 navigation very difficult, without entirely destroying it. 
 These different obstacles are known under the names of 
 Rebojo de Joao Pinto and Rebojo da Onc,a, distant two 
 leagues from one another. The second of these obstacles 
 passed, the current flows to the right in the summer, but 
 to the left in the rainy season. Half a league onward the 
 bed of the river describes some diagonal lines, which imitate 
 a capital M, the name which is commonly given to it ; and 
 one league lower down various reefs, called the Cachoeirao, 
 intercept navigation, so that it is necessary to transport by 
 land the canoes and goods. Two leagues below the fall of 
 Cachoeirao are encountered three other whirlpools, which are 
 not dangerous, and the island of Natividade, which divides 
 the Rio Doce into two unequal arms. In the dry season the 
 canoes are unloaded at this island to pass the great Cacho- 
 eira called the Escadinhas, because for one league it is 
 formed of steps of stone. The canoe-men carry the goods 
 on their backs as far as Porto de Souza ; but when the 
 waters abound the canoes descend without difficulty to the 
 Registo de Lorena, near the confluence of the river Mandu 
 (Guandu), which comes from the south and serves as the 
 limit to the provinces of Mirias Geraes and Espirito Santo, 
 and ascend it also with cargo, albeit with some difficulty. 
 Between the mouth of the Mandu and Porto de Souza sev 
 eral rapids are encountered, which are descended with ease, 
 
90 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 but which cannot be ascended except by hard rowing, or by 
 towing the canoes." 
 
 The Rio Manhuassu is quite a considerable little stream, 
 which enters the Doce from the south, opposite the island 
 of Natividade. Very little is known about it, as it flows 
 through a forest region inhabited by the Botocudos. The 
 Guandu is, as above described, only a little stream, and of 
 very little importance. The country bordering the river 
 near its junction with the Doce is very flat, with deep allu 
 vial, clayey soils. It is heavily timbered, and affords most 
 fertile lands for the agriculturist ; but the Botocudos, 
 who have been until late years hostile, have prevented the 
 successful colonization of the country. A little colony of a 
 few families has been established on the Guandu, but it was 
 not, at the time of my visit, flourishing. 
 
 The rocks exposed here in the river channel are gneiss. 
 I observed some heavy veins of milky-white quartz near 
 that place, but I had not time to examine them. Opposite 
 Guandu, on the north side of the river, there are several 
 quite high gneiss hills, and opposite Porto de Souza is a 
 bare hill not far from the river. Gerbcr has laid down on 
 his map a little river entering the Doce opposite the Guandu, 
 which appears to be a mistake. At Porto de Souza the 
 river valley is very narrow, and . the river was, at the time 
 of my visit in December, 1865, not more than 250 feet wide 
 opposite the port. 
 
 The port is situated at the foot of a series of rapids, at the 
 head of navigation. Even in the dry season it would be 
 possible at all times, I was assured, for a small steamer to 
 reach it. Gneiss is exposed in the bed of the river and 
 along the banks. It is gray, coarse, and homogeneous. 
 Strike N. 60 D E. Dip 45 Northeastward. In the river- 
 
PROVINCE OF ESPIR1TO SANTO. 91 
 
 banks the surface of the gneiss is rugged, and not smooth as 
 in glaciated surfaces, showing that it has been determined 
 by water action. Above it are beds of coarse, yellowish 
 sands and quartz gravel, the latter often very coarse. At 
 the Quartel, or Barracks, these beds reach a level of about 
 twenty feet above that of the river. Above these is a 
 layer of brown, clayey earth, with scarcely any admixture 
 of sand, but full of little silvery mica flakes. This affords 
 a good, deep, fertile soil, suitable for coffee, corn, beans, 
 castor-beans, &c., &c. The forests of this region are 
 quite luxuriant, and are rich in valuable woods and 
 game.* All this region is finely adapted for agricultural 
 purposes, and some day must become the home of an agri 
 cultural population ; but the Indians have so far held sway 
 over it. 
 
 At the port the left-margin of the river is occupied by 
 some immense sand-banks, covered during the rainy season, 
 when the river rises some twenty feet above its ordinary 
 level. Descending the river from Porto de Souza, the river 
 continues very narrow, deep, and rapid, with rocky banks 
 for a mile or more, the banks becoming lower as one de 
 scends. The gneiss is exposed in the bottom of the banks 
 on both sides of the river, and over this, as at the Porto, 
 are alluvial deposits, making the banks eight to fifteen feet 
 in height. At a distance from the river are gneiss hills. 
 A quarter of a mile below the Ilha da Esperanc.a the river 
 widens to at least 600 feet. The Rio Mutum,f laid down 
 
 * I observed in the forest between Porto dc Souza and Guandii great 
 numbers of Barrigndo trees (Bombax), some of considerable size. 
 
 t The llio Docc is very rich in fish, a large collection of which was made by 
 my companion and myself for Professor Agassiz. Prinz Max. zu Neu Wied 
 speaks of the occurrence of a species of sawfish (Pristis Scrra?) in the lower 
 waters of the Docc, and says that it enters the Lagoa Juparanaa. Mr. Copeland 
 
92 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 RIO DOCE AT PORTO DE SOUZA. 
 
 on Gerber s map as being some three or four leagues below 
 the Porto, is not more than three miles. It is but a brook ; 
 but it has a fall near the mouth, and, being in the midst of 
 a very fertile country, may be of importance by and by for 
 its water-power. On the Mutum the land is in some places 
 flat, and thirty to forty feet high, and may be in part ter 
 tiary ; but I doubt it. On the north side of the river are 
 many forest-clothed gneiss hills, the most of which range 
 from 600 to 700 feet in height. East of these, a few miles 
 from the Mutum, and standing back from the Doce, is a 
 prominent hill called the Morro do Padre, the greater part 
 of which is bare, and which must be at least 2,000 feet in 
 height. On the opposite side of the river is the Morro do 
 
 and I took specimens at or near the mouth of the Mutum. They certainly ascend 
 to Porto de Souza. The canoe-men said that they ascended to the Guandu. The 
 fishermen all said that this fish secured its prey by striking a lateral blow with its 
 long snout or jaw. On one of the saw-teeth of a large specimen we took there 
 was impaled the large, tough scale of a curumata, which could only have been 
 pierced by a smart blow, when attached to the side of the fish. The fisherman 
 said that a true shark, caissao de denies, was found in the Doce. 
 
PROVINCE OF ESPIRITO SANTO. 93 
 
 Lage among gneiss hills.* Leaving the gneiss hills below 
 the little river Sao Joao, the land becomes lower, and the 
 hills flat-topped, or with flat outlines, and a rather gentle 
 slope toward the river. Some of the hills appear to be 
 gneiss, others tertiary. The low lands bordering the river 
 are alluvial. So exceedingly dense is the forest covering 
 this whole country, that, except it be occasionally a bare 
 gneiss rock by the river-side, a bare gneiss hill which is 
 never to be mistaken, or the exposures of the alluvial deposits 
 of the river, there is no guide for the working out of geo 
 logical features other than the general topographical out 
 lines of the country. The clothing of forest tends very 
 much to exaggerate the height of the lower lands. No 
 where in Brazil, not even at Para, have I seen a more lux 
 uriant forest growth than that of the Doce. The trees, all 
 bound together by llianas, and filled in with a dense under 
 growth of palms and shrubs, crowd down to the water s 
 edge, and stretch their great vine-draped branches out over 
 the river, as if in want of light and air. The forest forms a 
 dense wall along the river, so dense that the eye does not 
 penetrate into its shade, and one must be armed with a 
 strong wood-knife who would enter it. 
 
 In these forests flourishes the Jacaranda, or rosewood {Big- 
 nonia Brasiliana Lam.), which once abounded along the river, 
 but, having been extensively cut for exportation, is now to be 
 met with of sufficient size for cutting only at a distance from 
 the stream. The principal article of export from the Doce 
 has been rosewood, which has the reputation of being of a 
 fine quality. The Cupiuba ( Copaifera officinalis} , furnish 
 ing a valuable wood and an abundant oil used in medicine 
 
 * Right bank opposite Ilha do Lage. Gneiss bands very siliceous. Strike 
 N. 80 E. Dip northward 50. 
 
94 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and the arts, abounds here, together with the Pao Brazil 
 ( Cfzsalpinia echinata Fr. AIL), noted for its once costly 
 tint; the Sapucaia (Lecythis*), furnishing food to the Boto- 
 cudos ; the Cedro (Cedrela*), Ipe* (Teccwm), Pao d Arco 
 (Big-nonia), Peroba (Aspidospermum ) , Putumuju (Ptttu- 
 mvju), Yinhatico (Acacia), and species of Genipa, MachcE- 
 rum, Ing-a, Bowditchia, <fcc., abound. The names of the 
 trees furnishing valuable woods for construction and cabinet 
 work, many of great beauty and durability, are legion, and 
 when the country becomes inhabited, these must become 
 a source of wealth. 
 
 Several species of palms, among them the Airi (Astro- 
 caryuni) and Palmetto (Euterpe) ,. flourish in the forest, 
 while the Embaiiba (Cecropia), the food of the sloth, with 
 the Ubas and Heliconias, form one of the marked features 
 of the vegetation of the river-banks.* 
 
 Game is exceedingly abundant, among which may be 
 mentioned the Anta ( Tapirus Americanus) , whose tracks, 
 together with those of the Capabara (Hydrochccrus Capa- 
 bara), are everywhere seen by the river margins. The Paca 
 (Caslogenys Paca) and Cutia (Dasyprocta) are very com 
 mon, and are valued for food. There are, at least, two spe 
 cies of sloth found here, Bradypus tridactylus and B. tor- 
 quatus, together with species of opossum (Didelphys) and 
 Cuati (Nasua*). Armadillos (Dasypus) abound as else 
 where, and I have seen at least two species. Wild pigs 
 Caititus and Queixadas (Dicotyles), are found in herds in 
 
 * Prince Max. zu Neu Wied gives a plate representing a view on the Rio 
 Doce. It was evidently not drawn from nati;re, but it gives quite a good idea 
 of the shore vegetation. An immense alligator is represented in the foreground. 
 One may spend a month on the Doce and not see a single alligator, and those 
 of the Doce are very small. 
 
PROVINCE OF ESPIRITO SANTO. 95 
 
 the forest, and are hunted for food. At least four species of 
 Fells occur here, Felis Onga, F. concolor, F. pardalis, F. 
 macroura, together with the fierce black jaguar, which 
 may be only a variety. One or two species of Cervus are 
 not uncommon, together with hares and squirrels. Of 
 monkeys there are the following : Ateles hypoxanthus, My- 
 ce.tes ursinus, species of Cebus and Callitkrix, and Jacclms 
 (Hapale) leucocephalm. Jacupembas (Penelope mar ail 
 Linn.), Mutuns (Crax), Araras (Psittacus macoa Linn.) 
 and other parrots, are very abundant. 
 
 The river, after leaving the gneiss lands, widens very 
 much, and is in some places 800 to 1,000 feet wide, 
 occasionally diversified by wooded islands, and affording 
 stretches of river scenery of very great beauty. Francyl- 
 vania is the name of a settlement established on the 
 northern bank of the Doce, at a short distance back from 
 the river, opposite the mouth of the little river Santa 
 Maria, and not far above the mouth of the Rio Panca. 
 Here, not many years ago, was established a Brazilian 
 colony, under the direction of one Dr. Franca Leite,* which 
 proved a failure ; its site being now marked only by a luxuri 
 ant second growth (capoeira), which on the Brazilian coast 
 
 * The history of this colony is briefly as follows : Dr. Franca Leite, some fif 
 teen or more years ago, conceived the idea of establishing a colony on the Eio 
 Doce, and at Francylvania, aided by government, he formed a settlement. 
 Mills for sawing lumber and grinding mandioca were erected, and a consider 
 able amount of ground was put under cultivation. Establishments were opened 
 at the Povoacao, at Monserras, and Ipyranga, where many cattle were raised. 
 But Dr. Leite s plan proved visionary. He failed to get the aid he demanded ; 
 the colonists, dissatisfied, and harassed by the Botocudos, who were very 
 troublesome neighbors, began to withdraw, and after an existence of some three 
 years, the Botocudos put an end to this colony by killing Ervalina, Leite s 
 brother-in-law, and a slave, and burning the settlement ; and thus failed another 
 attempt to settle the Doce. 
 
96 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 rapidly springs up on abandoned lands. The soils in the 
 vicinity of Francylvania are of the most productive kind, 
 and the locality was wisely chosen for the establishment of 
 a colony. 
 
 Descending from Francylvania to the hills of Santa An 
 tonio, the lands bordering the river are still lower, and the 
 hills have long, gentle slopes, but no rocks are seen, except 
 gneiss, on the borders of the river. The islands all consist 
 of sand, overlaid by a thick bed of brown soil. Alluvial 
 lands of the same character occur on both sides of the river, 
 forming irregular strips. These lands, which are liable to 
 be overflowed during the enchente, are of the highest fertil 
 ity, and are especially proper for the culture of sugar-cane. 
 
 A few leagues south of Francylvania the river passes 
 through a region diversified by hills of gneiss several hun 
 dred feet high. On the south side of the river the western 
 most of these hills form a range which, under the name of 
 the Serra de Sant Antonio, is seen stretching off south 
 ward, tying in with the coast mountains of Santa Cruz and 
 Nova Almeida. In this belt of country are many lagoas, 
 some of them of considerable size. Among these may be 
 mentioned the Lagua Sant Antonio do Norte, on the north 
 side, and the Lagoas Pao Gigante and Limao, on the south, 
 all of which communicate with the river by small, black- 
 water streams. These lakes and streams have been very 
 incorrectly laid down on the maps, owing to the fact that no 
 trustworthy survey of the river has ever been published.* 
 A few miles farther east, and opposite the Barra do Rio 
 Limao, is a range of high hills, which stretches off as far as 
 
 * I am informed that the river was surveyed and mapped for the English 
 Company of the Doce by Mr. Fred. Wilner, but I am not aware that the map 
 was ever published. 
 
PROVINCE OF ESP1RITO SANTO. 
 
 97 
 
 one can see from the river towards the northeast. This 
 hilly country ends on the river, just to the east of these 
 hills, in a sharp ridge about a mile long, escarped on the 
 eastern side, and known as the Morro da Terra Alta. 
 
 LOOKING UP THE DOCK FROM NEAR LTNITAEES. 
 
 Passing through this country the river is much diversified 
 by islands, and is in some places more than 1,000 feet wide. 
 The lands here vary much in character, and are noted for 
 their richness. Over the whole country spreads the heaviest 
 forest growth, but at the time of my visit it was unbroken 
 by a fazenda, and in the hands of the Indians and rosewood- 
 cutters. On leaving the hills the river soon widens to at 
 least half a mile, and is full of beautiful wooded islands, 
 while it is bordered by alluvial lands which, during the 
 enchenle, are liable to be overflowed. These lands extend 
 to Linhares. About two leagues above this town there are, 
 on the southern bank, a few clearings. I think that no 
 where in Brazil have I ever seen so rank and luxuriant a 
 vegetation. These lands are covered with the same brown 
 ish, clayey soil we have observed farther up the river. 
 5 o 
 
98 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 This soil, which is called mapape, or massape, is found on 
 all the alluvial lands bordering the rivers on the Brazilian 
 coast, and is noted for its great productiveness. 
 
 On the Rio Doce those massape lands which are of suffi 
 cient height to escape the effects of the enchente may be 
 used for the cultivation of almost any of the products of the 
 country, such as sugar-cane, tobacco, coffee, cotton, man- 
 dioca, &c., but a large part is likely to be submerged 
 every year when the river is full. The annual over 
 flow begins in December with the daily thunder-storms, 
 and lasts usually until March. During its prevalence the 
 river margins are overflowed for a greater or less length of 
 time, the freshets of March often being as high as those of 
 December. In the year 1833 occurred an extraordinarily 
 heavy freshet, since when nothing like it has been known. 
 The water of the river, even in the dry season, is very 
 turbid with sediment, and of a light yellowish-brown color. 
 During the enchente it becomes very much more turbid, and 
 a thin deposit of mud is thrown down over the flat lands 
 every year. On the subsidence of the waters, the vegetation 
 left decaying over the wet country is apt to breed fevers, and 
 the Rio Doce has had a very bad reputation for being very 
 unhealthy. It certainly is feverish, but I could not learn 
 that it was any worse than Sao Matheos, or even so bad. 
 My companion and I suffered nothing from our visit. The 
 river waters are bad, but if allowed to deposit their sedi 
 ment, and stand some time, become very potable and safe. 
 The climate of the Doce is warm and very damp, and it is 
 upon the distribution of rain throughout the entire year that 
 the luxuriance of the vegetation seems to depend. The cli 
 mate is damper than at Victoria or Sao Matheos. Came 
 secca, or dried salted beef, which keeps well elsewhere, soon 
 spoils here in certain seasons of the year. 
 
PROVINCE OF ESPIRITO SANTO. 99 
 
 The massape* grounds which are liable to be overflowed 
 are used for the culture of almost all the products of the 
 country, except mandioca, cotton, and coffee. The root of 
 the mandioca requiring more than one season to ma 
 ture, is likely to be injured by a freshet, so that its cul 
 ture is confined to the higher grounds. These massape* lands 
 are especially good for sugar-cane, which, as at Campos, is 
 not injured by an overflow, and also for beans, Indian corn, 
 rice, bananas, <tc. The corn and beans are planted usually 
 in March or April, so as to become ripe before the enchente. 
 Lands of this character are very extensive, and the Doce 
 region is adapted to sustain a very large population. 
 There can be no doubt that with the clearing of the forests 
 and the tilling of the ground the region would become more 
 healthy. 
 
 Linhares is built on the left bank of the river, on the top of a 
 bluff formed by the projection southward of a point from the 
 great tertiary plain lying north of the Doce. The bluff is, if 
 I remember rightly, about eighty feet high, and exposes beds 
 of white and red clays of the ordinary type. On the Doce 
 these clays occupy near Linhares only a small extension on 
 the northern bank. Their boundary line trends off north 
 ward to Sao Matheos and northwestward to the Lagoa Jupa 
 ranaa,* which is held in a basin scooped out of the beds of 
 this formation. The lagoa lies at a distance of about two 
 miles to the north-northwest of Linhares, and communicates 
 with the Doce at Linhares by a very narrow and tortuous, 
 but deep channel, called the Rio Juparanaa, w T hich flows over 
 the low wooded ground close to the edge of the bluffs. This 
 channel is about fifty feet wide, and, according to Senhor 
 
 * The word " Juparanaa," according to Prinz Neu Wied, is Tupi, and means 
 " sea." 
 
100 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Raffael P. do Carvalho, furnishes at all times at least four 
 to five feet of water, which would be sufficient for a little 
 steamer. It is exceedingly tortuous, but it would be a very- 
 easy matter to cut off some of the bends, and thus shorten 
 very much the distance between the Doce and the lake. 
 
 At the foot of the lake the river is bordered on the west 
 side by a narrow strip of low ground, north of which the 
 tertiary bluffs begin ; and, with the exception of this little 
 stretch of alluvial land, and of a similar stretch at the head 
 of the lake, the lake is bordered by the bluffs. Freireiss * 
 gives the length of the lake as seven leagues, width half a 
 league, and circumference sixteen to eighteen leagues. The 
 Dicclonario Geografico (vide Art. Juparanari) , which is here 
 manifestly inaccurate, gives the circumference as only five 
 leagues. According to my estimate, and the information 
 of Senhor Raffael, it must be at least twenty miles long and 
 in some places four miles wide. It is very deep, and in 
 some places, according to Freireiss, the depth is at least 
 eight to twelve klafter (fathoms) . The water of the lake 
 and river is of a light milky color. The banks of the lake are 
 eighty to one hundred and fifty feet high, as near as I could 
 judge, the height being greater at the head of the lake. 
 Along the eastern side, between the foot and the Fazenda do 
 Guaxe, white and pink clays are exposed in the bluffs, 
 and in many places the coarse red sandstone of the tertiary 
 crops out at water-level or thereabouts. About a mile north 
 of Guaxe, on the east shore of the lake, and opposite a very 
 small island of gneiss, a point called the Ponta de Ouro, a 
 quarter of a mile long, juts out into the lake. This point is 
 about fifty feet in height, and forms on the southern side a 
 line of picturesque cliffs. The strata composing this point 
 
 * Quoted by Neu Wied, Rdse nach Brasilien, Vol. I. p. 214. 
 
PROVINCE OF ESPIEITO SANTO. 101 
 
 are perfectly horizontal, and consist of a white or pink feld- 
 spathic clay, in some localities with no admixture of sand, 
 but for the most part with a large percentage of coarse, 
 angular, or slightly rounded sand and gravel, these 
 materials being scattered through the mass apparently 
 with no order whatever. The lowest beds are the most 
 sandy. The red color is due to ferric oxide, which is dis 
 tributed through the mass very unequally, sometimes ce 
 menting together portions of the beds into stalagmite- 
 like masses penetrating the clays. The tint is sometimes 
 yellow. Under the clays is the coarse red sandstone 
 which occurs in very solid and compact masses, in which 
 case the rock is regularly and evenly cemented. The 
 bluffs are steep and wooded, as also is the plain above. 
 Small farms are located along the lake, with an occasional 
 little fazenda. The slopes of the bluffs are found to yield 
 very abundantly, and to produce excellent coffee. The 
 LagQa Juparanaa is, like the Doce, very full of fish. I 
 found here two species of Cagados, probably the Emys 
 depressa and Emys radiolata of Max. zu Neu Wied. A 
 species of Unio is abundant in the lake, and is said to be 
 used for food. I saw heaps of the shells lying in front of a 
 fisherman s hut, but the animal may have been used for bait. 
 At the head of the lake there enters a little river called 
 the Sao Raffael. It rises in the forest, in the country of the 
 Botocudos, and has never been explored.* Gerber s map 
 represents a Rio Preto as flowing into the Juparanaa, but 
 no such river exists. The lake also, as laid down on his 
 map, is too small, and the island is too large. It is really 
 only a rock. The head of the lake cannot be distant from 
 
 * Senhor Raffael says that above it divides into three branches, and that it is 
 rich in fish, but that it contains no surubims (Platystoma). 
 
102 
 
 LAGOA JUPARANAA, LOOKING TOWARDS THE OUTLET. 
 
 the city of Sao Matheos, according to the best information I 
 have received, more than three to four leagues. The bar 
 of the Doce is so bad as practically to forbid the entry of 
 vessels, although they sometimes cross it ; but the river is 
 navigable for a little steamer, during the whole year, from 
 its mouth to Porto de Souza, a distance of ninety miles ; so 
 also is Lake Juparanaa. Until a railroad is built to Vic 
 toria it would seem best to construct a good wagon-road 
 through the forest, from the head of lake Juparanaa, over 
 the plain, to Sao Matheos, and make Sao Matheos the 
 port of the Doce ; but Sao Matheos can never offer the 
 same advantages as a port that Victoria docs. 
 
 The lands surrounding the lake are plains covered by 
 forest, but owing to the dryness and little fertility of the 
 soil, it is not very luxuriant. The soils of these high lands, 
 however, vary very much in quality, in some localities being 
 excellent for cotton, mandioca, &c., in others sandy and 
 barren. Rosewood abounds in these forests, and is quite 
 extensively cut. 
 
PROVINCE OF ESPIRITO SANTO. 
 
 103 
 
 The higher lands near Linharcs appear to be fertile, and 
 I have nowhere seen more vigorous and luxuriant crops 
 of sugar-cane, bananas, &c., than I saw growing on the 
 plantation of Senhor Alexandre, by the river-side, just 
 below the town, and situated on alluvial grounds. The 
 river here appears to be bordered by three terraces, but I 
 had no time to examine them closely. 
 
 Lying almost parallel with the river, and to the northeast 
 of Linharcs, is a very beautiful, narrow lagoon, which looks 
 like an old river valley, and is called the Lagoa do Aviso. 
 It is said to empty into the Doce to the east of Linhares. 
 There are other lakes in the vicinity. 
 
 LAGOA DO AVISO. 
 
 From Linhares the river runs off in a southeasterly direc 
 tion to the sea. The stream, is wide and interspersed 
 with many islands. The shores are low, forest-clothed, and 
 overflowed during the annual freshet. At the mouth the 
 river widens very much and enters the sea obliquely, from 
 behind a long sand-spit which extends southward from the 
 left bank. The river just inside the mouth forms a fine sheet 
 
104 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of water, but it is very shallow, the bottom being shifting, 
 so that the soundings vary with the quantity of water in 
 the river. So great is the amount of water, and so shal 
 low is the river, that the tide is not felt inside the mouth, 
 and the water is always fresh. Great numbers of drift 
 trees are brought down during the freshets, and the sea- 
 beaches of the vicinity are strewn with them. The 
 mouth is wide, shallow, and obstructed by a bar, on which 
 the waves break fearfully. It is always difficult, and 
 sometimes for weeks together impossible, to enter the 
 Doce, and very many vessels have been lost in the attempt.* 
 Just northeast of the bar are extensive banks, which ex 
 tend two or three miles out to sea. There is a little 
 hamlet on the right bank of the river, near its mouth, 
 and a short distance above, on the left bank, is a small set 
 tlement called the Povoagao. The lands here are sandy, 
 but they arc, for such soils, quite fertile, producing man- 
 dioca, cotton, mamona, sugar-cane, &c. The coast at the 
 mouth of the Doce projects considerably to the eastward, 
 and there is usually a strong current near the shore, de 
 pending for its direction upon the wind. 
 
 Many years ago an English company was formed to open 
 the river Doce as a highway for navigation into Minas, but, 
 from the above description of the river, any one can see how 
 unfit it is for navigation above the falls. The enterprise of 
 course proved a failure, and a curse has undeservedly fallen 
 on the Doce. At present a very small commerce in salt is 
 carried on between the coast and various points along the 
 river, the salt being transported in canoes. The journey 
 from the sea to Correntes, a town situated on the river 
 of the same name, an affluent of the Doce, a distance 
 
 * The coasting steamers occasionally enter the Doce. 
 
PROVINCE OF ESPIRITO SANTO. 105 
 
 in all of about two hundred miles, consumes over forty 
 days. From the Barra to Porto de Souza the journey 
 is easily performed ; but above that point the canoes 
 must be towed and poled with the greatest difficulty, 
 and at very short intervals the load has to be removed, 
 so as to allow the empty canoe to pass a waterfall. 
 On the coast a bag of salt of about sixty pounds costs 
 2 $000 (two milreis, or about a dollar). In Minas it costs 
 eight milreis or more. As nearly as I could learn, the 
 whole enterprise of the English company was conducted 
 with great extravagance and want of good management, 
 so that it is no wonder it failed. It is useless to think 
 of making the Rio Doce the highway to Minas, and of 
 using the mouth of the river as a port. The country bor 
 dering the Rio Doce and the Lake Juparanaa, and extend 
 ing westward for some distance beyond Porto de Souza, 
 must be treated as a great agricultural region by itself. 
 
 Since my visit to the Doce quite a number of Amer 
 ican families have settled on the river, forming a little 
 colony, which, so far as I can ascertain, promises to be a 
 success. The colonists came from the Southern States, 
 from a climate not so very different from that of Espirito 
 Santo, and those that I saw looked like men through whose 
 hands a better future might be worked out for the Doce. 
 We hope that their enterprise may be successful.* 
 
 Northward to the mouth of the Rio Sao Matheos stretches 
 a sand-beach, broken only by one or two river-mouths, 
 and back of which, between the shore and the tertiary 
 bluifs, is a wide area of swamps and lagoons, a region al- 
 
 * In the spring of 1868 there were on the Doce, according to Burton, four 
 hundred Americans, who were doing well, and were " studying coffee." 
 (Highlands of Brazil, Vol. I. p. 5.) 
 
106 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 most impassable, and never yet mapped. Just north of the 
 Doce and near the coast is a large lagoon called Monserras. 
 During the dry season this is separated from the sea by the 
 sand-beach, but when the rains come it opens for itself a 
 channel to the sea, which channel remains open until the 
 dry season returns.* A more waste and desert region than 
 the shore between the Doce and Sao Matheos can scarcely be 
 imagined ; but it is the high-road, and must be followed in 
 going to Sao Matheos. The weary sand-beach stretches ahead 
 to the horizon, dancing in the hot air, and dimmed by the drift 
 ing spray from the ocean breakers, that pour their thunder 
 ing, blinding surges on the desolate shore. A line of monoto 
 nous sand-heaps, like a great tumbling billow ready to burst 
 on the low grounds behind, runs parallel with the beach, 
 bare, or scantily covered by tufts of grass, dwarf palms, 
 &c., no shade, no water. The road is a strip out of Sahara. 
 On one side is the sea, on the other a miasmatic, pathless 
 swamp. Ordinarily the lagoons lie at some little distance 
 from the shore, and are separated from it by a dense, impene 
 trable thicket ; but at Pitanguinha there is a little lagoon by 
 the shore where water maybe obtained. The sands on these 
 beaches are coarse, and do not pack so as to afford a good 
 footing. Animals sink at every step, and the journey from 
 the Doce to the Sao Matheos is excessively fatiguing, the 
 traveller being, in addition, liable to suffer severely from 
 thirst. About thirty miles north of the Doce is Barra Secca, 
 where a little river, draining the marshes of the interior, 
 empties into the sea. Just where it takes its rise, and what 
 is its course, no one, even of the immediate vicinity, knew. 
 It is usually set down as draining a lake called Tapada ; but 
 
 * When I went to the Doce from Silo Matheos this bar was closed, but on 
 my return, in the latter part of December, it was open, and dangerous to cross. 
 
PROVINCE OF ESPIRITO SANTO. 107 
 
 that is not possible, for the lake lies only a few hundred 
 yards from the mouth of the river, and at a very much high 
 er level. The tide ebbs and flows in the Secca and evidently 
 runs up a long distance. The river comes from the south 
 and, just before reaching the sea, flows along in the rear of 
 the broad beach ridge, from behind which it escapes at the 
 barra, and, cutting a channel across the beach, flows into the 
 sea. The beach is constantly changing by the drifting of the 
 sands, through the action of the wind and waves, so that the 
 barra is as constantly shifting, never remaining long in one 
 place. With a northeast wind it shifts to the south, with a 
 southeast wind to the north, while an easterly storm some 
 times closes it entirely. The river is so shallow that it may 
 be forded at low tide, whence the name Barra Secca. At 
 the time of my visit, in 1865, there were exposed in the banks 
 of the stream at the barra strata of sandstone which were laid 
 bare by the washing away of the beach-sands. These sand 
 stones were formed, below high-water mark, by the cement 
 ing together of the sands of the lower part of the beach by 
 the lime of shells, &c. They preserved the characteristic 
 beach structure, and were full of shells ; but of these sand 
 stones I shall have more to say further on. 
 
 The lands behind the beach at Barra Secca are flat and 
 sandy, and, though cultivated to a very slight extent, are of 
 no real value. The lower grounds are damp, furnished with 
 a soil, and are largely covered with a rather luxuriant forest- 
 growth, back of which is the picturesque Lagoa Tapada, a 
 large, irregular, shallow sheet of clear water, margined by 
 grass-covered meadows and forest. Its waters drain off 
 northward into the Mariricu. 
 
 On the beaches between the Rio Doce and the Sao 
 Matheos the traveller sees, at frequent intervals, the shells 
 
108 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and skeletons of sea-turtles, and at certain times of the 
 year there is no sight more common than that of a flock of 
 Urubus feeding upon the decaying carcass of a turtle, re 
 cently killed by some hunter for the sake of its flesh, fat, or 
 ovarian eggs. Most abundant of the four species which 
 occur on the coast is the Loggerhead Turtle, Thalasso- 
 chelys cauana Fitz. 
 
 This species is very common on the Brazilian coast. 
 One may frequently see it floating lazily about on the sur 
 face of the sea, inside the Bay of Rio, as well as outside the 
 bar, but there are regions where it is especially abundant, 
 and to certain beaches it resorts in great numbers, at par 
 ticular seasons of the year, to deposit its eggs. Perhaps one 
 of the most noteworthy of these is the beach between the 
 Doce and Sao Matheos. In the month of November, 1865, 
 I found the turtles laying in the vicinity of Barra Secca, and 
 Mr. Copeland and I made a short stay at the place to capture 
 some specimens for the Museum of Comparative Zoology. 
 In the daytime the turtles remain out at sea, but in the night 
 the females come on shore to lay their eggs. According to 
 the statements of the fishermen, as well as to my own obser 
 vations, they come in shore when the tide is low, and having 
 gained the beach, creep to the upper part beyond high-tide 
 level before they make their nests. 
 
 I have repeatedly watched the movements of these animals. 
 They walk by means of their great flippers, assisted by their 
 short hind legs, with a very slow, hitching motion, the body 
 dragging on the sand. The flippers and feet make two irreg 
 ular grooves in the sand, three or four feet apart, which look 
 as if a great wagon with cogged wheels had been driven over 
 the beach. These tracks are so prominent that one may see 
 them even on a dark night, and they serve to tell the hun- 
 
PROVINCE OF ESPIEITO SANTO. 109 
 
 ter whether the turtle has returned to the sea or not, and to 
 guide him to her, or to her nest. While laboriously working 
 her way up the beach, the turtle keeps her head stretched 
 out, and from time to time snuffs and sighs as if fatigued, 
 and now and then she rests. So intent is the animal upon 
 the accomplishment of her mission to land, that one may 
 mount on her back without alarming her ; but if too much 
 disturbed, she hastily turns about and makes her way as fast 
 as possible to the sea. Arrived at the top of the beach, 
 sometimes just above high-water mark, sometimes a few feet 
 higher up among the sand-hills, at others, even on the land 
 ward side of the sand-ridge, she stops and prepares to make 
 her nest. This she accomplishes by means of her flat hind 
 feet, and after this manner : She digs up the sand with one 
 foot, and throws it to one side. Then she uses the other 
 foot in the same manner, working with one foot after the 
 other, alternately, like a machine, as Neu Wied has re 
 marked.* The sand which would be likely to fall back 
 again is moved out of the way by the foot, which moves 
 forward before it is thrust down to deepen the hole, and 
 scrapes the sand to one side. The whole is a very slow 
 operation, occupying several minutes. During this time the 
 animal remains with her head stretched forward, with very 
 little motion of the body, and occasionally giving a hiss or 
 breathing heavily. In this way I have seen a perpendicular 
 hole dug in the sand a foot and a half or two feet deep, 
 and a foot in diameter. The animal then remains quiet, 
 and the deposition of the eggs soon begins. While one 
 individual was laying, I caught the eggs in my hand, as they 
 fell. They were laid two by two, or one by one, at an inter- 
 
 * Neu Wied has in his Journey a plate of a turtle laying. The figure of the 
 turtle is very inaccurate, and the eggs are represented as of different sizes. 
 
110 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 val of about half a minute, falling in a heap into the hole, to 
 the number, if I remember rightly, of 143. A nest almost 
 always contains over 100 eggs, and usually from 120 to 150. 
 When the laying was going on, the animal appeared per 
 fectly unconscious of the presence of persons about her. 
 So soon as it was accomplished, she rested a moment, and 
 then, with her hind feet, scraped the sand back into the hole 
 until it was full. After this the body was raised a little, and 
 sand was scraped underneath it. The body then descended, 
 packing the sand tightly, and this operation was repeated 
 several times. It was an exceedingly interesting sight to 
 see this stupid reptile performing so strange an act, one to 
 which she seemed to be prompted by something more than 
 a blind instinct. The whole operation finished, she turned 
 round and started for the sea, when we captured her, by 
 taking hold of the shell behind, and upsetting her on her 
 back, in which position it is impossible for a sea-turtle to 
 turn over. In upsetting one of these animals two persons 
 are usually required. As soon as the animal is alarmed, 
 she thrusts her fore paddles into the sand and throws it 
 behind her, so that, if one does not take the precaution 
 to close his eyes, he is likely to be blinded. 
 
 The eggs are rather larger than those of a hen, round, 
 and covered with a tough, white, parchment-like skin. The 
 albuminous portion is clear, and does not become hard in 
 boiling. The yolk is very large, deep yellow, and is the 
 only part eaten. I found these eggs very palatable, though 
 they tasted somewhat fishy. The Brazilians are very fond 
 of them, and, while the turtles are laying in November, De 
 cember, January, resort to the shore to collect them, filling 
 in a short time the great panniers of their mules. They 
 discover the nests by thrusting a long stick into the sand, 
 
PROVINCE OF ESPIRITO SANTO. Ill 
 
 and then dig out the eggs with their hands. The nests 
 are, I am told, sometimes despoiled by the Teiii lizard- 
 (Teius monitor Merr.). The Brazilians cook the eggs in 
 various ways, but generally by boiling. The yolk is usually 
 mixed with sugar and farinha. They are not so satisfying 
 as the egg of the hen, and one may eat a dozen at a meal. 
 They make very good omelets. The eggs which remain 
 unmolested are hatched by the heat of the sun, in about 
 twenty days, it is said, when the young turtles dig their 
 way to the surface and escape to the sea, where the most 
 of them fall a prey to sharks and other fish. The eggs 
 are laid at such a depth below the surface as to insure a 
 uniform temperature, for on the surface the sands are 
 exposed during the day to a fierce, burning heat, while in 
 the night they cool down very much, the diurnal variation 
 of temperature amounting to thirty degrees or more. At 
 the depth of a foot and a half, or two feet, the temperature 
 is quite uniform, and would stand in December or January 
 at about eighty degrees. I observed that the dogs at Barra 
 Secca, when it was cold at night, scraped away the upper 
 layer of sand near my tent, and lay down in the warm 
 sand below, and I learned a useful lesson from them, 
 so that, when benighted and unsheltered, I could find a 
 warm bed on the sea-shore. The strength of the turtle 
 is enormous. A stout stick placed in its bill was crushed 
 like a straw. The flesh is dark red and coarse. We had 
 it cooked in various ways, but did not find it very good, 
 though we were obliged to use it for food. The animal is 
 usually very fat, the fat being of a greenish-yellow color, 
 and oil is tried from it for various purposes by the people 
 of the coast. The ovaries were always full of eggs in all 
 stages of development. These ovarian eggs are much es- 
 
112 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 teemed by the inhabitants, and the turtles are killed in 
 great numbers for their sake. All along the beaches one 
 may see carcasses from which the plastron has been removed 
 to allow the ovaries to be got at. Mr. Copeland and I 
 prepared some six specimens of the species of turtle above 
 described in the shape of skeletons, shells, skins, and alco 
 holic preparations of head and flippers, for the Museum of 
 Comparative Zoology. 
 
 The food of this turtle appears to consist of fish, shell 
 fish, sea-urchins, &c. I have observed growing on the cara 
 paces of these animals a large barnacle, Coronula, like that 
 which grows on the back of the whale on the same coast. 
 
 Prince Max zu Neu Wied, on his journey northward from 
 the Doce, saw a living sea-turtle near Barra Secca, and 
 watched her lay her eggs. He has referred this specimen 
 to the Green Turtle, Chelonia mydas, but I do not think that 
 his determination is satisfactory. The specimens which 
 Mr. Copeland and I collected have been examined by Mr. 
 J. A. Allen, of the Museum of Comparative Zoology, and 
 have all proved to be Loggerheads. I do not think I saw a 
 single mydas on the Brazilian coast ; at all events, all the 
 sea-turtles I examined were different from the green turtles 
 brought into the New York market. 
 
 Eretmochelys imbricata Fitz. (Tartaruga de Pente of the 
 Brazilians.) This species occurs quite abundantly on the 
 Brazilian coast. It is taken at the Abrolhos together with 
 the Cauana, and its thick scales are used to some extent 
 for the manufacture of ornaments. One may occasionally 
 find it in the markets of Pernambuco and Bahia, and at the 
 latter city my friend, Dr. Anto. de Lacerda, had one which 
 he kept captive in the pool of a fountain in his garden. 
 
 Sphargis coriacea Gray. (Leather-back Turtle.) Neu 
 
PROVINCE OF ESPIRITO SANTO. 113 
 
 Wied mentions hearing an immense sea-turtle with a leath 
 ery skin spoken of, and which he supposed might be Testudo 
 coriacea Linn. ; but he did not see a specimen. The fisher 
 men described to me a similar turtle, and their description 
 would tend to confirm Neu Wied s opinion. I did not see a 
 specimen. 
 
 Chelonia mydas Schu. (Green Turtle.) According to 
 Neu Wied and others, this species occurs on the Brazilian 
 coast, but I have never seen it. 
 
 The number of sea-turtles destroyed every year on the 
 beaches between the Doce and Sao Matheos is very large, 
 and the destruction, if persisted in, must ultimately drive 
 them from the coast. 
 
 Gerber s map of that part of Espirito Santo between Santa 
 Cruz and the Sao Matheos is very inaccurate, because the ma 
 terials from which his map was compiled are inaccurate, and 
 little trust is to be placed in the position assigned to the 
 lakes and minor streams. For this region the chart of 
 Mouchez, though giving in considerable detail the hydrog 
 raphy and topography of the coast lands, is worse than 
 nothing, for he seems to have disregarded all the previously 
 published maps of the province. 
 
 The beach, backed by a high ridge of sand, runs north 
 ward, for perhaps two miles, to a place called As Pedras, 
 where the same calcareous sandstone, as seen at Barra Secca, 
 occurs, exposed for some distance along the beach over a 
 considerable area. The beds dip seaward, and appear to 
 be very thick at low-water mark ; but they thin out before 
 reaching high-water level. The arrangement of the mate 
 rials in this sandstone is precisely like that of the beach, and 
 this formation is only the lower part of a beach ridge which 
 has been cemented by the lime of shells, &c., and then laid 
 
114 GEOLOGY AND PHYSICAL GEOGKAPHY. 
 
 bare. The sandstone is exceedingly hard. Two sets of 
 joints one parallel with the beach line, the other at right 
 angles to it divide it into great blocks, which, in those spots 
 where they have been undermined by the surf, lie upset and 
 in confusion along the edge of the reef. Along these joints 
 the rock is often harder than between them, so that when 
 the surface of a block is exposed to the action of the sea, the 
 edges wear less rapidly than the middle, and the cracks 
 seen on a worn surface are oftentimes bordered by nar 
 row ridges. These have evidently resulted from the pene 
 tration into the mass of water carrying lime in solution, 
 after the joints had been formed, and the farther solidifi 
 cation of the rock on each side of the joint. The waves 
 beat terribly against this reef, and it is badly broken up. 
 Continuing northward along the reef, which sometimes 
 forms a smooth pavement to the beach, at others obstructs 
 it by broken masses of stone, we come to a rocky point 
 where the shore bends in abruptly, and makes a little bay. 
 The beach ridge ends as abruptly, and the shores of the 
 cove have no ridge, but the reef rock continues straight on, 
 and forms a line of rock, stretching out some distance 
 across the bay, while at low tide one may see that the 
 reef, only just covered by water, is continued across the 
 bay to a point where the beach, furnished with a sand- 
 ridge, follows once more its normal northward trend. 
 This locality is very interesting, because it shows us 
 that the sandstone is confined to the beach, and that, 
 when the beach ridge and the sands behind have been 
 swept away, by a storm or otherwise, the bared reef may 
 stretch, like a wall or breakwater, across the coast inden 
 tation thus formed. 
 
 Still farther northward we find the beach bordered by a 
 
PROVINCE OF ESPIRITO SANTO. 115 
 
 fringing reef of this sort, a sort of irregular flagging to 
 its edge. Riding over these one sees the ordinary shells of 
 the coast, with their colors still fresh, imbedded in the rock. 
 There is a species of marine worm which constructs on these 
 rocks, near low tide, great, rounded, flattened agglomera 
 tions of sand-tubes, which are sometimes more than a foot 
 in height, and resemble immense sponges. The sand is 
 very compactly cemented together, as might be inferred from 
 their withstanding the continuous pounding of the Atlantic 
 surf. Sometimes these masses are broken across, when 
 they appear like fragments of some large Astrsean. At one 
 locality on the shore, opposite these rocks, there is a shallow 
 lagoon of considerable extent, called Mariricu, which lies 
 just behind the beach, and is separated from the sea only 
 by the beach ridge, which latter rises to a height of 
 but five feet above high-water mark. I found the surface 
 of this lake to be, on the 18th November, 1865, about five 
 feet and a half above the level of the sea at low water.* 
 Since it is separated from the sea by only a narrow ridge 
 of sand, there is a constant soaking of its waters through 
 the beach at low tide. It has seemed to me probable that 
 the waters of the lagOa percolating through the beach might 
 have something to do with the solidification of the sands. 
 The narrow beach which separates the lake from the sea 
 is the only dry ground between the swamps and the ocean, 
 and it is the road taken by the wild animals passing north 
 and south along the shore. At As Azeites, a little settle 
 ment near the shore, and about three leagues south of Sao 
 Matheos, there are several beach ridges, one inside of the 
 other. Inside of these, draining the lagoas and swamps of 
 the south, flows a little black-water stream called the Rio 
 
 * The tide here rises about six feet six inches. 
 
116 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Mariricu.* At As Azeites the swampy region is several miles 
 wide, and a considerable part of it is overflowed ; but the 
 lagoons and streams are so masked by floating masses of 
 water-plants (balsas ), and a dense swamp vegetation, that 
 the country is impassable, and consequently unknown to the 
 inhabitants of the district itself. At Azeites the Mariricu 
 is only a narrow, navigable channel leading through a 
 wide, overflowed region, and bordered by balsas of grasses, 
 arborescent arums, and trees. A narrow canal, several 
 hundred feet long, is here cut through the floating vegeta 
 tion to reach terra firma. On descending, the channel 
 frequently expands into broad lagoons, diversified by balsas, 
 and islands with clumps of trees, but the channel soon con 
 tracts, though the growth of floating grasses f on each side 
 makes it appear very much narrower than it is in reality. 
 In some places it is one hundred and fifty feet wide and ten 
 to fifteen feet deep. The lands bordering it are, for a large 
 part, only just above water, and are heavily wooded. In these 
 forests one sees the Gamelleira (Ficus*), and the Tucum palm 
 (Astrocaryum tucuma Mart.), furnishing a valuable fibre, and 
 the Inga, while on the open grounds flourishes the Cashew, 
 or Cajueira (Anacardium occidentalc) , some immense trees 
 of which I saw growing at As Azeites. These lands are 
 very excellent for the cultivation of rice. There are some 
 stretches of higher sandy grounds, which, though not fertile, 
 are more or less cultivated, producing mandioca, feijao, cot 
 ton, rice, and corn. The Mariricu empties into the Sao Ma- 
 theos a few miles above its mouth. The tide, which goes 
 
 * In all probability during heavy freshets some of the shore lagoas may 
 make outlets for themselves ; but the only barras I saw between the Doce and 
 Sao Matheos were those of Monserras and Secca. 
 
 t This is the resort of great numbers of alligators, capabaras, and of many spe 
 cies of water-birds, ducks, parras, cranes, &c., many of which breed there. 
 
PROVINCE OF ESPIRITO SANTO. 117 
 
 up the Sao Matheos some thirty-five miles, makes its influ 
 ence felt up the Mariricu for some distance. Sometimes, 
 during drouths, the salt water flows up the river and kills 
 the floating grasses, and many of the fresh-water fish die. 
 
 The Rio Sao Matheos rises in the province of Minas, 
 in the forest, south of the Colonia de Urucu ; but I have 
 no information as to the exact point, for the region of 
 its head-waters is a forest inhabited by savages and 
 quite unexplored. In a manuscript map kindly furnished 
 me by my friend, Herr Robert Schlobach, Imperial engi 
 neer of the Mucury, the Rio Sao Matheos is represented as 
 taking its rise a few miles south of Philadelphia. Its an 
 cient name was Cricare , or Quiricare". It is formed by the 
 union, at a distance of about sixty miles above its mouth, of 
 two branches called, respectively, Bra90 do Norte and Bra$o 
 do Sul. I made a horseback journey to the Fazenda do Capitao 
 Grande, distant some forty-five miles from the city of Sao 
 Matheos, and situated on the Brac,o do Norte, a few miles 
 above its junction with the Brago do Sul, and, descend 
 ing to the sea, mapped the river as far as Sao Matheos, 
 below which I was prevented from continuing my work. 
 At Capitao Grande the river, a stream some one hundred 
 feet wide, shallow and swift, flows in the bottom of a 
 valley, cut through the tertiary formation, which here has a 
 thickness, above river level, of three hundred feet, more or 
 less. On both sides of the valley the country is a plain, for 
 the most part heavily timbered, especially on the slopes. 
 At the fazenda, the soils on the slopes appear to be drift, and 
 are exceedingly fertile. The coffee on the slopes was vigor 
 ous and healthy, without blight, and it was very heavily 
 fruited with a berry of excellent aroma. I do not remember 
 having seen anywhere better coffee-trees than those at Capi- 
 
118 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tao Grande. The soil yields mandioca and the other products 
 proper to the climate in abundance. In the narrow valley 
 the climate is very hot, trying, and feverish, but the clearing 
 away of the forests will change its character. On the BraQO 
 do Norte clays are occasionally exposed in the bluffs, togeth 
 er with a kind of coarse white sandstone such as would re 
 sult from the hardening of the sandy clays of the tertiary. 
 There are but few inhabitants on this branch of the river. 
 Of the Bra^o do Sul I know nothing. The main river is a 
 respectable little stream, one hundred and fifty to two hun 
 dred feet in width above Sao Matheos, and bordered by 
 alluvial lands, which lie in the bottom of a valley cut 
 through the tertiary beds, and vary much in width. The 
 valley is bordered by bluffs, in which sandstones and clays 
 of ordinary type are occasionally seen. The tertiary slopes 
 are very fertile and largely cultivated, as are, to a consider 
 able extent, the lands on the upper plain,* so that the 
 country, with its numerous fazendas and cocoa-palm trees, 
 wears a very pleasant aspect. 
 
 The cocoa palm (Cocos nucifera) is, according to Wal- 
 
 * The soil of the upper plain, when uncleared, usually bears a heavy forest 
 growth. It is more or less sandy and clayey, and I should suppose would be 
 difficult to work ; but I am told that such is not the case. It is frequently 
 covered with a thin layer of loose sand on the surface, owing to the washing 
 away of the clayey portion by the surface water. It is especially fitted for the 
 cultivation of mandioca, which flourishes well in a sandy soil, as well as for 
 cotton, which also does well in a soil of that kind. Cane is planted to some 
 extent. When cleared, these lands make excellent grazing grounds, but grass 
 must be planted. On the Sertao flourish the Nmjd, Timbitrg, and Muri palms. 
 Northwest of Sao Matheos, on the north side of the river, are some extensive 
 barren and swampy plains covered by a vegetation composed of shrubs, 
 among which I observed a species of Vaccinium in fruit. The soil of the river- 
 borders is Massape, and very fertile ; but these are generally very low and liable 
 to be flooded. Near Sao Matheos are extensive swamps, which sometimes breed 
 very malignant fevers. 
 
PKOVINCE OF ESPIRITO SANTO. 119 
 
 lace and other good authorities, not a native of America, and 
 the early explorers of Brazil do not speak of it. Its home 
 appears to have been in the East Indies, but it has been in 
 troduced into America, probably by natural means, the im 
 pervious shell and thick husk fitting the fruit to bear long 
 transportation by ocean currents. It is now found every 
 where on tropical coasts, but in the East Indies it is more 
 largely cultivated than elsewhere, sometimes even forming 
 forests. It is a very valuable tree, furnishing food, oil, fibre 
 for cordage, arrack, <fec., &c.* 
 
 Yon Tschudi says that Villa Vi9osa is the southern limit 
 of this palm in Brazil. This is not quite correct. South of 
 Sao Matheos these trees are but rarely seen, but they, how 
 ever, grow at Rio de Janeiro, and great numbers may be 
 seen on the shores of Paqueta, while there are a few speci 
 mens on the islands of Cobras and Villeganhao. At the 
 Barra do Sao Matheos they grow very well, and at the 
 Fazenda do Campo Redondo, at a distance of several miles 
 from the sea, there is a fine grove on the edge of a bluff. 
 This palm, as has been frequently remarked, appears to 
 flourish best on the sands of the sea-shore, and it is very 
 rarely seen at any great distance inland. 
 
 A specimen was seen by Burton on the Sao Francisco at 
 Brejo do Salgado, three hundred and fifty miles from the 
 sea ; and, according to the same observer, the cocoa palm 
 occurs in occasional patches thence down the river. Burton 
 speaks of a large grove twenty-eight miles to the southwest 
 of Joazeiro, and he suggests that the saline character of the 
 soil may make up for the want of sea air. 
 
 * See Hamilton, Description of Hindostan, Vol. II. p. 210. Meyen, Botan 
 ical Geography, p. 331. Transactions of Royal Asiatic Society of Great Britain, 
 Vol. I. p. 546. Bennett s Travels in New South Wales, &c., Vol. II, Appen 
 dix, p. 295. 
 
120 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 FAZEXDA OF CAMPO REDOXDO, SAO MATHEOS. 
 
 Humboldt found this palm growing in the interior of 
 Venezuela, and the Duke Paul von Wiirtemburg reports it 
 as flourishing, at a distance from the sea, in the island 
 of Cuba. It occurs, also, sometimes in the interior of 
 India. 
 
 In Brazil I have seen the cocoa palm as far south as lati 
 tude 23 at Rio. It is grown abundantly in the provinces of 
 Bahia, Sergipe, Alagoas, Pernambuco, and along the coast 
 northward, but near the equator it is rarely seen. The 
 northern limit appears to lie within latitude 28 north of the 
 equator. May. en says, on the authority of Humboldt, that 
 in Venezuela it ascends to the height of 700 toises above 
 the plain, or in the vicinity of 4,500 feet above the sea. 
 
 The products of the Sao Matheos region are principally 
 mandioca meal (/arw/ia), together with some sugar, cacao, 
 beans, &c., <fec. The cacao is principally planted on the lower 
 grounds, where it yields well. The influence of the tide is felt 
 at a distance of some twenty-five miles, or more, above the 
 mouth of the river, and navigation for small vessels ends a few 
 
PROVINCE OF ESPIRITO SANTO. 121 
 
 miles above the town. The height of the bluffs decreases on 
 nearing the coast, and, at the city, they are not more than 
 eighty to one hundred feet. Sao Matheos, like Nova Almeida 
 and other Brazilian towns, is built partly on the edge of the 
 bluff, partly at its foot, by the river-side, at the point where 
 the bluffs leave the river and run southward to the Doce. 
 It has about two thousand inhabitants, and is a place of 
 some considerable importance, being reached by the coast 
 ing steamers and little schooners. The trade is princi 
 pally in farinha, feijao, &c. On the opposite side of the 
 river the bluffs run eastward for a few miles, and then 
 trend off northward to Itahunas, decreasing in height as 
 they near the sea. This decrease in height, I am inclined 
 to ascribe, in part, to the slope of the old sea-bottom, on 
 which the tertiary beds were deposited. 
 
 Below the town the river grows wider, shallower, and is 
 obstructed by sand-banks. The banks are low, with only 
 occasional plantations, and are, for the most part, covered 
 by a dense forest ; but, lying between the tertiary lands and 
 the sea on the northern side, at least, are quite extensive 
 sandy plains with their accompanying sparse vegetation. 
 The river-banks are generally muddy, and the aninga and 
 other brackish-water swamp plants grow abundantly on 
 them. Salt-water crabs and fish ascend to the town. The 
 city of Sao Matheos cannot be more than seven or eight 
 miles in a direct line from the sea, and is very incorrectly 
 located on the maps. According to my observation, it lies 
 about west-southwest of the barra. The river, descending 
 from the city, makes a large bend southward, receiving the 
 Mariricu ; it then runs north a few degrees east, and, just 
 above the barra bends round and enters the sea from the 
 northwest. The maps show a Rio Sao Domingos entering 
 
122 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the Sao Matheos, just above the villa, but I find no note of 
 it in my journal. The Villa da Barra do Sao Matheos is 
 situated on a ridge of sand, only a stone s throw from the 
 sea, but it is nearly two miles from the mouth, because the 
 river flows southward, behind the beach ridge for that dis 
 tance, before escaping into the sea. On both sides of the 
 river, but particularly on the southern side, there are exten 
 sive mangrove swamps, which furnish very interesting col 
 lecting grounds for the naturalist, for they are especially 
 rich in crustaceans. Among the mangroves two genera are 
 represented, Laguncularia and Avicennia. The common 
 red mangrove of the Brazilians is the Lag-uncularia racemosa 
 Gaert., of which the wood is used for burning, and the leaf 
 and bark for tanning. The second species is the Avicennia 
 tomentosa Linne", used for the same purposes. 
 
 One league, or thereabouts, north of the Barra do Sao 
 Matheos, is the mouth of the Rio Itahunas, or Guaxindiba. 
 This little stream, a black-water river, as has already been 
 noted by Max. zu Neu-Wied,* rises in the Sertao, and reaches 
 the coast at a point some three leagues north of the mouth ; 
 but a beach ridge, only a few hundred feet in width, prevents 
 its reaching the sea, and causes it to bend abruptly to the 
 southward, so that it flows along, through the low grounds, 
 
 * Prinz Max. zu Neu Wied, Reise nach Brasilien, Vol. I. p. 226, says that the 
 Peixeboi, or manati, Manatus americanus, occurs in a lai-ge grass-grown lake 
 south of the Sao Matheos and communicating with that river, and that it used 
 to be captured by the inhabitants for the sake of its blubber, flesh, and ear-bones ; 
 and in the second volume of his Beitrage aur Naturgeschichte von Brasilien, page 
 602, he states that it occurred plentifully in the river and lake aforesaid, or in 
 the vicinity of the Quartel Juparanaa, and was formerly frequently taken by 
 the inhabitants. He adds, however, that he had never seen a specimen. 
 While at Sao Matheos, and during a stay of three months in that part of 
 the coast, I never heard the animal spoken of, and I do not believe that it is 
 now to be found there. (See note on page 75.) 
 
PROVINCE OF ESPIRITO SANTO. 123 
 
 just behind the beach ridge, for two or three leagues. It 
 is rather narrow, but very deep, and vessels of consider 
 able size enter it. It is navigable for some distance into 
 the interior, and furnishes an excellent water highway. 
 Near its mouth a riacho * enters it from the southward, 
 and this affords navigation nearly to the Barra do Sao Ma- 
 theos. A canal is being opened to join this riacho with the 
 Sao Matheos. The low lands between the bluffs and the 
 coast are for the most part timbered ; but they furnish some 
 excellent pasturage. On the Itahunas, a strip of sand mar 
 gins the bluffs and appears to be the continuation of the sand 
 plain on the Sao Matheos. The bluffs bordering the plain 
 are only thirty or forty feet in height, but they grow higher 
 as we go westward, and, at a distance of two leagues from 
 Itahunas, they are at least one hundred feet high. They 
 have, as elsewhere, steep slopes, and, in part at least, are 
 covered by a very fertile soil. On the Fazenda of Senhor 
 Olindo Gomes dos Santos I saw most excellent crops, espe 
 cially of mandioca and coffee, which were almost, if not 
 quite, equal to anything on the Sao Matheos. 
 
 Off the shore, in front of the town of Itahunas, are some 
 ledges of rock washed by the waves. I could not examine 
 them, but I thought them to be sandstones. The bluffs 
 sweep round to the shore, just to the north of a little riacho, 
 called Doce, if I mistake not, and a short distance to the 
 south of the first point north of Itahunas ; they form along 
 the shore, for a distance of several miles, a line of low 
 cliffs, which, from their whiteness, have received the name 
 of Os LeiiQoes. These cliffs are nowhere more than thirty 
 to forty feet in height, this being the thickness of the forma 
 tion exposed above the sea. The most prominent feature 
 
 * This term is applied to small estuaries. It means literally a small river. 
 
124 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 in this line of cliffs, is a hard gray grit with an argillaceous 
 cement, and in some places twenty feet in thickness. This 
 rock is composed of materials precisely like those of the 
 sandy clays of Lagcta Juparanaa, and elsewhere ; but the 
 rock is so compact and hard as to be used as a building- 
 stone at Itahunas, although I believe that the stone actually 
 used is derived from quarries near the village. Associated 
 with this rock are white and red clays, and beds of the 
 common, coarse, red, lava-like sandstone.* The cliffs end 
 at, or in the immediate vicinity of, the Riacho das Ostras, 
 a little, black-water stream, of whose course I know nothing, 
 but the bluffs continue along the shore much further. Riacho 
 Novo is another black-water stream, which, before emptying 
 into the sea, runs for several miles in a deep channel just be 
 hind the beach ridge. The tide enters this channel, and it is 
 navigable for canoes for a considerable distance. The whole 
 country between Itahunas and the Mucury is wooded, the 
 lower and wet grounds very luxuriantly, the higher and drier 
 less so, the latter often supporting only a thick growth of 
 small trees and bushes. Dunes of sand occur on the shore 
 near the Mucury, and in one locality there are exposed on 
 the sand-beach the stumps of trees, rooted in their soil and 
 upright, something which points to an encroachment on the 
 shore by the sea. I have observed dead trees, apparently 
 mangroves, standing in the water off the shore below Cara- 
 vellas. Has there been a recent depression of this part of 
 the coast ? 
 
 * I was delayed one morning at a little settlement just south of the Riacho das 
 Ostras, and I undertook a careful examination of the beds for fossils, but I saw 
 not the slightest trace of organic remains in them. I am sorry that I was un 
 able to make notes on the arrangement of the materials in the cliff; but the 
 place was like a furnace, and I was driven away by the sun, blinded and ex 
 hausted. 
 
PROVINCE OF MINAS GEBAES. 125 
 
 CHAPTER III. 
 
 PROVINCE OF MINAS GERAES. THE MUCURY AND JEQUI- 
 T1NHONHA BASINS. 
 
 The Basin of the Mucury. Porto Alcgre. Description of the River below 
 Santa Clara. Luxuriance of Forest Vegetation. Santa Clara. Minas 
 Geraes a Land-locked Province. Want of Roads. The Philadelphia Road 
 and the Mucury Colonies. Difference in Topography and Soils between the 
 Tertiary and Gneiss Lands west of Santa Clara. Urucu, its Dutch Colony, 
 Soils, Climate, &c. Philadelphia and its German Colonies. Great Fer 
 tility of the Mucury Basin. Character of Country between Philadelphia 
 and the Head -waters of the Mucury. The Basin of the Jequitinhonha. 
 The Rio Pardo. General Geological Structure of the Jequitinhonha-Pardo 
 Basin. The Head-waters of the Setubal, their Geological Features and Ca- 
 tinga Forests. Geological Excursion from the Fazenda de Santa Barbara to 
 Alto dos Bois. Difficulty of geologizing in Brazil. The Brazilian Cam 
 pos. The Chapadas between Itinga and Calhao. The great Calhao- 
 Arassuahy Valley. Magnificent View over the Valley from the Chapada at 
 Agua da Nova. Calhao and the Geology of its Vicinity. Description of 
 the Country between Calhao and Sucuriu. The Chapadas. Minas Novas, 
 its Geology, Gold-Mines, &c. Occurrence of Gold in Drift. Gold-Mines 
 of the Arraial da Chapada ; their former Richness ; not yet worked out. 
 Decomposition of Clay Slates in the Minas Novas Region. The Rio Aras- 
 suahy. The Rio Jequitinhonha from its Confluence with the Arassuahy to 
 the Sea described ; its Geology, Vegetation, Commerce. The Salto Grande. 
 
 THE Rio Mucuiy takes its rise in the province of Minas, 
 about 150 miles west of Yilla YiQOsa, among the high lands 
 which form the water-shed bounding the basin of the Jequi 
 tinhonha on the east. Its course for the first seventy-five 
 miles, curves excepted, is approximately east-northeast, when 
 it meets the Rio Preto, a stream rising in the same water-shed 
 
126 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 some forty miles to the north-northeast. This last river 
 flows with a course almost parallel to the Mucuiy up to 
 about the same meridian, when it receives the little Rio das 
 Americanas coming from the north. It then bends abruptly 
 to the south, to join the Mucury, the combined waters of the 
 two streams flowing still southward until they reach the Rio 
 Todos os Santos, a stream rising at a point south of the 
 Mucury, and forty miles south of the source of the Rio Pre- 
 to, and flowing also parallel with the Mucury. The Mucury 
 soon bends gradually round to the east, and, making sev 
 eral broad curves, runs with a general southeast direction 
 to Santa Clara, on the boundary between the provinces of 
 Minas, Bahia, and Espirito Santo, a distance of thirty miles, 
 in a straight line, but much more, following the course of 
 the river. The Rio Urucu is another river flowing in a val 
 ley parallel with the Todos os Santos and Upper Mucury, but 
 emptying into the Mucury proper on the right bank, some 
 fifteen miles in a straight line below the mouth of the 
 Todos os Santos. On the opposite or northern side the 
 Mucury receives, about eight or ten miles farther down, the 
 Rio Pampao, which comes from the north, and has appar 
 ently a course of about sixty miles ; but its upper waters 
 have never been explored.* Above Santa Clara the Mucury 
 and its tributaries are swift and obstructed by rapids, but 
 from Santa Clara the main river runs with a very tortuous 
 course, as a rio dCareia, to the sea, into which it empties in 
 latitude 18 6 S., and ten and a half miles south of the par 
 allel of Santa Clara. Its waters come principally from the 
 province of Minas, where it drains an irregular triangular 
 
 * Some of the lands of the Mucury company were situated on the Pamp3o. 
 The only one who has visited them is my friend Mr. George Schieber, one of 
 the surveying corps of the Mucury. 
 
PROVINCE OF MINAS GERAES. 127 
 
 area, bounded as follows : by a line running due north along 
 the Serra dos Aymore s, eighty miles, another line, 135 miles 
 long, running a little east of northeast, and another on the 
 south about 110 miles long, and running east-west along the 
 water-shed dividing the basin of the Mucury from those of 
 the Doce and Sao Matheos. This triangle is a right-angled 
 one. From its southeast angle the area drained forms an 
 irregular strip some six or more miles wide, along the mid 
 dle of which flows the river. In the study of this river sys 
 tem, several points strike one as interesting. The parallel 
 ism of the Rios Preto, Mucury, Todos os Santos, and Urucu 
 shows that they flow in parallel valleys, which are evidently 
 determined by the trend of the foldings into which the gneiss 
 is thrown. The coincidence in direction between the courses 
 of the Rio das Americanas and that of the Rio Preto above 
 the Mucury, and the latter to its confluence with the Todos os 
 Santos, points to a valley running north-south, into which 
 the rivers Preto, Todos os Santos, and Mucury empty as side 
 tributaries. The Pampao flows in a similar valley. It is 
 interesting to compare the basin of the Mucury with that of 
 the Doce. In each, the greater part of the region drained 
 lies west of the coast cordilheira, while east of the cordilheira 
 the area dwindles down to a narrow strip bordering the 
 river on each side. Between these two rivers are intercal 
 ated the Sao Matheos, Itahunas, &c. 
 
 With these introductory remarks on the hydrography of 
 the Mucury basin I propose to give a somewhat detailed 
 description of the river basin in ascending the river to Santa 
 Clara, and then traversing the basin thence westward to the 
 head of the Mucury Pequeno, along the line of the Santa 
 Clara, Philadelphia, and Minas road. The Rio Mucury, con 
 trary to the general rule, enters the sea obliquely from the 
 
128 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 south. Its mouth is narrow and difficult to enter, being 
 much less practicable than the Sao Matheos. On the left 
 bank at the mouth, built on a sand-bank, and surrounded 
 by mangrove swamps, is the miserable little village to which 
 the ridiculous misnomer of Porto Alegre has been applied. 
 The place is of importance only as the port of the Mucury 
 district, which exports coffee, cotton, rosewood, &c. From 
 Porto Alegre large quantities of salt are sent into the inte 
 rior, where it finds its way into the very heart of the prov 
 ince of Min as Geraes. Dry goods are also imported through 
 the same channel. The town is one of the most wretched I 
 saw in Brazil. Its inhabitants are principally of Tupi origin. 
 From Porto Alegre to Santa Clara, a distance of forty-five 
 miles in a straight line, but at least fifty-five by the river, 
 the stream is rather shallow, very tortuous, narrow, and 
 affording navigation for pranchas, canoes, and a very small 
 steamer ; but the water is, for a considerable part of the 
 year, so shallow, and the river is so obstructed by sand 
 banks, that it is navigable constantly only for canoes.* Just 
 above Porto Alegre the tertiary bluffs appear, and sand 
 stones and clays are exposed in them. At first these bluffs 
 are not very high, in some places measuring only from 
 eighty to one hundred feet ; but, ascending the river, their 
 elevation increases, and at Santa Claraf they are some 330 
 feet above sea level. At Santa Clara the whole thickness of 
 the formation is not displayed in these bluffs ; for, back from 
 them the tertiary lands reach, in some places, an altitude of 
 
 * In the latter part of January, 1866, Mr. Copeland and I ascended the 
 river in the little steamer. The water was so shallow that we were constantly 
 running aground. A few leagues below Santa Clara we stuck fast, and the 
 rest of the distance we had to make in the prancha we had been towing. 
 
 t Santa Clara itself is 327 palmos above sea level, which would make the 
 level of the river below the rapids only a few palmos lower. 
 
PROVINCE OF MINAS GERAES. 129 
 
 360 feet above sea level, according to actual measurement. 
 The river valley is very narrow, and the sides have a steep 
 slope. The alluvial lands are small in extent. Both they 
 and the tertiary lands are covered by a very heavy and lux 
 uriant forest, and the scenery on the river is of surpassing 
 beauty ; for here, as on the Doce, the trees crowd down to 
 the water s edge, forming a dense wall of verdure. A host 
 of species of beautiful-leaved and bright-flowing climbing 
 plants hang a dense curtain from tree to tree, and sometimes 
 depend in folds from the outstretched branches, like the dra 
 pery from the arm of an antique. The gneiss first makes 
 its appearance below the tertiary rocks at a place called 
 Dous Irmaos, some eight leagues, more or less, below Santa 
 Clara. It has a northward dip. At Santa Clara navigation 
 is made impossible by a series of rapids, and thence into 
 the province of Minas the Mucury is rapid, and has many 
 falls. At Santa Clara, which is only a collection of a few 
 dwellings and warehouses, built on narrow alluvial flats 
 on the right bank of the river, the valley is very narrow, 
 and has steep banks. The rocks in the river are gneiss,* 
 which is much veined with granite. This locality is very 
 unhealthy, owing to the narrowness of the valley, and the 
 great heat of the day, which often gives way to damp 
 fogs by night, to the bad character of the river water, and 
 to swamps in the vicinity on the top of the chapada.f 
 
 * The Diccimario Geografico says that iron ore exists within the district of 
 Porto Alegre, but does not indicate the locality. Von Tschudi, Reisen durch 
 Sud-Amer., Vol. II. p. 338, says that he has found chrysolites in the river-sand 
 at the Barra. 
 
 t I was attacked by fever here, and only escaped by removing to the high 
 grounds of Minas. I owe a deep debt of gratitude to Signora Gazzinelli, who 
 took a mother s care of me, and also to my faithful and generous companion. 
 Mr. Schieber, whose kindness I never can forget. 
 
 6* I 
 
130 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Any one glancing at a map of Brazil will see that the rich 
 and populous province of Minas Geraes is land-locked, and 
 separated from the sea by serras and forests. The Serra da 
 Mantiqueira and the Serra do Mar skirt it on the south, and 
 on the east the coast mountains, collectively known under the 
 name of the Serra dos Aymords, clothed with forest, form 
 its eastern boundary line. None of its rivers are navigable 
 to the sea, though some of them are, for scores of miles, 
 navigable in their upper courses ; but all of them are ob 
 structed by heavy falls or rapids in their descent from the 
 plateau to the coast plains. Many of them, as is the case 
 with the Doce, Mucury, Jequitinhonha, and Sao Francisco, 
 are navigable in their lower courses, in some cases, even up 
 to the confines of the province. From Rio a railroad to the 
 foot of the Serra da Estrella, with a magnificent wagon-road, 
 which, crossing this Serra and the Serra da Mantiqueira, con 
 nects with the Barbacena district, about 152 miles in a direct 
 line from Rio. An excellent railroad crosses the Serra do 
 Mar to the northwest of Rio, and enters the valley of the 
 Parahyba, down which it extends many miles. It is to be 
 extended northward into the province of Minas. But the 
 greater part of Minas is destitute of wagon-roads, and the 
 traffic is almost wholly carried on on the backs of mules. The 
 coast forest- and mountain-belt bounding the province is al 
 most entirely uninhabited and impassable. A very small 
 quantity of salt and other articles of commerce finds its way 
 in canoes up the river Doce, as already stated, and a larger 
 quantity by the same means enters the province by the Je 
 quitinhonha ; but commerce with the sea-coast is carried on 
 with great difficulty and at much expense. Thus, a bag of 
 salt that costs two milreis at the sea-coast is worth eight or 
 even eleven milreis by the time it has reached the interior 
 
PROVINCE OF MINAS GERAES. 131 
 
 of Minas. The Senator Theophilo Benedicto Ottoni,* some 
 twenty-five years ago, conceived the project of opening a 
 good wagon-road from Santa Clara to Minas Novas, through 
 the broad forest region of the Mucury, and of colonizing that 
 region. A company was organized for the purpose of accom 
 plishing this object. Through agents in Europe, a consider 
 able number of German, French, and Swiss colonists were 
 secured, and two colonies were founded, one on the Rio Uru- 
 cu, the other at Philadelphia, on the Todos os Santos, the 
 Mucury colonies being founded in the year 1858. An ex 
 cellent wagon-road, now out of repair, was constructed from 
 Santa Clara to Philadelphia, and a mule-path was laid out to 
 Minas Novas ; but the colonists appear to have been, to a 
 very large extent, of very poor quality. Through the mis 
 representation of the agents of the company in Europe, the 
 colonists were led to expect to find themselves, on their ar 
 rival, put into the possession of a house and cultivated farm. 
 It was a bitter disappointment to them to be sent into the 
 virgin forest. Nevertheless, extensive clearings were made, 
 and the villages of Urucu and Philadelphia were built ; but 
 political opposition from the enemies of Ottoni was added to 
 the difficulties the colonists had to contend with. The com 
 pany failed ; the colonists, disappointed in their hopes, de 
 serted by wholesale,! and to-day the Mucury is dragging out 
 a miserable existence, Philadelphia is in decay, and the road 
 is out of repair. It was not because the lands of the Mucury 
 
 * To the Senator my companion and I are indebted for letters of introduc 
 tion, which secured for us friends and assistance everywhere along our whole 
 journey. 
 
 t I should state here that the worthless colonists were the first to leave. 
 Those that I had the opportunity of meeting in the Mucury seemed to me to 
 be of a good, industrious class, but they were crippled by the failure of the 
 company. 
 
132 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 were not fertile. They are exceedingly rich, while the cli 
 mate is healthy and agreeable. It is not that the project of 
 opening a road through the Mucury to Minas was unwisely 
 planned, but it is owing to bad management on the part of 
 the company, to the slanders of enemies, and to the bad 
 character of a large part of the colonists themselves, that the 
 enterprise has proved a failure.* 
 
 The Minas road, on leaving Santa Clara, runs for a few 
 miles through a hilly region bordering the Mucury. This 
 region is covered by the ordinary red drift soil, and is very 
 fertile. Quite a number of German families still remain 
 here. Beyond Barriado, where are a few settlers, one soon 
 leaves the river valley and rises to the top of the tertiary 
 chapada,! which, at a distance of eleven and three quarter 
 leagues from Santa Clara, and near the Riacho das Pedras, 
 is 1,226 feet above the level of the sea. It forms a plain 
 like that of the Sertao J below Santa Clara, and is well 
 but not densely wooded. The chapada is covered by a 
 layer of drift of a yellowish color, in which I saw, indis 
 criminately mingled with the clayey sand of which it is 
 composed, rounded and angular fragments of quartz, sand- 
 
 * The story of the Mucury is a long and sad one, and I do not wish to enter 
 into it here. Those who desire to read the history of the colony will find a 
 very fair statement of the facts in Von Tschudi s Reisen durch Siid-Amerika, 
 Vol. II. That of Dr. Ave Lallemant is prejudiced, unfair, and unreliable. 
 
 t Chapada means primarily a plain, but in Brazil the term is applied to ele 
 vated plains or small plateaus, usually consisting of horizontal deposits, and 
 separated by deep valleys of erosion. The term chapadao is applied to chapa- 
 das of great extent, as the chapadao de Santa Maria in Minas. 
 
 \ The term Sertao, plural Sertoes, so often used in works on Brazil, simply 
 means the interior of a country as opposed to the coast. It is applied, for 
 instance, to the lands in the vicinity of the city of Sao Matheos. The word 
 appears to have a somewhat indefinite signification in Brazil. The inhabi 
 tants of the Sertao are called Sertanejos. 
 
PROVINCE OF MINAS GERAES. 133 
 
 stone, and gneiss. This soil, like that of the plains below 
 Santa Clara, is weak, and much inferior to the gneiss soils. 
 At Riacho das Pedras, a little stream flowing into the Mu- 
 cury, the tertiary lands are left, and the road reaches a 
 rolling gneiss country. The rock, wherever I saw it, was 
 very coarse-grained and homogeneous, decomposed on the 
 surface, and covered with drift clay, which is usually very 
 fine in texture, and very red from the large percentage of 
 ferric oxide. The hills were low and rounded, with a 
 topography like that of the coffee region of the Parahyba 
 do Sul at the Barra do Pirahy. Indeed, the two regions 
 are precisely identical in soil, general topography, and 
 climate. The country is covered by a dense virgin forest, 
 far more luxuriant than that which clothes the tertiary 
 plains. The country continues with the same general 
 character to Urucu, where it becomes diversified by ab 
 rupt gneiss hills, many of which are bare and precipitous, 
 and give to the scenery a very romantic and pleasing air. 
 The soils of the Urucu are extremely fertile, and yield abun 
 dantly coffee, cotton, sugar-cane, mandioca, rice, etc. The 
 climate is warm, but not so hot as on the coast, and a 
 sea-breeze cools the air in the latter part of the day and 
 evening. The climate appears to be healthy ; and that is 
 the universal testimony of even the discontented settlers. 
 The hills of Urucu are all isolated masses, and form no 
 well-defined mountain range, though they appear to be the 
 remains of a range running about east-northeast, crossing 
 the Urucu. Westward of the colony the country rises 
 steadily in altitude. At a distance of eight leagues from 
 Philadelphia the road crosses a pass in the Morro do 
 Kupan at an elevation of 1,800 palmos. The Morro it 
 self must be at least 3,000 feet in elevation above the level 
 of the sea. 
 
134 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Philadelphia is a small village situated on the left bank 
 of the Todos os Santos, about forty miles above its conflu 
 ence with the Mucury, and twenty-eight and a half leagues 
 west of Santa Clara. The Todos os Santos is, like the rest of 
 the rivers of the Mucury basin, only a respectable brook, and 
 of no especial importance in itself. Within a few miles of 
 Philadelphia there empty into the Todos os Santos several 
 little streams, among which are the Rios S. Jacintho, S. An 
 tonio, and S. Benedicto, which flow through fertile, cultivated 
 valleys, and are settled by German and Brazilian colonists. 
 The ground on which the village stands is, according to 
 the measurement of Herr Schlobach, engineer of the 
 Mucury, 1,918 feet above the sea. Many of the neighbor 
 ing hills are 300 to 400 feet high, so that the general eleva 
 tion of the country would be considerably above 2,000 feet. 
 So far as the quality of the soils in the vicinity is concerned, 
 I can only reiterate what I have said in speaking of Urucu, 
 and repeat my comparison between them and the soils of 
 the coffee regions of the Rio Parahyba do Sul. In one word, 
 I may say that the whole country, from the Riacho das 
 Pedras to the head-waters of the Mucury, forms one of the 
 most extensive and uniformly fertile agricultural regions in 
 Brazil south of the Amazonas, and I cannot help expressing 
 my firm belief that nature having so abundantly blessed the 
 Mucury, a not far distant day will see it teeming with inhab 
 itants, and the highway of a commerce with the interior of 
 Minas.* The road from Sta. Clara to Philadelphia is well 
 
 * For an interesting and detailed description of Philadelphia and vicinity, 
 vide Tschudi s Reisen (lurch Siid-Amerika, Vol. II. His sketch of the village 
 is wretched, the hills to the south appearing like an Alpine mountain range. I 
 cannot speak too strongly of the Mucury as an agricultural region, and I would 
 call the attention of emigrants to it as one of the most fertile and healthy tracts 
 I have seen anywhere in Brazil. 
 
PROVINCE OF MINAS GERAES. 135 
 
 laid out with a very good grade, and with proper repair, 
 might be made an excellent carriage-road. The soil is, 
 however, very clayey, and the passage of the heavy ox-carts 
 in wet weather, not to speak of the gullying by the rains, 
 have cut it up fearfully. The bridges were in bad repair in 
 1866. The road from Philadelphia to Alahu is nothing but 
 a miserable mule-path, badly laid out, and obstructed by 
 bushes and fallen trees, and in wet weather most abominably 
 muddy. Just before reaching Agude it passes directly over 
 the top of the highest point, apparently for the purpose of 
 giving the traveller a view of the surrounding country. 
 
 Westward of Philadelphia the country is more hilly. 
 About one league from Philadelphia the path crosses a high 
 hill, from which one has a magnificent view over the low 
 swelling hills of the vicinity of Philadelphia, with the rugged 
 mountains of Urucu in the background. A short distance 
 west of the mill of Senhor Josd Maria, the gneiss becomes 
 very micaceous, passing into mica slate, and is very full of 
 quartz veins. In the latter occur large crystals of black 
 tourmaline. Crossing the head-waters of the Mucury, near 
 Pote, the country in the vicinity of Agude is much more 
 hilly than usual, and many of the hills are of considerable 
 height. The whole country is still most luxuriantly forest- 
 clothed, and the soils are extraordinarily fertile, and in some 
 places almost black. About twenty-seven to twenty-eight 
 miles west, a few degrees north from Philadelphia, the 
 water-shed dividing the basins of the Mucury and Jcqui- 
 tinhonha is passed, and one descends into the valley of the 
 Rio Setubal. For the last league or so bordering the valley 
 of the Setubal, the rock is chiefly mica slate, with much 
 quartz in veins and layers. The soil is redder than usual, 
 full of little flakes of mica, and boulders and fragments 
 
136 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of quartz, angular and rounded, are found abundantly in it. 
 Near the post, marking thirteen leagues from Philadelphia, 
 the mica slate appeared to have a strike east-west. Dip 
 vertical. 
 
 The Jequitinhonha, one of the most important rivers 
 of JVIinas Geraes, takes its rise in the knot of the Cordil- 
 heira do Espinhaqo, in the Serra Frio, three leagues west 
 of the town of Serro, and about the same distance south- 
 southwest of the Peak of Itambe", whence come the waters 
 of two or more of its little tributaries. 
 
 The area drained by it forms an irregular triangle, of 
 which one side, from its head-waters to Belmonte, is about 
 320 miles. An almost continuous range of Serras runs 
 with a zigzag course along this line to the confines of 
 the province. On the west it is bounded by the Serra do 
 Espinhago, the water-shed lying west of the Serra do Grao 
 Mogor. This side of the triangle, which runs approxi 
 mately north-northeast, is 165 miles in length. The remain 
 ing side, which marks a water-shed determined by a series 
 of elevated plains or chapadas, and which runs almost east- 
 west, is 210 miles long. The Jequitinhonha at first flows 
 with a general northeast course for about 130 miles, when 
 it receives from the northwest a small river called Itacam- 
 birussu, which rises in the Serra d ltacambira, in the south 
 ern part of an oblong region west of the Serra do Grao Mogor, 
 formed by a range of serras or highlands which leave the 
 Serra do Grao Mogor, and bowing out westward join the Grao 
 Mogor range again eighty miles to the north. The centre of 
 this region appears to be flat, and is diversified with a large 
 number of little lakes, which discharge their waters into the 
 Itacambirussu. This river crosses the Grao Mogor range 
 between the Serra do Grao Mogor proper and the Serra 
 
PROVINCE OF MINAS GERAES. 137 
 
 Sobrado, and reaches the valley of the Jequitinhonha. It then 
 runs with a southeast course for a few leagues, and empties 
 into the river of that name. The Jequitinhonha then changes 
 its course towards the east, and, some eight leagues farther 
 down, receives the Rio Vacaria, a small stream which has 
 its source in a number of lakes in the northern third of 
 the serra-enclosed region west of Grao Mogor. After this 
 it changes its course to the southeast, and in lat. 17 S., 
 long. 1 80 E. of Rio unites with the Rio Arassuahy. The 
 Arassuahy is a large stream which rises in the serras a few 
 miles northeast of the Peak of Itambe , and flows parallel 
 with, and on an average of fifteen to twenty miles, south 
 east of the Jequitinhonha. Its principal affluents are the 
 Rio Soledade on the left, and the Itamarandiba, Fanado, 
 Capivary, Agua Suja, Setubal, Gravata and Calhao, which 
 flow into it from the south, or right, in the above descend 
 ing order, all, with the exception of the Capivary and Agua 
 Suja, having their sources in the high lands separating the 
 Jequitinhonha basin from that of the Doce and Mucury. 
 
 The Jequitinhonha, after having been increased by the 
 waters of the Arassuahy, continues its course a few degrees 
 east of northeast to the sea. It is obstructed by many dan 
 gerous rapids and cascades, and on the boundary line, in the 
 very extreme northeast corner of Minas, there is a magnifi 
 cent series of falls, which, in the aggregate, must have an 
 altitude of 300 feet. Eight leagues farther down, at Caxoei- 
 rinha, it leaves the hills, and, reaching the coast tertiary 
 plains, flows, a broad, beautiful stream, to the sea. Above 
 Caxoeirinha, it is navigable of course only for canoes. The 
 Rio Pardo is so closely connected with the Jequitinhonha 
 in the general topographical features of the country through 
 which it flows, that, before describing the topography and 
 
138 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 geology of the Jequitinhonha basin in detail, I will first give 
 a sketch of its hydrography. This river, a much smaller 
 stream than the Jequitinhonha, takes its rise in the Serra 
 das Almas, to the north of the head-waters of the Yacaria, 
 and flows with a course almost west-east, emptying into the 
 sea a few miles north of the mouth of the Jequitinhonha, 
 and receiving by a side canal, as will be hereafter described, 
 some of the waters of that stream just before it reaches the sea. 
 Its basin is triangular in shape, long from west to east, but 
 narrow from northwest-south. It is bounded on the south 
 by the water-shed of the Jequitinhonha, which runs west- 
 east. The south side of the triangle coincides with this, 
 and is 210 miles long. The west side, running along the 
 Serra das Almas, is about sixty miles long, while the remain 
 ing side is about 195 miles in length. Gerber gives the area 
 of the basin of the Pardo as 420 square leagues, and that of 
 the Jequitinhonha as 2,200.* Hydrographically, above their 
 lower courses the two rivers form separate systems, but to 
 pographically and geologically they are very closely united. 
 The united basins are essentially eozoic and palaeozoic. 
 Gneiss, mica slate, sienite, clay slate, quartzite, and lime 
 stones form the bounding, mountain ranges, and the bottom 
 rock of the region. Through the kindness of my friend, Dr. 
 Anto. de Lacerda, of Bahia, I have in my possession a litho 
 graphed section across the country, from the Serra Congonha 
 across the Grao Mogor, extending into the valley of the 
 Jequitinhonha, a section constructed by the late Dr. Virgilio 
 Helmreichen. According to this section, the serras of Con 
 gonha and Grao Mogor are composed of metamorphic slates, 
 while the intervening country and the valley of the Jequitin- 
 
 * Nocoes Geographicas, $-c., da Prov. de Minus Geraes, por Henrique Ger 
 ber, p. 9, 1 863. 
 
PROVINCE OF MINAS GERAES. 139 
 
 honha is composed of primitive rock. I have never visited 
 the Serra do Grao Mogor, and can therefore say nothing of it 
 from personal observation.* At Calhao intelligent persons 
 informed me that it is composed of slates. The serra is dis 
 tinctly visible from near Minas Novas, though distant some 
 thirty miles, so that its height may be inferred. The outlines 
 of the hills are entirely different from those of the gneiss serras 
 of the coast. Gold occurs in this serra, together with the 
 ores of other metals, such as iron, which last is mined and 
 smelted at a locality called Tropinha, two leagues to the south 
 of the town of Grao Mogor. The region embraced between 
 Minas Novas and Calhao is, according to my own observations, 
 composed of clay slates, and this group of rocks undoubtedly 
 extends considerably to the west and southwest of Minas 
 Novas. From near the mouth of the Arassuahy to a little 
 below Caxoeirinha the rocks are gneiss, mica slates, and the 
 like. All these rocks have been folded, metamorphosed, and 
 denuded. 
 
 During the tertiary, as I shall further on attempt to show, 
 the plateau of Brazil was sunk so that the waters rose to 
 a height of more than 3,000 feet above their present level, 
 and flooded the great river basins of the whole country, 
 this submergence being of almost continental extent. In 
 the basins of the Jequitinhonha and Pardo, a great thick 
 ness of more or less arenaceous clays, sandstones, &c. was 
 deposited, filling up the valley to a height in some places of 
 fully 1,000 feet, converting it into an immense plain, whose 
 level above the sea must be on an average quite 3,000 feet. 
 
 * Spix and Martius visited the Serra da Grao Mogor, which they describe as 
 being only about 4,300 feet high. The prevailing formation of this region is 
 quartzose slate (quarz-schiefer). Boulders of white quartz (sometimes fibrous(?)) 
 are abundantly scattered over the surface, and contain asbestus. Gold and 
 diamonds occur here. 
 
140 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 These deposits I have called tertiary, because along the 
 whole coast they are undisturbed, nowhere participating in 
 the disturbance of the cretaceous, and because the drift sheet 
 extends over them. I believe them not to be drift, because 
 they were denuded by river action anterior to the formation 
 of the drift sheet, which descends their slopes, and extends 
 over the slate and gneiss hills left bare. Similar deposits 
 were at the same time laid down in the valleys of the Sao 
 Francisco, Parana, Parahyba do Sul. And indeed over the 
 whole plateau to the westward, as we shall see further on. 
 All these are older than the coast tertiaries. 
 
 In April, 1866, 1 entered the Jequitinhonha valley, on the 
 Setubal, and after making a detour to Alto dos Bois, crossed 
 the country to Calhao, from which place I found it prac 
 ticable to visit Minas Novas, after which I returned to 
 Calhao, and descended the Jequitinhonha to the sea. I 
 propose now to give the results of my explorations of this 
 region, following very nearly my line of travel.* 
 
 The tertiary clays are denuded away from the region of 
 the head-waters of the Setubal, and the wider valley of this 
 river is scooped out of these rocks, the river-bed being the 
 solid gneiss or slate, or excavated in alluvial deposits laid 
 down by the river. The Setubinho is a little river which 
 flows from the southwest in a valley bounded, on the one 
 side, by the hills of the water-shed between the Mucury and 
 Jequitinhonha basins, and on the other by the tertiary 
 plains, though the slopes on both sides are of the old mcta- 
 morphic rocks, the tertiary beds merely capping the hills on 
 
 * The months of February and March, 1866, were exceedingly rainy over 
 the Mucury region, and so was the month of April, which I spent in the basin 
 of the Jequitinhonha. During this time I was obliged to travel over the worst 
 possible roads in almost constant rains, so that my geological studies were made 
 under a great disadvantage. 
 
PROVINCE OF MINAS GERAES. 141 
 
 the northwest side. The slopes toward the Setubal, Seturna, 
 and Setubinho are all very steep, and for a large part bare 
 and excessively stony, the soil being full of rounded and 
 angular fragments of quartz often of large size. Over large 
 areas it is very barren, the only vegetation consisting of 
 low, gnarly branched, sparsely scattered shrubs and trees. 
 
 One observes immediately on entering the valley of the 
 Setubal from the Mucury that the forest thins out and dis 
 appears from the hillsides, though it extends down the wet 
 valleys and over such areas as may have rich soils, but even 
 there the forest has not the same luxuriance it had in the 
 valleys of the Mucury, and there are many trees confined to 
 each separate region. It is, however, on the open grounds 
 that the change in the vegetation is most marked. On 
 leaving the forest (sahindo do matto* as the Brazilians 
 say) one of the first plants to attract one s attention is an 
 arboraceous species of the order of Solanacese, called the 
 Boleiro. This tree attains to a height of fifteen to twenty 
 feet, and forms a conspicuous element in the landscape. 
 Its leaves are light-green and curly, its flowers bluish-purple, 
 and its fruit, which is of the size of a Baldwin apple, is 
 edible ; but of the flora of the campos more anon. 
 
 The high and steep hill north of the Setubal is almost 
 bare on the south and west ; but the northern side is covered 
 by a stiff drift clay, and is clothed with a thick wood, 
 densely filled with a luxuriant growth of a species of bam 
 boo, the slender-stemmed Taquara lisa of the Brazilians. 
 Thence to Corrego Grande the country is composed of highly 
 micaceous and schistose gneiss, and is covered by a thick 
 sheet of drift clay, in which are boulders and pebbles of 
 
 * The Brazilians speak of the plains as fora, outside, and of the forest as 
 dentro, or inside. 
 
142 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 quartz, and the country is wooded. At the Fazendas of 
 Santa Barbara and Santo Antonio, and in the neighborhood, 
 this forms a rich soil, which is very productive. Maize is 
 largely cultivated in this region, and takes the place of 
 mandioca for the making of farinha. Wheat grows well 
 here, and I saw some most excellent sheaves at a farm 
 house near the Setubal, but the farmers complain that it has 
 to be planted in clumps like rice, and weeded, which is very 
 troublesome. At the Fazenda of Santa Barbara the country 
 bordering the Rio Setubal is very hilly. The prevailing 
 rock is mica slate or schistose gneiss, with a general strike 
 of N. 80 E., and northward dip. I have recorded no south 
 ward dips. 
 
 Wishing to ascertain the character of the chapada west 
 of the Setubal, Mr. Copeland and I made an excursion 
 thither from Santa Barbara. The account of the journey 
 I transcribe with few changes from my note-book, in 
 order not only to give an idea of the country, but of the 
 disadvantage under which the geologist labors in exploring 
 in the rainy season. 
 
 For the last two months the rain had been constant, 
 and it was still raining when we reached Santa Barbara. 
 Our time was very limited, but to leave the Setubal with 
 out seeing the topography of the plains was not to be 
 thought of, so, on the last day of March, we set out on mule- 
 back, and unencumbered by baggage, for Alto dos Bois, a 
 point described as being so elevated as to overlook the 
 plains, and enable one, if the weather were clear, to see 
 across the valley of the Jequitinhonha, and discern the 
 mountains of Grao Mogor. It was raining heavily. We 
 crossed the Setubal by a bridge below Santa Barbara, finding 
 the stream very much swollen and turbid, and the meadows 
 
PROVINCE OF MINAS GERAES. 
 
 143 
 
 bordering it inundated. We followed up the valley a short 
 distance, passing through cornfields on the hillside, and 
 crossing a high hill by a miserable path leading through a 
 wood which was so tangled with bushes, unha de gato, bam 
 boos, &c., that it was with difficulty that we could burst our 
 way through it by main strength. We reached at last the 
 valley of the little river Santa Rita. Thus far the country 
 was of the same character as at Santa Barbara, and the 
 surface was covered by the same drift paste and boulder 
 deposit. For a distance of some two miles or more farther 
 on, after a long and steep ascent, we reached the foot of the 
 chapada, which presented a long, steep, even slope, as in the 
 following ideal section. 
 
 a, ACL 
 
 It was impossible to estimate satisfactorily the height of 
 the chapada top from the Santa Rita valley, because of the 
 wooded state of the country and the obscurity of the weather, 
 but after having ascended and descended the same slope, I 
 should judge that it is fully 800 feet, if not very much more. 
 The lower part of the slope is covered with the ordinary drift 
 paste, filled with boulders of quartz, gneiss, and mica slate. 
 The quartz of this region is so crystallized as to break up 
 into a coarse angular gravel. About half-way up I saw mica 
 slate in sitti traversed by a thick quartz vein. Farther up 
 
144 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the soil changed in character, and, when wet, was of an 
 umber brown color, and quite free from stones. The road 
 went straight up this slope, and being cut up into piloes * 
 by the feet of the mules, formed a sort of staircase of 
 stiff, adhesive, and slippery clay, which made the ascent 
 exceedingly wearisome. Arrived at the top of the chapada, 
 we found ourselves on a wooded plain. The soil seemed 
 very rich, and was of a dark brown color on the surface ; 
 but I observed that the material brought up from below 
 by the Saiiba ants was clayey and brick-red, like the ordi 
 nary drift paste of the gneiss regions. I saw a few quartz 
 boulders lying on the surface. I looked carefully in the 
 ascent for any cuttings which might enable me to ascertain 
 the material of which the chapada was composed, but I saw 
 none. The woods consist of large trees, but they are rather 
 sparsely sown, and they belonged to the catinga class, 
 losing their leaves in the dry season. There is an abun 
 dant undergrowth of bushes and Samambaia ferns (Pteris 
 caudata (?). The beautiful Indaid palm (Attalea) is very 
 abundant here, and its plumy coronals add very much to 
 the picturesqueness of the scenery. It was long after 
 nightfall when we reached the open campos at As Trovoadas 
 and crossed a very deep valley to the place called Batatal, 
 where, jaded and wet, we spent the night. The next morn 
 ing we visited the Alto, or the highest point of land in the 
 vicinity. It had been raining ; but, providentially, for a 
 half-hour it cleared up, and we had an almost uninterrupted 
 and most magnificent view of the country on every side. 
 The Alto is the name given to the highest point of the 
 
 * It is well known that mules travelling over a bad road follow in one an 
 other s footsteps, cutting up the road into a series of transverse muddy troughs, 
 separated often by high ridges, over which the animal carefully steps. 
 
PROVINCE OF MINAS GERAES. 145 
 
 swelling ridge between the valleys of the Fanado and 
 Capivary. From this point the country descends with 
 very long sweeping curves to the river valleys on both 
 sides, the ridge itself growing gradually lower toward Mi- 
 nas Novas. The valley of the Capivary is very broad, 
 and in profile rounded, not angular. The long and gentle 
 curves of this landscape are very noteworthy, and are 
 very different from those which result solely from water 
 denudation or erosion. The Fanado valley is of the same 
 character. East of the Capivary the land rises in a high 
 ridge called the Caixao, which runs northward, present 
 ing the same topographical features. This whole country 
 is covered on the surface Avith red drift clay and pebbles, 
 and this layer is twenty or more feet in depth. No solid 
 rock is to be seen, but on the sharp descent into the valleys, 
 and in certain gullies, especially at the foot of an isolated, 
 dome-shaped mass which rises above the general level of the 
 country at As Trovoadas, the rock, in a very decomposed 
 state, is seen to be crystalline and metamorphic, and in 
 beds highly inclined. At As Trovoadas, as nearly as I could 
 make out, the strike was N. 60 W., and the dip 40 north 
 ward, but the observation was taken from decomposed rock, 
 and may not be very reliable. The rock appears to be com 
 posed of feldspar and mica, with quartz in rounded grains, 
 but when decomposed it is red, very soft, and crumbling. 
 The soil in the gully, where the above observation was taken, 
 is full of fragments of very limpid quartz and crystals of 
 kyanite and black tourmaline.* This rock must be very deeply 
 
 * The Diccionario Geografico says (Vol. II. p. 98), that antimony occurs at 
 Alto dos Bois, and I heard many persons on the Setubal say that gold had been 
 found in the ridge separating the Setubal from the Capivary. Saint Hilaire 
 also says that antimony is found at the Alto. But the antimonio of the Bra 
 zilians is only a pyrites. 
 
 VOL. I. 7 J 
 
146 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 decomposed, for the freshly cut gullies are excavated to a 
 great depth, and the river valleys are evidently cut through 
 soft material. Their shape is not such as usually results 
 from river erosion, for in soft materials river valleys have 
 sides either bluff or with angular slopes of about thirty de 
 grees or more, and bluffs of that kind are to be found along 
 the edge of the valley recently washed out by torrents. 
 
 The tertiary clays and decomposed metamorphic rock 
 being of so nearly the same consistency, it is not at all won 
 derful that in their denudation they should both wear down 
 together, and that the metamorphic hills should pass almost 
 insensibly into the plains. I believe that the wide upper 
 valley of the Capivary is one of erosion anterior to the drift, 
 and that the present swelling outlines and long curved slopes 
 of its bounding ridges are due to glacial action over a 
 surface deeply decomposed. 
 
 The hillsides and slopes of this region are sparsely covered 
 with coarse grass and small flowering plants. It was like 
 a garden. Trees are few and scattered, and are noted for 
 their gnarly branches and rough bark. A little palm ( Cocos 
 flexuosa) is not uncommon on the campos. Another con 
 spicuous little tree of these campos is the Pao de Paina, which 
 has a small stem covered by a thick growth of a woolly sub 
 stance, used to fill pillows, &c. It is very abundant in the 
 neighborhood of As Trovoadas. Sometimes the trees form 
 clusters (capQes) in low and wet places, and along the river 
 side, in the valley of the Capivary. In the dry season these 
 campos are dried up, and vegetation is withered and appears 
 dead. The lands, though rich, are worthless, except as fur 
 nishing pasturage for herds of cattle which abound on these 
 campos.* The Ema, or American ostrich (Rhea Americana), 
 
 * According to Spix and Martius, from the Arraial do Rio Manso there 
 
PROVINCE OF MINAS GERAES. 147 
 
 is found on the campos, but appears now to oe rare in this 
 vicinity. It occurs more abundantly farther north on the 
 campos of the Rio Pardo, and on the campos of the Sao Fran 
 cisco basin. It ranges from Cear4 to Buenos Ay res. In 
 Patagonia another species, Rhea Darwini, is found. There 
 is also a species of deer (Cervus campestris), called by the 
 Mineiros Veado campeiro, which is not uncommon on the 
 plains. But after a half-hour s enjoyment of the magnificent 
 view, down came the heavy rain with a strong chilling wind, 
 and we left the Alto to return. The steep slope from the 
 chapada to Santa Rita was so slippery and untrustworthy, 
 that we were obliged to make the descent on foot in the 
 deep mud, leading our mules. 
 
 From Santa Rita we crossed a steep, high ridge by a road 
 terribly cut up, and so full of loose quartz stones as to afford 
 a very insecure foothold for the mules, and at night reached 
 the river Setubinho wet and sore, and there we spent the 
 
 stretches northward a low plateau, in a north-south direction, for several 
 leagues, apparently tying in with the great chapada forming the dividing 
 line between the Jequitinhonha and Arassuahy, opposite Minas Novas. " The 
 vegetation of these uniform, elevated plains, which extends from Tejuco to 
 Minas Novas, and gently decreases in height, shows a form which we had not 
 before observed to a similar extent." Low crooked-branched and broad-leaved 
 trees lift themselves here and there amongst a dense thicket of many kinds of 
 bushes, which alternate now with bare rock-sheets, now with thirsty open fields, 
 or, in the low places and beds of streams, with a somewhat higher and sappy 
 wood (Capao). The bush they call here Serrado, or, when it is lower and 
 without trees, Carrasco. The plants belonging to it do not all lose their leaves 
 during the dry season, and in a note our authors add : " Particularly those of the 
 genera Sida, Ochna, Mimosa, Acacia, Qualea, Coccoloba, Kielmeyera, Laurus, 
 Nycterosition, Arrazoa, Barnadesia, Albertinia, Anona, Banisteria, Malpighia, 
 Aspidosperma. The stemless palms, Astrocaryum campestre and Diplothemiun 
 campestre, and the low Cocos flexuosa, one sees here." (Vol. II. pp. 473, 474.) 
 A very interesting article in Danish, on the campos region of Brazil, from the 
 pen of Eugen. Warming, is to be found in the Tidsskrifl for pop. Frcem. af 
 Naturvidenskaben, 3 d " Rcekke, 5" Bind, !* Hefte, 1868. 
 
148 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 night. The morning found the little river swollen to its ut 
 most capacity, and running like a mill-race, turbid and yel 
 low, and it was with much difficulty that we forded it. The 
 Rio Setubal we found also in as swollen a state ; mules and 
 horses gave out, and the weary travellers waded through the 
 mud many a long mile, and arrived drenched at the Fazenda 
 de Santa Barbara. 
 
 Leaving Santa Barbara for Calhao, I observed near a little 
 brook at the Fazenda da Lagtia rounded quartz boulders, 
 overspread by drift clay. Beyond this the road, after 
 passing a low flat, ascends a gentle slope, on which the same 
 are seen at a height of one hundred feet above the brook ; 
 and a short distance farther on and higher up the road is 
 full of coarse gravel, intermixed with angular boulders. 
 Mica slate is seen occasionally cropping out on the hillsides, 
 and on the top of a chapada, before reaching the Fazenda 
 of the Tenente Honorio Ottoni, I saw in the drift paste 
 rounded pebbles. From the chapada one descends into the 
 valley of one of the tributaries of the river Gravata. The 
 hills of the valley are mica slate and compact dark gray 
 gneiss, with a strike of N. 40 W. and a dip apparently to 
 the northeast, or vertical. Near the fazenda are numerous 
 road cuttings through the hillside, showing the red drift 
 clays to be many feet in depth, and containing an abundance 
 of large fragments of gneiss, quartz, and mica slate. The 
 river Gravata, where the path to Calhao crosses it, runs in 
 a deep, narrow valley bordered by gneiss slopes, rising to 
 chapada plains above. I observed, in ascending the slope 
 to leave the valley, that, for some five hundred feet, the 
 slope was strewn with quartz boulders, which are very 
 numerous and large. About half-way up the slope there is 
 an exposure of a white or brownish schistose rock, much de- 
 
PROVINCE OF MINAS GERAES. 149 
 
 cayed, which seems to be wholly made up of very minute 
 and rounded grains of limpid quartz, without visible cement 
 in the specimens I examined. This rock had apparently a 
 dip of 20 to the southward. The relation of this rock to 
 the gneiss, or to the materials of the chapada, I did not make 
 out ; but I much suspect that the dip was only local. This 
 chapada extends from the Gravatd to a little valley called 
 Estrella, north of which is a plain of very wide extent, and 
 perfectly level. I find in my diary a note that on the top 
 of this plain I saw quartz boulders lying. Two little streams, 
 the Agua da Nova and Diamantino, take their rise on the 
 plain, in quite extensive, shallow, marshy lagoons, but they 
 soon cut for themselves valleys down to the older rocks be 
 neath. These plains are more or less thickly covered with 
 bushes, gnarly-branched trees, and occasional thickets. I 
 saw no Indai&s here, but a little crooked-stemmed palm, 
 called Licuri, the bases of whose leaf-stalks were persistent 
 for some distance down the stem, was quite common. 
 Ferns are rare. Among the trees are several species which 
 produce edible fruit, especially the Mangabeira, Bacupari, 
 Piqui, &c. Great numbers of cattle are pastured on these 
 plains, and grazing is one of the principal occupations of 
 this part of the country. The cattle are allowed to roam 
 over the plains, and are taken care of by mounted vaqueiros, 
 who dress from head to foot in leather, that they may be 
 able to break through the thickets in their chase after the 
 cattle. 
 
 The valley of the Agua da Nova not only deepens, but 
 grows wider in descending, and opens out broadly, on leav 
 ing the chapada, into the great valley of the Calhao and 
 Arassuahy. Running along the edge of the chapada, at the 
 top of the slope, are occasional perpendicular bluifs, in which 
 
150 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 is exposed a thick horizontal bed of sandstone, which is 
 seen forming similar bluffs on the opposite side of the valley. 
 This bed forms the upper stratum of the chapada formation, 
 and may be seen forming bluffs of the same sort along the 
 valley of the Calhao. The sandstone is white, very com 
 pact, and rather fine-grained, but there are some beds which 
 are coarse, containing pebbles, and with a hard, opaque, 
 white cement, resembling that of the tertiary sandstone of 
 Itahunas. Veins of milky quartz traverse these rocks. 
 They certainly bear an altered and old look, but they are 
 here surface-layers, and have never been disturbed. The 
 valleys of the Calhao, and of some of the little rivers west, 
 coalesce several miles before reaching the Arassuahy, leav 
 ing the metamorphic rocks, over a very large area, denuded 
 of the formation of the chapadas. This forms a great de 
 pression like a lake valley, some 800 or 1,000 feet below the 
 plain, and which is surrounded on all sides by high, level- 
 topped chapadas, which project in capes and promontories 
 between the river valleys. The bottom of the depression is 
 diversified by low, rounded, wooded hills. From the top 
 of the sharp spur of the chapada, on the western side of the 
 valley of the Agua da Nova, one has a magnificent view over 
 this great valley. It was near the close of a clear afternoon 
 that we rode out on the edge of this spur to descend, and 
 suddenly, leaving the bushes of the plain, saw before us the 
 beautiful valley. The level-topped chapadas beyond the Aras 
 suahy extended like a wall to the north of the depression, 
 blue in the far distance, while below us lay the billowy sea 
 of foliage which clothed the bottom of the valley. Weeks 
 of sore, weary forest wanderings, beneath a rainy sky, were 
 forgotten, and the heart, homesick, tired, and often disap 
 pointed, gladdened as the eye revelled in the beauties of the 
 
PROVINCE OF MINAS GERAES. 151 
 
 landscape ; but the sun was rapidly nearing the level horizon, 
 and our camarada warned us that we must descend. So, 
 turning our mules into the steep path, we soon passed into 
 the thicket, and the landscape was lost to view. It was as 
 when, after the curtain has dropped at the close of the last 
 act of an opera, and the memory of the brilliant scenery 
 and the rich music still lingers in the heart, one wakes to 
 feel the sorrows and the realities of life again. 
 
 The country bordering the Calhao River, for several 
 miles above its mouth, is composed of slates, which are 
 seen exposed in the banks of some of the little brooks 
 flowing into the Calhao.* These slates, on the right bank 
 of the Arassuahy River at Calhao, are fine-grained and sili 
 ceous, and have the slaty structure well developed. They 
 dip to the N. 70 D W., at angles varying from 50 to 80. 
 The planes of cleavage dip to the south 20 W., but I have 
 omitted to note the angle. The country forming the bottom 
 of the valley is much more uneven than one would suppose 
 when looking at it from the chapada, and some of the hills 
 are several hundred feet high. They are everywhere cov 
 ered by the characteristic red drift-clay on the surface, under 
 which occurs usually a sheet of pebbles, as at Rio. This 
 pebble-sheet is sometimes very thick, and being exposed on 
 hillsides by the washing away of the clays, leaves them very 
 barren. As a general thing the country is sparsely wood 
 ed, but ordinarily the forest (catinga) does not bear the 
 
 * At the head-waters of the Calhao Spix and Martius found the rock to be 
 coarse-grained, whitish, unstratified, with little white mica, but with much 
 black schorl, often in long prismatic crystals. They state that it is covered by 
 a layer of gray or white pebbles of quartz containing yrisolitas (crysoberyl) 
 of a greenish-white, pale ochre, or citron-yellow color, and others of an olive, 
 grass, or blue-green color (Acjoas marin/ias), precious garnets and white and 
 bright blue topazes. (Vol. II. p. 502.) 
 
152 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 same luxuriant look as that of the Mucury, and resem 
 bles a second growth. I have observed immense arbo 
 rescent cactuses (Cereus^) growing in the woods near 
 Calhao.* The higher lands are apt to be dry, though the 
 soil would otherwise be fertile, and during the dry season 
 the trees lose their leaves. The river-borders, or varzeas, 
 are very productive. Cotton seems to be the principal pro 
 duct, and it is of excellent quality. One thousand canoe- 
 loads were sent down the Jequitinhonha to the sea in one 
 year, but a very considerable quantity of the cotton is manu 
 factured at home into coarse cloths, &c. Large quantities 
 of corn, beans, &c. are raised here. Calhao is a village of 
 respectable size, situated on the right bank of the Aras- 
 suahy, at its junction with the Calhao, which is so small, 
 and ordinarily so shallow, that the negrowomen wade across 
 it to fill their water-jars in the Arassuahy. At Calhao the 
 latter river is about the size of the Mucury below Santa 
 Clara. Calhao derives its importance from being a sort 
 of centre of the salt-trade with the coast via the Jequi 
 tinhonha. 
 
 At the point where the path from Calhao to Minas Novas 
 crosses the Rio Setubal gray quartzites are exposed, with a 
 strike of N. 65 E., dip 85 southeastward. At the passage 
 of the Corrego de Sao Joao fine-grained siliceous gray 
 schists are seen, strike N. 60 J E., dip 88 to 90 southeast 
 ward ; and at the passage of the Sucuriu the same rock is 
 seen, and an observation gave strike N. 60 E., dip 80 N. 
 
 * Saint Hilaire says : " In general the cactuses in the province of Minas 
 appear to belong to the catingas in the neighborhood of the Arassuahy and 
 Jequitinhonha, for I have not met with a single species either in the Geraes, 
 properly so called, or in the carrascos. (Vol. II. Part I. p. 103.) The same 
 author calls attention to the number of Barrigudo trees (Bombax), and the 
 absence of Melanostomaceous plants in the catingas. 
 
PROVINCE OF MINAS GERAES. 153 
 
 At the villa of Sucuriti the same rock is seen, strike N. 
 30 E., dip 50 S. On the west side of the Corrego de Sao 
 Joao there is an outlying chapada, the southern side of which 
 presents a red and white cliff, in which are exposed horizon 
 tal white beds, which, I was informed, are composed of ta- 
 batinga, or clay over which is a thick bed of red drift-earth, 
 such as is seen everywhere covering the country. The drift, 
 as well as the sand and gravel of the streams in the vi 
 cinity of Sucuriu, contains gold, but in small quantities. I 
 saw a few old abandoned workings. West of Sucuriu the 
 road passes over a chapada, and descends into the valley of 
 the Sucuriu, which is bordered by high slate hills, and then 
 ascends to a chapada which, perfectly level and covered by 
 carrasco, extends for a league to the Rio d Agua Suja.* 
 This chapada is precipitous along the edges, and is covered 
 by a thick bed of red drift-clay, under which appears to lie 
 a sheet of gravel, which in some places is cemented by 
 oxide of iron, and forms a conglomerate. From the borders 
 of this chapada one has the most extended views of the sur 
 rounding country, and in clear weather the higher points of 
 the Serra do Grao Mogor are distinctly visible. 
 
 The valley or canon of the Agua Suja in some places cuts 
 through the whole chapada formation to the metamorphic 
 rocks below, and is very deep and narrow. This river flows 
 northward into the Arassuahy. After crossing a narrow 
 
 * The grass and bushes of the campos are infested by the carapato (Ixodes 
 ricinus], a wood-tick which, swept off by contact with the garments, attaches 
 itself by hundreds to the skin, and can only be detached by the application of 
 tobacco or something of the kind. Strangers are apt to suffer severely from the 
 irritation caused by these disgusting creatures. Even the Jaboti tortoise is at 
 tacked by them. When allowed to remain, the animal feeds on the juices of its 
 prey until its body becomes as large as a castor-bean, to which in shape and 
 color it hears a close resemblance. 
 7* 
 
154 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 chapada another canon is reached, that of the Rio d Agua 
 Limpa or Mae d Agua, a little river flowing into the Agua 
 Suja.* The banks of these valleys are wooded, and possess 
 a fertile soil. At the ford slates of ordinary quality are 
 seen. Strike N. 50 E., dip 70 southeastward. Between 
 the Capivary and Minas Novas the country is very hilly 
 and barren, the vegetation being of the character of the 
 campos. In the numerous rain-gullies in the mule-paths 
 the drift is cut through, and decomposed slates are ex 
 posed. They are as soft as the drift-clay, and were it not 
 for the different tints of the laminae and the quartz veins 
 which traverse them, it would be difficult to recognize them 
 as a metamorphic rock decomposed in situ. The Ribeirao 
 do Meio is a brook emptying into the Capivary, from the 
 sands of which gold has been obtained. Spix and Mar- 
 tius have left us the following graphic picture of this part 
 of the country : 
 
 " The thick wood appeared to us a wide grave, for the 
 dry season had stripped off all ornament of leaves and 
 flowers ; only once in a while thorny species of smilax, or 
 cord-like twists of cissus, set with single leaves, climbed up, 
 or the stately flower-panicles of Bromelias stretched them 
 selves out from among the branches Thorny acacias, 
 
 many-branched Andirce and Copaiferce, and fig-trees rich in 
 milk, appeared here in exceeding plenty ; but what most 
 pleased us were the giant stems of Chorisice ( Chorisia ven- 
 tricosa), which, contracted above and below, were swollen in 
 the middle like huge casks, their cork-like bark being beset 
 with stout shining brown spines. Here huge bunches of 
 parasitic plants depended from the branches. Here myriads 
 
 * I give the names on the authority of my guide. According to Gerber s 
 map, the eastern stream is the Agua Limpa, the other Agua Suja. 
 
PROVINCE OF MINAS GERAES. 155 
 
 of ants have hung from the branches their nests full of 
 Daedalian windings, and which, with a circumference of sev 
 eral feet, contrasted strangely in their black color with the 
 bright gray of the leafless branches. The autumnal torpid 
 wood echoed with the cry of many kinds of birds ; especially 
 croaking araras and piriquitos. Shy armadillos and anteat- 
 ers {Dasypus septemcinctus and Myrmicophaga tetradactyla) 
 met us, .... and sluggish sloths (Bradypus tridactylus ) 
 hung stupidly from the white branches of the embauba 
 (Cecropia peltata), which here and there rose among the 
 rest of the trees. Herds of howling monkeys were heard 
 in the distance. The high dry grass was covered with 
 crowded balls of little carapatos, which, when we accident 
 ally disturbed them, scattered themselves with lightning-like 
 rapidity over us, and excited a painful itching. Not infre 
 quently a snake was heard in the thicket by the traveller 
 riding hastily by." * 
 
 I passed through this same region in the wet season, 
 when the trees were all in leaf, and the woods looking 
 gay and pleasant. I saw scarcely any animals. I heard 
 some guaribas howling ; but neither armadillos, sloths, nor 
 snakes of any kind were seen. It is a very mistaken idea, 
 carefully spread abroad by our geographies and popular 
 works and pictures, that one may everywhere expect to see 
 in the Brazilian forests great boas wreathed about the trees, 
 and all manner of birds and beasts in profusion. I have 
 ridden day after day through the virgin forest without seeing 
 or hearing anything worth shooting, and nothing more dan 
 gerous than a wasp ! 
 
 In the year 1727 Sebastiao Leme do Prado, with a band 
 of Paulistas, travelling northward through the province of 
 
 * Spix and Martins, Reise, Vol. II. pp. 499, 500. 
 
156 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Minas, discovered gold in the river Bom Successo, and gave 
 it the name which it bears He established here regular 
 mining operations, and founded the city now known as Minas 
 Novas, which grew to be a flourishing town. The precious 
 metal was also discovered elsewhere in the vicinity, and in 
 especial abundance on the hills bordering the Rio Capivary 
 near the Arraial da Chapada, where it was very exten 
 sively mined. The gold was principally obtained from the 
 sands and gravel of the river, and from the gravel sheet 
 underlying the drift-clays on the slopes and tops of the 
 hills. Very little gold was extracted from the veins of the 
 quartz, some of which were known to be richly auriferous. 
 The hills are dry, and water is to be found only in the 
 rivers, which during the greater part of the year aiford a 
 good supply, so that the washing of the gravels on the 
 high grounds was attended with much difficulty. Ditches, 
 or reg-as, were dug round the hills to collect rain-water, 
 which was brought into tanks, and in some of the washings 
 all the water used was derived from this source ; and at 
 Minas Novas and Chapada, washings said to be rich were 
 pointed out to me as abandoned because of the scarcity of 
 water, when just below, a hundred feet or more, tumbled a 
 dashing stream. Notwithstanding the disadvantages under 
 which these old miners labored, a large extent of ground 
 was, as we shall see, worked over, and an immense quantity 
 of gold was extracted, according to one authority 300 arrobas 
 (=9,600 Ibs. avoirdupois) being sent to Bahia alone. Many 
 large nuggets were discovered in these mines. In the Lavra 
 do Batatal a lump weighing 28 Ibs. was found. In 1746 dia 
 monds were discovered in the vicinity of Diamantina, and 
 government prohibited the extraction of gold in order to 
 encourage the search for diamonds. This prohibition put a 
 
PROVINCE OF MINAS GERAES. 157 
 
 stop to the gold-mining of Minas Novas,* and, though the 
 prohibition has been removed, the blow has been fatal, for 
 little gold-mining has since been carried on, and the present 
 inhabitants content themselves with agricultural pursuits, 
 or help to swell the number of miners who wash for dia 
 monds on the Rio Jequitinhonha. The gold-mines are 
 to-day practically abandoned, but the idea that they were 
 worked out is very erroneous. At Minas Novas and Cha- 
 pada the rocks are slates and quartzites, and resemble very 
 closely those of the gold region of Nova Scotia. Indeed, it 
 was the strong resemblance borne by the slates of Calhao 
 and the vicinity to the Nova Scotian gold-bearing rocks 
 that aroused my interest, and led me to turn out of my 
 way to visit Minas Novas. These rocks evidently overlie 
 the mica slates which flank the gneiss of the coast belt, 
 and I believe they will prove to be Lower Silurian in age.f 
 At Minas Novas their strike is N. 42 to 50 E., and their 
 dip is vertical. They are traversed by great numbers of 
 milky-quartz veins, some of which are well known to be 
 auriferous. Some of these veins are of considerable dimen 
 sions. In an enormous gully cut out by the surface waters 
 in the hillside above the cemetery on the Bom Successo at 
 Minas Novas are several fine veins of corrugated quartz. J 
 These veins run nearly vertically through the rock, and may 
 be beds instead of veins. As they are exposed in the cliff 
 they present the appearance of vertical fissures, in which 
 cylindrical masses of quartz are piled in a single row, their 
 
 * The city is in decay, and is to-day of very little importance. The cotton 
 raised in its vicinity has a very excellent reputation in Brazil ; similar lands 
 in Bahia and Pernambuco produce a good quality of cotton. 
 
 t Perhaps Quebec group. 
 
 J These appear to have precisely the same structure as the " Barrel quartz" 
 of Nova Scotia. 
 
158 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ends projecting like logs. Some of these cylinders of quartz 
 are two feet in diameter. In section they appear as repre 
 sented by the accompanying woodcut. I spent some time 
 in an examination of this vein for gold, 
 but could detect none. The large size of 
 the quartz veins of the vicinity may be in 
 ferred from the dimensions of the quartz 
 boulders scattered over the surface, some 
 of which weigh many tons. I am not aware 
 that any auriferous vein has been worked 
 at or near Minas Novas, but at the Arraial 
 da Chapada several were anciently more or 
 less worked. A rich vein, according to universal testimony, 
 crosses the praga, and it is well known that one miner fol 
 lowed it in secret until he undermined his neighbor s house, 
 when his secret was let out. There one hears the terms 
 " vein " and " gravel " gold, and I saw many beautiful spe 
 cimens of crystallized gold in the hands of the inhabitants, 
 some of which were taken directly from quartz veins, though 
 it is true others were obtained from quartz boulders. There 
 can be no doubt that rich auriferous lodes exist in the neigh 
 borhood, which have never been explored, and which one day 
 must be developed, for all the gold which so richly abounds in 
 the drift must have come from the underlying rocks.* In the 
 Minas Novas region I have seen no signs of gneissosc rocks, 
 itacolumite, or itabarite associated with the auriferous slates. 
 The slates, <fec., of the valley of the Jequitinhonha are 
 decomposed to a great depth, and are as soft as earth, and 
 can be easily worked with a spade. This decomposed rock, 
 which is of a bright red color, preserves its lamination, and 
 the quartz veins traverse it as in the solid rock. Excellent 
 
 * For an account of the gold-mines of other parts of Brazil, see further on. 
 
PROVINCE OF MINAS GERAES. 159 
 
 opportunities for the examination of it are afforded at Minas 
 Novas and elsewhere by the enormous gullies swept out in 
 it on the hillsides by the mountain torrents. Some of these 
 gullies are more than 100 feet deep, and show at the same 
 time the very finest sections of the drift. Near the Arraial 
 da Chapada the bright red cliffs of these ravines are very 
 conspicuous elements in the landscape, and some parts of 
 the country appear as if scored by a giant plough. Burton 
 describes similar gullies in other parts of Minas and in Sao 
 Paulo. In the latter country they have received the name 
 of vossorocas. Burton supposes they were formed by the 
 giving way of a hillside under the hydrostatic pressure 
 caused by the soaking of the mass by water ; and he says 
 that the ground breaks away suddenly with the force of an 
 eruption, the hollow in the hillside thus formed being after 
 ward excavated to a greater size and depth by rains and 
 streams, which sometimes gush out of the head of these 
 gullies. The gullies which I saw did not strike me as 
 having been formed in this way. I supposed that they had 
 been hollowed out with more or less rapidity by the ac 
 tion of surface water, perhaps aided by springs, and with 
 out a regular land-slide. The surface of the undisturbed 
 decomposed rock is always well marked, and has a regu 
 larly rounded contour like that of the gneiss, and is never 
 irregular and jagged like a water- worn surface. The de 
 composed rock is immediately overlaid by a sheet of cas- 
 calho, or quartz pebbles, whose thickness varies from a few 
 inches to eight or more feet. The pebbles are of all sizes, 
 and are more or less rounded. I observed in several 
 localities that there were large boulders lying in this gravel 
 just above the rock. The cascalho is often so cemented 
 by ferric oxide as to form a conglomerate, which requires 
 
160 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 to be broken up before it is washed for gold. Like the 
 drift pebble sheet of the coast, it forms a concentric layer 
 wrapped over the whole rock surface of the hills, and it 
 is found lying on very high slopes and piled up in masses 
 such as water never deposits. 
 
 It is in the cascalho that the greater part of the gold 
 of Minas Novas and vicinity occurs. Over this gravel lies 
 a mass of red drift-clay, varying very much in thickness, 
 from a few inches to fifty feet or more. This is, like the 
 drift-clays of the coast, a homogeneous mass, through which 
 are scattered from time to time angular and rounded quartz 
 boulders of large size. Over large tracts between Minas 
 Novas and the Arraial da Chapada this sheet of clay is so 
 thin that the cascalho bed lies on the surface, and the 
 country is consequently stony and barren. The clay con 
 tains sometimes more or less gold. It is, however, to the 
 cascalho sheets that the search for the precious metal has 
 been principally confined. The gold occurs disseminated 
 through the gravel in flattened grains, and occasional nug 
 gets of considerable size, which are always in a crushed and 
 battered state. The process of extraction was similar to 
 that described by Mawe as employed at the mines of Jara- 
 gud in Sao Paulo, and which I shall further unfold in the 
 description of that province. It consisted in stripping off 
 the clay sheet down to the gravel, which was broken up and 
 washed on the spot in rude trenches to separate the peb 
 bles, when the auriferous mud and sand were washed in 
 the bateia, or wooden washing-pan. A great number of 
 the washings were situated on the tops of hills, or slopes 
 at some height above the water of the stream, and in these 
 cases the washing was performed through the aid of rain 
 water. In several localities water was conducted to the wash- 
 
PROVINCE OF MINAS GERAES. 161 
 
 ings from streams. Some of the old regas, or ditches, are 
 still visible running for miles around the hills. The supply 
 of rain-water was of course sufficient only during the rainy 
 season, so that washing operations had to be suspended for 
 the rest of the year. In the old washings, as in that above 
 the cemetery at Minas Novas, or the Lavra da Santa Cruz, 
 at the junction of the Rios Fanado and Bom Successo, the 
 gravel lies in great piles. At the Arraial da Chapada the 
 same thing is seen, but there the whole tops of hills have 
 been deprived of their clay coating and washed over, so that 
 to-day they are hoary with the quartz boulders that remain, 
 the testimony of a departed industry. I was informed that 
 the custom was with the miners, as a general thing, to wash 
 the gravel on the spot. It seems wonderful that when the 
 washing was near a river or stream the gravel was not sent 
 down to this stream to be washed. To-day the washings, 
 though owned by private individuals, who to some extent 
 know their value, are unworked, the owners finding it more 
 profitable to pursue agriculture or wash for diamonds in 
 the Jequitinhonha. The abundance of gold over this region 
 may be seen from the nuggets in the possession of the peo 
 ple, and which have been picked up on the hillsides or in 
 rain-gullies. After rains one sees in the ravines the prints 
 of the feet of those who regularly go in search of gold 
 washed out by the surface waters, and in the streets of 
 Minas Novas and Chapada little dams are built across the 
 small rain-gullies by the children, to collect water to wash 
 the soil for gold, which they collect in quills, and larger 
 dams are built by the elder members of the population 
 for the same purpose. No one who has been over the 
 ground as I have, and has seen the irregular way in 
 which the mining has been performed, and the immense 
 
162 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 area of drift which has yet been untouched, drift rich in 
 gold, as the occasional recent washings testify, can doubt 
 that the region is far from exhausted. It has only been 
 forgotten. My friend Mr. J. S. Mills, of New York, an 
 excellent geologist, who has discussed these observations 
 with me, has suggested that the gold probably occurs in 
 bands in the drift, the direction of which might be worked 
 out by a careful topographical survey. Senator Theophilo 
 Benedicto Ottoni, of Rio, about two years ago obtained from 
 the Emperor a concession of the area of the Comarca of the 
 Jequitinhonha, to explore it for gold and other minerals ; 
 and an attempt, which we hope may yet be successful, has 
 been made to organize an American company for the pur 
 pose of thoroughly exploring and developing the gold-fields 
 of Minas Novas and vicinity. With modern mining meth 
 ods and appliances I have the fullest confidence that they 
 would prove very remunerative. The system of washing by 
 hose-pipes could be employed successfully in many local 
 ities. 
 
 Gold also occurs in the gravel and sands of the streams, 
 these loose materials being derived in part from the drift, 
 in part from the decomposed rock. Near the Arraial da 
 Chapada is an outlier of the tertiary called the Serra do 
 Macaco, which forms a very picturesque flat-topped moun 
 tain, with escarped sides, in which the horizontal layers of 
 red and white clays are beautifully exhibited. 
 
 Now that we know that gold may occur in any formation, 
 why may not the lower beds of this series be found to be 
 auriferous in some places ? 
 
 I have had no opportunity of making an examination of 
 the gold of the Minas Novas region, and I know of no 
 analyses ever having been made of it. 
 
PROVINCE OF MINAS GERAES. 163 
 
 The sands of the Arassuahy above the Rio Setubal, or 
 thereabouts, are rich in gold. I have never heard of their 
 affording diamonds. 
 
 From Calhao I took passage in a canoe, and descended to 
 the sea. On that voyage the following observations were 
 made on the rivers Arassuahy and Jequitinhonha. 
 
 At Calhao the Arassuahy is about as large as the Sao 
 Matheos. Its current is strong, and even during the dry 
 season it contains much water. The country on both sides 
 of the river below Calhao is, generally speaking, low and 
 uneven up to the foot of the chapadas, while varg-ens, more 
 or less wide, border the stream. These consist of alluvial 
 deposits, and afford a rich soil. Their height above the av 
 erage level of the river is about twenty feet. From Calha o 
 to the mouth of the river the country is sparsely settled. 
 The river-bed is much obstructed by ledges of slate, but there 
 are no rapids, and canoe navigation is not very difficult. 
 At the Pontal, at the mouth of the river, these slates appear, 
 from the canoe, to lie very flat, and to be traversed by heavy 
 veins of a crystalline rock, like granite, the outcrop of one of 
 which crosses just above the Pontal. At this place, in the 
 angle between the two rivers, is a little settlement, which 
 the inhabitants hope may one day rival Calhao in its com 
 merce in salt. 
 
 The traveller who has heard the Jequitinhonha constantly 
 spoken of by the Mineiros as a " majestoso rio" feels much 
 disappointed when he reaches it at its junction with the 
 Arassuahy, for it is but little larger than the latter river. 
 It is, however, much deeper. 
 
 Above the Arassuahy the Jequitinhonha flows in a wide 
 canon, separated from the valley of the Arassuahy by a 
 long, narrow chapada, which extends from the western 
 
164 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 limits of the chapada formation well down into the angle 
 formed by the union of the two rivers. The chapada 
 forming the eastern boundary of the valley of the Calhao 
 comes down into the corresponding angle on the other 
 side of the Arassuahy, so that that river really escapes into 
 the canon of the Jequitinhonha through a cut across the 
 chapadas. Below the mouth of the Calhao the river valley, 
 comprised between the chapadas, is quite wide, uneven, and 
 composed of mica slate, gneiss, <fec. Some six or eight 
 miles down the river there is a high hill, the Morro do 
 Ariao, which presents the smooth, black-stained, even rock 
 surface so characteristic of the gneiss hills of the coast. 
 
 Fifteen or twenty miles below the Calhao the little river 
 Piauhy enters the main river from the south. This stream 
 takes its rise in the Serra do Chifre, a short distance to the 
 north of the source of the Calhao, from which river it is 
 separated by a strip of chapada called the Chapada do 
 Piauhy, on the plains of which herds of cattle are pastured. 
 The Piauhy is noted for its affording grisolitas (peridote or 
 chrysoberyl) , pingoas d agua (white topaz or limpid quartz 
 pebbles), and other valuable stones, like those found in the 
 Rio das Americanas in the Mucury. The chrysoberyls are 
 used in jewelry and by watchmakers, and at the time of my 
 visit to Minas were selling for 11 $ 000 per pound, or about 
 $ 5.50 American currency. The demand of late years for 
 them has been very small. A few years ago, according to 
 Senator Ottoni, several hundred-weight were extracted and 
 exported, which drugged the market, and made it for a long 
 time unprofitable to wash for them. 
 
 Two miles below the Barra do Piauhy, the mica slates dip 
 to the N. 45 W., and at the Ilha do Cubango there are 
 heavy vertical veins of granite, which extend in walls almost 
 
PROVINCE OF MINAS GEEAES. 165 
 
 across the river, while, a couple of miles farther down, there 
 are some high gneiss or mica-slate hills. The river is full 
 of rocks, and the banks are rocky, though the banks of the 
 river are generally low. At the Arraial d ltinga the mica 
 slates still show themselves, with a strike of N. 45 E., and 
 vertical dip, and are traversed by granite veins. 
 
 The Arraial is a considerable little town, built on a ridge 
 of quartz gravel bordering the river on the northern side, 
 and which, being considerably higher than the river border, 
 itself about twenty feet high, is not covered during the 
 enchente. It derives its importance from its trade in salt, 
 which is brought up the river from the sea, and is sent into 
 the interior to the Sertao do Rio Pardo, together with mer 
 chandise, <fec. On both sides, but at a considerable dis 
 tance, the chapadas skirt the river, but they are rarely seen 
 by the voyager by canoe, because of the intervening gneiss 
 hills, which are sometimes 500 to 800 feet in height above 
 the level of the river. Just below Itinga one has a distinct 
 view of a chapada on the south side of the river, and in the 
 cliffs at its top the characteristic white rock is seen. The 
 height of the chapada top above the level of the river must 
 be over 1,000 feet. 
 
 The rocks exposed in the river-banks between Itinga and 
 the " Estreito " are gneiss, a compact variety. The hills have 
 the ordinary topography of the gneiss regions of the coast, 
 and often present bare, blackened cliffs and slopes. Back of 
 the hills the flat tops of the chapadas are seen, and occasion 
 ally they accompany the river. The slopes of the chapadas 
 invariably show gneiss almost to the top, where there are 
 usually lines of white cliffs. The thick red bed at the top 
 of the chapadas of Minas Novas I have not observed here, 
 nor is it to be seen in the cliffs of the chapadas at the June- 
 
166 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tion between the Arassuahy and Jequitinhonlia. The hills 
 and chapada slopes are thickly wooded, but the trees are all 
 small. A small species of Barrigudo (Imbare , Bombax, or 
 Chorisia)., with an enormously swollen trunk, is very abun 
 dant on the margin of the river. The course of the river is 
 rapid, and its breadth is about equal to that of the Parahyba 
 do Sul at Sao Fidelis. At the " Estreito " the river passes 
 through a narrow gorge across a gneiss ridge. This chasm 
 is in some places not more than 150 feet wide, and is a most 
 romantic spot. The sides are bold, rounded masses of rock 
 piled up one upon the other in picturesque confusion. 
 
 When the river is swollen, the "Estreito" is a fearful 
 place to pass ; the waters rush through with great fury, and 
 below it are dangerous whirlpools, where canoes are frequent 
 ly lost. Between the " Estreito " and the Pedra do Bode 
 the river-banks are low and flat, and the country behind 
 is often marshy and interspersed with shallow lagoons. In 
 one of these I found an abundance of Ampullarias, but I 
 could find no other shells. The Pedra do Bode, one of the 
 noted landmarks on the river, is a gneiss hill on the north 
 bank, presenting a smooth precipitous face to the river. It 
 is of some considerable altitude, but is not so high as the 
 chapada behind it. Thence to Sao Miguel the river is 
 bordered by gneiss hills and chapada spurs, and back of 
 these, on both sides of the river, are seen the level tops 
 of the chapadas which accompany the river. The hills 
 are often very abrupt, and present many bare surfaces. 
 Some, which may not be wholly composed of gneiss, 
 are very regular in their curves, and have steep slopes 
 covered with a low vegetation. The Indai4 palm is very 
 common on some of the hills, going to make up the 
 greater part of some of the woods. A short distance above 
 
PROVINCE OF MINAS GERAES. 167 
 
 Sao Miguel is the Caxoeira do Labyrintho, a series of rapids 
 extending for more than a mile. In some states of the 
 river these rapids are very dangerous, owing to the incli 
 nation of the river-bed and the numerous rocks which 
 obstruct the river, and canoes are wrecked and lives lost 
 here almost every year. In descending, the proeiros, or 
 oarsmen of the prow, row vigorously to give the canoe 
 a good headway, so that it may obey the steering-oar, 
 which must be handled very dexterously. Sao Miguel is 
 a miserable hamlet on the right bank of the river at the 
 mouth of the river Sao Miguel, which takes its rise in 
 the same serra with the Rio das Americanas. It is impor 
 tant principally because of its commerce in salt. There 
 are some large fazendas in the vicinity on both sides of 
 the river, on which very large herds of cattle are raised. 
 There are said to be some fertile lands here. Below Sao 
 Miguel is the Caxoeira de Dorma, a series of rapids usually 
 easily passed. On the right bank of the river, and some two 
 or three miles below Sao Miguel, is a range of irregular gneiss 
 hills, which have apparently a general north-south trend, 
 and present a precipitous front to the west. A small stream 
 springs from the top of one of these precipices, and hangs 
 a white thread of water against the black wall of rock. 
 Irregular gneiss hills occupy the right bank of the river for 
 some four and a half leagues. On the left bank the great 
 chapada stretches along, its sides descending with rounded, 
 smooth slopes, often destitute of forest and green with low 
 herbage. The stream here is full of rapids with a strong 
 current, and is about 500 feet wide. The scenery on this 
 portion of the river is exceedingly grand. Just above the 
 valley of Sao Simao is a little village called Farrancho, in 
 habited by civilized Machacalis. Below this the chapada 
 
168 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 slopes advance to the river-side and border it for a league 
 or more, forming the narrow valley of Sao Simao. The 
 chapadas are of great elevation, 1,200 feet or more above 
 the river ; their sides descend with steep, smooth slopes to 
 the river. The lower part of the slopes is thickly wooded, 
 but toward the top the vegetation generally becomes low 
 and scrubby ; in some parts, however, they are wooded to 
 the top. 
 
 The regularity of the slopes would justify one in assign 
 ing to the chapada formation here a great thickness. At 
 the entrance of the valley a very siliceous gneiss is seen 
 underlying the chapada. Leaving Sao Simao the chapadas 
 recede from the river, and the country thence to the Salto 
 Grande is gneiss. Immediately below Sao Simao are pic 
 turesque groups of hills, the Serra da Vigia on the right, 
 and the Serra das Panellas on the left, and below these 
 are the rapids of the Panellas. In descending we pass the 
 Serra do Feijoal on the left, and other hills on both sides 
 of the river, shoot the Caxoeira do Angelim, the Caxoeira 
 da Farinha, and other rapids, and reach the eastern extrem 
 ity of the fine Serra da Lua Cheia, which, coming from the 
 southwest, breaks down near the river. Between the hills 
 of the Feijoal and the Serra da Lua Cheia, the lands border 
 ing the river are generally flat and low, so low that they are 
 easily flooded by the enchente. The soils of these low 
 lands, or vargens, are in general composed of a fine sand 
 with but little admixture of clay. They appear to be very 
 fertile, as the vegetation they support is very luxuriant. 
 These flat lands are full of swamps and shallow lagoons, which 
 are flooded during the enchente, and are left full of water 
 when the freshet subsides. This water sometimes becomes 
 putrid, from the decay of the vegetables abounding in the 
 
PROVINCE OF MINAS GERAES. 169 
 
 swamps, and fills the air with miasma, while the water en 
 tering the river poisons it. It was in the beginning of May 
 that I descended the river. Between Sao Miguel, sezoes, or 
 fever and ague, were exceedingly prevalent, and I left all 
 my canoe-men sick at Salto Grande. There was scarcely a 
 house on the river where there were not cases of fever, and 
 canoes on the up voyage were delayed at the Salto and 
 elsewhere along the route because of the sickness of the 
 crews. This general prevalence of fever among the canoe- 
 men is principally attributable to their exceeding impru 
 dence. They drink freely of the Avarm muddy river-water 
 when overheated. They bathe in it under the hot sun, and 
 go with dripping garments a great part of the time, spend 
 ing frequently night after night under drenching rains with 
 no other shelter than a woollen blanket. I had risen from 
 a sick-bed to make the voyage, and was constantly exposed 
 to the rain and cold ; but I avoided the river-water, and es 
 caped, as did my fellow-traveller, a merchant from Calhao.* 
 
 The hills comprising the Serra da Lua Cheia are of con 
 siderable altitude, much broken up and very irregular in 
 outline. In this serra are several conspicuous needles visible 
 from a long distance, two of which are named respectively 
 the Enchadao and Enchadinho. From this serra to the 
 dangerous Caxoeira de Santa Anna, formerly called the 
 Caxoeira do Inferno, the river is very rapid, full of islands, 
 and there are some places difficult to pass. 
 
 The Caxoeira de Santa Anna is at all times so dangerous 
 that the cargo is always carried round the rapids, and re- 
 embarked below, the canoe descending empty. At the head 
 
 * I cannot let pass this opportunity of acknowledging the kindness of this 
 gentleman, Scnhor Baretto, a mulatto, who gave me my passage to the Salto 
 from Calhao, and was of the greatest service to me. 
 VOL. i. 8 
 
170 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of the rapids, which extend for nearly a mile, is a large isl 
 and. Both channels are practicable in some states of the 
 river, but with low water the northern is the only safe one. 
 The Caxoeira consists not only of a series of rapids, but also 
 of several bancos, or low cascades. Canoes constantly de 
 scend, a most exciting feat. The ascent is accomplished 
 only with the empty canoe and with great difficulty. Between 
 this Caxoeira and the Salto Grande the river is very swift, 
 full of rapids, and obstructed by rocks, while in some places 
 it is very narrow, and bordered by a wide margin of rocks 
 covered by the annual floods. Islands are numerous. At 
 the town of Salto Grande, a wretched little place, on the 
 right bank, a quarter of a mile above the Salto, or falls, 
 and celebrated for its trade in salt, &c., the river is only 
 eighty to one hundred feet in width, but on each side are 
 low margins of bare gneiss * rock and sand-banks. 
 
 At the Salto the river reaches a point whence, within the 
 distance of a mile or thereabouts, it descends some three 
 hundred feet, more or less,f in a splendid series of cascades 
 and rapids. At the head of the Caxoeira, when the river is 
 not swollen, it is suddenly narrowed to forty or fifty feet, 
 and plunges down a very steep incline into a gorge with 
 perpendicular banks, making a wild and most romantic fall 
 of about fifty feet. Below this are other falls, which, owing 
 to the state of the weather, I was unable to visit. On each 
 side of the rapids is a wide strip of rocky ledges, swept bare. 
 When the enchente prevails, the stream swells too big for 
 its channel, and pours in a terrible flood over the rocks 
 
 * This gneiss is composed of feldspar, quartz, and hornblende, and is well 
 bedded. Strike, N. 30 D W., dip vertical at the upper Porto do Salto. 
 
 t I am entirely unable, from the character of the country, to form a very re 
 liable estimate of the total height of these falls, but I believe that 300 feet is 
 much within the truth. 
 
PROVINCE OF MINAS GERAES. 171 
 
 on each side, making a series of rapids to which those of 
 Niagara are as nothing. The Salto Grande, during the 
 floods, must be a sight worth a pilgrimage to see. The 
 Diccionario Geografico says that the fall is twenty bracas 
 in height, and that the noise of the waters may be heard at 
 a distance of four leagues, which is not very correct. The 
 Salto consists of several falls and rapids, as above described. 
 The Caxoeira is of course an effectual barrier to navigation, 
 commerce requiring a transport of goods by mules around 
 it, which have to be re-embarked above or below the falls. 
 On the road from the village to the port below the coun 
 try is seen to be covered as usual with drift-clay, in which 
 are large boulders of the hornblendic gneiss, together with 
 rounded and angular fragments of quartz. Below the Salto 
 the river leaves the province of Minas Geraes, and enters 
 that of Bahia ; but to make my description of the river com 
 plete, I continue it here to the sea. 
 
 Between the Salto and the Caxoeirinha the river is nar 
 row, with high gneiss banks. It is much obstructed by 
 rocks and rapids ; but this part of the river I am unable to 
 describe in detail, because I was obliged to run the greater 
 part of it, rapids and all, in the night. At the Caxoeirinha 
 the river leaves the rocks, and becomes a no d areia. Up 
 to this point the canoes bring from the sea very heavy 
 loads of salt, &c., but here their cargoes have to be divided 
 and rearranged. Here has sprung up a little settlement, 
 which bears the same name as the rapids, but it is of no 
 importance. 
 
 The river, on leaving the rocks, becomes immediately shal 
 lower, less rapid, and widens into a fine broad stream, com 
 parable to the Doce, and from 800 to 1,000 feet in width or 
 more. The lands also grow lower, and the river valley is 
 
172 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 cut through the coast tertiary band. The whole country is 
 heavily wooded, but the vegetation did not bear to me the 
 luxuriant air of that of the Doce. There are a few settlers 
 along the river, and one or two large fazendas. At a place 
 called Zinebra, a few leagues above the mouth of the river 
 there is an old fazenda, with which is connected a good saw 
 mill. Below this a short distance, on the same or right 
 bank of the river, an American colony has been established, 
 and on the occasion of my visit I found two Southern 
 ers, Messrs. Ogclen and Thompson, engaged in cutting a 
 clearing in the forest. The locality they have chosen is 
 a fertile one, but it seems to me doubtful whether, single- 
 handed, they can ever succeed. Below Zinebra the tertiary 
 lands leave the river, an isolated patch being found on the 
 Po-assu, a channel on the north whence some of the waters of 
 the Jequitinhonha escape into the Pardo. Thence to the 
 sea, low alluvial lands, with a heavy forest growth and 
 swamps, border the broad, beautiful river. It is, however, 
 very shallow, and full of sand-bars. The river would be 
 navigable for a little flat-bottomed river steamer, but it 
 would have to be of very light draught. At the mouth the 
 river becomes exceedingly broad and shallow, and is to such 
 an extent obstructed by sand-bars, that the level of the 
 river is always higher than that of the sea, and the salt 
 water never enters, as is the case with most other rivers. 
 So heavily does the surf beat on the bar, that vessels enter 
 with great difficulty, and when once they have entered it 
 often happens that weeks or even months may elapse before 
 it may be safe to pass the bar again. Cargoes of corn laden 
 at Belmonte have often to be relanded after lying in the 
 hold of a vessel for weeks. Nor is this all. The sand-banks 
 are constantly shifting, and a vessel at anchor may be heaped 
 
PROVINCE OF MINAS GERAES. 173 
 
 round by sand and detained for a long while. The result 
 is, that the port is rarely ever resorted to by coasters. 
 Belmonte is a little town situated on the alluvial border 
 of the river, in a grove of cocoanut-trees, on the right 
 bank, a short distance above the mouth. During the fresh 
 ets it is liable to suffer from the eating away by the river 
 of the bank on which the town stands. It is of scarcely 
 any importance, doing very little trade, its inhabitants being 
 principally fishermen. Cattle are raised on the plains of the 
 vicinity, but there is small opportunity for agriculture. 
 
 From Caxoeirinha to the fazenda of Zinebra I saw next to 
 nothing of the geology, owing to a part of the journey having 
 been made in the night, and because of the prevalence of 
 very heavy rains ; but near Zinebra I saw a small exposure 
 of shales, w r hich appeared to be of the same character as 
 those hereafter to be described in speaking of the Rio Pardo, 
 but owing to the height of the river I could make nothing 
 of them. The Diccionario Geografico says, in speaking of 
 the river, that in 1840 beds of rose-colored marble were dis 
 covered. Through the kindness of Senhor Piraja I have in 
 my possession a specimen of this marble. It is exceedingly 
 fine in texture, and of a delicate pink tint, compact and 
 hard, and would take a fine polish. If it occurs in sufficient 
 quantities, it would make a beautiful building-stone. For 
 the present let us leave the Jequitinhonha. When treating 
 of the geology of the Province of Bahia we shall have to 
 return to it again. 
 
174 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER IV. 
 
 THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 
 
 The Geology of the Abrolhos. Trap-bed, Fossil Plants, &c. Land Fauna and 
 Flora ; Spiders, Lizards, and Sea-Birds. The Cemetery of the Frigate-Birds. 
 The Whale and Garoupa Fisheries. Importance of these Fisheries. 
 The mythical Brazilian Keef. The Coral Reefs and Consolidated Beaches 
 confounded by Travellers and Writers. The Author s Discovery of the 
 Porto Seguran Coral Keef. Coral-building Corals found almost wholly to 
 the north of Cape Frio. The Fringing Reef of Santa Barbara ; its Structure 
 and Life. Corals found on the Reef. Star-fishes, Ophiurans, &c. 
 Resemblance between the Echinoderms of the Abrolhos and West Indies. 
 The Chapeiroes. The Parcel dos Abrolhos; its Appearance; forms a 
 serious Obstacle to Navigation. Safe Canal west of the Islands. The 
 Parcel dos Paredes. The Recife do Lixo. Its great Extent. The Sub 
 merged Border and its Coral-Growth. The Coral Fauna of Brazil. The 
 Millepores and their Stinging Properties. The Reefs of Timbebas, Ita- 
 columi, Porto Seguro, Santa Cruz, Camamu, Bahia, Maceid, and Pernam- 
 buco. The Roccas. 
 
 THE islands of the Abrolhos * lie about midway between 
 the cities of Rio and Bahia, a little south of the parallel of 
 
 * The general impression seems to be that the name is derived from the Por 
 tuguese words meaning, " Open your eyes," a name which would be exceed 
 ingly appropriate, for the islands, whitened by the dung of sea-birds, have a 
 spectral look, and, in addition, the reefs with which they are surrounded are so 
 dangerous, that, before the lighthouse was erected, it required much vigilance 
 to enable vessels to pass them in safety, and they have been always justly 
 dreaded. The author of the odd old Dutch Reys-boeck van het rijclce Brasilien, 
 published in 1624, says that they are very periculeus, and adds : " Daerom als 
 hy dese passagien passeren willen so nemen sy eerst met al haer vole hct Sac 
 rament ende wanneer sy die ghepasseert hebben bedrijven sygroote blijdtschap- 
 ghelijck al by alle Jurnalen soo wel vande spaensche als van de onse te sien is ! " 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 175 
 
 ISLAND OF SANTA BARBARA DOS ABROLHOS. 
 
 Caravellas, and at a distance of about forty miles from the 
 mainland. The position of the lighthouse on the island of 
 Santa Barbara is, according to Mouchez, lat. 17 57 31" 
 S., long. 40 58 58" west from Paris. These islands are 
 situated apparently near the middle of the submerged bor 
 der of the continent, which here, over a very large area, lies 
 at a depth of less than one hundred feet. They are four in 
 number, with two little islets, and they are arranged in an 
 irregular circle, three of them close together. All are rocky 
 and rather high, Santa Barbara, the principal one, being 
 33.22 metres in height. The length of this island is about 
 three quarters of a mile. Its outline is irregular, and 
 it is very narrow. It is composed of beds of sandstone, 
 shales, and trap, which dip approximately north-northwest, 
 at an angle of from ten to fifteen degrees. Owing to this 
 
 Captain and crew took the Sacrament before passing them ! The name, how 
 ever, means rocks, and is so defined in Fonseca s Dictionary. There is a little 
 group of reefs and islands lying on the western coast of Australia, in lat. 28 
 S., and known as Houtman s Abrolhos. These are, in great part at least, 
 composed of coral 
 
176 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 northward dip of the strata, the northern side of this island 
 presents a steep slope to the sea, while on all other sides it 
 is precipitous. The island is almost divided in two in the 
 middle by a cove indenting it on the southern side. 
 
 THE ISLANDS OF THE ABKOLHOS FROM THE SOUTH. 
 
 In the cliff below the lighthouse, the lowest beds seen are 
 an arenaceous limestone (?), a, of the following diagram, 
 
 a rock so hard as to form a platform below the cliff. Over 
 this is a hard, blue shale, 6, with undeterminable organic 
 markings, some of which appear to be the scales of Teleos- 
 tian fishes. This is again overlaid by a thick bed of a yellow 
 ish sandstone, c, rather fine in texture, and sometimes more 
 or less shaly, on the surfaces of some of the layers of which 
 there are obscure impressions of plants. This sandstone is 
 harder than the underlying shales, and so forms an over 
 hanging cliff. The sandstones are overlaid by a bed of 
 basaltic trap that occupies the greater part of the surface of 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 177 
 
 the island, as is seen in the following little map, in 
 which the darkly shaded portion represents the trap-bed. 
 
 r 
 
 This trap-bed is divided by numerous joints into polyhedral 
 masses of all sizes. On the upper surface of the bed these 
 masses have lost, by decomposition, one concentric coat 
 after another, until, in some cases, they have been rounded 
 down to irregular spheres, like cannon-balls, and the greater 
 part of the island is strewn with these 
 boulders of decomposition. I have al 
 ready called attention to similar boul 
 ders at Tijuca. The trap of Santa 
 Barbara is traversed by but few veins, 
 so far as I could see, the only mineral that I could find be 
 ing chalcedony, incrusting cavities, and having the color 
 and general appearance of Prenhite. Cracks and crevices are 
 sometimes filled up with guano and phosphates from the 
 dung of the sea-birds that frequent the islands ; but I saw 
 no regular deposit of guano. The surface of the rocks is 
 sometimes covered in patches by an incrustation of a hard, 
 brownish substance, which I have supposed to have been 
 deposited by the surface waters, and to have been derived 
 
 8* L 
 
178 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 from the birds dung. Darwin, who visited the islands on 
 his celebrated voyage round the world, also speaks of it, and 
 describes a similar substance as found on the island of 
 Ascension, and on St. Paul s rocks.* The same geologist 
 mentions having observed a columnar structure in the trap 
 of Santa Barbara, but I find no note of it in my journal. 
 The underlying stratified beds are somewhat altered, and 
 indurated from their proximity to the trap. 
 
 In lithological characters the Abrolhos beds resemble the 
 sandstones, <fcc., of the Rio Sao Francisco at Penedo, to be 
 described farther on, and which contain similar plant re 
 mains. They have been disturbed by the same upheaval, 
 and I have little hesitation in referring both to the creta 
 ceous. 
 
 If my identification be correct, it is interesting to observe 
 the cretaceous rocks on the eastern border region of South 
 America disturbed and associated with volcanic deposits, for 
 along the eastern border region of North America the cre 
 taceous and tertiary rocks have suffered no disturbance. I 
 have seen no trace of secondary rocks on the coast oppo 
 site the Abrolhos and to the southward. On the Mucury 
 and elsewhere the tertiary clays are everywhere seen to rest 
 immediately on the gneiss ; but the submerged border of the 
 continent seems to be more or less overlaid by cretaceous 
 strata, as in the eastern border region of North America. 
 
 The Ilha Redonda, lying just west of Santa Barbara, is 
 composed of rocks of the same character as those of Santa 
 Barbara, but I observed no basalt. Near the top of the cliff, 
 on the eastern side of Redonda, there is seen a thick bed of 
 a white or yellowish material which looks like chalk, and is 
 easily cut with a knife when wet, but on drying it grows 
 
 * Darwin, Geological Observations, Part II. p. 33. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 179 
 
 harder. It does not, however, effervesce with acids, and it 
 appears to be an aluminous product of the decomposition of 
 some rock. Mr. Henry Hughes, of Cornell University, ex 
 amined a specimen of the rock, and reports it as containing 
 quite a percentage of phosphoric acid, which has doubtless 
 been brought into the rock by the percolation through it of 
 rain-water from the dung-strewn surface of the ground above. 
 The other two islands are composed of stratified rocks which 
 appear to underlie conformably those of Santa Barbara and 
 Redonda ; but I was unable to visit them. On the shores of 
 Santa Barbara I found fragments of pumice scattered about 
 and much rolled by the waves. These have been observed 
 elsewhere on the Brazilian coast, and it is somewhat puz 
 zling to account for their origin. Darwin found pumice 
 pebbles on the coast at Bahia Blanca, in the southern part of 
 the Argentine Republic ; but these, he says, had been brought 
 down by the rivers flowing from the cordillera.* 
 
 The beaches of the Abrolhos Islands are formed in part 
 of the debris of the rocks comprising the islands, but they 
 consist largely, and in some places entirely, of coral and 
 shell sand. It is very interesting to see how these materi 
 als are cemented together through the action of the sea 
 water, and even the shingle is soldered into an exceedingly 
 firm mass.f 
 
 * Darwin, Geological Observations, Part III. p. 4. 
 
 t I have seen the same in the island of St. Thomas, W. I. Darwin says : 
 " On the shores of Quail Island in the Cape Verdes I found fragments of 
 brick, bolts of iron, pebbles, and large fragments of basalt united by a scanty base 
 of impure calcareous matter into a firm conglomerate. To show how exceedingly 
 firm this recent conglomerate is, I may mention that I endeavored with a heavy 
 geological hammer to knock out a thick bolt of iron which was embedded a 
 little above low-water mark, but was quite unable to succeed." Geological 
 Observations, Part II. p. 21. 
 
180 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 On the island of Santa Barbara was erected a few years 
 ago an excellent lighthouse with a flash light, and the only 
 human inhabitants of the island are the lighthouse-keeper 
 and his assistants. A few goats were introduced on Santa 
 Barbara some time since, which have multiplied until there 
 is now a flock of several hundreds. These animals have 
 almost deprived the island of vegetation, and can now only 
 barely subsist. Redonda is covered with coarse grasses, with 
 dwarf mimosas and a few ferns, &c., a very meagre flora. 
 The island Siriba boasts in addition a single tree of the 
 same name, together with two dwarf cocoa-palms planted by 
 the whalers from Caravellas. The land animals consist of 
 little lizards of several species, which are extraordinarily 
 abundant, more so than in any other locality I ever visited. 
 An immense Myg-ale, the Aranha caranguejeira of the Bra 
 zilians, abounds in like manner, living under stones, but I 
 did not observe that it made any nest. This huge spider 
 preys on the lizards. It has been known to attack and kill 
 young chickens on the island and suck their blood, and it 
 is not improbable that it may destroy the young of the 
 sea-birds, so common on the island. 
 
 Breeding-places for sea-birds are few along the Brazilian 
 coast north of Cape Frio, and during certain seasons of the 
 year several species resort to the Abrolhos in great num 
 bers. Among these are the frigate-bird ( Tachy petes aqui- 
 lina), the Piloto, the Grazina (Phaeton), Beneditos, gulls, 
 &c. Since the occupation of the island by men and goats, 
 and the establishment of the lighthouse, these birds have 
 resorted to the island of Santa Barbara less abundantly than 
 formerly. One fact with reference to the frigate-bird is 
 worth mentioning. At the southwest extremity of Santa Bar 
 bara is a little islet composed of a heap of large trap boulders 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 181 
 
 of decomposition, and joined to the main island, as I shall 
 have occasion hereafter to remark, by a fringing coral reef. 
 This islet, whitened by the bird-dung, is called " O Ceme- 
 terio" or the cemetery. I was assured that to this sepul 
 chral-looking spot the frigate-birds of the vicinity resorted 
 on the approach of death, and that the place was strewn 
 with their bones. At low water one day I visited the 
 " cemetery," and I found such to be actually the case. 
 There were remains of hundreds of these birds, some freshly 
 dead, but the most of the skeletons were disarticulated and 
 bleached. Nowhere else did I see a dead frigate, and it 
 would seem that for generation after generation they 
 had gone there to die. I do not know of a better station 
 for an ornithologist desirous of studying the habits and 
 embryology of the sea-birds of Brazil than the island of 
 Santa Barbara. He can find as comfortable lodgings at the 
 lighthouse as he could possibly desire, and he may at the 
 right season of the year collect as many specimens, young 
 and adult, of the birds frequenting the island as he may 
 wish. 
 
 Before I go on to speak of the coral reefs of the Abrolhos 
 and vicinity, a few remarks on the fisheries may not here 
 be out of place. On the coast of Brazil are found several 
 species of whales and smaller cetaceans, but these animals 
 have not been carefully examined by competent naturalists, 
 and I am unable to give as accurate an account of them as 
 I could desire. These animals are captured at various sta 
 tions, from Santa Catheriiia northward to Bahia. At pres 
 ent the two most important stations are Caravellas in the 
 Abrolhos region and Bahia. I visited the Abrolhos during 
 the whaling season, and during my cruise I saw several hump 
 backs (Meg-aptera*), all apparently of the same species; but 
 
182 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 I was not so fortunate as to see a fresh specimen brought 
 in. I made the most diligent inquiries of the fishermen in 
 relation to the different kinds they were accustomed to take, 
 but they seem to confound the species, and I could obtain 
 nothing very satisfactory from them. 
 
 From all that I could learn, three species are taken in 
 the vicinity of the Abrolhos. The noruega is a humpback, 
 which has the belly white and smooth, back very dark 
 bluish, length fifty to fifty-five feet. This whale gives more 
 oil than the mystipa, which the whalers said differed from 
 the above in having the back black, and the belly and throat 
 furrowed. Sometimes they have white spots on the sides. 
 
 The ca$el6 (cachelot) is distinguished by being wholly 
 dark colored or black, and without spots or furrows. The 
 fishery begins at Bahia, according to Castelnau,* about the 
 13th of June, and lasts until the 21st of September. At 
 Caravellas I was assured that the whales always appeared 
 later than at Bahia, and the fishery does not begin until 
 the last week in June,f continuing through the month of 
 September. This seems strange, since the whales, ac 
 cording to the fishermen, come from the south in June, 
 and return in September, and one would naturally expect 
 that they would arrive at the Abrolhos before they arrive 
 at Bahia. The first whales appear in the Abrolhos waters 
 at about the end of May, and they linger until October. 
 The females often bring young calves with them, and ap 
 pear to seek the shelter of the reefs. The head-quarters of 
 the Abrolhos fishery is at Caravellas, or rather at the 
 mouth of the river Caravellas, where are situated the ar- 
 ma^oes, or trying-houses. In the year 1867 there were 
 
 * Expedition dans I Amfrique du Sud, Tome I. p. 1 50. 
 t So also Von Tschudi. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 183 
 
 employed in this fishery seventeen launches. These ves 
 sels are large, well-made, pink-sterned, open boats, carry 
 ing one large square sail. The same build of launch is used 
 at Bahia, and they are very good sailers. One of these 
 launches costs, the hull alone, about 500 8 000 (8 250, more 
 or less).* The crew consists of from fourteen to sixteen 
 persons. Each launch takes in tow a whale-boat. These 
 are of the ordinary build, and cost about 230 8 000. The 
 crew of the boat consists of seven men, a harpooner and a 
 steersman, the rest being rowers. The launch and boat 
 usually belong to one person, who hires the crew for the 
 season, or safra, furnishing them rations every ten days. 
 When a large whale is captured the harpooner receives 
 120 8 000, but if small only half that sum. The steersman 
 receives half the prize money of the harpooner, the rowers 
 each 248000 if the whale be large, and 128000 if small. 
 The master of the launch receives 408000 if large, 20 8 000 
 if small, and the crew 168000 if large, 88000 if small. 
 The dead whale is towed in to land by the launch, aided by 
 the boat if necessary. The distance is often great, and 
 when the wind is adverse the whale often comes to land in 
 a wretched condition, and frequently badly mangled by 
 sharks, which abound in these waters. A small tug would 
 be of much service in this fishery, not only to .bring the 
 whales promptly to the shore, but occasionally to tow 
 the launches in case of a head wind. The whales are 
 brought to the shore, beached in front of the trying-houses, 
 and cut up. There are several of these trying-houses. 
 The one that 1 visited was well constructed, and was fur- 
 
 * The reader will please bear in mind that the 1 $000 or mil-reis (not mil- 
 rei, as foreigners will call it), has an approximate value of about fifty cents 
 American currency. 
 
184 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 nished with five cutting-tanks, which would accommodate 
 the blubber of two large whales, together with ten tanks 
 for oil having a capacity of about 15,000 gallons. There 
 were twenty-six trying-pots. 
 
 The armagoes are hired by the owner of the launch cap 
 turing the whale, who furnishes the men necessary for the 
 cutting up and trying out. The large females accompanied 
 by young, madrijosf are very fat, and are the most valuable 
 prizes. There are killed every year and cut up at the Ponta 
 da Balea from thirty to ninety whales ; but were the fishery 
 pursued more vigorously, with proper economy of time and 
 the use of a small steam-tug or two, the yield might be 
 more than doubled. I learn through M. Bornand, of Villa 
 Yigosa, that a company has been formed at that town for 
 the prosecution of the fishery. A good-sized whale ought to 
 afford 1,000 to 1,600 canadas of oil, the Canada containing 
 about ten bottles, or one and eight ninths gallons, the large 
 whales giving much more. The oil, whose quality might, 
 it seems to me, be improved by more care in the trying out, 
 sells on the spot at from 1 $ 600 to 3 $ 000 per Canada. 
 
 The whalebone is short, but sells well, but I have omitted 
 to note the price it brings. The beach on which the whales 
 are cut up is covered during the season by huge masses of 
 rotting flesh, and is strewn with bones. There must be 
 on the spot the bones of over 500 whales. These, with 
 the flesh and the refuse from the trying-pots, would, prop 
 erly and scientifically prepared, make an excellent manure, 
 which, if judiciously applied, would go far towards re 
 juvenating the soils of the plantations of the vicinity, which 
 are rapidly becoming exhausted. At present it seems 
 ridiculous to hear the complaints of the planters, while 
 
 * Madrijas ? I do not find the word in Fonseca. 
 
THE ISLANDS AND CORAL BEEFS OF THE ABROLHOS. 185 
 
 hundreds of tons of the most valuable kind of manure are 
 left to rot on the sands or are poured into the sea ! The 
 Abrolhos region is rarely visited by whalers, though I was 
 informed that an American vessel some nine years ago 
 spent a season on the ground, taking twenty whales. 
 
 The fishery at Bahia is carried on on a much larger scale 
 than at Caravellas. Castelnau* estimated in 1850 that it 
 gave occupation to 2,000 persons, and from 100 to 120 
 boats, giving a revenue of 200,000 francs. The same author 
 estimated that on the whole coast of Brazil from 10,000 to 
 12,000 persons were engaged in this fishery, and that it 
 produced a capital of 1,000,000 francs, but it seems to me 
 that that estimate would be far too high for the present 
 time. Castelnau speaks of the fact that whale-flesh is 
 used as food by the lower classes in Bahia, and I saw 
 it exposed for sale ; but Dr. Antonio de Lacerda assured 
 me that it was not healthy, and tended to produce ele 
 phantiasis. Castelnau states that, according to the fisher 
 men, the whales enter the bay every morning, but always 
 return to the open sea to spend the night, and I heard 
 the same report. Whales are frequently taken very near 
 the city, and one may sometimes enjoy the rare sight of 
 sitting at a restaurant in the upper city and watching the 
 chase and capture of a whale in the bay below ! 
 
 The other fishery carried on in the waters of the Abrolhos 
 is that of the garoupa, an excellent fish exceedingly abun 
 dant, and taken with the hook and line. The head-quarters 
 of this fishery is at Porto Seguro, a town some seventy 
 miles to the northward of the Abrolhos. At this town is 
 owned a fleet of thirty-five or forty small vessels, each car 
 rying from seven to ten men. The fishery really extends 
 
 * Expedition dans I Amerique du Sud. Histoire du Voyage. Tome I. p. 152. 
 
186 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 from Barra Secca northward to the Commandatuba, but the 
 best grounds lie between lat. 17 and 18 S. The cruise is 
 usually for twenty-five or thirty days. The fish taken are 
 principally g-aroupas, but there are also several other kinds, 
 such as the meiro, vermelho, &c. The fish are salted down 
 in the hold, but, owing to the heat, they arrive almost in 
 variably with a very strong and disagreeable odor. They 
 are dried on shore and sent to Bahia. The yearly product 
 of this fishery is from 160,000 to 200,000 arrobas (2,560 to 
 3,200 tons). The garoupa is a delicious fish, and with 
 proper care might be prepared so as to be quite equal 
 to the cod. The names of the fish taken in the vicinity 
 of the Abrolhos, and which are used for food, are legion, 
 and among them are some of the most delicious of ma 
 rine fish. The Abrolhos Islands offer an excellent station 
 for drying and curing fish, and there has been some talk 
 of establishing there the head-quarters of a company to 
 carry on this fishery. Immense quantities of codfish are 
 now sent to Brazil, together with European sardines and 
 canned fish from Portugal, and every venda is full of them. 
 Some enterprising American should form a company for the 
 development of this fishery. The Brazilian fish are as 
 cheaply obtained as the Portuguese, they are nowhere to 
 be excelled, and Brazil ought to be exporting her delicious 
 fishes, canned, or otherwise prepared, to Europe, besides 
 supplying her own market. The government would favor 
 any undertaking of the kind proposed, and there are wealthy 
 Brazilians who would aid in carrying it out. 
 
 In works on Brazil, from those of the old explorers to the 
 present time, we find the uniform statement that a reef or 
 consolidated beach, like that of Pernambuco or Barra Secca, 
 extends around the greater part of the Brazilian coast. 
 
THE ISLANDS AND CORAL EEEFS OF THE ABROLHOS. 187 
 
 There has been much confusion as to what this reef really 
 was, some describing it as of coral, others as composed of 
 sandstone, but in scientific works it is generally stated that 
 no coral reefs exist on the coast of Brazil. 
 
 I am not sure who first expressed the opinion that the 
 stone reef surrounded a large part of the coast, but I find 
 it in Piso, whose first volume bears date 1648.* Since 
 his time this general reef has been described over and over 
 again almost in the same words, and it is even occasionally 
 to be found laid down on maps. 
 
 Prince Max. zu Neu Wied has nothing to say concerning 
 the true coral reefs, and, strangely enough, he does not 
 describe the Porto Seguran or Santa Cruz consolidated 
 beaches, notwithstanding he gives drawings of both. Von 
 Martius,t however, observed coral banks at Camamu and 
 near Ilhdos, and referred some of the corals to Lamarckian 
 species. 
 
 Darwin, who just touched at the Albrohos, observed 
 corals growing on the shore, but he did not see the reef. In 
 his Geological Observations J he says : " Round many inter- 
 tropical islands, for instance the Abrolhos on the coast of 
 Brazil, surveyed by Captain Fitz Roy, and, as I am informed 
 by Dr. Gumming, round the Philippines, the bottom of the 
 sea is entirely coated by irregular masses of coral, which, al 
 though often of large size, do not reach the surface and form 
 proper reefs." Darwin speaks also of having received infor- 
 
 * Piso says : " Maximam Brasilia partem, nuno interrupts nunc continuato 
 ducto tuctus. Ejus latitude planissima cst et quasi arte in superficii levigata ad 
 viginti, subinde triginta passus et ultra se extcndit. Tantas vero altitudinis ut 
 vix summo aestu inunditur." Hist. Nat. Brasilice. Guilielmi Pisonis, M. D. 
 de Med. Brasil. Liber primus. 1648. 
 
 t Reise nach Brasilien, Band II. Seite 684, 685. 
 
 J Part I. p. 58. 
 
188 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 mation of the coral reef at Maceio, which further on I shall 
 describe ; and in another place in the same work, referring 
 to the Pilote du Bresil, by Baron Roussin, a work I have 
 never seen, he says : " A few miles south of the latter city 
 [Pernambuco] the reef follows so closely every turn of the 
 shore that I can hardly doubt that it is of coral." Dana 
 also says : * " About Pernambuco, as I am informed by Mr. 
 Titian R. Peal, there are some patches of growing corals, 
 and they are said to extend along to 20 or 21 south lati 
 tude," which is not quite correct, as we shall see further on. 
 Staff-Commander Penn,f in treating of Cape Sao Roque, 
 says that " the coast of Pititinga and the Cape is skirted 
 by a reef which, between two and two and a half miles 
 southward of the former, in front of two small villages, forms 
 a curve with its outer edge and runs thence a mile from the 
 shore, having two and three quarters fathoms of water inside 
 of it." These reefs are represented on the map of Rio 
 Grande do Norte by Almeida, and they appear to be coral 
 reefs. Penn speaks of other reefs between the Punalm and 
 the Touro which are of the same character. A little farther 
 on J he says : " The recife, a singular ridge of coral rock, 
 borders the coast, generally at a distance of about a half to 
 three miles, but in some places much farther off, and ex 
 tends more or less from the northeast part of Brazil as far 
 as Bahia. Traces of it may be found farther southward 
 and along the north coast to Maranhao. The reef, which is 
 about sixteen feet in breadth at the top, slopes to the sea 
 ward, is perpendicular on the shore, and said to be generally 
 covered, but sometimes rises from distance to distance 
 
 * Coral Reefs and Islands, p. 108. 
 
 t South American Coast Pilot, Vol. I. p. 22. 
 
 J Op. cit. p. 25. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 189 
 
 nearly three feet out of water. It is nearly always bordered 
 by rocky banks, and forms a natural breakwater, having 
 smooth water and shallow inside of it, with navigable chan 
 nels for coasters, &c. It is broken occasionally, and forms 
 by the openings entrances to the greater part of the ports, 
 rivers, and creeks on the coast." Now such a description 
 could never have been written by any intelligent seaman 
 who had examined the coast. It is Piso s account of the 
 recife told over again, and it is the more erroneous since it 
 gives more detail. 
 
 Gardner not only mistook the structure of the Pernam- 
 bucan stone reef or consolidated beach, but he describes the 
 mythical coast reef in the same general terms. So no 
 wonder that the whole structure and character of the coast 
 reefs of Brazil have remained a puzzle to the geologist and 
 the geographer, and that it has been a serious question as to 
 what the Abrolhos reefs really were, one author declaring 
 that they were formed of decomposed gneiss ! The fact is, 
 that the reefs of Brazil are of two kinds, the coral reefs 
 and the consolidated beaches, which last are occasionally 
 separated from the coast line, and sometimes run across 
 the mouths of rivers, as at Porto Seguro, Pernambuco, <fcc., 
 like narrow rock walls resembling artificial breakwaters. 
 These, so far as I have observed or can learn, are never 
 found at any great distance from the shore, neither are they 
 continuous over any great distances. The Brazilian calls 
 them recifes. That at Pernambuco, owing to the great 
 trade with that port, has become famed, and many travel 
 lers have seen it and been puzzled over it. 
 
 From the Abrolhos northward to the shore of Maranhao, 
 at very irregular and often very long intervals, are scat 
 tered true coral reefs, which lie in patches at a short dis- 
 
190 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tance from the shore, there being usually navigable channels 
 between them and the mainland. It is very rare that one 
 of these reefs is dry other than at very low tide, and the sea 
 constantly breaks on its outer edge. These reefs are known 
 by the Brazilians as recifcs* Coral and coral-rock are 
 called pedra de cal, or limestone. The whole confusion has 
 evidently arisen in this way : A traveller has visited Pernam- 
 buco, and has seen the reef. He hears it called the recife, 
 and is told that the coast of Brazil is bordered by recifes. 
 On his way up or down the coast he sees from time to time 
 the sea breaking against the coral reefs in a long line of surf. 
 His pilot tells him that is the recife. He perhaps asks if it is 
 made of coral ; but this word in Brazil is almost exclusively 
 applied to the precious red coral (Corallium rubrum ), and 
 the pilot says, " No, it is made of pedra, or pedra de cal" 
 Some of the coral reefs are laid bare at low water, but their 
 great width is not visible from the deck of a ship sailing at 
 a distance, and they look like walls. So the coral-reefs of 
 Brazil have come to be confounded with the consolidated 
 beaches ; indeed, I should never have suspected the real 
 character of the coral reef of Santa Cruz, close to which I 
 passed in a steamer, had it not been that I had previously 
 examined the coral reef at Porto Seguro. It seemed a low, 
 narrow wall, and there was nothing that I could elicit from 
 the pilot or captain that would have led me to suppose 
 that it differed from the inner consolidated beaches at Porto 
 Seguro or Santa Cruz. 
 
 I made my first acquaintance with the coral-reefs of 
 Brazil while at Porto Seguro, in 1866. I had been for 
 
 * This word, as I shall show further on, is derived from the Arabic word 
 razif, which means literally a pavement. Sometimes the form arrecife is used. 
 Compare reef, Eng., riff", Germ., and rtdf, French. See Recife in Index. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 191 
 
 several days collecting on the stone-reef or consolidated 
 beach, before I heard of the outer reef. There was nothing 
 that I could learn from the fishermen that could warrant 
 me in considering it as anything else than a consolidated 
 beach ; but my studies of the latter class of reefs had satis 
 fied me that the outer reef could not possibly be of that 
 character, and when, on a spring tide, I visited it, in com 
 pany with Mr. Copeland, I was not astonished to find it 
 composed of coral. On that short visit I collected all the 
 principal corals found on the coast, and made out quite 
 satisfactorily the general structure of the reef, and of the 
 chapeiroes which surround it. I soon felt satisfied, from all 
 that I could learn from the garoupa fishermen, that the Ab- 
 rolhos rccifes were true coral reefs, and my companion and 
 I were on the point of visiting them when we received let 
 ters from Professor Agassiz, desiring us to come immedi 
 ately to Rio, to return home with the expedition. 
 
 At Rio I found Mouchez s chart of the Abrolhos, on 
 which is a note describing the reefs of the Abrolhos so 
 clearly as to leave no doubt as to their being immense coral 
 reefs. In order to settle the question I returned to Brazil 
 the next summer, and went over the reef grounds of the 
 Abrolhos as thoroughly as my time and the slim means I 
 could command would allow.* 
 
 The coast of Brazil, north of Cape Frio, has quite a rich 
 polyp-fauna, but very few of the madreporian polyps cross 
 the southern tropic. The Bay of Rio offers only insignifi 
 cant representatives of that order. All the specimens I 
 could obtain were Astrangia3, growing in small scattered 
 
 * I regret exceedingly that I have not yet been able to use the dredge on the 
 Brazilian coast; but I hope that my studies of Nature in the tropics are only 
 the preface to more thorough and detailed explorations in the future. 
 
192 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 cells, on stones and dead shells in the shallow water off the 
 Ilha do Governador. Professor Agassiz tells me that a fine 
 species of Porites had been collected at Rio, and he also 
 informs me that corals have been found at Desterro, in the 
 bay of Santa Catharina,* a locality which, though extra- 
 tropical, so far as latitude is concerned, is not really so in 
 the character of its climate. Many species of Actinias are 
 found in the harbor of Rio, together with one species of that 
 curious locomotive halcyonoid, Renilla (R, Dance Verrill). 
 On the masonry of the new Custom-House docks at Rio I 
 collected in abundance a slender, branching, tender, nodose 
 halcyonoid undetermined. 
 
 As we go northward from Cape Frio the madreporians 
 become quite common on the rocky shores, though the spe 
 cies are not numerous, and they are associated with species 
 of Millepora, Zoanthus, and Palythoa, and various gorgo- 
 nians. I have already called attention to the coral fauna 
 of Guarapary and Victoria, and I have stated that I have no 
 evidence of the existence of any banks of living corals or 
 reefs south of the region of the Abrolhos. Here the condi 
 tions for the growth of coral reefs on a large scale are re 
 markably favorable. Over large areas the water covering 
 the great submarine shelf, on which the islands are based, 
 is much under one hundred feet in depth, and it is warm 
 and pure. So it is not to be wondered at that very large 
 coral reefs, both fringing and barrier, are found here. 
 
 When the tide goes out there is seen extending round 
 about one half the circumference of the island of Santa 
 Barbara a fringing reef, shown in the little map on page 
 177. f One may then walk out on its level surface as on a 
 
 * Collected by Mr. F. Miiller. 
 
 t This reef is also represented in the cut on p. 175. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 193 
 
 wharf, and from its ragged edge look straight down through 
 the limpid green water and see the sides of the reef and the 
 sea bottom covered with huge whitish coral-heads, together 
 with a wealth of curious things not to be obtained without 
 a dredge. 
 
 The surface of the reef, though flat, is somewhat irregu 
 lar. It rises but a short distance above low-water mark, 
 and it is overgrown with barnacles, shells, mussels, and 
 serpula-tubes, together with large slimy patches of the 
 common leather-colored Palythoa. The reef abounds in 
 small pools, some shallow and sandy, others deep, rocky, 
 and irregular. The former often contain scattered masses 
 of corals, particularly Siderastraia and Favia, and they 
 abound in small shells, crabs, OphiurcB, &c. ; but the deep 
 pools are the richest in life. These are usually heavily 
 draped on the sides with brilliantly tinted sea-weeds and 
 corallines, the bare rock being gay with bryozoa and hy- 
 droids. The most common coral of these pools is Sider- 
 astrcea stellata Yerrill.* This is a coral growing in round 
 ed or hemispherical masses with small cells. Professor 
 Verrill states that it " differs from S. radians in having 
 larger cells, which appear more open ; thinner septa, and 
 consequently wider intervening spaces ; and four complete 
 cycles of septa." This coral rarely ever forms masses more 
 than six inches in diameter, though I have collected speci 
 mens 8-12 inches in length. Its color when alive varies 
 much. Usually it is of a very pale pinkish tint, almost 
 white, and it is not unfrequently blotched with deepened 
 spots of the same color. 
 
 It is often seen growing in tide-pools in the reefs and 
 rocks, with only just water enough to cover it. On the stone- 
 
 * Trans. Conn. Acad., 1868, p. 353. 
 
 VOL. I. 9 11 
 
194 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 reefs, as at Guarapary and Porto Seguro, it is often seen in 
 the pools exposed for several hours to the full blaze of the 
 sun, and of course liable to great and sudden changes of 
 temperature. These pools are also likely to be very much 
 freshened by heavy showers while the tide is down. Near 
 Bahia I have seen it growing in tide-pools above sea level, 
 and to which the waves had access only at high tide. It 
 seems to stand exposure to the air with impunity ; for on 
 the reef at Porto Seguro I have observed it exposed to 
 a hot sun for an hour or more during a spring tide. It 
 is not confined to the tide-pools, but occurs also on the 
 submerged border of the reef, where I have collected it in 
 a depth of 3 - 4 feet at low tide. This species appears to 
 range from Cape Frio northward beyond Pernambuco. 
 Professor Verrill has separated under the name var. con- 
 ferta what appears to be a variety, and which is distin 
 guished by having in the central portion cells deformed 
 through crowding. These cells are irregular, and deeper 
 than the normal ones near the basal margin. Their septa 
 and dividing walls are more elevated and convex, and 
 sometimes adjacent cells are united by the breaking down 
 of these walls. Usually with Siderastrcea occur two spe 
 cies of Favia, F. gravida Yerr. and F. conferta Yerr. 
 The former a solid, heavy coral sometimes flattened, and 
 incrusting stones or dead coral, sometimes in round masses 
 rarely more than three or four inches in width is allied to 
 F. Ananas and F. Fragum of the West Indies ; but Profes 
 sor Yerrill shows that it has more spiny costa3 than either 
 of those species, while the septa? are narrower and sharper. 
 The other species, F. conferta, forms small hemispheri 
 cal masses of about the same size as in the former species. 
 It is interesting because of its affinity to Goniastrcca, while 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 195 
 
 it stands, according to Professor Vcrrill, in some respects 
 intermediate between the genera Favia and Mceandrina. 
 A hemispherical or almost globular coral with large cells, 
 Acanthastrcea Braziliensis Vcrrill, which is common on 
 the border of the reefs below low-water mark, is rarely 
 ever found in the tide-pools, though I have occasionally 
 collected it from the deeper ones. On the edge of the reef 
 it grows to a very large size. The color is a pale gray, 
 when seen in the water. This is one of the corals most 
 instrumental in building up the reefs. Occasionally an Aga- 
 ricia, closely allied to, if not identical with, the West Indian 
 A. Agaricites Edw. and Haime, is found in one of the little 
 pools. It is a thin, spreading coral, attached by one side, 
 and resembles the flat woody fungi growing on dead timber 
 or the stumps of trees. This species often occurs almost 
 at the water level. At Villa Velha, and elsewhere, it is 
 found attached to Mussce. It appears to extend along 
 the whole coast between Victoria and Cape Sao Roque. 
 The above are the principal madreporian corals found in 
 the tide-pools. I have only rarely observed millepores 
 growing in the pools, and these either in the deeper or the 
 broad, sandy-bottomed ones on the reef. The only species 
 was M. Braziliensis Verrill, a species easily recognized 
 among the Brazilian milleporae by the peculiar form of its 
 branches, which Professor Verrill has described as " erect, 
 angular, or flattened, or forming broad, convoluted, and 
 folded rough plates, with acute edges and summits ; the 
 sides covered with sharp, irregular, angular, crest-shaped, 
 and conical prominences varying much in size and elevation, 
 often becoming continuous ridges, usually standing at right 
 angles to the sides of the branches." Professor Vcrrill 
 suggests that this may, after all, be only a variety of his M. 
 
196 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 nitida, but I have never seen any intermediate forms. 
 This M. Brazillensis sometimes grows to quite a large size, 
 and it ranges along the whole coast from the Abrolhos to 
 Pernambuco. On the submerged border of this reef occur 
 the beautiful species of Mussa and Symphyllia^ with which 
 Professor Verrill has associated my name.* 
 
 While at Santa Barbara the weather was unfavorable for 
 an examination of the reef below low-water mark, and my 
 collections were principally made from the surface of the 
 reef and from the tide-pools. 
 
 Of gorgonians I collected the same species as I found 
 at Victoria, namely, Hymenogorgia quercifolia^ Eunicea 
 
 * Mussa Harttii Verrill is distinct from all others in its regular cells and 
 its costae, which are furnished with strong, sharp, and recurved spines. It 
 grows in great abundance on the submerged border of the reefs in 3 - 6 feet 
 water at low tide, forming beautiful hemispherical bouquets a foot or two in 
 diameter. The color of the coral when alive is pale whitish. It is very fra 
 gile, and seems to prefer sheltered localities, and I do not believe that it grows 
 on the outside border of the reefs, which is exposed to the heavy surf. It occurs, 
 also, at Pernambuco, Porto Seguro, and elsewhere. With it is found another 
 form, which, though closely resembling it, Professor Verrill considers to be 
 generically distinct, and has called Symphyllia Harttii, giving the same specific 
 name in case it should, after all, prove to be identical with the above. 
 
 t This Hymenogorgia is exceedingly abundant on the Brazilian coast, repre 
 senting in the Brazilian polyp-fauna the RhipidogorgicB of the West Indian 
 faurwx. Professor Verrill describes it as follows : " It forms broad, fan-shaped 
 fronds, often two feet high and a foot broad, consisting of broad, foliaceous 
 branches, often resembling oak-leaves in form ; but at other times large, oval, 
 and irregularly incised or palmate. The branches of the axis are slender and 
 rounded, and pass through the fronds like the midribs of leaves. The rather 
 conspicuous, flat cells are scattered over the sides of the fronds." The color 
 when alive varies from an ashen-gray to a light yellow or pink. The latter 
 color often deepens in spots. It grows on the rocks and stone-reefs in clear 
 water, and on the submerged border of the coral-reefs. It ranges downward 
 to a depth of 5 - 6 feet or more. It is apt to fade in drying. On the fronds a 
 small parasitic Ovulum (O.gibbosum) is often found. Professor Verrill has more 
 recently restored both this and the West Indian Rhipidogorgia flabellum to the 
 genus Gorgonia. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 197 
 
 humilis,* and Plexaurella dichotoma. f The shallow, 
 sandy-bottomed pools over the reef are more or less in- 
 crusted with irregular masses of a calcareous deposit, con 
 sisting of an agglutination of serpula-tubes, nullipores, 
 bryozoa, &c., which usually lie in rather loose flakes on 
 the surface, and are easily turned over by the hand ; 
 much of the reef itself is covered in the same way. The 
 under sides of these masses, which are generally concave, 
 are incrusted with beautiful bryozoa, and form hiding- 
 places for great numbers of species of marine worms, 
 chitons, little crustaceans, ophiurans, <fcc. Beneath this 
 crust nestle in great abundance several interesting species 
 of " brittle stars." The most common of these is the 
 large Ophiura cinerea Lyman ; almost quite as abun 
 dant are the pretty Ophiothrix violacea Miiller and 
 Troschel, and Ophionereis reticulata Liitken. The other 
 species collected on the reefs were Ophiomyxa flaccida 
 Liitken, Ophiactis Krebsii Liitken, and Ophiolepis pauci- 
 spina Miiller and Troschel. $ All of the above species are, 
 according to Professor Verrill, members also of the West 
 Indian fauna, and it is interesting to observe the occurrence 
 here of the huge West Indian starfish, Or easier g-igas 
 Liitken, of which I collected two fine specimens at Santa 
 Barbara, together with another West Indian species, Linckia 
 ornithopus Liitken. The sea-urchin, common at Santa 
 
 * Eunicea humilis Edw. and Haime is a gorgonian easily recognized by its 
 growing in low, densely branched clumps, with short thick branches, the color 
 being usually lemon-yellow. It is a very common species along the coast. 
 
 t Plexaurella dichotoma Kolliker is another gorgonian, with few, large, round, 
 blunt-tipped and thick-barked branches of a dark brown color when dry. It 
 is as common as the preceding. 
 
 J In the collection of radiates made by Mr. Copeland and myself, on the 
 Thayer Expedition, there are several additional species of Ophiurans. 
 
198 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Barbara and along the Brazilian coast, Professor Verrill 
 considers to be the same as the West Indian species Echino- 
 metra Michelini Desor, and I am unable to detect any 
 difference between the Abrolhos specimens and those 
 which I have collected at St. Thomas. Many of the species 
 of shells common on the Brazilian coast, south of Bahia, 
 appear to be identical with West Indian forms, and one is 
 astonished to find at Pernambuco, Bahia, the Abrolhos, 
 and Victoria the large Cassis Cameo* a shell so common 
 in the West Indian waters. Professor Verrill has called 
 attention to the number of species of Echinoderms in the 
 Brazilian fauna that are identical with West Indian forms, 
 in contrast with the almost complete distinctness of the 
 polyp-faunae of the two regions ; and he has suggested that 
 an explanation might be found in the fact that the Echino 
 derms remain longer in the swimming larval form than the 
 polyps, and may be carried to greater distances by currents. 
 There is no chance by which West Indian species could be 
 carried south of Cape Sao Roque, owing to the equatorial 
 current which sets along shore from that cape northwest 
 ward across the mouth of the Amazonas, whose fresh waters 
 must have long presented a barrier to the migration south 
 ward of shallow water species. f It is interesting to observe 
 that on the Brazilian coast, south of Pernambuco at least, 
 we find no Diademas or Tripneustes, forms so exceedingly 
 characteristic of the West Indian Echinoderm fauna. 
 
 The material composing the reef is an exceedingly hard, 
 whitish limestone, ringing under the hammer, and, so far as 
 I had an opportunity to examine it, for the Brazilian reefs 
 
 * Formerly called Cassia Madagascariensis. 
 
 t Professor Verrill suggests, however, that the species found in both faunte 
 mav have migrated northward from Brazil. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 199 
 
 are never broken up by the surf, showing no distinct trace 
 of organic structure. The Santa Barbara reef extends around 
 about one third of the island, and on the northwestern side 
 it reaches across to the " Cemetery," so that when the tide 
 is down that islet is joined to the main island by a broad, level 
 platform of rock,* diversified by tide-pools, and forming an 
 excellent collecting-ground for the naturalist. The reef, 
 built up principally of Acant/iastrcea, Siderastrcea, Ac., has 
 completed its growth on arriving at low-tide level, the up 
 per surface being still farther added to by serpulas, bryo- 
 zoa, corallines, barnacles, &c., together with the coral-sand 
 and debris of shells accumulating on the reef. 
 
 So far I have spoken only of fringing reefs, but there are 
 other coral structures of greater interest in these waters. 
 Corals grow over the bottom in small patches in the open 
 sea, and, without spreading much, often rise to a height 
 of forty to fifty or more feet, like towers, and sometimes 
 attain the level of low water, forming what are called on 
 the Brazilian coast chapeirVes.^ At the top these are usually 
 very irregular, and sometimes spread out like mushrooms, 
 or, as the fishermen say, like umbrellas. J Some of these 
 chapeiroes are only a few feet in diameter. A few miles 
 to the eastward of the Abrolhos is an area with a length of 
 nine to ten, and in some places a breadth of four miles, over 
 which these structures grow very abundantly, forming the 
 well-known Parcel dos Abrolhos, on which so many vessels 
 have been wrecked. 
 
 * The reef is represented in the woodcut on page 175, but the cemetery has 
 unfortunately been omitted. See also map on page 177. 
 
 t Singular chapeirilo, pronounced sha/)-a-roim*, the n g representing a strong 
 nasal. The word means literally a biy hat. In the plural it is chapeiroes, pro 
 nounced shap-a-rd 1 -erfls. 
 
 \ The Dutch used to call them " Jesuits." 
 
 The word parcel means shoal or hidden rock, plur. parcels. 
 
200 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 I visited in my launch the northwestern part of this reef, 
 where the chapeiroes were sufficiently scattered to allow me 
 to sail about among them. 
 
 Among these chapeiroes I measured a depth of sixteen to 
 twenty metres, and once, while becalmed, I found twenty 
 metres alongside one chapeirao and three metres on top. 
 The chapeiroes, as a general thing, are rarely ever laid bare 
 by the tide. They are here, as elsewhere, of all heights 
 and dimensions ; but in no case do they reach low-water 
 level, nor, according to the testimony of the fishermen and 
 whalers, arc they ever in any part uncovered. They do not 
 coalesce here to form large reefs as they do to the west of 
 the islands. When the weather is clear and cloudless and 
 the water calm, these chapeiroes can be readily distinguished 
 at a considerable distance. The surface of the sea appears 
 to be flecked by shadows from a sky full of scattered 
 cloudlets, producing a striking effect. The water, being- 
 shallow and clear, and with a sandy bottom, is of a very 
 light greenish tinge, like that of the Niagara River at Buf 
 falo. The general color of the chapeiroes is brown, from 
 their being incrustcd with patches of Palythoa, and their 
 position is marked by brownish spots on the surface of the 
 sea. In the daytime a launch may sail in safety among 
 tbcm in calm weather, and a small vessel may traverse 
 some of the chapeirao grounds without danger, but large 
 ships are likely to find themselves in a labyrinth from 
 which escape is not easy.* In windy weather the waves 
 break over the chapeiroes, but if there are white caps be 
 side, and a cloudy sky, their position cannot be made out, 
 and it is safest to keep well away from them. In stormy 
 weather there is nothing to mark their position, and they 
 
 * See Chart of Abrolhos by Mouchez, note. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 201 
 
 are very dangerous. Sometimes vessels striking heavily on 
 small chapeiroes break them off, and escape without receiv 
 ing any serious injury, as has been remarked by Mouchcz. 
 At other times a vessel may run upon one of these struc 
 tures and stick fast by the middle of the keel, to the amaze 
 ment of the captain, who finds deep water all around, the 
 vessel being perched on the chapeirao like a weather-cock 
 on the top of a tower. Ordinarily in passing the Abrolhos 
 vessels and steamships go outside of these reefs to the east 
 ward in sight of the islands. It is not easy, however, to 
 calculate one s distance from a point at sea, and especially 
 from a light by night, and many vessels, notwithstanding 
 the lighthouse, have been wrecked upon them. West of 
 the islands there is deep water, and no chapeiroes, and 
 between the islands and the Paredes there is a channel 
 about eight miles in width, with plenty of water and no ob 
 structions. The best way in passing close to the Abrolhos 
 is to go to the westward of the islands, where one may run 
 close to them with safety even in the night-time. There is 
 then no danger whatever, and the sea is smoother. On 
 the return voyage from Rio, September, 1867, the American 
 steamship " South America" was, at the suggestion of the 
 writer, taken through this channel. In case of necessity, 
 good anchorage may be found close in by the island of Sta. 
 Barbara on the northern or southern side, as the direction 
 of the wind may determine. 
 
 Eight miles northwest of the islands, between them and 
 the mainland, is the Parcel das Paredes, literally, The Shoal 
 of the Walls, an irregular area about seventeen miles long 
 from north to south, and some nine miles in width, occupied 
 by very extensive reefs and chapeiroes. Mouchez has only 
 given the general outline of the Parcel, which was all that 
 
 9* 
 
202 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 was really necessary for his chart. The reefs within the 
 Parcel are not drawn from an actual survey, and have no 
 approach towards accuracy. 
 
 In the northern part of the Parcel the chapeiroes so 
 closely unite as to form an immense reef, which has grown 
 upward to a little above the level of low water, and is quite 
 uncovered at low tide. This reef, as are all the others, is 
 exceedingly ragged in outline, full of indentations, and 
 abounds in shallow pools. The fishermen describe two 
 channels that enter the reef from the north, and almost 
 separate it into three parts. My captain, Jacob Torgjusen, 
 an intelligent Dane, says that the water in these chan 
 nels is quite deep. The northeastern part of this reef is 
 called the Eecife do Lixo, because of the abundance of a 
 shark-like ray called the lixo, which is furnished with large 
 crushing teeth, and frequents the reef in search of shell 
 fish, on which it feeds. 
 
 RECIFE DO LIXO. 
 
 I spent one tide on the Recife do Lixo, during the full 
 moon of the 13th of August, 1867, when the reef was un 
 covered, and examined it quite carefully. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 203 
 
 The surface of the reef was remarkably even in height, and 
 covered largely by calcareous sand, on which were patches 
 of dead coral incrusted with millepores, barnacles, serpula?, 
 &c., with occasional living corals, such as Sidcrastrcea stellata 
 and Favia, and perhaps a Porites solida. The dead corals, 
 nullipores, <fcc., usually forming incrusting masses over the 
 sand, so loose as to be easily turned over, affording a rich 
 harvest of Ophiurans, among which Ophiura cinerea, Ophi- 
 onereis reticulata, and Ophiothrix violacca were especially 
 abundant. In some situations sea-urchins, Echinometra 
 Michelini, were very plentiful. Volutes and Cassis Cameo 
 may frequently be picked up. The reef is not very rich 
 in shell-fish, but abounds in crustaceans.* A large naked 
 
 * On the Thayer expedition Mr. Copeland and I collected large numbers of 
 crustaceans at all the principal localities on the coast between Rio and Bahia. 
 On my second visit to Brazil I was too much engaged in geological studies, 
 and in my examination of the reefs, to make extensive collections. The few 
 crustaceans brought home from the second journey, my friend, Mr. S. I. Smith, 
 of the Sheffield Scientific School of New Haven, has been kind enough to 
 examine and describe ; and since writing the above he has published a paper in 
 the second volume of the Transactions of the Connecticut Academy of Arts and 
 Sciences, entitled " Notice of the Crustacea collected by Professor C. F. Hartt on 
 the Coast of Brazil in 1 867, together with a List of the described Species of Bra 
 zilian Podophthalmia." Mr. Smith enumerates the following species as occur 
 ring on the reefs of the Abrolhos : Milnia bicornuta Stimp., Mithraculus coronatus 
 Stimp., Mithrax liispidus Edwards, Xantho denticulata White, Cldorodius Flari- 
 dianus Gibbes, Panopeus politus Smith, Panopeus Harttii Smith, Eriphia 
 gonacjrn Edwards, Goniopsis cruentatus De Haan, Dromidia AntiUensis Stimp., 
 Petrochirm rjranulatus Stimp., Calcimts sulcatus Stimp., Clibanarius AntiUensis 
 Stimp., Alpheui htterochelts Say, Gonodactylus chiragra Latreillc (?). 
 
 From other localities were described the following : Callinectes Dance Smith, 
 Pernambuco. C. ornatus Ordway, Caravellas. C. larvatus Ordway, Bahia. 
 Ache.Ioufs spinimanns De Haan, Bahia. A. Ordicayi Stimp., Bahia. Uca corda- 
 ta, Bahia. Cardiosoma quadratum Saussure, Pernambuco. Clibanarius vittatus 
 Stimp., Caravellas. C. sdopetarins Stimp., Caravellas. Sryllarus wquinoxialis 
 Fabr.. Bahia. Panulirus echinatus Smith, Peruambuco. Palcemon Jamaicensis 
 
204 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 mollusk (Aplysia) * four to five inches long, grayish, and 
 ornamented with dark rings and spots, may sometimes 
 be found very abundantly on these reefs. It gives out a 
 very copious deep purple fluid when handled. Octopi are 
 very common in the cracks and crevices of the reef. The 
 shallow pools are often very rich in life. 
 
 The uniform level of the surface of the reef laid dry is 
 very remarkable. So very even and unencumbered is it 
 that a loose coral a foot in diameter turned over, or a cassis 
 lying on the surface, attracts attention at a long distance. 
 The reef is so protected that the waves have no power to 
 break off its edges and encumber its surface, as is the case 
 with the coral reefs of the Pacific. Sand, resulting from 
 
 Olivier, Penedo. P. forceps, Edwards, Para. P. ensiculus Smith, Para. Peneus 
 Brasiliensis Latreille, Bahia. Xiphopeneus Ilarttii Smith, Caravellas. 
 
 The list of the Brazilian Podophthalmia is too long to be inserted in this 
 volume. A very large number of the species examined by Mr. Smith have 
 been identified with West Indian or Floridan forms. Mr. Smith suggests that 
 one reason why my collection is so much richer in proportion in these forms 
 than the Brazilian collections heretofore made, may be because my collections 
 were made on the reefs and rocky parts, while the others were made at Rio, 
 where there are no coral-reefs. The collecting-grounds at Rio are rocky as well 
 as sandy. I suspect that the true reason is to be found in the fact that the crus 
 tacean fauna changes its character south of Cape Frio. 
 
 I called Mr. Smith s attention to the name Cardiosoma Guanhumi, in which 
 the specific name appeared to have been derived from the word Guayamu or 
 Gudinumu, the Tupi name for the species. The former is the way it was written 
 for me by a Brazilian, but Fonscca gives the latter form. I feel quite sure that 
 the name Guayamu was applied to several distinct species. In reference to Uca 
 una, Mr. Smith has slightly misunderstood me. I do not remember the present 
 vulgar name. Vca-una I found in Piso. It is Tupi, and means simply black 
 crab, vya, or more properly Uca, meaning crab and una, black. 
 
 * Mr. J. G. Anthony kindly informs me that the species, of which I furnished 
 numerous specimens to the Museum of Comparative Zoology, is probably A. 
 Argo d Orb. Sander Rang describes a large species, A. Braziliensis, from the 
 Bay of Rio. Hist. Nat. des Aplysies, p. 55, PI. VIII. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 205 
 
 the decay of coral, the breaking up of shells, &c., accumu 
 lates very slowly. The reef has grown as high as is possi 
 ble, and is now dead, and at the lowest tide it is not more 
 than two feet out of water. An ordinary tide would not 
 uncover it completely.* 
 
 An irregular raised border, consisting principally of a 
 growth of millepores, serpulae, barnacles, <fcc., sometimes a 
 foot, more or less, in height, separates this part of the reef, 
 which is uncovered at low water, from that which is always 
 submerged. It is here that the waves break at low water, 
 and this favors the growth of these animals more than else 
 where. From the border this reef generally slopes off 
 gently towards the edge under water, where it drops down 
 perpendicularly into deep water, as at the islands. This 
 submerged border at the Lixo has only a few feet of water 
 on it at low tide, and one may usually wade out to its edge 
 and collect. It is a perfect garden covered with growing 
 corals of large size. Here grows Acanthastr&a in large 
 heads, more abundant on the edge of the reef. Millepora 
 nitida Yerrill forms pretty rosettes. This interesting 
 species is thus described by Professor Verrill : f 
 
 " Corallum forming low rounded clumps, four to six 
 inches high, consisting of short, rapidly forking, rounded 
 or slightly compressed branches, about .4 to .8 of an inch 
 in diameter, which have remarkably smooth surfaces, and 
 are obtuse, rounded, or even clavate at the ends. The 
 larger pores are small, very distinct, round, evenly scattered 
 over the surface, at the distance of .06 to .1 of an inch 
 apart. The small pores are very minute, numerous, scat- 
 
 * The height of the reef is probably in part due to the recent uprise of the 
 land. 
 t Trans. Conn. Acad., 1868, p. 362. 
 
206 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tered between the larger ones, and often show a tendency 
 to arrange themselves in circles of six or eight. The tissue 
 is, for the genus, very firm and compact." The color, when 
 alive, is light pinkish. This species is abundant on the sub 
 merged border of the Recife do Lixo, growing in from 
 three to four feet of water at low tide. Some of the ro 
 settes I collected lay quite loose and without any attach 
 ment to the reef. 
 
 Among the millepores I obtained on the Brazilian coast 
 Professor Verrill has distinguished three forms which so 
 closely agree with Millepora alcicornis Linnasus, that he 
 has separated them as varieties of that species. One of 
 these, var. cellulosa Verrill, I found at Pernambuco, but I 
 did not see it alive. Professor Verrill describes it as fol 
 lows : " Corallum consisting of numerous, irregular, rath 
 er short branches, arising from a thick base, branches pro 
 liferous or digitate at the ends, the last division short, 
 mostly compressed and acute at the tips. Some of the 
 branches occasionally coalesce, so as to leave small open 
 ings. Cells numerous, crowded, rather large for the genus, 
 each sunken in a distinct depression, the wall rising up 
 into an acute ridge between them, texture rather open and 
 coarsely porous." 
 
 Another form, which differs from the other in its " some 
 what more porous texture, and the greater regularity and 
 more scattered arrangements of the cells," together with its 
 round and digitate branches and branchlets, having three to 
 five short compressed divisions at the end, Professor Ver 
 rill has referred with doubt to the variety digitata of Esper. 
 The third variety distinguished by having the branches 
 in the same plane, and coalescing in such a way as to leave 
 frequent openings, Professor Verrill has referred to the 
 M.fenestrata Duch. and Mich. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 207 
 
 Both of these last occur abundantly along the coast from 
 Cape Frio northward, as far as I have examined. They 
 seem to prefer the edge of the reef, where they form beauti 
 ful broad frills of a light yellowish-brown or pinkish color. 
 In collecting these millepores, I was struck with their pow. 
 erful stinging properties, and they burned me sometimes 
 like hot iron, producing a sensation precisely like that 
 caused by the Physalia, or our Northern jelly-fish, Cyanea. 
 I was stung in the same way by the millepores of St. 
 Thomas. The fishermen who were with me on the Bra 
 zilian reefs handled these corals with impunity, but they 
 called them " sea-ginger," and told me that they were 
 accustomed to play practical jokes on land-lubbers by per 
 suading them to taste them. I presume that a thin and 
 delicate skin makes me more sensitive than most persons 
 to the stinging properties of these animals.* 
 
 The Sidcrastrcea and Favice, already described in speak 
 ing of the Santa Barbara reef, are found on the Recife do 
 Lixo, both in the pools and on the submerged border, and 
 associated with them are a few forms which appear to be 
 somewhat rare, as I could find but few specimens of them. 
 Among these is the species of Favia, described by Professor 
 Verrill under the name F. leptopkylla, an interesting spe 
 cies which forms large hemispherical corals, easily recog 
 nized by " the very open, deep, rounded cells ; few, thin, 
 projecting septa ; and thin distinct walls." Another and 
 beautiful coral is Heliastrcca aperta Verrill, which the 
 following description, almost in Professor Yen-ill s own 
 words, will serve to distinguish : The corallum is large, 
 more or less regularly hemispherical, sometimes subspher- 
 
 * This stinging property of the millepores is in accordance with their acele- 
 phian structure, first announced by Professor Agassiz. 
 
208 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ical, and often a foot and a half in diameter. The text 
 ure is open and light, which character, together with the 
 thinner and more acute septas, serves to separate it from 
 H. cavernosa Edw. and Haime, which it resembles in the 
 large size and prominence of its cells. These in H. aperta 
 are circular, large, moderately deep, with a broad central 
 area, the margin projecting about .08 inch above the gen 
 eral surface. Septa in three complete cycles, narrow, thin, 
 subequal, the summits considerably projecting, angular, 
 acute, the inner edges nearly perpendicular, finely toothed, 
 often with a distinct paliform tooth at the base. Columella 
 well developed, of loose, open tissue. Costse elevated and 
 thin, rising obliquely upward to the summits of the septa, 
 finely serrate. "Walls very thin, inconspicuous. This spe 
 cies seems to be more abundant in the Bay of Bahia than 
 in the Abrolhos region, and I have frequently seen it there 
 in the heaps of corals brought from the Island of Itaparica 
 to the city for burning into lime. 
 
 A very pretty Pectinia (P. Braziliensis Edw. and 
 Haime) is another of these apparently rare forms, of which 
 I have found only a single specimen, which was growing on 
 the reef borders at the Lixo, in about two feet of water at 
 low tide. 
 
 On the Recife do Lixo I collected a few specimens of a 
 massive Forties, resembling P. Guadaloupensis Duch., 
 which is very abundant on the Porto Seguro coral-reef. 
 It is sometimes of a bright sulphur-color, though it varies 
 very much in tint. Professor Verrill has described it as a 
 new species under the name P. solida, and states that it 
 differs from the "West Indian forms in its larger and deeper 
 cells, thicker walls, wider and more crispate septa, and 
 more solid structure. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 209 
 
 The beauty of the polyp growth on the submerged reef 
 border is much enhanced by the great luxuriance of the 
 gorgonians, which are the same as those already ob 
 served elsewhere, but on the Lixo reef I discovered a 
 beautiful new species, described by Professor Verrill un 
 der the name of Gorgonia gracilis. It grows in little 
 tufts, about six to eight inches high, with few slender and 
 very delicate branches. The color is yellow or purple, and 
 in rough water the species is apt to be overlooked by the 
 collector, from its resemblance to a sea-weed. Towards the 
 edge, as the water deepens, the reef grows more and more 
 irregular. It is full of holes, and almost wholly composed 
 of live and growing corals, which furnish a very insecure 
 footing. On the edge and sides grow immense coral-heads, 
 and the Mussce are especially abundant. The outline of 
 the border is exceedingly ragged. 
 
 The height of the perpendicular edge on the western 
 side of the Porto Seguro and Lixo reefs varies very much, 
 being in some places three or four feet, in others ten feet or 
 more. I could not examine it on the eastern side, owing to 
 the surf. I have introduced a diagram, showing the reef 
 as seen in section, with the distribution of the different 
 species of corals indicated. Just alongside the reef, at least 
 on the western side, wherever I have examined it, and 
 the same holds good of the Porto Seguro reef, the bottom 
 slopes rapidly away from the reef-edge, and is composed of 
 a soft, bluish, calcareous mud, washed from the top of the 
 reef, which makes the reef appear much lower than it really 
 is. A short distance away, in some places, a depth of sev 
 enty to eighty feet may be found. The diagram on page 
 211 represents a sketch of the edge of the reef of the Lixo, 
 with soundings made by myself. The large reefs appear to 
 
GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 SECTION ACROSS BORDER OF LIXO REEF. 
 
 a. Tide-pool with Sideraslrcea stellata and Favia gravida. 
 
 b. Low dike-like border of serpula-tubes, barnacles, &c. 
 c c. Low-water level of spring tide. 
 
 d. Siderastrcea stellata. 
 
 e. Acanthastrcea JBraziliensis. 
 
 f. Heliastrcea aperta. 
 
 g. Porites solida. 
 
 h. Millepora Brazillensis. 
 
 i. ^Millepora nitida. 
 
 . Varieties of Millepora alcicornis. 
 
 m. Gorgonia (Hymenogorgia) quercifolia. 
 
 n. Plexaurella dichotoma. 
 
 o. Gorgonia gradlls. 
 
 p. Eunicea humilis. 
 
 r. Mussa narttii. 
 
 have been formed not only by the upward growth of large 
 patches, but by the filling in and coalescence of chapeirocs, 
 which is a feature not hitherto spoken of in the growth of 
 coral reefs. The ponds on the surface of the reef probably 
 mark the intervals between the chapciroes where the filling 
 in is nearly complete ; though they may in some cases 
 mark spots where the corals have been killed by the drift 
 of sand. 
 
THE ISLANDS AND CORAL REEFS OF THE ABROLHOS. 211 
 
 I visited the eastern side of the Lixo, but the waves were 
 breaking with too much force to allow me to see anything 
 distinctly. The boatmen said that this reef drops down 
 perpendicularly into deep water. I observed no sand 
 banks upon it. On the western side of the Lixo there are 
 
 \ *: -_=- _ isu u -_-. 
 
 but few chapeiroes, but the other sides are bordered by an 
 abundance of them, and they stretch off southward, forming 
 with two other reefs, called the Recife de Leste and Recife 
 da Pedra Grande, the rest of the Parcel. The Recife da 
 Pedra Grande was described by Jacob as being like a wall, 
 quite straight, some three miles in length, and only two or 
 three hundred feet wide. Southwest of the Paredes are 
 several other quite large reef-grounds. One of the reefs, 
 Coroa Yermelha, has become converted into an island by 
 the heaping up of sand in the centre. There are no reefs 
 or chapeiroes south of the Mucury. ^orth of the Recife 
 do Lixo are some small reef-patches, and about ten miles 
 
212 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 northward, a little to the west, is a very dangerous reef- 
 ground some three miles long, and one and a half to two 
 miles wide, called the Timbebas, between which and the reefs 
 to the south is a wide open channel called the Itanhaem. 
 This, as Mouchez says in his chart, is the most dreaded of 
 all the reefs, because it is situated just beyond the limit of 
 visibility of the coast, and there is nothing to mark its po 
 sition, so that even the pilots of the coast keep well away 
 from it. I once passed close to it in the steamer Santa 
 Cruz, and saw the waves beating over it. Small coral 
 patches occur close in shore along the coast northward as 
 far as to Point Carumba, just off which is a reef-ground 
 eight miles long and three to five miles wide. Mouchez 
 says that the reef is uncovered at low tide in the western 
 part, and that the rest of the ground is covered by chapei- 
 r5es. These reefs are the celebrated Itacolumis. The 
 next important reef-ground stretches across the mouth of 
 the Bay of Porto Seguro. I found the reef uncovered over 
 an area as far as I could see north and south while standing 
 on the reef, with a width in some places of a mile ; but it 
 was exceedingly difficult to judge of distance on so plane and 
 monotonous a surface. This reef is surrounded by chapeiroes. 
 The same line of reefs extends northward, passes the low 
 point north of Porto Seguro, leaving a deep channel, allow 
 ing the passage of steamers, and is continued across the bay 
 of Santa Cruz, as is represented in the sketch-map on p. 233. 
 In the summer I passed close to the northern part of the 
 reef off Santa Cruz. It has, like the other principal reefs, 
 finished its growth, and is in part converted into an island, 
 on which I observed a few mangrove-trees growing. A 
 schooner had struck on the outer side of this reef, and had 
 afterwards been carried over and sunk erect, just behind the 
 
THE ISLANDS AND CORAL BEEFS OF THE ABROLHOS. 213 
 
 reef. Its masts were standing, showing a depth alongside 
 of the reef of about thirty feet. Reef patches occur along 
 the shore in the vicinity of Camamu. Quieppe Island is 
 surrounded by them on all sides, and a little bay between 
 Camamu and Boyapeba is full of chapeiroes, while the en 
 trance to the Barra Carvalhos, according to the " South 
 American Coast Pilot," is similarly encumbered. Along 
 the shores of Itaparica Island there are extensive coral- 
 banks, from which coral is taken for the purpose of being 
 burned into lime ; and at low tide coral is largely collected 
 from banks lying off Periperi, on the Bahia Railroad, to 
 which place it is brought and burned. I saw large quanti 
 ties of it at Bahia and Periperi. The corals were of the same 
 species as are common on the Abrolhos reefs, but Heliastrcea 
 was more common. Limestone is very scarce on the Bra 
 zilian coast, and corals are largely used for making lime. 
 They call the coral pedra de cal* in Brazil, for this reason. 
 Still going northward, I am not aware that any coral- 
 patches are to be found until we reach Macei6, but in the 
 harbor of this town I examined quite an extensive one, which 
 lies some distance, perhaps a mile, off the town, and at low 
 tide is uncovered for a large irregular space. Its surface is 
 flat, but irregular, composed of dead corals, and is full of 
 holes. Walking out to the edge of the reef, where the sea 
 was breaking heavily, I found it exceedingly rough. I was 
 unable to see much, as the tide rose rapidly, flowing over 
 the reef, and I had to wade back half a mile over the 
 rough surface to my boat. I saw very few live corals in 
 the ponds. f There were a very few little Siderastrcece, and 
 
 * It would not be worth stating here that the red coral does not occur on the 
 Brazilian coast, if it were not that from my use of the term coral the report had 
 been circulated that I had discovered a reef of red coral at the Abrolhos ! 
 
 t Professor Agassi/, tells me that he has fine millepores from Maceid. 
 
214 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 an occasional dwarfed Acanthastrcea. Echinometrce were 
 abundant, and English sailors were spearing the large 
 Aplysica for the sake of spilling their purple ink. In the 
 hands of a sailor I saw a beautiful Valuta like V. musica, 
 and I collected a specimen of Linckia ornithopus. My 
 boatman said that on the outer edge of the reef large 
 coral-heads grew, and that these were collected and brought 
 on shore for burning into lime. 
 
 From the lighthouse at Macei6 the reefs could be seen 
 extending in an irregular line along the shore northward 
 many miles. In the vicinity of Pernambuco are patches of 
 growing coral, as Dana has remarked. I have specimens 
 from these patches, but I was prevented from visiting them. 
 I know of no one who has examined the reefs north of Per 
 nambuco with a view to scientific results. 
 
 The Roccas are a very dangerous cluster of reefs in the 
 latitude of Fernando de Noronha, noted for its annular 
 shape. From the charts of the Roccas, together with what 
 I have learned from those who have visited them, they 
 must be true coral reefs. 
 
 Though there is a general resemblance between the Brazil 
 ian and "West-Indian polyp faunae in the representative spe 
 cies of Siderastrcca, Favia, Porites, Plexaurella, &c., &c., yet 
 one who has collected in the West Indies, as at St. Thomas, 
 for instance, or Florida, is much struck with the absence 
 of Madrepora, Mccandrina, Diploria, Manicina, Cladocora, 
 Oculina, and other genera characteristic of the West-Indian 
 fauna. The Brazilian reefs are built up by a very few spe 
 cies, among which Acanthastrcea Brazlliensis appears to be 
 the most common, together with Siderastrcccc and other 
 massive forms ; but the Milleporcc and Musses must con 
 tribute more or less to their growth. 
 
PROVINCE OF BAHIA. 215 
 
 CHAPTER Y. 
 
 PROVINCE OF BAHIA. COAST SOUTH OF SAO SALVADOR. 
 
 The Tertiary Lands between the Rivers Mucury and Peru hype ; their Vegeta 
 tion, &c. Colonia Leopoldina and its Coffee Plantations. Villa Vicosa. 
 The Canal joining the Rivers Peruhype and Caravellas. Formation of 
 Beaches and Beach Ridges. Coast between Caravellas and Porto Seguro. 
 Monte Pascoal. Porto Seguro and its Recife, or Consolidated Beach. Santa 
 Cruz and its Reef. Coast northward to the Jequitinhonha ; the Lagoa do 
 BrafO, Campos, &c. The Canal Po-assii and the Rio da Salsa. Mangrove 
 Swamps between the Jequitinhonha and Pardo. Cannavieiras. The Salt 
 Trade of the Jequitinhonha. Description of the lower Part of the Rio 
 Pardo; Cacao Plantations, &c. Coast northward to Ilhc os. Prince Neu 
 "NVied s Description of the Country between Ilhe os and Conquista, Possoes, 
 and Cachoeira ; the Forests, Campos, Social Plants, &c. Ilheos. Rio and 
 Lngoa Itahype. Dead Coral Banks. Rio das Contas. Bay of Camamu. 
 Turfa Deposits. Villa de Camamu. Coast northward to the Bahia 
 de Todos or Santos. The Bay of All Saints described. Ilha Itaparica. 
 Rio Jaguanpe and Nazareth. Rio Paraguassii. Description of River 
 below Cachoeira. The Tram-road Sant Amaro and the Agricultural 
 Institute. 
 
 THE Rio Peruhype has its source in the Serra dos 
 Aimores, a few miles north of Santa Clara, and flows in 
 a deep, narrow valley, worn for the greater part of the dis 
 tance through the coast tertiary belt. It is not more than 
 fifty miles in length, and is consequently an insignificant 
 stream. Between the Peruhype and the Mucury the lands 
 are almost wholly tertiary, though between the mouths of 
 these two rivers there is a strip of sand and marsh. I owe 
 to the kindness of my friend Sr. Schlobach a profile of a 
 railroad line surveyed by him to connect the Philadelphia 
 
216 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Road at Santa Clara with the Peruhype above Villa Vic,osa, 
 which profile is of much interest, since it shows that the 
 surface of the tertiary beds slopes very regularly from the 
 serra to the sea, a slope which I believe to depend on 
 the slope of the sea bottom on which the beds were de 
 posited, though it is, in part at least, due to a thickening 
 of the beds towards the serra. The country between the 
 two rivers forms a great wooded plain ; but the forests 
 are by no means so luxuriant in their appearance as on the 
 Sao Matheos or Doce. The soil is sandy, weak, and dry, 
 though on the slopes of river valleys it is productive, and 
 the bottoms of the valleys are covered by a very rank growth. 
 This region is traversed by many little streams, all of 
 which have cut for themselves deep valleys or canons with 
 very steep sides. The upland forests furnish many valua 
 ble woods, and rosewood is extensively cut both on the 
 Peruhype and Mucury. In some parts the forest is very 
 sparse of underbrush, but in others it is so matted and tan 
 gled with bamboos (Taquaras) and young Airi palms 
 (Astrocaryum Airi Mart.) as to be almost impenetrable. 
 There are large marshy areas on this plain, and these are 
 in part flooded, some of them only during the rainy season, 
 when they form shallow lagoons. The vegetation of the al 
 luvial lands bordering the little streams is luxuriant beyond 
 description. Here one finds an abundance of Palmitos 
 (Euterpe edulis Mart.) and beautiful tree-ferns. The Boto- 
 cudos still hold the country, and I saw their deserted 
 ranches every few miles along the path. The forest 
 abounds in game. Queixadas, coititus, antas, onc,as, mon 
 keys, and birds of every description are very numerous. 
 
 Nowhere on the route did I see any good exposure of the 
 tertiary beds, but in the descent into the river valleys 
 
PROVINCE OF BAHIA. 217 
 
 I repeatedly saw the drift, which here as elsewhere is com 
 posed of red clay full of fragments of quartz. All the 
 streams are " black-water," and so is the Peruhype.* On 
 this last river a colony of Germans, together with some 
 natives of other countries, was established many years ago. 
 It is called Colonia Leopoldina, and consists of a consider 
 able number of fazendas situated on both sides of the river 
 for several miles above Sao Jose*, the head of navigation for 
 steamers. Some of these fazendas are very large and valu 
 able, as that of the late Sr. Joao Flach for instance. The 
 cultivated lands are situated on the very edges and slopes 
 of the chapadas, and they were formerly very productive.! 
 Coffee is the staple product, and is noted in Brazil as being 
 of very superior quality, passing by the name of Cafe" de 
 Caravellas. The trees are not allowed to grow to their full 
 height, but are trimmed down, so that the picking is easily 
 done by hand from the ground. Six feet or thereabouts 
 appears to be the average height of the coffee-trees of 
 this region. The trees, when trimmed in this way, are 
 flat-topped with pendant branches, and a cafezal, or coffee 
 plantation, at the colony is a very pretty sight. Mandioca, 
 maize, cotton, and other Brazilian products are cultivated 
 largely, and one riding through the fazendas sees here as 
 elsewhere the orange, banana, lime, citron, pine-apple, &c. 
 The mamao, a species of Carica, furnishing a large and 
 savory fruit full of seeds, is frequently seen, though it is to 
 be found almost everywhere along the coast, growing in 
 corners of fields and gardens. The climate of the colony 
 
 * Von Tschudi says (Reisen, &c., Vol. II. p. 357) that the water of fhe 
 Peruhype at Villa Vicosa is deep brown, and that, churned up by the paddle- 
 wheels of a steamer, it looks like foaming porter. 
 
 t These soils become rapidly exhausted, and many of the old fazendas are 
 very nearly worn out. 
 
 VOL. i. 10 
 
218 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 is very unhealthy, especially for foreigners, but it is far 
 superior to that of the Mucury below Santa Clara. The 
 Peruhype is navigable as far as Sao Josd for coasting 
 steamers, which, however, do not enter the river at its 
 mouth, but by means of a natural canal which, through a 
 very extensive swampy tract, connects the Peruhype with 
 the Caravellas River. The bar at the mouth of the river 
 is very bad, and is rarely crossed by any except very small 
 vessels. The entrance at Caravellas is good. Since the 
 navigation of the Mucury is so difficult, and its bar so bad, 
 it has been proposed to extend the Minas Road across the 
 plains to the Peruhype below the colony, and ultimately 
 make it a railroad. The only difficulty in the way of build 
 ing such a road would be that of bridging the deep valleys 
 of the streams, which would require very high viaducts or 
 bridges. It is to be hoped that this project will one day 
 be accomplished. Villa Vic,osa, a town of some five hundred 
 or one thousand inhabitants, is situated on the right bank 
 of the Peruhype, at a distance of about four miles from 
 the sea, and five or six below Sao Jose" da Colonia Leo- 
 poldina. The lands of the vicinity are very productive in 
 mandioca, and a large quantity of farinha is exported.* 
 From the mouth of the Peruhype to that of the Caravellas 
 the shore consists of a long sea-beach broken only by one 
 river-mouth, Barra Nova. From Villa Vic,osa to the 
 Villa de Caravellas a rather narrow tidal canal extends 
 
 * The water used for drinking at Villa Vir-osa is very bad. Yon Tschudl 
 says that it is collected in shallow holes sunk in a sandy grass-grown plain 
 southwest of the town. When first drawn its taste is very disagreeable, but it 
 becomes potable after standing a day or two. All the towns along the coast, 
 built on the sands of the sea-shore, as a general thing have bad water. This 
 is the case at Caravellas. It is needless to say that the want of pure water is 
 one cause of the prevalence of sickness in these places. 
 
PROVINCE OF BAHIA. 219 
 
 parallel with the coast. "Westward of this, as Von Tschudi 
 has remarked, is the terra firma, the tertiary lands 
 trending off northward at a greater or less distance from 
 the canal. This stream, in which the tide ebbs and flows, 
 and whose waters are salt, is sufficiently deep to allow the 
 passage of the coast steamers. Between the canal and 
 the sea the lands, which are but little above the water, con 
 sist in part of sandy flats covered with forest or cocoa-palm 
 groves, but the greater part appears to be an immense 
 mangrove swamp, similar to that which lies between the 
 Jequitinhonha and Pardo to the north, and, like this last, 
 it is intersected by a network of narrow canals, which 
 have never been mapped. Gerber s map represents this 
 area as a cluster of small islands, with the sea penetrating 
 from the coast through numerous canals, which probably 
 misled Von Tschudi, who also describes it as an archipelago. 
 The so-called Rio Caravellas is only a narrow estuary, which 
 penetrates into the interior for a distance of about twelve 
 miles. From this estuary near Caravellas the canal extends 
 southward to Vic,osa, while at its head empty the Rio Cara 
 vellas and the Rio da Fabrica. 
 
 The water off this coast is very shallow, and along shore 
 it is usually very turbid. Outside are the very extensive 
 coral-banks, described in the preceding chapter, which not 
 only break the force of the ocean waves, but give rise to a 
 system of currents whose force and direction depend almost 
 entirely on the winds. Owing to the protection of the shores 
 from such powerful wave action as obtains almost everywhere 
 else on the coast, we find here a sloping sand-beach, rising 
 to but a few feet above high-water, and unaccompanied by a 
 beach-ridge. The water along the coast is ordinarily as 
 smooth and calm as that of an inland lake. Near the mouth 
 
220 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of the Caravellas and a few miles to the southward, I ob 
 served dead mangroves standing in clumps in the water just 
 outside the land. A short distance south of the Mucury, as 
 already remarked, I observed dead trees still erect, and 
 rooted below high-water mark. It would seem that there 
 has been, within a few years, a slight sinking of this part 
 of the coast. The mangrove swamps are sometimes formed 
 in depressions caused by the wearing away of the lower 
 lands by the sea ; but this is very rare. They almost al 
 ways occupy tracts protected from the waves by sand-beaches 
 or otherwise, and rilled up with sand-banks, or which, by 
 the gradual rise of the land, have been brought so near the 
 surface that the seeds of the mangrove take root. If one 
 examines the sea bottom along the coast he will find that, 
 especially in the vicinity of large rivers, such as the Sao 
 Francisco, Pardc, Jcquitinhonha, Doce, Parahyba do Sul, 
 and, as a general rule, off flat lands everywhere, the bottom 
 slopes away very gradually, and consists of beds of sand, 
 and but rarely of mud, because such fine material is almost 
 invariably swept out with the current into deep water. 
 A beach must of course be formed within the limits of 
 wave action. If these correspond with the edge of the 
 land, then the beach will skirt it, but where, as is frequently 
 the case, the water is so shallow that the waves break at a 
 distance from the shore, there will shortly be formed, along 
 this line, a ridge of sand, which will gradually increase in 
 height until at last it will appear above water, forming a 
 narrow strip parallel to the shore. This may become so 
 high as to form a permanent barrier, enclosing behind it a 
 lagoon of shallow water. Sometimes the formation of these 
 beaches is due to the action of storms of extraordinary 
 violence, which have caused the waves to disturb the bottom 
 
PROVINCE OF BAHIA. 221 
 
 farther than usual from the shore. In the slow rise of a 
 coast bordering shallow water the line of surf action would 
 be gradually removed from the shore. If the slope of the 
 bottom is uniform, and there are no storms, the effect may 
 be to add slowly to the coast by a constant throwing up of 
 sands by the waves ; and this is beautifully illustrated at 
 the mouth of the Jequitinhonha by the plains stretching 
 southward from Belmonte. These plains consist of a great 
 number of parallel beaches, one lying in front of the other, 
 and traceable for miles ; but even here the growth of the 
 coast has not been uniform, and occasionally an increased 
 violence of the surf has thrown up a ridge a few yards or 
 rods outside of that last formed, making a narrow ditch- 
 like lagoon, like a river, running parallel with the shore. 
 A lagoon of this kind runs behind the beach along the 
 present shore line like a strip of insertion. If the water is 
 very shallow, there will be no regular beach ridge, but there 
 may be dunes if the sand is light enough to be raised by 
 the wind ; but if the water is so deep that the whole force 
 of the waves is expended on the shore line, breaking in 
 a single line of surf, then we may expect to find a high 
 ridge accompanying the beach, a ridge which has owed 
 its origin primarily, it may be, to some very heavy storm, 
 and secondarily to the joint action of the winds and waves 
 in piling up the sand. This is beautifully seen at the 
 mouth of the Jequitinhonha. At Belmonte the sea is ex 
 ceedingly shallow, consequently the shore is low and with 
 out a ridge. Going southward the water deepens, the surf 
 is heavier, and a well-defined ridge begins, growing higher 
 and higher the farther south we go. 
 
 Along the coast of Long Island, and of the Middle and 
 Southern Atlantic States, we have these phenomena well 
 
222 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 exhibited. "When between the rivers the shore is flat, such 
 a beach may extend from the mouth of one to that of 
 the other, while the lagoon behind may form a communi 
 cation like a canal between the two, or the mouths of the 
 two rivers may be united by a more or less wide strip of 
 marsh and lagoon. These lagoons are liable to be filled up 
 by sand and silt carried down by streams from the higher 
 grounds. "When the bottom is brought up to the level of 
 low tide, or a little above, the seeds of the mangrove take 
 root, and the shoal soon becomes covered with vegetation. 
 Among the roots of the mangroves the silt of the water is 
 deposited, and the sand-bank is overspread by a layer of 
 soft mud, which may increase in thickness until the bank 
 is covered only at high water. The mangrove in Brazil, 
 Florida, and elsewhere is a very efficient agent in the silt 
 ing up of marsh lands and swamps ; but it is not alone in 
 this work. It flourishes only in salt or brackish water, but 
 after a time much of this swamp land becomes covered 
 with fresh water, when arborescent arums, reeds, rushes, 
 coarse grasses, and other aquatic plants tend all the more 
 to choke up and stagnate the water, and often form rafts of 
 vegetation, on which small trees grow, as is the case in the 
 fresh-water lakes and lagoons of the vicinity of Barra Secca 
 and the rivers Sao Matheos and Doce. 
 
 On the coast north of Rio the course of the rivers is at 
 right angles with the coast. At the mouth of very many 
 of them a branch runs off both north and south, parallel 
 with and close to the shore, and in some cases uniting with 
 a similar branch from a neighboring river, though usually an 
 impassable swamp prevents a complete communication. In 
 some cases these are* only lagoons fed by the tide and rains ; 
 but occasionally it is a little river, which, having come 
 
PROVINCE OF BAHIA. 223 
 
 down to the shore, is obliged to flow for several miles be 
 hind the beach ridge before it can find an exit to the sea in 
 the mouth of a larger river. An example of this kind is the 
 Itahunas, already described. All the rivers flowing from the 
 interior are white-water rivers, but the Itahunas, probably 
 having its source in the swamps of the chapadas east of the 
 serras, is a black-water river. So also is the Mariricu, and so, 
 in general, are all the little rivers which rise either in the 
 fresh-water swamps along the shore or come from the cha 
 padas eastward of the serras. Advantage has been taken, 
 in numerous instances, of the streams and lagoons of the 
 low lands to cut canals uniting settlements lying on two 
 rivers ; and a favorite project with some Brazilian states 
 men has been that of opening a line of canals, extending 
 along the coast from Santa Cruz, in the province of Espirito 
 Santo, northward to Caravellas. I consider the project as 
 impracticable, and that good roads over the plains of the 
 chapadas would, in most cases, be far preferable. 
 
 The town of Caravellas is of small size. It owes its im 
 portance to its being the port of the surrounding country, 
 and to its whale-fishery, described in the last chapter. The 
 town is built on a sand-bank a few miles above the mouth 
 of the river, and on the northern side. In the vicinity are 
 large groves of cocoa-palms. 
 
 Contrary to the general rule, the river Caravellas just be 
 fore reaching the coast makes a bend to the northward, and 
 enters the sea very obliquely. The channel, narrow and 
 marked by poles, continues for several miles northeastward 
 beyond the Ponta da Balea. There is another narrow chan 
 nel running eastward, and still another southward, which 
 last doubles sharply round the point on the southern side 
 of the river. Between these channels are large sand-banks, 
 which are, in some cases, visible even at high water. 
 
224 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The northward bend of the river at its mouth seems to be 
 owing to the prevalence of a northward-setting current. 
 The variation of the current through the tides is most prob 
 ably the cause of the existence of the other two channels. 
 Off this coast lie the coral reefs of the Parcel das Paredes 
 and the islands of the Abrolhos, already described. 
 
 The shore northward to Prado is low and flat, a long 
 monotonous sand-beach, which is broken by the Barra Velha, 
 the mouth of an inconsiderable stream, and, at a distance 
 of about nine miles north of the Ponta da Balea, by the 
 Barra do Rio Itanhaem,* which is a small river like the 
 Peruhype, arising in the Serra dos Aymore s. At its mouth 
 this river bends abruptly southward, and flows for at least 
 a couple of miles almost parallel with the sea-coast before 
 it escapes into the ocean. According to Prince Neu-Wied 
 manati have been captured in this river. Alcobaga, a small 
 town of very little importance, is situated on the seaside 
 between the river and the beach. Prince Neu-Wied says 
 that the country about AlcobaQa is healthy, but that the 
 climate is unpleasant from the frequency of strong winds 
 and storms. 
 
 Twelve miles farther north, the Rio Jucurucu empties 
 into the sea. This stream is formed by the union of the 
 two branches called respectively the Bra^o do Norte and 
 Brago do Sul, which take their origin in the Cordilheira 
 dos Aymore s. This river, like the Itanhaem, on reaching 
 the coast is obliged to flow southward for two or three 
 miles, behind a beach ridge, before escaping into the sea. 
 
 * According to Dr. Jose* Candido da Costa, in his pamphlet entitled " A 
 Comarca de Caravellas," a copy of which I owe to the kindness of Sr. Licino 
 Lessa, the Itanhaem is narrower and shallower than the Peruhype. Near its 
 mouth it forms a little basin. It is subject to freshets. 
 
PROVINCE OF BAHIA. 
 
 225 
 
 The bar of the river allows the entrance of schooners and 
 small vessels, and the river is said to be navigable for 
 small vessels (sumacas) for a distance of some six leagues. 
 The valley of this stream is rich in lumber, and in 1857, 
 according to Da Costa, almost all the inhabitants of the mu- 
 nicipio of Prado were occupied in cutting it. The town of 
 Prado, another unimportant place, is built between the river 
 and the sea on the left bank. A few miles north of the 
 Prado the tertiary bluffs, which from the Peruhype have 
 skirted the coast at a greater or less distance in the interior, 
 come down to the sea, and thence to Porto Seguro form 
 a long stretch of picturesque perpendicular red cliffs, alter 
 nating with steep slopes covered with verdure, and occa 
 sional patches of sands or swampy ground. 
 
 From the sea the horizontal bright red and white bands 
 of clays are distinctly seen. Near Porto Seguro, as well 
 as elsewhere, I observed that the valleys of the rivers enter 
 ing the sea did not have angular sloping sides, but their 
 profile as seen in a cliff was as in the following sketch, 
 
 which is a kind of topography that we should expect to see 
 in a glaciated surface. The bluffs would, I should judge, 
 average two hundred feet in height. 
 
 About forty miles north of Caravellas, and a few miles 
 inland, is the beautiful group of hills noted as being the 
 first land of America seen by Cabral. The most conspicu- 
 10* o 
 
226 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ous of these hills is an irregular conical peak which bears 
 the name of Monte Pascoal.* Mouchez sets the hill down 
 on his map as 536 metres in height, but I should judge that 
 it was much higher. These hills are undoubtedly gneiss, 
 but they lie far back in the virgin forest, in the as yet un 
 explored home of the savage Botocudo. The appearance 
 of the group as seen from the sea is represented in the 
 following woodcut. Along this coast several small rivers 
 rising in the Serra dos Aymores empty into the sea, but they 
 are of very little importance, and are usually incorrectly 
 laid down on maps. 
 
 MONTE PASCOAL FROM THE SEA. 
 
 * Monte Pascoal was the first point seen when Cabral discovered Brazil, on 
 the 21st of April, A. D. 1500. Pedro Vaz dc Caminha, one of the companions 
 of Cabral, has described the events connected with the discovery in a letter writ 
 ten on the 1st of May, 1500. This writer speaks of the mountain as very high and 
 round, with other serras to the south, together with flat land covered with large 
 trees. " E neste dia a oras dc vespora ouvemos vista de terra, a saber : primei- 
 ramente de huum gramde monte, muy alto e redondo, e doutras serras mais 
 baixas aho Sul dele, e de terra chaa com gramdes arvorcdos ; aho quaal montc 
 alto ho Capitam pos nome ho Montc Pascoal, e aa terra ha Terra da Vera 
 Cruz." Corofjrafia Brazilica, Tome I. p. 13. A French translation of this 
 letter is to be found in the Art de Verifier les Dates, Tome XIII. p. 441. The 
 original, which is of great interest, is preserved in the government archives at 
 Rio. Cazal says, speaking of the Serra dos Aymore s : " Cuja porcao mais 
 alta he o Monte de Joam de Liam [Jo5o de Leao] e mais fora o Monte Pnscodl 
 que se avista de muitas lequas ao mar." This Jo2o de Leao is a noted land 
 mark, but I can give no idea of its height 
 
PROVINCE OF BAHIA. 227 
 
 In front or to the east of Monte Pascoal lies the great reef- 
 ground of the Itacolumis. This is separated from a low 
 projecting sandy point called Ponta Corumba by a very nar 
 row but deep channel. A few miles north of this point 
 is the Rio Craminuan or Caxoeira, which is noteworthy 
 from the fact that, contrary to the general rule, on reach 
 ing the coast, instead of immediately entering the sea, it 
 flows northward nearly a mile close to the sea, but sepa 
 rated from it by a sand-bank, showing that the wash of the 
 coast sands is here northward instead of southward, a fact 
 determined by the reefs lying off the shore. North of the 
 Craminuan aretheRios Joassema, Frade,* and Taipe". The 
 village of Trancozo is situated on the coast a few miles 
 south of Porto Seguro, but it is of no importance what 
 ever. 
 
 At Porto Seguro enters the Buranhaem River, a stream of 
 moderate dimensions, which, according to the maps, rises 
 to the southwest in the province of Minas, a few miles from 
 the boundary line. Gerber represents this as a considerable 
 stream flowing through a large lake called Gravata, distant 
 some thirty miles from the sea, and communicating with an 
 other considerable lake about half-way between this and the 
 sea. At Porto Seguro I was informed that this was all very 
 incorrect ; but Since I have never ascended the river, I can 
 not speak authoritatively. The stream is a black- water one, 
 
 * In speaking of the coast between Prado and Rio do Frade, Max. zu Neu 
 Wicd says : "Als ich im November dieses Jahres noch einmal diese Reise 
 macbte, fand ich bei starker Ebbe weite Banke von Sand- und Kalk-Felsen, 
 die sich tief in die See hinaus erstrccken, und wohl grossentheils durch Corallen- 
 Thierc gebildet wordcn sind. Ihrc Oberflachc ist in regelmassige parallele Risse 
 getheilt ; in den vom Wasser darin ausgcwaschenen Luchern leben Krabben und 
 andere See-Thicre ; die Oberflache diescr Felsbanke iiberzieht zum Theil cine 
 griine Byssus-artige Masse." Prinz Max. zu Neu-Wied, Reise nach Brasilien, 
 Erster Band, 297" Seite. 
 
228 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and flows in a deep narrow valley cut in the tertiary beds. 
 The flat lands bordering it are said to be fertile, and favor 
 ably situated for agricultural purposes.* The forests are 
 rich in valuable woods, Pao Brazil, Jacaranda, <fcc. Porto 
 Seguro is quite a large and commercial town, situated on 
 the left bank by the seaside. It really consists of two 
 towns, one built by the river and seaside on a broad, flat, 
 sandy, and marshy tract, and part on the top of the bluff on 
 the northern side of the valley. The lower town is the 
 business portion ; the upper contains the ruins of ancient 
 churches, monasteries, &c. Porto Seguro is noted as the 
 head-quarters of the garoupa fishery of the Abrolhos. It 
 has an excellent harbor, protected in front by the coral reefs, 
 which break the force of the Atlantic waves, and by a reef 
 like that of Pernambuco, consisting of a solidified beach. 
 This last begins close to the shore a short distance north of 
 the mouth of the river, runs like a wall of rock in front of 
 the river, passing close to the point on the southern side, 
 and continues on to the southward with occasional breaks 
 for a distance of several miles. Its course is remarkably 
 straight, and its height and width are very regular. It 
 forms a more efficient breakwater to the harbor than the 
 reef at Pernambuco does ; but this is owing partially to the 
 fact that the sea is broken by the outside coral reef which 
 stretches across the mouth of the bay of Porto Seguro. 
 The river escapes around the northern end of the reef. On 
 the inner side it is overhanging, on the outer perpendicular 
 and much undermined, as represented in the woodcut. 
 As elsewhere its surface is diversified by ponds, in which 
 several species of corals grow. The northern end of 
 the reef is much shattered, cracked up, and dislocated, 
 
 * Lindley says that gold was found on one of the branches of this river. 
 
PROVINCE OF BAHIA. 
 
 229 
 
 MgTtWcLter. 
 
 SECTION ACROSS bTOME REEF AT PORTO SEGURO. 
 
 and it shows that the cementing of the beach-sands has 
 taken place to a depth of many feet below low-water level. 
 Beyond the reef, the rock slopes out with a smooth, rather 
 even surface to a considerable distance, and I have waded out 
 on it in some places a hundred feet or more. Its limits I 
 was not able to determine, owing to the surf. On the out 
 side the reef below low-water mark is covered with a growth 
 of corals. Stony corals, Hymenogorgice, Eunicice, and the 
 common polyps of the Abrolhos, are very abundant here. 
 I found a single specimen of Mussa Harttii in a deep in 
 dentation at the northern extremity of the reef. 
 
 As at Pernambuco, Bahia, Barra Secca, and elsewhere, 
 the rock is obliquely laminated, as in a sand-beach, the 
 lamina dipping seawards at a small angle. It is composed 
 of rather fine sand, with occasional small pebbles, compactly 
 held together by a calcareous cement. It contains an abun 
 dance of recent shells, Venus, Cerithium, Chama, &c., &c., 
 such as now live on the sea-beaches of the vicinity. On the 
 inside the reef-rock is of little thickness, and the reef is 
 flanked by a sloping bank of mud, on which a few man 
 groves have taken root. Oysters grow here on the rocks 
 and mangroves, and a large species of Littorina is very 
 common, being found even in the trees at a height of four 
 or five feet above the level of high water. This same 
 species occurs elsewhere, and I have observed it in great 
 
230 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 abundance higher than I could reach in the mangroves at 
 Santa Cruz, a few miles north of Port Seguro. Crustaceans 
 are abundant on the reef. 
 
 Little fiddlers ( Gelasimus palustris Edwards)* are very 
 common on the sandy beaches in some localities, boring 
 holes in the sand. There are larger species of the same 
 genus (Gelasimus Maracoani Latreille) with a nut-brown 
 body, and one of the hands enormously developed, looking 
 like a pair of broad-bladed shears. f 
 
 On the shore of the point on the southern side of the 
 mouth of the river is a small patch of beach, which is only 
 partially consolidated. This is separated from the reef by 
 a narrow channel almost laid dry at low water. Outside 
 the main reef, and opposite this, there are the remains 
 of probably an older reef, which has elsewhere been almost 
 entirely obliterated. Southward of the river the channel 
 soon widens into quite a broad sheet of water, and the reef 
 is left running at a distance of several hundred feet from 
 the shore. The water inside the reef is shallow, and at 
 low water one may wade about over a large part of the 
 area and collect. The bottom is sandy, but on it grows 
 an irregular crust, composed of millepores, corals, bryozoans, 
 <fec., which forms the nestling-place for holothurians, ophiu- 
 rans, crustaceans, and a thousand interesting animals. A 
 large naked mollusk, probably Aplysia Arg-o D Orb, is very 
 common here, and I have collected a dozen specimens in the 
 same little pool. Inside the reef the water is deep enough to 
 
 * This species bears, in every respect, an exceedingly close resemblance to 
 our northern G. vocans Dana. Mr. Sidney I. Smith considers the Brazilian 
 species as identical with one occurring in the Gulf of Mexico, and even so far 
 north as South Carolina. 
 
 t These fiddlers are called by the fishermen Charna mar? (call tide), because of 
 their congregating at low-water mark and waving their big claws. 
 
PROVINCE OF BAHIA. 231 
 
 admit vessels of ordinary tonnage and the coasting steamers. 
 The reef is higher than that at Pernambuco, if I judge 
 rightly, and less shattered ; but, as above remarked, the 
 Pernambucaii reef is exposed to the full action of the sea. 
 At high water the waves also break completely over the 
 Porto Seguran reef. Prince Neu-Wied gives a sketch of 
 Porto Seguro from the southern point. The width of the 
 river is very much exaggerated. 
 
 The line of the tertiary slopes continues on without inter 
 ruption to Santa Cruz ; but the shore on leaving Porto 
 Seguro soon separates itself from the chapadas, and forms 
 a considerable point of land, of which the interior seems to 
 be, for a considerable part at least, swampy. On the south 
 ern side the beach is backed by mangrove swamps, from 
 which the water is drained by a little stream that flows 
 across the beach. In passing very close to the point at low 
 tide, I observed that the waves were breaking along a line 
 at some distance from the shore, as if against the edge of a 
 reef. It would appear that the point has been formed by 
 the filling up of the channel behind a coral reef with sand. 
 Santa Cruz, one of the most ancient towns of Brazil, and at 
 the same time one of the most miserable, is built on the 
 southern side of the mouth of the little river Santa Cruz, 
 partly on the top of the bluff and partly on the sands at 
 the base, between the bluffs and the sea. Though it has 
 a fine harbor, and many natural advantages, it is of not 
 the slightest importance, and is only a miserable little fish 
 ing village. It is situated in a shallow bay, about seven 
 miles long, which is protected by coral reefs, that extend 
 across it, offering anchorage for large ships. 
 
 The Rio Santa Cruz, anciently called Joao de Tiba, be 
 longs to the same class of rivers as the Buranhaem, but it 
 
232 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 is a smaller river. It is said to rise in the Serra dos 
 Aymores, to have a course of about ten leagues, and to be 
 navigable for a considerable distance for canoes. It is laid 
 down incorrectly on maps. At Santa Cruz I was informed 
 that its course was such as to cause it to approach the Je- 
 quitinhonha, from which, at Zinebra, it is separated by only 
 a very short distance, and I was informed that it would ad 
 mit of navigation up to that point by a small river steamer. 
 Prince Xeu-Wied says that the river has two branches, and 
 that the head-waters of one of them lie so near the Jequi- 
 tinhonha, that the report of a gun can be heard across the in 
 tervening space. The valley of the Santa Cruz is of the same 
 character as that of the Buranhaem. It is fertile and 
 richly wooded, furnishing building timber and some Jaca- 
 randa and Pao Brazil. It is noted for the abundance of 
 canna fistula (Cassia nigra), a tree valuable for its medici 
 nal properties. 
 
 The river on reaching the sea is prevented from flow 
 ing immediately into it by a recife, or consolidated beach, 
 which, beginning on the shore just to the south of the vil 
 lage, continues in the trend of the beach, which is north 
 a few degrees east, with an occasional break for a distance 
 of about two miles, the river flowing behind it and escaping 
 around its northern extremity. At low water the breakers 
 show that the reef is continued under water with the same 
 general trend northwards, tying in with a reef which, be 
 ginning at a point about a mile north of the river-mouth, 
 fringes the beach for more than half a mile, as represented 
 in the accompanying sketch-map of the bays of Santa Cruz 
 and Cabral, in which not only the stone but coral reefs are 
 shown. Mouchcz has very incorrectly represented the har 
 bor in his chart. Through the kindness of my friend 
 
PROVINCE OF BAHIA. 
 
 233 
 
 Hugh "Wilson, Esq., superintendent of the Bahia Steam 
 Navigation Company, I was enabled to revisit Santa Cruz 
 in 1867, and to correct Mouchez s chart by drawings made 
 from the top of the old church on the bluff, and these 
 corrections I have introduced in the map. In front of the 
 
 town the reef clings to the shore, though there is a channel 
 behind it filled at high water. The reef is much shattered, 
 and great blocks lie tumbled about in confusion, broken 
 from it by the waves. For a distance of a mile or so 
 north of the village the reef is backed by a narrow strip 
 of mangroves. There is then a narrow break in the reef, 
 forming a sort of bar, into which very small vessels may 
 enter at high water. Thence to the barra the reef is backed, 
 almost to the end, by a wider strip of mangroves. At the 
 extremity the reef is double, the remains of an older reef 
 being visible on the outer side. An example of a triple 
 reef is represented in the chart of the mouth of the Rio 
 Parahyba do Norte, published in the old work of Barlasus, 
 
234 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 in 1660.* I have never seen this reef, and know nothing 
 of its present appearance. The bar of the Santa Cruz 
 is good, and may be safely entered by coasting steamers or 
 large vessels. At the time of my last visit I saw a brig 
 lying behind the reef at the bar. There is sufficient water 
 to enable steamers and vessels of considerable size to go 
 up to the town. This harbor could be very much improved 
 by stopping up the channels by which part of the river water 
 escapes near the town as well as the break in the reef above 
 described. This would cause the whole force of the stream 
 to be expended on the proper channel of the river, and it 
 would in this way be kept from being filled up by mud and 
 sand banks. The country in the vicinity of Santa Cruz, 
 embracing the region of the Jequitinhonha, is so naturally 
 rich, that as the coast becomes settled it must one day be 
 developed, and one cannot doubt that Santa Cruz will ulti 
 mately become the port of the Jequitinhonha, and a place 
 of much importance. From Santa Cruz the bluffs run 
 northward, following close to the shore, with occasionally 
 some sandy tracts in front, but they gradually trend off 
 from the coast and cross the Jequitinhonha between Zinebra 
 and the Po-assu. 
 
 A few miles north of the Santa Cruz is the Eio Sant 
 Antonio, a little black-water river of no account. About a 
 mile and a half north of this river I saw exposed at low- 
 water, in a little bay, an area of several acres of dead coral, 
 principally Acanthastrcea, &c., in situ, but much worn by 
 the waves. These corals must have grown when the sea 
 bottom over this region stood at a considerably lower level 
 than at present, for I have seen them nowhere living at a 
 
 * For a reduced copy of a part of the chart of Barlasus, as well as a recent 
 sketch map after Almeida, see chapter on the Province of Parahyba 
 
PROVINCE OF BAHIA. 235 
 
 level where they would be likely to be exposed at low water. 
 When they were growing the sands forming the low grounds 
 bordering the coast here could not have had so wide an 
 extension seawards. This reef patch has evidently been 
 killed by the joint rise of the coast and the encroachment 
 of the shore sands. I collected on the shore here a great 
 number of specimens of the pancake sea-urchin (Encope 
 emarginatus^) , so common in the Bahia do Todos or Santos. 
 Neu Wied speaks of the great abundance of this echinoid 
 on the shore near Sant Antonio. He refers it to Echinus 
 pentaporus, or as he writes it, Eschinus pentaporus. 
 
 Thence northward the coast is bordered by a narrow 
 strip of plain, which widens as the bluffs recede from the 
 coast, and finally opens out into broad, sandy campos, which 
 extend from the Mugiquisaba River, a small black-water 
 stream, to Belmonte. It is interesting to observe how this 
 whole plain has been added to the coast, by the throwing 
 up by the sea of the sands of the Jcquitiiihonha, and more 
 over that this growth has taken place to the south of the 
 river rather than to the north, because of the sweeping of 
 the sauds southwards by the oblique beat of the waves, and 
 perhaps by the drift of a southward flowing current. In 
 going northward from the Mugiquisaba to the Jequitinhonha 
 the water gradually shallows. The result is that the surf 
 beats more heavily on the southern part of this coast, 
 and we find a high beach-ridge developed, while, near the 
 mouth of the Jequitinhonha, the water growing shallow, 
 the bottom is disturbed farther out by wave action, and the 
 sands are constantly being driven in shore to widen the 
 plain, and of course extraordinarily heavy storms are here 
 likely to throw up beaches outside the present shore line. 
 The plain consists of a large number of parallel beaches, 
 
236 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 one lying in front of the other, sometimes united, at others 
 separated by miniature valleys, occasionally only a few feet 
 in width, but often continuous for a considerable distance. 
 Many of these beaches have their slopes almost as perfect 
 as if it were but yesterday that they were swept by the 
 waves. In the depressions water accumulates, sometimes 
 forming shallow lagoons. The present beach on going 
 southward from the barra separates itself from the plain, 
 and at the same time increases in height as the surf action 
 becomes heavier. Behind this beach is a narrow river-like 
 lagoon, called the Lagcia do Brae/). This is at first narrow 
 and shallow, but to the south it deepens and widens as the 
 beach-ridge becomes separated from the plain. Its waters 
 are fresh and clear, and very rich in fish. Camboas, or 
 fish-weirs, are common along its shores, which are muddy, 
 and to a very large extent support a luxuriant growth of 
 mangroves and guaxuma-bushes. In making a voyage to 
 Porto Seguro my companion and I became separated from 
 our baggage troop, which was following the beach, and, 
 taking a road on the western side of the Lagoa, were un 
 able to cross it anywhere, because of the swampy nature of 
 the banks and the depth of the water. Prince Xeu-Wied 
 says that about one half of the way between Mugiquisaba 
 and the Jequitinhonha is the barra where a clried-up arm 
 of the latter river once emptied into the sea. By the 
 dried-up arm he probably means the Lagoa do Brago, which, 
 as he travelled along the sea-beach, he probably did not ex 
 amine, as he says nothing about it. I saw no outlet to the 
 Lagua. When swollen it may break through the beach 
 ridge, as is the case with other seaside lagoons. It ap 
 pears to be drained principally by the slow percolation of 
 its waters through the beach into the sea. 
 
PROVINCE OF BAHIA. 237 
 
 The plain is to a large extent open and very sparsely 
 covered with coarse grass, bromeliaceous plants, cactuses, 
 <fec., with here and there clumps of trees. The Aricuri 
 palm is very common on this plain, together with the 
 pretty dwarf Guriri. This latter, all along the whole coast, 
 is found growing on and just back of the sea-beaches, and 
 is one of the marked elements of the beach flora. The 
 Piassaba palm (Attalea funifera Mart.) flourishes on the 
 Mugiquisaba, as remarked by Neu-Wied. I have not seen 
 it south of this point. Cactuses, with procumbent pris 
 matic stems, form large patches on the sands. Another 
 very characteristic plant of the coast sands is the Ipomosa 
 littoralis, a convolvulaceous plant, with long, thick, cord-like 
 creeping branches, pink blossoms, and large thick oval 
 leaves. This plant grows sometimes on the beach almost 
 within reach of the waves, and its prostrate stems are often 
 buried by the sands. 
 
 The right bank of the river, for some eight or ten miles 
 above its mouth, is very low and liable to be flooded. A 
 very large part consists of mangrove swamps. The Po- 
 assii * is a very narrow ditch-like canal which runs through 
 the alluvial grounds, leading off part of the waters of the 
 Jequitinhonha into a small black-water river called the 
 Salsa, which flows into the Pardo. So much higher do 
 the waters of the Jequitinhonha stand above those of the 
 Pardo, that, though the tide rises and falls in the latter 
 river where the Po-assu enters it, there is a constant flow 
 of water from the Jequitinhonha into the Pardo. The 
 Po-assu is so narrow, tortuous, and filled up with trees, that 
 
 * This name is Tupi. Pd-a$u means left hand, so that it might mean left- 
 hand channel. Yg-apti means a swamp. I am inclined to think that the name 
 was first Yg-ap6-acu, or the great swamp, which has since passed to the canal. 
 
238 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 its current is measurably slackened, else it would sweep out 
 for itself a broader channel and draw off a larger quantity 
 of water from the Jequitinhonha. What the consequence 
 of this would be I shall state further on. The soil of the 
 grounds bordering the Po-assu is massape", and supports a 
 very vigorous and thick forest growth.* From the Salsa, after 
 having received the Po-assu, side arms stretch out eastward 
 into the flat grounds lying between it and the sea. The 
 whole area embraced between the Pardo and Jequitinhonha, 
 the cross-stream and the sea, is one vast swamp, compar 
 able to that which lies between the Peruhype and the 
 Caravellas ; the Po-assu being comparable to the Bra^o de 
 Vic,osa, both the Po-assu and Brac,o being channels by 
 which the waters of a river with an inefficient mouth- 
 opening are enabled to escape into the sea through the 
 mouth of another river. There are some quite extensive 
 sandy tracts in this area and along the coast. These are 
 planted with groves of cocoa-palms, from the fruit of which 
 cocoa-oil is to some extent manufactured. A large tract 
 lying northward of the Salsa, and embraced between it 
 and the Pardo, is of the same character. The Pardo is a 
 large river, and has a bar on which the waves beat fearfully, 
 but it usually admits at high water of the passage of 
 steamers under the proper pilotage. The Pardo does not 
 bring down so much sediment as the Jequitinhonha, its 
 
 * In the forest of the Po-assu the Quitdra (Dcsmoncus), a trailing palm, 
 is exceedingly common, interwoven with the trees. I saw it in fruit early 
 in May. The fruit is round, about the size of a small cherry, bright red, 
 and in clusters like grapes. This is a common species found in the Catinga 
 woods on the low grounds along the coast, sometimes in such great abundance 
 as to be a nuisance, for it hangs its pendant leaves, terminated with hooks, over 
 the mule-path. The name Quitara is the one given it by my guide. On the 
 Rio Negro the Desmoncus macroacanthus is called Jadtdra. 
 
PROVINCE OF BAHIA. 239 
 
 channel is deeper, and the tide runs up the river for quite a 
 number of miles. 
 
 Just before entering the sea the river makes a bend south 
 ward, and the barra is really cut through a strip of sand-beach 
 which forms the shore for a long distance north and south. 
 A wide channel, or riacho, leaves the river on the south just 
 before it cuts through the sand, and runs through the man 
 grove swamps for a considerable distance southward. In 
 journeying from Cannavieiras* to Belmonte I followed this 
 route. My companion and I took canoe up this riacho as far 
 as was possible, then coming to a sandy tract, we chose the 
 sea-shore for some distance, when we embarked in a canoe on 
 another riacho which led southward to the Jequitinhonha, a 
 dismal, but to the naturalist exceedingly interesting journey. 
 There was a perfect network of channels leading through 
 the dense mangrove growth, and for some distance little 
 land was to be seen. The mangroves standing in the 
 mud or dirty water, with their bared basal roots, their fre 
 quent aerial roots hanging ropelike from the branches, and 
 just before reaching the soil forking like tripods, their dense 
 green foliage meeting overhead, and their curious cigar- 
 shaped seeds pendent in the tree or occasionally sticking 
 upright in the mud, were objects of wonder and admira 
 tion ; but among their roots innumerable fuzzy and muddy- 
 legged Guayamti crabs glared at us with expressionless eyes, 
 and then hustled away into their holes in the black mud. 
 Beautiful orange Aratus, with white fingers, lay like rich 
 fruit fallen from the trees; but they, too, took fright and 
 ran nimbly off to their holes, or hid in crevices under the 
 roots, while myriads of fiddlers marched away, waving their 
 
 * This is the usual orthography, but one meets with Canavieira, Canavieiras, 
 or even Canasvieiras. 
 
240 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 big arms, and took refuge in the sandy spots. Parrots * 
 screamed among the branches ; but, peste ! mosquitoes and 
 all the horrible blood-sucking race of Maroims, Piums, 
 &c., swarm ; and, how ardent soever the naturalist may 
 be, his patience can ill stand the plague of flies that falls 
 upon him in these swamps. The air of the swamps is 
 very unhealthy, and the vicinity of Belmonte and Canna- 
 vieiras is feverish. The town of Cannaviciras is situated 
 on the left bank of the river, just above the mouth. It 
 is built on an old beach, an island surrounded by a chan 
 nel, which, leaving the river some distance above the town, 
 enters it again just before the river reaches the sea. The 
 town is quite a considerable place, numbering, perhaps, two 
 thousand inhabitants. It derives its chief importance from 
 the fact that it is the port of the Jequitinhonha and Pardo, 
 and is one of the stations of a line of steamers. Canoes de 
 scending the Jequitinhonha do not go to Belmonte, but come 
 here by way of the Po-assu and Salsa, to exchange the 
 cotton, corn, and other articles of export of Minas and the 
 lower river for salt, dry goods, &c. The salt trade is very 
 large. In descending the Jequitinhonha, we passed every 
 day several large canoes on the up voyage, the most of 
 them carrying salt. 
 
 It has been proposed to straighten and widen the Po-assu, 
 so as to facilitate the navigation between the two great 
 rivers. At present the waters of the Pardo during the annual 
 freshets wear away the unstable land on which Cannavieiras 
 stands. Were the Po-assu widened and a free passage opened 
 for the waters of the Jequitinhonha, the result would prove 
 disastrous to Cannavieiras as well as to Belmonte. At pres- 
 
 * Psittacus ochrocephalus Linn, builds its nest in the mangroves. Sec Ncu 
 Wicd. 
 
PROVINCE OF BAHIA. 241 
 
 ent the struggle between river and sea is only sufficient to 
 keep the bar of the Jequitinhonha open. Draw off any con 
 siderable part of the water of that river, and it is very 
 doubtful whether its bar would not become a permanent 
 obstruction to navigation, while the sands of the Jequi 
 tinhonha, thrown into the Pardo, would probably shoal the 
 water and render its bar worse. Santa Cruz is the natu 
 ral port of the Jequitinhonha, and it would certainly seem 
 that the commerce of the Jequitinhonha would be vastly 
 improved by using the river Santo Cruz so far as it is 
 navigable by steamers, and then building a good wagon- 
 road thence to the Jequitinhonha. 
 
 The Pardo is joined with the Poxim, a small river said 
 to take its rise in a large lake, and whose barra is a few 
 miles north of that of the Pardo, by a narrow arm running 
 parallel with and close to the shore, and called the Patipe. 
 Cannavieiras itself is situated on an island formed by a 
 channel called the Eio Sip6 which leaves the Pardo and 
 joins the Patipe. The Poxim, just before it unites with this 
 last river, subdivides and enters the sea by two mouths, one 
 of which is called the Barra do Patipe, and the other on 
 the north the Barra do Poxim. One or two channels 
 continuing northward from the Poxim like the Patipe, 
 running just inside the coast line, empty into the sea a 
 couple of leagues to the north at the Barra de Com- 
 mandatuba. This whole coast has been very indifferently 
 mapped inside the coast line. The Jequitinhonha and 
 Pardo are often shown as entering the sea by the same 
 mouth. 
 
 The map of Senhor Henrique Gerber, so excellent for the 
 provinces of Minas Geraes, Rio de Janeiro, and Espirito 
 Santo, is exceedingly defective so far as the province of 
 
 VOL. I. 11 p 
 
242 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Bahia is concerned ; but it must be remembered that his 
 map does not pretend to be a map of Bahia. 
 
 This part of the coast is better shown on the map of the 
 Tenente Manoel Ernesto de Souza Franca, published in the 
 report of the President of Bahia in 1866 ; but even this 
 only gives the general features, and is not based on a care 
 ful survey. 
 
 According to Prince Neu-Wied s edition of Arrowsmith s 
 Map of the Brazilian Coast, no connection is shown as exist 
 ing between the Jequitinhonha and Pardo, and the latter 
 river is represented as dividing into three a long distance 
 from the coast, two of which unite before entering the sea, 
 while the Poxim and Commandatuba, though joined to 
 g-ether, have no connection with the Pardo. 
 
 o 
 
 The Diccionario Geographico declares that the Rio Pardo 
 divides into two streams, one of which enters the sea under 
 the name of the Rio Pardo, the other emptying into the Bay 
 of Ilheos under the name of Cachoeira or Patipe, which 
 is all simply ridiculous. 
 
 Mr. Copeland and I ascended the Pardo to the head of 
 canoe navigation, a journey of about three days, and the 
 following notes were made on the return voyage. 
 
 At the Caxoeirinha do Rio Pardo the river reaches the 
 low country and becomes a rio de arcia. Here navigation 
 is obstructed by a series of rapids caused by the river falling 
 over ledges of slate conglomerate, of which the dip was 
 found to be 45 to the S. 10 W., a dip corresponding with 
 that of no other strata I have studied on the Brazilian 
 coast. The material is a very highly altered conglomerate, 
 composed of pebbles of milky quartz, granite with quartz in 
 lamellae, c., imbedded in a slaty mass. The rock is ex 
 ceedingly hard, and appears to resist decomposition. The 
 
PROVINCE OF BAHIA. 243 
 
 stratification is not very distinct. On the water-worn sur 
 faces the pebbles stand out very prominently, but a fracture 
 passes straight through pebbles as well as cement. At 
 the Caxoeirinha these rocks are overlaid by tertiary beds, 
 which form plains elevated three hundred feet, more or 
 less, above the river, and which descend with very steep 
 slopes to the stream. The alluvial deposits along the river 
 reach a height of about twenty-five feet above the river 
 level. During the enchente these are sometimes at least 
 overflowed. The lands here are very fertile, and the whole 
 country appears to be densely wooded, but uninhabited. 
 
 Slate conglomerate and sandstone continue to show them 
 selves in the river-bank for a mile or more below the Caxoei 
 rinha, when to these rocks succeeds a fine-grained bluish 
 slaty rock, an altered shale, in which I could find no trace 
 of fossils. 
 
 A few rods above the fazenda of Sisterio there are some 
 ledges of a fine-grained, very hard, bluish, compact, altered 
 sandstone, with occasional bands of grit and conglomerate, 
 and often very beautifully obliquely laminated. In this rock 
 I observed remains of plants, and one surface of rock had the 
 impressions of several stems of a thick-noded equisetaceous 
 plant. One stem showed three nodes and another two, the 
 length of the internodes being about three inches. Owing 
 to the hardness of the rock, I was able to bring away only 
 the impression of a single node. This plant resembles in its 
 swollen nodes a plant not uncommon in the upper Devonian 
 rocks of St. John, New Brunswick, and called Astcrophyl- 
 lites ? scutigera Dawson. These beds appear to me to be 
 palaeozoic. 
 
 Below this point the river valley becomes wider and the 
 alluvial lands more extensive, though the river itself is not 
 
244 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 so very wide, as it averages for a considerable distance only 
 about three hundred feet. The country is abundantly for 
 est-clothed, and the river-banks, which are very fertile, are 
 more or less cultivated, several large fazendas being scattered 
 along the river. Some of these are very picturesquely situ- 
 ted, and are surrounded by orange, banana, jack and cocoa 
 trees. Cacao ( Theobroma Cacao} is quite extensively cul 
 tivated here. This plant is a native of the Amazonas, where 
 it flourishes almost without culture ; * the cacaoeiros, when 
 once they have begun to bear fruit, requiring little care. It 
 needs a warm damp climate and a rich alluvial soil, and 
 appears to suffer nothing from an occasional freshet. South 
 of the Amazonas it is cultivated but rarely, though even in 
 the province of Bahia there are some quite large plantations, 
 especially at Ilhdos on the Pardo, and at Valenga. I saw 
 no cacao in the province of Espirito Santo. A little is cul 
 tivated in the province of Rio de Janeiro. In the Amazon 
 ian region the fruit, when ripe, is collected twice a year 
 and dried in the sun ; the seeds are then separated from the 
 shells, and are employed principally for the manufacture of 
 chocolate or other preparations for beverages. They furnish 
 a thick yellow fatty substance which is sometimes extracted 
 and used for various purposes. 
 
 Coffee does not produce well here. The trees grow very 
 rank and high, with spreading branches ; but the berries ripen 
 very unequally, and not unfrequently have to be gathered six 
 times during the year. This is owing to the peculiar climate 
 of the region. In Rio de Janeiro, Minas, and elsewhere, 
 there are well-defined wet and dry seasons during the year, 
 and these appear to be necessary for the successful culture 
 of coffee ; but here on the Pardo a very large quantity of 
 
 * See Bates, The Naturalist on the Amazons, pp. 87 and 162. 
 
PROVINCE OF BAHIA. 245 
 
 rain falls, and is distributed throughout the entire year, 
 making the climate very damp, so that while it is especially 
 adapted for the culture of cacao it does not do for coffee. 
 On the higher chapada lands, however, where the soil is 
 sandy and drier, coffee and cotton may be cultivated. 
 Alluvial, flat, damp lands are nowhere proper for coffee, 
 which flourishes best on hillsides. 
 
 The jack (Artocarpus Braziliensis Gom.) is cultivated 
 in Brazil, particularly in the province of Bahia and to the 
 north, though I have seen it at Sao Matheos, and occasion 
 ally as far south as Rio. It also occurs in Minas Geraes, 
 but in the province of Bahia it becomes of considerable im 
 portance. The timber is valuable for building purposes, being 
 very durable and strong. The fruit is immense, being some 
 times a foot and a half in the longer diameter. It consists 
 of a stringy, mucilaginous, sweet, and nutritious pulp, in 
 which are imbedded large seeds, which, when cooked, are 
 edible and nourishing, and are largely used for food. In 
 some parts a kind of farinha is prepared from them, but 
 its use is by no means general. The Fruita pao or bread 
 fruit, Artocarpus incisa, is also cultivated in Brazil, and 
 may be seen very frequently in the province of Bahia. 
 
 The forests on the banks of the Pardo remind one of the 
 Doce in their luxuriance. The trees by the river-side are 
 loaded with parasites and interlaced with pendent rope-like 
 cip6s. Ferns are very numerous, and one species, with 
 beautiful fringed pinnae, climbs up the trees to a height of 
 forty feet or more. One sees here a species of grass called 
 Capim da Colonia, (an exotic ?) which has been introduced 
 on the Pardo, and has spread within the last few years 
 quite extensively over the river-banks. Jit the Estreito 
 there is an isolated patch of tertiary, but below that the 
 
246 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 lands bordering the river are all low. A short distance be 
 low the Furado, a channel cut across a bend in the river, 
 the lands become lower and less heavily wooded, and 
 aninga and Guaxuma bushes appear on the banks, an 
 infallible sign of the approach to salt or brackish water. 
 There is a fern common to the mouths of the rivers on the 
 coast which appears to prefer this kind of water. It has 
 a tall, erect, narrow frond, with stiff, long, narrow pinnules 
 bent upward toward the stem.* It grows abundantly in the 
 vicinity of the Furado, on the muddy banks reached by the 
 brackish water. 
 
 From Cannavieiras to the Rio Poxim runs a canal, sepa 
 rated from the sea by a beach-ridge, and thence northward 
 this same canal extends to the Commandatuba, and is said 
 to afford water communication to the latter river from the 
 Pardo. The Commandatuba flows into this channel, which 
 is prolonged just behind the beach for some distance north 
 ward before it opens out into the sea. At the Barra da 
 Commandatuba a colony has been established, and there is 
 a little village of thirty or more houses there. The shore be 
 tween the Commandatuba and the Una is low and flat, but 
 a short distance inland one sees the tertiary slopes, which 
 stretch along northward from the Jequitinhonha and Pardo. 
 
 The Una is a little river which, according to Prince Neu 
 Wied, is so dry at ebb tide as to be easily forded. There 
 are some excellent lands on this river, and the forests are 
 rich in jacaranda and other valuable woods. This river is 
 noteworthy for bending northward just before reaching the 
 sea, and flowing a little distance behind the beach-ridge. 
 
 Opposite the mouth of the Pardo several high hills, lying 
 t 
 
 * This must be a species well known, but in the absence of specimens I 
 cannot give its name. 
 
PROVINCE OF BAHIA. 247 
 
 some twenty miles inland west of Commandatuba, are visible 
 at sea, and appear to be gneiss, but when off the Una a 
 heavy mountain range, with outlines like those of the gneiss 
 hills north and south, is seen stretching oif northward, and 
 traceable beyond Ilheos, probably tying in with the moun 
 tains of Camamu. This range is called the Serra de 
 Itaraca.* Some of the hills in this range must be three thou 
 sand feet in height or more. I insert here an outline sketch of 
 this coast, to show the character of the topography. I was 
 
 much struck with the outline of the table-topped hill on the 
 right, which appears to be composed of soft horizontal rocks. 
 May it not be an outlier of the great tertiary sheet that 
 covers so large an area in the interior, and which in some 
 places must have extended down to the coast before the 
 deposition of the coast tertiary beds ? I would suggest a 
 comparison between it and the hills of Monte Alegre on the 
 Amazon as. 
 
 From the Una northwards the tertiary plains descend to 
 the coast and extend, with high, steep, wooded, rarely per 
 pendicular slopes, seaward to Ilheos. A few insignificant 
 little streams empty into the sea along this coast. As a 
 general thing, a strip of flat sandy or marshy ground 
 runs along under the bluffs. The Piassaba palm (Attalea 
 funifera) grows very abundantly on the slopes and the 
 plain above, as Prince Neu-Wied has remarked. This 
 author says that he did not see it anywhere north of Ilheos. 
 It now covers large tracts on the Rio Paraguassu, below 
 
 * Prince Neu-Wied says that gold is found here. 
 
248 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Caxoeira, and along the Bahia and Sao Francisco, and its 
 fibre is quite an important article of commerce. Prince Neu 
 Wied speaks of finding rounded fragments of pumice * on the 
 shore near this place, and states his belief that they were 
 drifted to the Brazilian coast from the island of Ascension.! 
 He speaks of finding them on the shore near Porto Seguro, 
 and I have picked up large quantities on the island of Santa 
 Barbara dos Abrolhos. 
 
 The little village of Olivenca is situated on the top of a 
 hill by the shore, about nine miles south of Ilheos. Its in 
 habitants, civilized Indians, employ themselves in making 
 straw hats, baskets, <fec., which they send to Ilheos by way 
 of the beach, on mule-back. They plant only enough to 
 support themselves. 
 
 As above remarked, in speaking of the Rio Pardo, it was 
 believed that that river divided, part of its waters flowing 
 into the sea at Cannavieiras, while another part, under 
 the name of the Rio Cachoeira, flowed into the bay of Sao 
 Jorge dos Ilhe os. There is a road, or rather trail, through 
 the forest following this river into Minas, and Prince Neu 
 Wied travelled over it, crossing the head-waters of the 
 Cachoeira. He says nothing about that river being only 
 an offset from the Pardo, and in his map of the coast be 
 tween the twelfth and fifteenth parallels, after Arrowsmith, 
 he represents the Rio Cachoeira as taking its source very 
 near the Pardo, at some considerable distance below the 
 Villa da Vareda, and this is probably correct. Just above 
 the mouth of the river a small river called the Rio do Fun- 
 dao empties into it. 
 
 * Rei se, Vol. II. p. 85. 
 
 t This hardly seems probable, since there is a southward-flowing current 
 setting along the Brazilian coast south of Cape St. Roque. 
 
PROVINCE OF BAHIA. 249 
 
 In the year 1817 Prince Max. zu Neu-Wied made a 
 journey into the interior from Ilhdos to the borders of Minas, 
 and then went by land through the interior to the Bay of 
 Todos os Santos. His voyage is full of interest, but I ex 
 tract from it only a few points which bear upon the geogra 
 phy, natural history, &c. of the country. 
 
 His picture of the forest on the river near Ilh^os is so 
 grand, and so true to nature, that I translate a portion 
 of it : 
 
 " Life and luxuriant plant growth is spread everywhere. 
 Nowhere is there a little spot without plants. On all stems 
 species of Passiftora, Caladium, Dracontium, Piper, Bego 
 nia, and Epidendrum, besides many ferns, lichens, and 
 mosses of different kinds, bloom, climb, luxuriate, and at 
 tach themselves. The forest is made up of the genera Cocos, 
 Melastoma, Bignonia, Rhcxia, Mimosa, Ingd, Bombax, Ilex, 
 Laurus, Myrthvs, Eugenia, Jacaranda, Jatropha, Visinla, 
 Lecythis, Ficiis, and a thousand other, for the most part, 
 unknown species of trees, whose fallen flowers one sees 
 lying on the ground, and can hardly guess from which giant 
 tree they came. Others covered with blossoms shine afar, 
 white, bright yellow, bright red, rose-red, violet, sky-blue, 
 &c., and in swampy places there rise, thickly crowded to 
 gether, the great elliptical leaves of the Heliconias on 
 long stems often ten to twelve feet high, and noteworthy 
 for their bright red or fiery flowers. On the highest 
 trunks, high up in the forks of the branches, grow immense 
 tufts of Bromelias, with large clusters \_Blumenkolben und 
 Traubeii] of light cinnabar-red flowers, or of some other 
 color ; from these fall great bundles of roots like cords, 
 which hang down to the earth and form a new obstruction 
 for travellers. Such bromelia-clumps fill all the trees, 
 11 * 
 
250 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and when with the lapse of years they die, and are dislodged 
 by the wind, they fall with a crash. A thousand llianas, from 
 the slightest thread to the thickness of a man s thigh, and of 
 hard, tough wood (Bauhinia, Banisteria, Paulinia, <fcc.), 
 entwine the trunks, climb even to the highest top of the 
 tree, where they blossom and bear fruit where no human 
 eye can see them. Many of them are so wonderfully 
 formed that one cannot look upon them without amazement, 
 as, for instance, certain species of Bauhinia. From many of 
 these the trunk around which they had wound decays, and 
 here stands a giant coiled serpent, whose origin makes itself 
 easily understood in this way." * 
 
 In penetrating into the interior the country becomes grad 
 ually higher, and at the same time drier, while the forest 
 becomes less and less luxuriant, and finally passes, on the 
 higher ground, into a catinga, which begins, on the road to 
 Minas, at Porto da Canoa, on the Rio Cachoeira. Bromelia- 
 ceous plants become more abundant, with several species of 
 Solanacece, Mimosas, and the stinging Cansancao (Jatroplia 
 urens*). In the valleys, however, the forest is still dense and 
 thick. In the catinga grows a cactus (Cereus) with im 
 mense stems reaching a height of fifty or sixty feet, with a 
 diameter of two feet. On the western side of the Ribeirao 
 da Issara is a range of hills, called the Serra da Sussuarana, 
 which are not very high, but covered with masses of 
 loose rocks and stones, with a thick growth of catinga. 
 The country onward to the Giboya, a little stream flowing 
 southward into the Pardo, is covered with catinga. Here, 
 as elsewhere, the low grounds of the valleys are filled with 
 high forests, but the woods on the slopes and high lands 
 grow lower, and come under the class of catinga. The 
 
 * Raise, Band II. Scite 106. 
 
PROVINCE OF BAHIA. 251 
 
 Prince says that the Giboya flows over Granit-tafeln 
 (gneiss). I should judge from his description that the 
 country between the Giboya and Ilh^os was of the same 
 character. A short distance to the westward of the Giboya 
 is a range of mountains, " whose hills, of a considerable 
 height, have a rounded outline, and are strewed over with 
 masses of rocks and granite blocks in which especially very 
 large pieces of white quartz occur. The whole vicinity is 
 overgrown with very thick forest or catinga. These 
 mountains bear the name of Serra do Mundo Novo. The 
 first mountain is the highest ; it rises, it is true, with gentle 
 slopes, but it requires a full hour to ascend it. Thence on 
 ward hills and valleys alternate, until one at last descends 
 into a considerable depression. The Rio Pardo flows to the 
 
 left, in a deep valley parallel with the road So soon 
 
 as we had left behind us the fatiguing Serra we found the 
 wood still more changed into catinga, for in the depression 
 itself it was only 40 - 60 feet in height, filled with many 
 bromelia and cactus clumps, hung with moss tufts ( Tillan- 
 dsla), arid intermingled with many kinds of trees which 
 reached only an inconsiderable height. Here is found the 
 Pao de Leite (in all probability a Ficus), which is feared 
 because of its corrosive milky juice, but nowhere was seen 
 the beneficent, nourishing milk of the Palo de Vaca, which 
 Humboldt has described ; this milk would in our situation 
 be a great comfort. Farther on we found the cork-like 
 Barrigudo-tree (Bombax), which here grew to only a small 
 height, many species of Mimosa, of Bignonia, and so forth, 
 and between them rock masses and granite blocks. All this 
 shows that one has gradually ascended from the wet, dark 
 region of the coast forests to higher and drier country." 
 This description is very interesting, because it shows that 
 
252 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 here we have a region which has not been so much affected 
 by decomposition as the coast belt, and it shows us how the 
 forest zone narrows down as we go northward. 
 
 Beyond the Barra da Vareda one enters a catinga wood, 
 and gradually ascends, the hills being gently rounded, and, 
 as the Prince remarks, announcing the open plains and 
 ridges which make up so large a part of the interior of Bra 
 zil. " The wood has in many places lagSas grown up with 
 swamp reeds, in others extensive naked places which have 
 been burned over so as to produce grass for the cattle. Such 
 places become covered immediately with high ferns (Pteris 
 caudata), whose horizontally placed fronds wear an agree 
 able look. With the end of the wood one reaches pleasant 
 green fields, which, despite the dry climate, appeared to have 
 the fresh green of our European meadows." 
 
 Going westward the country becomes more and more open, 
 and there are extensive open flat tracts of great extent, 
 covered with a sparse catinga vegetation, immense can- 
 delebra-like cactuses, and ant-hills, and diversified by shal 
 low lagoas. From Tamburil to the boundary of Minas 
 Geracs one passes through a monotonous and somewhat 
 hilly country, cut through by deep ravines, and covered by 
 catinga. So soon as one has climbed the ridges which uni 
 formly command one another, and throughout are covered 
 in the same manner with catinga or carrasco, small narrow 
 fields, grown up with many rush-like grasses, are reached. 
 
 In some places near Ressaque the Prince found mica 
 slates with staurotides in single crystals, together with 
 hornblende. This succession of gneiss and mica slate cor 
 responds to what I observed on entering the basin of the 
 Jequitinhonha, but I have seen no staurotides in Brazil. 
 
 The carrasco is a more or less thick, usually matted, 
 
PROVINCE OF BAHIA. 253 
 
 growth of bushes, with stiff, gnarly stems, which grow to a 
 height of ten or twelve feet, and this is the character of the 
 vegetation of a large part of the campos of the interior, par 
 ticularly of the wide elevated plains. 
 
 The country grows flatter and flatter in going into the in 
 terior, and at the same time the bushes grow lower until 
 the campos geraes are reached, where, " far as the eye can 
 reach, open wooded plains, or gently rounded hills and 
 ridges, are spread out, covered with dried grass or scat 
 tered bushes." 
 
 " In the valleys which intersect these wide, naked ridges 
 and plains, one finds the banks of the rivers and brooks 
 bordered by woods. Here also single clamps of bushes are 
 found, hidden here and there in the deep places, particularly 
 as one approaches the borders of Minas Geraes, and this 
 kind of wood is in part one of the peculiar characteristics of 
 these open places. One often believes that he has a contin 
 uous plain before him, when he comes suddenly upon a nar 
 row, steep-sided valley, hears deep below him the murmur 
 of a stream, and looks down upon the forest-trees which, va 
 riously colored with numerous flowers, line its banks. Con 
 stant winds prevail here during the cold season, with, for 
 the most part, a cloudy sky, and in the dry months a burn 
 ing, oppressive heat, whereby the grass is dried up, the re 
 gion is glowing hot, and there is a scarcity of potable water." 
 
 In the winter these high lands are quite cool, and hail not 
 unfrequently falls. Prince Neu Wied describes a large tract 
 over which the vegetation was dead and leafless, and he 
 was informed that it had been killed by frost, though it may 
 be that the effect had been produced by excessive dryness. 
 To one accustomed to the climate of the coast, that of these 
 high regions is apt to be very disagreeable. The Prince 
 
254 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 found the temperature early on a foggy morning 14 
 Reaumur = 63.5 Fahr., and at noon on a dry sunshiny 
 day 19-| Reaumur, = 75.87, and this was just at the end 
 of the rainy season, in February. On the 22d January, at 
 Catole", he found in the shade between two and three 
 o clock, P. M., a temperature of 24|- Reaumur, = 87.12 
 Fahr., and he states that it sometimes reached 30 Reau 
 mur, = 95.50 Fahr., in the shade. 
 
 These campos stretch westward to the Serra das Almas, 
 and southward to the valley of the Jequitinhonha,* and are 
 only sparsely inhabited, principally by herders of cattle. 
 Agriculture is confined to the bottoms of valleys and moist 
 places. 
 
 The Etna, or American ostrich (Rhea Americana), occurs 
 in abundance on the campos of the basin of the Pardo, to 
 gether with the celebrated seriema (Palamedea cristata 
 Linn., Dicholophus cristatus Illiger), a large bird the size 
 of a crane, very swift of foot, and noted for its shrill voice 
 like that of the peacock, whence the English name crested 
 screamer, sometimes applied to it.f 
 
 With reference to the climate of this part of the coast, 
 Neu-Wied says that generally the months of February, 
 March, April, and May are the rainy months. The four 
 months following are the cold season, while the hottest 
 weather is in October, November, December, and January. 
 Our author says that he never observed a lower temperature 
 than 13 Reaumur, nor a much higher one than 30 Reumur. 
 
 * See description of the campos of the Comarca da Jequitinhonha in Chap. 
 III. of this work. 
 
 t See Eitg. Warming, Skildringer af Naturen i del tropislce BrasUien. V. Cam- 
 posdyrtne, Tidsskriji for Pop. Fremst. af Nat., Trcdie Rsekke, Femte Binds, 
 Tredie Hefte, 1868, p. 231. This is a very interesting article on the fauna of 
 the campos region. 
 
PROVINCE OF BAHIA. 255 
 
 On the 5th March, one of the hottest days of the voyage, 
 Neu-Wied observed a temperature of 28.50 Reaumur, and 
 in the twilight of the same day 15 Reaumur, and an hour 
 later, when the dew had begun to fall, 14 Reaumur, which 
 observations are interesting as showing the diurnal varieties 
 of temperature. 
 
 Arraial da Conquista, or Victoria, is the name of a little 
 village lying a few leagues to the north of the Barra do 
 Vareda, and noted for its cotton and large herds of cattle, 
 which are sent to Bahia to be sold. 
 
 On the voyage to Bahia, near a place called Urubu, lying 
 between the Rios Caxoeira and Contas, Neu-Wied describes 
 brooks whose waters were salt, discolored, and whitish, and 
 he speaks of others near the valley of the Contas. Thus 
 far a great part of the country between the Arraial da Con 
 quista and the Rio das Contas has been hilly and more or 
 less wooded. 
 
 Between the Arraial da Conquista and Os Possoes (Po^oes) 
 the country is very uneven and covered with low woods. 
 Neu-Wied calls attention to a very interesting fact stated by 
 Humboldt,* that the number of species of social plants in 
 
 * Alex, de Humboldt, De Distributione Geographica Plantarum, 1817, pp. 
 51, 52: 
 
 " Rarissimrc autem sunt plantae sociatae (Plantes societies, gcsellige pflanzen ) 
 in plaga oeqiiinoctiali. Difficiliter enim, ex genere arborum Silvis Orinocensibus 
 nomen ponas, quippe in quibus magnus specierum numerus a?que commixtus 
 sit. Neque in locis planis sub zona torrida Novi Orbis, plantas sociatas fere 
 alias ullas enumeres prater Rbizopboram Manglen, Sesuvium Portulacastrum, 
 Croton argenteum, Bambusam Guaduam, atque propter capita fluvii Amazo- 
 num et in calidis Provincial Jean de Bracamoras, amoenissima nemora Bouguin- 
 villea et Godoya repleta. Augentur vero stirpes catcrvatim nascentes quo 
 magis per Mexicanum imperium versus Cancrum procedis, vel per cacumina 
 Andium te tollis, ubi altitudine 1800 haxapodarum repcries Escalloniam myr 
 tilloidem, Brathim juniperinam et multijugas Molinae species." 
 
256 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the tropics is very small, and says that not unfrequently 
 we find large tracts taken possession of by the samambaia 
 {Mertensia dichotoma or Pteris caudata ?) to the exclusion 
 of all other shrubs. This is the case all along the coast. 
 It is apt to spring up in abandoned dry fields, and bury 
 them with a thick mantle of foliage. I observed it growing 
 at Itabapuana, Sao Matheos, and in numerous other places 
 on the coast, where it was so abundant as to be a nuisance, 
 and I have noted it as being very common in Minas. It 
 seems to have the same habit as our Pteris aquilina, which 
 in the same way takes possession of fields and drives 
 everything else out. The rapid growth of the samam 
 baia in Brazil is often aided by fires set in the bushes and 
 grass over dry places, which deplete the flora and give the 
 fern a more open field to grow in. The most remarkable 
 social plants of Eastern Brazil are the mangroves, Cono- 
 carpus and Avicennia, in addition to which and the Pteris 
 caudata are some species of Rhexia, Cecropia, Big-nonia, 
 together with the Uba,Taquarassu, some grasses, a bamboo, 
 and the dwarf palm of the coast, Guriri. The Piassaba and 
 Carnahuba palms would also appear to merit being included 
 under this head. 
 
 Between the Arraial da Conquista and Os Possoes, Neu 
 Wied describes a locality with high, gently rounded hills 
 covered with the samambaia, and he states that sometimes 
 such tracts are burned over so as to produce a growth of 
 grass. The whole country here is exceedingly dry, and 
 during the hot season the vegetation is withered and 
 scorched. Water then fails, and cattle die if not removed. 
 In this dry region is found a beautiful Bignonia, eight to ten 
 feet high, with large bright citron-yellow blossoms, and a 
 Cassia, together with the Licurl palm, a species which I 
 
PROVINCE OF BAHIA. 257 
 
 found growing over the taboleiras at Alagoinhas on the 
 Bahia and Sao Francisco Railroad. The soil is of a red 
 dish-yellow color. The deep valleys are filled with dense 
 forest. Everywhere one sees the round yellow hills of the 
 white ant scattered about. Carapatos* are exceedingly 
 numerous, incrusting twigs so as to make them fairly red, 
 and worrying the traveller by day and night. Between Os 
 Possoes and Urubu the country is of the same desolate 
 character. In this region many cattle are pastured, and in 
 some localities a little cotton is planted. Of the journey 
 from Urubu to the Fazenda da Cachoeira Neu-Wied draws 
 a striking picture of the country in these words : f "I fol 
 lowed the way through an inhospitable deserted wilderness, 
 in which, crowded together, mountain after mountain rose 
 behind one another ; all lay before us, monotonous, covered 
 with thick-woven brushwood, rough and wild, and mingled 
 with projecting rock masses. Some of these mountains are 
 naked and consist of variously formed masses of rock, as a 
 general thing gently rounded above ; in the places bare of 
 wood the soil shows itself as a red-yellow clay. Bushes of 
 finely-plumed thorny mimosas, mingled here and there with 
 beautiful flowering plants, amongst which I will mention 
 only one splendid plant, a new species of Ipomoea, with 
 large, brilliant fiery blossoms, made on both sides a border 
 to the way." 
 
 " The rock masses of the strangest forms, often like 
 towers or pulpits, standing singly out above the bushes, are 
 everywhere in these mountains inhabited by a little cavia," 
 called the Moc6, Ccelogenys rupestris. " In these dry rocky 
 
 * Bates says that there are two species on the Amazonas. See Naturalist 
 on the Amazons, p. 173. 
 t Reise* Vol. II. p. 236. 
 
 Q 
 
258 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 woods there reigned a heat beyond belief; not a breeze 
 stirred, and the rays of the sun were reflected from every 
 side ; only the proud Araras [macaws] in our neighborhood 
 appeared to enjoy it. They flew screaming about, while 
 the most of the other birds took their siesta on the shady 
 branches." 
 
 The mouth of the river at Ilhe os forms a very good har 
 bor, and is entered by the coasting steamers. Where the 
 valley of the river opens out into the sea are two isolated 
 gneiss hills, standing one about a quarter of a mile east of 
 the other. The western of these hills, once an island, has 
 been joined to the bluff on the north by a broad strip of sand, 
 on which is built the villa of Sao Jorge dos Ilhe os. This 
 causes the river to make an abrupt bend southward. The 
 eastern hill, also formerly an island, has been joined to the 
 low lands on the south by a strip of sand, which compels 
 the river to turn towards the north, when it enters the 
 sea between the two hills. The mouth of the river is shal 
 low, but is usually entered without much difficulty. The 
 waves beat very heavily on this coast, and it is interesting to 
 see how they are striving to throw up a spit across the river- 
 mouth on the northern side. At high tide the water flows 
 through a channel cut across the sand-beach, uniting the 
 eastern island with the shore ; but the waves tend to increase 
 the height of this beach, and the river is obliged to escape 
 around the western side of this hill. To the northeast are 
 the reefs Sororoca, Itapitinga, and Itaipins, which I have 
 not examined, and north of these are the islets Ilha Grande 
 and Ilha Pequena, distant about two miles north, a few de 
 grees east of the barra. 
 
 The town of Ilhe os is about as large as Caravellas, but 
 is of much more importance. The banks of the Rio Cax- 
 
PROVINCE OF BAHIA. 259 
 
 oeira and of its tributaries near Ilheos are thickly set 
 tled, and there are many large fazendas for the cultiva 
 tion of cacao, sugar, &c., which products are exported to 
 Bahia. 
 
 The rocks of the hill on the western side of the barra 
 consist of a well-bedded gneiss like that of Bahia, but much 
 disturbed and broken up. The approximate strike is north 
 a few degrees east. 
 
 A few miles to the north of Ilheos the little river Itahype 
 empties into the sea. The banks of the river are largely 
 settled. The river itself is very narrow, but deep, and alto 
 gether is only about twenty-eight miles long. On the north 
 ern side is a little lagoa communicating with the Itahype 
 by a narrow canal. Neu-Wied says that sea-shells are found 
 on the banks of this lake, and that kettle-formed holes, like 
 those hollowed by the action of the sea, are to be seen in the 
 rocks bordering it. 
 
 Spix and Martius found coral banks in the lake, show 
 ing that it was formerly a bay which has been cut off from 
 the sea by the throwing up of a beach across its mouth. 
 " These banks show themselves in many places in the lake 
 at a depth of from six to twelve feet," and furnish material 
 for the manufacture of lime. The, corals observed by Spix 
 and Martius were referred by them to the old species Mad- 
 repora cavernosa, hexagona, and astroites, Lam.* 
 
 On the sea-shore near the Serra Grande, south of the Rio 
 de Contas, Spix and Martius found " banks, five to six feet 
 high, of a soft, coal-black substance which soiled the finger 
 when pressed, and which, carefully examined, seemed to 
 be made up of coal and quartz grains." This appears to be 
 some recent formation. 
 
 * Von Martius, Reise, Band II. Seite 684. 
 
260 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 From Ilhe os northward to the Rio de Contas the coast 
 lands, as a general thing, are about two hundred feet in height, 
 and level topped, with abrupt, steep slopes to the sea, and 
 there can be no doubt that they are tertiary. The appear 
 ance of the coast just south of the mouth of the Contas is 
 represented in the following sketch. I am told that the 
 
 VIEW OF THE COAST SOUTH OF THE RIO DE CONTAS. 
 
 soils of these plains are sandy, but are in part quite fertile. 
 On the slopes cacao is planted. 
 
 The Rio de Contas, or Jussiape, is a considerable river, 
 which, according to the Diccionario Geographico^ rises eight 
 leagues to the northwest of the Villa de Rio de Contas, or at 
 a distance of some one hundred miles from the sea and to 
 the west of the Serra de Sincora.* The course of the main 
 stream is almost east-west. It is navigable only for a dis 
 tance of some four leagues above its mouth. 
 
 The lands on the southern side at the mouth of the river 
 are moderately high, and two or more rocky islands and a 
 reef of rocks project northward, so that the channel of the 
 river is bent toward the north, as is the case at Ilhe os. There 
 are other rocky points on the same side of the river, in which 
 a rock like that of Bahia is exposed. The town is situated 
 in a little cove, just inside the bar, on the southern side. I 
 found the rock in the hill and point west of the town to be 
 
 * According to Almeida, it takes its rise in a lake forty or fifty miles north 
 of the Villa do Rio de Contas. Burton says that this name should be Rio 
 das Contas. I follow Cazal, Almeida, and the common usage of Brazilian 
 writers. 
 
PROVINCE OF BAHIA. 261 
 
 of the same character as that of Bahia, and exposed with a 
 strike of N. 45 E., and a vertical dip. The town is a small 
 one, and of so little importance that the coasting steamers 
 rarely stop there. Its commerce consists principally in fa- 
 rinha, of which something like 50,000 sacks are exported 
 every year ; cacao, yearly exportation more than 4,000 arro- 
 bas; coffee, 800 to 1,000 arrobas; together with a little rice 
 and sugar, and large quantities of woods, especially jaca- 
 randd, cedro, putumujd, vinhatico. Along the river above 
 the town are many fazendas. The population is said to be 
 increasing. On the opposite side of the river the lands 
 bordering the coast are flat, sandy along shore, but inside 
 they are largely overgrown with mangroves. The beach 
 extends southward in a sharp point, which tends to close up 
 the river off the point just east of the town ; but the current 
 keeps it open. The bar is not difficult, and there is good 
 anchorage for vessels off the town. The Rio do Contas ap 
 peared to me to be about the size of the Mucury. The ter 
 tiary bluffs trend off northward of the river, gradually leav 
 ing the coast. A narrow channel coming from the north 
 and flowing through the low lands, parallel with the coast, 
 empties into the river opposite the town. I have seen in 
 some Brazilian work a Rolciro, if I rightly remember 
 a statement that there had been found on the banks of this 
 river the bones of some immense extinct quadruped. Spix 
 and Martius also refer to them. (See Chapter VII. of this 
 work.)* 
 
 * Cazal speaks of the existence, in different part of Brazil, of the bones of 
 an immense extinct animal, which he suggests might he the behemoth. I 
 translate a few sentences from his note on it. He says : " Morse gives to this 
 quadruped the name of Mammoth, and says that the Indians of North America 
 pretend that its species still exists in the woods which are to the north of the 
 great lakes This beast must have been of a slow inarch, not proper 
 
262 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 In 18G6 I touched at Camamu, but unfortunately during 
 bad weather, so that I was unable to see much. I have 
 never revisited the spot. I regret much that I am unable 
 to give a more precise and detailed description of the bay 
 on which it is situated, for it is of great interest, both to 
 the geologist and the zoologist, because on the borders of 
 the bay are the turba deposits which have attracted so much 
 attention, while off the mouth of the bay are extensive coral 
 reefs. This bay, and the streams which flow into it, have 
 been most erroneously represented on the maps and charts. 
 Arrowsmith represents it as a deep wide bay opening broad 
 ly to the sea, half as large as the Bahia de Todos os Santos, 
 and sown with little islands ; and Mouchez s chart gives one 
 really no idea of if. My friend, the Rev. Mr. Nicolay of 
 Bahia, kindly furnished me with a copy of a recent map by 
 Sr. Jose Nascimento, which is here subjoined, not only be 
 cause of its value as a contribution to the hydrography of 
 the coast, but because Mr. Nicolay has indicated on it the 
 localities where the turba occurs, as well as several other 
 points of interest. 
 
 The shores of the bay and its arms are for the most 
 
 for a hunting or carnivorous animal, and with a belly so capacious that only 
 vegetables could suffice to nourish it. 
 
 " Among the many skeletons which have been encountered in the different 
 provinces of the New World, perhaps none may help to form a better idea of the 
 animal than the skeleton which was discovered, at the close of the past century 
 in the Termo da Villa de Rio de Contns, by the persons engaged in clearing out 
 a hollow in the rock (caJdeirao de pedra), in order to make a tank for the cattle. 
 .... This skeleton, considerably injured, occupied a space of more than 
 thirty paces in length ; the ribs were a palm and half broad ; the shin-bones 
 were of the length of a man of medium stature ; the tusks were almost a brara 
 in length ; a molar tooth, without the root, weighed four pounds ; in order 
 to remove the lower jaw, the strength of four men was necessary." Coroifrnjia 
 Brazilica, p. 67, note. See D Archiac, Palebntolotjie Stratiymphique, p. 231. 
 
PROVINCE OF BAHIA. 263 
 
 part flat, and largely of recent origin, and there are exten 
 sive mangrove swamps bordering them. From over this 
 region the tertiary clays have been almost entirely denuded. 
 Gneiss occurs at the town, and just to the westward is the 
 considerable range of hills called the Serra do Condoru, 
 which is a continuation northward of the coast serras. It 
 is in the low grounds of the Marahu, a broad river-like arm 
 extending off for some distance to the south of the bay, that 
 the turba deposits are found. The specimens of turba which 
 I have seen were of a very light material, grayish or brownish 
 in color, and felty in texture. The material burned readily 
 when ignited in a candle, affording an abundant smoky flame, 
 leaving the mass, however, of the same dimensions and form 
 as before. The material appears to be merely a mud im 
 pregnated with bitumen ; and as it appears to exist in large 
 quantities, it would be very valuable for gas-making or the 
 manufacture of kerosene. Prof. Arthur M. Edwards of New 
 York, the microscopist, informs me that a Mr. Southworth 
 brought home with him some specimens of carbonized wood, 
 and Prof. Edwards * has some fragments of leaves from the 
 locality, but they are unfortunately too badly preserved for 
 identification. Mr. Nicolay personally examined the turba 
 deposits, and has kindly furnished me with the following 
 observations, which throw some little light on the mode of 
 occurrence of the bituminous layers, though their age is still 
 left in obscurity : 
 
 * Since the above was written, I learn that some of the turba distilled in 
 New York yielded of first quality, one hundred gallons light, clear oil to the ton ; 
 second quality, seventy-five gallons. Prof. Edwards says, in speaking of the 
 specimens he examined : " I consider the turba a sand impregnated with bitu 
 men, but it has evidently not been deposited under water, because it contains 
 no diatoms, and the vegetable remains in it are wood, a few leaves, and fibres 
 like fine roots." An English company has recently been formed to work this 
 deposit for the above-mentioned purposes. 
 
264 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 " The Camarnu series does not appear in any way connected 
 with that of Bahia or with those of the South, unless, as has been 
 reported, turba is found on the Rio de Contas. 
 
 " This series may be designated as follows : 
 
 " In a basin of gneissose rocks are bituminous schists, sands, and 
 marls containing fossils, (fresh water 1) and it is presumable above 
 that the turba. 
 
 " There are also sandstones, principally red concretions, not 
 apparently connected with the bituminous strata, but lying nearer 
 the sea, and possibly outliers of the Bahia series. 
 
 " The following section is from the pit sunk by Sr. Joao da 
 Costa, filho : 
 
 " 20 feet in. Argillaceous and arenaceous schists. 
 
 3 " " Bituminous clay. 
 
 4 " " Ferruginous and arenaceous sandstone. 
 " 9 " Schist with lignite. 
 
 " 9 " Bituminous stratum. 
 
 1 " 6 " Micaceous schists. 
 
 1 " 6 " Schist with lignite and bitumen below. 
 
 2 " 6 " Schistose strata. 
 
 2 " " Bituminous strata, some quite pure. 
 
 12 " " Schistose rocks. 
 
 15 " " Bituminous strata. 
 
 45 " " Gneissose rocks. 
 
 108 feet." 
 
 Mr. Nicolay states that the bituminous strata vary much, 
 from pure bitumen to an arenaceous kind interstratified 
 with bituminous shales ; " but it is to be observed that, at the 
 depth of eighteen feet, two veins of pedra molle, an imper 
 fect turba occur. These are in many cases present on the 
 surface, and the connection should be ascertained, but ex 
 cept in this instance pedra molle and turba only appear above 
 
PROVINCE OF BAHIA. 265 
 
 the limestone, i. e. so far as is known, which forms the 
 bottom of the estuaries which unite to form the Camamu 
 Bay." 
 
 The town of Camamu is only a small one, but is of some 
 little importance, exporting to Bahia coffee, farinha, rum, 
 rice, cacao, and woods for building purposes. Spix and 
 Martius speak of the occurrence of coral banks in the Bay of 
 Camamu, and they refer some of the corals to the old La- 
 marckian species, Madr. Uva, M. Astroides, and M. Acropora. 
 The Ilha de Quieppe lies just off the mouth of the bay, 
 and is surrounded by coral reefs. The dangerous reef of 
 Sorocosti (Sororocossu ?) lies a short distance to the north 
 west. About ten miles north of the mouth of the Bay of 
 Camamu lies the large island called Boyapeba. Mouchez, 
 in a note to his chart says that the pilots represent the 
 gulf lying between this island and the Bay of Camamu as 
 being full of corals, probably chapeiroes. The island of 
 Boyapeba is about five miles in diameter, moderately high, 
 with a few prominent hills, and separated from the main 
 land by a narrow channel. On the southeastern side of 
 this island is the dangerous point Castelhanos, on which 
 the French ship Beam was wrecked a few years ago. The 
 village of Boyapeba, near the northern extremity of the 
 island, is noted for its little commerce with Bahia in pias- 
 saba, rice, and mangrove bark for tanning. 
 
 Separated from this island on the south by narrow chan 
 nels, and from the mainland by a channel which unites 
 them with the Barra do Rio Una, is a much larger island, 
 called Tinhare , some ten or fifteen miles long and five or 
 six wide. Wedged in between these two islands and the 
 mainland is the little island Tupiassu. The northern portion 
 of the island of Tinhard appears to be high ; while the rest, 
 
 VOL. I. 12 
 
266 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 together with the island of Tupiassu and the mainland oppo 
 site, is low and in great part swampy. Mouchez represents 
 doubtfully the island of Tinhare as crossed by two or three 
 channels. The channel west of this island is very narrow 
 but deep. Into it flows from the west the Rio Jequie , a 
 small stream of little importance. The northeastern ex 
 tremity of the island of Tinhare is prolonged northward 
 into a sharp promontory called the Morro de Sao Paulo, 
 which, according to Dr. Anto. de Lacerda, is composed of 
 gneiss.* West of this is a deep bay some two miles wide, 
 and four or five deep, into which the Rio Tinhar empties 
 from the south, and the Rio Una from the west. This bay is 
 remarkable for a long sharp sand-spit, the continuation of a 
 sand-beach, which projects southeastward into the deep 
 curve west of the Morro de Sao Paulo. In this bay are 
 several small islands. Mouchez s chart appears to be very 
 inaccurate. Prince Neu-Wied, after leaving Ignez, reached 
 a river, which he calls the Jequeri^a, which he descended 
 for some distance, when he was arrested and carried across 
 the country to Aldea, a little place near Nazareth. In his 
 edition of Arrowsmith s map he has laid down his route, 
 which he represents as following the Una instead of the 
 Jequiriga, which flows into the sea some ten miles north 
 of the Una. One cannot doubt from his description that it 
 was the Jequiricd which he descended, and that, owing to 
 that river having been represented as emptying into the 
 bay of the Una, he laid down his itinerary incorrectly. The 
 town of Valenga is situated at the mouth of the Una, and 
 is noted for its large cotton-factory, and its commerce in 
 woods and coffee. From the Una to the Jequiri^d runs a 
 
 * Pissis also says that it is gneiss. Me m. de I Institut de France, Tome X. 
 p. 357. 
 
TODOS OS SANTOS, 
 
PROVINCE OF BAHIA. 267 
 
 long sand-beach, behind which the lands are at no great 
 distance moderately high and hilly. North of the Jequiric,d 
 the beach is interrupted by a few rocky points. 
 
 The Bahia de Todos os Santos, leaving out the islands, is 
 a quadrilateral figure, with unequal sides, as follows : * A line 
 from the Ponta de Sant. Antonio to the Ponta Garcia runs 
 approximately southwest, and measures a trifle over eighteen 
 marine miles ; a line running along the western side of the 
 bay would run nearly north-northeast, and measure thirty- 
 three miles ; the northeastern side of the figure representing 
 the width of the bay, which is remarkably uniform through 
 out, would run about southeast, length fifteen miles ; while 
 the other side of the figure, bringing us back to the point of 
 departure, would be approximately, south-southwest, sixteen 
 miles. A large island, of a sigmoid shape, called Itaparica, 
 and about four or five miles wide in the widest part, lies in 
 the bay with a general northeast trend, its axis lying several 
 miles to the west of that of the bay, and its southern end dis 
 tant about two miles northeast of the Ponta Garcia. About 
 three miles northeast of the Ilha Itaparica is a considerable 
 island called the Ilha dos Frades, which lies just off a point 
 projecting southward from the head of the bay. This point is 
 deeply indented by numerous little estuaries, and several 
 large islands lie off its west side. The Ilha Itaparica and 
 the Ilha dos Frades divide the bay into two strips, a long 
 and narrow one to the west, a shorter and wider one to the 
 east. "With this general description of the bay, let me now 
 enter into a little more detail. Between Ponta Garcia and 
 the Ponta da Caixa de Pregos of the Ilha Itaparica, the river 
 Jaguaripe finds an outlet to the sea, although its waters are 
 free to enter the channel west of Itaparica. 
 
 * These measurements I take from the best maps and charts in my posses 
 sion. 
 
268 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 This so-called river is an estuary, which at the mouth is 
 some two miles wide, but narrows rapidly down, being at 
 Nazareth, a town some ten to fifteen miles up the river, only 
 a few hundred feet in width. Steamers ascend to Nazareth, 
 but navigation ends a few miles above that town. The river 
 rises about fifty miles west of Cachoeira, and has a south 
 eastern course. All its branches below Nazareth appear to 
 have the estuary character, opening broadly into the river. I 
 have never ascended the river, and can therefore give no defi 
 nite information with respect to the geological character and 
 agricultural capabilities of the country bordering it. Naza 
 reth and Jaguaripe, a town about a league above the mouth, 
 are noted for their extensive manufactories of tiles and 
 earthenware, which are sent to Bahia. Opposite this last 
 town a canal runs northeastward, joining the Jaguaripe with 
 the estuary of the Jacoruna, separating a large tract from 
 the mainland called the Ilha de Santa Anna. 
 
 If the coast should rise fifteen or twenty feet, the island 
 of Itaparica would be joined to the mainland at its southern 
 extremity ; for at low tide the channel now separating it 
 from the mainland at the junction of the canal with the 
 estuary of the Jaguaripe is only some six feet deep at low 
 water, but it deepens gradually on going northward. The 
 west side of the channel is bordered by a line of islands called 
 the Ilhas de Parajuhia, which extend almost to the mouth of 
 the Paraguassu. It is interesting to compare the eastern 
 exposed side of Itaparica with its long stretches of sea-beaches 
 uniting rocky points with the sheltered western side, which 
 is exceedingly irregular, and cut up by bays and estuaries. 
 Of the geology of Itaparica I know nothing, except that I 
 am informed by Dr. de Lacerda and Mr. Nicolay that the 
 high hills are gneiss. In Dr. de Lacerda s cabinet at 
 
PROVINCE OF BAHIA. 269 
 
 Bahia is a pebble from Itaparica full of what appear 
 to be tertiary shells ; but it bears so strongly the appear 
 ance of a European formation, that I more than sus 
 pect that it has found its way across the Atlantic in 
 ballast.* Darwin speaks of deposits of tertiary shells, and 
 Elie de Beaumont says that M. Pissis found tertiary de 
 posits resembling the European mollasse in the bay of 
 Bahia. I have seen none, and it is singular that, if they 
 exist, Mr. Nicolay, who has explored the bay most indus 
 triously, should not have found them. In all probability 
 both Darwin and Pissis mistook (most pardonably) the 
 recent consolidated beaches for tertiary. f Coral banks exist 
 along the shores of the island, and the corals are collected 
 and burned for lime. The species are the common ones 
 of the Abrolhos region, though Heliastra3a appears to be 
 more abundant than at Bahia. 
 
 The Rio Paraguassu is the largest and most important 
 stream that waters the province of Bahia. It rises in the 
 Serra da Chapada, in the diamond district west of the Bahia 
 de Todos os Santos, and, obstructed by many falls, readies 
 at last, a few miles below the city of Cachoeira, the head of an 
 estuary which empties into the Bahia de Todos os Santos on 
 the western side and to the northwest of the northern extrem 
 ity of the Ilha de Itaparica. Through the kindness of my 
 friends Dr. Antonio de Lacerda and Mr. Hugh Wilson, I was 
 
 * One must be on his guard on the coast of Brazil against collecting speci 
 mens from rocks brought as ballast not only from other parts of the Brazilian 
 coast, but also from other countries. At Maceio I saw palaeozoic rocks which I 
 was told came from North America. So the Irish carboniferous limestone 
 brought to St. John, New Brunswick, finds its way all along the coast of that 
 province. 
 
 t I think it is Darwin who speaks of tertiary fossils as occurring at the head 
 of the bay. I know of no reason why fossils should not exist as the tertiary 
 rock of that locality. 
 
270 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 enabled to ascend the river, or rather estuary, as far as the 
 city of Cachoeira, on the occasion of the fete of the open 
 ing of the new steam tramroad just commenced by an Eng 
 lish company, and intended to unite the city of Cachoeira 
 with the Villa do Urubu on the Sao Francisco, and the fol 
 lowing observations were made at that time. The entrance 
 of the Paraguassu is quite narrow, and bordered by high 
 lands. The water is deep. Some of the hills at the entrance 
 may be five hundred feet high. Bluffs on the southern side 
 showed that they were composed of horizontal beds of a 
 soft, light yellowish brown sandstone. I believe this to be 
 an extension of the tertiary formation of the coast. 
 
 On entering the narrow mouth the river immediately 
 widens into a lake-like expansion, in which is a long, narrow 
 island. High bluffs on the shore of the island and on the 
 left bank of the river opposite show red sandstone with 
 white streaks, and a beautiful oblique lamination. About a 
 league from the mouth, just above the Barra do Rio Camu- 
 rugipe, the river narrows abruptly, and is here more than one 
 hundred feet deep. The land on each side of the river is 
 three to four hundred feet high, flat-topped, with steep 
 slopes toward the river, and occasional bluffs which show 
 red sandstone in horizontal strata, as below. The soil is 
 red, but thin, and the vegetation is very scanty, consisting 
 of piassabas and low bushes, with a few cocoa-palms ( Cocos 
 nucifera) and Dendes* (El<zis Guineensis L.). At one 
 place near the battery, -called the Fortinho, the sandstone 
 beds appeared to have a slight dip northward, but this is 
 evidently local, as the whole formation is quite undisturbed 
 and horizontal. 
 
 * This palm, an African species, goes by the name of Caiauhe on the Ama- 
 zonas. The fruit gives an orange-colored oil, much used in Brazil for culinary 
 purposes. It is also employed in the manufacture of fine soaps. 
 
PROVINCE OF BAHIA. 271 
 
 Above the fort the river opens out into another broad ex 
 panse, and a beautiful view is presented. The whole country 
 is an elevated plain, with steep slopes to the estuary and its 
 branches. These slopes have the same character as below, 
 and are covered largely by piassaba-palms. An occasional 
 sugar-factory may be seen on the river. From this lake-like 
 expansion of the river a broad arm extends off to the 
 southwestward for some six miles in a broad valley, into 
 the head of which empties the Rio Capanema, and on which 
 is situated the village of the same name. This arm is called 
 the Rio de Capanema, and it is said to receive the waters 
 of a lake. Maragogipe is a prettily situated and flourish 
 ing village, built at the base of the chapada, at the junction 
 of the Capanema with the Paraguassu.* The slopes of the 
 chapadas are here more or less cultivated with corn, <fec., 
 and there are large groves of cocoa-palms on the island in 
 front of the town. 
 
 From the northern part of the expanse of the Paraguassu 
 extends northwards another arm called the Iguape,f into 
 which a little river empties. The valley of the Iguape is 
 broad and exceedingly fertile, and there are very many ex 
 tensive sugar-plantations situated in it.J The tertiary cha 
 padas grow higher as we ascend, and on the Iguape they 
 must be at least seven hundred feet in elevation. Passing the 
 Iguape the river narrows very suddenly, and thence to the 
 city of Cachoeira it is only a few hundred feet wide. More 
 
 * Cazal, Coroyrajia, Tom. II. p. 125, says that in the vicinity is found 
 antimonio and bolo-armenlo. 
 
 f This name is rather uncommon, being found only a few times on the Brazil 
 ian coast. It appears to be another form of Yijarnpe, canoe-path or channel, a 
 name applied to the side channels of the Amazonas. (See Index.) 
 
 \ Cazal says that the soils of the valley of Iguape are the best known for the 
 culture of cane, bein<r of black massap^ and strong. Cor. Draz., Tom. II. p. 125. 
 
272 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 properly speaking, the mouth of the Paraguassu is here, and 
 the irregular sheet of water below, to the bay, is a tidal 
 estuary. At the point where the river flows into the estuary, 
 rock shows itself in the river-banks, and, according to Senhor 
 Przewodowski, who accompanied me on the voyage, is the 
 corafdo de negro rock, like that of Bahia. The strike is 
 N. 40 E. Dip vertical. The country bordering the river 
 consists of rounded hills, back of which are the elevated 
 chapadas. Bricks, tiles, and pottery are largely manufac 
 tured on the river, from the clays of the alluvial river-banks. 
 The water of the river is very turbid and brownish. 
 
 The stream narrows as we ascend, and is so shallow that 
 two of the little steamers with a part of the excursionists 
 from Bahia got aground. There are a few sugar fazendas 
 on the river below Cachoeira. Above Cachoeira the river 
 is obstructed by falls and rocks. The town of Cachoeira, 
 a respectably sized village, is noted more particularly for its 
 export of sugar and cigars. 
 
 The country bordering the Bahia de Todos os Santos is 
 in general terms very productive, and it has long borne the 
 name of the " Reconcavo" Sugar is the principal product, 
 though mandioca, &c., are cultivated. The tertiary de 
 posits extend across from the Paraguassu to Sant. Amaro, a 
 flourishing city on the river Serigi, some three leagues above 
 the mouth, according to my friend Dr. Brunet, the Director 
 of the Agricultural Institute of Bahia. Dr. Brunet informs 
 me that the lands of the vicinity of Sant. Amaro are very 
 fertile. Sugar and farinha are the principal products. Here 
 is situated the Agricultural Institute above spoken of. East 
 ward the tertiary beds extend to the Bahia and Sao Fran 
 cisco Railroad. 
 
 It is interesting to compare the Bahia de Todos os Santos 
 
PROVINCE OF BAHIA. 273 
 
 with the Bahia do Rio de Janeiro. The latter is also divided 
 almost into two parts by the Ilha do Governador and the 
 islands to the northeast ; but the most interesting point is 
 to observe the difference in the character of the rivers emp 
 tying into the two bays. Those of Rio are all small, and at 
 their mouths they are bordered by flat grounds and mangrove 
 swamps, owing to the rapid building out of the land ; for 
 the lands about the bay are mountainous, and the streams, 
 swollen by heavy rains, bring down an immense quantity of 
 sediment into the bay. The lands bordering the Bahia de 
 Todos os Santos are high, and stretches of sands and man 
 grove swamps are rare ; while the rivers, bringing down less 
 sediment, are not so contracted at their mouths, but open 
 out broadly into the bay. 
 
 12* 
 
274 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER VI. 
 
 THE SAO FRANCISCO BASIN. 
 
 The Explorations of Halfeld, Liais, St. John, Allen, Ward, Burton, &c. Gen 
 eral Shape of the Basin. Its uniform Width. The Sao Francisco Valley 
 hollowed out of a Series of horizontal Beds of Limestone and Sandstone. 
 The Chapadas. The so-called " Serra" separating the Sao Francisco from 
 the Tocantins Basin an irregular Strip or Table-land of Sandstone. The 
 Serras of Araripe and Dous Irmaos. Table-topped Hills in the Valley of 
 the Sao Francisco ; Outliers of the Chapadas. Doubts about the Age of the 
 Sandstones and Limestones. Limestones of the Rio das Velhas. Remains 
 of Extinct Quadrupeds in Brazil, spoken of by Cazal, Spix and Martius, &c. 
 Claussen s Discoveries in the Caves at Curvelo. Dr. Lund s exhaustive 
 Researches at Lag6a Santa. Caves described ; their Number, Extent, Sta 
 lactites and Deposits of Bones in Saltpetre Earth. Immense Quantities of 
 small Bones brought in by Owls, &c. Large Number of Fossil Animals dis 
 covered by Lund. Former Existence of Megatheria, Mylodons, Mastodons, 
 immense Armadillos and Cats, Horses, &c., in Brazil. Remains of a Race 
 of Man of high Antiquity. Rcinhardt s Generalizations. The Rio de Sao 
 Francisco of the Sixteenth Class among the Rivers of the World, but third 
 in Rank in Brazil. General Description of the Stream. Its Affluents, 
 the Rios Para, Paraopeba, and Das Velhas. The Rio das Velhas alone 
 capable of being made navigable for Steamers. The Sao Francisco naviga 
 ble with but few Interruptions for two hundred and sixty-four Leagues below 
 the Rio das Velhas. Cost of removing Obstructions. Proposed Railway 
 from Joazeiro to Piranhas. Fertility of Low Lands of Sao Francisco Val 
 ley. Liais s Picture of the Campos. 
 
 BEFORE we take up the geology of the interior and western 
 portion of the province of Bahia, which latter embraces a 
 considerable part of the basin of the Sao Francisco, I 
 propose to give, in a few words, a sketch of the geology 
 and hydrography of the whole basin ; and this is the more 
 
SAO FRANCISCO BASIN. 275 
 
 needed, since further on we shall have to visit the lower 
 part of the great river, and it will be necessary to consider 
 some questions bearing on its navigation, &c. 
 
 No river in Brazil has been so carefully studied and 
 mapped as the Sao Francisco and its tributary the Rio das 
 Velhas. Halfcld explored the main river from the conflu 
 ence of the two streams to the sea, and published a few years 
 ago a magnificent chart of the river in atlas form, accom 
 panied by a detailed description of every league. He also 
 made an estimate of the cost of removing the obstacles to 
 navigation, so as to render it a great interior water high 
 way. But, as Burton and Liais have remarked, the chart 
 of the river is rather a detailed plan than a scientifically 
 accurate map, as it wants the meridians and parallels. 
 
 M. Emmanuel Liais, the author of L Espace Celeste, 
 made a most elaborate survey of the Rio das Velhas and 
 Upper Sao Francisco. In his report* on this survey the 
 obstructions, such as rocks, rapids, &c., are shown in dia 
 gram and described, and estimates are given of the proba 
 ble expense of removing them. Almost all the well-known 
 explorers of Brazil have visited some part of the Sao Fran 
 cisco basin. Among the later are Messrs. St. John and 
 Allen of the Thayer Expedition, the former a geologist, the 
 latter an ornithologist, but a good geological observer. 
 Mr. St. John made an exceedingly careful examination of 
 the valley of the Rio das Velhas, and of the Sao Francisco 
 as far as the Barra do Rio Grande, whence he crossed over 
 into the basin of the Parnahyba, and continued his explora- 
 
 * Ilydrographic du Haul San-Francisco et du Rio das Velhas, ou resultats 
 au point de vue hydrographique d un voyage effectue dans la province de Minas 
 Geraes, par Emm. Liais. Ouvrage public par ordre du gouvcrncmcnt impe 
 rial du Bresil, et accompagne dc cartes levees par 1 autcur avec la collaboration 
 de MM. Edw. Jose de Moraes ct Ladislas de Souza Mello Netto. 1865. 
 
276 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tions across the provinces of Piauhy and Maranhao to the 
 city of Sao Luiz. Mr. St. John had the good fortune to ex 
 amine the geology of a route along which the geological 
 features of the country, though somewhat monotonous, are 
 more exposed to view than along almost any other that 
 could have been assigned him. He did his work thorough 
 ly, and his report will be of great interest when published, 
 and we hope that it will not be long delayed. Mr. Allen, 
 suffering from ill health, parted with his companion and 
 went on to Chique-Chique, where he took mules and 
 crossed the province of Bahia to Cachoeira. Mr. Allen 
 has made a valuable report on that part of the country 
 which he visited alone, and he has been kind enough to 
 give me some notes on the geology and physical geography 
 of the province of Bahia, which I shall insert in a subse 
 quent chapter. 
 
 Finally we have Burton, who in 1867 explored the Rio das 
 Velhas and Sao Francisco to the sea, and he has given us 
 an account of his journey in the second volume of his 
 " Highlands of Brazil." This book is like a series of sharply 
 drawn photographs of nature and life along his route. It 
 is exceedingly full of facts of every kind ; but his peculiar 
 style, and wholly unique geological language, render his 
 geological observations in many cases quite valueless. 
 
 The basin of the Sao Francisco is a peculiar one. It is 
 long, of very equal width, the lower half making a strong 
 curve, the convexity of which faces the northwest.* 
 
 The head of the basin, bounded by high serras of eozoic 
 and older Palaeozoic metamorphic rock, narrows rapidly to- 
 
 * Liais calls attention to the very uniform width of the Sao Francisco val 
 ley, and says that it varies from fifty to eighty leagues. See his paper, Le Rio 
 San-Francisco au Bre ti/, Bull, de la Soc. de Geographic, Fifth Se rie, 11. p. 389. 
 
THE SAO FRANCISCO BASIN. 277 
 
 ward the south, and at the apex, near Ouro Preto, are some 
 of the highest elevations in Brazil. It is separated from the 
 coast river basins by the ridges of the great metamorphic 
 coast belt and the line of chapadas running along a part 
 of its summit, the whole collectively called the Serra do 
 Espinhac,o by Baron von Eschwege. This metamorphic belt 
 flattens down between the two provinces of Bahia and Per 
 il ambuco, and the basin bends round to the east and crosses 
 it, opening out to the sea. 
 
 The whole length of the basin is not far from twelve hun 
 dred miles, and its greatest width not more than two hun 
 dred and forty. It is separated from the Parana basin by 
 metamorphic serras, which become lower as one goes north 
 ward, finally dipping under a sheet of sandstone corre 
 sponding to that of the chapadas along the Serra do Es- 
 pinhacjo, and which run northward to the province of 
 Piauhy, forming a table-land, flat topped, without mountains, 
 and of a varying width, which sends out broad spurs of 
 chapadas between the affluents of both the Sao Francisco 
 and the Tocantins basins. 
 
 Mr. Ward tells me that the valley of the Sao Francisco, 
 along the western flank of the Grao Mogor, is skirted by 
 high chapadas composed of sandstones and limestones, hav 
 ing precisely the same character as those of the opposite 
 side of the valley ; the chapadas sending out spurs into 
 the valley. These chapadas are, when dry, covered by 
 the ordinary campos vegetation, but magnificent groves 
 of Buriti palms are found in the damp shallow valleys. 
 Mr. Ward estimates the height of these elevated plains at 
 2,500-3,000 feet. They break down abruptly on reach 
 ing the Sao Francisco valley. Mr. Ward describes the 
 flat-topped hills of the valley as outliers of the sandstones 
 
278 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and limestones, and he, I think, rightly ascribes their lower 
 height to their having suffered denudation. 
 
 On nearing the limits of Piauhy the basin reaches the 
 metamorphic belt, and, being prevented from extending 
 northward by the sandstone-crowned metamorphic ridge of 
 the Dous Irmaos, bends round as above described and flows 
 to the east. The chapadas I have mentioned continue on, 
 with more or less breaks, along the northern side of the 
 basin forming the serra of Araripe. The upper part of the 
 basin is occupied by horizontal limestone deposits, while 
 farther down are sandstones, shales, <fec. The lower half 
 of the valley, almost to the sea, is cut down to the bottom 
 metamorphic rocks. The basin was primarily determined 
 by the ancient denudation of the metamorphic rocks ; but as 
 it stands to-day it is worn in the great sandstone sheet, 
 which I shall in the course of the succeeding chapters 
 attempt to show covered the whole plateau of Brazil. It 
 is the fashion among map-drawers to throw in a range of 
 mountains separating two great river basins, especially if 
 there is known to be any high ground between them, and 
 this mountain range is carefully drawn along the sum 
 mit line of the water-shed. Brazil is usually represented 
 as traversed in all directions by mountain chains, drawn 
 as if they were all alike narrow ridges. But rivers may 
 take their rise on elevated plains, and the water-shed may 
 be only an insignificant bulging. It is so with the great 
 divide between the La Plata and Amazonian systems. In 
 the case of the Tocantins and Sao Francisco the streams 
 traverse an elevated plateau of sandstone, which forms on 
 top a plain. The branches taking their rise on the divide, 
 and flowing in opposite directions, have cut for themselves 
 valleys that widen out towards their respective main rivers, 
 
THE SAO FRANCISCO BASIN. 279 
 
 so that the high lands separating the two basins, instead of 
 being a narrow mountain chain, consist of an elevated plain 
 sending out jagged spurs between the valleys of the tribu 
 tary streams. Outliers of these chapadas form isolated table- 
 topped hills and ridges bordering the main stream of the 
 Sao Francisco as far down as the Great Falls.* These 
 high lands, usually drawn on maps as narrow ridges, appear 
 like ranges of upheaval, and the maps are only calculated 
 to mislead the geologist and physical geographer. 
 
 The stratigraphical relationship of the limestones has yet 
 to be worked out, but they appear to underlie the sandstone 
 of the Tocantins-Sao Francisco divide, for both Gardner 
 and Ward speak of them as appearing in the lower part of 
 the slopes in descending from the chapadas into the Sao 
 Francisco valley, f I shall leave the detailed description 
 of these deposits to Mr. St. John, to whose report it prop 
 erly belongs, and confine myself only to a few points well 
 determined by other geologists, and which I need to state 
 here in order to complete that general sketch of Brazilian 
 geology and physical geography which I am attempting to 
 give. 
 
 According to Reinhardt, the limestone of the Rio das 
 Yelhas is dark gray in color, fine-grained and crystalline. 
 It splits into thin slabs, and is so sonorous that plates of it 
 were used formerly as bells to the churches. Lund, Bur- 
 meistcr, Reinhardt, St. John, and in fact all the geologists 
 who have examined the limestone, testify that it is with- 
 
 * I think that I may safely say that all the great north-south ranges of high 
 land in Brazil, north of the latitude of Diamantina, except the Grao Mogor, 
 usually described and mapped as mountain chains, are ranges of chapadas or 
 narrow plateaus resulting from denudation. 
 
 t Mr. St. John tells me that they underlie the sandstones of the Sao Fran 
 cisco valley. 
 
280 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 out fossils, and it has been supposed, if I rightly under 
 stand Lund, to be very old, and probably Palaeozoic. Rein- 
 hardt would make it Devonian. I see no good reason for 
 coming to such a conclusion. The want of fossils is no cri 
 terion of the age of a formation ; it is no proof that a rock 
 is old simply because it contains no fossils ; nor would the 
 apparently metamorphosed and somewhat crystalline con 
 dition of the rock necessarily show that it was very ancient. 
 The metamorphism of rocks is largely due to the action 
 of water ; and in the decomposition of the rocks of Brazil 
 we see what a powerful agent meteoric water is in working 
 changes in the rocks. We know that the limestones of the 
 Rio das Velhas are plentifully soaked with water by the heavy 
 rains, and the metamorphism of the limestone is doubtless 
 due to this cause. Compact limestones without fossil re 
 mains are not by any means uncommon all over the globe, 
 and they may be of any age. These limestones were depos 
 ited at the head of the Sao Francisco basin, in a bay shel 
 tered on the east, south, and west by high lands. 
 
 In these limestones are the celebrated bone caverns, of 
 which I will give some account before I describe the great 
 river and its navigation. 
 
 It appears to have been Cazal who, in 1817, first called 
 attention to the existence of bones and skeletons of giant 
 extinct quadrupeds in Brazil ; * but similar remains had 
 been found even as far back as 1G02, or earlier, by the first 
 explorers in Bolivia and on the Pampas. Spix and Mar- 
 tius have described many localities in the provinces of 
 Minas Gcraes and Bahia where mammalian remains were 
 found ; and Eschwegc, St. Hilaire, and other travellers 
 spoke of the occurrence of those remains, not only in the 
 
 * Sec p. 261. 
 
THE SAO FRANCISCO BASIN. 281 
 
 deposits in shallow hollows in the rock, but also in the 
 saltpetre caverns of Minas Geraes ; but no one of these 
 travellers made a systematic examination of any of the 
 localities, and we are indebted almost entirely to the dis 
 tinguished Dane, Dr. P. W. Lund, for what we know of 
 the fossil fauna of the bone caverns of Brazil. Lund 
 had been travelling many years in Brazil in company 
 with Riedel, the botanist, and was on his return with him 
 to Rio when he accidentally heard of Claussen, another 
 Dane, who was residing near Curvelo, in the valley of the 
 Rio das Velhas. Claussen had been examining the salt 
 petre caves of the vicinity, and collecting bones from them, 
 and Dr. Reinhardt* tells us that when Lund visited him 
 he was trying to study them out with the aid of Buck- 
 land s Reliqufs Diluviance. This was in 1834. Lund saw 
 that here was an immense field to explore ; and as soon 
 as he could honorably withdraw from Riedel, he returned to 
 Cachoeira do Campo to examine some caverns there ; but 
 he soon rejoined Claussen, and worked with him for some 
 time. The two, however, seem not to have agreed well to 
 gether, and in 1835 Lund withdrew to Lagoa Santa, and he 
 has remained there ever since devoting his time to an ex 
 haustive examination of the bone caverns of the vicinity, 
 reaping, as we shall see, a rich harvest for science. Bur 
 ton tells us that the distinguished geologist is confined 
 to Brazil by consumptive tendencies, and is bedridden by 
 rheumatism. 
 
 The region in which Lagda Santa is situated is composed 
 
 * For most of the farts in this account of Lund s researches in the bone 
 caves of Brazil I am indebted to the very interesting paper of Professor J. 
 Reinhardt, in Lutken s popular Tidsskrift, entitled De Brasilianske Knoglehuler 
 og de i dem forekommende Dyrelevninger. 
 
282 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of beds of limestone and shales, the limestone lying be 
 low and the shales above, the whole being covered by a 
 bed of red earth, which Reinhardt describes as resulting 
 simply from the decomposition of the shales, but which 
 Lund, if I rightly understand him, believes to be the 
 same red clay which covers the whole country, and which 
 Professor Agassiz and I would refer to the drift. These 
 beds are, as already remarked, horizontal, and are traversed 
 by narrow, often ramifying channels, caused by the widening 
 of the joints of the limestone by the penetration of surface 
 waters which sometimes form therein subterranean streams ; 
 for in the limestone region the streams sometimes disappear 
 and pursue an underground course, often for a long distance. 
 Professor Reinhardt gives a ground-plan of the principal 
 ramifications of one of the most noted caverns, called the 
 Lapa Vermelha, situated about a mile from LagOa Santa, 
 and this plan I have reproduced below.* According to 
 
 GROUND VIEW OF THE LAPA VKKMKLHA. 
 
 * It was drawn by Lund s former assistant, the late Peter Andreas Brandt, 
 and gives only the larger galleries of the cavern. Besides these, there are in 
 numerable smaller ones, some of which are only mere cracks where the joints 
 have been widened by the water. 
 
THE SAO FRANCISCO BASIN. 283 
 
 Reinhardt, this cavern extends over two thousand feet into 
 the rock, growing narrower and narrower, until it becomes^ 
 only a mere canal. The floors are usually horizontal. Some 
 times the caverns are only narrow cracks, at others they are 
 wide arched galleries, which not infrequently open into large 
 halls. The walls and roof are smooth and without sharp 
 corners. Reinhardt is inclined to believe that the excavation 
 of the caverns has been partially due to the surface waters, 
 which have in soaking through the rock dissolved away the 
 surface of the walls. That this has been the case to a con 
 siderable extent is proved, as Reinhardt has remarked, by 
 the projection from the smooth limestone wall of very thin 
 sheets of clay, which would certainly have been worn away 
 if the whole hollowing out had been performed by running 
 water. One thing seems quite certain, that these caverns 
 were excavated before the valleys of the region in which they 
 occur. The roof and sides of the caves are often covered 
 by very large and beautiful stalactites of a great variety 
 of forms. A stalagmitic crust sometimes covers the earthen 
 floor, and in some caves there are large pillars. These 
 stalactites, formed by the exceedingly slow deposition of cal 
 careous matter by the water trickling through the rock, 
 since the time when the clay of the floor was deposited, 
 bear some testimony as to the great age of the bones therein 
 buried. 
 
 The earth covering the floor is a yellowish-red clay, 
 which is, according to Lund, like the superficial soil of 
 the country. All authors describe it in very much the 
 same way. It is however very clear, from its mode of 
 occurrence, that it is not drift, and that it is a deposit in 
 troduced into the caves by the action of the surface waters ; 
 but precisely how may be a question. The earth often con- 
 
284 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tains fragments of quartz and other rocks. Reinhardt is 
 of the opinion that it has been introduced from above by 
 the water flowing into the caves through the overlying red 
 earth and decomposed shales. It appears, in some cases, 
 to have once filled the caves from floor to roof, and to have 
 been subsequently more or less completely washed Out. 
 This red earth is strongly impregnated with saltpetre, and 
 its extraction is so profitable, that the Brazilians have 
 removed it entirely from many of the caves.* Reinhardt 
 says that a small cartful sometimes produces as much as 
 two stones, or an arroba, of the salt. Bones of extinct ani 
 mals occur buried in this clay in almost all the caverns, but 
 in such small quantities, in the majority of them, that they 
 do not reward the pains of the collector. Lund told Bur- 
 meister that he had examined at least one thousand caves ; out 
 of these only sixty contained bones in any quantity, and but 
 half that number really paid for working. The number of 
 caves is astonishing, and Burmeister tells us that almost 
 every bank has its cavern. They are not confined to the 
 immediate vicinity of Lagoa Santa, but are found in great 
 numbers throughout the limestone region. The skeletons 
 found in them are usually disarticulated. The bones are 
 often much broken and almost invariably scattered about, 
 so that the discovery of an entire skeleton is hardly to be 
 thought of. Besides this, the earth in which the bones 
 occur is much cemented together, and has to be broken up 
 to allow of their extraction. The bones are not all of the 
 same age, and a large proportion of those in some caves 
 belong to now existing animals. But in other caverns 
 there are found remains of extinct animals of high an 
 tiquity. 
 
 * See note on saltpetre, near end of Chapter VII. 
 
THE SAO FRANCISCO BASIN. 285 
 
 Some of the caves contain immense quantities of small 
 bones belonging to bats and small animals of existing spe 
 cies. Near Caxoeira do Campo is a cave about 120 feet 
 long, 30 to 40 feet high, and 6 to 9 feet broad. Over a part 
 of the bottom lay quite a thick bed of earth filled with small 
 bones. Lund carried out half a cubic foot of this earth, 
 and counted all the half-underjaws he found in it. Of small 
 opossums (pungrotter) there were 400, and about 2,000 of 
 different kinds of mice, besides bats, porcupines (pigrotter ) , 
 and small birds. Another interesting instance is related by 
 Reinhardt. Lund had the whole of the clay brought from 
 a cave at the fazenda of Escravania, which was only 24 feet 
 deep. This earth filled 6,552 firkins. Lund determined 
 the number of half-underjaws found in a certain measure, 
 and calculated that in the whole mass there were the 
 remains of not less than 6,881,500 individuals of cavias, 
 opossums, porcupines, and mice ! Beside these there were 
 immense quantities of bones of small birds, lizards, frogs, 
 &c. And all these bones had been brought into the cave 
 by owls ! Now owls are unsocial birds, and we cannot 
 resist the conclusion of Reinhardt, that the deposit must 
 have been gathering for many thousands of years. While 
 these bones belong to the present geological epoch, those 
 buried in the red clay below the stalactite accumulations 
 belong to a more ancient time, and are for the greater part 
 of extinct forms ; and it is from this source that the bones 
 of the Megatherium and other giant animals are derived. 
 
 Of these animals there have been discovered by Messrs. 
 Lund and Claussen 115 species of mammals, belonging to 
 58 different genera, distributed as follows : * 
 
 * See U Archiac, Geologic et Pal^ontologle, p. 722, from which the table on 
 the next page is taken. 
 
286 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 
 Genera 
 
 Species. 
 
 Quadrumana . . 
 
 4 
 
 6 
 
 Cheiroptera .... 
 
 . 3 
 
 7 
 
 
 9 
 
 18 
 
 Rodentia .... 
 
 . 15 
 
 32 
 
 Edentata .... 
 
 13 
 
 28 
 
 Pachydermata .... 
 
 . 9 
 
 10 
 
 Ruminantia .... 
 
 4 
 
 7 
 
 Marsupialia .... 
 
 . 1 
 
 7 
 
 58 115 
 
 With bones of extinct animals occur those of now living 
 species, as, for instance, Cervus rufus, Cervus simplicicor- 
 m s, Sciurus (estuans, Echimys Cayennensis, Myrmecophaga 
 tetradactyla, Lepus Brasiliensis, Fells concolor, and Felis 
 mitis.* Among the extinct quadrupeds may be mentioned 
 the Mastodon, whose remains have been only rarely found 
 in caverns, but more often in pits and holes. There were 
 bones of species of Macrauchenia, Toxodon, Chlamydothe- 
 rium, and of the gigantic Glyptodon (Hoplophorus), Mylo- 
 don, and Megatherium. Among the carnivora were wildcats 
 and jaguars, and a species of Smilodon ($. neog-ceus ), 
 an immense cat-like animal with enormous knife-like canine 
 teeth in the upper jaw, allied to the fossil European species. 
 Of monkeys Lund found but few, and they belonged to the 
 genera Callithrix, Hapale, Ateles, and Protopithecus, the 
 latter being an extinct genus. In six or seven of the holes 
 Lund found stone implements and remains of man so buried 
 with the remains of the extinct fauna, as to leave no doubt 
 that man was contemporaneous with it in Brazil as in Eu 
 rope. In the Sumidouro cavern they were found mingled 
 
 * Reinhardt, op. cit., p. 315. 
 
THE SAO FRANCISCO BASIN. 287 
 
 with bones of the extinct cavern jaguar (Felis protopanther) 
 an immense Capibara (Hydroch&rus sulcidens), together 
 with remains of llamas and horses, which last certainly 
 existed in Brazil, as in North America, long before the 
 conquest. According to Reinhardt, the race of men whose 
 remains Lund has found appear to have been well built, but 
 slender. The same writer states that a skull he examined 
 was dolichocephalic and somewhat prognathous. It was of 
 medium size and ridged with a very prominent cheekbone, 
 a small forehead, and eyes wide apart. The walls of the 
 skull were extraordinarily thick. 
 
 Reinhardt* has come to some interesting conclusions with 
 reference to the history of the cave fauna, and I translate 
 them in full. They are : 
 
 " 1. That Brazil, in the post-pliocene time, was inhabited 
 by a very rich mammalian fauna, of which the present may 
 be said to be a fraction or stunted remainder, since many 
 genera, nay, even large systematic groups, such as families 
 and suborders, have disappeared, and only very few have 
 come down to our day. 
 
 " 2. That the Brazilian mammalian fauna, in the whole 
 post-pliocene time, had the same peculiar stamp which in 
 the present distinguishes the South American fauna in 
 comparison with that of the Old World, while the extinct 
 genera belong to families and groups which still to-day par 
 ticularly characterize South America. Only two of these 
 genera, one extinct, the Mastodon, the other still existing, 
 the horse, belong to families which are entirely confined to 
 the Eastern hemisphere, and form exceptions to the rule. 
 
 " 3. That the mammalian orders were far from being 
 richer in genera formerly than now. The Ruminants, Pa- 
 
 * Liitken s Tidsskrift, 3 d " R., 4 de Bind, 4 de Hcfte, p. 351. 
 
288 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 chyderms, Elephants, and the Carnivora have suffered the 
 greatest loss. Some orders, as the Cheiroptera and Mon 
 keys, number perhaps to-day more genera than formerly. 
 
 " 4. That in South America the post-pliocene mammalian 
 fauna was more distinct from the present fauna, and was 
 more especially rich in peculiar and now extinct genera than 
 was the case with the corresponding fauna in the Old World. 
 
 " 5. That the poverty in large animals, one may almost say 
 the dwarfish character, which in our day the South American 
 mammalian fauna, in comparison with the mammals of the 
 Eastern hemisphere, was far from obtaining, or rather did 
 not obtain at all in the prehistoric fauna. The post-pliocene 
 Mastodons, Macraucheniae, and Toxodons, those giant arma 
 dillos and sloths, could well compete with the Elephants, 
 Rhinoceroses, and Hippopotami which at the same time 
 lived in Europe." 
 
 Liais shows that the Rio de Sao Francisco is, so far as 
 length is concerned, to be counted as belonging to the six 
 teenth * class among the rivers of the world, since its length 
 is about 2,900 kilometres, or a little more than 1,802 miles, 
 and he tells us that in Europe there is only one longer 
 river, namely, the Volga. In America it is surpassed by 
 only the Amazonas, the Mississippi, the combined Parana 
 and La Plata, the St. Lawrence, and the McKenzie, while 
 in South America it occupies the third rank. The Sao 
 Francisco takes its source in the highlands between lat. 20 
 and 21 S., and flows almost due north to its confluence with 
 the Rio das Yelhas, in lat. 17 11 54" S., and long. 1 43 
 35" west of Rio. It receives two considerable affluents 
 on the right bank before reaching the Rio das Velhas, 
 the Pard, which unites with it in about lat. 19 10 S., 
 
 * Burton says seventeenth or eighteenth. 
 
THE SAO FRANCISCO BASIN. 289 
 
 and the Paraopeba, a much larger stream, which enters in 
 about 18 49 S. Both of these streams rise in the same 
 highlands with the Sao Francisco, and flow northward, 
 inclining toward the west, entering the main stream very 
 obliquely. The Rio das Yelhas is the main branch of the 
 Sao Francisco. It takes its source in the Serra da Mae 
 dos Homens, near Ouro Preto, and runs almost parallel to 
 the Sao Francisco, from which it is separated by a little 
 chain of limestone hills called the Serra do Espirito 
 Santo. Between Sabara and its mouth the river has to 
 descend nearly 263 metres, but it makes so many turns 
 that the descent per metre is very much lessened, not only 
 through the increased distance, but through the friction of 
 the river against its banks. Liais makes the descent of the 
 river 0.394 m per kilometre, while the velocity of the current 
 varies from 0.30 m to l m . The river is some 80 m in width, 
 and, were a few obstacles removed, it would be navigable by 
 steam from its mouth to Sabara, 120 leagues. To remove 
 these obstacles in the way of navigation, Liais calculated that 
 an expenditure of 260,000 would be required. Burton 
 thinks that it could be done for 55,000. The Rio das 
 Velhas flows in a narrow valley, cut through the limestone, 
 and bordered by bluffs like an lowan stream ; and Liais s 
 map shows it doubling sharply about narrow ridges, some 
 times isolated, at other times having the character of spurs 
 to the main line of bluffs. Here and there older rocks form 
 high ridges and peaks, but these are rare. Among these 
 is the Serra da Piedade, eastward of Sabara, a mass com 
 posed principally of iron ore. It is 1,774 metres in height.* 
 (Buril.) 
 
 * For a graphic account of a visit to the Piedade, see Herr Eug. Warming s 
 En Udfluyl til Brasiltens Bjerge, Liitken s Tidsskrift, &c., l" e Bind, l" te Hefte. 
 
 VOL. I. 13 g 
 
290 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The low lands bordering the Rio das Velhas are alluvial, 
 rich, healthy, and suitable for cultivation with the plough. 
 The country back from the river is wavy campos land, fit 
 only for grazing. 
 
 On the main Sao Francisco also, and its affluents, there 
 is much valuable land. The two rivers are quite well 
 settled, and from one end to the other there is seen a suc 
 cession of fazendas, hamlets, and not a few considerable 
 towns. 
 
 Liais has drawn a beautiful picture of the scenery of the 
 campos and of the Rio das Velhas, and rather than spoil it 
 by a translation, I beg the reader to allow me to give it in 
 his own words : * 
 
 " La presence d un 6pais tapis de graminees sur toute la 
 surface du sol donne, au premier abord, 1 idee d une grande 
 uniformite d aspect. Cependant il n en est pas ainsi, et les 
 paysages des Campos sont des plus varies. Des bouquets 
 d arbres dans lesquels les feuillages les plus divers s allient 
 aux fleurs de toutes couleurs portees par les guirlandes des 
 lianes ou par de superbes orchide es ou bromeliacees para 
 sites, rompent la monotonie du tapis de verdure, et Ton se 
 croirait dans un pare admirablcment cultive*. D autres fois, 
 sur le bord de petits ruisseaux, croissent des groupes de 
 gigantesques Mauritia vinifera, palmiers precieux de ces 
 regions. Leur tronc Sieve*, surmont d un magnifique para 
 sol form6 par de vastes feuilles en eVentail, produit un effct 
 des plus pittoresques, lorsque surtout une immense prairie 
 est parseme e ^a et la de ces ve getaux gracieux. D autres 
 fois, et ce fait s observe surtout dans les regions les plus 
 seches, des arbustes tortueux couvrcnt tout le terrain, et 
 dans ces parties des Campos se font remarquer les belles 
 
 * Butt, de la Soc. G&g., 5 Scric, XI. pp. 396, 397. 
 
THE SlO FRANCISCO BASIN. 291 
 
 fleurs des Car iocar, des Coclilospermum, des Vochysia. 
 Enfin, souvcnt, au milieu d une vaste plaine, on voit surgir 
 une de ces curieuses chaines de montagnes de gres rouge- 
 atre ou verdatre, a sommet coupe" en table, et si abondantes 
 dans tout le Bresil, ou M. de Castelnau les a dcja signale es. 
 Les flancs arides de ces collines, parfaitement aligne es et 
 qui se prolongent sur plusieurs lieues de longueur en gar- 
 dant le meme niveau et presentant 1 aspect d un toit, sont 
 converts par des Melocactus et par de magnifiques Kiel- 
 meycra, dont les grandcs fleurs roses rappellent celles des 
 Camillas. Lorsqu on monte sur ces collines, qui parfois 
 atteigneiit jusqu a 500 metres au dessus du niveau de la 
 region environnante, uu admirable panorama se ddroule 
 sous les yeux du spectateur. Je me rappclle en particulier 
 un magnifique tableau de ce genre que j ai apergu en gravis- 
 sant les flancs de la serra de Curumatahy. Le regard em- 
 brassait toute la largeur de la vallce du Rio das Yelhas. 
 Son fond offrait 1 apparence d une immense plaine, d ou on 
 voyait sortir comme des ilots les serras du Parauna, de 
 Buenos-Ayres, da Garga et du Bicudo. La riviere, accom- 
 pagnee sur ses deux rives d un cordon de grands arbres, 
 dessinait son cour au fond de la valle e par une ligne 
 d une verdure fraiclie qui tranchait sur la teinte rougeatre 
 des graminees dessecliees et ^clairdes par les feux du soleil 
 couchant. De belles teintes violettes couvraicnt les flancs 
 des montagnes rapprochees, et dans le lointain, a une 
 dnorme distance, une chaine de montagnes bleu pale se 
 montrait a 1 horizon. C etait la serra da Mata da Corda 
 avec ses domes dioritiques, qui limite a 1 oviest le bassin dvi 
 San-Francisco. * 
 
 * A similar panorama is to be seen from the edge of the chapadas bordering 
 the Calhao-Arassuahy valley. 
 
292 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The Upper Rio de Sao Francisco flows with a more direct 
 course, and its current is consequently more swift than that 
 of the Rio das Velhas. It is also much impeded by rapids 
 and falls.* Notwithstanding that its general level is higher 
 than that of the Rio das Velhas, its banks are very un 
 healthy, and terrible fevers, called carnadeiras, from time 
 to time drive away the population from the vicinity, so 
 that Nature has made the Rio das Velhas more fit to 
 sustain a population and be a water highway than the 
 Upper Sao Francisco. From the mouth of the Rio das 
 Velhas the Sao Francisco would be navigable for steam 
 boats, with some interruptions on account of obstructions 
 which might be removed, as far down as the Villa da B<3a 
 Vista, a distance of about 264 leagues. From that point to 
 the Porto das Piranhas, a little over 70 leagues, the river 
 is not navigable. From the Porto to the sea steamers al 
 ready ply. To remove the obstructions from the main river, 
 and make it navigable for steamers, Mr. Halfeld estimates 
 the probable cost at about X 108,900. A canal has been 
 proposed to unite Boa Vista and Porto das Piranhas ! This 
 is certainly not advisable when a railway could be con 
 structed at vastly less cost. Burton estimates that an ex 
 penditure of 203,000 would be sufficient to open the 
 Rio das Velhas and Sao Francisco, and build a railway 
 around the obstructions of the Paulo Affonso to the Porto 
 das Piranhas. 
 
 The opening of steam navigation in 1867 below Porto 
 das Piranhas has given an immense impetus to the trade of 
 the whole country adjacent to the Lower Sao Francisco, and 
 Burton says that its effects are visible even in the neighbor- 
 
 * The worst of these is a series of rapids and falls called the Pirapora, which 
 forms an obstacle that it would cost enormously to remove. 
 
THE SAO FRANCISCO BASIN. 293 
 
 ing provinces of Piauhy and Ceard. But what a future is 
 in store for the great Sao Franciscan valley when it shall 
 receive the gift of a steamboat and shall hear the scream of 
 the locomotive ! When any one stops to consider how much 
 Nature has done towards furnishing Brazil with a great in 
 terior water highway, it seems wonderful that it should not 
 long since have been improved.* 
 
 The railways of the province of Rio have already been 
 described in the " Journey in Brazil." Government is 
 pushing the Dom Pedro II. line northward into Minas, with 
 the view of continuing it over the Mantiqueira and across 
 the highlands into the valley of the das Velhas. A tram- 
 road has been commenced from Cachoeira, on the bay of 
 Bahia, to Urubu, on the great river, and there is hope for 
 the Sao Francisco, even if the Pernambuco and Sao Fran 
 cisco and the Bahia and Sao Francisco railways fail, as they 
 probably will, in reaching the river. 
 
 * The difficulty seems to be not Paulo Offonso, nor Sobradinho, nor Pira- 
 pora, but politics, and the jealousies of those who have had anything to do with 
 the matter. A steamer some time ago was built in sections, and started on 
 its overland journey to the Sao Francisco ; but I cannot learn that it has yet 
 reached its destination. Burton says that a M. Dumont brought to Rio from 
 Bordeaux two small steamers, which were to be transported in sections to the 
 Rio das Velhas, and commence running in 1869, so that it is probable that 
 steam navigation has been by this time opened on that river. I have tried in 
 vain through my Brazilian correspondence to inform myself on this as well as 
 other matters relating to Brazil, but it seems wellnigh as difficult to keep one s 
 self posted in the progress of affairs in the interior of Brazil as it is to obtain 
 news from the heart of China. 
 
294 GEOLOGY AND PHYSICAL GEOGKAPHY. 
 
 CHAPTER VII. 
 
 THE PROVINCE OF BAHIA, INTERIOR. 
 
 Journeys of Spix and Martins, Nicolay and Lacerda, Allen, and other Explor 
 ers. Geological and Physical Features of Country between Malhada and 
 Cachoeira, described by Von Martius. Sandstones. Remains of Masto 
 dons found near Villa do Rio de Contas. Immense Copper Boulder from 
 Cachoeira. Rev. Mr. Nicolay s Report of Journey from Cachoeira to the 
 Chapada Diamantina. Occurrence of Diamonds in Sandstones. Lime 
 stones. Sterile Plains. Diamantiferous Sands of the Chapada. The 
 Diamond Mines of Sincora and Lencoes. Annual Yield of the Provinces 
 in Diamonds. Mr. Allen s Report of a Journey from Chiquc-Cliique, via 
 Jacobina, to Cachoeira. Country between Chique-Chique and Jacobina an 
 immense Limestone Plain. The Chapada at Jacobina a detached flat- 
 topped Mass of Sandstone. Gneiss Hills. " Lake Plain," east of Jaco 
 bina. Knobs Potholes, probably of Glacial Origin. Eastern Sandstone 
 Plain. Climate, Vegetation, &c. of Route. Difference in Topography 
 between Gneiss Regions of Bahia and the Mucury described and accounted 
 for Former greater Extension of Forests. Von Martius s Description 
 of the Country between Cachoeira and Joazeiro. Country near Feira da 
 Conceicao Scrra do Rio Peixe. Rio Itapicurii. Want of Rain at 
 Queimados. Serra de Tiuba. Tanques and Fossil Bones near Coche 
 d Agua, Barriga Molle, and Neighborhood. Monte Santo. The Great 
 Meteorolite of Bemdego. Rock Inscriptions. Villa Nova da Rainha. 
 Joazeiro to be the Terminus of Bahia and Sao Francisco Railroad. Rio de 
 Salitre. Salt Licks. Mr. Allen s Note on the Salt of the Sao Francisco 
 Valley. Saltpetre. Geology of Country between Carunhanha and Uru- 
 bvi. Change in Geological Structure, Climate, Vegetation, &c., below 
 Urubu. 
 
 THE interior of the province of Bahia, notwithstanding 
 its rich diamond-mines, is almost a terra incognita to the 
 geologist and geographer. It forms, however, so important 
 
PROVINCE OF BAHIA, INTERIOR. 295 
 
 a part of the empire, that I have deemed it worth while to 
 collect the most important facts that bear on its geology 
 and physical geography, and with these before us, I think 
 that we shall be able to come to some trustworthy conclu 
 sion as to its general structure. Though Spix and Martins 
 explored the province while geology was yet in its infancy, 
 they made many interesting observations. The Rev. Mr. 
 Nicolay, a few years ago, visited the diamond district in 
 company with Dr. de Lacerda, and he has kindly furnished 
 me with some notes on the route he followed. Mr. J. A. 
 Allen, ornithologist on the Thayer Expedition, crossed the 
 province from Chique Chique, on the Sao Francisco, to Ca- 
 choeira, and I am indebted to him for a very interesting 
 sketch of the country he traversed. That part of the Sao 
 Francisco valley comprised in the province has been ex 
 amined by Von Martius, St. John, and others, and finally 
 by Burton, so that we know its general geological features. 
 These observers furnish us with three complete sections 
 across the country between the Sao Francisco and the sea, 
 and Mr. Nicolay gives another incomplete one. In examin 
 ing this material, we shall take up these sections in their 
 order, going from north to south, and we will first follow 
 Von Martius in his journey from Malhada to the coast. 
 
 This little town is situated on the Rio Sao Francisco, op 
 posite the mouth of the Rio Carunhanha, in the province of 
 Bahia, at the extreme northern angle of Minas Geraes. Von 
 Martius says that the vicinity " is composed of limestone, 
 which the burning of the woods not infrequently changes 
 on the surface into a white chalk-like crust. This rock 
 formation we left, on the third day s ride, between the 
 fazendas Curralinho and Pe da Serra, where we observed 
 granite, and on it here and there layers of a porous iron 
 
296 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 sandstone, in part weathered to iron ochre." He describes 
 this whole region as being covered by a catinga growth 
 with Cerei and Cnidoscoli, &c. The water is bad and slimy. 
 The population is principally engaged in raising cattle and 
 horses. Leaving this part of the country, high banks of 
 red granite, some bare, others covered thickly with cacti, 
 were met with. Near the Serra dos Montes are rounded 
 hills and mountains, composed of diorite, and destitute 
 both of soil and vegetation. The rock forming the Serra 
 dos Montes Altos is gneiss and granite. The soil is in 
 many places highly impregnated with saltpetre, but this 
 salt is extracted to no important extent. In the Serra de 
 Caytele quartzose shale or quartz rock, like that found so 
 extensively through Minas, abounds. It is flesh-red in 
 color, almost horizontal, and frequently traversed by heavy 
 veins of white quartz. East of Caytele is a level high land 
 of this same quartz rock. On leaving this, one descends 
 to reach a hilly country composed of gneiss, covered with 
 catinga forests. 
 
 The mountains in the vicinity of the fazenda of Joazeiro 
 are of granite and gneiss granite. The road thence to the 
 Villa do Rio de Contas " rises gradually, and leads finally 
 into a valley shut in on both sides by high mountains. The 
 Serra da Villa Velha rises at least 1,200 feet above the villa. 
 The base of the mountain is composed of mica slate, on 
 which rests red quartzites (Quarzschiefer), and over these 
 white rocks of the same kind." The strike, according to our 
 author, is from N. N. W. to S. S. 0., with a westerly dip, 
 which is higher in the upper beds than in the lower. The 
 foot of the mountain is covered with light vegetation, which 
 resembles the flora of Serro Frio ; on the top it resembles 
 that of Tejuco. Quartz rock, thinly laminated and clas- 
 
PROVINCE OF BAHIA, INTERIOR. 297 
 
 tic, was observed. Gold occurs in veins in the rock, and 
 also in the sands and gravels of the Brumado and other 
 streams, where it is found in grains and nuggets. Spix 
 and Martius speak of one nugget having been found weigh 
 ing eight pounds. Two leagues north of the villa are 
 other gold deposits. The great sandstone formation is 
 rightly described as extending northeastward under the 
 names Morro das Almas, Serra de Catule*, Serra. da Cha- 
 pada, c., to Jacobina. 
 
 Spix and Martius describe the top of the Morro Redondo 
 as flat, and speak of the occurrence there of a hard, white 
 sand rock, on which were drawings in red paint, supposed 
 to have been made by the Indians. This rock rests upon 
 granite, which in some places contains augite. North of 
 this is the Serra de Tiuba, and between it and the Sao 
 Francisco they found green pistacite in the granite. 
 
 Over the quartz rock lies a red sandstone, concerning 
 which Von Martius shall speak in his own words : 
 
 " The third formation, which we met here, is that of the 
 so-called red Todtliegende, or older sandstone. It occupies 
 the highest point of the mountain, as even at Brumadinho, 
 and shows, without distinct stratification, here and there a 
 thickness of several hundred feet. This rock is here com 
 posed of grayish white quartz grains, in which pieces of red 
 dish quartz sandstone and of red Grauwacke slate are im 
 bedded, and it is not infrequently intermixed with much 
 silver white mica. To this formation or to one of the over 
 lying clays belong probably certain nodules of clay iron 
 stone which are hollow inside, and contain a very fine red 
 powder, which, according to the results of an examination 
 made by my honored colleague, Hofr. Vogel, is composed of 
 iron oxide, argillaceous and siliceous earth, with some lime 
 
 13* 
 
298 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and magnesia, and is used by the inhabitants as a tonic. 
 The highest mountain of this district, the Serra de Itaubira, 
 probably presents the same formation of the red Todtlie- 
 gende on its conical head. "We saw it northwestward from 
 the Morro Redondo, rising high into the blue ether, and are 
 of the opinion that it is at least 5,000 feet high." 
 
 From the vicinity of the river Sant Antonio a beautiful 
 variety of alabaster is obtained, and, according to Spix 
 and Martius, it occurs in large quantity. It is sent to Bahia 
 to be made into images and ornaments. Dr. Lacerda kindly 
 presented me with a fine specimen of this mineral. 
 
 Bones and teeth of the Mastodon* occur apparently in 
 plenty in the vicinity of the Rio de Sant Antonio, near the 
 Villa do Rio de Contas, and near the old fazenda de Bom 
 Jesus de Meira, eight leagues from the Villa, buried in the 
 soil. 
 
 Between the Villa and the Rio de Contas are quartzites, 
 where mica slates passing into granulite make their appear 
 ance lying in granite. The granulite " holds here and there 
 masses (J&nauern) of a very hard coarse-grained cellular 
 gray quartz. On this formation we observed layers of a 
 light green, somewhat porous, very compact sandstone, 
 which appears exactly like that which in Germany is here 
 and there interstratified with the Quadersandstein. Parched 
 woods, leafless in the dry season, stretch out in immeasur 
 able extension over the hilly or even mountainous land ; 
 large tracts are covered with bushes of the Ariri palm, 
 (Cocas schizophylla Mart.), .... and here and there a 
 lighter clump of the Aricuri palm (Cocos coronata Mart.)." 
 
 The base of the Serra das Lages is composed of clay and 
 mica slates, chiefly greenish-gray in color, some approach- 
 
 * See note to p. 261. 
 
PROVINCE OF BAHIA, INTERIOR. 299 
 
 ing chlorite slate, and containing octahedral crystals of iron. 
 Higher up quartzites appear, and on the top the vegetation 
 resembles that of Minas. Near the fazenda of Lages, on 
 the top, are heavy deposits of iron ore, in the form of mag 
 netic iron, specular iron, and brown iron-stone ; " the last 
 furnishes not infrequently considerable quantities of stil- 
 posederite (phosphate of iron)." The prevailing strike of 
 the quartz rock is from north to south im Stunde 22, 23, 
 and 24, the dip of the strata at high angles from 40 - 60 
 towards the east. 
 
 In the Serra de Sincord the quartz rocks (Quarz- 
 schiefer) have a strike of N. S. im Stunde 22, 23, and 24, 
 and dip with high angles toward the east. It forms the 
 division between the high and low lands of the province of 
 Bahia ; east of it obtains a changeable wet climate, while 
 to the west there is a dry climate. 
 
 Leaving the Serra of Sincord one meets with granite, 
 hornblende, and clay slates and diorite. These are over 
 laid with layers of clayey sand of an ochre-yellow color, 
 which in some places is even ten feet thick. " Near Cara- 
 bato there overlies the granite an older sandstone ( Graues 
 Todtliegendes), which is composed of fine-grained quartz, 
 feldspar, and mica, and approaches feldspar porphyry. In 
 this are imbedded rounded masses of quartz." 
 
 At Olho d Agua great blocks of white quartz are ex 
 posed. 
 
 " The soil, which already at Olho d Agua began to be 
 hilly and mountainous, continues with similar irregularity, 
 and covered with catinga, until finally in the vicinity of the 
 Fazenda do Rio Secco, which we reached at the end of the 
 fifth day, the road sank gradually between some high, bare, 
 granite mountains, where the traveller reaches a plain, which, 
 
300 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 covered merely with dried shrubs a few feet high, presents 
 
 a more free aspect At Rio Secco there rested on the 
 
 granite, which when bedded showed a strike of W.W.W. - 
 S. S. 0., and a westerly dip (Einschiesseri) under a high 
 angle, a fine-grained hornblende rock and iron-stone." 
 The country between Villa da Pedra Branca and Cachoeira 
 Spix and Martius found to be composed of gneiss and gran 
 ite, with occasional beds of hornblende rock and mica 
 slate. 
 
 In the last century there was found near Cachoeira a 
 huge mass of native copper, which was carried to Lisbon. 
 Spix and Martius * visited the locality where this mass was 
 found, and could discover nothing that would justify them 
 in believing that the copper was derived from the rocks of 
 the vicinity, which consist of gneiss. They afterwards saw 
 the specimen in the museum at Lisbon and examined it. 
 It bears the following inscription : 
 
 " Maria I et Petro III imperantibus, cuprum nativum 
 minerae ferri mixtum pondcris libr. MMDCXYI in Bahiensi 
 Prsefectura prope oppidum Cachoeira detectum et in Prin- 
 cipis Museo P. MDCCLXXXII." According to Yandelli, 
 in the Memorias da Academia P^cal das Sdcncias tie Lisbda, 
 Vol. I. p. 261, the outside of the mass is of a hardened 
 dark yellow color. A portion of the surface was analyzed, 
 and gave ninety-seven per cent of pure copper, with no trace 
 
 * According to Von Martius, Reise, Band II. Seite 746, copper occurs at 
 the following places in Brazil : 
 
 Riberao de Sao Domingos, near Pe do Morro, in the Comarca do Serra 
 Frio, Minas Geraes, where it is found in greenstone. 
 
 Primeiros Campos, in the Serra Curaca, Province da Bahia, chloride (Saltz- 
 saurcs) and sulphuret, occurring in granite. 
 
 Arn-ial do Pinheiro, Cattas Altas da Itaperava and Inficionado in Minas 
 Geraes. 
 
PROVENCE OF BAHIA, INTERIOR. 301 
 
 of either gold or silver. Vadelli says that a second and 
 smaller piece was found near it. Spix and Martius appear 
 to have considered the mass as a meteorolite, but I have 
 seen fragments of amygdaloidal trap from the vicinity of 
 Cachoeira, and I am inclined to consider it an erratic, de 
 rived from this trap. 
 
 Mr. Allen gives me the following note : " The country 
 between Malhada and the coast, in all its leading geological 
 features, as given by Yon Martius, bears a most striking 
 resemblance to that traversed by myself some 100 to 200 
 miles to the north of this line. Some minor features, as 
 the occurrence here and there of clays and slates, etc., I 
 noticed at only one or two points, and only as insignificant 
 patches." 
 
 " At Chique Chique I observed very small patches of 
 magnetic iron ore, at times apparently in place and resting 
 on the limstones, but commonly occurring as detached 
 patches and irregular fragments of large size. It is un 
 doubtedly similar to that spoken of by Burton as occurring 
 in large quantities a little below Chique Chique." 
 
 Mr. Nicolay says that the country rises toward the Chapa- 
 da Diamantina by a series of terraces, and he estimates the 
 height of the chapada at 3,000 feet above the sea, which 
 would coincide with Mr. Allen s estimate of the height of 
 the chapada at Jacobina, and of my own estimate of the 
 height of the chapada at Minas Novas. At the " chapa 
 da," says Mr. Nicolay, there are " shales, sandstones, and 
 conglomerates. The sandstones vary much in quality, both 
 as to composition and hardness, but are all evidently the 
 direct products of primitive rocks. Upon these sandstones 
 there is Cor was) a stratum of quartzite, in many places 
 still very distinct, in which are found crystals of magnetic 
 
302 GEOLOGY AND PHYSICAL GEOGRAPHY 
 
 and other pyrites, and among the sands created by the 
 disintegration of this rock, as marked by these crystals, 
 diamonds are usually found. 
 
 " The superposition of the harder upon the softer strata 
 is the cause of the presence of those caverns called g-runa, 
 which frequently perforate the hills, and in which many 
 diamonds are found. They are all formed by the percola 
 tion of water through the rock, and the disintegration of 
 the softer strata ; but in the larger number of cases not a 
 cavern but a ruin is formed, and the surface presents a wild 
 confusion of enormous blocks or slabs of conglomerate sixty 
 to seventy feet square, and from ten to fifteen feet thick, for 
 the larger examples. As yet I am not aware that any 
 fossils have been found in this district. The chapada forms 
 the eastern limit of the barrier of the great river Sao Fran 
 cisco, and I can trace it from the sources of the Paraguassu 
 into Goyaz. 
 
 " The chapada is separated from the next division by the 
 valley of the river Sao Josd on the south, a tributary on the 
 left bank of the Paraguassu. The next range, which may 
 be called the limestone range, that rock being developed in 
 magnificent cliffs, especially on the eastern side, and pre 
 senting numerous caverns, is distant about twelve miles. 
 
 " I am not aware that this limestone has been more than 
 casually examined by Vivian near Joazeiro, by myself at 
 Mocambo, and by Cato at Rio Una, an affluent of the 
 Paraguassu, right bank, nor have I heard that any fossils 
 have been found in it. It is very distinctly bedded.* Im- 
 
 * Mr. Allen, who has carefully read Mr. Nicolay s report, says : 
 
 " The limestones mentioned by Nicolay greatly puzzle me. They seem to 
 
 occ ir on tributaries of the Paraguassu, and hence must be east of the chapada. 
 
 If so, beds of limestones must occur on both sides of the divide which sep- 
 
PROVINCE OF BAHIA, INTERIOR. 303 
 
 mediately to the east of this occurs a belt of violent dislo 
 cation, say twelve leagues in breadth, or more, presenting 
 irregular hills of primitive rock [gneiss, c. F. H.] with 
 valleys between them, not having 1 usually any outlet* and 
 for the most part covered with a forest of ancient growth. 
 
 " Here the road is strewed with large quartz pebbles, 
 and boulders of all sizes, qualities, and colors. The brejos, 
 or hollows between the hills, are sometimes lakes, more 
 often swamps, and some occasionally are quite dry. This 
 is the Serra do Mocambo, Serra da Calderao da Onc,a, and 
 Serra da Saude, and is marked on the south of the river 
 Paraguassu by the Mato dos Macacos. 
 
 " To these hills succeeds a zone of taboleiros or a taboleiro 
 or table land, where gneissose rocks are often exposed on the 
 surface, which is nearly level, but varied by occasional small 
 lakes or ponds, and riachos or watercourses, having no final 
 issues for their waters, and often dry during a part of the 
 year. This is crossed by the deep cut, formed by the Rio 
 Paraguassu, which, like other rivers to the north, presents, 
 to the extent of its enchcnte, or overflow, sometimes a mile 
 in width, a belt of verdure ; all the rest is arid, a region of 
 cacti and prickly and aromatic plants. 
 
 " Upon this surface, however, at long intervals, appear iso 
 lated masses of primitive rock at Balm, about 150 feet high 
 [above the plain], and a range of similar rocks or hills, 
 known as the Serra das Pedras Brancas, from which some 
 outliers, singularly rounded on the surface, are presented 
 at Pedra Redonda. This crosses the Taboleiro at about 
 
 arates the valley of the Sao Francisco from the sea. If so, it is a patch be 
 longing undoubtedly to the limestones developed so extensively in the Sao 
 Francisco valley. In his gneiss taboleiro I recognize my lake plain. (See 
 report of my journey.)" 
 
 * This is one of the features of a glaciated surface. C. F. II. 
 
304 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 four leagues from its eastern extremity, but without entirely 
 breaking its continuity. Beyond this the Serra da Boque 
 rao, also of primitive rock, for the most part bare, and im 
 mediately to the east the Serra Mangabeira, where I ex 
 pect on further examination to find sandstone ; beyond this 
 is undulating wooded ground for six leagues, to another 
 taboleiro of the same geological character as the other, but 
 presenting a superior vegetation, and which is again bounded 
 to the east by a chain of nearly continuous elevations, which 
 forms the main buttress of the system, the west limit of the 
 Lagoa do Rio Paraguassu, and to the east of which, only so 
 far as I know, excepting at the chapada, sandstones are 
 developed." 
 
 " Throughout the entire district the bottom rocks are 
 gneissose, varying occasionally to porphyry and granite on 
 the one hand, and hornblende and quartz rock on the other, 
 occasionally presenting micaschist." 
 
 Mr. Nicolay further remarks that " not only near the 
 limestone ranges, but on the edge of the great taboleiro, 
 saline streams are found." 
 
 From Mr. Nicolay s report, as well as from what he has 
 stated to me in conversation, there can be no doubt that the 
 diamonds of the interior of Bahia occur in a sandstone bed, 
 forming part of the great sheet which once overspread the 
 whole country, tying in with the sandstones and clays of the 
 Jequitinhonha basin ; and this sandstone, as we shall see 
 from Mr. Allen s report, is found also at Jacobina, at which 
 place, in 1755, diamonds were first discovered in the prov 
 ince of Bahia. 
 
 I saw specimens of the diamantiferous rock from the 
 chapada in the hands of Mr. Nicolay. It was not itacolu- 
 mite, but it seemed to me to bear a very close resemblance 
 
PROVINCE OF BAHIA, INTERIOR. 305 
 
 to the sandstone bed overlying the clays in the Jequitin- 
 honha basin. It also bore a remarkable resemblance to the 
 tertiary sandstones on the Bahia Railroad near Pitanga, 
 where diamonds also occur. The diamantiferous sands I 
 saw in the possession of Dr. de Lacerda at Bahia appeared 
 to have resulted from the disintegration of the chapada 
 sandstones.* 
 
 It is much to be regretted that the diamond mines of the 
 Chapada Diamantina have never been critically examined, 
 for I feel convinced that from their study the mystery of the 
 origin of the diamond is to be solved. 
 
 The metamorphio basis of the province of Bahia presents 
 a long and, on the whole, a gentle slope towards the sea, 
 and a shorter and equally gentle incline towards the Sao 
 Francisco valley. Along the highest part of the province 
 there runs an irregular strip of sandstones, occasionally 
 sending off spurs in various directions, and not infrequently 
 forming isolated patches. These sandstones, lying nearly 
 if not quite horizontally, form a series of chapadas or table 
 lands, and flat-topped hills of greater or less extent, and 
 with an elevation of about 3,000 feet. On the eastern side 
 of this line of chapadas is the diamond district, embracing 
 
 * Specimens of diamantiferous sands sent from Bahia to M. Damour were 
 found to contain <he following minerals : hyaline quartz, jasper and silex, 
 itaoolumite, disthene or cyanite, zircon or hyacinth, feldspar, red garnet, mag- 
 nesian garnet, mica, tourmaline (green and black), hyalotourmaline (feijao), 
 talc, wavellite (cabodo), yttric phosphate, titaniferous yttric phosphate, diaspore, 
 rutile, brookite, anatase, hydrated titanic acid, tantalite, baierine or colum- 
 bite, litaniferous ferric oxide, stannic oxide, mercuric sulphide, and gold. 
 (Bulletin de la Soci& Ge ologique de Paris, 2 a Serie, seance du 7 Avril, 1856, 
 p. 542.) Another paper on the diamantiferous sands of Bahia, by the same 
 author, is to be found in the Bulletin de la Soci&e Philomathtque, 5 Fevrier, 
 1853. I have been able to consult neither of these papers, and I quote 
 through Burton. 
 
306 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the head-waters of the Paraguassu and Itapicurii, forming 
 an irregular area nearly 150 miles in length from north 
 to south. 
 
 The Serra or Chapada do Sincora lies many leagues to the 
 southeastward of the Serra da Chapada proper, of which it 
 is a spur or outlier, and it appears to cross the Paraguassu 
 with a northeast trend, but the maps vary to so great a de 
 gree that I can form no satisfactory conclusion as to its 
 extent. The only description 1 can find of the serra is con 
 tained in a letter from the geologist Helmreichen.* Ac 
 cording to him the serra " bears the same raw and un- 
 hospitable character to the eye as that of the Grao Mogor ; 
 extensive campos form the country between its western slope 
 and the Serra da Chapada, while the country from its east 
 ern slope toward the coast is covered by thick woods." He 
 says that there is a close analogy between this serra and the 
 Grao Mogor in geological structure, and that probably it is 
 composed of itacolumite. " The first discovery of diamonds 
 was here made on the banks of the Macujd, and the Com- 
 mercio (the chief place), distant ninety miles from Bahia, is 
 on the Macuj6 on the lands belonging to the Fazenda de Sao 
 Joa o. Diamonds were found in the serra of Sincora over 
 an extent of twenty leagues. The washings on the west 
 side of this serra have up to the present turned out to be 
 poor. Considerable quantities of diamonds were, however, 
 washed from the Macujd itself, and from the points where 
 the Paraguassu and Andarahy cut through the serra. On 
 the Andarahy the principal washings are confined to the 
 brooks of the vicinity, which flow into it on its right banks. 
 
 * Quoted by Von Tschudi, Reisen durch Siid America, Zweiter Band, 154" 
 Seite. Helmreichen did not himself visit the Serra do Sincora, but he obtained 
 his information from a traveller in whom he put confidence. 
 
PROVINCE OF BAHIA, INTERIOR. 307 
 
 Here there are many snakes, much fever and ague, and 
 many diamonds." A very rich deposit has been discovered 
 within the last few years at Sincora, and the city has grown 
 to a very large size. 
 
 The city of Lei^oes, which is the government head-quar 
 ters of the diamond district, is situated about thirty miles 
 to the north a little east of Macuj^ or Santa Isabel do Pa- 
 raguassu, and is a large and very important place, and in 
 the vicinity great quantities of diamonds are washed.* Cas- 
 telnau says that along the course of the river of Lenses 
 there are pot-holes, some of which are of the depth of 
 twenty-five with width of one or two bra$os ! In these cal- 
 deiroes, as they are called, a considerable number of dia 
 monds have been found. The same author says that these 
 pot-holes are found also in the chapada and are always rich. 
 
 Diamonds also occur at a locality not far southeast of 
 Chique Chique, at a locality called the Corrego de Santo 
 Ignacio, visited by Burton, who describes the vicinity as 
 composed of itacolumite.(?) He says that there occurs here 
 as at the chapada " a bouldery, not pebbly conglomerate, 
 which resembles that of the Scottish Old Red," f so that it 
 would seem that the hills among which the Santo Ignacio 
 
 * "Lcs plus beaux diamants de la chapada viennent dos Lencoes (les draps), 
 lieu situe a vingt lieues de Santa-Isabel, cette bourgade tire son nom d un gros 
 ruisseau, enclave dans une gorge profonde ; il se precipite de sommets eleves sur 
 de large dalles, et apres y avoir parcouru environ trois cents metres, il se jette 
 en formant des cascades dans le rio Sao Jose. Tout a 1 entour de ces mines, des 
 montagnes enticres, des blocs dnormes composes en grande partie de cailloux 
 roule s et cimentes par une pate ferrugincuse et presque noire, temoignent de 
 grandes revolutions geologiques. En ge neral, les pierres ont des formes tres 
 re gulieres, et celles qui pre sentent la cristallisation en octaedre forment la 
 grande exception." Castelnau, Histoire du Voyage, Tome deuxieme, p. 343 
 (note). 
 
 t Burton, Highlands of Brazil, Vol. II. p. 336. 
 
308 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 diamond diggings are located belong to the same formation 
 as that of the chapada.* Burton describes the Serra do 
 Pintor as table-topped. 
 
 I regret that I have been unable to obtain precise infor 
 mation concerning the annual yield of the diamond-mines of 
 Bahia as well as to the size,f quality, and relative value of 
 the stones. According to the report of the President of the 
 province, published in 1866, there were exported through 
 the custom-house as follows : during the year 1862 - 63, 
 diamonds to the value of 1.647 :450|000 ; during 1863-64, 
 1.476:9001000; and during 1864-65, 1.381 :500$000 ; 
 which figures show a decrease during these three years, but 
 they cannot be considered as giving even an approximate 
 idea of the annual yield, for only a very small proportion of 
 the diamonds actually exported pass through the custom 
 house, so great are the facilities for smuggling. So nearly 
 
 * Mr. Allen, in looking over my MS. has kindly added the following 
 note : 
 
 " The Serra do Assurua I did not visit, but I saw it at a distance and infer 
 red it to be of sandstone. As it overlies the limestone, it seems to be evidently a 
 part of the sandstone formation noticed by me to the eastward, as already men 
 tioned. I am now fully convinced of the truthfulness of your generalization in 
 respect to the former great extent and subsequent denudation of the sandstones. 
 The occurrence of gold and diamonds in the above-named serra, as also at Jaco- 
 bina, was repeatedly spoken of to me by many trustworthy persons." 
 
 In 1858 the government conceded to the Companhia Metallurgica do Assurua 
 the right to mine gold and other metals within the space of four leagues. Oli- 
 veira, Exploracao de Mineraes, published as an Annexe in a government report 
 for 1866. My copy has no title-page. 
 
 t The diamonds of the chapada are often of considerable size, and Burton 
 says : " The Chapada of Bahia also produced a stone weighing 76J carats, and, 
 when cut into a drop-shaped brilliant, it proved to possess extraordinary play 
 and lustre. It was bought by Mr. Arthur Lyon of Bahia for 30 contos 
 [S 15,000], and it is now, I am told, in the possession of Mr. E. T. Dresden." 
 (Burton, Highlands of Brazil, Vol. II. p. 153.) 
 
PROVINCE OF BAHIA, INTERIOR. 309 
 
 as I am able to ascertain, the annual production in dia 
 monds of the province cannot fall far short of three millions 
 of dollars. 
 
 I am indebted to Mr. Allen for the following : 
 
 Notes on the Geological Character of the Country between Chique- 
 Chique, on the Rio de Sao Francisco, and Bahia, Brazil. By 
 J. A. ALLEN. 
 
 Chique-Chique is a small village, situated on the Rio de Sao 
 Francisco, about fifty miles below the mouth of the Rio Grande ; 
 it is a little north of the parallel of Bahia. My journey thence to 
 the latter place was by the route usually taken by mule-trains in 
 passing from the Villa da Barra do Rio Grande and Chique-Chique 
 to the coast, namely, by way of Engenho Velho, Jacare", Jacobina, 
 Arraial do Riacho do Jacuhipe, Villa da Feira da Sta. Anna, and 
 Cachoeira. As I found it necessary to perform the journey over 
 this unsettled and poorly watered district in company with the 
 large eastward-bound tropas, I was obliged to pass on hurriedly, 
 and had not time to explore the country adjacent to my route, or 
 for a satisfactory examination of many of the interesting localities 
 immediately upon it. The following is a summary of such geo 
 logical observations as I was able to make, the geology of the 
 country being to me at the time a matter of secondary interest. 
 
 The country between the Sao Francisco at Chique-Chique * and 
 the coast at Bahia presents three natural regions, which are 
 
 * Mr. Allen says : " This name is always -written by the inhabitants of the 
 village as above ; never, so far as I observed, Xique-Xique. The place takes 
 its name from the abundance of a low branching form of Cereus, called by this 
 name that grows here." Burton uses Chique-Chique for the town and Xique- 
 Xique for the cactus ; a distinction, it seems to me, without a difference, since 
 both are pronounced alike. The vicinity of Chique-Chique appears to be a 
 perfect paradise of cactuses. The name is applied to several species of the 
 plant. See Burton, Highlands of Brazil, Chap. XXII. 
 
310 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 LEVEL 
 CACHOZIFSA 
 
 plateaus, differing widely from each other in geo 
 logical and other characteristics. The first or west 
 ern extends from the river above named to the 
 vicinity of Jacobina, a distance of rather more 
 than two hundred miles. It is a vast limestone 
 plain rising almost imperceptibly from the level 
 of the Sao Francisco River to the summit of the 
 divide which separates the waters flowing west 
 ward and northward into this noble stream from 
 those that reach the sea by other channels. Here 
 and there large exposures of the underlying rock 
 occur, commonly scarcely rising above the general 
 level, but occasionally forming irregular, pinnacled 
 hummocks or low serras.* The first so-called 
 " serra " met with occurred at Sta. Euzebia ; 
 thence eastward they were frequent. At Sta. 
 Euzebia these elevations rarely attained a height 
 of fifty to one hundred feet, above the general 
 level of the country ; farther east, and particularly 
 in the Volta da Serra, they rise much higher than 
 this, and in some cases probably reach an altitude 
 of nearly a thousand feet above the plain. The 
 limestone for many leagues to the eastward of 
 Sta. Euzebia is very compact, dark blue, and dis 
 tinctly stratified, resembling lithologically some 
 that I had previously seen on the lower portion 
 of the Rio das Velhas and at Lagoa Santa, as also 
 
 * Burton says that the limestone at Chique-Chique is 
 largely quarried for burning, and he suggests that it would 
 make a good hydraulic cement. According to him, " Chique- 
 Chique annually sends up and down stream between the 
 Villa da Barra and Joazeiro fifteen hundred to two thousand 
 alqueires." Mr. Allen remarks that " the weathering of the 
 
 limestone often leaves the surface studded with acicular or small attenuated 
 
 points." 
 
PROVINCE OF BAHIA, INTERIOR. 311 
 
 on the Sao Francisco at Urubu. Further eastward this stratified 
 variety passed into an earthy, light-colored, unstratified kind, 
 which was frequently greatly decomposed at the surface, becoming 
 soft, white, and chalky sometimes to the depth of several feet.* 
 This gave to the distant, nearly verdureless hills in many places 
 the appearance of being covered with snow. Bands of the strati 
 fied rock, which was sometimes quite shaly, alternated with those 
 that were unstratified ; in some instances the passage of the one 
 into the other was easily traced. The strike of the limestone 
 strata varied from E. and W. in the western part of the plateau 
 to W. N. W. and E. S. E. f in the eastern. The dip was at first 
 nearly vertical or somewhat to the southward, but afterwards an 
 inclination to the northward was observed. Large caverns were 
 reported to be of frequent occurrence throughout this limestone 
 district, from some of which, I was told, very large bones had 
 been taken. Many of the caverns are doubtless very rich in 
 palseontological treasures, but want of time and other circum 
 stances would not allow me to visit them, though I greatly desired 
 to do so. The extent northwards of the limestone I had no means 
 of determining. Its southern limit I once saw at a locality known 
 as Olhos d Agua, about seventy-five miles west of Jacobina. At 
 this point we turned southward a few miles from our usual course 
 to visit a spring of water, and found an extensive exposure of a com 
 pact quartzose sandstone, which was horizontally stratified and over 
 laid the limestone unconformably, the dip of the latter beneath it 
 being at a considerable angle. The sandstone here stretched away 
 to the southward for miles like a vast level floor, its surface cov 
 ered only with detached angular blocks of the same rock, and sup 
 porting a few cacti. A distant low, even serra seen in the same 
 
 * Von Martins thought that this chalky crust resulted from the burning over 
 of the surface; but the heat would not be great enough to produce such an 
 effect. 
 
 t This strike is very remarkable, for it usually varies in Brazil from N. N. E. 
 to E. N. E. C. F. H. 
 
312 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 direction, and called Serra das Pedras d Agua, was doubtless of the 
 same formation. At Jacobina a similar horizontal sandstone was 
 observed, which gave rise to a beautiful level grassy plain, called 
 the Taboleira de Jacobiua, and which occupied the summit of the 
 water-shed. 
 
 Near the Volta da Serra two large hills of hornstone, or chert 
 ("pedra de fogo " of the Brazilians), were crossed in the few 
 leagues intervening between this considerable serra and the Jaco 
 bina taboleira. One of them attained the estimated height of 
 several hundred feet and was several miles across, while the other 
 was about half these dimensions. 
 
 In respect to the relative age of the limestone of the western 
 plateau I obtained but few data. It must, however, be much 
 older than the sandstones already referred to. The compact 
 stratified portion has a very striking lithological resemblance, as 
 previously observed, to limestones seen on the lower part of the 
 Rio das Velhas, which rested conformably upon very old clay 
 slates. No fossils were seen in it, but nodular concretions were 
 frequent at a few places. The Taboleiro de Jacobina is the most 
 elevated part of the region under consideration. On several sides 
 the country slopes gradually, but to the eastward the descent to 
 the Jacobina valley is abrupt, through a narrow precipitous defile, 
 called the " Tombador " (literally the " tumble down "). The con 
 trast of the scenery here with that of the taboleiro, which the 
 traveller may have left but an hour or two before, is very great. 
 Almost vertical walls of rock, nearly a thousand feet in height, 
 quite surround the head of the valley, while lower down are iso 
 lated, dome-shaped peaks within these enclosing walls.* 
 
 These peaks are composed of compact gneiss, though a few of 
 the smaller appear granitic, but the upper portion of the walls is 
 quartzose sandstone. The valley seems to have been formed by 
 a rent in the sandstone, which was subsequently eroded to its 
 
 * This scenery must bear the closest resemblance to that of parts of the 
 Arassuahy-Calhao valley. C. F. II. 
 
PROVINCE OF BAHIA, INTERIOR. 
 
 313 
 
 present size and form. From the village of Jacobina the level 
 summits of the distant valley walls to the westward are conspic 
 uous features in the landscape. 
 
 Leaving Jacobina and winding among the hills for a few leagues 
 we soon enter upon the second or middle plateaii, which extends 
 thence eastward to the Serra da Terra Dura, a point midway be 
 tween Jacobina and Cachoeira, at the head of Bahia Bay. Through 
 out this distance of nearly two hundred miles I found generally 
 only different varieties of gneiss, usually very compact, and some 
 times so granitic in structure as only here and there to present 
 well-defined stratification. At one or two points hornblende rock 
 was noticed, of which there was a considerable exposure on the 
 Riacho de Jacuhipe, near the arraial of that name. The dip was 
 always eastward, and usually very great ; the strike varied some 
 what at different localities. From Jacobina nearly to the Serra 
 da Terra Dura it was generally N. N. W. and S. S. E. ; sometimes 
 N. W. and S. E., and at one or two points N. and S. East of the 
 Arraial do Riacho de Jacuhipe it was nearly E. and W., as in the 
 Morro da Lagoa do Boi and Morro do Curral Velho, which have 
 this trend. In the Serra da Terra Dura, but a few miles further 
 
 VOL. I. 14 
 
314 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 east, it varied at different points from N. N. E. and S. S. W. to 
 N. N. W. and S. S. E., averaging about N. and S. 
 
 Over this whole region there is an almost entire absence of 
 loose materials on the surface. Vast exposures of nearly bare 
 rock, sometimes of acres in extent, occur almost constantly, and 
 nowhere is there more than a very thin stratum of soil. Slight 
 knolls and shallow basins alternate, which rarely differ more than 
 twenty or thirty feet in elevation.* In the rainy season many of 
 these basins become filled with water, forming shallow lagoas, 
 varying in area from less than one to more than fifty acres, from 
 the most of which the water evaporates in the dry season. They 
 are filled with rank aquatic vegetation, the dead parts of which, 
 accumulating year by year, already form large deposits of partially 
 decomposed vegetable matter. So numerous were these lagoas for 
 more than fifty miles that it seemed natural to speak of this 
 region in my notes as the " Lake Plain." Almost everywhere the 
 elevations are evenly rounded, indicating that the rocky crust has 
 been exposed to severe and probably long-continued abrasion. 
 But the absence of abraded materials seemed most remarkable. 
 Very rarely were even loose boulders observed, though a few such 
 were repeatedly noticed. At frequent intervals there were singu 
 lar holes in the rocks, usually nearly filled with water, to which the 
 inhabitants give the name of " caldeiraos." t These " caldeiraos" 
 are of frequent occurrence, but I was unable to learn whether all 
 
 * The country just below the falls of Paulo Affonso, at Piranhas, for in 
 stance, though composed of gneiss, is worn down in the same way almost to 
 a plain ; but while it agrees with Mr. Allen s " lake plain " in the thinness of 
 the soil, the part I saw was abundantly strewn with loose rocks. See reference 
 to Piranhas, in index. 
 
 t The term caldeirao has the same derivation as our English word caldron 
 (chaudicre), and it has the same signification. It is applied to true pot-holes, 
 but sometimes to rock basins in which water collects ; but these last are more 
 frequently called po$os, and when excavated they form tanqites. Mammalian 
 bones (Mastodon, &c.) are not infrequently found in the caldeiraos of the lake 
 plain. C. F. H. 
 
PROVINCE OF BAHIA, INTERIOR. 315 
 
 were of a similar character. Nearly all of the considerable number 
 examined proved to be genuine pot-holes, and some of them were 
 of great size. The largest one I measured was elliptical in outline, 
 eighteen feet long, nine or ten in width, and twenty -seven deep, 
 with smoothly worn sides. Beneath the water that partially filled 
 it there must have been many feet of materials that for ages have 
 been falling into it, so that its whole depth must be much greater 
 than my measurements indicate.* Near tho Serra da Terra Dura 
 the country becomes somewhat diversified by the presence of ab 
 ruptly rising points or knobs that at intervals dot its surface, as 
 
 * Mr. Allen tells me that these pot-holes often occur out on the plain, far 
 away from any high land, and that they are sometimes found excavated in 
 the summits of slight bulgings in the plain, or even on the top of a hill, as 
 in the case of the Morro do Caldeirao. These holes must have been exca 
 vated by falling water. There is only one suggestion that I can make as to 
 their origin, and that is that they were formed by glacial waterfalls, in the 
 same way as the pot-holes found over the glaciated regions of North America, 
 as, for instance, in New Brunswick and Nova Scotia, where I have had an 
 opportunity of examining them. It is well known that glacial waterfalls, 
 notwithstanding the constant movement of the ice, are very often stationary, 
 and in the Alps they hollow out enormous pot-holes in the rocks. The lake 
 plain is noted for the small amount of decomposition which has taken place 
 over it, owing, I believe, largely to the fact that it has never been covered bv 
 the virgin forest, having always been dry. C. F. H. 
 
316 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 illustrated in the accompanying sketch, taken from an eminence 
 known as the Morro do Caldeirao-assu, there being a large pot 
 hole near its summit. This elevation reached the height of about 
 one hundred feet, and was the most considerable elevation crossed 
 in this part of the journey. It overlooked the plain for many 
 miles in every direction, and the knobs rising from it looked not 
 unlike small rocky islands in the sea. Those examined were com 
 posed chiefly of quartz, the great hardness of which may have pre 
 vented their being so much abraded as the softer strata enclosing 
 them. At its eastern border the middle plateau becomes more 
 broken, and merges gradually into the Serra da Terra Dura. 
 These hills are all composed of gneiss, varying somewhat in char 
 acter, but generally very hard and compact, with a rather small 
 amount of mica. The average trend of the strata has already 
 been given. 
 
 From the Serra da Terra Dura a considerable descent is made 
 in reaching the third, eastern, or coast plateau. This is char 
 acterized, so far as observed by myself, by the general absence of 
 rock exposures, and a deep sxiperficial deposit of compact sand, 
 probably detritus from the naked abraded plains to the westward. 
 Its extent and general features are too well known to require 
 a detailed description here.* 
 
 The three plateaus described above are separated from each 
 other by low mountain chains and belts of broken country. The 
 sandstone serra of the Tombador and the gneiss range of Jaco- 
 bina divide the western from the middle one, while the Serra da 
 Terra Dura separates the latter from the eastern. In some 
 respects these several districts are somewhat alike, but geologically 
 they widely differ. The first, as previously observed, is a vast 
 limestone plain, two hundred miles in extent, rising gradually to 
 
 * This plain, which Mr. Allen has represented on his profile, he describes as 
 continuing down to Cachoeira, and it is undoubtedly composed of the tertiary 
 sandstones which extend over so large an area at the head of the bay of Todos 
 os Santos. C. F. H. 
 
PROVINCE OF BAHIA, INTERIOR. 317 
 
 the eastward ; the second, of equal breadth, is gneissose, and 
 apparently everywhere of nearly uniform altitude; the third is 
 narrower, somewhat lower, and sandy. The whole region be 
 tween the Sao Francisco and the sea is covered usually with a low 
 forest, or catinga, except a narrow belt along the coast, where 
 a moister atmosphere cherishes a more luxuriant growth. The 
 country everywhere wears a barren aspect, the vegetation is 
 dwarfed and scanty and the aridity of the climate is excessive. 
 The greatest aridity and the highest temperature obtain in the 
 limestone district, where for nine months of the year little or 
 no rain falls and all the herbaceous vegetation annually withers. 
 Cacti in great variety, including some of gigantic proportions, with 
 various species of bromeliacese, are leading forms in the vegeta 
 tion. With the exception of a few species, the trees are leafless 
 throughout the long dry season, and the streams either become 
 dry or form mere chains of brackish pools. The convolvuli and 
 other vines that overrun the catinga, though dead at the time of 
 my journey, indicate an excessive luxuriance of foliage and flowers 
 during the short rainy period. 
 
 The middle or gneiss district differs but little in its climate and 
 vegetation from the preceding ; it is, however, less arid, and cacti 
 are proportionately less frequent, though still a leading feature of 
 the vegetation. The eastern or sandy plateau is also quite arid in 
 its western part, but gradually becomes moister towards the sea, 
 where the vegetation exhibits the ordinary luxuriance character 
 istic of the Brazilian coast. In the vicinity of Jacobina, however, 
 where the great altitude of the land arrests the currents of damp 
 air from the sea and condenses their moisture, mists and light 
 rain-falls occur at frequent intervals throughout the year, and the 
 forests are not only of larger size, but their verdure is perennial. 
 In the Serra da Terra Dura a nobler forest growth is also seen, 
 resulting from causes similar to those existing at Jacobina. 
 
 The population of the middle and western plateaus is extremely 
 sparse. The settlements consist of but a few families each, and 
 
318 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 occur only at long intervals. Nothing approaching the character 
 of a village is met with between the Sao Francisco and Jacobina. 
 The middle district is more thickly settled, little hamlets being 
 more or less frequent, and there are a few small villages. The 
 eastern is comparatively well settled and largely under cultivation ; 
 towards the coast the soil is very productive. 
 
 A journey across the limestone plain is always tedious and 
 difficult. Extra animals must be taken to transport food for both 
 men and beasts, and in the dry season water must in like manner 
 be provided for use in crossing the long stretches where none can 
 be obtained. During the rainy season the swollen streams and 
 noxious exhalations from the temporary pools render the journey 
 equally troublesome and far more dangerous. 
 
 The topography of the arid, gneiss country of the interior 
 of Bahia, Sergipe and AlagGas, with its great plains and 
 bare surface, is in striking contrast with that of the forest- 
 clothed gneiss region of the coast of these provinces and of 
 those to the south, where the surface of the gneiss never 
 forms plains, but is always hilly and ridgy, and covered by 
 a thick bed of drift-clay. This difference in topographical 
 features has resulted, at least in so far as the last surface- 
 moulding is concerned, from the different climatic influences 
 to which these provinces have been subjected. Over both 
 of these parts of the country an extraordinary amount of 
 erosion has taken place. One might at first be inclined 
 to consider that the amount of denudation had been greater 
 over the gneiss plains than in the hill-roughened basin of the 
 Mucury, since on the former the upturned gneiss strata are 
 planed down to a more even level ; but it seems to me that 
 this feature furnishes no criterion. The peculiar topography 
 of the wooded gneiss region is owing to the prevalence of a 
 very moist climate, giving rise to numerous streams, which 
 
PROVINCE OF BAHIA, INTERIOR. 319 
 
 have furrowed the surface with an intricate system of water 
 courses, such as we do not find over the gneiss plains of the 
 north, for the grand physical features of the northern coun 
 try in question are such as produce a dry climate, and pre 
 clude the possibility of that unequal erosion such as is pro 
 duced by the flow of surface waters fed by heavy periodical 
 or constant rains. Decomposition must have played its part 
 before the drift as well as afterward, for else how could 
 those feldspathic sandy clays of the tertiary, spread over 
 the coast plains and the great Amazonian valley, have been 
 formed ? I cannot resist the conclusion that the present 
 forest belt was wooded before the drift, and that this forest 
 and its attendant peculiarities of climate, were the causes of 
 decomposition as at present, which decomposition has aided 
 immensely in rounding down the hills and producing topo 
 graphical features which received their finishing touches 
 from the glaciers. 
 
 In the interior of Bahia, behind the arid region, where, as 
 in the Chapada, or the Serra de Tiuba, the country throws 
 up barriers to the air-drift and causes the condensation of 
 moisture, we find a different kind of topography, and the 
 surface is deeply furrowed. There are forests, and decom 
 position has taken place to a greater or less extent. 
 
 The limits of the forests, of the belt of decomposition, 
 and of the area over which copious rains fall, coincide very 
 remarkably, and show a dependence upon each other, but 
 the forest belt has a smaller area than that of decomposition 
 or of the rains. The wooded belt seems to have narrowed 
 greatly within comparatively recent times, losing its foot 
 hold in the west, where immense regions, now campos, over 
 which the climate and soil would normally be proper for the 
 growth of forests, have dried up, the climate has become 
 
320 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 hot, less rain now falls, and the forest cannot regain its lost 
 place. Doubtless there are many natural physical causes to 
 be taken into consideration in studying the distribution of 
 the forest, catinga, and campos florae; but there is one 
 agency which has been at work in Brazil, whose effects we 
 can hardly overestimate, and that is the burning over of 
 wood and campos lands by man.* The very physical fea 
 tures of the highlands of Brazil determine a difference of 
 luxuriance in the florae of different regions, and there are, 
 as I have already shown, regions where for ages the climate 
 has been such that forests could scarcely have had any 
 noteworthy extension, so that there must have always been 
 in Brazil, naturally, virgin forests, catingas, campos, and 
 barrens. On the coast, where the forest is dense and moist, 
 and the climate is wet, forest fires are next to impossible, 
 and one never sees a scorched and dead wood, such as cov 
 ers so large an area in the province of New Brunswick, for 
 instance. But in the interior, where the catinga forests 
 drop their leaves, and are as dead for several months in the 
 dry season, fires are easily kindled and the wood killed ; and 
 fires set in open fields or campos, for the purpose of pro 
 ducing a new crop of grass, may spread to the neighboring 
 catingas. It is the opinion of many writers that a large 
 part of the catinga and campos region of the Brazilian high 
 lands was once covered by forests, and that their present 
 bare appearance and the character of their florae is in very 
 great measure due to frequent and extensive burning over 
 of the country. Every year the Brazilian campos lands are 
 systematically and almost entirely burned over, for the pur 
 pose of producing a new crop of grass. This burning of 
 
 * For a very interesting article on the effect of the burning over of the 
 campos, see Tidsskriftfnr Pop. Frem. of Naturvid., Camposflaraen og Campos- 
 brandene, by Eug. Warming. 
 
PROVINCE OF BAHIA, INTERIOR. 321 
 
 course has destroyed all those trees and shrubs and plants 
 of all kinds that cannot bear the scorching, and has wrought 
 a great alteration in the character of the whole flora of the 
 region ; the climate also has suffered a change, for with the 
 destruction of the woods and forests it becomes hotter, the 
 unprotected earth is like a furnace, streams run dry a few 
 days after a shower, and the springs disappear. The whole 
 sale and careless destruction of the forests on the Brazilian 
 coast, unless put a stop to, will in the end work a sure 
 ruin to the country. Brazil owes her climate and fitness 
 for agricultural purposes to her forests, and it is absolutely 
 necessary that they should be preserved over a very large 
 part of the country, especially on the coast. The climate 
 of the Bahia has already suffered from the destruction of 
 the forests of the Reconcavo, and the burning over of the 
 plains. But I fear that Brazil will learn this fact only 
 when it is too late. 
 
 This whole subject of the former wooding of the Brazilian 
 campos is the same as that relating to the North American 
 prairies, which many suppose to have once been wooded. 
 Dana has shown that the existence of forests depends upon 
 moisture, and any climatic change which may lessen the 
 amount of moisture over a region may cause the thinning 
 out and final disappearance of its forest ; and it is more than 
 probable that some such influence, beside that of forest clear 
 ing and burning, has been at work in Brazil. 
 
 Taking the Estrada de Capoeirassu on their journey from 
 Cachoeira to Joazeiro, Spix and Martius ascended the steep 
 slope of the same name as the road to a height of about 
 seven hundred feet, when they reached the top of the dry 
 plateau. 
 
 14* TT 
 
322 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The rock in the vicinity of Cachoeira is gneiss, reddish or 
 yellowish in color, with a N. S. or N. E. S. W. strike 
 and westerly dip. In some parts specular and magnetic 
 iron were observed to take the place of mica in the rock. 
 
 Two leagues from Cachoeira the country becomes barren 
 and uninhabited. Here is a little place called Feira de 
 Concei^-ao. " The plain, as a general thing elevated from 
 six to seven hundred feet above the sea, forms here and 
 there shallow hollows, in which during the rainy season 
 brackish water, often unfit for the use of cattle, collects. 
 In other places one sees in several directions rows of hills 
 with gently sloping sides. The only rock we found was 
 gneiss, gneiss granite, or granular granite, for the most 
 part of a reddish or yellowish color, though sometimes also 
 blackish or white. This rock lies entirely bare over a large 
 extent of surface, or is covered by a thin coat of a heavy 
 red clay, which appears to originate from the decomposition 
 of the same rock. Besides this, fragments of granite and 
 fine granite lie scattered about. In the low lying and wet 
 places one finds little woods very much like the capoes of 
 Minas Geraes ; the higher plains and the hills are in some 
 cases bare of all vegetation, in others clothed with single 
 cactus stems and plants or with thick bushes and low trees. 
 All these plants belong to the catinga group, for they shed 
 their leaves in the dry season, and for the greater part clothe 
 themselves only on the entrance of the rainy season. Only 
 in the low wet places do the leaves remain on during the 
 
 whole year The wood is never wholly without sap 
 
 during the leafless season." 
 
 The trees leaf out with marvellous rapidity, and a short 
 rain suffices in two or three days to clothe a wood with 
 spring-like verdure. Yon Martius speaks of the roots of the 
 
PROVINCE OF BAHIA, INTERIOR. 323 
 
 Imbuzeiro (Spondias tuberosa Arr.) which extend under 
 ground near the surface, and are full of swellings as large 
 even as a child s head, hollow and full of water, a pro 
 vision against the drought. During the dry season these 
 hot plains are almost destitute of life. The same kind of 
 country extends on to the Rio do Peixe, and during the dry 
 season it is almost wholly without water. When Spix and 
 Martius reached the Rio do Peixe, they found it only a 
 string of brackish-water pools. 
 
 Crossing the Rio do Peixe our travellers passed over a 
 range called the Serra do Rio do Peixe, which is described 
 as being composed of gneiss and granite, and strewn with 
 gigantic and isolated blocks of gneiss. In some places horn 
 blende rock was observed. 
 
 The Rio Itapicuru was, like the Rio do Peixe, dried up, 
 forming only a string of pools. Over all this region this is 
 the state of the streams during the dry season ; but a week s 
 rain fills their dry beds and converts them into torrents, 
 which if the rain does not continue, soon become dry again, 
 for the surface water runs speedily off from the bare rocks 
 and exceedingly scanty soil. Such a country is of course 
 barren and unfit for culture. Rain falls abundantly enough 
 on the sea-coast, but the air soon parts with its moisture, or 
 becomes so heated that but little or none is condensed in the 
 interior. The Arraial das Queimadas lies only about one 
 hundred and thirty miles from Bahia. But very little rain 
 falls there, and Von Martius relates that the inhabitants 
 assured him that in some parts in their vicinity it had not 
 rained for three years. A little cotton and maize are culti 
 vated there. One league N. W. from this place red gneiss 
 was found, strike N. N. E. At Bebedor, one league farther, 
 white granite, with a N. W. -S. E. strike, was seen. Ap- 
 
324 GEOLOGY AND PHYSICAL GEOGEAPHY. 
 
 preaching the Serra de Tiuba light green pistacite became 
 more and more abundant in the granite, at first in grains, 
 taking the place of mica, afterwards in bands through the 
 rock. Thin layers of a slate-like hornblende-stone were 
 also observed in the rock. 
 
 The Serra de Tiuba, where our travellers passed it, is 
 about twelve hundred feet high above its base, and is com 
 posed of reddish granite passing into sienite. It is wooded 
 to its summit, probably owing to the condensation of the 
 moisture in passing over the serra, producing a damper 
 climate ; and I am strengthened in this opinion, since Von 
 Martius says that the trees increase in height in going up 
 the serra. 
 
 West of the serra a more level country succeeds, and con 
 tinues to Villa Nova da Rainha, southwest and north of 
 which are mountains which show in their valleys high 
 catinga woods and a comparatively thick covering of fer 
 tile soil. From Villa Nova Spix and Martius made a rapid 
 journey to Monte Santo, to visit the great meteorolite of 
 Bemdego. 
 
 The way ascends gradually, with occasional patches of 
 catinga, to the fazenda called Coche d Agua, on the west 
 side of the Serra de Itauba. The hills at the base of the 
 mountains are, like the serra itself, of gneiss-granite. 
 " The rock is for the most part bare, but here and there 
 a thickness of from four to five feet of reddish clay lies in 
 the shallow hollows which the Sertanejos excavate to a 
 depth of several feet in order to make tanks for the keeping 
 of the rain-water. In these hollows numerous bones of 
 ancient animals are found, for the most part in a broken 
 state, and so scattered that one hardly dare hope to find a 
 complete skeleton. The bones in a recognizable condition 
 
PROVINCE OF BAHIA, INTERIOR. 325 
 
 which we had the opportunity to gather are the under jaw, 
 a vertebra, ard a part of a shoulder-blade of a mastodon." 
 
 At the fazenda called Barriga Molle similar bones were 
 found, and at Mundo Novo and Pedra Vermelha Spix and 
 Martius found in a tanque the head of a femur. Other 
 localities for these fossil remains mentioned by Von Martius 
 are the Fazenda de Sao Gonzalo and Caldeiroes, and at the 
 Fazenda Canganc.ao, near the Monte Santo. 
 
 Monte Santo is an isolated hill of mica slate which has a 
 N. - S. strike. The height of the mountain is, according to 
 a barometrical measurement, about seventeen hundred and 
 sixty feet above the sea. The rock is said to contain dis- 
 thene. In the vicinity are several serras characterized by 
 Von Martius as resembling one another in their " round ed- 
 off, long-drawn-out ridges, without steep sides, gaps, or rug 
 ged cliffs," the whole being covered with catinga vegetation. 
 
 At a place called Bemdego, near Monte Santo, Spix and 
 Martius examined an enormous block of meteoric iron, al 
 ready visited by Mornay * in 1811, and Von Martius gives a 
 long description of it in his Reise, to which the reader is 
 referred. It was discovered in the year 1784 by a man who 
 was searching for a lost cow. Coining under the notice of 
 the Governor, an attempt was made to carry it off, under 
 the impression that it was silver ; but the cart broke down, 
 and Spix and Martius found it long afterwards lying in a 
 brook nearly buried with sand. They give the greatest 
 length at eighty Paris inches, the greatest breadth 43^", 
 and the greatest height 34 ^".f The specific weight was 
 
 * Phil. Transactions, 1816, p. 270. 
 
 t Mornay, loc. cit., gives its dimensions as 7 X 4 X 2 feet, the cubic contents 
 at 28 feet, and the weight of the whole mass at 14,000 pounds. See Dana s 
 System of Mineralogy, p. 16. Mornay s paper is accompanied hy a plate, 
 showing the shape of the meteorite. Von Martius also figures it in his atlas. 
 
326 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 7.731. They estimated the weight of the whole block at 
 17,300 Paris pounds. Fragments of the mass were cut off 
 and carried to Europe, where they were examined by 
 Fickentscher. Wollaston had already made a chemical 
 analysis of it. The latter found it to contain, iron, 96.1 ; 
 nickel, 3.9 ; while Fickentscher obtained 
 
 Iron 96.10 
 
 Nickel 5.71 
 
 pron ^ 
 
 I Nickel I 
 Mixture of S SiHcon ? .46 
 
 v Carbon ) gg QJT 
 
 The prepared surface showed the Widmannstadtian fig 
 ures. 
 
 In the vicinity of Bemdego were found some strange 
 characters painted on a rock, apparently by Indians, and 
 Von Martius gives a sketch of them in the atlas accom 
 panying his work. 
 
 Burton * gives a long list of localities in Brazil in which 
 inscriptions occur either engraved in or painted upon 
 stone, and he figures a number of the hieroglyphics he 
 observed on the banks of the Sao Francisco, a short distance 
 above the rapids of Itaparica. He gives also copies of cer 
 tain other glyphs, observed by Mr. C. H. Williams, of Ba- 
 hia, on the Rio Panema, one of the influents of the Lower 
 Sao Francisco. Characters of this sort, which his Majesty 
 the Emperor of Brazil has supposed to be the work of ma 
 roon negroes or Quilombeiros, but which Burton refers to 
 the Indians, appear to be very common in numerous locali 
 ties on the Lower Sao Francisco and in the northwestern 
 
 * Highlands of the Brazil, Vol. II. Chap. XXVII. 
 
PROVINCE OF BAHIA, INTERIOR. 327 
 
 part of the province of Baliia.* Von Martins speaks of hav 
 ing found, in the immediate vicinity of the locality he de 
 scribes, fragments of Indian pottery, showing that there 
 had anciently been an Indian encampment on the spot. 
 He afterward found rock sculpturings on the banks of the 
 Japura. 
 
 Near the Villa Nova da Rainha is the Serra do Gado 
 Brabo, which is described as a granite hill covered in places 
 with a layer of red clay affording gold. 
 
 From Villa Nova towards Joazeiro the road leads for some 
 six leagues through catinga. The rock for the whole dis 
 tance appears to be gneiss, often with pistacite, and the 
 country is quite even. Granite blocks lie strewn about over 
 the surface, with loose pieces (Fundlinge) of verdegris-col- 
 ored quartz, fibrolite, schorl, and common opal.f 
 
 From Joazeiro, Spix and Martins made an excursion to 
 the Rio de Salitre to visit a locality where salt was extracted. 
 This river flows into the Sao Francisco from the south, about 
 a league to the west of Joazeiro. According to Von Martius, 
 the rock in the vicinity of the town is granite. Going west- 
 southwest towards the Rio de Salitre this rock was soon left, 
 and whitish-yellow dolomite succeeded. Burton says that 
 at its mouth the Rio de Salitre has tall banks, white with 
 
 * Mr. Wallace, in his Travels on the Amazon and Rio Negro, describes sim 
 ilar picture-writings as occurring at Monta Alegre, Scrpa, at several localities 
 on the Rio Negro, and on the Uaupes. 
 
 t The Bahia and Sao Francisco Railroad is to terminate at Joazeiro. Ac 
 cording to the survey of Vivian, Joazeiro is set down as 936 feet above 
 the sea. Halfeld gives the height of the river at this point at 1383 palmas (998 
 feet). 
 
 From a MS. map, furnished me by Mr. Nicolay, showing the line of Vivian s 
 survey, I take the following heights of places along the line : 
 
 Alagoinhas, 300 feet. Agoa Fria, 763 feet. Coite , 1,145 feet. Faz. da Sta. 
 Luzia, 1,106 feet. Queimadas, 888 feet. Faz. da Arueira, 1,997 feet. 
 
328 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the finest limestone. Passing the limestone band, Yon Mar- 
 this describes a band of finely bedded mica slate composed 
 of crystalline quartz grains and white or bright brown mica. 
 A clay slate of a dark green color, highly laminated, and 
 containing crystals of magnetic iron ore or pale flesh red or 
 bluish, with chlorite, occupies a large area along the river. 
 This rock sometimes passes into mica slate. 
 
 Associated with it are greenstone and gray limestone, 
 with garnet, and folia of chlorite, and mica slate. At the 
 Fazenda Aldea, between the hills of the last-named forma 
 tion and the river, Von Martius describes a flat piece of 
 ground, about 60,000 square feet in area, over which the 
 soil is highly impregnated with salt. There are many other 
 similar localities. The soil is alluvial, and deposited by the 
 river. It is ochre yellow in color, and contains more or less 
 of pebbles and vegetable material. After a rain or freshet 
 has covered this soil, and the sun has dried it, an efflorescence 
 of salt appears on the surface. It is not pure, and it ap 
 pears to contain sulphate of lime, chloride of lime, chloride 
 of magnesia, and saltpetre. 
 
 The basin of the Sao Francisco, from the Rio Verde north 
 ward to the Rio de Salitre, is extensively covered by saline 
 deposits, and some of the streams, as, for instance, the Rio 
 Verde,* a river navigable for some distance for canoes, are 
 brackish. Von Martius says : " To the west the mountains 
 withdraw themselves still farther from the stream, and the 
 country consists of a uniform, dry plain, grown up with grass 
 and low bushes. Here one sees, especially in the low places, 
 and particularly after rain, white crusts of salt weather out, 
 and the places where it makes its appearance most abundant- 
 
 * Burton remarks that the fish seemed to be attracted in swarms by the brack 
 ish streams. 
 
PROVINCE OF BAHIA, INTERIOR. 329 
 
 ly (lagoas, salines) are the salt-mines of the inhabitants." 
 These mines lie sometimes at a great distance from the 
 river. The salt is collected by scraping up the crust of 
 salt from the earth, mixing it with water to separate the 
 earthy impurities, and then allowing the salt to crystallize 
 out by the heat of the sun. 
 
 Burton says that sometimes the liquor is " strained in 
 bangues (coppers or hides) evaporated over the fire and al 
 lowed to crystallize." Salt made in this way must of course 
 vary very much in quality, and there is every gradation 
 from almost pure salt to a useless dirty variety bitter with 
 magnesian salts. The origin of the salt of the Sao Fran 
 cisco valley is unknown. So far as we know, no deposits 
 of rock-salt * occur. The amount of salt manufactured on 
 the Sao Francisco is insufficient to meet the demand, and 
 much sea salt finds its way overland by way of Joazeiro. 
 
 Mr. Allen has been kind enough to give me the following 
 note on the salt of the Sao Francisco valley. 
 
 " A saline efflorescence occurs at innumerable localities in 
 the drier portions of the Brazilian plateau, as in other arid 
 districts, but chiefly along the banks of the streams. At 
 Jacard, situated about midway between Chique-Chique and 
 Jacobina, the efflorescence arising in the dry months from 
 the annually overflowed banks or bottom lands of the Riacho 
 do Jacar4 is scraped up at intervals, of course with more or 
 less earth, and the whole leached. The lye thus obtained 
 is placed in small troughs to be evaporated by the sun, by 
 which means a small quantity of impure common salt is ob- 
 
 * According to the Engineer Nesbitt, rock-salt occurs below Chasuta, on the 
 Rio Huallaga, one of the great tributaries of the Amazonas, and he says that 
 the banks of the river for more than a league are pure rock-salt! (Brazil and 
 Brazilians, p. 578.) 
 
330 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tained. At times the amount of other accompanying saline 
 compounds, as sulphate of magnesia, &c., is so great as to 
 render it quite unfit for use. Though the river bottoms, or 
 the overflowed portions of their valleys, aiford the principal 
 sources of native salt, the borders of the half-dried lagoons 
 not unfrequently abound with a similar efflorescence. Such 
 salt licks occur as far south at least as Januaria, where 
 they are numerous. From this fact the city is more com 
 monly known in the neighborhood as Salgado than by the 
 name of Januaria. 
 
 " The banks of the lower portion of the Rio das Yelhas 
 are also remarkable for a similar thick incrustation of what 
 appeared to be nearly pure sulphate of magnesia. This 
 incrustation is often of considerable thickness, appearing 
 not unlike thick hoar-frost. To the presence of these saline 
 impurities in the waters of the streams of many portions of 
 Minas is attributed the great prevalence of the disease 
 known as goitre, that occurs there with such frequency. 
 
 " Most of the streams of Bahia are brackish, at least in 
 the dry season." 
 
 " At Jacard the apparatus I saw in use in the manufac 
 ture of salt was extremely rude, a section of a hollow tree 
 serving for the leaching-tub, and small logs hollowed out 
 for evaporating-vessels. 
 
 Saltpetre, as above remarked, occurs with the salt over a 
 large area in the provinces of Bahia and Minas Geraes ; but 
 in the limestone region of the Sao Francisco valley, where 
 caves are abundant, it is found, as we have already seen, 
 mixed with the earth in the bottom of the caves. Von 
 Martius says that fifteen leagues up the Rio do Salitre there 
 are extensive caverns excavated in limestone and filled with 
 a black earth containing -J^ of saltpetre. The process of 
 
PROVINCE OF BAHIA, INTERIOR. 331 
 
 extraction is very simple. The earth is lixiviated with hot 
 or cold water, and the lye is reduced in strength by evapo 
 ration until the saltpetre crystallizes out.* When salt 
 occurs mixed with the saltpetre the lye obtained by the 
 lixiviation of the earth is first evaporated down sufficiently 
 to allow the salt to crystallize, after which the saltpetre is 
 obtained by further evaporation. 
 
 Saltpetre is quite largely extracted in some parts of the 
 Sao Francisco valley, and on the Rio das Velhas. Burton 
 says that on the Upper Rio das Velhas it sells for 10 $ 000 
 (about $5.00) per arroba. 
 
 From Carunhanha to Urubu the Sao Francisco flows 
 through a flat country, bordered here and there, at a 
 greater or less distance from the river, by isolated hills and 
 ridges. Just below the Rio Carunhanha is the Serra da 
 Lapa, composed of limestone f of a bluish color, and hori 
 zontally stratified. Below Urubu the river valley, accord 
 ing to Burton, becomes more contracted and is bounded by 
 " Serras," which on one side or the other accompany it at 
 a short distance. These serras are masses of horizontally 
 stratified sandstone, with which the valley was doubtless at 
 one time filled, but which has suffered very extensive denu 
 dation. Burton says that below Urubu, with the change in 
 the geological structure of the valley there is ushered in a 
 change in the climate and the vegetation, which Mr. Allen 
 assures me is very marked. The sandstones in many lo 
 calities are described by Burton, Halfeld, and others as ita- 
 columite. The sandstone lying near the surface of the 
 
 * Burton, Vol. II. p. 291. 
 
 t Mr. Allen says : " I well remember the serra of horizontal blue lime 
 stone at and below Urubii mentioned, as you state, by Martins and Burton. 
 The limestone for many miles east of Chique-Chique was lithologically of the 
 same character, but the strata there were inclined. See my report." 
 
332 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tertiary chapadas of the Jequitinhonha basin also resemble 
 itacolumite.* 
 
 Below the Barra do Rio Grande the river valley is de 
 scribed by Burton as broadening and forming a dead flat, 
 which in places, as at Chique-Chique, is covered with 
 patches of blowing sand, reminding one of an African desert. 
 
 A short distance below Chique-Chique, near Tapera da 
 Cima, are heavy deposits of magnetic iron ore, which Bur 
 ton compares with the Itabirite and Jacutinga of Gongo 
 Socco and vicinity. The relation of these deposits to the 
 other rocks he does not give, but Halfeld states that they 
 have a north south course. 
 
 At Pilao Arcado Burton speaks of finding a conglomerate 
 underlaid by soft green shale, traversed by quartz veins. 
 Here gneiss makes its appearance. The Serra do Tomba- 
 dor, near the Ilha Grande do Zabele", Burton describes as 
 composed of magnetic iron ore resting on limestone. The 
 river for a great part of the way between Chique-Chique 
 and Joazeiro is bounded by conical hills and ridges. Below 
 Joazeiro outliers of the great horizontal sandstones, &c., ap 
 pear constantly on both sides of the river, and they con 
 tinue, as we shall see further on, even below the fall of 
 Paulo Affonso. The bottom rocks are gneiss, slates of va 
 rious kinds, limestone, &c., but no competent geologist has 
 been over the country, and its structure has yet to be 
 worked out. 
 
 * Mr. Allen gives me the following note : " Respecting the sandstone occur 
 ring below Urubii, Burton s observations and my own also agree. I mention 
 them in my note as quartzitcs and quartzose sandstone. I could see no 
 difference between them and those observed 200 miles to the eastward at 
 Olhos d Agua and in the Jacobina Taboleiro. Near Jacobina I observed them 
 also disintegrated, forming beds of white quartz sand on the declivities of some 
 of the hills." The term " itacolumite " is very loosely used by travellers, and 
 is applied to compact schistose sandstone as well as to the true itacolumite. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 
 
 CHAPTER VIII. 
 
 PROVINCE OF BAHIA, GEOLOGY OF THE VICINITY OF SAO SAL 
 VADOR AND THE BAHIA AND SAO FRANCISCO RAILROAD. 
 
 Topography of the Vicinity of Sao Salvador da Bahia. The Upper and 
 Lower Cities. The Population, &c. The Harbor. The Commerce of 
 the City and Province. The Climate, &c. The Bahia Steam Navigation 
 Company. The Bahia and Sao Francisco Railroad. The Paraguassii 
 Steam Tram-road. The Gneiss of Bahia. Decomposition. Drift De 
 posits Consolidation of Beaches. Stone Reef at Rio Vermelho. 
 Blown Sands covering the Drift of the Hills. Mr. Allport s Description of 
 the Cretaceous Beds of Monscrrate and Plataforma. Fossil Fishes, Croco 
 diles, &c. Description of several species of Fossil Mollusks. Cretaceous 
 Beds of Plataforma and Vicinity. Prof. Marsh s Notice of the Reptilian 
 Remains Fossil Fishes at Agua Comprida. Gneiss at the Rio Johannes. 
 Tabolciros and Sand Plains of Camassari. Peculiarities of the Topo 
 graphy of the Tertiary Hills. Tabatinga Clay. Sand Plains and Ta- 
 boleiros of the Imbu9ahy. Peat-Bog. Drift. Diamond-washings at 
 Pitanga. Cretaceous Strata at Pojuca. Piassahas. Campos of Ala- 
 goinhas. Tertiary Hills. Character of Vegetation. 
 
 THE sea-coast line going southward doubles sharply north 
 ward on itself on reaching the mouth of the Bahia de Toclos 
 os Santos, forming a sharp peninsula or cape directed south 
 ward and terminating in the point St. Antonio. This point, 
 which is composed of gneiss, is about four miles long, and 
 has a mean elevation of about two hundred feet. On the 
 esaward side the land is hilly, the hills being often hemi 
 spherical or hemi-elliptical, and presenting very remarkably 
 regular outlines. On the coast, as in the case with the Morro 
 do Conselho, and the hills in the vicinity of Rio Vermelho, 
 
334 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 these hills are bare of trees, so that their form is beautifully 
 seen. On the western side, which is occupied by the city 
 of Bahia, this cape is very even in height, and is bounded 
 by a steep, in some cases precipitous slope,* which, with a 
 somewhat zigzag course, is continued many miles to the 
 northward, apparently marking the line of a fracture. 
 
 The city of Sao Salvador da Bahia, usually known as 
 Bahia, stretches along the edge of the bluff for several 
 miles. The cidade alta is irregularly but well built for a 
 South American town, and there are some beautiful resi 
 dences, especially in the southern part of the city. There 
 are many churches, some of which are very fine, together 
 with schools of various kinds. Besides these there is a 
 medical college, a public library of several thousand vol 
 umes, a large theatre, a public garden, and a museum ; but 
 the latter is no credit to the city. 
 
 Below the bluff is the cidade baixa, which occupies a very 
 narrow strip along the bay. This is the business portion, 
 and though it for the most part consists of but one or two 
 streets, it is closely built up with warehouses and stores. 
 Here are the Alfandega, or custom-house, the markets, 
 the marine arsenal, the consular offices, banks, several 
 hotels, &c. It is a hot and busy place. The city stretches 
 along the shore for several miles to the northeast, form 
 ing the suburbs of Monserrate and Itapagipe, the latter 
 known as the head-quarters of the Bahia Steam .Naviga 
 tion Company. The two cities are united by very highly 
 
 * "In 1671, as a result of heavy rains, there occurred a destructive slide 
 from the summit of the bluff, which precipitated a large quantity of earth upon 
 the lower town, destroying houses, burying thirty persons alive, and filling up 
 
 half of the Praia Similar slides have frequently occurred in the history 
 
 of Bahia, notwithstanding the expenditure of immense sums in endeavoring to 
 prevent them." Kidder s Brazil, Vol. II. p. 39. 
 
GEOLOGY OF THE VICINITY OF SlO SALVADOR. 335 
 
 inclined streets, passable with great difficulty by carriages 
 or teams.* 
 
 The whole city numbers from 160,000 f to 180,000 $ 
 inhabitants, principally of Portuguese and negro descent, 
 though there are many foreigners, as at Rio. Quite a num 
 ber of Englishmen are engaged in business there, but Ameri 
 cans are few. The society is almost thoroughly European- 
 ized, and there is much real culture among the people. 
 
 The Bay of Bahia forms one of the best harbors on the 
 Brazilian coast, and, next to that of Rio, it is the most re 
 sorted to. It is, however, so wide, and the entrance is so 
 open, that the ocean-swell rolls in, preventing shipping 
 lying at the quays, which is also the case at Rio. In ordi 
 nary weather the shipping is protected by the high land on 
 which the city stands, but southwest storms cause a heavy 
 swell. The city owes its importance chiefly to its harbor, 
 which eminently fits it to be a port for foreign trade, while 
 it makes the city the centre of the trade of the coast for a 
 long distance north and south, as well as of the interior. 
 The products of the Sao Francisco and of Sergipe find their 
 way for the most part to Bahia. The borders of the bay 
 itself, or the Reconcavo, so called, are highly cultivated, and 
 produce much sugar, tobacco, piassaba fibre, &c. The princi 
 pal product of the interior is cotton, which is cultivated to a 
 
 * A street with a moderate grade to connect the two towns was in process 
 of building in 1867. By this time it is probably completed. Dapper, in his 
 America, published in 1673, gives a very curious and interesting copperplate 
 engraving of the city. He represents the upper city as built on a plain, sur 
 rounded behind by a narrow crescent-shaped lake, and with high mountains in 
 the background. Two inclined planes for carriages elevated by machinery 
 are seen uniting the upper and lower towns. 
 
 t Pompe o, Geographic/, 1864. 
 
 J Dr. Candido Mendes de Almeida, Atlas, 1868. 
 
336 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 considerable extent in the less arid regions. Cattle are 
 raised in great quantities, and large herds are sent into 
 the province from the country west and north. Perhaps 
 no better idea of the commerce of Bahia could be afforded 
 than by giving the official tables of the exports during the 
 year 1864 65 * : 
 
 Articles. Official Values. 
 
 Agoardente 3 72:813$! 20 
 
 Cotton 1.303:277$553 
 
 Sugar 6.316:627$583 
 
 Cacau 173:225$356 
 
 Coffee 1.614:063$450 
 
 Cigars 45:839$000 
 
 Hides 356:008$300 
 
 Diamonds 1.381 :500$000 
 
 Other Articles 81:029$049 
 
 Tobacco 2.060:833$745 
 
 Woods 237:2660997 
 
 Piassaba 141:437$653 
 
 The cacau comes principally from Ilhe os, and the coffee 
 from the Colonia Leopoldina. Large quantities of good 
 cigars are made in Bahia, and are sold at a very low price. 
 At Cachoeira, or rather at Sao Felix, opposite, is a very 
 large manufactory of cigars (charutos) and cigaritos. The 
 latter are much esteemed in Brazil. The hides come from 
 the interior of the province and from the Sao Francisco. 
 
 I have in preceding pages called attention to the whale- 
 fishery of the port. 
 
 The following table will show the value of national pro- 
 
 * Relatorio of the President, the Commendador Manoel Pinto de Souza 
 Dantas, Bahia, 1866, 2, Quadro No. 11. 
 
GEOLOGY OF THE VICINITY OF SlO SALVADOR. 337 
 
 ducts exported to other provinces during the latter half of 
 the year 1865 : 
 
 Ports. Values. 
 
 Alagoas 1 15:05 2$949 
 
 Ceara 26:062$318 
 
 Espirito Santo .... 9:545$877 
 
 Maranhao 27:812$877 
 
 Pernambuco 1.048:050$020 
 
 Para 60:008$970 
 
 Parahyba 1:485$500 
 
 Rio de Janeiro 368:492$986 
 
 Rio Grande do Sul . . . . 212:865$511 
 
 Sergipe 150:930$837 
 
 2,020:307$845 
 
 The climate of Bahia is hot, but not unhealthy.* In the 
 lower city, where the streets are narrow and protected from 
 the sea breeze by the high ground behind, it is very warm 
 and uncomfortable, but I have never suffered so much in 
 the lower town from the heat as I have in early summer in 
 New York. In the upper city the climate is exceedingly 
 pleasant and healthy for a tropical city. Bahia has suf 
 fered at times from yellow-fever,f but it has been for many 
 years remarkably healthy. 
 
 Bahia is connected with Europe by two lines of steam 
 ships, and with New York by the Brazilian mail steamers. 
 It is the head-quarters of the Bahia Steam Navigation Com- 
 
 * For several years past Dr. Antonio de Lacerda has kept a journal of 
 meteorological observations for M. Arago, but I do not know that they have 
 yet been published. The climate of Bahia appears to be more moist than that 
 of Rio, and rains are more frequent, being distributed through the entire year. 
 
 t Pompe o, writing in 1864, says (Geographic., p. 449): "Until 1849 Bahia 
 was sufficiently healthy ; but since that time the yellow-fever became almost 
 endemic, attacking with preference the Europeans." Pompeo is not quite ac 
 curate here. 
 
 VOL. I. 15 T 
 
338 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 pany, under the able direction of my esteemed friend, Mr. 
 Hugh Wilson. This company has at present a fleet of six 
 teen steamers. Of these one runs regularly between the 
 city and Maceid, in the north, calling at Penedo, and at 
 various ports in the province of Sergipe. Regular steam 
 navigation has been established 011 the Sao Francisco, below 
 the Porto das Piranhas. Several steamers of the company 
 are engaged in running between Bahia, Cachoeira, St. Amaro, 
 and various other points in the Reconcavo, and there is a 
 regular line between Bahia and Caravellas, or the Colonia. 
 
 A railroad has been commenced to Joazeiro, but after 
 extending ninety miles in a northerly direction from Bahia,, 
 it ends in a desert sand-plain at Alagoinhas, as we shall see, 
 and there is no prospect of its ending anywhere else for 
 some years to come. 
 
 From Cachoeira to Urubu, on the Sao Francisco, a steam 
 tram-road has been projected. It was formally opened in 
 July, 1867, but I have not learned what progress it has 
 made up to this time.* The road will run up the Para- 
 guassu valley, through the diamond region, with side 
 branches to Feira de Sta. Anna and Lenc,6es. 
 
 This railroad has a future before it, as it is to run through 
 a much more important tract of country than the Bahia 
 and Sao Francisco Railroad. 
 
 The gneiss composing the point on which Bahia is built 
 is composed of a very compact, sometimes trap-like variety 
 of gneiss, not infrequently without mica and with very in 
 distinct planes of stratification, as has already been observed 
 by Allport f and Darwin. The latter has given the following 
 description of the Bahian gneiss : ^ 
 
 * Major James informs me that it is proposed to make this a railroad. 
 t Quarterly Journal Geological Society, Vol. XVI. Part 3, p. 263. 
 t Geological Observations, p. 140. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 339 
 
 " The prevailing rock is gneiss, often passing, by the disappear 
 ance of the quartz and mica, and by the feldspar losing its red 
 color, into a brilliant gray primitive greenstone. Not unfre- 
 quently quartz and hornblende are arranged in layers in almost 
 amorphous feldspar. There is some fine-grained sienitic granite, 
 orbicularly marked by ferruginous lines, and weathering into 
 vertical, cylindrical holes almost touching each other. In the 
 gneiss, concretions of granular feldspar, and others of garnets with 
 mica, occui . The gneiss is traversed by numerous dikes composed 
 of black, finely crystallized, hornblendic rock, containing a little 
 glassy feldspar and sometimes mica, and varying in thickness from 
 mere threads to ten feet ; these threads, which are often curvi 
 linear, could sometimes be traced running into the larger dikes. 
 One of these dikes was remarkable from having been in two or 
 three places laterally disjointed, with unbroken gneiss interposed 
 between the broken ends, and in one part with a portion of the 
 gneiss driven, apparently whilst in a softened state, into its side 
 or wall. In several neighboring places, the gneiss included angu 
 lar, well-defined, sometimes bent masses of hornblende rock, 
 qxiite like, except in being more perfectly crystallized, that forming 
 the dikes, and, at least in one instance, containing (as determined 
 by Professor Miller) augite as well as hornblende. 
 
 ". . . . The folia of the gneiss within a few miles round Bahia 
 generally strike irregularly, and are often curvilinear, dipping 
 in all directions at various angles ; but where best denned, they 
 extended most frequently in a N. E. by N. (or East 50 N.) and 
 S. W. by S. line, corresponding nearly with the coast line north 
 wards of the bay. I may add that Mr. Gardner found in several 
 parts of the province of Ceara, which lies between four hundred 
 and five hundred miles north of Bahia, gneiss with the folia ex 
 tending E. 45 N. ; and in Guyana, according to Sir R. Schom- 
 burgk, the same rock strikes E. 57 N. Again, Humboldt de 
 scribes the gneiss granite over an immense area in Venezuela and 
 even in Colombia, as striking E. 50 N., and dipping to the N. W. 
 
340 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 at an angle of fifty degrees. Hence all the observations hitherto 
 made tend to show that the gneissic rocks over the whole of this 
 part of the continent have their folia extending generally within 
 almost a point of the compass of the same direction." 
 
 On the surface the gneiss is in some places decomposed 
 to a great depth, and so soft as to be easily removed by a 
 spade or mattock. It is well exposed in the numerous cut 
 tings on the streets in different parts of the city and vicin 
 ity. On the Ladeira do Bom Gosto, on the southern side 
 ascending the hill towards the cemetery, is a very heavy ex 
 cavation, in which the decomposed gneiss in situ is beauti 
 fully shown, and where, at the time of my visit, workmen were 
 cutting it away with hoes. The decomposition consists of 
 a loss of alkalies on the part of the feldspar, reducing it to 
 the state of kaolin. The iron of the rock is oxidized and 
 stains it a deep brick-red, though the color is rarely uni 
 form, but likely to be in streaks, portions of the rock being 
 white. Between the lighthouse at the barra and the first 
 little point above to the westward, the rock is the dark 
 gneiss above described ; it is well bedded, but the stratifi 
 cation is much disturbed. The strike here is N. 60 E. Dip 
 N. W. 35. Nearer the lighthouse the rocks become much 
 distorted ; there are here several large veins traversing the 
 rock ; one trap vein about nine feet wide has a N. S. direc 
 tion, while a granite vein I observed runs N. E., S. W. 
 
 The lighthouse is built on a rock which is joined to the 
 mainland by a narrow and low isthmus. The road which 
 skirts the shore crosses this isthmus, making a considerable 
 excavation necessary, while the road branching off to go to 
 the lighthouse runs through a similar cutting. The greatest 
 thickness thus exposed is twelve feet. The upper part, 
 c c c, consists of the ordinary red earth one finds every- 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 3-41 
 
 where spread over the surface of the hills at Bahia. In 
 it are seen occasional loose irregular masses of quartz and 
 Bahian gneiss, the latter decomposed. The under part, b b, 
 consists of an irregular sheet of boulders of Bahian gneiss 
 more or less decomposed, sometimes a foot and a half in 
 diameter, in some cases angular, in others rounded, to 
 gether with others of that peculiar friable quartz so com 
 mon in the quartz veins of the Bahian peninsula, also either 
 angular or rounded milky quartz pebbles, of all sizes, the 
 
 whole being thrown together, without any arrangement, in a 
 confused mass. The greatest thickness of the boulder bed 
 exposed is about four feet, but it thins out at a to only six 
 inches, and is seen to lie immediately upon the decomposed 
 gneiss. 
 
 This is the most remarkable gravel deposit which I have 
 seen on the coast. According to my estimation, the height 
 of the bottom of this sheet above sea level is twenty to 
 twenty-five feet. 
 
 On the eastern side of the lighthouse the gneiss is much 
 broken up and faulted. There are here some interesting 
 granite veins, in which the materials are seen to be very 
 coarsely crystallized on the side of the fissure, while the 
 crystals of the middle of the vein are very small. Near 
 this place I found gneiss of a light color, full of garnets, 
 almost vertical, and with a strike of N. 25 E. 
 
 Going up the beach and passing the first little point, one 
 observes a sheet of a soft dark brownish or yellowish sand- 
 
342 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 stone lying against the hillside, and extending from near low- 
 tide mark to a height of some fifteen to twenty feet. This 
 sand has been evidently washed or blown up from the shore 
 over the hillside. The grains of sand in this rock are very 
 fine and uniform, and the Avhole is loosely cemented by 
 oxide of iron. 
 
 I had observed that the red drift coating in no place came 
 down to the surface of the sea, but that it was invariably 
 worn away up to about that level to which the waves must 
 have reached at the time when the late rise of the land 
 began. Below that line the hillsides were bare and the 
 rock surface broken and irregular. I found that the sands 
 just described overlapped the drift and in part protected it, 
 but the waves had washed away a portion of it, so that the 
 drift is seen underlying the sands and extending to below 
 half-tide mark, showing, I think, conclusively that the land 
 was more elevated when the drift was deposited than it is 
 at present. 
 
 The seaward sides of several gneiss hills are bared and 
 much broken up by the waves, which beat heavily on this 
 coast. It is interesting to observe the exceedingly irregu 
 lar surfaces developed by this action, and how the harder 
 rocks become prominent while the softer are washed away. 
 This gneiss is of the same general character as that at 
 the Barra, though it varies much in color and general ap 
 pearance. 
 
 At the lighthouse, and especially to the eastward, the 
 beach is partially solidified by the cementing together of its 
 materials by the lime derived from shells. The shingle of 
 the beach and the sand and gravel packed away in cracks 
 in the rocks become cemented solidly together. Where the 
 beach is composed of sand, it is converted into a very hard 
 
GEOLOGY OF THE VICINITY OF SiO SALVADOR. 343 
 
 sandstone up to a level considerably above half-tide, and in 
 this sandstone the structure and irregular bedding of the 
 beach are beautifully preserved. Very commonly, after this 
 solidification has taken place to some depth, the newly made 
 rock is washed bare, and broken up by the waves. In this 
 case it forms detached masses, sometimes an extensive 
 sheet of sandstone, which appears to crop out on the beach 
 with a seaward dip of a few degrees, whatever the trend 
 of the beach may be. Masses of this kind we find at inter 
 vals on the shore between Bom Fim and Bahia, and on the 
 coast thence to Rio Vermelho. The sandstone of this reef 
 or solidified beach often contains layers of coarse materials 
 as well as an abundance of shells, the latter usually broken, 
 but sometimes perfectly preserved, and with their colors 
 quite fresh. 
 
 It is very interesting to see how this consolidation goes 
 on. On a long sand beach one may trace it, sometimes, in 
 every stage of progress. It frequently happens that, owing 
 probably first to a heavy storm, a ridge is thrown up behind 
 the beach. This opposes a barrier to the sea for a long time 
 thereafter, and may increase in height and extent from the 
 drift of the sand by the wind, which is very likely to take 
 place, especially when the shores have a northerly trend, on 
 account of the prevalence of northeast winds. Along this 
 line shells are thrown up in great quantities and buried 
 in the sands. These shells become dissolved by the water 
 soaking through the sands, whether salt or fresh, and the 
 carbonate of lime thus derived is deposited as a cement 
 to the materials of the beach to a height somewhat above 
 half-tide, rarely ever to high-tide mark. The copious rains 
 must tend notably to this result, and the waters from marsh 
 lands soaking through the beach must also assist in the 
 
344 GEOLOGY AND PHYSICAL GEOGRAPHY/ 
 
 solidifying process. Sometimes, after the solidification of 
 the lower part of the beach, the loose parts are swept away 
 by a storm, it may be, sometimes, by the bursting of the 
 barrier by the freshet of a river, which has inundated the 
 low grounds behind. In this case the beach is left stand 
 ing like a wall running in the water parallel to the coast. 
 Such a reef is seen at Rio Vermelho, where, after skirting 
 the shore for some distance, it projects partially across the 
 mouth of a little bay, like a wall or breakwater of rock. 
 
 The reef at Rio Vermelho illustrates very well the general 
 character of these consolidated beaches. It is composed of 
 layers of calcareous sandstone and conglomerate, often some 
 what irregular, dipping seaward, the dip being only a few 
 degrees, or about that of an ordinary sand beach. The height 
 of the reef is very uniform. In thp finished and isolated 
 reefs, as that of Pernambuco and the one under considera 
 tion, the recent rise of the land has brought this level some 
 what above that of the sea. 
 
 The solidified portion is seen to be but a sheet of varying 
 thickness lying on the surface of the beach. On the inner 
 
 MigliWater. 
 
 ItoarWccte 
 
 side it is quite thin, and from the action of water behind it 
 is undermined and broken off, until, at last, it forms a low 
 perpendicular wall, undermined below and sometimes pro 
 jecting several feet. Usually this side of the reef is flanked 
 by a slope of sand or mud, and sometimes by large oblong 
 blocks of the sandstone. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 345 
 
 The surface of the reef is, broadly speaking, horizontal, 
 but it is marked by longitudinal ridges and much worn 
 away, sometimes honeycombed and exceedingly rough, with 
 large, shallow or deep, irregular pools of water, the homes 
 of several species of corals, <fcc. The whole mass is divided 
 by joints into great blocks. On the seaward edge the reef 
 is often worn away by the waves and undermined, present 
 ing always a perpendicular wall to the sea. The upper bed 
 almost always projects a little, and great blocks broken 
 from it lie in front, which afford some protection to the reef. 
 One often finds a depth of twelve to fifteen feet or more at 
 low tide along this side of the reef. In most cases corals 
 grow on the faces of these reefs below low water. 
 
 The reef rock is quarried quite largely at Bahia, as well 
 as at Pernambuco, for building purposes, and in both cities 
 it has been used for flagging. In the sidewalks of the 
 upper town at Bahia one may see it in large blocks which 
 are full of shells of species now common on the coast. 
 
 Going northward beyond the Morro do Conselho, the 
 gneiss hills recede more and more from the coast, and low 
 lands come in between them and the sea. These I have 
 examined for a short distance north of the Morro do Con 
 selho. The coast is fringed with patches and strips of sands 
 which, in part forming dry plains with a scanty vegetation, 
 in part swampy, do not rise to a greater height, as a general 
 rule, than eight to ten feet above the sea. There are many 
 patches of this kind near Bahia, and some are of considerable 
 extent. On the sea-coast they occupy several little bays be 
 tween the hills. As we go northward beyond the Morro do 
 Conselho these sands become wider in extent and occupy 
 large tracts. The sands on this coast are very fine, and 
 are easily raised and carried by the winds, so that exten- 
 
 15* 
 
346 . GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 sive sand-hills are formed, and the shores of Itapuan, a few 
 leagues north of Bahia, are white as snow. 
 
 In an excursion made among the dome-shaped gneiss hills 
 to the east of the city of Bahia, and bordering the sands, I 
 observed that the light, white sand was blown up on the 
 seaward side of some of the hills over the rich red drift, so 
 that one half of a hill is snow white and the other brick red. 
 These hills arc cultivated in places, and it was very interest 
 ing to see a red field set in a framework of white sands, the 
 drift soil being turned up in the process of tilling. 
 
 The city of Bahia, as already described, is built in part 
 on a low strip of land, which, beginning some two miles 
 above the lighthouse on the bay side, accompanies the 
 shore for a mile or two farther on, when it sweeps round 
 westward, separating itself from the gneiss ridge, forming 
 a long, irregular tongue of land occupied by the suburbs 
 Jequetaia, Bom Fim, and Itapagipe. 
 
 From Jequetaia to Bom Fim this tongue of land is flat 
 and composed of recent sands. At Bom Fim and Monser- 
 rate the land is higher and more irregular, the height being 
 eighty to one hundred feet. It is composed of cretaceous 
 strata. Allport says : * 
 
 " The rocky cliff forming the southwest of the hill, on 
 which the fort of Monserrat is built, presents to view sev 
 eral alternations of conglomerate, sandstone, and shale. 
 Towards the northeast these beds pass into a gritty shale 
 of a bluish-gray color, and full of pebbles ; the latter 
 gradually disappear, and the upper strata, as far as the sea 
 ward exposed portion extends, consists of beds of shale, al 
 ternating with bands of sandstone, both of which contain 
 the same species of fossil shells. The entire series of these 
 
 * Quarterly Journal Geological Society, Vol. XVI. Tart 3, p. 263. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 347 
 
 deposits is covered with the usual red loam, and have the 
 general inclination to the northwest. 
 
 " The seaward exposed position of the cliff of Monserrat, 
 about thirty feet in height, consists chiefly of conglomerate, 
 with irregular wedge-shaped bands of shale and clay, and 
 also bands of sandstone. The conglomerate is composed of 
 more or less rounded pebbles of gneiss, granite, quartz, and 
 other crystalline rocks, and occasionally of sandstone, the 
 whole forming an extremely hard rock. The pebbles vary 
 in size from the finest gravel to large boulders. 
 
 " In the shale, near the base of the cliff, were found the 
 fossils about to be noticed, consisting chiefly of scales and 
 other portions of fish, bones and teeth of saurians, together 
 with lignite, a few Mollusca and some Entomostraca. 
 
 " Two miles from the above hill, in a northeast direc 
 tion, is the Plataforma, another hill of the same formation, 
 but loftier ; the conglomerates and shales have here the 
 same lithological character, and in the latter are found sev 
 eral fossils similar to those found at Monserrat. 
 
 " The geological position of the above formation is unde 
 termined, as they have not been traced in connection with 
 other deposits ; but a probable inference may, perhaps, be 
 made from an examination of the fossil remains. 
 
 " With regard to the fish remains, Sir P. Egerton, Bart., 
 F. G. S., to whom the specimens have been submitted, 
 states that " the scales are those of Lepidotus. The spe 
 cies appear to be a new one. The nearest approach to it 
 is an undescribed species from the lithographic stone of 
 Pappenheim [middle oolite]. 
 
 " Numerous fish-bones were found associated with the 
 scales ; and probably the greater portion belong to Levido- 
 tus also. But these and the crocodilian teeth and bones, 
 
348 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 which are also common in these clays from Monserrat and 
 Plataforma, have not yet been systematically examined. 
 Professor Owen, on a cursory view of the large vertebra, 
 figured in PL XVII.,* suggested that it would be a dorsal 
 vertebra of a Dinosaurian reptile allied to the Mcg-alo- 
 saurus" 
 
 At Monserrat Mr. Allport collected a Melania described 
 by Morris as Melania terebriformis, and species of Unio, 
 Paludina, Neritina (?), together with seven species of cyp- 
 rids, enumerated and in part described by Jones, namely, 
 Cypris (?) concultata, Candona Candida Mull, Cypris (?) 
 Monserratensis, Cypris (?) Allportiana, Cypris sp. non de- 
 script. 
 
 I was unable to visit the locality at the fort, but I ex 
 amined the beds quite carefully near the Pedra Furada. 
 On the shore, south of this last-named locality, are exposed 
 thick beds of sandstone, shale, conglomerate, and limestone, 
 the continuation of the same series as Allport describes. 
 This limestone is of a compact texture, and mottled with 
 brown, gray, and green. It contains some sand, together 
 with little pebbles of gneiss and quartz, generally angular. 
 Freshly broken, this rock shows the fossils very indistinctly, 
 but on the weathered surfaces they stand out in fine relief. 
 At the same time the surface becomes granular, as if it 
 were composed of coarse, round, or oval grains of sand. 
 These are, however, calcareous, and the structure may be 
 oolitic, though the grains look as though they had been 
 rounded mechanically. 
 
 Associated with this rock is a fine-textured argillaceous, 
 light slate-colored limestone, in which there is no trace of 
 this oolitic (?) structure, but instead one finds occasionally 
 
 * Quarterly Journal, loc. oil. 
 
GEOLOGY OF THE VICINITY OF SlO SALVADOR. 349 
 
 the cone-in-cone structure beautifully developed, a some 
 what unusual circumstance in rocks of this age. In these 
 rocks, which, at the most, form a bed but two feet thick, 
 several species of shells occur in immense numbers. Of 
 these Melania terebriformis and Paludina ( Vivipara) La- 
 cerdce sp. nov. are the most abundant. Species of Unio, 
 Planorbis, teeth and bones of crocodiles, and bones of 
 Dinosaurian reptiles are also to be found here, together 
 with scales and bones of Lepidotus and other fishes. 
 
 This limestone with its fossil shells resembles very closely 
 specimens of fresh-water fossiliferous limestones from the 
 Weald of England. 
 
 There are also thick beds of shale, in places black and 
 finely laminated, but for the great part not Avell laminated 
 and very soft, of a very light color, and full of little flakes 
 of mica. In this shale are to be found layers abounding 
 in entomostracan remains, of which the most interesting is 
 an esthcrian with its valves marked with concentric ridges 
 like an Astarte, and apparently new. Fish scales and skele 
 tons are not uncommon. 
 
 There are some heavy beds of sandstone in this section, 
 which are seen near the Pedra Furada. This sandstone is 
 fine-textured, soft, and of a light, greenish-gray tint. It is 
 a rare thing that it affords fossil remains, and these are 
 usually carbonized stems of plants. Von Martius speaks of 
 rich beds of Blatter- Kohl near the mouth of Itapagipe and 
 of a brown coal in the sandstone near the city of Bahia. 
 This last is said to have been worked for a short time by 
 the direction of the government, by Feldner, a German, 
 early in this century. I saw nothing to indicate the exist 
 ence of coal or lignite at Bahia, and I do not believe that 
 any deposit of the slightest value exists there.* 
 
 * Gardner says that Dr. Parigot found the Bahian coal to be lignite. 
 
350 GEOLOGY AND PHYSICAL GEOGRAPHY- 
 
 The Monserrate fossils clearly indicate a fresh-water origin 
 for the beds in which they occur. The following are the 
 mollusks thus far obtained from this locality : 
 
 Melania terebriformis Morris, Geological Joiirnal, Vol. XVI. p. 266, 
 PL 16, Figs. 3 a, 3 b, 3 c. 
 
 " Shell subulate, consisting of 7 8 flattened whorls, marked 
 with numerous oblique, somewhat prominent, rounded ribs, which 
 are in some specimens stronger towards the anterior part of the 
 shell. In some individuals the posterior part of each whorl is 
 slightly raised, making the suture more distinct. The last whorl 
 is somewhat constricted. The aperture is ovate. The lip of the 
 columella is somewhat thickened and reflexed." Length 18 mm. 
 Mr. Morris gives a figure of a smooth variety, with- a less cylin 
 drical shell. The top of the spire of the common forms of M. 
 terebriformis is rarely perfect in adult specimens. The younger 
 shells are sharp-pointed, and might be mistaken as belonging to 
 another species. 
 
 The test is usually well preserved, and the shells are sometimes 
 hollow. 
 
 Locality, Monserrate, and vicinity, Bahia. Age, Lower Cretaceous. 
 Collectors, Allport, Nicolay, Williams, Lacerda, and C. F. H. 
 
 Melania Nicolayana, sp. nov. 
 
 Shell minute, subulate, nearly cylindrical, with six or more 
 whorls which are flatter and more oblique than in M. terebriformis, 
 Sutures distinct. Surface smooth. Length 2 mm. 
 
 This form occurs associated with the preceding, but it is appar 
 ently a very rare shell. I dedicate it to the Rev. Mr. Nicolay, to 
 whom I am under many obligations. 
 
 Locality, same as preceding. Collector, C. F. H. 
 
 Vivipara (Pahidina) Lacerdce, sp. nov. Geological Journal, Vol. XVI. 
 
 Plate 14, Fig. 2. 
 Shell about 20 mm. in length, ovate conical, usually with four 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 351 
 
 very ventricose whorls separated by deep sutures, subumbilicate, 
 smooth, shell rather thick. 
 
 This species is figured by Morris as above, and is referred to by 
 him in a note to Mr. Allport s paper as having " a smooth shell 
 subumbilicate, and showing four ventricose whorls, deeply sutured." 
 This shell appears sometimes to have had a truncated spine like 
 our modern V. excisa. It is exceedingly common in the limestones 
 at Monserrate, and is occasionally found in the shales. I have 
 never seen it at Plataforma. I take much pleasure in associating 
 with the species the name of my esteemed friend Dr. Antonio de 
 Lacerda. 
 
 Collectors, Allport, Nicolay, Lacerda, C. F. H. 
 
 Vivipara (Paludina) Williamsii, sp. nov. 
 
 Shell smaller than that of V. Lacerdre, from which it is also dis 
 tinguished by being more conical and having whorls much less 
 ventricose and shouldered. Length 9-10 mm. 
 
 Quite common with the preceding species at Monserrate. 
 
 Collectors, Allport, Williams, Nicolay, C. F. H. 
 
 Plo.norbis Monserratensis, sp. nov. 
 
 Shell minute, flat above, concave below, whorls two and a half 
 in number, flattened from above, rounded, increasing rapidly in 
 width, and apparently flaring a little near the mouth, which is 
 oblong, wider than high. Surface smooth. 
 
 This pretty little shell is quite common associated with the 
 Vivipara, Melania, &c. at Monserrate. In the conglomerate at 
 Plataforma, in which the reptilian bones occur, there is a form 
 which much resembles this, but the spiral seems more open and 
 the shell is oblique in outline, apparently from the rapid widening 
 of the body whorl. This last may be a species of Valvata. The 
 width of the shell of the species just described is about 2 mm. 
 Collector, C. F. H. 
 
 Unio (Anodon ?) Totium-Sanctorum, sp. nov. 
 Shell small, ovate-elongate, compressed, wider behind than be- 
 
352 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 fore, the ends and lower margin forming a very regular curve, 
 which is somewhat straightened below. Hinge line much shorter 
 than shell. Umbo quite prominent, strongly and broadly flat 
 tened by a wide and shallow but well-marked depression which 
 runs downward but obliquely backward across the valves, giving 
 to the shell the appearance of having two rather prominent but 
 wide ridges extending from each side of the umbo, one obliquely 
 forwards widening and growing lower toward the margin, the other 
 much more obliquely backwards with the same characters. The 
 shell is rather thick, and is marked on the outside by numerous 
 fine, concentric lines or wrinkles. 
 
 Collectors, Allport, Nicolay, and C. F. H. 
 
 During my last visit to Bahia I undertook a careful and 
 systematic examination of the rocks exposed in the cut 
 tings along the Bahia and Sao Francisco Railroad,* and I 
 walked over the greater part of the line to Alagoinhas, ex 
 tending my observations for several miles beyond among the 
 taboleiros. 
 
 The observations made on this excursion I give in detail, 
 taking them with little change from my field-book. 
 
 The Bahia and Sao Francisco Railroad, leaving the station 
 at Jequetaia, follows the low sandy ground skirting the 
 gneiss hills, and runs along the northern side of the bay of 
 Itapagipe. At a distance of about two miles from the sta 
 tion at Jequetaia (Calgada), on passing a little gully, there 
 are seen exposed in a cutting beds of a loose-grained sand 
 stone and a shale, the latter being of a greenish-gray tint, 
 and very finely laminated, but too much decomposed on the 
 surface to offer anything of much interest. These beds 
 
 * I have to thank Mr. Howry, the superintendent of the railroad, for a 
 pass over the line, and Messrs. Tiplady, Turner, and Orecchi for many favors 
 received during the excursion. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 353 
 
 have a strike of about northeast, as near as I could make 
 out, and a dip of a very few degrees to the northwest. 
 
 Going on towards the long bridge, we meet no beds well 
 exposed in the railway cuttings, but in a quite deep one only 
 a short distance from the bridge we find thick beds of finely 
 laminated, dark-colored shale, much decomposed, in which 
 are fossil fishes like those at Monserrate, together with a 
 great abundance of cyprids. 
 
 Crossing the long bridge, the land on the opposite side 
 becomes much higher and more irregular, and there are 
 some heavy cuttings, in which, as well as on the sea-shore 
 at low tide, the rocks are very well exposed. These consist 
 of alternate beds of conglomerate and dark shale, with 
 occasional layers of sandstone, which have an average strike 
 of N. 60 E. and a dip of 30 N. W. 
 
 The surface of the beds of shale, exposed on the shore be 
 tween tides, is very much burrowed into by a marine worm, 
 which excavates a tube that enters the shale perpendicularly 
 to the depth of an inch or more, and then, bending round 
 rather abruptly, comes to the surface again. A little crab 
 also burrows into the shale in places, making deep tubes, 
 whose sides are corrugated after a manner that makes them 
 resemble the empty moulds of fossil stems of plants or 
 corals, and they are well calculated to puzzle an observer, 
 who did not know how they were formed. 
 
 The decomposed shale has been used in the works of the 
 Bahia Steam Navigation Company at Itapagipe as a fire 
 clay, and Mr. Ford, the chief engineer, to whom, by the 
 way, I am indebted for many kindnesses, assured me that 
 it answered well for that purpose. 
 
 In the cutting at Plataforma, which is a very heavy one, 
 there is seen a thick bed of shale with occasional bands of 
 
354 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 sandstone, affording fossil fish, and cyprids, over which lies 
 a heavy bed of conglomerate. This is composed of frag 
 ments of the principal rocks of the neighborhood, presenting 
 the same appearance as that of the beds whence they were 
 derived. The pebbles of this conglomerate, mostly of 
 quartz and gneiss, are of all sizes, even up to eighteen or 
 twenty inches in diameter. They are but slightly rounded, 
 and are more or less angular. It appears to be a deposit 
 rapidly accumulated, and from its hardness it now forms 
 quite a high ridge. This conglomerate is a massive bed, 
 showing in itself few traces of stratification. It has afforded 
 a few reptilian bones. The same bed makes its appearance 
 on the shore at low tide, where it may be examined. All 
 these beds are well displayed on the shore, the whole dis 
 tance from Plataforma to the little bay of Periperi, and, as 
 the strike of the beds is tangcntal to the curve of the shore 
 between these two places, and their dip is away from the 
 shore, they may be examined both in ascending and de 
 scending series, and over a considerable area. Above 
 water-mark disintegration makes their examination very 
 unsatisfactory. Below that line it has not obtained to so 
 large an extent, and, save a thin coating of half-decomposed 
 material on the surface, the rocks are nearly in their nat 
 ural state. 
 
 About a half-mile from Plataforma, and where the rail 
 road passes close to the water s edge, there is exposed in one 
 place a section like that on the opposite page, of which the 
 height is about ten feet. 
 
 In many of the other beds, especially in those which arc 
 finer in character, reptilian and other remains are quite abun 
 dant ; but, owing to the compactness of the rock, they are 
 difficult to extract. These fossils consist in spines of fish, 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 
 
 355 
 
 Shale. 
 
 High-water level. 
 5 Shale. 
 
 with occasionally a vertebra, bones of dinosaurians, and a 
 few teeth, both of reptiles and fishes. Of these reptiles 
 teeth the most interesting are those of crocodiles. These 
 vertebrate remains are especially abundant in a calcareous 
 conglomerate, forming a bed a foot or two in thickness, and 
 which I have designated in the above section as a bone-bed. 
 
 There are a few beds of a coarse arenaceous shale, which 
 afford some very perfect specimens of a Teliostean fish, 
 apparently different from the species found at Monserrate.* 
 It is worthy of remark that this shale, as well as some of 
 the other rocks of this series, are bituminous as well as cal 
 careous. 
 
 The reptilian remains from Plataforma and Monserrate I 
 placed in the hands of Professor 0. C. Marsh of Yale Col 
 lege for description. The following with a few unimportant 
 omissions is a notice of them published by him in the Amer 
 ican Journal of Science and Arts, Vol. XL VII., May, 
 1869 : - 
 
 "The most interesting of the reptilian remains collected by Pro- 
 
 * Several fragments of the jaws of fishes have been found. One of these, 
 the right ramus of the lower jaw of a little Pisodus, had the minute shot- 
 like teeth preserved. All the fish remains from this locality are in the hands 
 of Professor Agassiz. I regret that his illness prevents him from furnishing 
 descriptions of them for this work. 
 
356 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 fessor Hartt in the Bahia deposit is the tooth of a large Croco 
 dilian, from the arenaceous shale near Plataforma station, on the 
 Bahia railroad. This specimen is in an excellent state of preser 
 vation, and indicates a species new to science. It is larger, more 
 slender, and more pointed than the teeth of existing crocodiles, 
 resembling most nearly those of some extinct American species. 
 It is conical in form, round at the base, and slightly compressed 
 at the apex. The crown is two inches and three lines in length, 
 along the outer side, and ten lines in diameter at the base. One 
 edge is somewhat more convex than the other, and this is also 
 true of one of the sides, and hence the tooth appears slightly 
 curved in two directions. On either edge of the crown there is 
 a sharp ridge, most prominent near the apex, over which it passes, 
 but gradually disappearing before reaching the base, resembling in 
 this respect the teeth of Thoracosaurus, from which, however, this 
 specimen differs in being longer, and less curved than the teeth of 
 that genus usually are. The sides of the crown are covered with 
 fine, interrupted, undulating striae, which appear to be different 
 from the dental sculpture of the Crocodilia hitherto described. 
 These striae are most distinct near the middle of the tooth, be 
 coming much more delicate at the base, and nearly obliterated at 
 the apex. 
 
 " In size and general appearance, this specimen resembles some 
 what the teeth of Crocodilus antiquus Leidy, from the Miocene of 
 Virginia, but differs from that species in being less tapering, and 
 in having the ridge on the edges extend farther downward. It re 
 sembles still more closely the teeth of a new species of crocodile 
 discovered by the writer at Squankum, N. J., in the tertiary 
 green-sand, which will soon be more fully described under the 
 name Thecocampsa Squankensis Marsh. Both species have essen 
 tially the same proportions, and similar dental striae, but the cut 
 ting ridge of the New Jersey specimens is more prominent, and 
 extends nearly or quite to the base of the crown. The two species 
 were apparently about the same size, both being considerably 
 larger than existing Crocodilians. 
 
GEOLOGY OF THE VICINITY OF SiO SALVADOR. 357 
 
 " Other parts of the skeleton of the Brazilian species would per 
 haps show generic characters to distinguish it from the modern 
 proccclian crocodiles, but in the absence of these, it may for the 
 present be placed in the same genus. Its form, cutting edges, and 
 especially its peculiar striae, readily distinguish it from any species 
 with which it is liable to be confounded, and it may appropriately 
 be named Crocodilus ffarttii, in honor of its discoverer, whose 
 recent researches have thrown so much light on the geology of 
 Brazil. 
 
 " Several specimens of reptilian teeth collected by Mr. Allport 
 at Montserrate, a locality in the same deposit about two miles 
 southwest of Plataforma station, evidently belong to this species, 
 as the illustrations accompanying his paper (Plate xvi., figures 1, 
 2, 3, and 5) clearly indicate. The explanation of the plate refers 
 to the specimens as, Teeth of crocodile with delicately wrinkled 
 surface, but no further description is given. 
 
 " In the same paper Mr. Allport has given figures of several 
 Crocodilian teeth from the localities at Plataforma and Montser 
 rate, which are quite different from those above described. These 
 are represented in Plate xv., figure 5, and Plate xvi., figures 4, 6, 
 7, and 8, and are referred to on page 268 as, Teeth of crocodile 
 with strong continuous striae, and coarse riblets. These speci 
 mens, taken in connection with some imperfect remains in the 
 collection made by Professor Hartt, indicate the existence in this 
 deposit of a second and smaller species of Crocodile, probably 
 allied to the modern gavials. The teeth are not so large as those 
 of Crocodilus Harttii, and are more tapering and more curved. 
 They also differ widely in the striae and lateral folds. These spe 
 cimens may provisionally be referred to the genus Thoracosaurus, 
 and, as the species is evidently new, it may be called T. Bahiensis. 
 
 " An interesting fossil, found by Professor Hartt at Plataforma 
 station, is a fragment of a bone, evidently reptilian, but the exact 
 affinities of which it is difficult to determine from this specimen 
 alone. It resembles in some respects the extremity of an ulna, 
 
358 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 but after a careful comparison the writer is inclined to consider it 
 the proximal end of a rib. It is much flattened at the articular 
 extremity, and tapers gradually to the broken end, which is some 
 what triangular in outline. Its length is about four inches, the 
 transverse diameter of the perfect end two and a half inches, and 
 of the other, one and a quarter inches. The larger extremity is 
 divided into two articular facets lying oblique to each other, the 
 smaller one being elevated about half an inch above the other, and 
 covering rather more than a third of the entire terminal surface. 
 In form and general proportions this specimen is not unlike the 
 upper end of a right dorsal rib of some of the amphiccelian croc 
 odiles, especially a rib in which the head and tubercle have so 
 closely approached each other that their articular surfaces are 
 nearly confluent. The size and other characters of the specimen, 
 however, seem to exclude it from that order, and it probably be 
 longed to a Dinosaurian reptile, possibly the same as a large 
 vertebra from Monserrate, which Mr. Allport figured in his paper 
 in Plate xvn., and which Professor Owen suggested might prove to 
 be allied to Megalosaurus. 
 
 " The only other specimen in this collection that need be par 
 ticularly mentioned here is a small flat bone, about two inches in 
 length, with one articular extremity partially preserved. This 
 appears to resemble most nearly the fibula of a tortoise, and 
 probably should be referred to that group of reptiles." 
 
 I have not yet observed at this point any of the species of 
 Melania and Vivipara of Monserrate. In some of the beds, 
 however, a minute gasteropod is very abundant, and a unio- 
 likc shell also occurs, though somewhat rarely. These, 
 with the estherians, point to a fresh-water origin for these 
 beds, while the alternations of shales with sandstones and 
 conglomerates show that there were intervals of quiet depo 
 sition of fine material with times when currents spread 
 coarse material over their surface. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 
 
 359 
 
 This locality, to a patient and painstaking collector, would 
 yield a fine harvest. 
 
 From Periperi to Matto de Sao Joao I went over the rail 
 road several times both by cars and with a trolly ; but I 
 have had an opportunity of examining a few points between 
 these two stations, and with the aid of the information of 
 my friend, Mr. Tiplady, the engineer of the first section of 
 the railroad, and to whom I am much indebted for aid and 
 hospitality, I am able to give the following observations. 
 
 At the tunnel at Periperi the rocks consist of shale and 
 sandstone, with a slight northward dip, and contain but few 
 fossils. 
 
 Between Olaria and Mapelle stations there are heavy beds 
 of shale and sandstone, the latter a rather soft, bluish kind, 
 which is quarried for building purposes along the railway. 
 It is not very durable, as it weathers very rapidly. Fossils 
 are very rare in it. The tunnel at Mapelle is through shale, 
 the dip being northward and slight. 
 
 At a place called Cotigipe, between Mapelle and the next 
 station, Agua Comprida, thick beds of shale and conglom 
 erate are cut through. 
 
 a. Shale, b. Conglomerate. 
 
 In the shale of bed, a, of the above diagram, which varies 
 much in character from a thin, soft, black, almost papyra 
 ceous shale to a coarser greenish variety, fish remains and 
 perfect fishes are quite common, but they appear to be of 
 the same species that is found at Monserrate ; together with 
 these occur bones and teeth of reptiles, and an abundance 
 
360 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of cyprids. I have collected nothing from the conglom 
 erates. The dip of the beds is northeast, at a small angle. 
 Thus far the country has been irregular, the hills being 
 rather angular in their outline, and presenting altogether 
 different topographical features from those of the gneiss 
 regions. The cretaceous hills here are generally round 
 topped, with long slopes. 
 
 Passing Agua Comprida, at Sapucaia, there are several 
 cuttings through shale, and here we find the conglomerate 
 in the hills on the west side of the road. Farther on no 
 more of this rock is seen on the line. At Moritiba are 
 found beds of sandstone of the same series, but their dip, 
 according to Mr. Tiplady, is south. There are here some 
 horizontal beds of pinkish sands of a much later formation. 
 
 Cutting No. 82, not far from Moritiba, is through a heavy 
 bed of red clayey sand, sometimes used for moulding in the 
 railroad foundry at Periperi. This bed is quite horizontal, 
 and appears to belong to the same series as the sands to 
 be described farther on. 
 
 Just before reaching the Rio Johannes there is a cutting 
 through decomposed gneiss. Mr. Tiplady informed me 
 that in the river-bed, a few rods above the railroad, the 
 gneiss is exposed, while the rocks below the railroad are 
 sandstones. In this vicinity the hills are rounded and more 
 or less dome-shaped, yet not more than two to three hun 
 dred feet high ; but in going on towards Parafuso they 
 become lower, and the surface is rolling. At Parafuso 
 there is a long, low cutting, in which I made the following 
 section. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 361 
 
 The lower bed consists of a soft, yellowish sandstone, 
 with beautiful oblique lamination. I have looked long and 
 carefully in this and other similar localities for fossils in 
 these sandstones, but without success. Over this bed, 
 which is quite horizontal, lies a thin sheet of hard, red 
 iron-stone, and over this a bed of clay and pebbles. 
 
 In cuttings 8, 9, and 10 of the Third Section, which are 
 through long and low banks, are seen similar beds of light 
 pink-colored sand, slightly consolidated, and showing beau 
 tiful false bedding. 
 
 At Camassari I spent several hours in an examination of 
 the vicinity. The station stands on a plain composed of the 
 sands just described, over the surface of which is spread a 
 sheet of clayey gray or white sand, often containing pebbles 
 and broken pieces of rock. In the vicinity are a few low 
 hills, composed of beds of arenaceous pinkish and white- 
 mottled clays, and sandstones which vary much in charac 
 ter, being sometimes soft and pinkish in color, while at 
 others they are cemented by ferric oxide, dark red, and very 
 hard. These beds are horizontal and very thick, and be 
 long to the great coast tertiary formation. In some places 
 they lie in an unbroken sheet, except by rivers, forming very 
 extensive elevated plains, like the sertoes and chapadas of 
 Sao Matheos, the Mucury, and elsewhere along the coast ; 
 but sometimes, as on the Bahia Railroad, they are denuded 
 in such a way as to form deep and wide valleys, in which 
 stand isolated hills that rise abruptly from, or rather pierce 
 through, the thick beds of sands and clays formed from 
 their debris, and which occupy the bottom of the valley. 
 In the denudation of these horizontal tertiary beds the ten 
 dency is to form, at first, a mass with a flat top, and sides 
 steep and abrupt, as represented in the following diagram. 
 
 VOL. I. 16 
 
362 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 X \ 
 
 Such is the character of the slopes that border valleys. 
 These are rarely perpendicular, and are almost always 
 covered by vegetation. When, however, the sea or a river 
 washes the foot of a slope, it is always perpendicular ; as 
 is the case with the red clay cliffs, which at intervals gird 
 the shore of Brazil, from the Amazonas to Rio. As the 
 valleys widen and approach one another, the hills preserve 
 the same topographical features, and we may find all grada 
 tions between the broad chapada, the narrow hill with its 
 flat top, and the roof-shaped or conical hill. These de 
 tached hills are called oiteiros, or taboleiros. At Camassan 
 the topography of these hills may be easily examined. The 
 valley here is quite wide, and there are a great many small 
 scattered hills, that rise like islands from the plain. I ob 
 served that most of the hills had rather a long smooth slope 
 to the east, and an abrupt one to the west. The hills 
 are covered thickly with fragments of rocks. This seems 
 a little singular, because it is not owing to the dip of the 
 beds, for. as already said, they are horizontal. I have 
 questioned whether it might not be in some way due to 
 glacial action. The w r hole hill, in every case, was once 
 covered with a sheet of clay and fragments of rock, as I 
 shall describe more fully hereafter. The finer materials of 
 this superficial coating have been washed from the steeper 
 slope, leaving the coarser masses behind, while on the long 
 slope these last still lie buried below the surface. The 
 sands occupying the valleys are not deposited horizontally, 
 but form a series of sand-banks sloping more or less gradu 
 ally toward the middle of the valley. The surface is not a 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 363 
 
 plain, but rolling. The following is the outline of the 
 country across the hills from Camassari station, going 
 west for a little more than a mile. It is intended to show 
 the topographical features of the older tertiary hills and 
 sand plains. 
 
 CAMASSARI 
 
 a. Tertiary clays and sandstones. b. Sands. 
 
 While at Camassari station a man brought in a mule-load 
 of " Tabatinga," * or pure milk-white feldspathic clay, which 
 he said came from not far off, probably from a bed inter- 
 stratified with the sands. The material is exceedingly 
 fine and chalky, and becomes a pure white clay on wetting. 
 It is used instead of lime for whitewashing walls. This 
 Tabatinga clay is also found in the tertiary beds of the 
 coast, where it is often pinkish or yellowish in color. I 
 had observed the same material, as a cement to the super 
 ficial sands and gravel, in the cuttings in the vicinity of 
 Camassari. 
 
 Leaving Camassarf and crossing the taboleiros f one soon 
 enters the hills, which are one hundred and fifty feet or 
 more in height, composed of the tertiary sandstone and clays 
 above described, and have some of their sides strown with 
 large blocks of the red sandstone. A long cutting through 
 a heavy bed of quartz gravel, of which the cement appears 
 to be sand and clay, is soon reached, and one then goes 
 on to another sand plain, which is more extensive than 
 
 * This word is of Tupi origin, and is derived from the two words, Taud, 
 clay, and tinga, white. Tabatinga is a corruption, 
 t This word is also applied to the plains. 
 
364 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 that first described. Another bed of gravel is seen at a 
 place called Embira Branca, and farther on, when fairly 
 out on the taboleiros, at a place called As Pedras, there 
 is what appears to be an old beach, composed of rounded 
 quartz pebbles, filled in with snow-white sand. The beach 
 is irregular in outline, and presents many projecting points. 
 The surface of the sand, which is almost level, abuts against 
 this gravel bank, and lies in the indentations in its outline, 
 like a sheet of water. Just here this gravel is naked, but 
 not far distant it is covered by a dark grayish soil, similar 
 to that which covers the taboleiros. It appears to have 
 been washed off here. 
 
 a. Sand and soil. 6. Gravel. 
 
 A little farther on we find a cutting in which a cap of 
 gravel is seen overlying the sand. 
 
 a. Gravel. b. Sands. 
 
 Thence to the river Imbugahi the land slopes very grad 
 ually, and for miles on each side of the railroad is remark- 
 bly even, being interrupted only by the isolated tertiary 
 hills that show themselves above the surface of the sands. 
 From the ImbuQahi to Feira Velha the sands rise with 
 a very gentle slope. On this side they are bounded by 
 tertiary hills and chapadas, some of which are roof-shaped 
 ridges. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 365 
 
 Both at Camassari and on the Imbuc.ahi the sands are 
 barren, and form slightly rolling plains, supporting only 
 a scanty vegetation, which consists of several species of 
 grasses, and a multitude of small flowering plants. Trees 
 are very few. In the lower grounds, where moisture gath 
 ers, and where there are in some places streams, a long 
 coarse grass grows most luxuriantly. A quarter of a mile 
 south of the Imb^ahi is such a grass-covered area, and 
 here excavations by the side of the railroad show that a bed 
 of peat has accumulated, which is two feet thick in some 
 places. 
 
 Leaving the taboleiros we enter a valley among the hills, 
 which are at first low, but become higher farther on. They 
 belong to the tertiary series, are often very irregular in shape, 
 and about 350 feet in relative height. As we go up the valley 
 it narrows more and more. The bottom is occupied by beds 
 of sands and clays, which form a series of undulations 
 whose outlines are in contrast with those of the bordering 
 hills, as exhibited in the following diagram. 
 
 a. Tertiary clays and sandstones. b. Sands. 
 
 In order to illustrate the structure of these lower grounds, 
 I will describe what is seen in a number of railway excava 
 tions in this valley. 
 
 Two cuttings towards Bahia of a cutting called Jacu- 
 merim, the road just pares away the southern side of a 
 low hill, about twenty-five to thirty feet high, as seen in 
 the following sketch. 
 
366 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Tertiary-. 
 
 This liill is composed of sand, which is exposed for a 
 depth of twenty feet. The same material forms the lower 
 ground alongside. The hills in the background are com 
 posed of the tertiary clays and sands. The most interest 
 ing feature exhibited in this section is a sheet of fragments 
 of red sandstone from the tertiary hills extending under 
 the soil over the surface of the. sands, not only on the low 
 ground, but also over the whole hill. 
 
 The following cut represents a section across this same 
 hill, but at right angles to the first. The boulders of sand- 
 
 a. Soil containing pebbles and boulders of sandstone. 
 
 b. Stratified sands. 
 
 stone are here seen scattered over the low flat by the side 
 of the railroad. These boulders must have come from the 
 adjacent tertiary hills. 
 
 Going up the road we soon come to another cutting in 
 which a sheet of gravel overlying a bed of yellow sands is 
 exposed, as seen in diagram on preceding page. This gravel 
 consists of large, well-rounded pebbles of quartz or decom 
 posed syenite, with a white paste. In the next cutting, 
 Jacumerim, we see the following section. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 
 
 The pebbles in the gravel are of the same kind as those 
 just described. They are very large, the material being 
 rather a coarse shingle, and they are very closely packed. 
 Over this elevation, as in the other, is still seen the layer 
 of broken pieces of sandstone, and the same may be exam 
 ined in the next cutting, which is a very instructive one. 
 
 SECTION AT CAMPO. 
 
 a. Fine brick-red clayey sand, in some places with scarcely consistence enough 
 to be moulded ; in others very clayey. 
 
 b. Whitish and yellowish sand. 
 
 c. Soft white sands, no consistency under pressure. A thin streak or two in the 
 lower part of this bed is consolidated. 
 
 d. White or reddish sands. 
 
 e. An irregular mass of boulders, of dark red tertiary sandstone, some of which 
 are three to four feet in diameter. 
 
 In this diagram the dip is exaggerated. At the post 
 marking ten leagues from the city of Bahia is a long low 
 cutting, and in a trench by the roadside a stiff sandy clay, 
 mottled with dark red and white, and overlaid by a dark 
 brown sandy soil. 
 
 Between this point and Matta station there is much 
 swampy ground bordering the road at intervals, and on this 
 great numbers of piassaba palms grow. 
 
368 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 A very interesting section occurs in a cut about a mile 
 and a half below Matta station. The gravel sheet here is 
 very thick, ten feet, more or less,* and is irregularly and 
 obliquely stratified, and mixed with layers of sand, as if the 
 whole had been laid down by a current coming from the north. 
 The sheet is very irregular, and varies much in thickness. 
 
 Matta station, and the village of Matta de Sao Joa o, are 
 situated on a plain, formed by the widening of the valley 
 which the railroad follows. The tertiary hills still border 
 it, and these are, in some cases, much denuded. 
 
 Above Matta, at the cutting at the fazenda of Coronel Si- 
 puda, there is a decayed, fine-grained sandstone or arenaceous 
 shale seen, probably cretaceous, and thence on to Pitanga 
 station, rocks of the same general character. Just below 
 the station at Pitanga I made the following section. 
 
 a. Sandstone. 
 
 b. Shale, slightly arenaceous. 
 
 c. Shale ; very argillaceous, and so soft as to be easily moulded by the hand. De 
 composed, it is of a gray or light slate-color, with bright red patches. It contains 
 apparently the same Estherian found at Pojuca tunnel, to be presently described. 
 
 (/and e. Shale decomposing yellow and red. 
 
 f. Sandstone; soft and ochre yellow. 
 
 g. Shnle decomposed to a pink and white clay. 
 A. Sandstone. 
 
 k. Layer of concretionary iron-stone a foot or two in thickness. The surface is 
 very irregular, and some of the projecting knobs have a sort of septaria structure- 
 m. White sandstone. 
 n. Layer of iron-stone. 
 
 o. White tabatinga, with red iron-stone nodules. 
 p. Decomposed shale. 
 q. Sandstone (soft). The dip is exaggerated in this section. 
 
 The railroad here skirts a flat along which flows the Rio 
 
 * I have omitted to note it exactly. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 369 
 
 Pitanga, which a short distance below the station expands 
 into a large pond. At the head of this pond I found a 
 gentleman with six men washing for diamonds. An ex 
 cavation several yards square had been made in the flat 
 alluvial ground. The lowest stratum seen was a stiff blue 
 clay, called by the miners pizarra, over this was a sheet of 
 a few inches in thickness of cascalho, or gravel composed 
 principally of quartz and fragments of the different rocks of 
 the neighborhood, this cascalho being mixed with a mottled 
 reddish clay. This was the stratum from which the diamonds 
 were washed. Over the cascalho was a thin stratum of clay. 
 The washers were seated on benches level with the water of 
 a little pond ; a portion of the gravel was thrown into a 
 large shallow wooden pan, water was mixed with it, and 
 the pan with its contents was whirled about on the surface 
 of the pond, the muddy water being allowed to escape until 
 only the gravel and sand remained. This was now placed 
 in a sort of sieve, the bottom of which was made of copper, 
 pierced with round holes about an inch in diameter, which 
 allowed all the little pebbles and sand to pass through. 
 Not long since, at one of the Brazilian diamond washings, 
 a diamond too large to pass through the openings of the 
 sieve was found in the heap of coarse refuse. It had 
 been picked up by a child as a plaything. After washing, 
 the pebbles are picked out, when the overseer, looking 
 carefully over the sand, selects the diamonds with ease, 
 should there be any ; but during the weeks which they 
 washed at Pitanga only five minute stones had been found. 
 A few years ago quite extensive washings were carried 
 on beside the river nearer the station, and a large num 
 ber of fine stones were discovered ; but the owner of the 
 land has put a stop to further operations. On the west side 
 
 16* X 
 
370 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of the railroad I visited some other old washings, from 
 which diamonds had been obtained. I found that they had 
 been dug from the cascalho or gravel sheet on the slope of 
 the hill near the river, which gravel was covered with the 
 common red drift-clay of the vicinity. I take this gravel to 
 be drift. The diamonds appear to me to come from the 
 tertiary beds of the neighboring hills, and this seems to 
 be the opinion of Mr. Nicolay, who shows that the diamonds 
 of the Chapada Diamantina come from a conglomerate and 
 sandstone, which, from his descriptions and specimens, ap 
 pears to be a tertiary rock of the same kind as that which 
 forms the chapadas of the valley of the Jequitinhonha. Mr. 
 Nicolay says that the cascalho in the vicinity of Bahia dif 
 fers from that of the chapada, in that it is more siliceous 
 and less metaliferous. There is no Itacolumito in the 
 vicinity of Pitanga. The gravel is made up principally of 
 fragments of quartz and of pebbles of a sandstone like that 
 of the tertiary chapada, though somewhat harder than the 
 kind usually seen along the road. I do not believe that the 
 diamond ever occurs in the true palaeozoic Itacolumite in 
 Brazil, but that it is derived from the tertiary sandstones. 
 
 At a sugar fazenda beyond Pitanga there is a cutting 
 under a bridge across the railroad, and in a gully made by 
 the rain I found a shale full of cypris, and evidently creta 
 ceous. The beds appear to be horizontal, or with a slight 
 inclination southward. The overlying soil here is red 
 mottled with white. 
 
 The next cuttings are not very satisfactory, and appear 
 to be through decomposed sandstones and shales. 
 
 I introduce a sketch here to give some idea of the topog 
 raphy, and of the general appearance of the tertiary hills 
 bordering the valley on the west, just below the tunnel. 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 
 
 TERTIARY HILLS NEAR POJUCA TUNNEL. 
 
 The sandstone which to some extent, at least, forms the 
 chapadas of the vicinity is a somewhat soft, coarse-grained, 
 reddish and mottled kind, with an argillaceous cement. It 
 has been occasionally used on the railroad for building pur 
 poses, but it cannot be very durable. 
 
 A ridge crosses the railroad, and a tunnel (" Pojuca tun 
 nel ") is made necessary. This ridge runs nearly north- 
 south, and is composed mainly of cretaceous strata, which, 
 according to information received from Mr. Turner, en 
 gineer of this section of the road, have a high westward dip. 
 In the cutting at the eastern entrance to the tunnel hori 
 zontal beds of a soft brown sandstone are exposed, and 
 this is so soft as to be easily crumbled by the hand. In this 
 occurs a bed, fifteen centm. in thickness, of a yellow clay 
 (decomposed shale ?) in which I have found a few Esthe- 
 rians. The stuff thrown out of the tunnel is a bright bluish, 
 very soft shale, containing fish remains, together with a 
 finely laminated blue-gray sandstone, also containing fish 
 remains and fragments of plants. I did not see these 
 strata in place. 
 
372 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 In the cutting at the western entrance are thick beds of 
 shale and sandstone, so decomposed that their character 
 cannot well be determined. The strike, as near as I could 
 make it out, is N. 65 E. and dip 73 S. E., but it is to be 
 remembered that Mr. Turner says that in the tunnel the 
 dip is W. The beds of the tunnel and western entrance 
 are cretaceous, I think. Those at the east end appear to 
 overlie the cretaceous beds unconformably, and may be ter 
 tiary ; but I would not speak positively, on account of 
 the great difficulty of recognizing some of these rocks 
 when they are decomposed. This ridge is covered with red 
 drift. 
 
 From Pojuca to Sitio Novo I have been over the road 
 in the train several times, but I was prevented from ex 
 amining the rest of the road on foot, because of the rain. 
 From Sitio Novo to Alagoinhas I went over the road in a 
 trolly, but in rain and under unfavorable circumstances. 
 Soon after leaving Pojuca the valley becomes narrower. 
 The railroad follows its bottom, so that there are no rocks 
 exposed, and the cuttings are of very little interest. The 
 low lands are very fertile, and are here and there covered 
 by forest, though they are planted to a considerable extent 
 with sugar-cane. 
 
 As for the lands of Bahia, those of the gneiss are fat, 
 deep, and exceedingly fertile ; the slopes are generally 
 somewhat steep, and in the southern part of the province 
 they are favorable for coffee. The soils of the cretaceous 
 along the railroad are rich ; but the rains make fearful 
 havoc with them, carrying them away completely after a 
 few years of cultivation, leaving them barren, as is the 
 case over large tracts along the railroad. Of the low lands, 
 those bordering the rivers are very fertile. The tertiary 
 
GEOLOGY OF THE VICINITY OF SiO SALVADOR. 373 
 
 high lands are sometimes valuable on the slopes, but above, 
 as a general thing, they are dry and barren. The taboleiros 
 are useless, except for grazing. Immense tracts along the 
 railroad are allowed to grow up with the piassaba palm. 
 Considerable difficulty has been experienced by the engi 
 neers, in different parts of the road, from the instability 
 of the rocks, and the liability of their sliding on one 
 another, especially when the shale beds become wet. In 
 one case, where the beds dipped with the slope of a hill 
 toward the railway, the upper beds over a considerable 
 area slid down some distance, throwing the track out of 
 shape. 
 
 The Pojuca tunnel caved in from the same reason, about 
 two years ago. Some distance south of Alagoinhas the 
 valley opens out and an extensive series of taboleiros is 
 reached, in the middle of which is the station of Alagoin- 
 has. I made a long detour over these taboleiros on foot, 
 and found them to be essentially like those at Camassari. 
 They were great, gently undulating plains of sand, scat 
 tered about over which were irregular isolated hills of the 
 horizontal tertiary clays and sandstones. The cutting at 
 the terminus of the railroad is a long one, about five feet 
 deep, through white clayey sands, under which is the ever- 
 appearing pebble line. 
 
 In a cutting for a road through a slight elevation near 
 the station, I saw beds of clayey sand slightly consistent and 
 of a white or pinkish color. Here the Saiiba ants had 
 formed immense mounds of a clayey sand of a warm pink 
 ish tint, the material having been brought up from a con 
 siderable depth by these busy creatures in the excavation 
 of their galleries. 
 
 The village of Alagoinhas is on the sand plains, about a 
 
374 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 league eastward of the station of the same name. The 
 surface of the plains is not level, and there are differences 
 of elevation amounting to fifty feet more or less, but the 
 slopes are very long, exceedingly smooth and even, and as 
 a general thing their direction is from the hills towards the 
 centre of the plain. The surface is loose white sand, of 
 course a most unfertile soil. The vegetation of the sand 
 plains and taboleiros differs in the most marked way from 
 that of the soil-covered hills. Trees are few, scattered, small, 
 and very often with rough bark, and stiff and contorted 
 branches. One of the most conspicuous of them is the 
 Carahyba, a tree about twenty feet in height, which is scat 
 tered all over the plains. 
 
 The Murici (Byrsonima) is a small scrubby tree, about 
 eight feet high, with bunches of large elliptical leaves cov 
 ered with hairs like the mullein borne at the ends of the 
 branches. The fruit is about the size of a large cherry, 
 yellow, very fragrant and much esteemed. The Mangaba 
 (Hamornia speciosa) is another small tree, with weep 
 ing branches and small leaves. Its fruit is of the size of 
 a plum, and very delicious. The Perico is a bush produ 
 cing a fruit as large as a gooseberry, and very pleasant to 
 the taste. The Bahianos are fond of it, and in the season 
 of fruitage the berries are sold in large quantities in the 
 city. 
 
 A small tree called Sambahiba is remarkable for its 
 curled leaves, the upper side of which is so rough and 
 hard as to scratch wood like sandpaper. The Janahuba is 
 a characteristic shrub of the taboleiros, and it is noteworthy 
 on account of bearing at the end of its stem a cluster of 
 large leaves, giving out an abundant milky sap when broken. 
 Among the other common plants of the taboleiros one ob- 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 375 
 
 serves the Alecrim, with its fragrant leaves ; the Almescar, 
 furnishing a sweet-scented resin ; the Macella, producing a 
 material used in stuffing mattresses ; the Purga do Campo 
 and Orelha da On^a, both used in medicine, and the curious 
 Barrigudinha, with its swollen stem. Creeping about over 
 the bushes, one sees here and there tangled skeins of the 
 yellow thread-like stems of the Cipo de Chumbo, a species of 
 Cuscuta, or Dodder. A large number of the plants of the 
 taboleiros are medicinal, and very many have aromatic 
 leaves. Small and beautiful flowering plants abound among 
 the tufts of coarse grass with which the plains are covered. 
 These grasses grow in widely separated clumps. My guide 
 indicated the three principal kinds as Capim agreste, Ca- 
 pim de C/ieiro, and Capim pubo. I observed one or two 
 species of Melocactus, but I do not remember having seen a 
 single large Cereus, a plant so common on the dry rocky 
 sertoes of the Rio de Sao Francisco. A little palm, Licori, 
 is very common, and is the only species I saw on these 
 campos. 
 
 Ant-hills are common on the taboleiros, looking like scat 
 tered boulders. Under one of these I found a large scor 
 pion, an insect which, like the centipede, is common enough 
 in Brazil ; but one might travel a year in the country with 
 out seeing a specimen of either, unless he made special 
 search for them. 
 
 As elsewhere, these campos are burned over from time 
 to time, and the flora has consequently suffered great 
 modification. 
 
 To give an idea of the general character of the vegetation, 
 as well as of the topography of the sand plains and the ter 
 tiary hills near Alagoinhas, I introduce the sketch on the 
 following page. 
 
376 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 THE TABOLEIROS NKAK ALAGOLNHAS. 
 
 These hills are composed of a soft, reddish, argillaceous 
 sandstone, with some layers, however, of a coarse kind, very 
 hard, of a dark-red color and with a cement of ferric oxide. 
 In some places these rocks form cliffs. Here, as in the 
 taboleiros of Camassari, the sands bathe the foot of the hills 
 like a sheet of water. The hills are covered from top to 
 bottom with fragments of the coarse red sandstone, with 
 some quartz pebbles. Usually round the base of the hills 
 the sands have a coating of red soil washed down from the 
 hillside. Near Jacare the soil of some parts of the hilly 
 land becomes better in quality, and is soft and loose. Here 
 we find a liberal forest coating, and some poor settlers cul 
 tivate fields of mandioca, tobacco, corn, &c. The soil is 
 sandy, and grayish-brown in color. Farther back comes 
 a belt of forest beyond the taboleiros, and a considerable 
 quantity of sugar is raised in that region. 
 
 The height of Alagoinhas station above the sea, according 
 to the survey of Mr. Vivian as laid down on the manuscript 
 
GEOLOGY OF THE VICINITY OF SAO SALVADOR. 377 
 
 map of Mr. Nicolay, is three hundred feet. The tertiary 
 hills of the vicinity I should judge to be one hundred and 
 fifty feet higher, more or less. 
 
 As to the extension of the cretaceous beds on both sides 
 of the line, I have very scanty information. The island of 
 Madre de Deus, in the Bay of Bahia, belongs to the same 
 series, as I have seen in Mr. Nicolay s collection several 
 specimens of sandstone and other rocks from this locality 
 precisely like those on the railroad ; and the island of Itapa- 
 rica is also, to a large extent at least, cretaceous. I have 
 not heard of their extension farther south. I believe that 
 the Bahia cretaceous series is confined entirely to the bor 
 ders and islands of the Bay of Bahia, and that it is for the 
 most part an accumulation within a closed fresh-water basin.* 
 
 The sands and gravels of the plains and connecting 
 valleys are certainly newer than the coast tertiary beds, 
 which are denuded in order to form the basins in which 
 these were deposited. They appear to me to be of la 
 custrine and fluviatile origin, and I believe that the plains 
 were once covered by large lakes, which have been drained 
 by the cutting through of the high lands on the south 
 by the streams flowing from them. From what I have 
 been able to observe, these beds are uniformly overspread 
 by a sheet of clayey sand, mixed with fragments of rock 
 from the neighboring hills, which sheet I consider to be 
 drift, so that the sands and gravels may possibly be of 
 very late tertiary age. It is very probable that they may 
 turn out to belong to the same series as certain similar de 
 posits observed by Mr. St. John in the valley of the Rio de 
 Sao Francisco, which are overlaid by drift in the same way. 
 
 * In Professor Agassiz s collection I find a fragment of greenish shale with 
 cyprids from Parahyba do Norte, which appears to be from a formation similar 
 to that of Bahia. 
 
GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER IX. 
 
 THE PROVINCES OF SERGIPE AND ALAGOAS, AND THE RIVER 
 SAO FRANCISCO BELOW THE FALLS. 
 
 The Province of Sergipe. Its Division into Mattas and Agrestes. The Eio 
 Real. Estancia ; New Red Sandstone, Sugar Plantations, c. Sand 
 Dunes. The Rio Vasabarris. The Rio Cotinguiba. Aracajii. Cre 
 taceous Beds with Inocerami at Sapucahy. Maro im. Cretaceous Lime 
 stone with Ammonites. " Fossil Turtles." Sugar Plantations. Messrs. 
 Schramm and Company. The Bar of the Sao Francisco. Sand Dunes 
 of the Pontal. Character of the River below Penedo. Aracare. Villa 
 Nova and its Cretaceous Sandstones. The City of Penedo and its Geol 
 ogy. Its Commerce and Fair. Notes on the Piranha and its Habits. 
 Propria. Morro do Chaves and Cretaceous Fossils. Traipu. Iron 
 Ore. Campos, Vegetation, Cactuses, &c. Pao de Assucar. Cattle 
 Fazendas. Piranhas. County flat and covered by Boulders. The 
 River Valley a Narrow Gorge in a Gneiss Plain. The Falls of Paulo Af- 
 fonso. Halfeld s Description. Liais s Description. Comparison be 
 tween Paulo Affonso and Niagara. Mastodon Remains from near the 
 Falls. Climate of the Sao Francisco below the Falls. Steam Naviga 
 tion. Character of the Coast of the Province of Alagoas, South of 
 Maceid. The Lagoas. The City of Maceid and the Geology of its 
 Vicinity. Tertiary Beds. Harbor and Reefs. 
 
 THE Province of Sergipe is a very small one, wedged in 
 on the coast between the province of Bahia and the Rio 
 de Sao Francisco. Its coast line is only about ninety miles 
 in extent. The eastern part of the province is low and 
 uneven, and there are extensive tracts of sands along the 
 coast. In this region are some lands fit for cultivation. 
 The western half of the province is higher and somewhat 
 mountainous, the principal mountain range being the Serra 
 
PROVINCES OF SERGIPE AND ALAGoAS- 379 
 
 d ltabayana, of which I shall have occasion hereafter to 
 speak. The eastern half of the province passes by the name 
 of Mattas, because of its forests, while the term Agrestes is 
 applied to the western portion, because it is destitute of 
 forest, and is to a large extent barren and dry, being in 
 some parts fit only for pasturage, though during the dry 
 season water fails, the vegetation dries up, and cattle suffer 
 from hunger and thirst. 
 
 The principal rivers of the province are the Rio Real, 
 which separates it from the province of Bahia on the south, 
 and empties into the sea at the same mouth with the Rio 
 Piauhy, the Rio Yasabarris, the Cotinguiba, and the Japa- 
 ratuba. 
 
 The Rio Real is a small river navigable for a distance of 
 some nine leagues above its mouth. In its upper course it 
 flows through the dry belt, and it is bordered by cattle fa- 
 zendas. A little cotton is raised along the river. 
 
 The Piauhy, another small river rising in the western 
 part of the province, and to the north of the Real, unites 
 with the latter river just before it empties into the sea. On 
 my return from a voyage on the Sao Francisco I touched at 
 Estancia, and made a hasty visit to the town. Estancia is 
 built on a rolling country, where the heights of the imme 
 diate vicinity are not more than two or three hundred feet. 
 The hills are rounded, and the rocks composing them are 
 coarse red micaceous sandstones, quite indistinguishable 
 in the hand specimen from the triassic red sandstone of 
 New Jersey. This sandstone covers a large area, and must 
 be very thick. I examined it in several places, but found 
 no signs of fossils. The dip, as a general thing, appears to 
 be but a few degrees to the eastward. This sandstone is 
 covered by a red clayey soil, which bakes very hard, so that 
 
380 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the vicinity of Estancia is very arid and of little fertility, 
 and the vegetation is low and sparse. There are, however, 
 very productive lands lying farther to the westward, and 
 probably inside of the sandstone range. Estancia exports 
 every year 7,000 boxes of sugar (48,000 cwt.), together 
 with some cotton, tobacco, etc. 
 
 Below Estancia I found a poor exposure of limestone, 
 containing shells, and which was quarried for burning into 
 lime. I could discover no signs of bedding. The Dic- 
 cionarlo Geog-raphico says that in the neighborhood of 
 the Rio Piauhy a mine of coal was discovered, and that 
 since 1840 no other coal has been used in the forges of 
 the vicinity ; but while at Estancia I made careful in 
 quiries about minerals, and heard not a word said about 
 the existence of coal in the neighborhood. 
 
 The river is narrow, and at the time of my visit, in Au 
 gust, 1867, it was very shallow. The water below the falls 
 was fresh, but very muddy. As the hot season advances 
 the river falls very low, and the salt water flows up to the 
 port. The lower part of the river is bordered by extensive 
 mangrove swamps ; its mouth is obstructed by a bad bar. 
 On shore, on the south side, are some magnificent sand- 
 dunes, forty to fifty feet high, as regular in their outlines 
 and as white in color as snow-drifts. 
 
 SAND-DUNES AT THE MOUTH OF THE RIO REAL. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 381 
 
 North of Estancia low, irregular hills begin to make their 
 appearance, and seem to be tertiary outliers ; but away in 
 the interior, twenty or thirty miles from the coast, is seen 
 the blue outline of the Serra de Itabayana. It presents a 
 low, very evenly rounded form, and must be over two thou 
 sand feet in height. It is composed of gneiss and mica 
 slate, as I had an opportunity of seeing on the Rio de Sao 
 Francisco. 
 
 The Vasabarris, anciently denominated Irapirang, rises 
 in the province of Bahia, in the Scrra Itiuba, according to 
 the Diccionario Geographico, and is of very little impor 
 tance except near the sea, where it is navigable for a dis 
 tance of about twenty miles. The same authority says that 
 the Rio Sergipe flows into it near its mouth, which is mani 
 festly a mistake. Not far from the mouth is the city of 
 Sergipe d El-Rei, or Sao Christovao, on the bank of a minor 
 stream emptying into it from the north. This city was, for 
 many years T the capital of the province, but so very bad is 
 the bar of the Vasabarris that the river was not to be 
 depended upon for navigation. The capital was conse 
 quently removed to Aracaju, on the Cotinguiba, a few 
 leagues farther north. The city is now in decay. 
 
 The Cotinguiba, or Cotindiba, is a smaller river than the 
 Vasabarris, and takes its rise, as near as I can ascertain, 
 in the Serra d ltabayana. It is navigable, at high water, 
 for smacks only, as far as Maro im, a distance of some ten 
 or twelve miles. The river has several branches, on which 
 are a number of towns of more or less importance. At 
 the mouth the river is very wide, presenting a beautiful 
 sheet of water ; but it appears to be, after all, only a sort 
 of estuary. Between Maro im and the sea the banks are 
 largely covered by mangroves, but there are some hills and 
 hisrhcr lands. 
 
382 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The bar at the mouth of the river is very dangerous, 
 and the surf beats on it with great fury, sometimes pre 
 cluding the possibility of entering. Almost across the 
 mouth, from the northern side, stretches a line of sand 
 banks of a crescent-like shape. Three of these are joined 
 
 Bar of 
 
 Rio 
 
 together, their convexities being turned toward the sea, 
 while the extremities are produced up the river as long 
 spits. A similar sand-bank is attached to the northern 
 bank. These hook-shaped spits of sand are made in the 
 struggle between the river and the waves of the sea, in 
 the same manner as the hook at the mouth of New York 
 Bay is made. Like the other rivers on this coast, the 
 Cotinguiba enters the sea very obliquely with a southeast 
 course. The left bank is flat, composed of sand, and 
 continues low for several miles up the river. On the oppo 
 site side, at the mouth, there are extensive dunes forty 
 to fifty feet high, flanking a tract of recently elevated sands, 
 stretching along several miles, covered with cocoa-palms 
 
PROVINCES OF SERGIPE AND ALAG<3AS. 383 
 
 as far as the city of Aracajii, a small and uninteresting 
 town, the capital, built on a flat of lower alluvial ground, 
 bordering the river at the base of the sands. In the upper 
 part of the town the dunes are very large and conspicuous. 
 In company with Dr. Brunet, Director of the Agricul 
 tural College at Bahia, I walked over the stratified sands 
 for some distance up the river to a hill on which there is 
 a church and little village. I saw no shells in the sands. 
 The hill referred to is tertiary, and we saw in the soil cov 
 ering it irregular masses of the common dark-red tertiary 
 sandstone. The height of the sand plain above water level 
 was, if I remember rightly, about fifteen feet. 
 
 A few miles above the city, and also on the right bank of 
 the river, at a place called Sapucahy, there is quite a large 
 quarry in a little hill composed of a white flaggy and shaly, 
 rather soft and chalky limestone, used quite extensively 
 for building purposes in Aracaju. Of this limestone a 
 thickness of over a hundred feet is exposed. The stratifi 
 cation is remarkably regular, and the quality of the stone 
 is very uniform. As a general thing it is almost pure, white 
 and somewhat soft, but there are bands of a grayish variety, 
 much harder, which have, at first sight, the appearance of 
 a lithographic stone. The thinner shales resemble strongly 
 those of Solenhofen. A large part of the rock is flaggy, 
 and is readily quarried out in large thin slabs. There are 
 some lines of flints in these beds, but these are not rounded 
 as in the English chalk, but tabular and angular. I have 
 submitted some specimens of these rocks to my friend Mr. 
 Arthur M. Edwards, the microscopist, of New York, who 
 has been unsuccessful in discovering in them any micro 
 scopic remains. On the surface of some of the layers of 
 limestone I have found great numbers of valves of a pretty 
 
384 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 InoceramuSj most probably new, together with a little Am 
 monite and some teliostian fish-scales. I am told that per 
 fect fishes have been obtained here as well as at Laran- 
 geiras, and some specimens were collected a few years ago 
 in this vicinity by his Majesty the Emperor of Brazil, a 
 shrewd geological observer. This white limestone appears 
 to represent the white chalk. The beds at Sapucahy have 
 a moderate dip to the southeast approximately. 
 
 Between Sapucahy and Maroim, a few miles up the river, 
 the grounds are low, and the boatmen informed me that lime 
 abounded in the vicinity. The limestone of Sapucahy is 
 said not to furnish good lime. At Maroim I was gratified to 
 find the streets paved with large blocks of a coarse yellowish 
 limestone, and to see on them the impressions of some large 
 Ammonites and Ceratites, some of which I took up and 
 brought away with me.* Besides these fossils I obtained, 
 
 * I visited the quarry that supplied the stone, hut in th", short time at my 
 disposal I found hut little. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 385 
 
 through the kindness of Mr. Adolph Lane,* several speci 
 mens of a large Natica. Mr. Nicolay once showed the 
 drawing of a Cidaris sent to him from that locality. 
 
 At the Salem meeting of the Association for the Advance 
 ment of Science, in August, 18G9, I exhibited my collection 
 of Maroiin fossils, and expressed the opinion that they were 
 cretaceous. Professor Alpheus Hyatt, in examining the 
 specimens, was struck with the remarkable peculiarities 
 possessed by some of them, and which recalled Jurassic 
 forms. Professor Hyatt kindly consented to study the 
 specimens critically, and I am glad to be able to present 
 the following report upon them, which he has prepared for 
 this work. I am especially glad to have the opinion of this 
 naturalist on the Ceratites and Ammonites, because he has 
 made such exceedingly careful studies of these groups. 
 
 Report on the Cretaceous Fossils from Jfaro im, Province of Sergipe, 
 Brazil, in the Collection of PROFESSOR HA.RTT. By ALPHEUS 
 HYATT, S.B., Curator in the Pecibody Academy, Salem, Mass. 
 
 In the small collection of fossils from the above locality, placed 
 in my hands for examination, I have been able to make out the 
 following species : 
 
 Natica prwlonga Leymerie. 
 
 Natica prcdonga Leymerie, 1842, Mem. de la Soc. Ge"ol. t. v. pi. 16, fig. 8, p. 13; 
 D Orb., Voy. dans 1 Am^r. Mer. t. 3, p. 73, pi. 18; Pal. Fran., Terr. Cr4t. p. 152, 
 pi. 172, fig. 1. 
 
 This species is quite closely allied to Natica Pierdenalis of 
 Roenier, collected in Texas, but has a longer and more acute spire. 
 The French specimens were found at Thieffrain and Yandoeuvre, 
 in the Lower Neocomien, by M. Leymerie, and the identification 
 
 * Since my return home Mr. Laue has been so kind as to send me small lots 
 of these fossils. 
 
 V.OL. I. 17 Y 
 
386 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Avas made by a comparison between the Brazilian specimens and 
 the figure given by D Orbigny in the PaUontologie Franqais. 
 D Orbigny s figure of the specimen from Columbia, in his Voyage 
 dans FAmerique Meridionale, is that of a young specimen found 
 by Boussingault on the Rio Suba, one of the affluents of the Rio 
 Suarez. This agrees perfectly with the figure of the French 
 specimen. The French, Columbian, and Brazilian specimens are 
 all casts of the interior. 
 
 Locality : Cretaceous of Maroi m, C. F. H., 1867, and Mr. Adolph 
 
 Lane. 
 
 Ceratites Harttii n. sp. 
 
 This specimen I was at first disposed to consider as a Goniatite 
 in spite of the septa, the characteristics of which are unquestion 
 ably ceratitic. It is evidently a fossil cast which has been trans 
 ported to the beds in which it was found, from some older stratum 
 of precisely the same lithological composition. Serpulse have in- 
 crusted the surface, stretching their long cornucopia-like shell- 
 cases across, or fitting closely into the abraded depressions which 
 mark the former edges of the septal partitions. The presence of 
 these serpulse show conclusively that the specimens rrmst have 
 been a fossil long before they began to grow upon its surface, and 
 that it had suffered much from the wear and tear of the elements 
 before they could have fitted themselves so accurately into the 
 depressions of its rough and pitted exterior. The umbilicus is 
 filled with the tough calcareotis matrix, but its form must neces 
 sarily be that of a very deep funnel, the whorls enveloping the 
 greater portion, if not the whole, of the sides of the young. This 
 is certainly the case with regard to the last whorl, which covers 
 nearly the whole breadth of the sides of the immediately preced 
 ing volution, leaving only a narrow band exposed. This peculi 
 arity, and the great transverse breadth of the volutions, gives to 
 this species a remarkably close resemblance to Goniatites crenistria 
 and other allied forms, and this it was which, at first sight, led me 
 to imagine that the older Goniatites, as well as the Ceratites, had 
 
PROVINCES OF SERGIPE AND ALAGoAS. 387 
 
 "colonized" the cretaceous shores of Brazil. The longest diameter 
 of the cast is nearly five (4|) inches. The latter half of the 
 last volution, occupied by the terminal chamber, is devoid of sep- 
 tal markings, and has suffered too much from abrasion to allow of 
 accurate measurements. Near the last septum, however, the ful 
 ness of the original shell is very well preserved, and here the 
 dorso-abdominal breadth of the last whorl was found to be two 
 sevenths shorter than the dorsal breadth, measuring from shoul 
 der to shoulder of the umbilical edge. The abdomen and sides 
 are symmetrically rounded, reaching their greatest width or de 
 parture from each other on the umbilical edge. The curvature of 
 the sides is so rapid, however, and the apparent elevation of the 
 abdomen so great, that, without actual measurement, no one would 
 be apt to suspect that the base of the arch was longer than its 
 axis of elevation. The edges of the septa, though worn away to 
 a considerable extent, are sufficiently distinct to allow of the de 
 termination of their general characteristics. They are evidently 
 more closely allied to those of Ceratites (Ammonites) Robinii, as 
 figured by Von Buch,* than any other species. The abdominal 
 lobe, however, has a siphonal cell of ordinary size in place of the 
 large broad cell occupying the abdomen of that species. The 
 lateral lobes and cells have very nearly the same general outline 
 as in Ceratites Robinii, and the superior lateral cell has a similar 
 triple division of the base line, due to the presence of two minor 
 lobes of equal size. The superior lateral lobe is about one third 
 deeper than the abdominal lobe. The inferior lateral cell is very 
 much broader than in Ceratites Robinii, and the base, instead of 
 being smooth, is broken by two or more minor lobes, almost obliter 
 ated, however, in the cast. In other respects these cells are very 
 like those of Ceratites Robinii, having precisely the same hump-like 
 form rising gradually on the abdominal, and bulging out on the 
 dorsal or umbilical sides. The inferior lateral lobes in both spe 
 cies are, on account of the contiguity of the septa, in contact, and 
 
 * Utler Ceratiten, Abhand, d. Kong. Akad. Berlin, 1848, p. 176, fig. 4. 
 
388 GEOLOGY AXD PHYSICAL GEOGRAPHY. 
 
 set one into the other like a pile of coffee-cups ; this lobe, how 
 ever, in Ceratites Hartlii is either deeper or of about the same 
 depth as the superior lateral. The first auxiliary cells and lobes 
 are equally imitative, but the basal outlines of the former are di 
 vided by two minor lobes similar to those of the superior and 
 inferior lateral cells. 
 
 The first auxiliary lobe, also, reaches the umbilical edge, instead 
 of being situated at some distance therefrom, as in Ceratites Robinii, 
 and the smaller auxiliary lobes and cells are upon the unexposed, 
 inturned portion of the border, and are not seen upon the sides, as 
 in Von Buch s figure of the latter. Probably no very close com 
 parisons can be made between the form of the shell in these two 
 species, since Ceratites Robinii has a shallow umbilicus and nar 
 rower whorls. There seems to be no doubt, therefore, that this 
 fossil is undescribed, and as it is the remains of a very remarkable 
 cretaceous animal, its dedication to its discoverer cannot be classed 
 among compliments of a similar sort which are too often common 
 place or misapplied.* 
 
 Ceratites (Ammonites) Pierdenalis Von Buch. 
 
 No. 4 of the collection is so closely allied to this species in the 
 form of the whorls and the extent to which they envelop each 
 other that I have no doubt of their specific identity. The septa 
 are not apparent, but the hollowing in of the umbilical sides of 
 the volutions and the acuteness of the abdomen are well marked 
 and evidently the same as in Ammonites Pierdenalis. 
 
 Locality. In cretaceous beds at Maroi m. C. F. H. and Mr. 
 Adolph Laue, collectors. 
 
 Ammonites Hallii Meek and Hay den 1 ? 
 
 No. 1 is probably a fragment of a large whorl of Ammonites 
 Hallii of Meek and Hayden, or a closely allied species. Although 
 
 * This Ccratite was obtained from the cretaceous beds of Maroiiu. C. F. 
 H. and Mr. Adolph Laue, collectors. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 389 
 
 none of the inner whorls are preserved, the umbilicus was evi 
 dently deep and comparatively narrow, a characteristic caused by 
 the broad dorsum, small number, and very rapid increase in size 
 of the volutions. 
 
 The costse, or pilre, as I prefer to call them, cross the abdomen 
 without interruption, and not more than one in four reach the 
 umbilical edge, where they develop large, coarse tubercles and dis 
 appear. An outline or section of the whorl would curve like a 
 Roman arch, the abdomen being rounded far down on the sides, 
 the sides flattened only when near to the umbilical edge, which is 
 very abrupt, and in the whorl examined measured about four fifths 
 of an inch from the edge to the side of the preceding whorl. The 
 base of the whorl in its broadest part, from edge to edge, measures 
 three and two fifths inches, and its height, from a line connecting 
 the umbilical edges, three and one fifth inches. The septa are not 
 sufficiently well preserved to afford an accurate description. 
 
 No. 2 is a species of the Ligati group. The ligature-like depres 
 sions constricting the whorl are plainly visible, and the form 
 reminds one of Ammonites semistriatus D Orb., at least the general 
 aspect of the last whorl somewhat resembles that species. 
 
 No. 3 appears to be identical either with Ammonites Peruviamis 
 Von Buch, or Ammonites acutocarinatus Shumard. The fragment 
 is very much compressed, and the true characteristics of the ab 
 domen obliterated. 
 
 No. 3 a is a fragment of a young specimen of No. 3. This is 
 not compressed, and shows the prominent keel and broad pilao of 
 this species much more plainly. 
 
 Locality : Maroi m, in cretaceous beds. C. F. H. and Mr. Adolph 
 Laue, collectors. 
 
 Ammonites Gihbonianus Lea. 
 
 No. 5 is probably the young of No. 6. The piloe (ribs or costae) 
 make their appearance on the first quarter of the second whorl. 
 
 After this first period the whorls are obscured more or less until 
 the last quarter of the fifth whorl. From this time until the com- 
 
390 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 pletion of the sixth volution there are large tuberculated piltc 
 which alternate with others of lesser height and thickness, though 
 in a very irregular manner. The larger pilse begin to lose their 
 greater proportional height on the latter part of the sixth whorl, 
 and appear about to assume the same form as those of No. 6, de 
 scribed below. 
 
 The dorsum of the sixth whorl is much broader than the abdo 
 men ; the umbilical edges are rounded and the sides slope evenly 
 to the base of the keel. The pilse have squarely cut geniculse, 
 probably tuberculated on the shell, and which bend forward on to 
 the abdomen and terminate close to the keel. 
 
 The keel is very prominent, thin, and sharp, and the sides be 
 tween the pilse are evenly rounded upon the edge of the abdomen, 
 in those parts not affected by compression. 
 
 The septa were too obscure to be observed with any approach 
 to accuracy. 
 
 No. G. The keel is very nearly perfect, and shows to the fullest 
 extent its great breadth and the entirely external position of the 
 siphon. In this respect it resembles Nos. 5 and 7, in both of 
 which the siphon is not present at all in the internal casts of the 
 whorl, but disappears with the removal of the shell. 
 
 The latter part of the sixth and the first half of the seventh 
 whorls are exposed, and tolerably well preserved. The umbilical 
 edge is rounded, and the sides slope evenly to the base of the keel. 
 The umbilicus itself rather deep. The abdomen is not quite so 
 broad as the dorsum, measured from edge to edge. The pilae are 
 depressed on the umbilical border, but the geniculoc are slightly 
 more prominent, but not tuberculated, and bend forward on to 
 the abdomen, terminating near the keel. 
 
 They continue to remain straight until near the second quarter 
 of the seventh volution. Here a double curvature begins to be 
 apparent. The lower part bends forward over the umbilical edge 
 with a salient curvature, and is continued by a reentrant curve, 
 which, also, takes a forward direction over the edge of the abdo- 
 
PROVINCES OF SEEGIPE AND ALAGoAS. 391 
 
 men to the base of the keel. The pilso lose something of their 
 former prominence near the geniculec, and conform more decidedly 
 to the curvature of the sides of the whorl. 
 
 The keel itself on the seventh volution measured nearly one 
 half of an inch, and the whorl nearly two inches. 
 
 No. 7. The lai gest of this lot hardly reaches beyond the sixth 
 whorl. Though differing considerably at first sight, they are really 
 still younger specimens of No. 6 than No. 5, with the casts of the 
 pilse better preserved. The umbilical edge slopes sharply inwards, 
 the sides incline outward to the edge of the abdomen, and then 
 slope with a reentrant curve to the base of the keel. The envel 
 opment extends only so far as to cover the abdomen and perhaps 
 the tubercles on the geniculee. The pilse conform to the curves 
 of the umbilical edge and then rise gradually to greater promi 
 nence and acquire tuberculated geniculse upon the edge of the 
 abdomen, with extensions which raise folds upon the abdomen 
 reaching nearly straight across to the base of the keel. The keel 
 itself is very prominent, and possesses the same remarkable thin 
 ness and prominence obsei ved in No. G. All the specimens, with 
 one exception, have the pilee evenly developed and equally promi 
 nent, but in this one the same alternation of large and small pilse 
 may be observed as in No. 5. 
 
 Thus there can be little doubt of all four of these varieties 
 belonging to one and the same species. For the specimen last 
 described differs from all other specimens of No. 7 only so far as it 
 agrees with No. 5, namely, in the alternation of the pilee ; and 
 No. 5 differs from it only in those characteristics which it shares 
 in common with No. 6, namely, untuberculated genicuhe much 
 less prominent than in No. 5, and rounded umbilical edges with 
 sides which slope evenly to the base of the keel as in No. 6. 
 
 The general aspect of a section of one of No. 5, the amount of 
 envelopment, and the outline of the whorl, closely resemble Mar- 
 cou s figure of Ammonites GMonianus, found in Texas, and there 
 is a faint resemblance to Lea s original, but miserably inadequate 
 
392 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 figure and description of a fragment of the same species found in 
 New Grenada. 
 
 It may, perhaps, excite surprise that the Ammonites noticed in 
 this report are not published under different generic appellations 
 from those usually employed, and this indeed calls for some expla 
 nation on my part. All the genera described by me in the Bulletin 
 of the Museum of Comparative Zoology were collected from Liassic 
 beds, and their characteristics were determined by careful com 
 parisons of the }*o\mg and adult specimens throughout large series 
 of specimens. This kind of work has led to the conclusion that 
 it will not in most instances promote the knowledge of palaeon 
 tology to describe isolated genera in other formations. When a 
 series of connected generic groups can be delineated standing in 
 their serial relations to each other, and illustrating natural laws 
 of arrangement, or when the diagnoses of new genera, even 
 though isolated, may indicate important facts of stratigraphical or 
 geographical distribution, the readjustment of the older and more 
 comprehensive names and groups may become imperative. The 
 Brazilian specimens, however, evidently belong to new genera, ac 
 cording to my views of the relations of species among the Ammo 
 nites, but for the present any change of their well-known names 
 seems unnecessary. 
 
 All of the Brazilian Ammonites are either identical with, or so 
 closely allied to, species already described from the Texas beds by 
 Roemer and others, that they cannot be safely separated. 
 
 The presence of such well-characterized species as Natica prce- 
 lonc/a, Ammonites Peruvianus, and perhaps of other species on the 
 western as well as eastern side of the Andes-Rocky Mountain 
 chain in Brazil and Texas, indicates the connection between these 
 slopes, either across the Isthmus or west of Brazil, while a creta 
 ceous ocean still flowed over the whole of the northern portion of 
 South America. These facts, when considered in connection with 
 the discovery of a fossil Ananchytes on the Isthmus, as recorded 
 by Mr. Alexander Agassiz, have a direct bearing upon a very im 
 portant question. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 393 , 
 
 The expeditions of the Coast Survey, as is now well known to all 
 naturalists, have established the fact of a remarkable similarity be 
 tween the present deep-sea fauna and the species of cretaceous 
 genera ; and it has been shown that the sui face or littoral animals 
 were more or less represented by identical or closely allied species 
 of the Pacific side of the Isthmus. Thus the question has arisen 
 whether or not the closely allied or identical forms are the de 
 scendants of Gulf species, which, having migrated through some 
 ancient channel subsequently closed by the rise of the neck of 
 land forming the Isthmus of Darien. Of course the first step to 
 wards the solution of this problem would be to prove the existence 
 of a channel affording a free passage to marine animals at some 
 preceding period. This gives great interest to facts like the above, 
 which appear to confirm the conclusion of Mr. Alexander Agassiz, 
 that during the cretaceous period the Gulf of Mexico and the 
 Pacific Ocean were really continuous seas. 
 
 I heard it frequently reported, that large fossil turtles 
 were found at Maro im. One of them I saw at Bahia, and 
 it was nothing but a huge Septarium. 
 
 The Maro im limestone is evidently upper cretaceous, and 
 belongs lower down in the series than the white flaggy lime 
 stone of Sapucahy. I saw.no trace of red sandstone like 
 that of Estancia, which may perhaps underlie the Maro im 
 limestone, though it may be wanting, as I have seen it no 
 where else, and it seems to be a local formation. The 
 country about Maro im is hilly, but low. The soil is very 
 rich, and a large trade is carried on in sugar, the greater 
 part of which is in the hands of the wealthy house of Messrs. 
 Schramm & Co. 
 
 The shore between the mouths of the Cotinguiba and Sao 
 Francisco Rivers is low, with a few scattered hills, and is of 
 little interest. 
 
 17* 
 
394 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The barra of the Rio de Sao Francisco is, like that of the 
 Cotinguiba, obstructed by sand-banks, and at times the en 
 trance is dangerous, even for steamers.* A hook-shaped 
 sand-spit is seen extending from the southern side of the 
 entrance. 
 
 On entering the river, the southern side is seen to be 
 swampy and for a considerable distance covered by man 
 grove-trees. The opposite side is sandy, and there are 
 some large areas covered by fine dunes of blown sand. 
 Since Halfcld s survey this has much changed its appear 
 ance, the dunes have increased in height, burying some 
 of the cocoa-trees to their crowns, and encroaching on 
 the river.* Strongly in contrast with the smooth sweep 
 ing outlines of the growing dunes are the irregular con 
 ical hills of sand which occupy a large part of the Pontal. 
 After a dune has grown, some of the coarse trailing plants 
 take root, and shrubs and little trees spring up on its sur 
 face. These protect the area over which they grow, while 
 the remainder may suffer removal by the wind, forming 
 little conical hills with tufts of vegetation on their tops. 
 The sand is very fine and of a light brown color. Among 
 the sand-hills are large flats, partially occupied by marshes 
 and ponds, and the resort of great numbers of wading and 
 water birds, cranes, plover, and the long-toed Parra Jacana. 
 
 Just about the mouth a channel called the Rio Parapuca 
 
 * Liais says that near its mouth, after the dry season, the Suo Francisco de 
 livers 2,800 cubic metres of water per second. (Bull, de la Soc. de Gcog., 5 me 
 Serie 2, p. 390.) Gardner in his Travels, p. 104, draws a very discouraging 
 view of the prospects of rendering the Sao Francisco navigable, and says that 
 here is seldom more than four feet of water on the bar. Though dangerous, 
 it is regularly crossed by the large coast steamers. 
 
 t Gardner describes similar dunes at Peba, five leagues north of the Barra 
 do Sao Francisco. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 395 
 
 leaves the main river and empties into the sea about a 
 league to the south. 
 
 From the Pontal to Penedo, a distance of about twenty- 
 five miles by the river, the shores of the river are very 
 low and flat, and there are many large islands. Ascend 
 ing some distance the mangrove swamps disappear, and a 
 growth of aninga is common, while the giant uba grass 
 sometimes covers considerable tracts. These islands and 
 the neighboring flat lands are partially wooded, but the 
 growth is neither so luxuriant nor so thick as that which 
 characterizes the Mucury and Doce in the south, or the 
 Amazonas in the north. The river-banks are quite low, 
 and large tracts are every year covered for a time by the 
 waters of the annual freshet, which, however, deposit a fresh 
 coating of mud over the surface, adding to the fertility of 
 the region. The banks are, in some places, seen to be 
 composed entirely of sand, but usually there is a superficial 
 stratum, more or less thick, of massape , or yellowish and 
 brownish alluvial clayey soil, which is very fertile. These 
 flat lands are exceedingly well adapted for cultivation, and 
 are especially suitable for sugar-cane, of which, however, 
 little is as yet planted. There are several settlements on 
 these lands, of which Piassabossu is the largest, where 
 there are several engenhos for the manufacture of sugar 
 and cachac,a. 
 
 The higher lands begin a short distance below Penedo on 
 the right bank of the river at Porteira, and consist, so far 
 as I have been able to see, of cretaceous rocks and outliers 
 of the great coast tertiary sheet. 
 
 At Aracare*, a prominent rocky point just below Villa 
 Nova, I found a series of beds much broken up, and about 
 whose stratigraphy, from my time having been occupied in 
 
396 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 a search for fossils, I do not feel quite sure. The spot is of 
 considerable interest, because rocks are found there that I 
 have not seen elsewhere. 
 
 There arc beds of a light yellowish or brownish fine 
 grained, shaly, micaceous sandstone, in which I found a 
 multitude of fossils which are almost, if not quite, undeter 
 minable. Most abundant is a little bivalve which has filled 
 some layers, but which has left only empty moulds of the 
 valves ; in addition to these are what appear to be the spines 
 of fishes and fragments of plants. There are some layers 
 of a light-colored shale, in which, however, I found no fossil 
 remains. These beds are considerably inclined, but I omit 
 ted to take an observation of dip and strike. 
 
 The shore is encumbered by great masses of a consider 
 able variety of rocks, some of which I did not see in situ. 
 Of these is a light, porous, argillaceous, warm red sandstone, 
 which resembles somewhat the sandstone of the tertiary 
 hills near Pitanga on the Bahia Railroad, and with this are 
 associated large masses of a coarse sandstone and conglom 
 erate coated by clay and oxide of iron, in which quartz and 
 agate pebbles are found, which rock also appears to be ter 
 tiary, so that I am inclined to think that we have here over 
 lying the fossiliferous sandstone and shale fragments from 
 the now generally denuded tertiary sheet. The fossiliferous 
 beds I believe to be the upper members of the series of 
 sandstones of Villa Nova and Penedo, about to be described, 
 and which I regard as cretaceous. 
 
 I found here numerous fragments of a rock with a sort of 
 oolitic structure which is very interesting ; when a fresh, 
 undecomposed species is broken, it is seen to be made up 
 of round or irregularly spherical masses of a granular brown 
 quartz, about the size of coarse duck-shot, filled in with a 
 
PROVINCES OF SERGIPE AND ALAGoAS. 397 
 
 cement of a bluish, translucent chalcedony, in which are 
 bedded very much smaller masses. These shot-like grains 
 appear to have been formed by the filling up of globular 
 cavities by quartz, for some of them are minute gcodes, 
 still hollow in the centre. Each has a thin concentric coat 
 ing of milky chalcedony. In decomposing, the cement of 
 the grains becomes white, and sometimes is first removed, 
 leaving the grains projecting. At others the material com 
 posing the grains is dissolved out, leaving a honeycombed 
 surface. It is a curious rock, whose formation I do not 
 feel prepared to explain. It looks more like a pseudomorph 
 after oolite than anything else. 
 
 At Villa Nova we find the low, rocky bluff point on which 
 the village is built composed of thick beds of a fine, sharp, 
 whitish or slightly yellowish sandstone without fossils, with 
 well-marked oblique and irregular lamination, and a strike 
 N. 50 E., a dip of 15 -20 S. 40 E. 
 
 The city of Penedo is built on the Alagoas side, at the 
 foot and on the side of a ridge which runs from the left 
 bank of the river towards the northeast. This ridge has a 
 steep slope to the southeast, while for some distance on the 
 northwestern side it is precipitous and about fifty feet high. 
 The rocks exposed on the shore and in the cliffs at Penedo 
 are of the same general character as those of Villa Nova. 
 A layer of decomposed shale or clay is seen in the cliffs 
 skirting the town, together with some thin bands of a fine 
 shaly micaceous and ferruginous rock. The whole, like the 
 Villa Nova beds, dip to the southeast at a small angle.* 
 I have carefully examined these sandstones for fossils, but 
 have seen only some badly preserved remains of plants. 
 
 * Gardner says that the sandstones incline from east to west, which is cer 
 tainly incorrect. 
 
398 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The hill at Penedo is covered by red drift-clay, between 
 which and the rock I found a sheet of quartz pebbles, min 
 gled with angular fragments of sandstone. 
 
 The Penedo sandstones are very porous, and of a fine and 
 even sharp grain, which makes them suitable for sharpening 
 tools. The stone is highly esteemed for that purpose, and, 
 owing to the want of sandstones of the same kind elsewhere 
 along the Brazilian coast, it is exported to a small extent, 
 finding its way even as far as Rio. 
 
 The same sandstone shows itself at Boassica,* about three 
 miles farther up the river, on the Alagoas side. It has here 
 the same dip and strike and oblique lamination as at Pe 
 nedo, and contains occasional layers of pebbles. 
 
 On the Sergipe side it is seen again at Coqueiro and Villa 
 Nova, and at one of these places it has been quarried for 
 building purposes at Penedo. 
 
 At Carrapixo, on the Sergipe side, the native civilized In 
 dians manufacture a very good quality of earthenware from 
 the clay of the low grounds. 
 
 The city of Penedo is quite a respectable little town of 
 some 3,000 or 4,000 inhabitants. It carries on a considerable 
 commerce in rice, corn, hides, mandioca-farinha, cotton, <fec., 
 &c., which are sent to Pernambuco and Bahia. Every week 
 a fair is held, and a great concourse of people from up and 
 down the river assembles there, bringing hides, coarse sugar, 
 pottery, tobacco, and a host of other articles, which are ex- 
 
 * This name, which is applied to a little stream and lake as well as to the 
 settlement, is of Tupi origin, and is derived from Boya, a serpent, and assica, 
 mutilated, the name of a species of serpent which appears as if it were muti 
 lated. See Tupi Dictionary and Clirestomathia da Lingua Brazilica. I do 
 not know what species is meant. The same name is applied to a lake in the 
 province of Rio de Janeiro situated between the Rio Macahe and the Rio das 
 Ostras. Die. Geoy., sub voce Boassica, 
 
PROVINCES OF SERGIPE AND ALAG6AS. 399 
 
 posed for sale in booths on the broad sand-beach bordering 
 the town. Penedo is a calling-place for steamers connect 
 ing with Bahia, and it has steam communication with the 
 river above as far as Piranhas. 
 
 The city is exceedingly well supplied with fish. Among 
 these are Tubaranas, Curimataes,* Piaus, Sarapos (Cara- 
 pus), Piabas of several species, Cachimbaus or Acaris, Pi 
 ranhas, Pirampebas, &c., &c., of which I made a large 
 collection.! 
 
 I extract from my journal some notes on the color, hab 
 its, &c., of the Piranha $ (Pygocentrus), which seem to be 
 of interest. 
 
 This species of Piranha, according to the testimony of the 
 natives, is confined entirely to the Sao Francisco and its 
 tributaries, though other species of the same genus (or Serra- 
 salmo) occur elsewhere in South America; but I cannot 
 vouch for the truth of this statement. 
 
 The Piranha of the Sao Francisco is strictly a fresh-water 
 fish, and it occurs not only below the falls of Paulo Affonso, 
 but also above them. It descends the river quite to the salt 
 water, but never goes into the sea. None of those I saw at 
 
 * I have spelled this name as I heard it commonly pronounced. Bates writes 
 it Curimatd, and so does Fonseca in his dictionary. The Tupi dictionary gives 
 it Curymald. This fish belongs to the genus Anodus, and several species are 
 very common in the Brazilian rivers. 
 
 t All the fishes collected by me on both journeys are in the hands of Pro 
 fessor Agassiz. When the above was written Professor Agassiz intended to con 
 tribute to this volume a scries of articles and notes on the fresh-water faunae 
 of the coast I explored, but illness has prevented his preparing them. 
 
 J In the Tupi dictionary the word " Piranha " is translated scissors, and most 
 writers seem to suppose that the name was given to the fish because of its scis 
 sors-like jaws. The Tupis knew nothing about this implement before the com 
 ing of the Europeans. Piranha (root Pirn, fish) is an ancient Tupi name, 
 and it was doubtless afterwards applied to the scissors because they bit like the 
 Piranna. 
 
400 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Penedo were more than twenty inches long, but the fisher 
 men say that they sometimes grow to the length of two feet. 
 The upper half of the body and head of a specimen just 
 caught were, when seen from above, of a dull and rather 
 dark bluish-gray or lead color. The lower half of the head 
 and body have as a ground color an opaque white, over 
 which is a wash of a clear bright gamboge-yellow, deepen 
 ing in some spots to a rich orange. All the young Piran 
 has I saw had the belly of a rich red-orange or blood 
 color. On the sides the yellow is sometimes shaded with 
 light gray, and the yellow or orange tint extends itself up 
 ward in irregular lines over the dark gray of the back. 
 The pectorals are light orange-yellow, the tint deepening in 
 the lower and middle part of the fin. The anal fin is on 
 the thickened base grayish, washed with a light clear yellow 
 tint. The border is light purplish-brown. The dorsals and 
 caudal are a dark, dull bluish-gray. The iris is pearly white, 
 with a cloudy patch of black above and below the pupil.* 
 During the freshets the water overflows the low grounds and 
 swamps, and the different kinds of fish leave the river proper 
 and enter the lagoons and quiet overflowed places to spawn. 
 
 * Gardner (Travels in Brazil, p. 96) describes the Piranha fish as follows : " It 
 is commonly about a foot in length, but sometimes it is as much as two feet 
 long, being very much compressed laterally and very deep; the back is of a 
 dark brownish color, and the belly yellowish- white, both being thinly marked 
 with reddish spots ; the lower jaw projects a little beyond the upper, and both 
 are armed with about fourteen flattish triangular-shaped teeth, upwards of a 
 quarter of an inch in length and very short." Thrs description appears to 
 refer to a different species from the one I describe. Humboldt in his Travels, 
 Vol. II. p. 167, speaks of the Piranhas, or Caribcs of the Orinoco, as having 
 "the belly, gill-covers, and the pectoral, anal, and ventral fins of a fine orange 
 hue." My specimens from the Silo Francisco still preserve their orange color. 
 The Sao Franciscan species appears to be much larger than those of the Ama- 
 zonas and Orinoco. 
 
PROVINCES OF SERGIPE AND ALAGtiAS. 401 
 
 The fishermen said that the Piranha also leaves the river 
 and chooses a shallow spot with a sandy bottom to depos- 
 ite its eggs. Stooping down, the fisherman with whom I 
 was conversing one morning took up a Piranha lying be 
 fore him, and showed me just how the eggs were laid. 
 The fish having chosen the proper spot, sweeps away the 
 sand with its tail and anal fin, so as to make a little saucer- 
 shaped depression four or five inches wide. The eggs, of 
 about the size of mustard-seed, are then laid in the nest in a 
 ball, two or three inches in diameter. This accomplished, 
 the mother fish takes up her position near the nest, and 
 keeps watch over the eggs until the young are hatched, for 
 the Piranha has enemies in the hungry little Piabas, which 
 swarm about in countless numbers, and from which she 
 must defend her eggs. Looking over a large lot of Piabas 
 in a canoe, I found it difficult to obtain a perfect specimen, 
 on account of the mutilation of the tail and fins. " This is 
 the Piranha s work," said the fisherman, " and the marks 
 of the terrible teeth of the monster." The larger fish of 
 the river also bear similar scars. 
 
 The laying, according to the fisherman, takes place prin 
 cipally in October, or soon after the freshets set in. Dur 
 ing this time the Piranhas are especially fierce. 
 
 One fisherman described in a vivid manner his finding a 
 Piranha watching its nest in a shallow place by the river 
 side. " I thought to catch it," said he, " and waded softly 
 into the water to put a basket over it, but I was not quick 
 enough. The fish darted at me and took a piece out of my 
 leg. Look there ! " and rolling up his pants he showed a 
 pair of crescent-shaped scars left by the fish s jaws. The 
 fishermen are often bitten, and almost every man present 
 had scars to show, either on the arms or legs. 
 
 z 
 
402 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The fishermen united in saying that just as this fish is 
 about to spawn the color of the belly changes from yel 
 low or orange to the same color as the back, but that soon 
 after the eggs are laid the original color returns. 
 
 The Piranha frequents rather the deeper parts of the 
 river, and abounds in the eddies among the rocks, but I 
 have seen it caught, as at Penedo, close in shore, where the 
 water was not very deep and the bottom was sandy. 
 
 During my stay at Penedo a poor little idiot sitting on 
 the pier, having been frightened by a cannon, fell over into 
 the river. The next morning the Piranhas caught in the 
 vicinity were found to contain portions of his body. There 
 are numerous well-authenticated cases where persons have 
 been attacked by the fish while bathing and devoured. Only 
 a short time before my visit to Penedo a young lady was 
 thus attacked and eaten. A horse fording the river slips 
 and wounds himself on a stone ; the Piranhas, attracted by 
 the blood, crowd about in great numbers, each cutting out 
 mouthful after mouthful of flesh, until, in many instances, 
 the voracious creatures have been known to devour the 
 entire animal in a few hours. They sometimes throng about 
 their prey in such numbers that they may be seen leaping 
 one on top of the other out of water, in their eagerness 
 to get at it.* 
 
 So far as I could learn, these fish appear to be particularly 
 
 * Bates speaks of the great swarms of Piranhas in the Amazonas. Hum- 
 boldt (Travels, Vol. II. p. 167) says of the Piranhas of the Orinoco: "The 
 Indians dread extremely these caribes ; and several of them showed us the 
 scars of deep wounds in the calf of the leg and in the thigh made by these 
 little animals. They swim at the bottom of the rivers ; bat if a few drops of 
 blood be shed on the water they rise by thousands to the surface, so that if a 
 person be only slightly bitten it is difficult for him to get out of the water with 
 out receiving a severer wound." 
 
PROVINCES OF SERGIPE AND ALAG^AS. 403 
 
 dangerous only during the spawning season. During my 
 voyage on the Sao Francisco I saw everywhere bare-legged 
 women standing in the water on the banks washing, and 
 that not only in sandy and quiet localities, but among the 
 rocks, as at Proprid, while I repeatedly saw men wading in 
 the water and boys bathing in the river. 
 
 The Piranha is much esteemed for food, and may ordina 
 rily be found for sale in the market at Pcnedo. 
 
 There is a species of Serrasalmo (?) found at Penedo called 
 Pirampcba, the name being evidently a compound of the 
 two Tupi words Piranha and peba, or the flat Piranha. This 
 species is smaller than the last, very much flatter laterally, 
 and of a silvery- white below. I did not learn that it pos 
 sessed any of the voracious propensities of the Piranha.* 
 
 Between Penedo and Propria,f a distance of about six 
 Brazilian leagues up the river, the Sergipe side is bordered 
 by low hills, some of which are irregular and isolated, and 
 composed of the sandstone above described. The country 
 back from the river is flat, moderately elevated, and appears 
 to be, in part at least, tertiary. On the AlagSas side, for 
 some distance above Boassica, there is a considerable stretch 
 of low meadow-land. At Morro Vermelho, on the Alagoas 
 side, cretaceous sandstones show themselves on the shore, 
 but the dip here is approximately to the N. W. 15. They 
 are covered by red drift-earth and great quantities of quartz 
 pebbles, and at Prazeres, half a mile farther up, we again 
 see the same sandstones with a dip of about 18. In this 
 vicinity the massape lands bordering the river are about fif 
 teen feet in height. 
 
 * Gardner speaks of seeing dried sturgeon exposed for sale in the market at 
 Penedo. There are no sturgeons in South America. 
 
 t This is, I believe, the correct orthography, but it is often written Propia. 
 
404 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Mandioca flourishes well below Propria, and is planted 
 as well on the alluvial river-banks as on the hillsides ; but 
 on the low grounds covered by the enchente it cannot come 
 to maturity. It is usually planted in February or March, 
 and is generally fit for use in a year, when planted on the 
 uplands. It is also cultivated to a considerable extent on 
 the sloping river-banks. Just before the enchente it is 
 pulled up, and is eaten sometimes when not more than 
 six inches long. The river begins to rise in October. 
 
 One mile below the town of Propria, on the Sergipe side, 
 is a small hill called the Morro do Chaves, or, as Halfeld 
 has it, Morro do Eusebio. This hill, which has rocky sides 
 toward the river, I examined in company with my friends 
 Drs. Brunet and Lacerda. The rocks composing it consist 
 of a series of limestones, conglomerates, shales, and sand 
 stones, the whole of which have a strike E. 15 S., and a dip 
 of about 20 to the N. 75 E. The lower strata consist of 
 thick beds of limestone, calcareous sandstone, and conglom 
 erate, some of the layers of which are made up of shells. 
 These beds are well exposed on the side of the hill nearest 
 the town. 
 
 This limestone is usually more or less arenaceous, and often 
 contains grains and pebbles of the underlying metamorphic 
 rocks, so as to form a calcareous conglomerate. In some 
 places it is crystalline and metamorphosed. It is used to 
 some extent for burning into lime, but the most of it is very 
 impure. 
 
 The shells so exceedingly abundant in some parts are 
 lamellibranchs, about one half to three quarters inch in 
 diameter, and with very thick valves. I fear that even the 
 genus is wholly unrecognizable. Overlying these limestone 
 beds are shales, rather soft, not well laminated, calcareous, 
 
PROVINCES OF SERGIPE AND ALAG6AS. 405 
 
 micaceous, and of a greenish color, in which I found a large 
 quantity of bones of teleostean fishes, and an impression 
 which had the outline of the tooth of a Notidanus. A fur 
 ther examination of these shales may reveal something of 
 more interest. 
 
 Going around the hill we come to a break which extends 
 from the top of the hill to the bottom, and in which we find 
 a confused mass of broken pieces of sandstone. 
 
 This series of rocks I have referred to the cretaceous. 
 Farther up the river we see no mesozoic rocks. 
 
 On Halfeld s map a note says that in the Morro do 
 Euscbio there is a camada de cal em gneiss granite, a bed of 
 limestone in gneiss-granite. This is not correct, as is seen 
 from the above description. The limestones, shales, &c., 
 overlie beds of clay slate, which I saw very badly exposed 
 on the river-side. There appears to be a distinct slaty 
 structure to the rock, but I did not see enough of the for 
 mation to enable me to make out the bedding. I have seen 
 no rock like this elsewhere on the coast of Brazil. 
 
 Proprid itself is built on the river-bank, and has in front 
 of it ledges of flaggy gneiss and mica slate, with a north 
 west strike* and a dip toward the southwest. It is only 
 a small village of little importance, because the lands of 
 the vicinity are not very productive, yet it exports some 
 cotton, hides, &c. A little lake near the town is said to 
 abound in fish, and to furnish a considerable revenue to 
 the Tillage government. Opposite Proprid the lands are 
 
 * This strike is almost at right angles with that of the rocks of the Scrra 
 do Mar, but it corresponds to that of the gneiss hills between Cape Corrientes 
 and Tapalqucn, and of the gneiss near Montevideo, in which the direction is 
 W. 25 -30 N. D Orbigny thinks that the system of upheaval by which 
 these extra-Brazilian rocks were disturbed was very nearly as old as that which 
 disturbed the gneiss of the Serra do Mar. 
 
406 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 very low and flat over a very large area. Behind them are 
 seen the tertiary chapadas lying a few miles off. When 
 the enchcnte comes on the river rises some fifteen to twenty 
 or more feet here, and converts the low grounds just 
 spoken of into a magnificent lake-like expansion. 
 
 Between Proprid and Sao Braz the land is still low, but 
 the hills are of gneiss and rounded. The rocks frequently 
 show themselves along the river-side, the dip being usually 
 up stream, though at Agua Comprida I observed a dip down 
 stream. 
 
 The hills in the vicinity of Lagua Comprida are 300 - 400 
 feet high, with rounded outlines and steep sides, which are 
 well wooded. Farther up the river-banks grow higher on 
 the Sergipe side. There are very irregular cliffs of gneiss, 
 broken and rough, and almost like those of our northern 
 gneiss regions. The country in the vicinity is, generally 
 speaking, not very fertile, but, on the Sergipe side, two or 
 three miles from the river, the hillsides are highly culti 
 vated, presenting an unusual and pleasant appearance. The 
 same gneissose and mica-schistose rocks, often traversed by 
 large veins, are well seen in the bluffs. Their strike varies 
 little, and the dip is southwest, varying from nearly hori 
 zontal to forty-five degrees. The river thus follows the same 
 general direction as the strike of the rocks. 
 
 In front of Traipu, on the opposite or Sergipe side, there 
 is a range of high hills, with steep bluff sides towards the 
 river, and a long slope from it. It is composed of gneissose 
 and schistose beds, which, inclined at a rather high angle, 
 dip away from the river. Lying on the southern slope of 
 these hills is a thick bed of a compact rock, which, in the 
 steep rocky sides bordering the river, as well as in the trans 
 verse valleys separating the hills, forms a line of bluffs. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 
 
 407 
 
 Traipti is a little village of no importance on the AlagSas 
 side. At this place a specimen of specular iron ore, said 
 to have come from near Pao d Assucar, was presented to 
 me. It is equal in richness to the Swedish iron ores, and 
 if in workable quantity would be very valuable. 
 
 TEAIPU FROM NEAR MARCACAO. 
 
 At Marcaqao flaggy gneiss, very regular in stratification, 
 is exposed on the shore. Strike N. 35 W. Dip 35 S. W. 
 A short distance above Traipu the same rocks with a north 
 east dip are seen on both sides of the river. 
 
 The country above the Serra de Tabanga becomes every 
 mile more and more rocky and barren, while the vegetation 
 becomes more sparse, consisting of small bushes with a great 
 abundance of bromeliaceous plants of several species. Of 
 these last there is a very common one known as macambira, 
 with narrow leaves, bearing along their margin long hooked 
 spines arranged wide apart. This plant furnishes a strong 
 fibre, and during the dry season its roots contribute largely 
 to support the cattle. 
 
 Several species of Cereus, some of which attain the size 
 of large trees, grow thickly over the rocky hillsides, and 
 
408 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 form one of the most characteristic features of the vegeta 
 tion. One of these is the Cliique-Chique, so common on 
 barren and arid lands in the interior, and so often described 
 by travellers. 
 
 One never sees the Chique-Chique in the zone of forest 
 and fertile lands bordering the coast, but as the forests dis 
 appear it makes its appearance, and sometimes grows to 
 the height of forty feet. Melocacti* also abound, together 
 with opuntias.f With these are associated many species of 
 croton and of sapotaceous plants. There is a species of 
 JBignonia, B. Tecoma Mart., very common on the hills and 
 shores, sometimes growing to a height of forty to fifty feet, 
 with a trunk three or four feet in diameter. Its foliage is 
 very light and the flower is yellow. 
 
 The Joazeiro (Zizyplms Joazeiro) , a very beautiful tree 
 with dense heavy green foliage, is quite frequently seen on 
 the river-banks, together with species of Azolla, Mimosa, 
 Geoffroya, Peltophorum, &c. 
 
 Along the river here the meadow lands have become very 
 much narrowed down, and occupy only little bays among 
 the hills, though sometimes these are only shut off from the 
 river by beaches, and form lagSas, numbers of which we find 
 all along the river. The elevated country is of use only for 
 pasturing. The thinness of the soil is not the only reason 
 why the vegetation is so sparse and peculiar. The surface is 
 yearly burned over during the dry season, when the cattle 
 
 * One of the most interesting of these curious cactuses is a large species 
 described by Gardner under the name of Mclocactus Ilookerianus. 
 
 t Gardner speaks of the occurrence of a species of cochineal insect on the 
 leaves of these plants. Dr. Brunei called my attention to the same fact, and 
 assured me that there was no reason why the insect should not be successfully 
 cultivated. 
 
PROVINCES OF SERGIPE AND ALAGOAS. 409 
 
 feed on the cactuses and the roots of the bromeliaceous 
 plants. 
 
 The meadow grounds are sandy and not very fertile, but 
 during the time when the river is low, mandioca, rice, beans, 
 cotton, niandubi * (Arachis hypogcea Linn.) or the pea 
 nut, are planted, and in front of the numerous fazendas 
 cle gado, or cattle estates, and villages the green plots of 
 these plantations bordering the river appear in cheerful con 
 trast with the hills behind, which are scattered all over with 
 loose blocks of stone, and bristle with cactuses. 
 
 The country in the vicinity of Curral das Pedras is roll 
 ing, and not very high. 
 
 At Jacobina there is a lagoa where much rice is planted. 
 One man makes more than 1,000^000 per year by renting 
 it for this purpose. 
 
 At Intaes there is a group of high hills near the river, 
 while others are seen in the distance on both sides, but the 
 country continues with the same general character nearly to 
 Pao d Assucar. At Lagoa Funda, AlagOas side, strike N. 
 30 E., dip. N. W. 3 - 40.f Flaggy gneiss. 
 
 Looking ahead from this point there is a very fine view, 
 the country still presenting the same low flat lands, but be 
 yond we see the irregular tops of higher hills, which appear 
 to be of a different structure. At Cajueiro (Alaguas) strike f 
 X. 36 E., dip 30 N. W. 
 
 Passing the island of Sao Pedro the river narrows while 
 
 * This curious and well-known plant is a native of Africa. In Brazil it goes 
 by the name Amendoim, Mandubi, and Afandubim. The second form appeai-s to 
 be the correct one. The name is of African, not Tupi, origin. This plant 
 is largely cultivated in Brazil, and is used for the making of sweetmeats. It 
 produces an oil employed for burning and the manufacture of soap. 
 
 t Both these observations were made as we passed close along shore in the 
 steamer, and were carefully taken. Hulfeld makes the strike northwest. 
 VOL. i. 18 
 
410 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the hills become much higher and excessively rocky, their 
 sides being covered with blocks of stone. The rock is still 
 gneiss, traversed by many veins. 
 
 A little beyond, the view opens out, and one looks up to 
 the town of Pao d Assucar, built on the low ground bor 
 dering the shore, with a wall of high hills in the background, 
 and over them, seen away in the distance, are the blue tops 
 of the Serra do Pao d Assucar. 
 
 Pao d Assucar is a considerable village built on the Ala- 
 goas side of the river, on a high, narrow strip of alluvial 
 ground. In this vicinity these alluvial lands have a con 
 siderable extension, and are well planted. 
 
 LOOKING DOWN THE RIVER FROM PAO D ASSUCAR. 
 
 The rocks at Pao d Assucar are gneiss, but siliceous and 
 flaggy. At the upper end of the village there is a high, 
 sugar-loafed hill, which gives the name to the locality. This 
 is one of a number of hills which together form a range 
 running southeast, crossing the river here, being continued 
 on the opposite side. The gneiss is vertical, and has a strike 
 of N. 40 W. It is seen in the same position in a prom 
 inent bluff opposite the town on the Sergipe side. Between 
 the town and the hills behind are some quite extensive 
 lagoas, on the borders of which rice is planted. I ascended 
 the Pao in company with Mr. Brunet. From the top is 
 one of the finest views I have seen in Brazil. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 
 
 411 
 
 In the vicinity of Pao d Assucar are very many cattle 
 fazendas, to which belong great numbers of cattle, and on 
 which cheese (requeijao} and hides * are manufactured. 
 
 CATTLE FAZENDA AND GARDEN NEAR PAO D ASSUCAR. 
 
 Immediately above Pao d Assucar the river narrows 
 much, and the banks become still more steep and craggy, 
 varying in height from two to three or four hundred feet. 
 The hills bordering the river seem to be conical or dome- 
 shaped, but are really in most cases the ends of ridges 
 cut across obliquely by the river. 
 
 In proportion as the river narrows, the alluvial deposits 
 on its banks grow higher, and at Entre Montes, an exceed- 
 
 * The hides are tanned in stone vats, of which each fazenda possesses one 
 or more. The process is as follows : The hides are cut in two lengthwise and 
 soaked in the river. They are then placed in vats in alternate layers of hides and 
 the ashes of Barauna (Melanoxylon) or Catinga de Porco. At the end of three 
 days the hair is removed. They are then scraped and placed once more in 
 the vats with the bark of Angico stamped with water. This bark is removed 
 three times in the tanning of the best skins, the bark remaining each time a 
 fortnight on the hides. The process is finished by washing the skins and ex 
 tending them over poles in the sun, when they are pressed into boxes for expor 
 tation. Many goat-skins are prepared in the same way. At Penedo I saw a 
 shoemaker blacking leather by rubbing it over with mud from the bottom of 
 a pond near by. He assured me that he used no other preparation, and that 
 the mud alone gave a rich black color. 
 
412 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ingly picturesque little village built in a notch among hills 
 some seven hundred feet in elevation, these flats are, I should 
 judge, at least fifty feet high. At Allegria and Coleite 
 the rock is red, very homogeneous, and compact. The river 
 
 VIEW NEAR ALLEGEIA. 
 
 becomes so narrow that in some places it is not more 
 than four hundred feet across, while the precipitous rocky 
 walls, three or four hundred feet high, make the scenery 
 of this part of the river exceedingly fine and beautiful. 
 
 Porto das Piranhas is a miserable little village built on a 
 sand and gravel bank on the Alagoas side of the river, at 
 the foot of the hills, which rise with steep sides to a height 
 of about seven hundred feet, if not more. The river here 
 is somewhat tortuous, with rocky shores and occasional 
 rocks and ledges. At the upper end of the village is a sugar- 
 loafed hill which I found to be composed of gneiss, strike 
 N. 20 W., dip vertical. 
 
 I had been told that the country on top of the hills 
 is flat. I climbed in a miserable rain-storm to the sum 
 mit of the steep slope behind Piranhas, a height of wcllnigh 
 seven hundred feet, from which in the intervals between 
 
PROVINCES OF SEKGIPE AND ALAGSAS. 413 
 
 the showers I was able to look over the whole country. 
 Instead of finding a chapada or absolute plain like that of 
 the tertiary regions, as I had been led to expect, I found the 
 general surface country remarkably even, but consisting of 
 
 VIEW LOOKING UP TOWARDS PIRANHAS. 
 
 a great number of very low ridges whose summits all came 
 to nearly the same level. In the distance on both sides 
 of the river serras or short chains of high, irregular hills 
 were visible. The whole seemed to be formed of gneiss 
 and other metamorphic rocks. The surface of the slope 
 and top was thickly strewn with blocks, usually angular, and 
 of the same material as the surface on which they lay ; how 
 ever, I did observe a number of boulders of a red syenite lying 
 on gneiss ; but they could not have come from far, for I saw 
 the same material in place quite close to the edge of the 
 river valley above. Soil there was scarcely any, and the 
 red clay and pebble sheet were absent. Rounded boulders 
 were abundant. It is not possible that these boulders could 
 have been the result of decomposition, for this action has 
 obtained here to only a very slight extent. My stay at 
 Piranhas was necessarily limited to a few hours, which I 
 
414 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 very much regretted, for I have seen no region on the 
 Brazilian coast where polished and scratched surfaces would 
 be more likely to be preserved than here. 
 
 Halfeld gives it as his opinion that gold may be found 
 in the vicinity of Porto das Piranhas, but I do not consider 
 it probable. The same author describes the country from 
 Piranhas to the celebrated falls of Paulo Affonso as being 
 remarkably even, and composed of granite, which is, I sup 
 pose, compact gneiss, or, in part at least, syenite ; the rock 
 over which the falls precipitate themselves being of the 
 same material. My good friend Mr. Franz Wagner of Ba- 
 hia, who tarried behind our party to visit the falls, said 
 that the rock was sandstone. Burton speaks of sandstone 
 as occurring there, and of a conglomerate at the rapids of 
 Itaparica, a short distance above the falls. 
 
 On the road to the falls from Porto das Piranhas is a 
 serra called Serra d Olho d Agua, where occurs a sandstone. 
 This rock, according to the reports of some friends who 
 visited the falls shortly after I left Piranhas, is white and 
 much denuded. Halfeld represents it as dipping irreg 
 ularly from the hills northward and northwestward, and 
 he indicates a locality where sandstone occurs at the Ca- 
 choeira Cancamunhi d Acima. 
 
 The Falls of Paulo Affonso are situated about 56 leagues, 
 or 168 miles, from the sea.* I have not been able to visit 
 them, so I translate a few selections from Halfeld s descrip 
 tion of the cataracts and of the neighboring country : f 
 
 "The first fall, 44 palms 6 inches [about 33 feet] in height, 
 throws itself into a basin garnished by granitic rocks almost 
 perpendicular, and sometimes even overhanging the river; from 
 
 * Liais says 300 kilometres. 
 
 t I quote from his Explora^ao do Rio de Sao Francisco. 
 
PROVINCES OF SERGIPE AND ALAGOAS. 415 
 
 this basin the river makes a sharp turn to the left at a right an 
 gle, and precipitates itself between steep rocks into the bottom of 
 an abyss 66 palms and 1 inch in height, transforming themselves 
 in consequence of this leap apparently into milk foam, casting and 
 
 CACHOEIKA DE PAULO AFFONSO.* 
 
 blowing up, similar to the explosion of a mine, great masses of 
 water into the air, which are turned into a vapor that rises still 
 higher Transferred by this fall into a river of milk, the wa 
 ters precipitate themselves in great billows and waves, and be 
 tween towering masses of granite, beating at a right angle against 
 the left bank of the river. This side consists of a native granite 
 rock, which is 365 palms [about 250 feet] high above the surface 
 of the water, and having still 120 palms of depth, f 
 
 * I regret not to be able to give a better view of these falls, but the photo 
 graph from which it was drawn was not a good one. I am assured, however, it 
 gives a better idea of the falls than any other sketch yet published. 
 
 I " Resscrre entre deux immenses murailles de pierre, il coule d abord en 
 torrent et sur un fond dont la de clivite accroit la vitesse, puis tout a coup il se 
 precipite en trois chutes conse cutives dont la hauteur re unie et de 84 metres. 
 
416 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 " The impetus with which the waters precipitate themselves 
 against this wall makes them constantly ascend and descend from 
 the point of contact with the rock. On the right they descend 
 in a right angle to the bed of the river below ; but to the left, as 
 they have no way of egress, they produce, in consequence of their 
 advancing and retiring movement, a come-and-go like the waves 
 of the sea on the shores, from which has resulted, for thousands 
 of years up to the present time, not only the wearing away of the 
 rock, and the formation of a little bay, but even of a cavern in the 
 rock, which is 444 palms in length, and whose mouth is 80 palms 
 high and 40 broad, divided in the interior into two great halls, the 
 dwelling-place of thousands of bats, and for this reason called 
 Furna dos Morcegos 
 
 "The rock in which ihisfurna is formed, as well as in all the ex 
 tension of the Cachoeira, is of the hardest granite, of a fine grain, 
 and in truth it is incomprehensible how it should be possible for 
 the waters to form such a cavern in a rock of so great hardness. I 
 am inclined to attribute this fact to the circumstance that the 
 granite in the direction of the cavern, as well as in that of the 
 river, from the mouth of the cavern below to the Riacho da Gan- 
 gorra, presents many veins of calcareous spar, of flesh-colored feld 
 spar and quartz, which have a thickness of to 5 inches. The 
 granite by the sides of these veins is less hard, and sometimes is 
 decomposed and saturated with muriate of soda [salt], in such 
 abundance that those living near the fall mine this stone on a 
 
 small scale to extract the salt I am inclined to think that 
 
 the circumstances now indicated with respect to the ready decom- 
 
 La derniere de ccs chutes, la plus grande des trois n a pas moins de 10 metres 
 
 d altitude La compression de 1 air a la surface des eaux apres la chute est 
 
 telle, qu une pierre lancee avec la plus grande force nc pout re sister au vent 
 resultant, de sorte quo sa vitesse est ane antie apres un parcours de 6-7 metres. 
 Cctte particularite* a rcpandu, parmi les habitants des environs 1 opinion que le 
 lieu de la cascade est enchante ." (Liais, Bulletin de la Soc. de Gtog., 5 me Serie, 
 XI. pp. 390-392.) 
 
PKOVINCES OF SERGIPE AND ALAGOAS, 417 
 
 position of that rock, over the breadth which comprises all the 
 veins mentioned, probably determined the excavation of the cav 
 ern below, a circumstance which would have given place to the 
 formation and present existence of the falls, whose bed is really 
 excavated in the rock, since, for a great distance on both sides of 
 the steep banks of the river, the soil presents a plain without hills 
 or serras, which could have determined the cataract of Paulo 
 Affonso. 
 
 " From Paulo Affonso down the river to Porto das Piranhas, 
 the waters of the river are narrowed and run with many falls 
 between steep rocks of 350 to 800 palms in height, which, 
 with the exception of a very few places, as at Porto do Salgado, 
 Monto Escuro, Sitio Novo, &c., are inaccessible to allow of descent 
 to the river. These cliffs are commonly called here Talhado [or 
 Talhadao], since the width of the river is generally reduced to a 
 few hundreds of palms and sometimes less, as in the falls of the 
 Garganta, where the river is only 85 palms wide, and runs like 
 a mill-race between perpendicular walls of rock 350 palms in 
 height." 
 
 The falls of Paulo Affonso are of the same class as those 
 of the Jequitinhonha at Salto Grande. In both cases a large 
 river just before it reaches the edge of the plateau, and 
 flowing a broad stream with a wide flood plain, dashes first 
 down a slope forming a series of rapids, and then, a few miles 
 farther on, contracting suddenly, plunges in a magnificent 
 series of cascades and rapids into a narrow valley, in which, 
 with a swift and rock-impeded course, it descends to the 
 coast plains, where it spreads out widely and flows calmly, 
 proudly on to the sea. The river is, so to speak, strangled 
 in the Estreito, or Talhadao. As the rocks of the regions 
 cut through by the falls of Salto Grande and Paulo Affonso 
 are both crystalline and highly inclined t it is manifest that 
 
 18* AA 
 
418 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 their retrocession must have operated in a very different 
 manner from that of the falls of Niagara, and that the time 
 occupied by the excavation of the gorge below the Brazilian 
 falls cannot be estimated according to the same rule. I 
 should judge that the falls of Paulo Affonso are finer than 
 those of Sal to Grande at ordinary times ; but when the 
 Jequitinhonha is swollen the Salto must be a grand sight. 
 Both these falls, as navigation on the coast and on the rivers 
 becomes more prompt and commodious, will erelong be 
 come well known to travellers. Burton, who visited Paulo 
 Affonso in 1867, says, that if Niagara be the monarch of 
 falls, Paulo Affonso is the king of rapids. Liais thinks that, 
 seen close at hand, the latter exceed the former in magnifi 
 cence.* 
 
 Halfield, speaking of the Riacho da Yaca, says : f "At its 
 mouth on the south side of the river is the Lagoa da Pedra, 
 where I found the fossil bones of a mastodon. The lagoon, 
 which consists of a concavity or basin, is surrounded by great 
 cliffs of the said rock, and is seventy paces in length, ten 
 in width, and ten to twenty-five palms in depth. It was full 
 of earth, sand, and gravel in beds. Of these the lower about 
 twenty years ago contained the bones of a mastodon. The 
 neighbors, residents of the Lagfia da Pedra, had commenced 
 
 * " Vue a distance la cascade de Niagara 1 emporte done en magnificence 
 sur celle de Paulo Affonso, mais dc pros, 1 avantnge est pour le San-Francisco, 
 dont les eaux furieuses se rclevent avec plus de violence et forment unc serie 
 d immenses vagues chargecs d ecume. L cffet de cos grandes vagues, d ou sort, 
 comme de la chute elle-meme, une gigantesque colonne de vapeur, ajoute a la 
 splendour du spectacle, et la force expansive de 1 air que les eaux, dans cet 
 etroit canal, entraincnt ct comprimcnt au pied dc la chute, produit une sorte 
 d ouragan dont la puissance contribue a accroitre 1 extension de cette immense 
 colonne de poussiere aqueuse." (Liais.. Bulletin de la Soc. (Jeog., 5 me Serie, XI. 
 p. 391.) 
 
 t Description of the 328th league. 
 
PROVINCES OF SERGIPE AND ALAGoAS. 419 
 
 to dig out this cavity, in order that it might be made to 
 serve as a reservoir for the water of the rains, and as a 
 drinking-place for cattle ; there then appeared at the cutting 
 the bones of a mastodon, which they threw outside the basin, 
 but the intended excavation was never completed, and two 
 thirds of its length remained full of soil in which, at the 
 head of the ancient opening still appeared the points of 
 bones of large dimensions." From this vicinity, a few years 
 ago, an immense collection of bones, teeth, &c., of masto 
 dons were collected and sent to Rio, where they may now 
 be seen in the Public Museum. 
 
 " The country for a considerable distance around the hill 
 mentioned presents an extensive plain, over which are 
 found dispersed thousands of enormous loose rocks of gran 
 ite, and sometimes superimposed one on the other, and set 
 sometimes only upon a point or very small base without 
 falling." 
 
 I have already remarked the character of the country 
 above the Serra de Tabanga, the want of a soil, the way 
 in which the surface is strewn by blocks of stone, and also 
 the little amount of disintegration which it has suffered . I 
 cannot conceive how these loose masses could have resulted 
 from decomposition, without showing some evidence of it 
 in a much decomposed surface and an abundant soil. The 
 surface of the country looks precisely like that of our drift- 
 covered regions of the North. 
 
 Halfeld states that granite occurs as the foundation rock 
 of the country for many leagues up the river beyond the 
 falls. The term as used by him is rather too comprehensive, 
 and may comprise gneiss and syenite. 
 
 The Serra de Itaparica, which forms a long, sharp-backed, 
 narrow ridge 720 palms high, which runs from the river nine 
 
420 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 leagues above the Oachoeira de Paulo Affoiiso to the west 
 ward for several miles, is, according to the same author, com 
 posed of sandstone. " The sandstone in the top of the 
 serra is of a fine, but the lower beds are of coarse grain to 
 the base, which is situated upon granite of extreme hard 
 ness, where it forms beds of coarse gravel." * This sand 
 stone is doubtless of the same series as that of the Serra 
 d Olho d Agua. Many of the other hills of the vicinity are 
 also composed of this sandstone. 
 
 The whole country was evidently once overspread by a 
 sheet of this rock, and has subsequently suffered very exten 
 sive denudation. 
 
 The climate of the Lower Sao Francisco presents some 
 interesting features. On the coast rains are frequent and 
 plentiful, and along the shore, as already remarked, there 
 is a belt of forest. Inland, however, the climate becomes 
 more dry ; three or four months of the year June, July, 
 and August are usually without rain, and everything dries 
 up. The heavy rains, as a general thing, begin in March. 
 The river rises to a great height, and sometimes overflows the 
 high banks on which the towns of Propria and Piranhas are 
 built, inundating the lower stories of the houses. At both 
 places I saw the muddy line along the fronts of the houses 
 left by the last freshet. Such an immense volume of water 
 pouring tumultuously through the racecourse-like channel 
 between the falls and Piranhas, and, as at Propria, spread 
 ing out over and inundating the extensive low grounds, 
 converting them into a great lake, must be a grand spectacle. 
 The annual overflow takes place in October and lasts until 
 March, during which time it is raining in the highlands. 
 
 Nothing strikes one more strongly on the Sao Francisco 
 
 * Exploraydo do Rio de Sao Francisco, p. 44. 
 
PROVINCES OF SERGIPE AND ALAG(Us. 421 
 
 than the regularity with which the winds rise and blow. In 
 the morning I used to look out of my window at Penedo ; 
 all was still, and the river was without a ripple. Canoes 
 and montarias dropped down on the current, and all was 
 repose. About nine o clock * a breeze from the sea stole 
 over the water, ruffling its surface with ripple-patches ; 
 this increased gradually, until at noon a stiff wind was 
 sweeping inland ; the montarias spread their picturesque 
 sails and scudded before it up the river ; thus steadily the 
 breeze continued until well into the night, when it hushed 
 down, and a calm morning again dawned on the rippleless 
 river. This sea-breeze is perfectly regular. Boats can sail 
 up the river, but they must drift or be rowed down. 
 
 The Lower Sao Francisco below Piranhas admits of navi 
 gation by small steamers and sailing craft during the whole 
 year. In August, 1867, the river above Penedo was formally 
 opened to steam navigation by the Bahia Steam Navigation 
 Company, but Penedo has been in communication with Pcr- 
 nambuco and Bahia for some time. Through the politeness 
 of my friends, Mr. Hugh Wilson, the able superintendent 
 of the company, and Dr. Anto. de Lacerda, I was enabled 
 to participate in the fete ; and to them I owe the oppor 
 tunities I afterwards enjoyed of making the observations 
 recorded above. To Dr. de Lacerda and Dr. Brunet, who 
 accompanied me on the voyage, I am indebted for valuable 
 contributions to my note-book. I take pleasure in express 
 ing here my gratitude to these gentlemen for their kind 
 ness. 
 
 The country below Proprid is very fertile, and there are 
 large areas of rich lands admirably adapted to the culti- 
 
 * Gardner says that the sea breeze reaches Penedo about noon, which is 
 not correct. 
 
422 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 vation of sugar-cane, cotton, mandioca, &c. The region 
 above Propria is proper for grazing. The whole Lower Sao 
 Francisco forms a district of much promise. 
 
 The coast of the province of Alagoas is formed by a 
 broad belt of tertiary lands of the same general character as 
 those south of Bahia. The country embraced in this belt 
 is a vast elevated plain two or three hundred feet above the 
 sea. The western or interior part of the province is a table 
 land of gneiss, the continuation northward of the gneiss 
 district of the Sao Francisco already described. It is dry, 
 barren, and fit only for grazing. 
 
 At Jequia, some miles south of Maceio, the tertiary bluffs 
 come down to the sea line, forming a long range of cliffs of 
 a bright red color, like the Barreiras do Siri. In the south, 
 I know of only one lake which occupies a basin of denuda 
 tion in this tertiary sheet, and that is the Lagoa Juparanaa 
 on the Rio Doce ; but in the province of Alagoas there are 
 several such lakes, and it is from them that the province 
 takes its name. These lakes in Alaguas are long and nar 
 row, and have usually a northwest trend. The valleys occu 
 pied by the LagQa do Norte at Macei6, and of the Lagoa 
 do Sul a short distance to the south of Maceio, open out 
 broadly to the sea ; but a strip of sand more or less wide, 
 extending across their mouths, bars out the sea. The 
 Lagoa do Norte at Macei6 is salt. It abounds in fish, and 
 it is said that sea-turtles are found in it. The Lagoa do 
 Norte at Maceio strongly reminds one of the lakes of Central 
 New York, Cayuga or Seneca, for instance. These lakes 
 of Alagoas, as well as Juparanaa, are very deep, and their 
 basins must have been excavated at a time when the land 
 stood at a greater height than at present. 
 
 The city of Macei6 is built at the mouth of one of the 
 
PROVINCES OF SERGIPE AND ALAG^AS. 423 
 
 largest of these valleys, and at the foot of the Lag8a do 
 Korte. The entrance to the valley is barred by a wide strip 
 of recently elevated sands, covered with a low and sparse 
 vegetation, consisting of clumps of bushes, among which the 
 aroeira (Sehinus terebinthifolius~), as Gardner has remarked, 
 is very abundant, together with species of Diospyrus, Esch- 
 weilera, Uriocaulon, Marcetia, Cereus, Melocactus, &c. The 
 lagua communicates with the sea by a narrow channel across 
 this flat. The tertiary plains are almost perfectly level, and 
 come down close to the coast line. They have steep slopes 
 towards the lakes, and the same towards the sea, where, at 
 their feet, lie sand plains ; but they are precipitous when the 
 sea washes their -base. The average height of these tertiary 
 plains along the coast is about one hundred and fifty feet, 
 though they evidently rise gradually toward the interior as 
 they do elsewhere. From the top of the lighthouse at 
 Maceio, built on the edge of this plain, the eye ranges for a 
 long distance both up and down the coast, and into the 
 interior. Save the depression of the lagua, the country, 
 appears from the sea to be level ; but inland, at the dis 
 tance of a few leagues, are visible the tops of a few hills, 
 evidently of some earlier formation, but probably gneiss. 
 The slopes of the tertiary lands, as elsewhere, are fertile and 
 heavily wooded, but the higher plain is as usual exceedingly 
 dry, supporting a dense growth of small bushes, with many 
 Licuri palms. The character of the vegetation is, as I have 
 already remarked in speaking of the Rio de Sao Francisco, 
 not wholly due to natural causes ; for this whole country 
 has been repeatedly burned over, and the virgin forest de 
 stroyed. During the political disturbances a few years ago 
 large tracts were fired purposely. 
 
 A stranger will occasionally observe in the pavement of 
 
424 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the streets palasozoic rocks containing a few fossils, which 
 might mislead one if he did not know that they were brought 
 thither from North America as ballast. Large quantities 
 of the finer qualities of Rio de Janeiro gneiss are carried 
 along the coast, and used both for building and paving ; but 
 it is of so peculiar a quality that one learns to recognize 
 it immediately wherever it is found. 
 
 The upper beds of the tertiary sheet are well exposed in 
 a cliff and cutting on the side of the spur on which the 
 lighthouse is built, behind the Matriz church. The lower 
 
 O 7 
 
 bed seen consists of a soft yellowish or reddish argillaceous 
 sandstone, very loose in texture and full of quartz pebbles, 
 which are arranged in layers and lenticular masses. There 
 are some pebbles of a white substance that looks like de 
 composed feldspar. The fragments of quartz, whether in the 
 form of pebbles or sand-grains, are more or less rounded. 
 Over this lies a thick mass, very indistinctly stratified, of 
 red, pink, and white variegated soft, friable argillaceous 
 sandstone, of the same general character as that found 
 in the southern part of the province of Espirito Santo, and 
 at Para on the Amazonas. The colors of these beds are 
 very warm, and are distributed unevenly. Over all is a 
 layer of clay and soil with pebbles beneath, like the usual 
 drift coating of the tertiary plains. 
 
 The tertiary coast belt extends, to my knowledge, some 
 thirty to forty miles above Macei6, and I have seen the 
 same kind of a coast fifty miles south of Pernambuco. 
 
 The flat-topped hills in the vicinity of Cape Sant. Augus- 
 tinho are portions of this sheet, which is much worn away 
 in the vicinity of Pernambuco. The country back of Per 
 nambuco is quite hilly, and it is probable that cretaceous 
 rocks will be found there. 
 
PROVINCES OF SERGIPE AND ALAGiUs. 425 
 
 Northward of Pernambuco I have seen the same kind 
 of coast in the province of Rio Grande, where the lands 
 are precisely like those of Maceio, the valleys having the 
 same steeply sloping sides so characteristic of the tertiary 
 plains. 
 
 The city of Maceio is a town of respectable size, built 
 a short distance inland, in part on a slight elevation at the 
 base of the bluffs on the north side of the valley. On the 
 sea-shore is built the town of Jaragud, which is the port of 
 Maceio. The flat lands in the vicinity of Maceio are cov 
 ered with cocoa palms, which give to the place a very pretty 
 appearance. Dende palms are also numerous, and I saw 
 a few date palms which bore fruit. A few specimens of the 
 Assahi {Euterpe oleracea) of the Amazonas are found here. 
 
 The harbor is formed by a line of coral reefs, which extend 
 off shore at a distance of half a mile or more, and protect 
 shipping from the northeast winds ; but the harbor is not 
 well protected during southerly storms, and, what is worse, 
 it is filling up by the drift of sands over the reef. The 
 shore is a sand beach, from which long piers are built out ; 
 but owing to the sea, the shipping cannot come along 
 side, and goods are landed in lighters. The principal trade 
 of the city consists in sugar, cotton, <fcc., which are princi 
 pally sent to Pernambuco. As above remarked, the upper 
 plain appears to be very dry ; and in the vicinity of Macei6, 
 at least, is not cultivated. The slopes of the tertiary lands 
 are very fertile, and usually very heavily wooded ; as also 
 are the alluvial lands bordering the lake, which last are 
 extensively planted. 
 
 The water of Macei6 is bad ; at the time of my visit steps 
 were being taken to supply the city with good water from 
 the river Bebidouro. 
 
426 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The Companhia Bahiana was about to put two little 
 steamers on the Lagoa do Norte. 
 
 While at Maceio, I heard a Schisto bituminoso spoken of 
 as occurring in the low lands near the Lagoa do Sul. I was 
 unable to visit the locality, but, from the information I re 
 ceived, it seems to be an inflammable vegetable deposit of 
 very recent date, underlying the sands of a plain similar to 
 the lower plain of Maceio. 
 
PROVINCE OF PERNAMBUCO. 427 
 
 CHAPTER X. 
 
 PROVINCE OF PERNAMBUCO. 
 
 The Limits, Area, &c., of the Province. Its Topography, Geology, Climate, 
 Soils, &c. The Rivers. Productions of the Province. The City of Per- 
 nambuco or Recife. Derivation of these Names. Situation of the City. 
 The Stone Reef. The Port formed by it. Shallowness of Water along this 
 part of the Coast. ternambuco a Calling Station for Foreign Shipping. 
 The Pernambuco and Sao Francisco Railroad. Table of Heights along the 
 Line. Island of Itamaraca. Fossilifcrous Limestones. Fertility of the 
 Island, Cocoa Palm Groves, &c. Fernando de Noronha. Darwin s De 
 scription of the Geology of the Island. Its Dryness and Sterility. 
 
 THIS rich and populous province has been so little ex 
 plored by the physical geographer and geologist, that it is 
 not possible to give more than a very general sketch of its 
 physical features, a sketch which the writer hopes future 
 explorations may help him to fill out with needed detail. 
 
 The province comprises the northern side of the basin of 
 the Rio de Sao Francisco from the point called Pao d Arara 
 to the Rio Moxoto, a few miles above the Cachoeira de 
 Paulo Affonso. East of that point the little province of 
 Alaguas is wedged in between it and the Sao Francisco. 
 The coast line of the province is only about forty-four 
 leagues in length. The superficial area of the province is 
 variously estimated at from 4,467 (Pompeo) to 7,200 square 
 leagues of twenty to a degree.* 
 
 The province is separated from that of Piauhy by the 
 narrow plateau known under the name of the Serra dos 
 
 * Dr. Almeida, in his Atlas, makes it 5,287. 
 
428 GEOLOGY AXD PHYSICAL GEOGRAPHY. 
 
 Dois Irmaos and Serra Vermelha. From Ceara it is 
 separated by the chapadas or plateaus known under the 
 name of Serra Araripe*, while the Serra dos Cairiris Vel- 
 hos separates it from the province of Parahyba, which lies 
 just to the north. Farther on, in treating of the province 
 of Piauhy and Ceara, I shall discuss the facts we possess 
 relative to the structure of the Serras Dous Irmaos and 
 Araripe, which, as already remarked, appear to consist in 
 the main of a narrow strip of horizontally disposed tertiary 
 sandstones lying along the summit of a ridge of metamorphic 
 rock. 
 
 We have already seen from Halfeld, Burton, and others, 
 that the part of the province bordering the Rio de Sao 
 Francisco is composed of much disturbed gneissose and 
 other metamorphic rocks, here and there overlaid by patches 
 of horizontal sandstones and associated rocks, like those of 
 the Sao Francisco valley farther up. The great mass of the 
 western part of the country appears to be formed of gneiss, 
 mica-schist, &c., and these rocks here and there afford gold, 
 but, so far as I can learn, not to any very large extent. Of 
 the mountain ranges of the province next to nothing is 
 known, and they are very inaccurately represented on the 
 maps. From the description of some who have visited the 
 Serra dos Cairiris Velhos, it would appear that they were in 
 part formed of horizontal sandstones like the Serra Araripe. 
 Some of the mountains of the interior of the province are 
 of considerable elevation, though I cannot learn that any 
 exceed 4,000 feet in height.* 
 
 Pompeo f says that the coast for a width of ten to fifteen 
 
 * Some of these highlands are visible from the sea, as for instance the Serra 
 Sellada, which one sees lying a few leagues back of Cape St. Augustinho. 
 t Geographia, p. 425. 
 
PROVINCE OF PERNAMBUCO. 429 
 
 leagues is low. This portion is covered in part by quite 
 heavy woods. It is very fertile, and is called mattas. Be 
 yond this there is another zone of uneven, undulating coun 
 try covered by carrasco, and dry, but it produces large crops 
 of cotton and vegetables. The interior, which goes by the 
 name of sertdo, is very mountainous, stony, and dry, being 
 fit for nothing but pasturage. 
 
 The same author says that " the interior, principally on 
 the borders of the Sao Francisco and of the province of 
 Piauhy is subject to droughts, like the provinces of the north ; 
 still it does not present the sandy deserts and the verdurous 
 oases seen by the traveller Koster." * 
 
 The coast is low, and, generally speaking, resembles that 
 of the province of Alagoas ; consisting along the shore of a 
 more or less wide strip of tertiary beds, which, though 
 sometimes extensively denuded, form high red cliffs pre 
 sented to the sea, as in the vicinity of Cape St. Augustiiiho. 
 Underlying these arc in some localities cretaceous rocks 
 which have never been carefully examined. The tertiary 
 beds have been swept away over considerable tracts which 
 were occupied by the sea just previous to the last rise of 
 the coast, forming deep indentations in the coast line. On 
 such an indentation, now filled up with sand and alluvial de 
 posits, the city of Pernambuco now stands. I am exceed 
 ingly sorry that, though I have three different times visited 
 Pernambuco, I have never been able to examine the high 
 lands in its vicinity. Mr. E. Williamson, in a short paper pre 
 sented to the Manchester Geological Society, and published 
 in the Proceedings of that society, says, that " at Caxinga, a 
 few miles out from Pernambuco, several fine sections of the 
 sands and marls " of the tertiary " have been exposed by 
 
 * Loc. cit. 
 
430 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 land-slips ; the strata here bear such a resemblance to the 
 new red sandstone of our own districts, that it would be 
 impossible by color and appearance alone to distinguish 
 one from the other." 
 
 The lands over large areas near the coast are very fer 
 tile, and produce excellent sugar-cane and cotton. 
 
 All the drainage of the middle and western part of the 
 province is into the Sao Francisco. Some of the streams 
 are quite large ; but, as has been remarked by Cazal,* they 
 disappear in the dry season, and the same is the case with the 
 majority of the rivers of the eastern part of the province, 
 which flow into the sea. During the rains the streams, like 
 those of Bahia, swell tremendously, but in dry weather they 
 dry up entirely. The principal rivers emptying into the sea 
 are the following : the Una, whose mouth is a few leagues 
 north of the boundary line of Alagoas ; between this river 
 and Pernambuco are the Serenhaem, Ipojuca, and Pira- 
 pama. The Capibaribe empties at Pernambuco. The Una, 
 Ipojuca, and Capibaribe are quite respectable rivers, if we 
 consider their length, the Ipojuca having a course of about 
 one hundred and fifty miles ; but during the dry season 
 they disappear, except in the immediate vicinity of the sea,f 
 so that they are consequently of little or no service for navi 
 gation. The climate, of course, varies in different parts of 
 the country. Along the coast it is damp and hot, though 
 
 * "Da Villa do Peneclo athc a da Barra do Rio-Grande, em cujo inter- 
 vallo os viandantcs contain acima de duzentas Icguas, nao sahc para o rio de 
 S. Francisco um so regato no tempo da secca." Corografia Brazilica, Vol. II. 
 p. 158. 
 
 t Dr. McGrath of Pernambuco kindly sent me a tracing of a map of the 
 eastern part of the province. In a note he says : " These rivers look very for 
 midable on paper, but, as you are probably well aware, they amount to almost 
 nothing above tide-water during the dry season." 
 
PROVINCE OF PERNAMBUCO. 431 
 
 refreshed by sea breezes (Pompe o). In the interior it is 
 very hot and very dry, especially during the rainy season, 
 which lasts from March to June. Von Tschudi says that 
 it is not so oppressingly hot at Pernambuco as at Rio. The 
 population of the province is principally confined to the 
 eastern part near the coast, and to the border of the Sao 
 Francisco. In the region of the mattas there are numerous 
 and well-conducted sugar estates, which produce a large 
 quantity of sugar, molasses, and rum. Pompe o says, that in 
 1857 the president of the province reported 1,106 sugar- 
 mills, 18 of which were operated by steam and 346 by water. 
 These produced, in 1856, 18,498:000^000 worth of sugar, 
 and the same year 1,341,354 canadas of aguardente (native 
 rum) were exported, worth 616:000$000. Pompe o, writ 
 ing in 1864, places the exportation of sugar at over four 
 millions of arrobas. Von Tschudi, in 1866, estimated the 
 yearly export at over one hundred and forty millions of 
 pounds. The cotton of Pernambuco is reckoned very good, 
 and, according to Von Tschudi, it brings the price of good 
 Louisiana cotton. The region of the Garanhuns, lying just 
 north of the middle of the province of AlagQas, is noted for 
 its cotton. Cazal * says that the Serra de Garanhuns is 
 covered by woods, and that streams descend from it, but 
 are soaked up and disappear on reaching the sandy campos 
 of the vicinity. This region produces also maize, mandi- 
 oca, feijao, and fruits of various kinds. 
 
 Among the fruits for which this province is famous are 
 the mango and the delicious giant pine-apple known as the 
 Abacaxi. Cocoa palms are planted in large groves on the 
 coast, and produce a very considerable revenue. In the 
 Sertao a large number of cattle are raised. Pompe o esti- 
 
 * Corografa Brazi/ica, Vol. II. p. 159. 
 
432 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 mated that in 1864 there ought to be 1,300,000 inhabitants; 
 but there has been no census for many years. Almeida, in 
 his Atlas, estimates the population of the province at 
 1^220,000 souls, of which the capital has 90,000. 
 
 Owing to the flatness of the coast and the small size of 
 the rivers, there are but few ports capable of admitting 
 large vessels. We find in use on this coast, principally for 
 the purpose of fishing, the Jangada, a narrow raft of light 
 logs, carrying a large triangular sail, a craft which may 
 be launched through the surf on the open coast. Of these, 
 in 1864, the province possessed between seven and eight 
 hundred. The city of Pernambuco,* or Recife,! owes its 
 importance to its consolidated beach or stone reef. Except 
 for this, it would offer no advantages for trade. Its posi 
 tion, in the very easternmost part of the empire, makes the 
 port exceedingly convenient of access ; and ships from 
 
 * The name Pernambuco is derived from the Tupi. Schalzo says that it 
 means mare sbucato. Cazal (Coro. Draz. II. 170) makes it a corruption of 
 Paranabuca, which is said to be the name given to the port by the Cahctes 
 who inhabited the place. The Roteiro Geral, cap. 16 (quoted in the Art de 
 Verifier les Dates, Vol. XIII. p. 256), says that the name was given because 
 the reef was broken through by the sea : " Se diz de Pcrnambuco, por sua 
 pedra junto delle esta furada do mar, e quer dizcr, Mar furada." D Orbigny 
 (L ffomme Amcricam, Tom. II. p. 280) makes it a corruption of Paranarnbu. 
 The French call the place Fernambouc or Pernambouc, and the English used 
 to call it Fcrnambuco. 
 
 t The name Recife was given to the city because of its reef. The Portu 
 guese for reef is recife, which word is not derived from the Latin recipere, as so 
 many authors would have it ; as, for instance, Barlaeus (Rerum per Octen- 
 nium, &c., p. 66), who says : "Ubi terminatur, pagus fuit, Reciffa dicta, forte 
 ab hoc, quod intra hunc et alium terra; similem tractum oblongum qnem Recif- 
 fam Lapidosam vocant recepi naves possint et soleant, accipiendis exponen- 
 disque oneribus." The authors of the Art de Verifier les Dates (Vol. XIII. 
 p. 33) make the same mistake. The word recife is, according to Fonseca, 
 derived from the Arabic racif or razif, signifying a pavement. See note on p. 
 190. 
 
PROVINCE OF PERNAMBUCO. 433 
 
 North America or Europe, bound for South American, East 
 Indian, or African ports, have to go but very little out of 
 their way to stop there. 
 
 I have already called attention to the fact that the city is 
 situated on a tract of low ground, occupying a deep inden 
 tation in the coast tertiary sheet, and extending from Olinda 
 nearly to Cape St. Augustinho.* It is built at the mouths 
 of the rivers Beberibe and Capibaribe, which unite, forming 
 a sort of delta, composed of a number of irregular islands 
 very difficult to describe, which are enclosed by a net-work 
 of channels. All of these islands are low, and some are 
 marshy. The Beberibe is the northernmost of the two 
 rivers. It takes its rise to the northwest of Pernambuco ; 
 at Olinda, about two miles north of Pernambuco, it meets a 
 very narrow strip of sea beach, which extends southward to 
 Pernambuco, a part of the city being built on the end of 
 the spit, which broadens in the manner represented in the 
 sketch-map on the following page. A channel from the Rio 
 Capibaribe joins the Beberibe just above its mouth, and this 
 cuts off from the mainland a large island, on which a second 
 division of the city is built, while the third quarter is situ 
 ated on the mainland, on the opposite side of the Capibaribe, 
 opposite its mouth. These three quarters are united by 
 bridges, several of which are of excellent construction. 
 
 To the southward a bay penetrates deeply into the land, 
 but its waters are very shallow. 
 
 That part of the city built on the extremity of the sand 
 beach above described is called Recife, though foreigners 
 almost invariably use the name Pernambuco. It is very 
 closely built up with warehouses, stores, custom-house build 
 ings, &c., and is the centre of the commerce. On the beach 
 
 * Kiddcr compares this tract to the Bahian "Reconcavo." 
 
 VOL. I. 19 BB 
 
434 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 PEHKAMBTTCO. 
 
 just to the north are the celebrated old fortresses of Brum 
 and Buraco. 
 
 Opposite the former, and distant about 250 bracas from 
 the shore, the stone reef* begins abruptly, and runs in al- 
 
 * The reef has been described over and over again by the old navigators 
 and travellers in Brazil. A very curious drawing of it is to be found in a Dutch 
 work published in 1624, and entitled Reys-boeck van het rycke Brasilien. I do 
 not know the author s name. The sketch is, however, so inaccurate as to be of 
 no value as showing the structure of the reef. Barlams not only speaks of the 
 reef, but he gives a large copper-plate engraving of it, together with an excel 
 lent map. There is another large and curious copper-plate of Pernambuco and 
 the reef in the " Historia delle Gverre del regna del Brasile, dal P. F. Gio. 
 Gioseppe di Teresa Carmelitano Schalzo," published 1 698. Speaking of the reef, 
 Schalzo says : " Vien reparato da vn come marauigloso molo iui formato dalla 
 natura il quale s inalza sopra 1 acque, distendendosi gran quantita di leghe tagli- 
 ato dall istessa natura con tanta egualta, come si veggono i moli con immensi 
 dispendij fatti daH arte." Dapper also figures it in his America. 
 
 This reef is represented in Kidder s "Brazil," and in Kidder and Fletcher s 
 "Brazil and the Brazilians," as if it were two or three miles distant from the 
 city ! A comparison between the views of Pernambuco in these works with a 
 good map will prove amusing ! What is the use of stating a thing in words, 
 and then giving the lie to it in an illustration ? 
 
PROVINCE OF PERNAMBUCO. 435 
 
 most a straight line southward, passing within a few hundred 
 feet of the city, extending across the bay, and then skirting 
 the shore for some distance south. I have been able to ex 
 amine it only in the vicinity of the city, and, since the maps 
 and charts differ so widely among themselves, I can form no 
 definite idea of its real length. Barkeus gives a map and 
 drawing of a similar reef extending across the mouth of a 
 bay to the south of Pernambuco, near Cape Saint Augustine, 
 if I remember rightly. 
 
 I have already described with much detail the stone reefs 
 or consolidated beaches of Porto Seguro, Santa Cruz, and 
 Bahia. The Pernambucan reef, so far as I have seen, has 
 precisely the same structure. It is the consolidated core 
 of an ancient beach which has been separated from the 
 mainland by the encroachment of tbe sea. It is remark 
 able for its great length and the straightness of its course. 
 It is exposed to a much heavier surf than the Porto Seguro 
 and Santa Cruz reefs, and at high tide and during heavy 
 weather it is usually deluged by the waves. The northern 
 end breaks down abruptly as at Porto Seguro, and forms a 
 wide opening for the entrance of shipping. Near this ex 
 tremity stands the ancient Dutch fort, known as the Picao, 
 together with a lighthouse. A part of the reef is artifi 
 cially built up with masonry, to make it a more efficient 
 breakwater. Ships of ordinary draught and small steamers 
 enter the port, but the larger men-of-war and the ocean 
 steamers usually anchor at a distance off the reef. 
 
 The water along this coast is very shallow, and the sound 
 ings continue for many miles out from the shore. The tides 
 rise at Pernambuco about six feet. 
 
 Pernambuco is one of the calling stations of the United 
 States, English, and French mail steamers, and it is the 
 
436 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 head-quarters of a steamboat company whose boats run to 
 various points on the coast to the north and south. 
 
 The Pernambuco and Sao Francisco Railroad is in about 
 the same condition as the Bahia and Sao Francisco line. 
 It extends southwestward from the city through the fertile 
 sugar lands, but it is, I believe, completed for a distance of 
 only about eighty or ninety miles, the present terminus being 
 Una on the river of the same name. I owe to the kindness 
 of Dr. McGrath of Pernambuco, and Mr. Mann, the super 
 intendent of the road, the following table of heights, along 
 the line : 
 
 Distances in Kilometres. Feet above the Sea. 
 
 Marca 9 no Arsenal de Marinha . . .10.00 
 Turntable de C. Pontas .... 17.96 
 
 2/j Ponte de Afogados 27.16 
 
 8/0 Estaqao de Boa Viagem . . . . 35.11 
 
 Ponte de Jaboatao . ..... 26.25 
 
 Ponte de Pirapama . . . . . 29.12 
 
 31^ Villa do Cabo 56.00 
 
 Centre do tunnel 184.00 
 
 Ponte de Utinga 210.00 
 
 45 Esta9ao de Olinda 340.00 
 
 Pedreiras de Timbo-assu . . . .373.00 
 Estate de Escada 314.50 
 
 70 " " Frexeiras 404.00 
 
 Contendas 413.00 
 
 Ponte de Amaragi ..... 296 18 
 
 Plana 387.30 
 
 113 T % Esta9ao d Agoa Preta 463.00 
 
 Ponte de Formigueiro .... 418.96 
 Excava^ao no Sitio Gomes .... 505.96 
 Esta^ao Una 437.96 
 
 About thirty miles north of Pernambuco is the large and 
 
PROVINCE OF PERNAMBUCO. 437 
 
 fertile island of Itamaracii, which is separated from the 
 mainland by a narrow but deep channel like a river, the 
 island being set into the mainland, and not lying off the 
 coast as it is usually represented on maps. It is about ten 
 or twelve miles long from north to south, and very narrow. 
 The land is low, and it is interesting geologically on account 
 of its limestones, which contain fossils, and apparently be 
 long to the cretaceous, if one may judge from a report made 
 to the Brazilian government by the engineer Paulo Jose de 
 Oliveira, who was sent to examine the island for coal, which 
 had been reported to exist there. Oliveira speaks of a bed 
 of chalky limestone, " containing some fossils of the turilite 
 and carditas family," and also of other fossils of the " anan- 
 chites family," found at Porco das Caixas. 
 
 Dr. McGrath of Pernambuco has kindly undertaken to 
 have a collection made from these limestones, and I hope 
 that we may soon have their age satisfactorily determined. 
 Lime is manufactured on Itamaraca from these rocks, and 
 also from corals which abound in the vicinity.* Oliveira 
 reports the existence of iron ore on the island. A small 
 quantity of salt is made on the coast from sea-water. The 
 island is very fertile, and is covered with sugar plantations 
 and cocoa-palm groves. The eastern side is thickly planted 
 from one end to the other with these beautiful and useful 
 trees. 
 
 Fernando de Noronha belongs to the province of Pernam 
 buco, though it is situated to the north of Cape Sao Roque. 
 It lies in lat. 3 55 S. and long. 32 40 W. of Greenwich 
 and is distant from the coast about two hundred miles. It 
 consists of one large island and several smaller ones, the 
 whole being, according to Darwin, nine miles long and three 
 
 * See Kidder s Brazil, Vol. II. p. 172. 
 
438 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 in breadth. This distinguished observer visited it, and I 
 cannot do better than to quote his own description of the 
 geological structure of the island : 
 
 
 ISLAND OF FKKNA2JDO Dli NOEOKHA.* 
 
 " The whole seems to be of volcanic origin, although there is no 
 appearance of any crater or of any one central eminence. The most 
 remarkable feature is a hill one thousand feet high, of which the 
 upper four hundred feet consist of a precipitous, singularly shaped 
 pinnacle formed of columnar phonolite, containing numerous crys 
 tals of glassy feldspar and a few needles of hornblende. From 
 the highest accessible point of this hill I could distinguish in 
 different parts of the group several other conical hills, apparently 
 of the same nature. At St. Helena there are similar great conical 
 protuberant masses of phonolite nearly one thousand feet in height, 
 which have been formed by the injection of fluid feldspathic lava 
 into yielding strata. If this hill has had, as is probable, a similar 
 origin, denudation has been here effected on an enormous scale. 
 Near the base of this hill I observed beds of white tuff, intersected 
 by numerous dikes, some of them amygdaloidal basalt and others 
 of trachyte ; and beds of slaty phonolite, with the planes of cleavage 
 directed northwest and southeast. Part of this rock, where the 
 crystals were scanty, closely resembled common clay slates altered 
 by the contact of a trap dike. The laminae of rocks, which un 
 doubtedly have once been fluid, seems to me to be a subject well 
 deserving of attention. On the beach there were numerous frag 
 ments of compact basalt, of which rock a distant facade of columns 
 seems to be formed." t 
 
 * This sketch I copy from Ulloa. The water-line has been inadvertently 
 omitted. 
 
 t Geological Observations, Part II. pp. 23, 24. 
 
PROVINCE OF PERNAMBUCO. 439 
 
 The island is almost deprived of vegetation, resulting 
 from the dryness of the climate, and Ulloa* tells us that 
 sometimes two or three years pass without rain. On this 
 barren rock the Brazilian government has established a 
 penal station. Flocks of sea-birds resort to the island to 
 breed, and sea-turtles in great numbers lay in the sands of 
 the shore during certain months. The harbor is an open 
 roadstead. If the reports of navigators are correct the 
 island is surrounded by coral reefs. 
 
 * "La csterilidad de esta Isla no precede de la mala calidad de su Tierra, 
 pues produce todo quanto se siembra en ella proprio de Paises calidos, sino de 
 la fulta de humedad ; porque passan dos, y tres afios sin Hover, ni verse el mas 
 leve aparato de Agua ; y su escase z es causa de que se sequen totalmente todas 
 las plantas, faltando la Agua a los Arroyos, y lo mas pingue de toda la Isla 
 quando las Nubes no la fecundizan con su riego, se vuelve tan arido, y desapaci- 
 ble, como los Penones, y Rocas ; en la casion, que llegamos, se havian passado 
 dos anos sin caer Agua alguna." ULLOA, Relacion Historica del Viaye, &c., 
 Parte II., Tomo Quarto, p. 416. 
 
440 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER XI. 
 
 THE PROVINCE OF PARAHYBA DO NORTE. 
 
 Limits of the Province. The Serra or Plateau of the Cairiris Velhos. The 
 Climate, Productions, &c., of the Province. Fertile Lands found only along 
 the Coast. The Rio Parahyba do Norte, its Navigability. The City of 
 Parahyba. The Consolidated Beach at the Mouth of the River. The River 
 and Town of Mamanguape. The Geology of the Vicinity of Parahyba. 
 Cretaceous Limestone with Fossils. Observations of Professor Agassiz 
 and Mr. Williams. Mr. Williams s Observations on the Geology of the 
 Country between Parahyba and the Gold-Mines of Pianco. Mode of 
 Occurrence of the Gold. " The Tasso Brazilian Gold-Mining Company 
 (Limited)." 
 
 THE Province of Parahyba do Norte lies to the north of 
 that of Pernambuco, from which it is separated by the Serra 
 dos Cairiris Velhos, which seems to be the continuation east 
 ward of the Serra Araripe. As laid down on the maps, the 
 serra is noteworthy for its west-east trend, though it pre 
 sents a gentle curve to the southward. This is in direct 
 contrast with that of the other serras on the coast, which 
 usually have a more or less northeast direction. 
 
 The principal mountain ranges crossing the province, as 
 the Borborema and Teixcira, have a direction considerably 
 to the east of north. These are composed of gneiss. This 
 abnormal trend attributed to the Cairiris Velhos would be 
 sufficient to lead one to suspect that the so-called serra be 
 longed to the same class as the Araripe and the " serras " 
 dividing the Sao Francisco and Tocantins basins, and the 
 suspicion is confirmed by Pompe o, who says that the Serra 
 
PROVINCE OF PARAHYBA DO NORTE. 441 
 
 Borborema forms in the south an extensive plateau. So that 
 it is more than probable that the Cairiris Velhos are very 
 erroneously laid down on the maps, and do not form a nar 
 row mountain chain, as represented. 
 
 The province forms a regular oblong about one hundred 
 and eighteen * miles in length from east to west, and about 
 ninety miles from north to south. The interior, as in Per- 
 nambuco, is uneven, and there are some considerable serras. 
 The climate is very dry. The country is very poorly watered, 
 and is consequently covered with a very scanty vegetation, so 
 that it is fit only for pasturage. Pompeo says that the cattle 
 subsist largely on the macambira, a bromeliaceous plant, 
 which not only furnishes food, but is sufficiently juicy to 
 quench their thirst. The coast is low, and much of it is 
 very sandy. f As in Pernambuco, there are some fertile 
 lands along the coast admitting of cultivation, and they 
 are in part covered by forest. These lands produce cotton, 
 sugar, ^ tobacco, &c. 
 
 The climate of this province is hot, but on the coast the 
 heat is modified by the breezes from the sea. It is very dry, 
 and from time to time severe droughts prevail, causing much 
 distress. The province is reputed healthy for Brazilians. 
 
 The principal river of the province is the Parahyba do 
 Norte, which takes its rise in the Serra de Borborema, near 
 the southern border of the province, and, skirting the prov 
 ince line, flows east a few degrees north to the sea. 
 
 At its mouth the Parahyba forms a sort of estuary, which 
 
 * Pompeo says one hundred and ten leagues, which is very far from being 
 correct. 
 
 t See Ivoster s Travels. 
 
 | Pompeo says that in 1864 there were one hundred and sixty sugar factories 
 in the province. 
 
 19* 
 
442 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 is quite wide, and opens into the sea from the south, as do 
 many of the other rivers along this part of the coast. 
 
 The Parahyba is navigable for large vessels up to the city 
 of the same name, the capital of the province, a distance of 
 three leagues above the mouth (Pompeo).* Small vessels 
 ascend about the same distance above the city, and canoes 
 are said to go as far as Filar, many leagues farther, but the 
 influence of the tide, according to Pompe o, is felt only six 
 leagues above the river mouth. 
 
 The city of Parahyba is situated on the right bank of the 
 river, and is divided into an upper and lower town. It 
 numbers, according to Almeida, 14,000 inhabitants.! Its 
 exports are principally cotton and sugar, which are sent to 
 Pernambuco. Pompe o says that during the year 1862 - 63 
 there were exported 201,890 arrobas of cotton, worth 
 3,021:124$800, and 620,270 of cotton, worth 821:120-3000. 
 
 Barlaeus J gives a chart of the mouth of the Parahyba 
 River, of which I give on the opposite page a reduced copy. 
 It is interesting because it shows a stone reef extending 
 across its mouth, leaving an entrance between it and the 
 point on the southern side of the river. The reef begins off 
 the point on which stands the old fort Cabedello, and runs 
 parallel with the shore in a southwest direction for some 
 distance. At its southern extremity it is, according to 
 Barla3us s chart, triple, two short reefs lying inside of it. 
 
 I introduce, for comparison with Barla3us, a copy of part of 
 a map of the mouth of the same river by Almeida, which 
 ought to be more accurate, though it does not represent the 
 reef with so much care. 
 
 * Judging from Mrs. Agassiz s description of the Professor s visit to Parahyba, 
 the steamer was unable to go up as far as the town. 
 
 t The population of the province is about 300,000 (Almeida), 
 t Rerum per Octennium in Brasilia, &c., 1647. 
 
PROVINCE OF PARAHYBA DO NORTE. 
 
 443 
 
 About eighteen miles north of the mouth of the Parahyba 
 is the mouth of the Mamanguape, a much smaller stream 
 than the Parahyba. It is said to admit of navigation for 
 smacks as far as the important town of Mamanguape, which 
 lies about four leagues from the sea. According to the 
 Dictionary) Geographico there is a stone reef at the mouth 
 of the river, and Almeida represents a line of reefs as run 
 ning along part of the coast. Cotton is one of the chief 
 exports of the district of Mamanguape. 
 
 Of the geology of the province of Parahyba we know very 
 little, except what is to be found in a short paper by Mr. E. 
 Williamson,* from which I condense the following : 
 
 " From Tambalic (Tambahu 1) to Parahyba the surface is 
 covered by thick beds of ferruginous conglomerate In some 
 
 * On the Geology of the Parahyba and Pernambnco Gold Kegions, by E. 
 Williamson, Proceedings of Manchester Geological Society. This pamph 
 let, which I owe to the kindness of Professor Bonamy Price, of Oxford, bears 
 no date. 
 
444 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 -orr HE 
 
 Rio PARAHYBA DO 
 
 * 
 
 tf 
 
 tf 
 
 places the conglomerate becomes so very coarse that it is wholly 
 made up of water-worn pebbles, of qxiartz, gneiss, and the harder 
 schistose rocks, cemented together by peroxide of iron. The size 
 of the pebbles varies from that of a small nut to boulders weighing 
 four or five pounds. This class is well represented below the 
 baths at Tambalic ; but as they approach the river towards 
 Parahyba they gradually become mixed with finer and more argil- 
 
PROVINCE OF PARAHYBA DO NORTE. 445 
 
 laceous beds, until, at last, at Sta. Rita, a few miles out from 
 Farahyba, they have become divided into regular bands of marls, 
 sands, and conglomerates." These deposits are probably tertiary. 
 " Immediately underlying the ferruginous conglomerates of Para- 
 hyba there occur beds of secondary limestone, having a strike 
 nearly north and south, and dipping gently to the east. For the 
 most part these limestones are siliceous, though at intervals beds 
 of nearly pure limestone and argillaceous bands occur." 
 
 Mr. "Williamson states that he found a cast of a fish tooth 
 and some estherians in this limestone, and that similar beds 
 abounding in fish remains occur at Minas da Cachoeira. 
 Professor Agassiz touched at Parahyba on his return from 
 the Amazonas and visited exposures of these rocks in the 
 vicinity. He found the limestone of a soft texture, of a 
 white or gray color, and destitute of recognizable fossils. 
 From the green shales he obtained species of estherians 
 which appear to be identical with some occurring in the 
 Bahian beds. This latter observation is of much interest, 
 since it would appear to indicate the existence, on this part 
 of the coast, of fresh-water deposits like those at Bahia. 
 
 Mr. Williamson s description of the gneissoid rocks and 
 of the gold deposits I give in full, as it is the only clear 
 account of the way in which gold occurs in this kind of rocks 
 in Brazil : 
 
 " The first clear outcrop of the rocks, from Parahyba, occurs at 
 Batalha on the river Parahyba ; it is a hornblendic rock, with 
 numerous small bands of quartz and feldspar much contorted. 
 
 " Between the river Parahyba and Pilar a very coarse gneiss 
 occurs, with large crystals of white feldspar and black mica ; at 
 Pilar the gneiss is interstratified with mica schists, generally fine 
 in texture ; at Mendonga, Mocheira, and Inga Velho beds of similar 
 character again occur, interstratified with gneiss ; at the last-named 
 
446 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 place the schistose beds become more frequent, until at Inga the 
 whole of the beds are micaceous and hornblendic schists. A little 
 past Inga a hard close-grained gneiss rock appears ; this flanks the 
 mountains of Lagradoura, which chiefly consist of a white por- 
 phyroid gneiss, holding large cleavable crystals of pure orthoclase, 
 interstratified with bands of syenitic and granitoid gneiss, much 
 resembling granite. On the northern flank the hard close-grained 
 gneiss rocks again occur. 
 
 " Between Lagradoura and Campinas a well-marked band 
 granitoid porphyry occurs, standing out some fifty and one 
 hundred feet higher than the softer rocks which surround them ; 
 this porphyry contains large crystals of white orthoclase. At 
 Campinas a series of micaceous beds occur, containing plates of 
 mica. The largest was about two inches in diameter ; but I was 
 told that plates a foot square had been found ; following this run 
 of micaceous schist is a band of porphyry, in which large cleavable 
 crystals of white orthoclase are embedded in a granular matrix of 
 quartz and feldspar. I could distinguish no true bedding lines in 
 this band, but from its dip and strike, an unbroken outcrop, I am 
 inclined to think that it might be interbedcled ; the succeeding 
 rocks are mica-schists, and gneiss. 
 
 " At Caximba Nova another band of hard granitoid rock occurs ; 
 following this there are a long series of mica schists and gneiss ; 
 near to Caracol occurs a series of black schists, alternating with 
 bands of granular black rock ; the schists are occasionally mica 
 ceous. At Caracol a small series of mica schists divide two broad 
 bands of granitoid rock, in places these much resemble the true 
 granites ; overlying the upper one is a small band of hornblendic 
 schist, this is followed by a long series of flaggy mica schists. At 
 Carnahuba these are succeeded by bands of hard, close-grained 
 gneiss ; this flanks the mountains of Teixeira (rocks similar in 
 character occur at Queimada on the opposite flanks) ; the rocks of 
 the Teixeira mountains bear such a strong resemblance to those of 
 Lagradoura, that I think they may be but a repetition of the same 
 
PROVINCE OF PAEAHYBA DO NORTE. 447 
 
 beds. Between Queimada and the Minas da Caxoeira, another 
 broad series of the same class occurs ; the remainder of the rocks 
 on the section are gneiss, alternating with bands of mica 
 schists. 
 
 " At various places on the section beds of quartz and quartzite, 
 with plates of mica, were interstratified with the harder rocks ; the 
 beds varied from a few feet to two hundred feet in thickness ; the 
 smaller bands were often beautifully opalescent, the larger bands 
 granular or amorphous. Hsematitic and titanic iron ores always 
 accompany them. 
 
 " On my journey from Parahyba to the mines I failed to detect 
 any beds of limestone interstratified with the Laurentian rocks, but 
 I was informed that limestone interstratified with the rocks had 
 been observed in other places where the limestones had not been 
 hidden by the covering of ferruginous detritus. 
 
 " The rocks at the Minas da Caxoeira, and the position of the 
 gold-bearing veins, will be best understood on reference to the 
 accompanying section, taken along the bed of the Bruscus River, 
 for a length of about six miles.* 
 
 " At the southern extremity, divided by a band of softer rock, 
 are two broad and well-marked bands of syenitic gneiss, one of 
 which forms the bed of the beautiful waterfalls of the Bruscus ; 
 underlying these is a series of schistose gneiss, and a narrow band 
 of syenite ; it is a bluish-gray crystalline rock, and bears a strong 
 resemblance to some of the Welsh upper Cambrian feldspar rocks. 
 Succeeding these are the gold-bearing series, which almost wholly 
 consist of fine-grained micaceous gneiss, passing imperceptibly 
 into mica schists. 
 
 " Crossing a bend of the river a little before reaching the Lima 
 lode, a small band of dark brownish-gray feldspar rock occurs ; it 
 is subtranslucent, and in places shows chatoyant colors ; a little 
 farther on is a band of white crystalline limestone, containing 
 hexagonal crystals of biotite ; in the bed of the river it is small, 
 
 * This section is omitted. 
 
448 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 but about a mile farther east from this point, at a place called 
 Piao, it is said to be a mile broad on the outcrop. 
 
 " A little east from where the Descubridora lode crosses the 
 stream there occur a few beds of earthy plumbaginous schists, in 
 which are two lenticular veins of graphite ; they seem to be of 
 small extent and of a very indifferent quality. 
 
 " At Cacimbinhas, a few miles farther on than the Boa Espera^a 
 lode, another broad and well-marked band of syenitic gneiss occurs, 
 quite as large as that of the waterfalls. 
 
 " The auriferous veins which traverse these rocks are very nu 
 merous; they appear as irregular lenticular masses, running parallel 
 to the strike, often dipping between, but rarely cutting through, 
 the beds. The matrix of the lodes is a coarse white semi-opaque 
 quartz, containing small quantities of the arsenides and sulphides 
 of iron, sulphides of copper, lead and zinc ; most of the galenas 
 contain antimony. The variety of minerals resulting from the 
 decomposition of these ores are very numerous, carbonate of 
 zinc, carbonate and chloro-phosphate of lead, phosphate, arseniate 
 and carbonate of copper, oxides of antimony and native sulphur 
 are common in some of the lodes ; sulphate of copper, sulphate 
 and chromate of lead were more rare ; native gold was sparingly 
 scattered throughout nearly all the lodes, and in the Boa Esperanga 
 veins, grains of platinum were found. 
 
 " The run of rocks in the valley of the Bruscus are very aurifer 
 ous, and quartz veins abundant ; and though the rocks are greatly 
 contorted, no trace of a true fault can be found anywhere in the 
 whole district ; this singularity appears to belong to all the altered 
 rocks which I examined in Parahyba and Pernambuco, for during 
 a ride of 1,000 miles I failed to detect any ; it is to this want of 
 true fractures that I ascribe the poorness of the quartz veins, there 
 being nothing favorable to the concentration of the ores ; the gold 
 has been equally distributed throughout all the veins. It is well 
 known to miners that no veins are so rich as those in which the 
 faces of dissimilar rocks are brought opposite each other on the 
 walls of the vein. 
 
PROVINCE OF PARAHYBA DO NOBTE. 449 
 
 " On my journey from the mines to Pernambuco I crossed the 
 same run of rocks as those marked on the section, and during my 
 ride was able to trace several anteclinals ; this accounts for the 
 vast extent of country covered by rocks of the same age. 
 
 " About seventy leagues from Pernambuco I found a band of 
 quartziferous porphyry ;* a specimen is exhibited ; it has a compact 
 base, composed of an intimate mixture of quartz and feldspar, 
 enclosing crystals of orthoclase and grains of quartz. 
 
 " Near to Jerimu there occur, within a few leagues of each other, 
 two bands of crystalline limestone ; one a narrow and highly 
 crystalline band, the other a very broad band : in some parts this 
 is micaceous, but none of the beds are so highly crystalline as the 
 small band. 
 
 " The country between Jerimu and Pernambuco bears a strong 
 resemblance in character to those marked on the section between 
 Parahyba and Campinas. 
 
 " The whole series of these rocks agree in every respect with the 
 characteristic features of the Canadian Laurentian rocks, as given 
 by Sir W. E. Logan, viz. : 
 
 " I. The total absence of anything like argillite or clay slate. 
 
 " II. That nothing corresponding to slaty cleavage has ever been 
 remarked. 
 
 " III. That the lamination of these masses is apparently in every 
 case coincident with, and dependent upon, the original stratification 
 of the sedimentary layers." 
 
 There was formed in 1865 an English company under 
 the name of the " Tasso Brazilian Gold-Mining Company 
 (limited), . to work the mines, not only of Parahyba, but also 
 of Pernambuco. Among the officers of this company ap 
 pear some very honorable names. I am entirely uninformed 
 as to what this association has done, or what its prospects 
 are. 
 
 Sr. Jose Jacomo Tasso, in an official report made to the 
 
 cc 
 
450 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 government not long since, said that seven distinct aurifer 
 ous veins had been discovered, but washing was carried on 
 with great difficulty, because water was scarce, and had to 
 be brought from a distance of some two or three miles. 
 
PROVINCE OF RIO GRANDE DO NORTE. 451 
 
 CHAPTER XII. 
 
 THE PROVINCE OF RIO GRANDE DO NORTE. 
 
 Limits of the Province, its Position, Mountain and River Systems, &c. The 
 Rio Piranhas. Vegetation. Productions. The Carnahuba Palm and its 
 Uses. Cochineal. Cattle. Climate. Natal. Geology of the Prov 
 ince. 
 
 THIS province occupies the extreme northeastern part of 
 Brazil, and forms an irregular quadrilateral about one hun 
 dred and eighteen miles in greatest length from east to 
 west, and between eighty and ninety miles in width from 
 south to north. From Ceara it is separated by the so-called 
 serra or plateau of Appodi. Two water-sheds running 
 north-south divide the province into three almost equal 
 parts. The eastern of these water-sheds is formed by a 
 continuation of the Serra Borborema or the Cairiris Novos, 
 east of which a number of little rivers run eastward to the 
 sea ; these rivers being of the same character as those 
 of Pernambuco and Bahia, disappear in the dry season. 
 Pompeo says that the Ceara-merim and Trahiry are per 
 ennial. None of these rivers are navigable for more than 
 a few miles above their mouths. 
 
 The Rio Piranhas, which originates in the province of 
 Parahyba, passes through the middle third of the province 
 of Rio Grande do Norte with a course almost due north, 
 emptying into the sea by several mouths, forming a con 
 siderable delta. This river is sufficiently large to allow 
 small vessels to ascend about seven leagues to the town of 
 
452 GEOLOGY AND PHYSICAL GEOGRAPHY, 
 
 Assu, an important place, noted for its cattle, its salines, 
 and commerce in salt. The Piranhas takes its name from 
 the fish of the same name, which abounds in its waters. 
 
 The western third of the province is traversed by the Rio 
 Appodi, of which the Upanema forms a branch. These 
 two streams take their source in the serras bounding the 
 province on the south and flow northward, uniting only 
 a few miles from the sea. Cazal says that the lands border 
 ing the Appodi are, for the most part, plains diversified by 
 numerous lagoons that dry up in the rainless years. The 
 river is navigable for canoes only as far as Santa Luzia, 
 six leagues from the ocean. 
 
 The greater part of the province is, like the province of 
 Parahyba, dry and largely covered by low, sparse vegetation 
 {Catingas carrasquentas}, but on the higher lands, where 
 the soil is good, there are considerable areas covered by 
 forest and adapted for cultivation. Forests are also found 
 in the eastern part on the low, swampy grounds, and along 
 the banks of some rivers. 
 
 Cotton and sugar-cane are the principal productions of 
 the country, and are cultivated on the serras and river 
 margins, particularly of the rivers Ceara-merim, Carimatau, 
 Potengi, Trahiry, &c. 
 
 Pompe o says that there were, in 1862, one hundred and 
 eighty-five sugar factories in the province, producing 375,000 
 arrobas. The carnahuba palm (Copernicia cerifera) is cul 
 tivated largely on the low grounds, and is one of the most 
 important vegetable productions of the country. This beau 
 tiful palm, which is met with all over the northeastern 
 provinces of Brazil, is of so much interest that I will give 
 a short description of it and its many uses. 
 
 It grows to a height of thirty or thirty-five feet. The 
 
PROVINCE OF RIO GRANDE DO NORTE. 453 
 
 leaf-stalks remain persistent to a height of six feet, more or 
 less, from the ground, the rest of the stem being smooth. 
 The fronds are fan-shaped, furnished with thorns, and dis 
 posed in a close ball-like head, so that the tree presents a 
 very different appearance from that of any other species of 
 palm. 
 
 The uses to which the different part of the plant can be 
 put are exceedingly numerous. 
 
 The roots are used as sarsaparilla. The stem, when about 
 six feet high, is furnished with a pith which, treated with 
 water, gives a sort of meal used for food in famine times. 
 The adult stem is an excellent timber employed in building. 
 The midribs of the fronds are used for making fences, &c., 
 the leaves for thatching, the fibre for cordage. The fruit, 
 properly cooked, tastes like boiled Indian corn, and is used 
 for food. The gum is edible. 
 
 Soon after the young leaves have opened, they are 
 cut and dried in the shade. Scales of a waxy substance 
 are then easily dislodged from their surface, and may be 
 melted over a fire into cakes. This is the caniahuba 
 wax. The same substance is said to be furnished also 
 by the berries. It is very brittle and brown in color, but 
 it may be bleached. Mixed with common wax or tal 
 low, it is made into candles of a fair quality. The pro 
 portion is usually three parts carnahuba to one part wax, 
 or one eighth to one tenth of tallow. (Burton.) This 
 palm is so very abundant that its wax is likely to be 
 come a very important article of export from the northern 
 provinces 
 
 The carnahuba may be occasionally seen in the prov 
 ince of Bahia. Burton met with it in the valley of the 
 Sao Francisco, just above the Barra do Rio Grande, but it 
 
454 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 is most abundant growing on the low lands bordering the 
 streams in the provinces of Pernambuco, Parahyba, Ceara, 
 and Piauhy. The wax is collected to a considerable extent, 
 and one may buy carnalmba caudles or wax all along the 
 coast.* 
 
 In addition to sugar and cotton and the carnahuba, rice, 
 inandioca, beans, tobacco, &c. are cultivated. A species of 
 cochineal insect is found in Brazil, living on the leaves of 
 cacti, and it has been supposed that it might be cultivated 
 successfully in this province, and made an article of com 
 merce. Pompeo says that it is found in the Sertao of 
 Serido, but he does not say whether it is made use of or 
 not.f 
 
 In the interior large herds of cattle are raised, and in 
 1862 there were 2,013 cattle fazendas, $ with about 59,630 
 head of cattle. 
 
 The climate of the province is very hot and dry, and 
 droughts of great severity sometimes prevail. 
 
 The most important place in Rio Grande do Norte is 
 Natal, situated at the mouth of the Potengy. It is a small 
 town, but of some little importance. The entrance to the 
 river is very difficult, owing to shoals and rocks. On the 
 southern side is a stone reef represented in the following 
 sketch-map from Almeida. 
 
 * See Von Martius, Palms, p. 49 ; Dr. Manoel Arruda da Camara in 
 Koster, Travels in Brazil, Vol. II. p. 311 ; Brando, An Account of a 
 Vegetable Wax from Brazil, Phil. Trans., 1811, p. 261 ; Boussingault, 
 Ann. de Chimie, Vol. XXIX. p. 330 ; Sir M. A. de Machedo, Notice sur le 
 palmier Carnauba, Paris, 1867 ; Agassiz, Journey in Brazil, p. 453. 
 
 t Spix and Martius (Travels in Brazil, Vol. II. p. 19) speak of the occur 
 rence of the cochineal insect in the province of Sao Paulo. 
 
 t Accioli, Corographia do Brazil, p. 61, says that cheese and butter are 
 manufactured and exported to a considerable extent. 
 
PROVINCE OF RIO GRANDE DO NORTE. 
 
 455 
 
 The coast is bordered by reefs, but they have never been 
 examined by a competent observer. 
 
 So far as the geology of the province is concerned, the 
 southern part appears to be largely composed of gneissose 
 and other metamorphic rocks, which form in the interior a 
 high and more or less mountainous country like that of the 
 provinces immediately south. Bordering this is a consid 
 erable band of tertiary rocks. The immediate vicinity of 
 the sea is flat, sandy, and often, as in the vicinity of Natal, 
 covered by dunes of blown sand. I know of no mines or 
 mineral deposits in the province, though gold is said to 
 occur there. In the Exposition Universelle of 1867, a 
 specimen of native sulphur, said to have come from this 
 province, was exhibited. I know of no localities in Brazil 
 which afford native sulphur, though it has been affirmed to 
 exist on the Itatiaiossu. One of the principal products of 
 the province is sea-salt, which is largely manufactured at 
 the Salinas d Assu and Mossor<5. 
 
456 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER XIII. 
 
 THE PROVINCE OF CEARA. 
 
 Geographical Position and Limits of the Province. The Serra da Ybiapaba. 
 
 Its Topography and Geological Structure. Serra de Araripe. River 
 Basins. General Sketch of the Geology of the Province. Climate. 
 Character of the Soil. Productions. City of Fortaleza. Population of 
 the Province. Gardner s Sketch of the Geology of Ceara. Character of 
 Country in the Vicinity of Aracaty. Description of Country between 
 Aracaty and Ico. Serra de Pereira. Villa do led and Vicinity. Coun 
 try between Ico and Crato. Gold Washings. Crato. Serra de Araripe 
 
 Villa da Barra do Jardim. Description of Fossil Fish Locality. The 
 Fishes noticed by Spix and Martins and Others, and described by Professor 
 Agassiz. Glacial Phenomena of Vicinity of Fortaleza spoken of by Pro 
 fessor Agassiz. Mammalian Remains. Minerals. Meteorolites. 
 
 THE province of Ceara is bounded on the north by the 
 Atlantic, on the east by the provinces of Rio Grande do 
 Norte and Parahyba, on the south by the province of Per- 
 nambuco, and on the west by Piauhy, and, according to Dr. 
 Pompe o, has an area of 4,681 square miles. Its western 
 boundary line is formed by a narrow range of high lands 
 called collectively the Serra da Ybiapaba.* Pompe o says 
 that this serra begins near the coast not far from the eastern 
 mouth of the Parahyba, and runs, under the names of 
 Serra Grande, Serra da Ybiapaba or Cratheus, in an al 
 most southerly direction, to the confines of the province of 
 Pernambuco, where it ties in with the great serra running 
 
 * This name is spelled in a variety of ways, as Ybiapaba, Hibiappaba, Ipia- 
 paba, &c. Some would make it mean land of precipices. According to the 
 Diet, da Lingoa Tupi, Ih/ apuba means terra talhada, or (/ashed kind. 
 
PROVINCE OF CEARA. 457 
 
 southwest, forming the water-shed between the basins of 
 the Parnahyba and Sao Francisco, and called the Serra dos 
 Dous Irmaos or Borborema. 
 
 In all this extension the serra presents steep slopes and 
 often precipitous sides towards the east, but its top is per 
 fectly flat, forming a chapada, according to Feijo, 2,000 
 to 2,400 teet high. This chapada is in some places from 
 thirty-two to fifty-six miles in width. On the western side 
 the slope is not so marked, and the country appears to grow 
 gradually lower all the way eastward to the river Parna 
 hyba. At Cratheus the serra is abruptly broken through 
 by the valley of the Poty. As to the exact structure of any 
 part of this serra in detail our information is very scanty, 
 but from the observations of Gardner, Capanema, Feij6, 
 Pompeo, and others, there can be no doubt that it is com 
 posed of comparatively modern stratified rocks lying hor 
 izontally on disturbed gneissose and other metamorphic 
 strata. 
 
 The horizontal strata appear to be principally sandstones. 
 Gardner supposed that the whole range was cretaceous. On 
 this head I shall have more to say farther on. The serra 
 separating the Province of Ceara from Pernambuco is called 
 Araripe. .This range is of the same general character as 
 the serra just described. At the end of the Termo do Jar- 
 dim it grows very low, and Dr. Pompe o describes the water 
 shed between the Blacho dos Porcos, which flows into the 
 Rio Salgado in Ceara from the Riacho do Mundo Novo, 
 which flows towards the Sao Francisco, as of very little ele 
 vation, and it has been thought practicable to connect over 
 it the waters of the Rio de Sao Francisco with those of the 
 Salgado in Ceara. Beyond this point the serra, after suffer 
 ing continual interruptions, stretches northward along the 
 
 VOL. i. 20 
 
458 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 eastern border of the province, until near the sea it meets 
 with the Plateau of Appodi. 
 
 The province is divided into two portions by a line of 
 serras which runs from the sea-shore near the capital south 
 west or south-southwest to the Serra da-Ybiapaba. The 
 southeastern half of the province forms a single river basin, 
 and is watered by the river Jaguaribe and its tributaries. 
 The western is watered by a host of little rivers, all flowing 
 directly into the sea. 
 
 The coast forms a broad belt of sands but slightly elevated 
 above the sea. This belt varies in width from four to six 
 leagues. (Pompe o.) The sands are light and piled up 
 more or less in dunes, while in some places they drift 
 inland to a considerable distance. The low coast lands, 
 which have a coating of alluvial matter, are very fertile, and 
 are planted with cane, mandioca, &c. 
 
 From the coast the land rises gradually towards the 
 serras, and is uneven though not properly mountainous, 
 being diversified by hills and plains. According to the 
 descriptions of Dr. Pompe o, the sertao, or the greater part 
 of the area of the lower country, is composed of mica-slate, 
 gneiss, and other metamorphic rock.* The country appears 
 to have been once covered by a sheet of sandstone similar to 
 the great sandstone formation of Piauhy, to be described in 
 the next chapter. This has been denuded to a great extent, 
 the sandstone in many cases being left capping the hills. 
 
 The climate in the interior is very hot and dry, but on 
 the sea-coast it is moist and more tempered by the refresh- 
 
 * Pompeo, Ensaio Estatistico da Provincia do Ceard, p. 145, says: "In the 
 sert2o limestone and marble occur in many places, as in the Serrote de Canta- 
 gallo (Caminho de Baturite"), where the rock is a primitive marble, Giboia, four 
 or five leagues from the capital, c., &c. Capanema reports graphite in con 
 nection with limestone. 
 
PROVINCE OF CEARA. 459 
 
 ing breezes that blow in from the ocean. The greatest 
 heat experienced on the sertao is 35 Cent. (95 Fahr.), and 
 the temperature does not fall below 18 Cent. (64.4 Fahr.) 
 
 The rainy season begins in January or March arid lasts 
 until June. The rest of the year is without rain, the rivers 
 and streams dry up, and occasionally the want of water over 
 certain tracts is such that the inhabitants are obliged to 
 leave for more favored districts. Accioli * says that in 
 1792 a drought prevailed for four months, and caused the 
 inhabitants of seven frequezias to abandon them entirely.! 
 Other noted seccas prevailed during the years 1825 and 1845. 
 
 Pompe o divides the soil into lelra-mar, which is fitted for 
 agricultural purposes ; montuoso, productive and covered 
 with forest; and sertao, dry, cut up by taboleiros, hills, &c., 
 and with brooks dry during the summer. 
 
 The sertao is valuable for pasturage, and Ceara is noted 
 for its cattle, of which, according to official papers, there 
 were, in 1854, about 310,000 head, with nearly 40,000 
 horses. Cheese is manufactured in considerable quantity, 
 not only for home consumption, but for export. 
 
 Among the natural vegetable productions may be men 
 tioned the following : The Quina, ipecacuanha, tatajuba, 
 jacaranda, cedro, pao d arco, and a host of other species of 
 woods valuable for building and dying purposes, and the 
 balsamo, jatoba, almescar, mane^oba, carnahuba, caju, man- 
 gaba, &c., &c. 
 
 Cotton, coffee, sugar-cane, and the other vegetable pro 
 ductions of Northern Brazil, are here cultivated. $ 
 
 * Cororjrafia do Brasil, p. 49. 
 
 t A very graphic account of the droughts of the northeastern coast of Brazil 
 is to be found in Roster s Travels in Brazil. 
 
 } Pompe o gives the following amounts of cotton, sugar, and coffee exported 
 between July, 1857, and April, 1858 : Cotton, 52,552 ; sugar, 121,697; Coffee, 
 304,009 arrobas. 
 
460 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The capital city is Fortaleza, or Ceara, a city of some 
 16,000 inhabitants, situated on the coast about two leagues 
 distant from the mouth of the river Ceara. Its principal 
 importance consists in its export of coffee. Besides the capi 
 tal there are seven other small cities in the province, namely, 
 Aracaty, eight miles above the month of the Jaguaribe on 
 the east bank, a place next to the capital in importance, 
 and noted, amongst other things, for its manufacture of 
 carnahuba candles, sole-leather, <fcc., &c ; Ic6, a fearfully 
 hot place, about thirty leagues from Aracaty, on the Rio 
 Salgado ; Crato, situated in a fertile district eighty leagues 
 from the sea ; Granja, at the mouth of the river Camocim, 
 a good port and stopping-place for steamers from Pernam- 
 buco ; Quixeramobim, a little city in a cattle district in the 
 sertao ; Baturite, sixteen leagues from the capital, and noted 
 for its activity in the cultivation of coffee, sugar-cane, &c. ; 
 and Maranguape, also noted for its coffee and sugar planta 
 tions. 
 
 The whole population of the province is probably about 
 540,000. 
 
 In the year 1841 Gardner published, in the " Edinburgh 
 New Philosophical Journal," * a short paper on the geology 
 of Ceara. This article gives so clear and minute a descrip 
 tion of the country, and contains so many important geologi 
 cal facts, that I think it worthy of being reproduced almost 
 in full, particularly since I shall have to examine critically 
 some of his general conclusions arrived at, not only in this 
 paper, but also in his since published " Travels in Brazil." 
 
 * Geological Notes made during a Journey from the Coast into the Interior 
 of the Province of Ceara, in the North of Brazil, embracing an Account of a 
 Deposit of Fossil Fishes, by George Gardner, Esq., Edinburgh New Philosophi 
 cal Journal, April, 1841, p. 75. 
 
PROVINCE OF CEARA. 461 
 
 Mr. Gardner says : 
 
 " I landed at the northeast corner of the province, at the town 
 of Aracaty, which is situated on the east bank of the Rio Jaguaribe, 
 at three leagues from the coast. The first thing that struck me 
 on my arrival was the flatness of the country around it, reminding 
 one of the descriptions which are given of the pampas of Buenos 
 Ayres. With the exception of a few low sand-hills towards the sea, 
 and a round, isolated one about eight hundred feet high, situated 
 two and a half leagues to the southwest of the town, called the Serra 
 de Arere, there is nothing to interrupt the uniform level. The soil 
 for many leagues around is of a sandy nature, and the character 
 istic vegetation is a beautiful species of palm called carnahuba by 
 the Brazilians. It is the Corypha cerifera of Martius, and is so 
 abundant that, on my journey south to the Villa do Ic6, 1 rode for 
 about two days through a forest of almost nothing else. Two and 
 a half leagues to the south of Aracaty I first met with rocks. 
 This was on crossing the river at a place called the Passagem das 
 Pedras. I found them to consist of thin strata of gneiss, almost 
 in a vertical position. The little inclination which they had was 
 towards the northwest, in the direction of the above-mentioned 
 Serra de Arere . From this place to the Villa do Sao Bernardo, a 
 distance of a little less than eight leagues, the country continues 
 perfectly flat, but the ground among the carnahuba palms, and in 
 several large open spaces almost destitute of vegetation, called 
 vargens, is covered with abundance of gravel ; and this, which 
 extends over large tracts, gives it the appearance of the dried-up 
 bed of an immense river. Intermingled with this gravel there 
 are numerous boulders of varioxis sizes, the largest I saw not being 
 more than four feet high. They are all more or less rounded, and 
 consist of granite, gneiss, and quartz. For the next ten leagues 
 the country continues nearly of the same character, with the 
 exception of a low range of gravelly hills running from east to 
 west, and wooded with shrubs and small trees, the most common 
 of which is a species of mimosa. During the next ten leagues a 
 
462 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 slight but perceptible rise of the country takes place. The soil is 
 generally a yellow-colored clay, in many places thickly covered with 
 gravel and boulders, while in others gneiss rocks are seen cropping 
 out and forming often long slightly elevated ridges covered with a 
 species of cactus and a large bromelia. Their strata, like those 
 farther down, are almost vertical. In this tract carnahuba palms 
 become less numerous, and small dicotyledonous trees and shrubs 
 more abundant, but all of them in the dry season, at which time 
 I passed, destitute of leaves. These deciduous woods, which often 
 cover large tracts of country, are called catingas by the Brazil 
 ians. These tracts are still farther characterized by three large 
 species of cacti, belonging to the genus Cereus. During the 
 next twenty leagues, which brought me to the Villa do Ico, the 
 nature of the country differs in again becoming more level, con 
 sisting of large open campos or vargens, the vegetation of which, 
 during the dry season, is quite burnt up, but they are said to 
 yield abundance of grass during the rains ; and the catingas or 
 deciduoiis woods are much larger than they are farther down. 
 The rocks are gneiss and quartz, and in several places large tracts 
 are covered with fragments of the latter, more or less rounded. 
 At about ten leagues below Ico, the monotonous level of the coun 
 try is varied by a mountain range, which makes its appearance to 
 the eastward. This is the Serra de Pereira. It runs from the 
 southwest to the northeast, It is sixteen leagues in length, but 
 its greatest height is not more than one thousand feet above the 
 level of the plains in which it is situated. The structure of its 
 southwest extremity at least is entirely primitive, but near its base 
 I observed a coarse red conglomerate, containing rounded fragments 
 of both primitive and secondary rocks. 
 
 " The Villa do Ico, which is one of the finest in the interior of 
 the north of Brazil, is situated on the east bank of the Rio Ja- 
 guaribe, in the middle of one of the large open campos which I 
 have already described, and during the dry season is one of tho 
 most miserable places imaginable to live in. The country around 
 
PROVINCE OF CEARA. 463 
 
 it is then so much dried up that not a green leaf is to be seen ; 
 and the river, which during the rains is of considerable size, 
 becomes quite dry. The houses are all built of brick, which are 
 made from a very good kind of clay found in the neighborhood, 
 and are all whitewashed on the outside with a white limestone, 
 which is found about ten leagues to the west of the villa. 
 
 " From Ico I went to the Villa do Crato, which is about thirty- 
 four leagues to the southwest of the former place. Between these 
 two places the country is of a more hilly, undulating character, 
 more abundantly wooded, the trees larger, and many of them ever 
 green. Owing to these circumstances but few of the large 
 campos which exist below Ico are met with. The carriage of 
 goods between Aracaty and Ico is effected in large wagons, gener 
 ally drawn by twelve oxen ; but the hilly nature of the country 
 between Ico and Crato does not admit of this mode of conveyance, 
 the backs of horses and even of oxen being made use of instead. 
 Shortly after leaving Ico I passed over the southwest end of the 
 Serra de Pereira at a place where it has but a slight elevation and 
 consists entirely of gneiss. From this place to the Villa das 
 Lavras da Mangabeira, a distance of about ten leagues, the country 
 is of a gently undulating nature, and in many places well wooded. 
 This villa, which is situated close to the Rio Jaguaribe, takes its 
 name from a number of small gold-workings (lavras) which, from 
 time to time, for many years past, have been wrought in its neigh 
 borhood. Nothing, however, was done to any extent till about 
 two years ago, when two English miners were sent for by a com 
 pany in the city of Ceara, the capital of the province. They con 
 tinued their labors till about two months before I passed through 
 the place, having been recalled by their employers. I could not 
 learn what amount of gold they had obtained, but the persons of 
 whom I made inquiries remarked, with apparently much truth, 
 that they did not believe it was sufficient to repay the expense, 
 or the work would not have been abandoned. The gold is here 
 found in small particles, in a dark-colored diluvial soil, at a con- 
 
464 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 siderable depth ; but the place being shut up, I had not an oppor 
 tunity of examining it. 
 
 " At about eighteen leagues below Crato I lost sight of the 
 gneiss rocks, and for the next four found them replaced by a gray- 
 colored primitive clay-slate. At the termination of this, the 
 secondary stratified series begins, the few rocks which I met with 
 from thence to Crato consisting of a white coarse-grained sand 
 stone. 
 
 " The small Villa do Crato stands in the middle of a large 
 undulating valley, which is bounded to the south, to the west, and 
 to the north by mountains which, in their highest parts, do not 
 rise more than from 1,200 to 1,500 feet above the level of the 
 town. The country around is very fertile, producing abundance 
 of cane, from which an impure sugar, in the form of small square 
 cakes, is made, mandioca, Indian corn, rice, cotton, and tobacco, 
 besides all the varieties of fruit which are to be met with on the 
 coast. The great cause of this fertility is the numerous springs 
 which exist along the foot of the mountains. The small streams 
 which proceed from these are divaricated in a thousand directions, 
 for the purpose of irrigating the plantations. The mountains are 
 branches of the long range which separates the provinces of the 
 coast from that of Piauhy to the west, which here receives the 
 name of Serra de Araripe. Their tops are perfectly level, and ex 
 tend so for many leagues to the westward and southward, forming 
 what the Brazilians call Taboleiras. I have ascended this range 
 in all directions, and have universally found it to consist of a 
 generally white-colored sandstone, but in many places it is of a 
 reddish tinge. In the bed of one of the largest streams which pro 
 ceed from it, where a section of the rocks to a considerable depth 
 is formed, I found a stratum of limestone about three feet thick, 
 immediately below the sandstone, and below it another of an 
 impure coal, two feet thick, resting on another stratum of lime 
 stone. Nothing seems to have disturbed the strata, as they all 
 lie in a perfectly horizontal position, and the level nature of the 
 
PROVINCE OF CEARA. 465 
 
 serra proves that this is general. In the limestone I could meet 
 with no fossil remains. The temperature of two of the springs, 
 which rise at the base of the serra I found on examination to be 
 75 Fahrenheit. 
 
 " That part of the serra which lies to the south of Crato is a 
 branch which runs about ten leagues to the eastward. On the 
 south side of it there is another small villa called Barra do Jardim, 
 distant from Crato about fourteen leagues. I went to this place 
 partly for the purpose of botanizing and partly to make a collec 
 tion of fossil fishes, which I was informed were found in great 
 plenty in its neighborhood. The road skirts along the base of the 
 serra in a southeast direction for about five leagues, at the termi 
 nation of which it is necessary to ascend it for the purpose of 
 crossing to the other side. The ascent is far from being good, it 
 being left entirely in the hands of nature. The only rock I 
 observed was sandstone, similar to that which exists at Crato. 
 The breadth of the serra here is nearly eight leagues, and during 
 the whole of this distance the road is as level as a bowling-green ; 
 and, as no water is to be found on it, travellers are obliged to 
 supply themselves with it before ascending. For small parties it 
 is carried in calabashes, but when many pass together a horse is 
 provided to carry two large leather bagfuls. These Taboleiras 
 are generally thinly wooded, with small trees, the principal of 
 which are a species of Caryocar called Pike, a small tree belong 
 ing to the natural order Apocynacese, which produces a delicious 
 fruit called Mangaba, a fine species of Brysonema, the Cashew 
 [caju] (Anacardium occidentale\ a purple-flowered Qualea, and 
 several small leguminous trees belonging to the division Rectem- 
 briae. 
 
 " The Villa da Barra do Jardim stands in a small valley, upwards 
 of a league in length, and in its broadest part about half a league 
 in breadth. It is bounded to the north and east by the branch of 
 the serra which I crossed over, and to the west by another, but 
 neither so broad nor so long. Having made inquiries for the place 
 20* DD 
 
466 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 where the fossil fishes were to be found, I was directed to a rising 
 ground which extends along the foot of the serra. On my arrival 
 at an open place of this gently sloping ridge to the north of the 
 villa, I found the ground covered with great abundance of stones 
 of various sizes, and I was informed that almost every one of 
 them on being broken presented some part or other of a fish. 
 These fragments I soon found to consist of compact fawn-colored 
 limestone. They are of all sizes, from pieces not larger than an egg 
 to blocks of several feet in circumference, and are almost all rounded 
 and smoothly polished, having apparently been for a long time 
 under the influence of a current of water. They, in general, split 
 very readily, and almost all of them present parts of a fish in a 
 more or less perfect state. But by far the greater number of 
 them are so much broken that it is with considerable difficulty 
 tolerably perfect specimens can be obtained. The spot which 
 these stones occupy is not above an hundred yards square, and 
 almost no other stone is mixed with them ; but on every side of 
 this deposit the ground is covered with little rounded sandstones, 
 similar to the rock of which the serra is composed. Besides this 
 I afterwards visited other deposits, one half a league to the south 
 of it ; one at a place called Maccape, five leagues to the east of 
 Jardim ; and another at Mundo Novo, three leagues to the west : 
 all perfectly similar to the one I have described, being all sitxiated 
 on the declivity of the low hills which stand between the valley 
 and the serra, and all occupying places which are almost altogether 
 free from other kinds of stone. From these places 1 have obtained 
 a suite of specimens, embracing upwards of a dozen species of 
 fossil fish.* They vary in size from those of a few inches in length 
 to others which must have been several feet; and all of them, so 
 far as my limited knowledge of the subject allows me to judge, 
 except two species, belong to the order Cydoideoe of M. Agassiz. 
 The most abundant species is one of those which do not belong to 
 
 * Mr. J. E. Bowman in a note says: "Agassiz makes them but seven 
 species, and refers three of them to the Ctenoid group." 
 
PROVINCE OF CEARA. 467 
 
 this order. Of it I possess a nearly perfect specimen about a foot 
 and a half long, but, judging from other fragments of the same 
 species, it must have attained a much larger size.* It has the 
 head very much elongated, and the scales of the back and abdo 
 men are angular, while those of the sides consist of but one row 
 of long, narrow ones, arranged vertically. Of the other species I 
 only possess the tail and a very small part of the body. It differs 
 from the last in appearing to be entirely covered with small 
 angular scales. Both of them I have no doubt belong to the order 
 Ganoideas of M. Agassiz.f 
 
 " On breaking these stones, some of them exhibit abundance of 
 a minute bivalve shell ; and at Mundo Novo I met with a very 
 perfect specimen of what I believe will prove to be a species of 
 Turrilites, about an inch and a half long, and a single valve of a 
 Venus about half an inch in length and in very excellent preser 
 vation. Both of them were found in the same fragment of lime 
 stone. I was informed by a person in Jardim that a few years ago 
 he found a small serpent coiled up in a stone which had been split, 
 but this, no doubt, was a species of Ammonites. In the several 
 hundred stones, however, which I broke in search of fish, I met 
 with nothing of this description. During my excursions in the 
 neighborhood of Barra do Jardim I nowhere met with limestone 
 in situ." 
 
 When we come to discuss, in the next chapter, the ge 
 ology of the neighboring province of Piauhy, we shall have 
 an opportunity of seeing how correct Dr. Gardner s conclu 
 sions are with reference to the age of the sandstone beds. 
 
 * " The fish here described is the Aspidorhynchus Comptoni Agass. 
 J. E. B." 
 
 t Specimens of these fish find their way all along the coast, and it may he 
 well for future observers to remember this. I had a specimen of Aspido- 
 rhynchus given me at Penedo, and specimens of another genus were sent me 
 recently by Mr. Laue from Mtiro im. 
 
468 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The above paper was followed in the same journal by 
 a communication by Professor Agassiz, in which the species 
 of fossil fish were named and described, the opinion being 
 expressed that they were of cretaceous age. 
 
 The first notice I have seen of the fossil fishes of Piauhy 
 and Ceard is made by Spix and Martius in their Travels, 
 one of the species being figured in the atlas accompanying 
 their work. 
 
 In 1838 Mr. Nicolet placed in the hands of Professor 
 Agassiz a few specimens he had received from Pernambuco. 
 
 Gardner s specimens, in the collections of Mr. Bowman, 
 the Marquis of Southampton, Lord Enniskillen, and Sir 
 Philip Egerton, were examined by Agassiz in 1840. Two 
 years later M. F. Chabrillac sent to M. Elie de Beau- 
 moiit from Pernambuco a few more specimens he had 
 obtained from Ceara. These were placed in the hands of 
 Professor Agassiz, who made a long report on them in a 
 letter addressed to M. Elie de Beaumont, published in the 
 Comptes Eendus, Vol. XVIII. p. 1007. In this letter Agas 
 siz enumerates seven species as known to him from the 
 province of Ceara, viz. : 
 
 Aspldorhynclius Comptus Ag. 
 
 Lepidotus temnurus Ag. 
 
 Rkacolcpis bvccalis Ag. 
 
 Rhacolepls Olfersii Ag. 
 
 Rhacolepis latus Ag. 
 
 Cladocyclus Gardncri Ag. 
 
 Calamopleurus cylindricus Ag. 
 
 Professor Agassiz in the above-mentioned letter restates 
 his belief in the cretaceous age of the fishes ; and recent 
 examinations of considerable collections in Rio confirm him 
 in this opinion. 
 
PROVINCE OF GEAR A. 469 
 
 From Ceara, the capital of the province, Professor Agassiz 
 made a journey to the serras, and studied the glacial phe 
 nomena there exhibited. Between the sea and the Serra 
 de Aratanha he reports that he found everywhere on the 
 higher lands a morainic soil with boulders. 
 
 " On this very serra of Aratanha," says Mrs. Agassiz, " at the 
 foot of which we happen to have taken up our quarters, the glacial 
 phenomena are as legible as in any of the valleys of Maine, or in 
 those of the mountains of Cumberland in England. It had evi 
 dently a local glacier formed by the meeting of two arms, which 
 descended from two depressions spreading right and left on the 
 upper part of the serra and joining below in the main valley. A 
 large part of the medial moraine formed by the meeting of these 
 two arms can still be traced in the central valley. One of the 
 lateral moraines is perfectly preserved, the village road cutting- 
 through it ; while the village itself is built just within the ter 
 minal moraine which is thrown up in a long ridge in front of it." * 
 
 At the close of the chapter from which I make the above 
 extract Professor Agassiz himself says : 
 
 " I spent the rest of the day in a special examination of the 
 right lateral moraine and part of the front moraine of the glacier 
 of Pacatuba ; my object was especially to ascertain whether what 
 appeared a moraine at first might not, after all, be a spur of the 
 serra decomposed in place. I ascended the ridge to its very origin, 
 and there crossed into an adjoining depression, immediately below 
 the sitio of Captain Henriques, where I found another glacial bot 
 tom of smaller dimensions, the ice of which probably never reached 
 the plain. Everywhere in the ridges encircling these depressions 
 the loose materials and large boulders were so accumulated and 
 imbedded in clay or sand that their morainic character is un 
 mistakable. Occasionally, where a ledge of the underlying rock 
 
 * A Journey in Brazil, p. 456. The Serra de Aratanha is composed of gneiss. 
 
470 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 crops out, in places where the drift has been removed by denuda 
 tion, the difference between the moraine and the rock decomposed 
 in place is recognized at once. It is equally easy to distinguish 
 the boulders which here and there have rolled down from the 
 mountain and stopped against the moraine. The three things are 
 side by side, and might at first be easily confounded, but a little 
 familiarity makes it easy to distinguish them. When the lateral 
 moraine turns toward the foot of the ancient glacier, near the 
 point at which the brook of Pacatuba cuts through the former, 
 and a little to the west of the brook, there are colossal boulders 
 leaning against the moraine, from the summit of which they have 
 probably rolled down. Near the cemetery the front moraine con 
 sists almost entirely of small quartz pebbles ; there are, however, 
 a few large blocks among them. The medial moraine extends 
 nearly through the centre of the village, while the left-hand lateral 
 moraine lies outside of the village, at its eastern end, and is trav 
 ersed by the road leading to Ceara. It is not impossible that 
 eastward, a third tributary of the serra may have reached the 
 main glacier of the Pacatuba. I may say that in the whole valley 
 of Hasli there are no accumulations of morainic materials more 
 characteristic than those I have found here, not even about the 
 Kirchet ; neither are there any remains of the kind more striking 
 about the valleys of Mount Desert in Maine, where the glacial 
 phenomena are so remarkable ; nor in the valleys of Loch Fine, 
 Loch Augh, and Loch Long in Scotland, where the traces of 
 ancient glaciers are so distinct." 
 
 From Dr. Felice, a land surveyor familiar with the Serra 
 Grande, Professor Agassiz learned that " there is a wall of 
 loose materials, boulders, stones, &c., running from east to 
 west for a distance of some sixty leagues from the Rio 
 Aracaty-assu to Bom Jesus in the Serra Grande " ; and this 
 wall Professor Agassiz believes to he a part of the lateral 
 moraine loft by a great Amazonian glacier. 
 
PROVINCES OF CEAEA. 471 
 
 Bones of huge Mammals, Mastodons and Megatheria, are 
 abundant in various parts of the province, and perfect skel 
 etons have been found. Some of the localities are the fol 
 lowing : Santa Catharina, (lagoa,) Sitio Cronzo, at the foot 
 of the Serra Ybiapaba at Inhamuns ; between Cratheus and 
 Quixeramobim ; Timbauba ; in the place called Sucatinga a 
 skeleton was found in an excavation, and part was sent to 
 Rio ; at Sta. Cruz, in a lagoa, another skeleton was found. 
 
 The following notes on the mineral productions of the 
 province are taken from Dr. Pompeo s Ensaio : 
 
 Amethysts. Serra do Taua. Some of the crystals from this prov 
 ince are very large and of a beautiful color. 
 
 Gypsum. Araripe (fibrous) Cairiri. 
 
 Saltpetre. Found all over the interior, but more particularly at 
 the following places : Tatajuba, where it was extracted by the 
 government ; Pindoba (government works) ; Tagycioca em Guru ; 
 Carnahubal em Sao Pedro de Villa Vic.osa ; Boassu na Granja ; 
 Conceicjio, Guru. ; Pirangi, Choro ; Uruburetama ; Ipu, fec. 
 
 Salt. In various parts of the Jardim, of the Sertao, and especially 
 in the Aracaty-assu, the waters are impregnated with salt. As 
 elsewhere there are, in argillaceous soils, salt licks much fre 
 quented by cattle. Salt is largely manufactured on the coast 
 from sea-water. 
 
 Alum. Feijo says that a considerable surface in the Inhamuns is 
 covered with alum. Capanema reports it from Araripe. 
 
 Magnesia. Cafundo, Inhamuns. 
 
 Carbonate of Potassium. Ipu, Serra Grande, Crato, S. Gonc_alo. 
 
 Amianthus. Cairiri, in veins. Quixeramobim near Lavras. 
 
 Lignite. Quixeramobim. 
 
 Gold* Granja, Baturite, Crato, Termo de Milagres, Ipu, Rio Sal- 
 
 * Feijo, quoted by Dr. Pompe o, Ensaio, &c., 1 52, says : " De ouro encontram- 
 f=e mais on menos vestiffios por todos os riachos, corregos e vertentes das mon- 
 tanhas, que formao as costaneiras da serra grande, desde a Timonha ate Can- 
 
472 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 gado from Missao-Velha to Lavras. In all these places the 
 gold occurs in grains or powder in sands, gravels, or clays ; its 
 source is not known. 
 
 Copper. Said to occur in the Serra Grande and elsewhere. 
 
 Zinc. Sao Pedro, near the Serra da Maosinha, Termo de Milagres. 
 " Dr. Theberge says that he encountered near Milagres a large 
 quantity of blende (sulphide of zinc), so abundant that in cer 
 tain localities it was only necessary to bum a clearing to reduce 
 the metal, which ran into the hollows, where pounds in weight 
 were collected." (Pompeo.) 
 
 Galena. Ipu, Quixeramobim. 
 
 Molt/Mate of Lead. (Capanema.) Near Villa Nova. 
 
 Sulphide of Antimony. {Capanema.) Near Villa Nova. 
 
 Graphite. Baturite, Quixeramobim, &c. 
 
 Specular Iron. Cangati. Iron ore occurs in many localities in the 
 province. 
 
 In the Comptes Rendus, Tome 5 me , p. 211, I find a statement 
 that on the llth of December, 1836, a large meteor passed over 
 Ceara and exploded over the village of Macao, at the entrance of 
 the river Assu, showering down over a large tract of country frag 
 ments of stone, many of which penetrated houses and destroyed 
 cattle. One of these fragments was sent by M. F. Berthou at 
 Paris for analysis. 
 
 ris, com particularidade nas vertentes do Salgado, Acaracu e Jaguaribe, no 
 Inhamuns, Banabuihii, Qnixeramobim, e cabcceiras de Jure. Em todas essas 
 vertentes e terrenes visinhos basta lavar a terra que se arha debaixo do cascalho 
 para pintar o ouro." 
 
PROVINCE OF PIAUHY. 473 
 
 CHAPTER XIV. 
 
 PROVINCE OF PIAUHY. 
 
 Geographical Position, Limits, c., of the Province. The Rio Parnahyba and 
 its Tributaries. Description of its Basin. General Geological Structure 
 and Topography of the Province. Table-topped Hills of Sandstone. The 
 Serra dos Dous Irmaos and its Structure. Discussion of Gardner s Obser 
 vations on the Geology of Piauhy and Ceara. Gardner mistaken in refer 
 ring the great Sandstone Sheet to the Cretaceous. Sandstones probably 
 Tertiary. Their great Extension over Brazil. Distribution of the Creta 
 ceous Beds in Brazil. Climate of Piauhy. The Campos Mimosos and the 
 Campos Agrestes. Peculiarities of their Vegetation. Productions of the 
 Province, Population, &c. 
 
 THE province of Piauhy forms a rather long and irregular 
 strip lying west of the province of Ceara, and to the north 
 west of Pernambuco and Bahia. Its area is about equal 
 to that of Bahia, and it embraces all the country watered 
 by the tributaries of the Parnahyba on the eastern side. 
 
 This river takes its rise in the Serra da Tabatinga in 
 Goyaz, nearly on the same parallel as that on which the 
 mouth of the Rio de Sao Francisco is situated, and in the 
 angle formed by the union of the two hydrographical basins 
 of the Tocantins and Sao Francisco. Its course is approx 
 imately north-northeast; and Pompeo gives its length as 
 330 leagues, which appears to me to be altogether too high 
 an estimate. The same author says that it is navigable for 
 a distance of 260 leagues. It has, according to Pompe o, six 
 mouths, but I cannot learn whether it has a regular delta. 
 
 On the west side, in its very upper course, a few small 
 
474 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 streams, namely, the Balsas, Balsinhas, Penitente, <fec., unite, 
 and with one or two other little rivers enter the main stream, 
 but for the remainder of the distance to the sea the Par- 
 nahyba flows so close to the western rim of the hydro- 
 graphical basin that it does not receive another affluent of 
 importance from that side. On the east, however, it re 
 ceives a host of little rivers which take their rise in the 
 Serra dos Dous Irmaos and its continuation the Serra da 
 Ybiapaba. Most important among these is the Gurgueia, 
 which drains the extensive lake of Paranagu4,* the Ca- 
 ninde, and the Poty. 
 
 The Rio Parnahyba is a white-water stream, flowing for 
 the most of its course through level, more or less swampy 
 lands, grown up with thick bushes and groves of carnahuba 
 and piassaba palms. It is without obstructions, and naviga 
 ble for a great distance. 
 
 The basin of the Parnahyba is a one-sided one, the drain 
 age being towards the east. 
 
 As the little coast streams east of the Parnahyba are com 
 prised within the limits of the province of Ceara, Piauhy 
 has a coast line of only about twenty miles, extending 
 from the mouth of the Parnahyba to the Barra do Igua- 
 rassu. The base of the country consists in the south and 
 southeast, at least, of gneiss and other metamorphic rocks, 
 much inclined, but the greater part of the country is over 
 spread by a thick sheet of sandstone in horizontal strata, 
 extending southward to the edge of the basin. 
 
 This sheet of sandstone has been very extensively worn 
 away by the rivers, and between them are isolated table- 
 topped hills or extensive chapadas. Such is the character 
 
 * The same name is applied to the bay on which the principal seaport of the 
 province of Parana is built. 
 
PROVINCE OF PIAUHY. 475 
 
 of the country in the vicinity of Oeiras, where the sandstone 
 is of a reddish tint, and is sometimes exposed in perpen 
 dicular cliffs. The Arraial de Sao Gon^alo is situated at 
 the foot of one of these sandstone hills, which is four hun 
 dred feet high.* The Serra da Topa is another sandstone 
 mountain, the rocks being of a white or pale red color and 
 disposed in terraces, the top being perfectly flat. 
 
 These sandstones lie on a basis of metamorpbic strata, 
 which has a gentle slope northward or northeastward, and 
 just south of the Lake Paranagua they lap up over these older 
 rocks, abutting against the Serra dos Dous Irmaos, reaching, 
 according to Mr. St. John, a level of about 1,500 feet. 
 Westward they rise toward the so-called Serra de Ybiapaba. 
 The question now is, whether they tie in with the sandstones 
 of the Ybiapaba, and of the Chapadao da Mangabeira, or 
 whether they are newer than the sandstones of Jacobina 
 and the Sao Francisco-Tocantins divide. From all that I 
 have been able to learn I think that they will be found to 
 form part of the great sandstone sheet of Minas, Goyaz, 
 <fcc., and that the table-topped hills of Piauhy will be found 
 to be the exact equivalents of those of Santarem, Monte 
 Alegre, &c. The clays and sandstones of the lower grounds 
 bordering the coast would seem to belong to the same series 
 as those of the vicinity of Pard. 
 
 The water-shed between the hydrographical basins of the 
 Sao Francisco and the Parnahyba is, according to the testi 
 mony of Spix and Martins, Gardner, and other travellers, a 
 low, very gentle swelling, composed of gneiss, mica-schist, 
 and other similar rocks, all more or less disturbed and 
 denuded down very evenly, as is the case with the similar 
 strata forming the divide between the Sao Francisco val- 
 
 * Alcide cl Orbigny, Voyage Pittoresque, 149. 
 
476 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ley and that of the Paraguassu, and with a gentle slope both 
 to the east and west. This gneiss ridge is about 1,250 feet 
 in height where the road from Oeiras to Joazeiro crosses it. 
 
 Gardner, in describing the journey from Paranagua, across 
 the water-shed, southward into the province of Pernam- 
 buco, says that, shortly after having passed the boundary of 
 the province of Piauhy, an elevated table-land is reached 
 called the Serra da Batalha, which he describes as being 
 covered oil its slopes with huge blocks of sandstone, of which 
 he supposed the serra to be composed. Two other serras 
 of the same character lie to the south, and Gardner supposes 
 them all to form part of one great range. 
 
 Padre Cazal, in his Corographia, says that in some parts 
 the serra dividing Pernambuco from Piauhy has two or three 
 leagues of chapada on top. 
 
 From Mr. St. John s observations it would appear that in 
 some places the sandstones are completely removed, as is 
 the case in Bahia also. 
 
 That the Serra dos Dous Irmaos should have its capping 
 of sandstone is not wonderful, when we find the Serra da 
 Ybiapaba so capped to a height of over two thousand feet, 
 and when on the highest land on the Sao Francisco-Para- 
 guassu divide we find patches of the same rock, forming a 
 series of chapadas traceable southward, and tying in with 
 the chapadas of the Pardo and Jequitinhonha valleys. 
 
 The age of the sandstones of the Serra da Ybiapaba and 
 the Serra de Araripe, and also of the great sandstone sheet 
 covering so large an area in the province of Piauhy, remains 
 to be determined. Gardner has called them all cretaceous. 
 Let us examine upon what grounds he has based his con 
 clusion. The whole matter turns upon the relation the 
 fish-bed bears to the strata composing the serra. He says 
 
PROVINCE OF PIAUHY. 477 
 
 in his "Travels" (p. 202) that "the place where these [fish] 
 were found was on the slope of a low hill about a mile from 
 the serra," and that the specimens all come from loose 
 masses, rounded, as he erroneously supposed, by the action 
 of water, and scattered over a very limited surface. He 
 consequently did not see them in place. Owing to his find 
 ing in the immediate vicinity pieces of sandstone * like 
 that which form the serra, he arrived at the conclusion that 
 the fishes come from the sandstone, and that because the 
 fishes were cretaceous, therefore the serra, and not only the 
 serra, but all the great extension of sandstones covering the 
 northeastern shoulder of Brazil, must be cretaceous also, and 
 he even went so far as to divide these rocks into a series of 
 groups, referring them to European horizons. 
 
 The fishes do not occur in rolled masses, as Gardner 
 says, but in concretions, as Mr. Bowman has remarked in a 
 note to Gardner s paper, and as I can also testify after an 
 examination of specimens in my own collections. The 
 occurrence of these concretions on a surface unassociated 
 with other rock would lead one to suspect that they had 
 weathered out of some softer rock, and this is confirmed by 
 a statement made by Pompe o, on the authority of Dr. The- 
 berge,that the fossils are found in " an extremely sticky clay." 
 Theberge says that they either occur in this way or at the 
 bottoms of deep valleys in brooks whose bed is a schistose 
 limestone, so that we have no evidence whatever that these 
 fossils occur in the sandstones of the serra ; on the contrary, 
 they occur in a band of rocks lying well below the sand 
 stones, and bared in the valleys by the denudation of the 
 sandstones, so that we may safely conclude that they are 
 
 * In his paper in the Phil. Trans., Gardner says that these were little 
 rounded sandstones. In his Travels he speaks of them as rounded blocks. 
 
478 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 older than the sandstones.* Gardner was misled, in the 
 first place, into the belief that the sandstones were creta 
 ceous from finding rocks having some resemblance to flint, 
 and by mistaking for chalk a white tabatinga clay on the 
 top of the serra, dug by the inhabitants for use in white 
 washing. That the cretaceous rocks form an extensive series 
 underlying the sandstones is very probable, because Gardner 
 speaks of a number of localities, all of which he describes 
 as being situated on the declivities of low hills skirting the 
 base of the serra. On the west side of the Serra da Ybiapaba 
 he found a similar deposit of fossil fishes, which would lead 
 one to suppose that this serra was also underlaid by the 
 same deposit. 
 
 Gardner, in his " Travels," says that the series of creta 
 ceous rocks in Ceara and Piauhy forming the serras, &c., 
 consists in descending order of 
 
 1. White chalk with flints exposed in pits, and partially 
 overlaid by red diluvial clay. 
 
 2. Sandstone with ichthyolites, equivalent to the English 
 upper green sand. 
 
 3. A series of marls, soft and compact limestones, and 
 lignite, equivalent to the English gault. 
 
 4. A ferruginous sandstone deposit, equivalent to the 
 lower green sand or Shanklin sands. 
 
 Now, of this series, the white "chalk" is a Tabatinga 
 clay. There is no evidence that it contains flints, and 
 he himself says that he found none in any of the " chalk 
 pits" he examined. In the second place, there is no evi 
 dence that the fossil fishes occur in the sandstones, and if 
 
 * In a paper read at the Salem meeting of the American Association, I ven 
 tured to state my strong suspicion that the beds affording the fishes would turn 
 out to be disturbed and inclined, like the other cretacean beds in Eastern Brazil, 
 when Professor Agassiz remarked that this had been reported by Dr. Coutinho. 
 
PROVINCE OF PIAUHY. 479 
 
 they did they would not prove the deposit to be upper green 
 sand, since Professor Agassiz has called attention to the 
 resemblance borne by Aspidorhynchus Comptoni and Lepi- 
 dotus temnurus to allied species of the chalk of Kent. So 
 far as the third and fourth series are concerned, Gardner 
 found no fossils in them, and their cretaceous age is im 
 proved. Pompe o, with Capanema most probably as author 
 ity, gives the following as the succession of rocks in the 
 serra of Araripe : 
 
 The uppermost beds, A, consist of beds of Psamenito, 
 sandstone of a reddish color, with bluish, sometimes black 
 nodules. Below these comes a bed, B, of an exceedingly 
 foliated limestone, under which lies a bed, C, of black 
 clay (?), (Taua), with layers, a palm in thickness, of a 
 bluish and very hard sandstone, containing veins of pyrites 
 and galena (?), or with a very bituminous schist containing 
 the same sulphides and spherical nodules. This is in turn 
 underlaid by a series of sandstones, D, less argillaceous in 
 character. In a sandstone similar to this occurring at St. 
 Pedro, Dr. Gongalves Dias found fossil wood. 
 
 Of this series, A corresponds to No. 2* of Gardner s 
 series, B and C to No. 3, while the lowest sandstones of his 
 series may or may not correspond to D. 
 
 Both of the above sections appear to have been made in 
 the Serra de Araripe, near Crato. Gardner says he saw no 
 limestone in situ at Jardim. 
 
 That these horizontal strata, and especially the sand 
 stones, have anciently had an immense extension over the 
 surface of Cear there cannot be the least doubt, as Dr. 
 Capanema has remarked, for the sandstones are often 
 
 * Gardner numbers the beds from below upwards ; accordingly this would 
 correspond to his No. 3. 
 
480 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 found capping isolated hills, at a long distance from the 
 serra. Indeed, there seems every reason to believe that the 
 provinces of Ceard and Piauhy were covered with it even 
 beyond the coast line. It has been swept away very largely 
 from the coast and the basin of the Jaguaribe almost to 
 Crato. 
 
 The upper part of the Serra da Ybiapaba for apparently 
 its whole length is composed of the sandstone, and the 
 plateau of Apodi seems to be formed in part of it. It is 
 true that where Spix and Martius crossed the Serra dos 
 Dons Irmaos they found, up to the highest point of the 
 pass, 1,250 feet, only metamorphic rocks ; but they found 
 overlying these rocks thick beds of laminated clay (scliiefer- 
 thon), and they report the Serra da Topa as composed of 
 Quadersandstein. Mr. St. John did not meet with the 
 sandstones, but Gardner did in crossing the Serra da 
 Batalha, and he describes the top as a flat chapada. The 
 Chapada de Sta. Maria, lying between the Sao Francisco 
 and the eastern branch of the Tocantins, is another dead- 
 level table -land composed of sandstone, south of which 
 comes the great Chapadao de Urucuia, evidently only the 
 prolongation of the Chapadao de Sta. Maria. On the oppo 
 site or southern side of the valley of the Urucuia we see 
 the chapadas once more. 
 
 In the southern part of Minas, as already remarked, there 
 are the chapadas between Piumhy and Passos on the Rio 
 Grande, east of which is the immense Chapadao de Taba- 
 tinga, some two hundred miles in length, between the two 
 branches of the Parana, the Paranahyba, and Rio Grande. 
 Then we have the elevated plains of the provinces of Sao 
 Paulo and Parana, composed in great part of argillaceous 
 sandstones. The valley of the Sao Francisco to the Cacho- 
 
PROVINCE OF PIAUHY. 481 
 
 eira de Paulo Affonso is filled with horizontal deposits, calca 
 reous in the upper part of the valley, silicious and calcareous 
 in the lower part. East of the valley of the Sao Francisco 
 we find the hills of the water-shed between the Sao Francisco 
 basin and the streams flowing eastward into the Atlantic 
 capped here and there with horizontal deposits, worn away on 
 every side ; the basins of the Rios Pardo and Jequitinhonha 
 are filled with these beds, and everywhere they rise to a very 
 uniform height over the country, a height of 2,000 to 3,000 
 feet above the sea. Westward we know that an immense 
 tract of country in the province of Matto Grosso and the 
 Amazonas is covered in like manner with similar deposits. 
 All these facts speak of a very uniform submergence of the 
 whole country to a depth of at least 2,000 or 3,000 feet 
 below the present sea level, during which the valleys were 
 filled up with beds of clays, sandstones more or less argil 
 laceous, limestones, &c., to a greater or less height. These 
 deposits appear in great part to have been rapidly formed 
 in the bottom of a muddy sea, the material being derived 
 from the decomposed rock crust covering the country. 
 
 The fossil fishes, according to Agassiz, resemble those 
 of the European senonien, so that if the sandstones are 
 cretaceous they must belong to the very uppermost di 
 vision of that formation. We have seen how this same 
 great formation extends over almost the whole Empire, but 
 nowhere that I have heard of aifording a single fossil. But 
 on the coast outside the edge of the plateau, lying on 
 the extension of the gneiss basis, we have, beginning at the 
 Abrolhos and extending northward through Bahia, Sergipe, 
 Pernambuco, and Parahyba, a great series of cretaceous 
 rocks belonging to different epochs in that period. Wher 
 ever I. have seen them they are always disturbed, while 
 
 VOL. I. 21 EE 
 
482 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 they moreover form border deposits abruptly abutting the 
 edge of the plateau, showing that the plateau was out of 
 water when they were deposited. These cretaceous beds 
 are compact sandstones, shales, limestones, conglomer 
 ates, &c. 
 
 The beds of the great sandstone formation approach the 
 coast in very many places, as on the Rio de Sao Francisco, 
 where they make their appearance near the Cachoeira de 
 Paulo Affonso. But they are everywhere horizontal, and are 
 nowhere disturbed by the foldings which tilted the beds of 
 the cretaceous a few leagues to the east. But the cretaceous 
 beds of the coast are overlaid by a series of clays, sandstones, 
 &c., which, though strictly a coast formation, bears in litho- 
 logical character a very close resemblance to the great sand 
 stone sheet of the interior. This latter, which is overlaid by 
 the drift I have referred, as already stated elsewhere, to the 
 tertiary ; but this coast formation is, I think, without doubt, 
 younger than the similar formation of the interior, so that 
 the latter must be either upper cretaceous or tertiary, and 
 I must give it as my firm conviction that it will be found 
 to be the latter. 
 
 As to the occurrence of drift in the province of Piauhy, 
 I have no information whatever. 
 
 I know of no workable mines, though gold is said to occur 
 at Olho d Agua, near Oeiras. 
 
 The climate of Piauhy is hot and, according to Pompe o, 
 rather damp. It is apt to be very prejudicial to foreigners, 
 especially in the low grounds along the banks of the Par- 
 nahyba, Poty, and other streams, where intermittent fevers 
 are prevalent. 
 
 The inhabitants make a distinction between the vegeta 
 tion of the eastern part of the province and the central and 
 
PROVINCE OF PIAUHY. 483 
 
 western part. That of the former region is called mimosa, 
 and is characterized by catinga forests, while its plants are 
 furnished with an abundance of hairs and prickles, stiff 
 leaves, small flowers, a very tender fibre, and very often a 
 milky juice. Gardner says that the grasses of the mimoso 
 pastures are annuals, their color is a brighter green, and 
 they have more pliant leaves than those of the agrestes. 
 Spix and Martius give a long list of grasses which are 
 characteristic of the campos mimosos. 
 
 The campos agrestes of Lower Piauhy consist in part of 
 woods, in part of quite open plains. The trees are, accord 
 ing to Gardner, almost all deciduous, and many are gnarled 
 and stunted. Swamps are not infrequent and support 
 clumps of Buriti palms. The grasses of the open plains 
 are coarse and perennial. 
 
 The rains begin in October and last until April, heavy 
 thunder-storms prevailing during that season. 
 
 The principal industry of the province consists in the 
 raising of cattle, and agriculture is pursued only to a small 
 extent. 
 
 The population amounts to about 250,000, of whom about 
 30,000 are slaves. The capital is Theresina, which has 
 about 6,000 inhabitants. The other cities are Oeiras and 
 Parnahyba. The latter is particularly unhealthy. 
 
484 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER XV. 
 
 THE PROVINCES OF MARANHlO, PARA, AND AMAZONAS.* 
 
 Sandstones of the Coast of Maranhao. The Interior composed of Metamorphic 
 Rocks. Gold-Mines of Tun and Maracassume. Climate of the Province. 
 Rains. Cities of Maranhao, Caxias, &c. Pororoca or Bore at the 
 Mouth of the River Mearim. Professor Agassiz s Sketch of the Geology, 
 of the Amazonian Valley. His Theory of the Mode of Deposition of the 
 Amazonian Beds. Discussion of this Question. Cretaceous Rocks in the 
 AmazoHian Valley. 
 
 THE coast of the Province of Maranhao is low and flat, 
 and consists of a tertiary ferruginous sandstone passing 
 into conglomerate, and overlaid, as in the vicinity of the 
 port of Maranhao, by a series of sandstones and clays. 
 Gardner, apparently on the principle that all sandstones 
 must be cretaceous, refers this series to that formation, 
 while Spix and Martius, as usual, call the rock Quader- 
 sandstein. The sandstones and associated rocks form a 
 line of high red cliffs along the shore of the island on 
 which Maranhao is built, just north of the city, east of 
 point Sao Marcos. On the mainland west of the channel a 
 similar line of cliffs stretches from the village of Alcantara 
 to the curious landmark, Mount Itacolumi. The same rocks 
 extend far up the valley of the Itapicuru, on the banks of 
 
 * That this name was given in commemoration of the supposed tribe of 
 female warriors described by Orellana there cannot be the slightest doubt. 
 The attempt to derive it from amassona, a word not to be found in the Portu 
 guese dictionary, falls into the same category as the derivation of Maranhao from 
 mar on nao ? or Alexander the Great from all etjtjs under the grate ! It is not the 
 Rio Amazonas, but the Rio das Amazonas, the river of the Amazons. 
 
PROVINCES OF MARANHAO, PARA, AND AMAZONAS. 485 
 
 which at Mangue Alto they were observed by Spix and Mar- 
 tius to lie on granite containing pistacite, and at Cachoeira 
 this rock passed into a syenitic form. 
 
 Mr. St. John, who descended the Itapicuru and made 
 a considerable stay at Maranhao, will doubtless, in his re 
 port on the geology of the country, furnish us with valua 
 ble information concerning the character and extent of the 
 sandstone deposits. 
 
 In the south and west the country is higher, more un 
 equal, and very largely composed of ancient metamorphic 
 rocks. 
 
 Gold occurs in the province, and is or has been worked 
 by a mining company ; but I have been unable to obtain 
 facts bearing upon the nature of the deposits or their yield. 
 The two principal mines are those of Turi and Mara- 
 cassume . 
 
 The climate of Maranhao, situated as the country is on 
 the edge of the great Amazonian valley, of which we are told 
 by Professor Agassiz it anciently formed a part, has the 
 same general character as that which prevails on the Ama- 
 zonas, being hot and damp ; the greatest heat is about 81 
 Cent., 97.8 Fahr., and the lowest 21 Cent., 69.8 Fahr. 
 (Pompeo.) " The rains begin with great regularity at the 
 end of December, although from October on showers occur, 
 commonly called the c/iuvas de cajd [the cashew-rains]. 
 The rains arc very abundant, and accompanied by much 
 thunder and lightning, becoming more frequent and heavy 
 in May, the end of the winter season." * From June to 
 December the winds called ventos g-eraes blow steadily from 
 the northeast or east-northeast during the day, and during 
 the night-time from the east. 
 
 * Pompeo, Geographia, p, 391. 
 
486 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The climate, except in the vicinity of the Paranahyba 
 River, is said to be quite healthy. 
 
 A great part of the country is heavily wooded with the 
 virgin forest, but in the interior there are some extensive 
 campos and alluvial flats often inundated during the rainy 
 season. The principal products of the country consist of 
 rice and cotton, although a little coffee is planted. A 
 large number of cattle are raised. 
 
 The population of the province amounts to about 390,000 
 inhabitants. 
 
 The capital is Maranhao, a beautiful city of 35,000 inhab 
 itants, built on an island lying off the mouth of the Itapi- 
 curu River. Alcantara is another considerable town sit 
 uated on the mainland opposite the capital, and in the 
 midst of a region noted for producing a most excellent 
 quality of cotton. 
 
 The city of Caxias, on the navigable river Itapicuru, 
 about three hundred miles from Maranhao, is a large town, 
 the centre of an important trade with the interior. Caro 
 lina, on the Tocantins, is a town of but little importance. 
 
 About twelve leagues west of the capital of the province 
 is the mouth of the river Mearim or Meary. This river has 
 so strong a current, and its channel is so shaped, that it causes 
 the tide to enter with a bore. Cazal says that the river 
 suspends for a long time the rise of the tide, then it comes 
 in with great fury, rising in a quarter of an hour the dis 
 tance it had taken nearly nine hours to fall, and then 
 running for three hours with the rapidity of a mill-race.f 
 
 * The river is navigated by steam. 
 
 t Speaking of this river, Cazal says : " Sen alvo he profnndo, e largo ; e sua 
 correntc tito rapida, que suspende a enchente da marc por largo tempo ; result- 
 ando desta opposirao ondus jcncapelladas, chamudas pororticas, que depois de 
 
PROVINCES OF MARANHAO, PARA, AND AMAZONAS. 487 
 
 It will be remembered that a similar phenomenon is wit 
 nessed at the mouth of the Amazonas, and in the vicinity of 
 Para, where, as in Maranhao, the bore is called pororoca. 
 
 Professor Agassiz has treated so largely of the physical 
 geography of the provinces of Par& and Amazonas, or the 
 Amazonian valley, in the " Journey in Brazil," that it is not 
 necessary for me to repeat here any of his conclusions ; 
 besides, the limits of this volume forbid that I should at 
 this time enter upon the discussion of so fruitful a subject. 
 I shall therefore confine myself, so far as the Amazonas 
 is concerned, to a very condensed statement of Professor 
 Agassiz s views with reference to the origin and strati 
 graphy of the various formations which occupy the Ama 
 zonian valley, and with a few remarks thereupon ; this seems 
 necessary in order to complete my sketch of the Geology of 
 Brazil. In the Bulletin de la Societe Greologique de France 
 (2 me S^rie, T. 25, p. 685) is a short article on the Geology 
 of the Amazonian valley, by Professor Agassiz and Dr. Cou- 
 tinho, presented by Professor Jules Marcou, which gives 
 Professor Agassiz s views with great conciseness and clear 
 ness ; and as it is not accessible to general readers, I have 
 reproduced the most important part of it here, together 
 with the section accompanying it. 
 
 Professor Marcou says : 
 
 " Mr. Agassiz thinks that the whole valley of the Amazonas was 
 formed at the end of the cretaceous period, which has left traces 
 of deposits in the province of Ceara and on the Upper Purus. Here 
 
 vencidas, tudo quanto vazou cm quazi nove horas, cnche era menos d hum 
 quarto ; ficando a mare caminhando para cima tres horas completas com uma 
 rapidez semclhante a calha d hum moinho. Este fenomeno occupa o espaco de 
 cinco leguas com grande roido. Ha sitios, denominados esperas, onde as canoas 
 esperao a decizao do combate, e continuao a viagem sem perigo." Cor. Braz., 
 Tom. II. p. 260. 
 
488 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and there, whether by denudations or by anterior dislocations, 
 one sees more ancient rocks. Thus Major Coutinho has found 
 palaeozoic brachiopods in a rock which forms the first cascade of the 
 Tapajos ; carboniferous fossils have been collected on the banks of 
 the rivers Guapore and Marnore", in Matto Grosso ; and finally, at 
 Manaos, Coutinho has recognized slates or phyllades in a very in 
 clined position, and beneath the formations of red sandstone of the 
 Amazonian valley." 
 
 Professor Agassiz supposed that during the tertiary the 
 Amazonian region was above water, and that the sandstones 
 and clays that now fill it are drift. 
 
 The following is a copy of the ideal section of these later 
 deposits by Professor Agassiz, forming a resumS of the ob 
 servations of M. Coutinho and himself : - 
 
 " I. Coarse sands (Sable grossier) forming the base of the drift 
 throughout where the level of the water has uncovered the lower 
 beds of plastic clays. 
 
 " II. The streaked plastic clay (Argile plastique bigarree) shows 
 itself on a large scale along the sea-coast at Para, at the Island of 
 Marajo, Maranhao, and here and there in the hollows along the 
 course of the Amazonas. 
 
 " III. Foliated clay in very thin beds, with frequent indications 
 of cleavage. This deposit appears to be more considerable in the 
 banks along the course of the Rio Solimoes than in the lower part 
 of the Amazonas. It is in these beds at Tonantins, on the Rio 
 Solimoes, that M. Agassiz has found leaves of dicotyledonous 
 plants, which appear to be identical with species at present living 
 in the valley of the Amazons.* 
 
 * These leaves occur in a fine, soft gray clay, resembling very closely the 
 recent alluvial clays of the Brazilian rivers. They are excellently preserved. 
 The leaf is partly carbonized, but it curls up from the surface on drying, and 
 may be detached, leaving a beautiful impression of the venation, &c. 
 
PROVINCES OF MARANHAO, PARA, AND AMAZONAS. 489 
 
 " IV. A crust of sandy clay, very hard, moulded in the inequali 
 ties of the foliated clay. 
 
 " V., VI., VII., VIIL, and IX. Sandstone 
 formation, sometimes regularly stratified and 
 compact, especially in the lower beds (V.), 
 such as one sees on the borders of the 
 
 igaraph of Manaos ; sometimes cavernous $ 
 
 s 
 and intermixed with irregular masses of clay 3 
 
 J k v 
 
 (VI.), especially well developed at Villa Bella ^ " 
 
 and at Manaos ; at others all the characters 
 of a torrential stratification (VII., VIII., and 
 IX.). The deposits of this last nature are 
 only seen in the elevated hills of Almeirim, 
 Erere, and Cupati, and in the most elevated 
 cliffs of the borders of the river, as at Tonan- 
 tins, Tabatinga, Sao Paulo, and on the bor 
 ders of the Rio Negro. 
 
 " X. The argilo-arenaceous, unstratified 
 drift, occupying all the inequalities of the 
 soil resulting from the denudation of the 
 sandstone with torrential stratification. It 
 is in this drift that MM. Agassiz and Cou- M 
 
 tinho have found true erratic blocks of dio. 
 
 M 
 
 rite, of a metre in diameter, at Erere. This * 
 
 formation is never met with on the cliffs ! 
 
 elevated several hundreds of feet in height. 
 There is not a trace of it on the summit of 
 the hills of Erere. 
 
 " The fact that the coarse sand, No. L, ap 
 pears throughout at the level of low water, 
 that it follows the general slope of the val 
 ley, shows incontestably that the deposition 
 of this formation does not reach back to an 
 epoch anterior to the excavation of the val- 
 21* 
 
490 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ley itself. The total thickness of the Amazonian drift does not 
 exceed three hundred metres ; it covers the whole basin of the 
 Amazonas, from the Andes of Peru and Bolivia to Cape Sao Roque ; 
 or, in other words, it is the most colossal drift formation known." 
 
 Professor Agassiz believes that the beds I., II., III., IY., 
 or the coarse sands and clays, were deposited in a lake or 
 sheet of fresh water occupying the valley of the Amazonas, 
 and sustaining on its surface a glacier, descending eastward 
 from the Andes, and furnished with a gigantic moraine in 
 front stretching across the mouth of the valley and convert 
 ing it into an inland fresh-water lake. After the ice had 
 broken up and become more or less disintegrated, and the 
 waters of the lake had swollen, the sandstone formation V., 
 VI. , VII., VIII., IX. was laid down, then the barrier was 
 burst ; the waters of the lake, suddenly released, furrowed 
 and wore down the sandstone beds, sweeping them entirely 
 away over an immense area, leaving only isolated hills, like 
 those of Erere", Obydos, Cupati, Almeyrim, &c., standing as 
 remnants of the once universal sandstone sheet. After this 
 period of turbulence and denudation came on an epoch 
 of quiet, and in the bottom of the diminished lake the 
 clays, No. X., were deposited, while ice-rafts floating on 
 its surface dropped here and there boulders, to be buried 
 in the accumulating material. Then the moraine was de 
 stroyed ; the drainage of the waters furrowed deeply these 
 clays, and even cut through them into the sandstone below, 
 in which the various channels of the system of the Ama 
 zonas are excavated. Professor Agassiz believes that the 
 great barrier stretched across the Amazonian valley far east 
 ward of its present extremity, and he has called attention 
 to the similarity between the formations found spread over 
 the coast of Maranhao and Piauhy and the Amazonian for- 
 
PROVINCES OF MARAXHAO, PARA, AND AMAZONAS. 491 
 
 mations here described, showing conclusively that these de 
 posits were once continuous. It is his belief that the Ama 
 zonian formation formerly extended a hundred leagues out 
 to sea beyond the present mouth of the Amazonas. There 
 can be no doubt that there is a rapid waste of land going on 
 along the sea-shores of the mouth of the Amazonas and of 
 the coast eastward for a long distance, a waste amounting 
 to even so much as two hundred yards in ten years in the bay 
 of Braganza, or a mile in twenty, as on the coast near Vigia, 
 where an island a mile wide disappeared in that time.* 
 
 I have three times visited Para, and have had an opportu 
 nity of seeing something of the Amazonian formation. The 
 rock underlying the town, and exposed at ordinary low- 
 water level at the base of the bluff under the fort, is a 
 coarse dark-red sandstone with an abundant cement of 
 iron oxide, and precisely like the red sandstone I have so 
 often described as occupying a similar level and under 
 lying the tertiary clays of the coast of Rio, Espirito Santo, 
 &c. Over this sandstone is a considerable bed of red, white, 
 and particolored felspathic clays, with a greater or less 
 admixture of sand, which clays present exactly the same 
 appearance and level as the tertiary clays of the provinces 
 just referred to. 
 
 Before I knew anything of the conclusions of Professor 
 Agassiz as to the age of the Amazonian deposits, I had sat 
 isfied myself that the clays and sandstones extending along 
 the whole eastern coast of Brazil, from the Bay of Rio to 
 
 * Since the tertiary, at least, and I believe for the greater part since the drift, 
 the whole eastern Brazilian coast has suffered denudation by the sea to an 
 immense amount, and a very wide sti ip of tertiary rocks has been removed. 
 I believe that these deposits once extended beyond the Abrolhos, and that south 
 of Cape Roque the sea has cut them away for a mean width of fifty miles or 
 more. 
 
492 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 the Amazonas, were older than the drift clays which over 
 spread them ; and as they are stratified deposits on an open 
 coast, there can be no doubt of their marine origin. At first, 
 misled by what I had read of the geology of Brazil, as well 
 as by the strong resemblance the sandstones bore to the 
 new red sandstones of the Basin of Minas in Nova Scotia, 
 with which I have been familiar since my boyhood, I was 
 disposed to regard the Brazilian formation in question as 
 triassic ; but I soon found that it was underlaid unconform- 
 ably by cretaceous rocks in Bahia, and I came to the only 
 conclusion possible, that it was older than the drift and 
 newer than the cretaceous. I can see no reason, therefore, 
 for considering the coast beds as anything but tertiary, 
 though they may be, and probably are, very late tertiary. 
 I have attempted no comparison between these beds and the 
 tertiary beds of the pampas, because, in the absence of 
 fossils, and having never seen the pampian tertiaries and 
 post-tertiaries, I have nothing to aid me in instituting such 
 a comparison. It has seemed to me that the fact of the 
 occurrence on an open sea-coast of clays and sandstones 
 precisely similar to those occupying the lower plains of the 
 Amazonas, as at Para, and in fact tying in with them, re 
 lieves one of the necessity of looking to a fresh-water origin 
 for the Amazonian beds. 
 
 There cannot be the slightest doubt that the beds forming 
 the mountains of Ercrd, Almeyrim, once covered the whole 
 valley, and have been enormously denuded. I should never 
 have doubted whether the red sandstones at Para really 
 belonged to the series of beds forming the Monte Alegre- 
 Ererd hills, if I had not found along the coast of Espirito 
 Santo the same sandstones with precisely similar overlying 
 clays, with no table-topped hills piercing them as at Ererd. 
 
PROVINCES OF MARAXHAO, PARA, AXD AMAZONAS. 
 
 Only once have I seen what I thought to be a table-topped 
 hill standing on the coast tertiary plain. That was on the 
 coast south of Rio de Contas, the observation was a doubt 
 ful one, and I have felt more like instituting a compari 
 son between the chapadas on the Amnzonas and those of 
 Ybiapaba and Minas Novas ; or, in other words, I am dis 
 posed to regard the chapadas of Erere" as the outliers of the 
 great tertiary sheet that once covered the great Brazilian 
 plateau, and now lies unbroken over such an immense 
 extent in the province of Matto Grosso. According to the 
 observations of Dr. R. P. Stevens and others, the plateau 
 of Guyana is covered by an extension of the same great 
 sheet, while the valley of the Orinoco is occupied with 
 clays precisely similar to those of the valley of the Ama- 
 zonas.* 
 
 It is with much hesitation that I express an opinion at 
 variance with so distinguished an authority as Professor 
 Agassiz ; but the facts have seemed to need a different 
 interpretation from that which he has given them. My 
 conclusions, after all, do not affect his theory of the former 
 existence of glaciers under the tropics, down to the present 
 level of the sea, a theory which I hold as firmly as he.f 
 
 * Professor Orton found west of Tabatinga tertiary shells in beds which he 
 considers to be a part of the Amazonian formation. These fossils were de 
 scribed by Professor Gabb in the Journal of Concltology. The species are Neritina 
 pupa, Turbonella minnscula, Messalia Ortoni, Tellina Amazonensis, Pachyilon ob- 
 liqna, and P. tenua. It is much to be regretted that Professor Orton has not 
 given a description of the locality where these fossils are found, and of their 
 mode of occurrence. 
 
 t I have confined myself in this chapter to a short discussion of the question 
 of the age of the Amazonian sandstones, and I have attempted to give no de 
 scription of the great river and its wonders. I would refer the reader desirous 
 of knowing more cf the Amazonas to the " Journey in Brazil." 
 
 It is indeed surprising that after several hundred volumes, classic in science, 
 
494 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Professor Agassiz has called attention to the fossils from 
 cretaceous beds discovered by Mr. Chandless on the river 
 Aquiry, one of the affluents of the Purus. These beds con 
 sist of hardened clay and "pseudo-conglomerate,"* the 
 latter being a sort of clay rock full of concretions, which 
 give to the formation the appearance of a conglomerate. 
 Associated with it is a sandstone. The fossils are said by 
 Mr. Chandless to be very abundant, and, according, I believe, 
 to the determination of Professor Agassiz, they consist of 
 the bones of Mosasauri f and turtles, together with fossil 
 wood. These remains appear to be principally confined to 
 the clays and " pseudo-conglomerate." 
 
 Professor Agassiz regards these fossils as indicating a 
 horizon like that of the Maastricht beds in Europe (J^lae- 
 striclitien^ upper chalk). Judging from the description of 
 
 have been written on Brazil, and some scores of works have been published on 
 the Amazonas, by such writers as De la Condamine, Humbolclt, Spix and Mar- 
 tius, Prince Adelbert, Bates, Wallace, Agassiz, and a host of others of greater 
 or less note, the idea should be so generally prevalent that the country is unex 
 plored, a perfect terra incognita, and that every year or two some traveller never 
 before heard of should astonish himself, if not the world, by rediscovering the 
 river. Perhaps after spending a month on its waters, the greater part of which 
 is consumed on board his canoe or the steamer, he writes a book, or at least a 
 magazine article or two ! Few countries have suffered more in America at the 
 hands of superficial travellers and writers than Brazil. I would especially 
 recommend to the attention of my readers the excellent little book by Mr. 
 Bates, " The Naturalist on the Amazonas." Its author is a good naturalist, 
 and his eleven years of residence in the country have enabled him to write with 
 great accuracy. 
 
 * Chandless s papers on the Rio Aquiry. (Jour. Roy. Geog. Soc.) 
 t Professor O. C. Marsh, in one of his papers read before the Salem meeting 
 of the American Association, called attention to the rarity of the Mosasauroid 
 forms in the European cretaceous and their great abundance in the cretaceous 
 of North America. It is interesting to observe the occurrence of this same 
 type in South America. 
 
PROVINCES OF MARANHAO, PARA, AND AMAZONAS. 495 
 
 Mr. Theberge, the Aquiry beds must resemble those of Ceara, 
 in which the fossil fishes occur. 
 
 On the Purus Mr. Chandless * found the same beds at 
 about ^t. 7 15 S., long. 66 W., with bones and an abun 
 dance of fossil wood. 
 
 * Ascent of Puriis, Jour. Roy. Geog. Soc., Vol. XXXVI. 
 
496 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER XYT. 
 
 PROVINCES OF GOYAZ AND MATTO GROSSO. 
 
 The Geographical Position of the Province of Goyaz. The Chapada da 
 Maugabeira. Geology of the Vicinity of Natividade. Gold-Washings of 
 the Serra da Natividade, the Arraial da Chapada, and the Arraial da Con- 
 ceicao. Structure of the Serra at the Town of Arrayas. The Serra 
 Geral. Subterranean Streams. Western and Southern Goyaz composed 
 of Metamorphic Hocks. Distribution of Gneissose and Granitic Rocks in 
 Western Brazil. The Monies Pyrenees and their Height. The Rio 
 Araguaya and its Navigation. Dr. Couto de Magalhaes. Ilha tlo Ba- 
 nanal. Note on Piranhas. Gold, Diamonds, Iron, and Chrome Ores. 
 Climate, Forests, Population, &c. The Western Part of the-Plateau of 
 Brazil composed of undisturbed Beds of Sandstone, &c. The Amazonas- 
 Paraguay Water-shed a Plain without Serras. 
 
 THE materials for writing a sketch of the Geology of the 
 Province of Goyaz are very meagre, since it has never been 
 explored by any competent modern geologist. St. Hilaire, 
 Pohl,* Burchell, and several other naturalists visited the 
 province in the early part of the century. Gardner made 
 a journey through the eastern part in the year 1840, mak 
 ing a few geological observations; but Castelnau, in 1844, 
 travelled very extensively through the length and breadth 
 of the province, furnishing us with connected and valuable 
 geological sections, though these sections and the accom 
 panying text rarely ever do more than indicate the litho- 
 
 * Pohl s (G. E.) Reise im Innern von Brasilien in den Jahren 1817-1821, 
 2 Bde., mit Atlas (Wein, 1831 -37), and his Beitrdge zur Gebirgskunde Brazil- 
 tens (Munchen, 1832) I have never seen. 
 
PROVINCES OF GOYAZ AND MATTO GROSSO. 497 
 
 logical character of the formations. The stratigraphy is 
 vaguely given, and no attempt is made to show the age of 
 the different deposits. 
 
 Goyaz lies west of the provinces of Piauhy, Bahia, and 
 Minas Geraes, and is very long from north to south, and 
 narrow from east to west. It comprises the basin of the 
 Tocantins above its junction with the Araguaya, the part of 
 the basin of the Araguaya east of that river, and the right 
 side of the basin of the Paranahyba, from the Rio Jacare 
 to the Rio Apure. 
 
 The Chapada da Mangabeira is, as already remarked, the 
 southward continuation of the table-land separating the 
 province of Piauhy from the valley of the Sao Francisco. 
 It is in some places quite forty miles in width. Its top 
 forms a plain, and it consists of horizontal beds of sand 
 stone lying on metamorphic rocks. Between the chapada 
 and Natividade the country is composed of these latter 
 strata, while the serra at Natividade has, according to Gard 
 ner, the centre of granite, overlaid by schistose rocks. The 
 western side of the serra is bounded by beds of a very com 
 pact grayish-colored limestone, which extends northwards for 
 several leagues, forming a range of low hills. The surface 
 deposits on this serra, which are largely composed of a fer 
 ruginous gravel, doubtless like the drift cascalho of Minas 
 Novas, contain gold, anciently mined to some extent. Gard 
 ner says that the view east and north from the serra is 
 bounded by several low ridges, but that to the west and 
 south the country appeared to be one vast plain. The 
 same traveller tells us that the whole country about the 
 Arraial da Chapada, a few leagues west, has been turned 
 over in search of gold. The Arraial stands on a low cha 
 pada, but Gardner does not describe its_ structure. Gold 
 
498 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 also occurs in the vicinity of tlie Arraial da Conceic,ao. 
 Gardner s account of the structure of the serra on which 
 the town of Arrayas stands is interesting. The rocks are 
 all metamorphic and almost vertical, the inclination tend 
 ing toward the east. Gardner says : " The most westerly 
 of these rocks have an arenaceous, schistose structure, and 
 these overlie a very compact, grayish-colored, stratified rock 
 very much resembling gneiss, in which are imbedded in 
 numerable rounded pebbles of granite and quartz, of all 
 sizes, from one to three or four inches, and which is probably 
 equivalent to the graywacke rocks of the Old World." Lime 
 stones, which occur both to the north and south, were not 
 observed here. Whether the limestones mentioned belong 
 in the same series with the slates does not appear from 
 Gardner s statement. The Serra Geral, eastward of Arrayas 
 is described as being of no great elevation, and present 
 ing a level top as far as the eye can reach, being evidently 
 the continuation of the Chapada da Mangabeira, and, like 
 it, composed of horizontal sandstone beds.* In speaking 
 of the road from Bonita to Arraial de Sao Domingos, we 
 are told that " the top of the serra still continued to be 
 level, with a precipitous face, the rock being of a reddish 
 yellow," and that " shortly after leaving Bonita an elevated 
 pyramidal peak of the same elevation as the serra is 
 descried to the southeast, presenting a remarkable resem 
 blance to some enormous work of art," so that there can 
 be no doubt about the general structure of the serra. 
 
 Castelnau represents on his map a little stream just north 
 of Sao Domingos as flowing in a subterranean channel. f 
 Gardner describes a river near the Fazenda de Sao Joao, 
 
 * This is confirmed by Mr. Thomas Ward, who has travelled over it. 
 t Castelnau, V mt Partie, GeogmpHif, Atlas, Planche 4. 
 
PROVINCES OF GOYAZ AND MATTO GROSSO. 499 
 
 that disappears in an opening in limestone strata, and 
 runs underground for several miles, when it reappears.* 
 
 On his journey to Sao Romao Gardner followed the Serra 
 Geral from near Sao Domingos to the head-waters of the 
 Urucuia, and he describes it as one great elevated plain or 
 chapadao. Of the geological structure of the southern part 
 he gives but few hints, but as he occasionally mentions the 
 occurrence of limestone it is very probable that it may be, in 
 part at least, composed of strata of that rock, the continua 
 tion of the horizontal limestone deposits of the Rio das 
 Velhas. Most maps represent a narrow mountain-chain 
 separating the basins of the Sao Francisco and the Tocan- 
 tins, and Gerber, in his map of Minas Geraes, though he 
 represents correctly the Chapadoes de Santa Maria and 
 do Urucuia, with their great level tops, draws along the 
 water-shed a mountain-chain on top of the chapada, call 
 ing it the Serra das Araras and the Serra do Paranan. 
 Along this whole region we have no evidence that I 
 have seen of the existence of any extensive elevations 
 breaking through the great table-land. Castelnau s map of 
 the southern part of the province shows the chapada as 
 extending southward nearly to Catalao. Gerber represents 
 a chapada on the opposite or south side of the valley of 
 the Urucuia. When we come, however, to the southern part 
 of the province, we find, just over the limits, in the Prov 
 ince of Minas Geraes, the immense Chapada da Tabatinga 
 lying between the Rios Grande and Paranahyba, and com 
 posed of horizontal beds of sandstones, &c. 
 
 According to Castelnau, Saint Hilaire, and others, the 
 
 * There are other examples of subterranean streams in Brazil. Gerber 
 represents the Rio Partlo, an affluent of the Suo Francisco on the left side, 
 as flowing under a ridge. 
 
500 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 foundation rock of western and southern Goyaz is every 
 where gneiss, mica-schists, clay-slates, and limestones, evi 
 dently belonging to the same metamorphic series we find 
 in the eastern part of the Brazilian plateau. These rocks 
 are much folded, rising in mountains comparing in eleva 
 tion with those of Minas Geracs. The ridge dividing the 
 basin of the Tocantins from that of the Paranahyba branch 
 of the Parana is of the same character. Clay slates and 
 other metamorphic rocks are seen at Cuiabd and Diaman- 
 tino in Matto Grosso, and Chandless * speaks of granite 
 seen in the bed of the river Tapajos, ten miles above the 
 Rio de Peixes, just below the Rio das Tropas, at the shal 
 low of Mangabal Grande, and at various points below on 
 the same river. All these observations go to show that the 
 great Brazilian plateau, like that of Guyana, was originally 
 wholly composed of gncissosc and schistose metamorphic 
 rocks, much disturbed throughout. It has been supposed by 
 some that, going westward from the Serra do Espinha^o, the 
 signs of metamorphism disappeared, and that rocks which 
 in eastern Minas Geraes might have been highly metamor 
 phosed spread out flatly westward, as in the eastern part 
 of the United States the palaeozoic rocks which were folded 
 along the Alleghanian region spread horizontally over the 
 west. This is not the case, so far as I can learn, in Brazil. 
 The metamorphic part of the Brazilian plateau, so high on 
 the east, in Minas slopes off to the north-northwest, and 
 southwcstward from the vicinity of Ouro Preto, and dips 
 under the great sheet of tertiary rocks, showing itself only 
 where these are denuded, or where an occasional and rare 
 prominence pierces these strata, but a ridge of these rocks 
 stretches off in a series of high lands from Ouro Preto and 
 Barbacena into Goyaz. 
 
 * Journal Royal Geographical Society, Vol. XXXII. 
 
PROVINCES OF GOYAZ AND MATTO GROSSO. 501 
 
 Mr. Thomas Ward, in a note addressed to the author, very 
 aptly describes the Province of Goyaz as a metamorphic 
 island in a sea of sandstone, and such seems indeed to be 
 the case. The sandstones have been swept away from the 
 greater part of the Araguaya-Tocantins basin, leaving the 
 irregular surface of the metamorphic rocks exposed. The 
 highest points in Goyaz are the Montes Pyrenees, near the 
 city of Goyaz, which are said to be over 9,500 feet.* 
 
 The highlands in Southern Central Goyaz, collectively 
 known as part of the Serra dos Vertentes of Baron von 
 Eschwcge, form the water-shed between the Tocantins- 
 Araguaya basin on the north and the Parana basin on the 
 south. The Araguaya and Tocantins above their junction 
 are both large rivers, but the Araguaya is very much longer, 
 and should rank as the main river. It flows for the 
 greater part of its length at a much lower level than the 
 Tocantins, and it offers much greater facilities for navigation. 
 The Araguaya has been several times explored. Castelnau f 
 in 1844 descended the river from the mouth of the Crixas to 
 its union with the Tocantins, and then ascended the Tocan 
 tins, making plans of both rivers. He found the Araguaya 
 navigable and with but few obstructions. In 1856 the 
 President of the Province sent Sr. Vallee to explore the 
 same river, and he reported that it might easily be made 
 navigable 4 
 
 * I find in the Bahia Interesse Publico for the 21st of November, 1868, a letter 
 from Sr. H. R. dos Genettcs, describing an ascent of the Pyrenees. This gen 
 tleman says that he ascertained the height of the most elevated point to be 2,932 
 metres, or about 9,619 feet, which is much greater than had been supposed. 
 
 t Castelnau, Expe d. dans I Amfr. du Sud, Hist, du Voyage, Tomes I. et II. ; 
 also Atlas. 
 
 J E. J. C. Vallee, Erploracao do Rio Araguaya. Published in the Anncxo P, to 
 one of the Government Reports, published, I think, in 1865. My copy wants 
 the title-page. 
 
1 
 502 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Dr. Couto de Magalhaes, formerly President of Goyaz, 
 espoused the cause of the Araguaya, and navigated it in a 
 little steamer from Jurupencem, fourteen leagues from the 
 capital on the Rio Vermelho, a branch of the Araguaya, to 
 Para. The president published not long since an excellent 
 memorial on the advantages to be gained by the navigation 
 of the river. Mr. Ward tells me that a steamer now makes 
 regular trips from Parti to Goyaz. I have told the story of 
 Araguaya to show that Brazil is not wholly without the 
 spirit of enterprise. She is exploring her great rivers and 
 establishing, slowly it is true, steam navigation upon them, 
 and in a few years the interior of Brazil, so long shut out 
 from the world, will be accessible to commerce. 
 
 The lands bordering the Araguaya are in great part flat, 
 low, and composed of sands, clays, and other very recent de 
 posits. An interesting feature in the river is the Ilha do 
 Bananal, formed by an arm leaving the main river on the 
 east in latitude about 12 80 (approx.), and entering it 
 again in about 9 30 . Castelnau determined the length of 
 the island to be seventy-five leagues. Almeida, in his map 
 of Goyaz, does not represent it as quite so long. 
 
 The Araguaya is very rich in fish,* and a species of 
 dolphin occurs in it. 
 
 Gold is found in many localities in the province. The 
 
 * Castelnau says that piranhas he calls them pirangas are very numer 
 ous and voracious. According to him, " leur voracite est telle, que presque 
 tous les oiseaux aquatiques que nous procurions avaient les pattes en partie 
 
 devore es par eux Un de nos compagnons de voyage pousse par 1 exces de 
 
 la chaleur, se mit imprudemment a 1 eau, et fut presque aussi tot attaque par 
 des legions de ces animaux ; immediatement les caux furcnt teintes de son sang 
 et il fut heureux pour lui qu il se trouvat tres pres du rivage, vers lequel il se 
 pre cipita avec rapidite, echappant ainsi aune mortcertaine et affreuse." Hist, 
 du VOI//IIJK, Tome I. p. 404. 
 
PROVINCES OF GOYAZ AXD MATTO GROSSO. 503 
 
 country in the vicinity of the capital is largely auriferous. 
 Castelnau speaks of the occurrence of the precious metal at 
 the following localities, Rio Vernielho, Rio Bagagem, Serra 
 Dourado, Districto de Ouro Fino, Morro do Calisto, Districto 
 da Anta, Thesouras, Rio Claro, Julgado de Crixas, Nativi- 
 dade, Trahiras, &c. 
 
 Diamonds have been found on the Rio Claro ; iron occurs 
 at Ouro Fino, Anta, Aldea de Sao Jose, and chrome at Ouro 
 Fino, where it is said to have been found by Pohl. 
 
 The Province of Goyaz is, generally speaking, dry, con 
 sisting of campos and catingas. Forests have but a small 
 extension. There is a large tract covered by the virgin 
 forest between the capital and Meia Ponte. The province 
 is especially adapted for grazing. The climate is dry, but 
 varies much, according to latitude. The population is, ac 
 cording to Almeida, 250,000 souls.* 
 
 The western part of the high lands of Brazil, forming the 
 Amazonas-Paraguay water-shed, and comprised in the prov 
 ince of Matto Grosso, is completely covered with undis 
 turbed tertiary beds, and forms a low swelling plateau, 
 on which the rivers take their source. 
 
 This is well shown in the maps and geological sections 
 accompanying the great work of Count Castelnau. The 
 rivers Xingti, Tapajos, and Paraguay all take their rise in 
 this plain f within a few miles of one another, near Diaman- 
 
 * Castelnau devotes Chapter XVII. of the second volume of his Hist, du 
 Voyage to a description of the Province of Goyaz. 
 
 t Chandless says that the water-shed between the Amazonas and the Para 
 guay, " though commonly called a serra, has nothing of a mountainous charac 
 ter. It is simply a high range of country, varying but little in its general 
 elevation, though deeply grooved by the valleys of the rivers. Around them 
 one finds more or less virgin wood ; the rest is campo, that is, pastures sprinkled 
 more or less thickly with stunted trees, in parts including the quina tree. This 
 
504 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 tino, and the water-shed is so low that wooden canoes ascend 
 the Tapajos from Santarem, cross over, and embark on the 
 Paraguay, descending to Villa Maria. 
 
 In descending the Tapajos, on his way from the Diaman 
 tino to the Amazonas, Chandless found the river bordered 
 by sandstones, which, as at Crepore", he describes as soft in 
 character. 
 
 On the road from Goyaz to Cuiaba one passes over an 
 immense plain of sandstone in horizontal beds. The valley 
 of the Paraguay at Cuiaba and Diamantino is excavated in 
 this sandstone sheet down to the metamorphic rocks lying 
 beneath. In the valley of the Paraguay, near Cuiaba and 
 Diamantino, diamonds and gold occur in considerable abun 
 dance. 
 
 So little is definitely known about the geology and physi 
 cal geography of the Province of Matto Grosso, that I con 
 tent myself with these few general remarks. Castelnau has 
 written more than any one else on the physical features of 
 the province, and the reader is referred to his Histoire du 
 Voyage for more details. 
 
 range seems to consist mainly of sand-rock and clay. In general it drops 
 steeply and often precipitously to the lower country, the plain below appearing 
 as a sea with deep bays and inlets." 
 
PROVINCES SOUTH OF RIO. 505 
 
 CHAPTER XVII. 
 
 PROVINCES OF SAO PAULO, PARANA, SANTA CATHARINA, AND 
 RIO GRANDE DO SUL. 
 
 The Serra do Mar of Sao Paulo a Plateau. Its Character. Drainage in the 
 Provinces of Sao Paulo and Parana to the Westward. Eastward-flowing 
 Streams of no Importance. The Sao Paulo Railroad. Description of the 
 Country, along the Railway between Santos and Scao Paulo by Major 0. C. 
 James. Geology of Vicinity of Sao Paulo. Mawe s Description of the 
 Gold-Mines of Jaragua, and the Method of extracting the Gold. Country 
 westward to Campinas. Iron-Mines at Ypanema Serra Arassoiava or 
 Guarayoiava. Climate, Products, &c., of the Province of Sao Paulo. 
 General Topographical Features of the Province of Parana, its Climate, 
 Productions, &c. Matte or Paraguayan Tea. Tea-Planting in Brazil. 
 Rivers. Colonies. Paranagua. Coal Basin on the Rio Tuberao in the 
 Province of Santa Catharina. General Description of the Physical Features 
 of the Province and that of Rio Grande do Sul. History of the Coal-Mines 
 of Brazil. Observations of Perigot, Bouleich, Ave-Lallemant, Plant, &c. 
 Description of the Coal-Fields of the River Jaguarao. Engineer McGinty s 
 Report on the Candiota Coal. Coal Basin on Rio Sao Sepe . Basin near 
 Sao Jeronymo. 
 
 THE so-called Serra do Mar, seen in sailing along the coast 
 of the Provinces of Sao Paulo and Parana, is the edge of the 
 great Brazilian plateau, which along the coast of Sao Paulo 
 has a height of 2,500 - 3,000 feet.* Towards the sea it pre 
 sents a very steep declivity, but on the opposite side there is 
 no corresponding slope. Climbing the serra at Santos, one 
 finds himself on an immense table-land of gneiss, roughened 
 by a line of considerable hills a few miles from its edge, but 
 
 * Mawe estimated the height of the plateau on the Santos and Sao Paulo 
 road at 6,000 feet. 
 
 VOL. i. 22 
 
506 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 soon growing gradually lower in going westward, until at 
 Campinas broad plains are reached, that stretch off with 
 more or less interruption toward the Parana, tying in 
 with the great plains of Paraguay and the Argentine Re 
 public. The united provinces of Sao Paulo and Parand 
 lie, like Ohio in North America, on the western slope of 
 the border of the great interior continental basin of South 
 America. As the bluff edge of the plateau so nearly coin 
 cides with the coast line, the drainage in these two prov 
 inces is principally westward into the Parana, while the 
 rivers flowing eastward are of very little importance. The 
 province of Santa Catharina, just south of Parand, lies 
 partly on the seaward side of the serra and partly behind 
 it. The streams flowing eastward are of no importance, 
 while those flowing westward form the head-waters of the 
 Uruguay. South of the province of Parana the water-shed 
 bends inland some two hundred and fifty miles, and then 
 runs southward through the province of Rio Grande do 
 Sul, ending as the high lands break down and disappear 
 on approaching the Rio de la Plata. 
 
 The city of Sao Paulo, the capital of the province of 
 the same name, is situated on the top of the plateau, 
 at a distance of forty-five miles from the sea. The prin 
 cipal port of the province is Santos, a considerable town 
 of some 7,000 inhabitants, and noted for its export of 
 coffee, which reaches 160,000 sacks yearly. From this 
 place to Sao Paulo runs a railway, which is continued west 
 ward beyond Campinas. This railway was constructed by 
 American engineers, and my friend, Major 0. C. James,* 
 was one of the corps. On my last voyage, while at Rio, I 
 
 * I wish I could fully express my indebtedness to Major James : I owe him a 
 thousand acknowledgments for his kindness. 
 
PROVINCES SOUTH OF RIO. 507 
 
 obtained so many facts of interest from him with reference 
 to the topography and geology of Sao Paulo, that, as he was 
 about to. return to Sao Paulo, I asked him to make obser 
 vations on the surface deposits along the line of railway, 
 which he did, furnishing me with a report, from which 
 nearly all the facts relating to the route of the railway have 
 been taken. Major James says that Santos is " within a 
 league or two of the foot of the great back-bone of moun 
 tains, a league or two of soft, oozy mud, a few feet above 
 the sea-level, the base of a sort of estuary, of which the 
 spur and main range form the bounds. This interlying 
 marsh or lagoon is overgrown with very shabby palms, 
 large moss-covered trees, looking like spectres in the land 
 scape, and with a very heavy, tangled undergrowth, 
 lazy, sinuous lines of water, having their beginnings and 
 endings in the sea, traverse it, as if it were just for a 
 stroll inland beneath the umbrageous foliage and back 
 again to the ocean refreshed." * Just before the railroad 
 reaches the base of the serra the mud is left, and a broad 
 band of gravel is crossed. This lies at a higher level than 
 the mud, and appears to slope away from the serra. The 
 materials are very coarse, and there were no sands seen. 
 Major James says : " We dug into it ten feet or so, to 
 get gravel for ballast. There are no shells in it. The 
 height above sea-level is not great, say ten or twelve 
 feet, not more than twenty." This deposit evidently corre 
 sponds to the raised beaches of Rio and northward. 
 
 A few small streams flow into the estuary, and the rail 
 road follows one of these until a deep gorge in the moun 
 tain is reached, a gorge formed by two spurs jutting out at 
 right angles to the line of the serra, placed like abutments 
 
 * It is hardly necessary to add that the locality is very unhealthy. 
 
508 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 to support the mountain. Taking the southern slope of the 
 northern spur, the railway creeps up at an angle of one in 
 ten until it reaches the summit, five miles from the plain, 
 having attained an elevation of twenty-six hundred feet. 
 
 CUTTING ON THE SAO PAULO RAILWAY, SHOWING DIUFT LYING ON DECOM 
 POSED ROCK.* 
 
 Beginning near the foot of the semi, an examination of 
 some of the cuttings through the massive corrugations of the 
 hillside reveals a yellow unstratified clay,f not very compact, 
 interspersed with pebbles, stones, and rocks, nearly all well 
 
 * The decomposed rock is seen gullied away underneath the drift. 
 
 t Major James says that this is generally the color of the drift-clay even to its 
 whole depth. He compares the drift-paste on the Sao Paulo Railroad on the 
 slope of the serra to that exposed in the deep cutting: near Rodeio on the 
 Dom Pedro Segundo Railroad. At Sao Paulo, however, it is more red,, as we 
 shall presently see. 
 
PROVINCES SOUTH OF RIO. 509 
 
 rounded, overlying the rock in situ, a decomposed gneiss. 
 Here and there in a few cuttings a thin sheet of pebbles may 
 be traced between the yellow clay and the rock, losing itself 
 now and then. This description will apply generally to all 
 the cuttings except a few where the excavation is made 
 through partially decomposed rock. 
 
 Major James informs me that the clay is thicker on the top 
 of the ribs or corrugations of the hills cut through by the 
 railroad than on the sides, which is owing to the denudation 
 of the slopes. This clay sheet may be fifty or more feet in 
 thickness. On reaching the top of the serra one finds him 
 self on an elevated country, a table-land diversified by hill 
 and valley, the hills being generally low and rounded, the 
 valleys wide and with a luxuriant vegetation flourishing on 
 marshy bottoms. On the old mule-road near the edge of the 
 plateau the country is gently rolling, and for some eight or 
 nine miles the soil, when not covered by a sparse vegeta 
 tion, is of a gray color, and presents a dreary aspect. On 
 the surface is a layer of white sand, only two or three inches 
 in thickness generally, but when drifted of course more. 
 Under the sand is an almost white clay, of about the same 
 consistence as potter s clay. Adobe houses built with it are 
 nearly white. Under this is the drift, which is exposed in 
 the banks of the Rio das Pedras, that flows through this 
 tract and runs toward the coast. The highest point in the 
 serra is to the west of this region. 
 
 Near Tamanduatahy the land spreads out between the 
 hills level as a lake, and about two miles wide, covered with 
 deep layers of black soil, which Major James describes as 
 " fibrous and woody like peat." He informs me that the rail 
 road was built over the surface of this bog, and he did not 
 know what kind of a soil underlay it, but he felt satisfied 
 
510 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 that it occupied a shallow valley in the drift, which he be 
 lieved extended underneath. Beyond this are some cuttings 
 through higher ground, in which the reddish drift-paste, 
 nearly one hundred feet in thickness, is cut through. In this 
 clay are "boulders like those of the serra, of a very hard, 
 bluish-gray rock scattered through the loose material, and 
 in one or two instances exposed on the surface." 
 
 The city of Sao Paulo lies at a distance of forty-five miles 
 from Santos. It is a large and important city of 20,000 
 inhabitants built on high ground, nearly surrounded by a 
 low plain, through which flows, on the west, the river 
 Tiete*, one of the affluents of the Parana. According to 
 Ma we, the hill on which Sao Paulo is built consists of the 
 following deposits in descending order. First of all, a coat 
 ing of red soil more or less thick, impregnated with iron 
 oxide ; under this is sand, together with other materials 
 associated with pebbles, the whole being from three to six 
 feet thick. Under this comes a layer of purplish or varie 
 gated clay, with thin layers of sand ; then follows a layer of 
 stratified materials, the whole resting on decomposed gneiss- 
 granite.* At Itu, a short distance from Sao Paulo, the Rio 
 Tietd is represented by Major James as cutting through hori 
 zontal deposits of red sandstone and conglomerate, and this 
 is the material which is used to pave the streets of the 
 capital ; as Fletcher and Mawe have remarked, the rock 
 contains gold.f 
 
 * M. Pissis p;ives a similar description of the deposit, and illustrates it by a 
 section. He calls the horizontal deposits of Sao Paulo and Itu fresh-water ter 
 tiary, and says that a similar lacustrine deposit occurs in the upper part of 
 the valley of the Parahyha do Sul. (Mem. de. V Institute de France, Vol. X.) 
 
 t Mawe (American edition, p. 73), speaking of the streets of Sao Paulo, 
 says : " The material with which they are paved is lamillary grit stone, cemented 
 with oxide of iron, and containing large pebbles of rounded quartz, approxi- 
 
PROVINCES SOUTH OF RIO. 511 
 
 Westward of Sao Paulo are some high hills ; the most 
 conspicuous among which is Jaragua, in whose vicinity 
 gold-mines were anciently worked. 
 
 Mr. Mawe, during his travels in Brazil, visited these gold 
 mines, and described them in his " Travels in the Interior 
 of Brazil." I am unable to refer to the English edition, but 
 I translate a few paragraphs from a French abstract,* since 
 it gives very neatly the mode of occurrence of the gold and 
 the ancient method of extracting it. 
 
 After speaking of the discoveries made by the Paulistas, 
 he goes on to say : 
 
 " The gold-mines of Jaragua, being situated at a distance of 
 four leagues from Sao Paulo, were the first that were discovered. 
 This part of the country is unequal and mountainous. The rock 
 forming the principal base of the soil rarely ever shows itself. It 
 appears to be a granite f passing into gneiss. 
 
 " This primitive rock is immediately covered over in many points 
 by a bed of a not very solid agglomeration, formed principally of 
 pebbles of quartz and gravel. This itself is covered only by the 
 vegetable earth. 
 
 " It is this conglomerate that is intermingled with grains of 
 gold. They give it the name of cascalho.\ 
 
 " The mining takes place in open cuttings, and the extraction 
 of the gold is carried on by washing ; negroes are employed in 
 this work. 
 
 mating to the conglomerate. This pavement is an alluvial formation, contain 
 ing gold, many particles of which metal are found in the chinks and hollows, 
 after heavy rains, and at such seasons are diligently sought for by the poorer 
 sort of people." 
 
 See also Spix and Martins, " Travels in Brazil," English Translation, Lon 
 don, p. 21. 
 
 * Annales des Mines, 1817, Vol. II. p. 202. 
 
 t In the German edition Mawe describes the gneiss as containing some 
 hornblende and an abundance of mica. 
 
 J Mawe, in the original English edition says, cascalhao. 
 
512 GEOLOGY AND PHYSICAL GEOGRAPHY, 
 
 " When a current of water can be found whose level is suffi 
 ciently elevated, steps are cut in the earth, each twenty to thirty 
 feet long, two or three wide, and a foot in height. At the base 
 a trench two or three feet deep is dug. 
 
 " On each step are placed six or eight negroes, who, as the 
 water slowly descends from above, stir up continually the earth 
 with shovels, until it is all converted into a liquid mud and car 
 ried lower down. 
 
 " The particles of gold contained in the soil descend into the 
 lower cutting at the bottom, to which they soon settle because 
 of their greater specific gravity. The workmen are continually 
 employed in removing the stones from the ditch, and in cleaning 
 the surface, an operation which is facilitated by the current of 
 water which falls there. After five days of washing they remove 
 the sediment from the bottom of the cutting. It is of a deep 
 carbonaceous tint, and composed of iron oxide, pyrites, ferruginous 
 quartz, and scales of gold. 
 
 " This sediment is transported to another current of water, there 
 to undergo another washing operation. For this purpose funnel- 
 shaped wooden bowls or gamellas, two feet large at the mouth and 
 five or six inches deep, are used. Each workman, standing upright 
 in the brook, takes about five or six pounds of the auriferous sedi 
 ment in his gamella. He then causes a certain quantity of water 
 to enter, and agitates it with dexterity, in such a way that the scales 
 of gold soon fall to the bottom and on the sides of the vessel, 
 uniting themselves together and separating from the other lighter 
 substances, which the water holds in suspension and carries away 
 little by little with it. He then rinses the gamella in another 
 of larger size, full of water. The gold is deposited there, and he 
 recommences a similar operation. The washing of a gamella takes 
 eight to nine minutes. 
 
 " The gold taken out varies in the number and the size of 
 the scales ; some are so small that they float, while others are 
 as large as peas and even larger. 
 
PROVINCES SOUTH OF RIO. 513 
 
 " This operation, of which the result is of the greatest impor 
 tance, is watched by inspectors. The gold-dust is carried to a 
 mint, where the impost of the fifth is taken out and the rest 
 melted 
 
 " The mines of Jaragua have been famous for two centuries for 
 their great yield. This district was regarded as the Peru of 
 Brazil ;* but its riches are to-day infinitely less." 
 
 The occurrence of gold in the gneiss regions of the Serra 
 do Mar, at Sao Paulo and Cantagallo and in the same 
 eozoic belt on the Mucury, as reported by the engineer 
 Schieber, is interesting, as these rocks are rarely ever rich 
 in the precious metal, and the whole gneiss belt of Brazil 
 is remarkably barren of metalliferous deposits of all kinds. 
 
 Gold is said to occur in the Villa de Guarapuava, to the 
 west of the river Tibaji and elsewhere in the province. 
 
 Major James reports that the cuttings beyond the Tiete 
 show the line of pebbles and stones overlying the rock. 
 
 The high hill called Cabellos brancos is bare of soil and 
 vegetation on top, whence the name.f 
 
 At Jundiahy, according to Major James, the rounded hills 
 are nearly bodily composed of the unstratified paste, but 
 behind the station the line of pebbles may be distinctly 
 traced, like a thin sheet or veil over the rock. A few 
 leagues farther on, near Campinas, the country becomes less 
 rugged, one might say it is heavily undulating, and here 
 from a slight eminence the gradual descent into the great 
 basin of the interior is plainly perceptible. The hills are 
 
 * Pompeo says that the gold-mines of Silo Paulo produced, up to the begin 
 ning of the present century, 4,650 arrobas of gold. ( Geoqraphia, p. 480.) 
 
 t Von Tschudi, Vol. III. p. 231 : " Sie fiihrt ihren Namen Das gebirge 
 des weissen Haares, weil dcr Kamm wahrend der kalten Jahreszeit in den 
 Friihstunde oft mil Rcif bedeckt ist." 
 
 22* GG 
 
514 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 now composed of a very dark-red ferruginous earth resem 
 bling in grain coarse brown sugar ; a mass squeezed by the 
 hand retains the impress of the palm. Major James says 
 that " travellers in dry weather get to look like Indians ! " 
 This earth goes by the name of terra roxa* The terra 
 roxa of Campinas Paulo is, according to Major James, the 
 continuation of the drift-paste of the higher lands and 
 seaward slope of the serra. It varies much in thickness, 
 and lies usually on the tops of ridges between the rivers, 
 not descending into the valleys, which, as in the basin of 
 the Jequitinhonha, are very deep, steep-sided, and narrow. 
 This red earth forms a most fertile soil, and the country 
 covered by it is clothed by an exceedingly luxuriant veg 
 etation. Bamboos are very numerous, but there are very 
 few cacti. No soil is better suited to the coffee-tree, and 
 in this part of the country it is extensively cultivated on 
 the upper lands, but never on the slopes or on the intervales. 
 The terra roxa rarely ever contains pebbles. It lies on 
 horizontal beds, which, according to my friend, consist of " a 
 soft friable rock, generally of a light-gray color, often hard 
 and laminated like that used for flagging our sidewalks ; 
 sometimes it is a red sandstone, very soft ; this latter is 
 what Fletcher says Sao Paulo is paved with." This same 
 material forms the great plain extending west to Sao Joao 
 do Rio Claro, and, as I have already remarked, tying in 
 with the formation of the plains of the Parana and the 
 south. This formation I believe to be the same as that 
 occupying the valley of the Jequitinhonha, and which I have 
 referred to the tertiary. f 
 
 * Literally, deep-red earth ; the ordinary red soil of liio goes by the name of 
 barro vermelho. 
 
 t Since writing the above I have received from Major James a specimen of 
 this rock, which is precisely like the clayey sandstone of the chapadas of the 
 Jequitinhonha. 
 
PROVINCES SOUTH OF RIO. 515 
 
 At Sao Joao de Ipanema, near Sorocaba and about twenty 
 leagues southwest of Sao Paulo, there are beds of sandstone 
 and limestone associated with diorites and porphyry, with 
 heavy deposits of magnetic iron ore.* This ore is mined 
 to a considerable extent, and is smelted almost on the 
 spot. 
 
 Von Eschwege says that the sandstones are modern- 
 secondary.! I would suggest a comparison between them 
 and the sandstone and iron deposits of the Sao Francisco 
 River, described by Burton. 
 
 These iron deposits were discovered in 1578 by one 
 Affonso Sardinha, who is said to have found at the 
 same time J " a vein of silver (?), of whose extraction the 
 government took charge ; but as the expense was great 
 all was soon abandoned, and these sites remained unpeopled 
 until 1803, an epoch in which some naturalists, exploring 
 the serras of the district of Sorocaba, came to recognize the 
 true importance of the iron-mines of Guara^oiava. After 
 seven years the prince-regent brought over from Sweden, at 
 no small expense, a company of miners, under the direction 
 of an individual of the same nation, named Hedberg, who 
 set up four forges, which, owing to their bad arrangement, 
 were of no service In 1815 new forges were con 
 structed by order of the same prince, together with a,fabrica 
 on a larger scale than the first, and the Count of Palma .... 
 was charged with the direction of the work of the engineers 
 and the superintendence of everything. This governor or 
 dered two enormous furnaces to be built beside those which 
 
 * A collection of specimens from this locality made by Olfer existed in 1828 
 in the Konigl. Mineralogische Museum at Berlin. 
 t Ann. des Mines, 8 me Vol. 182.3, p. 405. 
 | Die. Geo(j. t art. Sao Joao d Hipancma. 
 
516 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 still existed." * " The iron manufactured is excellent, and 
 the ore gives fifty to eighty-five pounds of metal to the 
 quintal. At present there arc two high furnaces measuring 
 eight metres in height, producing regularly 3,000 kilo 
 grammes of cast-iron in twenty-four hours with uninter 
 rupted work." 
 
 The region in which these iron-works are situated is cov 
 ered with forest, and wood is the fuel used. The flux em 
 ployed is limestone and diorite. In the immediate vicinity 
 of the furnaces is found an excellent quality of sandstone, 
 of which a refractory kind is employed in the lining of the 
 furnaces. 
 
 The serra Araassojava (Guara^oiava), according to Spix 
 and Martins,! is an isolated ridge, about one thousand feet 
 high above the level of the river Ipancma, and a league in 
 length from north to south. The ore occurs in a large de 
 posit, and our authors speak of seeing a mass of it forty 
 feet perpendicular. It is associated with a yellow sand 
 stone with a scanty argillaceous cement, and a clay slate 
 or shale of a dirty lavender color. Its strike is east-west. 
 In the same hill a porous quartz rock of a light-brown color 
 was observed, containing light-blue chalcedony. On the Rio 
 Tiete at Araraitaguaba Spix and Martius report a sandstone 
 similar to that of Ipanema. 
 
 The soils of the Province of Sao Paulo are exceedingly 
 fertile, and the climate is favorable to the growth especially 
 of coffee, sugar, and tobacco. Coffee flourishes exceedingly 
 well on the campinas west of Sao Paulo, and it is probable 
 that there is no more valuable coffee region in Brazil. The 
 climate is so mild in the vicinity of the city of Sao Paulo 
 
 * Cntaloyo da Seyunda Expoaicdo National de 1866, p G9. 
 t Reise, Vol. I. pp. 253, 254. 
 
PROVINCES SOUTH OF RIO. 517 
 
 that many European plants may be successfully raised, such 
 as flax, wheat, the vine, peaches, <fcc., c. The climate on 
 the higher lands is very healthy and agreeable, and well 
 suited to Europeans. The province has now nearly 800,000 
 inhabitants, and counts a considerable number of flourishing 
 
 7 o 
 
 towns. There are some German colonists in the province, 
 and I understand that a considerable number of Americans 
 from the Southern States have settled there. 
 
 The Province of Parana, which lies just south of that of 
 Sao Paulo, has very nearly the same topographical features, 
 being low along the coast, rising more or less steeply to 
 the summit of the plateau, and then sloping off towards the 
 Parana in extensive campinas. The capital is Curitiba, 
 situated, like Sao Paulo, at some distance west of the edge 
 of the plateau. The greater part of the country is covered 
 by heavy forests, though in the northwest there are extensive 
 plains. An important business of the province consists in 
 the raising of cattle, of which large numbers are exported. 
 Coffee, cotton, potatoes, sugar-cane, Indian corn, wheat, veg 
 etables of different kinds, <fec., are cultivated to a consider 
 able extent. The " herva mate" or matte of the Brazilians, 
 Ilex paraguay ensis, or Paraguayan tea, largely grown in 
 the Paraguayan republic, and used in lieu of Chinese tea, 
 is grown in large quantities and forms an article of export. 
 Bousquet* says that the Chinese tea grows well, but that it 
 is not as yet much cultivated. Tea will also grow at Rio, 
 Sao Paulo, and elsewhere in the south. Bousquet speaks 
 of the vanilla as growing spontaneously in the vicinity of 
 Paranagua. The province has never been explored by a 
 competent modern geologist. It is just possible, as admitted 
 
 * Note sur la Province de Parana, par M. Bousquet, Bull, dc la /Societe de 
 Geographic, 5 me Se rie, T. 9, p. 528. 
 
518 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 by Mr. Plant, that coal may be found on the low lands 
 between the Serra do Mar, or Cubatao, as it is commonly 
 called in Parana, and the sea. Bousquet says native mer 
 cury occurs near Paranagua. Gold and diamonds are found 
 on the borders of the Rio Tibagy.* 
 
 This province, like Sao Paulo, is well watered by large 
 streams. Between the two provinces runs the Parana- 
 panema, on the west is the Parana and to the south the 
 Uruguay, while the Ivahy and Tibagy, both rivers navigable 
 for canoes,! run through the province, one emptying into 
 the Paranapanema and the other into the Parana. The 
 Parana is navigable for about ninety leagues from the falls 
 of Urubupunga in the Province of Goyaz to nearly opposite 
 the Tiete in Sao Paulo, where navigation ends at the island 
 of Sete Qucdas. The Salto Grande on the Parand is de 
 scribed as being comparable to the Caxoeira de Paulo 
 Affonso. 
 
 So far as water highways are concerned, the province is 
 well furnished. 
 
 There is a little Brazilian-French colony, called Santa The 
 resa, established on the Ivahy, and another colony of some 
 five hundred inhabitants, called Supcraguy, near Paranagud, 
 of which the inhabitants support themselves by agriculture 
 and fishing. 
 
 The seaport of the province is Paranagua, a considerable 
 town, situated on a large and beautiful bay which forms an 
 excellent and spacious harbor. 
 
 The province of Santa Catharina lies just south of Paranft. 
 It is one of the most fertile of the provinces of Brazil, and is 
 
 * Olivcira, Exploraqao de Mincraes, p. 11. 
 
 \ Vereker " On Brazilian Province of Parana," Jour. Roy. Geog. Soc., Vol. 
 XXXII. p. 1.37. 
 
PROVINCES SOUTH OF RIO. 519 
 
 blessed with a delightful and temperate climate ; but except 
 on the seacoast it is not well settled, and the population is 
 probably not more than 150,000. The capital is Desterro, a 
 large town, delightfully situated on the western side of the 
 island of Santa Catharina. There are several flourishing 
 colonies in the province. 
 
 Between the edge of the plateau and the sea there exists 
 in Santa Catharina a coal-basin, in which along the banks 
 of the Rio Tuberao beds of bituminous coal of fair quality 
 are exposed. The Visconde de Barbacena is interested in 
 the development of this region, and we may hope erelong to 
 see some results of his labors. I am not aware that any 
 report on this field has yet been published. I understand 
 that the coal-beds lie quite flat, as elsewhere to the south. 
 The Province of Rio Grande do Sul is the most south 
 ern of Brazil, lying just north of Uruguay, and situated be 
 tween 27 50 and 33 45 south latitude. Much of the 
 country is hilly, particularly in the eastern and northern 
 portions, but towards the west and south it consists of plains 
 covered with herbs and forming the pasture-grounds of herds 
 of cattle. The northern and eastern portion is more or less 
 heavily wooded. The interior has never, so far as I can 
 learn, been scientifically explored, and I am unable to find 
 any reliable description of its physical features. 
 
 Of the geology of the province we have little information 
 except with reference to the coal-mines. 
 
 In an article in the Quarterly Journal of Science, No. 11, 
 1864, Mr. Edward Hull says that the first notice of the coal 
 deposits of the Province of Rio Grande do Sul was taken by 
 one Guilherme Bouleich in the year 1859. This is not quite 
 correct.* In May, 1858, Dr. Ave-Lallemant visited coal- 
 
 * The real discoverer of the Brazilian coal-fields was Dr. Terigot, of Flush- 
 
520 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 mines on the Arroio dos Ratos, which were at that time 
 worked on a small scale. He refers to an examination of 
 the locality made some time before by a Bacharel Vascon- 
 cellos.* Dr. Lallemant describes two horizontal seams of 
 coal, an upper, worked, about four feet thick ; a few feet 
 below this seam a second was found of the same thickness. 
 I have not had the opportunity of examining specimens of 
 this coal ; but I learn that it is a fair bituminous variety, with 
 a more abundant ash than the English coals sent to Rio. It 
 has been in use some ten years on the steamers of the Com- 
 panhia Jacuhy, and has been found more economical to 
 employ it than the English coal. That of the Arroio dos 
 Ratos sells for 138000- 171000 (86.50 -88.50 American 
 currency) per ton, which, as we shall presently see from Mr. 
 Plant s report on the Jaguarao coal-field, is a much lower 
 price than that of the English coals. There are three sep 
 arate coal-fields in the Province of Rio Grande do Sul,f 
 all of which Mr. Nathaniel Plant has studied with care. 
 This gentleman informs us that the basins are separated 
 from one another by rolling hills of syenite, mica-schist, 
 and granite, together with trachytic and basaltic rocks. 
 
 The largest of these basins is situated in the Jaguarao and 
 Candiota valleys, between lat. 31 and 32 S. and long. 53 
 and 54 E. 
 
 ing, Long Island, who was engaged in 1841 by the Brazilian government to 
 make geological observations in the Empire. This gentleman reported the ex 
 istence of a coal-field in Santa Catharina, some three hundred miles long from 
 north to south and twenty to thirty miles wide. 
 
 * Ave -Lallemant, Reise durch Siid-Brasilien im Jahre 1858, Theil I. p. 478. 
 
 t The Brazilian Coal-Fields, by Nathaniel Plant, F. R. G. S., &c., Geological 
 Magazine, Vol. VI. No. 4, April, 18G9. I owe a copy of this paper to the 
 kindness of Mr. Plant s brother, Mr. John Plant, Curator of the Royal Museum 
 of Salford. Much credit is due Mr. Plant for his long-continued exertions in 
 working out the structure and limits of these coal-fields, and for his persistence 
 in endeavoring to bring the coal into market. 
 
PROVINCES SOUTH OF RIO. 521 
 
 Mr. Plant presented Professor Agassiz in 1865 with a 
 short description of this coal-field, together with specimens 
 of rocks, fossils, and coals, which I have examined. The fos 
 sils are of true and characteristic carboniferous genera, and 
 no doubt can exist as to the equivalency of the deposits.* 
 I give Mr. Plant s paper in full.f One of the photographs 
 mentioned I have had engraved as a frontispiece to this 
 volume. 
 
 I may add that Mr. Plant has been for several years ex 
 ploring the coal-basins of the province with a view to having 
 them worked by a mining company. It is to be hoped that 
 his efforts may be abundantly successful. 
 
 Coal-Fields of the River Jaguarao, and its Tributaries, the Rivers 
 Candiota and Jaguardo-chico in the Province of Rio Grande do 
 Sul, Brazil. 
 
 The coal-basin of the river Jaguarao is situated in the southern 
 part of the province of Rio Grande do Sul, between lat. 31 
 and 32 S., and long. 324 and 325 (French meridian) in the 
 valley of the Jaguarao and its tributaries, the rivers Candiota 
 and Jaguarao-chico. It covers an area of" about fifty miles by 
 thirty, its greatest diameter being from north to south. The coal 
 strata which the geological section illustrates, and whence the ac 
 companying specimens have been obtained, and the thickness of 
 the beds determined, are exposed in an elevated escarpment on the 
 banks of the river Candiota, at a place called " Serra Partida," $ 
 where they appear in the following order of superposition : 
 
 * Strange to say, that, after all the explorations made of these coal-fields, their 
 carboniferous age was long in being recognized. Affonso Mabildc, in a report 
 made to the president of Rio Grande do Sul, says that the coal is a lignite of 
 tertiary age. 
 
 t This paper also appears in Fletcher s work. I presume Mr. Plant fur 
 nished him with a duplicate copy of the MS. 
 
 t See frontispiece, which is engraved from a photograph presented to Pro 
 fessor Agassiz by Mr. Plant. 
 
522 
 
 113 ft. 
 
 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 Soil 1 ft, 
 1. Ferruginous sandstone, 28 ft. 
 
 2. Coal shale, 9 ft. 
 
 3. Sandv shale, 5 ft. 
 
 4. Coal, 8ft. 
 
 6. White fossiliferous shale, 5 ft. 
 
 6. Coal, 11 ft. 
 
 7. Parting of blue clay. 
 
 8. Coal, 17 feet. 
 
 9. Fossiliferous clay, 9 ft. 
 
 10. Coal, 25 ft. 
 
 Thickness 
 
 not 
 determined. 
 
 11. Ironstone shales with fossil ferns. 
 
 12. Sandstone. 
 
 13. Limestone. 
 
 14. Mica- schist. 
 
 15. Metalliferous limestone. 
 
 No. 1. The uppermost bed is composed of sandstone of a highly 
 ferruginous nature, resembling in its appearance the " Ore s 
 Bigarre" " of Europe ; it contains nodules of a silicious peroxide 
 of iron, yielding from twenty-five to thirty-five per cent of metal. 
 It varies considerably in its thickness, in some places being com 
 pletely worn away, and in others attaining a depth of upwards 
 of two hundred feet. Immediately below this occurs a bed, 
 
 No. 2, of coal-shale, very argillaceous, and perhaps unfit for fuel ; 
 it possesses a thickness of nine feet, and can be seen cropping 
 out wherever the superincumbent bed has been denuded ; it 
 rests upon a bed, 
 
 No. 3, of sandy-ochreous shale containing septaria of an ochreous 
 oxide of iron, which, together with the ironstone found in the 
 sandstone, will, in all probability, be turned to profitable ac 
 count when the coal-beds are worked ; underneath this is, 
 
 No. 4, a bed of bituminous coal, three feet thick. The min- 
 
PROVINCES SOUTH OF RIO. 523 
 
 eral, although it leaves a high percentage of ash, will be found 
 useful in smelting the iron ores from the interstratifying beds ; 
 and there is every reason for supposing that it will be found 
 of a better quality when the bed is fairly worked. The samples 
 tested were taken from very near the surface, which may in 
 some measure account for its apparent impurity; it rests on, 
 
 No. 5, a bed of white clay or schist, containing innumerable im 
 pressions of fossil plants (perhaps aquatic), the general appear 
 ance of which would lead one to conclude that these carbonifer 
 ous deposits belong to a later period than that assigned to the 
 coal-measures of England and the United States, were such a con 
 clusion not confuted by the fossil ferns found in the other inter- 
 stratifying shales ; it has a thickness of five feet, and overlies a, 
 
 No. 6, bed of good coal eleven feet thick. This coal resembles 
 very much in its appearance the Newcastle, and may be traced 
 for many miles along the banks of the river Candiota, sometimes 
 forming the bed of that river and of the small streams falling 
 into it ; it is separated from another seam by a thin parting, 
 
 No. 7, of blue clay. 
 
 No. 8. The coal of the lower bed appears to be even of a better 
 quality than the one above ; it has a clean, shining fracture, 
 and in some places thin seams of pure cannel-coal may be 
 traced along the bed. It is highly inflammable, boiling up 
 like oil during combustion. This coal has been used as fuel 
 in various ways with marked success. It has been tried on 
 the steamers navigating the Lagos dos Patos in the province 
 of Rio Grande, and although it left a greater portion of ash 
 than the Cardiff coal, it was found to be a good caking coal, and 
 served every purpose of a steam fuel. Below this is another bed, 
 
 No. 9, of blue clay, containing vestiges of fossil plants ; in everything 
 else it is similar to the upper bed of the same mineral. It 
 reposes on, 
 
 No. 10, the thickest seam of coal exposed in the escarpment at 
 the Serra Partida. This is the lowest bed of coal exposed in 
 
524 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 any part of the coal-fields of Candiota, but, in all probability, 
 other beds will be found nearer to the centre of the basin, or 
 this, as well as the incumbent beds, may become thicker, judg 
 ing from the fact that all the beds appear to thicken as they 
 approach the middle of the valley of the river Jaguarao. The 
 great thickness (twenty-five feet), the good and homogeneous 
 character of the seam, are important features in this coal-field. 
 .The mineral (although taken from near the decomposed face of 
 the cliff on the river Candiota) was found to leave even less ash 
 than that from the seam above. It has frequently been used 
 on steamers with the same success as that obtained from New 
 castle coal. The coke obtained from this coal by Mr. W. G. 
 Ginty of the Rio Gas Works (vide Mr. Ginty s report) was even 
 better than that derived from Newcastle coal. It overlies a bed, 
 No. 11, of ironstone shale, which, in a scientific point of view, 
 is the most important deposit of the coal-measures of the 
 Jaguarao, from the fact of its containing impressions of or 
 ganic remains, by which the geological age of the coal-beds 
 can be determined ; the fossil plants found embedded in this 
 shale all belong to the same genera as those which charac 
 terize the coal-fields of Britain and the United States, the 
 most abundant belonging to the genera Lepidodendron * and 
 
 * Three species of plants have been described by Mr. W. Carruthers, viz. : 
 
 Flemingites Pedroanus CARRUTHERS. Geol. Mag., Vol. VI. No. 4, p. 5, PI. P. V. 
 
 " Stem lepidodendroid, scars small, obovate, without any markings ; base of 
 the petiole permanently attached to the stem ; leaf slender, linear. Fruit a 
 cone (?) the scales of which support numerous sporangia." 
 
 The stems of this plant at first sight resemble those of some species of 
 Lepidodtndron ; but they may be readily distinguished from Lepidodendron by 
 the character of the scars, which never show any impression from articulating 
 surfaces. Associated with the stems and leaves of this species are found in great 
 abundance minute flattened bodies which Mr. Carruthers considers to be the 
 sporangia of the plant under consideration. 
 
 Odontopteris Plantiana CARRUTHERS. Loc. cit, p. 9, PI. V. Figs. 2 & 3. 
 
 " Pinnules broad at the base, irregularly lobed, obtuse at the apex ; basal 
 
PROVINCES SOUTH OF RIO. 525 
 
 Glossopteris ; others have been recognized as being similar 
 to the ferns found in the very oldest secondary rocks, thus 
 leaving no uncertainty as to the true carboniferous charac 
 ter of the coal-measures of the river Candiota. This seam is 
 very prolific of fossils, and there can be no doubt that when 
 these immense beds of mineral treasure are worked many new 
 and interesting forms of vegetable life will be brought to light 
 to enrich our knowledge of the coal-fields of the Southern 
 hemisphere. The ironstone shale is very rich in metal, and 
 will doubtless be worked as an iron ore when the mines are 
 opened. Below this there occurs another bed, 
 
 No. 12, of sandstone, similar in all respects to the uppermost bed. 
 After which is a bed, 
 
 No. 13, of fine crystalline limestone, containing small fragments 
 of graphite, disseminated throughout the mass. It is trav 
 ersed also by veins of a very pure carbonate of lime in the form 
 of double-refracting spar, which, in some places, attain a con 
 siderable thickness. This limestone will not only be of immense 
 value for manufacturing into lime, but also as a flux in smelt 
 ing the iron ores. The three things essential for the erection of 
 smelting-works are thus found in the same district interstratified 
 with each other : the ore, the fuel, and the flux, all of the very 
 first quality, a combination of mineral riches (only waiting for 
 the hand of man to realize them) scarcely to be found together 
 in one spot in any other part of the globe. Evidently, 
 
 No. 14, the two lowest beds of these coal-measures are mica-schist, 
 
 No. 15, and another limestone rock of a very dark and compact 
 
 pinnules large, much and irregularly lobed ; nerves arcuately parallel, dichoto- 
 mous." 
 
 Nceggerathia obovata CARRUTHERS. Loc. cit.., p. 9, PI. VI. Fig. 1. 
 
 " Frond sessile, flat, entire, elongate obovate, attenuated towards the base ; 
 nerves dividing dichotomously, parallel." 
 
 I have seen Catamites and a Sphenophms (n. sp.) in specimens of coal shale 
 from one of south Brazilian coal-fields ; but the labels having been lost, I have 
 been unable to determine the exact locality. C. F. H. 
 
526 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 nature. It is scarcely possible to determine which is the lower 
 most, as in some places the mica-schist * is seen lying on the 
 syenite which surrounds the coal-basin, and in others the lime 
 stone ; the name of " metalliferous limestone " has been given 
 it, owing to the innumerable crystals and thin veins of sul- 
 phuret of iron which appear in it. In all probability other 
 metalliferous veins will be found in this limestone. 
 Nearly the whole of the coal-basin of the valley of the Jaguarao 
 is enclosed by syenitic hills of from two hundred to three hun 
 dred feet high ; the sides towards the coal-field slope gently down 
 wards till they disappear under the sandstone overlying the coal ; 
 on the other side the syenite, after presenting an uneven and undu 
 lating aspect for some three or four leagues, gradually subsides 
 into an even country, which continues on almost perfectly plain f till 
 the seaport city of Rio Grande do Sul is reached, so that the com 
 pany (already formed for making the survey for a railway to carry 
 the mineral riches of the valley of the Jaguarao down to a seaport, 
 where the coal can be shipped to the different ports along the coast 
 of Brazil, and the River Plate) will find no difficulty in discover 
 ing a route along which a cheap line of rails can be laid down. 
 
 The photographic views of the different escarpments in which 
 the coal-beds are shown along the river Candiota will show the 
 great facilities afforded for working the coal in almost any part of 
 the basin by open cuttings. Tramways can be laid down branch 
 ing off in different directions from the main trunk line, along 
 which the coal-wagons can be run right into the seams of coal, 
 thereby rendering the sinking of expensive shafts quite unneces 
 sary. 
 
 The general dip of the beds is from 5 to 10 S. "W., and in no 
 place are there signs of subsequent upheavals or dislocations of 
 
 * The mica-schist is without doubt much older than the carboniferous. 
 C. F. H. 
 
 t Mr. Hull speaks of a gently sloping plane of basalt stretching from the 
 coal-mines toward the port ou the Rio Gon9alo. (Quarterly Journal of Science, 
 No. 11, April, 1864.) 
 
PROVINCES SOUTH OF RIO. 527 
 
 strata visible, so that very little obstruction will be met with in 
 carrying the tramways along the seams as the working of them 
 goes forwards. 
 
 It is almost unnecessary to dwell upon the immense value of 
 these coal deposits as a commercial enterprise, when it has been 
 already ascertained, by a "running survey" of the country be 
 tween the seaport of Rio Grande do Sul and the coal-mines of 
 Candiota, that in all probability the coal will be delivered on 
 board vessels lying in the port of Rio Grande at, perhaps, less than 
 7$000 per ton, where it is at the present moment being sold at 
 24$000 ; and as soon as a bill is passed allowing vessels of all 
 nations to trade between the Brazilian ports, there will be no lack 
 of enterprising shipowners to carry the Rio Grande coal to Rio de 
 Janeiro in which port alone the enormous amount of 180,000 tons 
 of coal are annually imported for a price which will enable the 
 coal-mining company to sell the Candiota coal in the market of 
 the capital of the Brazilian Empire for about 15$000 per ton, 
 a price which will annihilate any competition from foreign mar 
 kets, seeing that the foreign coal is seldom sold for less than 
 22$000 per ton. 
 
 The consumption of coal in the river Plate is perhaps as great 
 as that of Rio de Janeiro, and the facilities for supplying the 
 markets of Buenos Ayres and Montevideo from the coal-mines of 
 the river Candiota are still greater than those for supplying Rio. 
 The coal can be sent from the mines, put on board colliers, and de 
 livered in Montevideo in three or four days, at about half the 
 cost of delivering the same article in Rio, which is a market where 
 coal is never less than fifteen dollars per ton (or 30$000). 
 
 The consumption of coal along the Brazilian coast and in the 
 River Plate increases yearly, and in all probability it will be 
 found, after the coal-mines of Candiota have been opened for a few 
 years, that a single line of railway will not be found sufficient to 
 carry the supply of coal to meet the increasing demands. 
 
 NATHANIEL PLANT. 
 Rio DE JANEIRO, 20th July, 1865. 
 
528 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Report on the Candiota Coal by W. G. Ginty, Engineer-in-Chief of 
 the Rio de Janeiro Gas Works. 
 
 MB. NATHANIEL PLANT : 
 
 DEAR SIB, I have received and examined your samples of 
 Brazilian coal from Candiota with great interest, and I am glad to 
 be able to congratulate you on its really good quality. 
 
 The samples you sent me were too small for complete and satis 
 factory analysis in the apparatus at my disposal. I found also 
 that the samples varied a good deal in appearance and quality. 
 This has arisen, no doubt, from their having been obtained from 
 various positions on the nearly perpendicular face of the immense 
 stratum, and from variable -periods of exposure, as, owing to the 
 crumbling away or disintegration of pieces under the incessant 
 action of the weather, these samples may have been exposed for 
 periods varying from each other as seconds do from centuries. 
 
 The Candiota coal resembles the Newcastle steam-coal (which 
 comes to this market at least) very much in structure, cleavage, 
 and general appearance, nor does it differ very much from New 
 castle coal in its useful properties, except that it contains more 
 than double the quantity of ash, which is detrimental to its heat 
 ing powers ; but this objection is likely enough to disappear alto 
 gether in samples from the deeper parts of the mine. 
 
 The coke from the Candiota coal is, however, very different 
 in appearance from that of the Newcastle coal, and resembles the 
 coke of (what is sold here as) Cardiff coal in its silvery-colored 
 laminations. 
 
 Some of this Candiota coal, however, especially that of the 
 lower seam is very friable, and is evidently what is called caking 
 coal ; that is, it boils or becomes molten during the process of 
 carbonization. However, all the qualities of the coke from the 
 Candiota coal are very good. 
 
 As you say the dip or inclination of the seams or strata of 
 this Candiota coal is 5 from the plane of the horizon, I think it 
 most reasonable to presume that a much finer, more compact, and 
 
PROVINCES SOUTH OF RIO. 529 
 
 equable quality of coal may be calculated upon at lower depths. 
 5 is a gradient of 1 in 11.4 or 8.77 per cent, or 462 feet per mile. 
 Thus in such an immense field as you have described to me there 
 is ample margin for obtaining other than surface coal, which for 
 obvious reasons in Brazil, as elsewhere, cannot be as pure, as com 
 pact, or as uniform in quality as that obtained at great depths. I 
 shall watch the prosecution of your explorations in this direction 
 with great interest. 
 
 The following are the results of my examinations (as far as they 
 went) on the Candiota coal, the samples of Newcastle, Cardiff, 
 and Wigan cannel, with which it is compared below, having been 
 tried at the same time in the same apparatus : 
 
 Specific Per Cubic feet Illuminating 
 
 Gravity. Cent of Gas Power in 
 
 Water of per Ton Standard 
 
 1,000. Coke. Candles. 
 
 Candiota coal (mean of three quantities) 1,240 63 6,900 5.00 
 
 " " Lower seam .... 1,230 60 8,198 5.80 
 
 Newcastle 1,250 62 
 
 Cardiff 1,275 80 
 
 Gas, or Cannel Coal (Case and Morris) 1,240 62 9,600 20.50 
 
 From the appearance of the lower seam I do not despair of 
 your finding a good gas coal for us in the Candiota district, and 
 thus freeing the Brazilian gas companies from the fearful tax they 
 have to pay in the shape of freights from England, amounting 
 to from 200 to 300 per cent on the value of the materia prima. I 
 send you labelled samples of the different qualities of coke above 
 referred to. I remain your obedient servant, 
 
 (Signed) W. G. GINTY, Mem. Inst, C. E., 
 
 Engineer Rio de Janeiro Gas Company. 
 
 Mr. Plant describes a second coal-basin exposed in the 
 valley of one of the tributaries of the Rio Jacuahy, called 
 the Sao Sepe*, in lat. 30 20 , long. 53 30 . 
 
 There are two beds of coal exposed in this basin, one 
 fourteen, the other seven feet in thickness. The sandstone 
 beds overlying these veins are disturbed and overflowed by 
 
 VOL. I. 23 HH 
 
530 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 trachytic eruptions. The area of this basin, so far as is 
 known, is about fifteen miles. 
 
 Near the town of Sao Jeronymo, on the banks of the 
 Jacuahy, is a third coal-basin, situated in lat. 30, long. 51 
 80 . At a depth of fifty-seven feet below the surface there 
 is at this locality a bed of highly bituminous coal six feet in 
 thickness. This is underlaid by a bed of ironstone shale 
 with fossils like those so common at Candiota. Several 
 other coal-beds of from two to six feet in thickness, and 
 interstratified with clay and ironstone, have been passed 
 through by the shafts of the mines, which have reached a 
 depth of one hundred and eighty-three feet. 
 
 Since the foregoing was sent to press I have had the op 
 portunity of examining the valuable paper of Weiss on the 
 collections made by Sellow in Southern Brazil.* 
 
 I regret that want of space prevents my giving a short 
 abstract of this paper. I confine myself, therefore, to a 
 quotation from the introduction.! 
 
 " On the 30th degree of south latitude," says Weiss, " there ex 
 tends from the sea in a direction west a few degrees south (in der 
 7 ten Stunde J) obliquely into the country for more than 5 of 
 latitude a range of hills of basalt, that is of amygdaloid and 
 black porphyry or melaphyr, but nowhere is the rock true 
 basalt. 
 
 * Uber dan siidliche Ende des Gebirr/szufjcs von Brasilien in der Provinz S. 
 Pedro do Sul und der Banda Oriental oder dem Staate ron Monte Video ; nach 
 den Sammlungen des Ilerrn Fr. Sellow, Abhandlungen dor Koniglichen Akade- 
 mie der Wissenschaften, zu Berlin, aus dem Jahre 1827. 
 
 t Loc. cit., p. 222. 
 
 J The German miner s compass, according to Gehler s Phyrikalifchet Wb rter- 
 buch (see article on Compass) was divided from north to south and south to 
 north into twelve hours (Stunden) instead of degrees. 
 
 Melaphyr is the name given by Von Buch to a species of porphyry found 
 
PROVINCES SOUTH OF RIO. 531 
 
 " On the southern side of its round and roof-shaped hills lies 
 the valley of the Guaiba or Jacuy, with Porto Alegre at the junc 
 tion of four rivers with the Lake Viamao. Nearer the Uruguay it 
 slopes to the west, but even there it is united with a range of the 
 same character extending southward (in der 12 ten Stunden) sepa 
 rating the head-waters of the Ibicuy from those of the Caavera, 
 forming between the Caavera, Ibirapuitam Grande and Ibirapuitam 
 Chico considerable ridges, and when it turns to the west toward 
 Salto Grande it sends streams southward to the Daiman and Rio 
 Negro, northward into the Arapey, Quaraim and Ibirapuitam. 
 This chain of amygdaloid is the source of the great quantities of 
 chalcedonies, agates, carnelians, rock-crystals, and amethysts which 
 cover the banks of Uruguay downward below the Rio Negro. 
 These uniting chains divide the country in a natural manner into 
 a northern and southern half, while farther south there is no 
 mountain-chain, as formerly supposed, to form the boundary be 
 tween the Portuguese and Spanish possessions. The amygdaloid 
 formation is probably continued on the Uruguay above, since it 
 forms reefs and cliffs on the river at Salto Grande and Salto Chico, 
 as well as fourteen leagues farther up at the Capella de Belem, and 
 among others the map of Nunez gives just above the Saltos Grande 
 and Chico a Monte Grande del Monteil. On the slope of the 
 amygdaloid formation there is spread an extensive clayey sandstone 
 formation extending into the country, and it ascends to the foot 
 of the granite coast mountains. It is certainly very young, and 
 much younger than was supposed by one traveller (Sellow), who 
 referred it principally to the Permian (Rothliegende). From its 
 character as well as its situation, it is extremely probable, not to 
 say certain, that it is tertiary, and may be provisionally referred 
 by us to the Molasse or Braunkohlen sandstone." 
 
 in the Alps. The ground mass is black or blackish gray with embedded crystals 
 of Labradorite (?) and Augite, with occasionally mica, hornblende, and pyrites. 
 Sometimes it becomes amygdaloidal. See Meluphyr in Handworterbuch der 
 reinen und anrjewandten Ckemie. 
 
532 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 CHAPTER XVIII. 
 
 THE GOLD-MINES OF BRAZIL.* 
 
 Geological Distribution of Gold in Brazil. Gold in Gneiss, at Jaragua, Can- 
 tagallo, Pianco, and elsewhere. The richest Deposits found in Veins 
 traversing Clay Slates. Character of Auriferous Quartz. Granular 
 Quartz, or Caco. Gold when associated with Sulphides rarely visible. 
 The auriferous Iron Ore, Jacutinga. Gold-Mines of Sao Joao d El-Rei. 
 The Morro Velho Mine, Mode of Occurrence of the Gold, Method of 
 Extraction, Yield, &c. The Congo Socco Mine. The Rossa Grande 
 Gold-Mining Company. Mines at Morro do Santa Anna, Congonhas do 
 Campo, Sao Vicente, Cata Branca. The Gold-Mines of Brazil not yet fairly 
 developed. 
 
 THE gold of Brazil f occurs in the ancient metamorphic 
 rocks, and in drift gravels and clays, and alluvial sands and 
 gravels derived from the wear of these rocks. 
 
 The eozoic gneiss of the coast-belt furnishes gold at 
 numerous localities along its whole extent. $ The mines of 
 
 * In this chapter I have not tried to treat exhaustively of this subject. I 
 have endeavored rather to present such facts as will enable the reader to obtain a 
 fair idea of the mode of occurrence of gold in Brazil, and of the character of 
 the mines. A great amount has been written on the gold-fields of Brazil, but 
 it is for the most part lacking in scientific accuracy, and much of it has rather 
 tended to obscure than throw light upon their real structure. 
 
 t For a short and interesting sketch of the early history of gold-mining in 
 Minas Geraes, see Burmeister s Reise in Brasilien, p. 590, " Zur Geschichte der 
 Goldminen und ilirer ersten Entdeckung." Consult also Von Eschwege s Pluto 
 Brasiliensis, and Mawe s " Travels in the Interior of the Brazils." Spix and 
 Martius, St. Hilaire, Castelnau, and almost all writers on the interior of Brazil, 
 have had more or less to say about the gold-mines. 
 
 J I think that the auriferous deposits occur in the upper part of the gneiss 
 
THE GOLD-MINES OF BRAZIL. 533 
 
 Jaragua in Sao Paulo, of Cantagallo in Rio de Janeiro, and 
 of one of the tributaries of the Itapemerim, are among the 
 most important in this region. Gold also occurs in the 
 gneiss of the Mucury Basin and in the north, as at Pianco. 
 Over this whole region the gold is found rather sparingly, 
 and appears to be derived from the quartz veins traversing 
 the gneiss ; but the only instance I know of where gold has 
 been extracted from a quartz vein in the gneiss-belt is the 
 mines on the Rio Bruscus in Parahyba. With this excep 
 tion mining has been confined entirely to the washing of the 
 cascalho * underlying the drift-clays, and of the gravels and 
 sands of the rivers. These washings have all been aban 
 doned. Between the coast gneiss-belt and the sea I know 
 of no auriferous deposits, but in very numerous localities in 
 the interior of the country the newer metamorphic rocks are 
 rich in gold. 
 
 The formations affording most gold are clay slates trav 
 ersed by auriferous quartz lodes, the itacolumite rock which 
 is also veined with auriferous quartz, and certain iron ores 
 variously known under the names of Itabirite and Jacu- 
 tinga.f All these formations are, I believe, of lower silu- 
 rian age. 
 
 In the clay slates the quartz veins sometimes show free 
 
 series. According to Mr. Wall, " Geology of a Part of Venezuela," Quart. Jour. 
 Geol. Soc., Vol. XVI., p. 460, gold is found disseminated in gneiss near Valen 
 cia, Venezuela. Gold occurs in the gneisses of the Itacama Mountains. (Tatc.) 
 
 * The auriferous cascalho of Brazil is quite a different thing from the cascajo 
 of Venezuela. The Spanish word has the same derivation as the Portuguese, 
 but according to Mr. Lc Neve Foster, it is applied in the Caratal gold-field to a 
 decomposed schist, on which the pay-dirt rests. 
 
 t This term is derived from two Tnpi words, Jacu, a kind of bird ( Penelope] , 
 and tinga, white. The name was given to the rock because of its resemblance 
 in color to the feathers of the above bird. Sometimes foreigners spell it jaco- 
 tinga, which is not correct. 
 
534 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 gold with very little pyrites associated. The auriferous 
 quartz varies much in character. Sometimes it is compact 
 and milky, at others, as at the Cata Branca mines, it is very 
 granular and sugary. A specimen of vein-quartz with gold, 
 from the mine of Sao Vicente in Minas Geraes in the 
 Museum of Comparative Zoology, is composed for the most 
 part of clear, colorless quartz in rather coarse granules, 
 giving to the rock the appearance of a pure quartz sand 
 stone, or of white lump sugar ; * but in the same specimen 
 the quartz passes into a more compact rock, which has a 
 bluish look. 
 
 In the quartz veins, as elsewhere the case, the rock is not 
 all auriferous, but the gold runs in streaks. The Sao Vicente 
 specimen above described shows well-marked streaks, rich 
 in free gold, which appear to have run parallel with the side 
 of the lode. 
 
 Where the vein rock is rich in sulphides the gold is, as 
 a rule, not visible, but intimately mixed with the rock. 
 This is the case at Morro Velho. The sulphides consist of 
 magnetic iron pyrites, which is the most abundant and 
 yields a little gold ; common iron pyrites is less abundant, 
 and gives more gold ; and the mispickel or arsenical iron 
 pyrites, which is the principal gold-bearer. 
 
 Of the ferruginous auriferous deposits none is more inter 
 esting than the so-called jacutinga formation. Heusser and 
 Claraz f say that the jacutinga is a pulverulent variety of 
 
 * This sugary quartz goes in Brazil by the name of Cacd. 
 I Heusser and Claraz, Ann. des Mines, Tome XVII. p. 290 : 
 " L itabirite est simplement une varie te de fer oligiste schisteux qtti est ac- 
 compagne e de quartz et de mica. Elle presente quelquefois des couches puis- 
 santes et tres-e tendues qui peuvent etre exploite es comme minerai de fer. Quand 
 elle est pulverulente, on la de signe sous le nom de jacotinga." 
 
THE GOLD-MINES OF BRAZIL. 535 
 
 Itabirite, a name given by Von Eschwege * to a rock com 
 posed of micaceous specular iron ore, compact specular 
 iron, rarely laminated, a little oxide of iron and quartz dis 
 seminated. It is the rock of which the Peak of Itabira 
 and the Serra da Piedade are composed. 
 
 Burton describes the jacutinga as follows : f " This sub 
 stance, of iron black with metallic lustre, sparkles in the 
 sun with silvery mica ; the large pieces often appear of a 
 dark reddish brown, but they crumble to a powder almost 
 black. The constituents are micaceous iron-schist, and 
 friable quartz mixed with specular iron, oxide of manganese, $ 
 
 and fragments of talc The floor rock at Cocaes is 
 
 fine micaceous peroxide of iron (specular iron), thin and 
 
 tabular Much of the jacutinga is foliated 
 
 It shows great differences of consistency ; some of it is 
 hard and compact as hasmatite, and this must be stamped 
 like quartz. In parts it feels soapy and greasy, not harder 
 than fuller s earth ; it is easily wetted and pulverized, but 
 it is hard to dry." 
 
 Gardner describes the jacutinga as a soft friable greenish- 
 colored, micaceous iron-schist. 
 
 The gold of the jacutinga is free. Castelnau || says, that 
 at Gongo Soco it is always confined to a little vein which 
 winds about in the rock. This is never more than five to 
 
 * " Itabirit, Eisenglimmer, Eisenglanz, meist dichter, auch blattriger, hin 
 und wieder magnetischcr Eisenstein und wenig Quarz erschcinen entweder als 
 festes, dichtes Gcstein, oder haben ein Kornigschiefriges Gefiige." VON ESCH 
 WEGE, Geognostisches Gemalde von Brasilien, 28 st * Seite. 
 
 t Burton, Vol. I. p. 301. 
 
 t Castelnau speaks of the occurrence of manganese in the jacutinga of Gongo 
 Soco. 
 
 Travels in Brazil, p. 373. 
 
 || Op. cit., p. 246. 
 
536 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 seven millimetres in width, and sometimes it is as thin as a 
 hair ; it contains much manganese. 
 
 In the gold region is found an auriferous superficial de 
 posit of broken fragments of ferruginous rocks cemented 
 together, and called Tapanhoacanga or canga* 
 
 I have already described in previous chapters the mode 
 of occurrence of gold in the drift gravels and clays, and in 
 the sands and gravels of the river bottoms. 
 
 With these introductory words, let us now examine some 
 of the more noteworthy gold-mines of Brazil. 
 
 At Sao Joao d El-Rei and Sao Jose, situated a few miles 
 west of Barbacena, on the Rio das Mortes, are auriferous 
 deposits formerly worked for many years with great profit. 
 In 1830 an English company called the " Sao Joao d El- 
 Rei Mining Company " was formed, and these mines were 
 leased ; but in 1834 it was found that they were unprofit 
 able, and they were abandoned. Captain Burton says the 
 gold was principally obtained from a lode, which however he 
 does not describe. The jacutinga formation is said to occur 
 here. 
 
 The country over a very large area in the vicinity of Ouro 
 Preto is very auriferous, and here are situated the richest 
 gold-mines of the empire. 
 
 The gold occurs primarily in quartz veins traversing 
 various metamorphic rocks, such as clay-slate, mica-slate, 
 iron-schists, &c., and also disseminated through the rock in 
 some places ; and secondarily it is found widely distributed 
 in drift and alluvial sands and gravels. 
 
 The celebrated Morro Velho mine is situated on the west 
 ern side of the valley of the Rio das Velhas, not far from 
 Sabara. It was long worked by native miners, but on the 
 
 * For a detailed description of this formation see page 559, next chapter. 
 
THE GOLD-MINES OF BRAZIL. 537 
 
 failure of the mines at Sao Joao d El-Rei, it was purchased 
 by the company of that name, and has been worked with 
 remarkable success ever since. The gold is extracted from 
 a lode of quartz enclosed in clay-slate. The following ac 
 count of the vein and the mine, with the accompanying 
 statistics, I extract from the work of Mr. Arthur Phil 
 lips : * 
 
 " The formation affording the gold is a strong, well-defined lode, 
 though irregular in direction, dip, and dimensions ; its inclination 
 or underlie has also been found to vary at different depths and in 
 different parts of its extent ; the vein-rock is mostly composed of 
 quartz, with iron pyrites disseminated more or less regularly 
 throughout its mass, and the lode is not unfrequently traversed by 
 clay-slate and barren white quartz. When pyrites is absent in 
 these rocks gold is seldom present, f In some places the vein is 
 cavernous and less close in its texture than in others ; but where 
 drusy cavities are frequent the yield of gold diminishes ; the most 
 productive matrix for gold is a compact mixture of quartz and 
 pyrites, with varying quantities of slate. The great metalliferous 
 deposit called the Cachoeira, Bahu, and Quebra Panella, is one 
 continuous, very irregular vein varying in width from seven to 
 seventy feet, and at one point reaching one hundred feet.;}: The 
 
 * J. Arthur Phillips, The Mining and Metallurgy of Gold and Silver, p. 82, 
 which see for additional information concerning these mines. 
 
 t " Arsenical, magnetic, and ordinary iron pyrites predominate at different 
 points and in varying quantities : carbonate of lime, dolomite brown spar, and, 
 very rarely, copper pyrites are also present in the vein." 
 
 t Burton, Vol. II. p. 234, says : " The breadth of the lode varies from four 
 to sixty feet. The general direction when worked is west to east with north 
 erly shiftings. The clip is 45, rising to a maximum of 40 30 or 47. The 
 strike is from S. 82 E. - S. 58 E. The cleavage planes of the killas are in 
 some places transverse to, in others parallel with, the lode. In certain sections 
 of the mine walls they bear N. 36 E., but the average is more easterly. The 
 
 direction is S. 46 E., and their dip is at angles varying from 43 - 70 
 
 The underlay or underlie dip, or inclination of the mineral vein, is 6 in the 
 23* 
 
538 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 average thickness, 176 fathoms perpendicular on the Cachoeira 
 and 165 fathoms on the BaM, is nineteen feet. The stoping 
 space extends over 807 square fathoms, .... the enclosing rock 
 
 is a clay-slate of tolerably uniform texture The mineral 
 
 brought to the surface is first freed from slate and other unproduc 
 tive stone on the spalling-floors, and the ore, after being broken to a 
 
 uniform size, is stamped fine The stamping-mills, as is also 
 
 the pumping and other machinery, are moved by water-power. 
 The pulverized ore, issuing from stamp coffers through finely per 
 forated copper grates, passes over bullock-skins in the first in 
 stance, and, lower down the inclined tables, over woollen cloth. 
 The bullock-skins are taken up and washed in vats every hour, and 
 the woollen cloths at longer intervals. The subjoined table shows 
 the quantities of rock raised and stamped, the amounts of gold 
 produced, and annual net profits made since 1868."* 
 
 The following account of the Morro Velho mines I trans 
 late from Gerber.f 
 
 " This company, after having worked for some time the gold 
 mines in the vicinity of Sao Joao d El-Rei, acquired the auriferous 
 lands of Morro Velho, which since then they have explored with 
 great profit to its stockholders, owing this happy result not only 
 to the richness of the formations, but especially to the perfection 
 of the method of extracting the gold, and to the great skill with 
 which the establishment is regulated. The auriferous lodes of Morro 
 Velho are in general pyrites, contained in argillaceous schist, 
 and inclined to the southeast, about 45. The principal mines 
 explored are those of Cachoeira and Bahti. The first has in hori 
 zontal section a length of 1,120 palmos, a width which varies from 
 
 Bahii and 8 in the middle Cachoeira. Its dip varies from S. 82 E. - S. 58 E., 
 and the inclination from 42- 47, but everywhere parallel with the striae." 
 
 * For a minute account of the process of separating the gold at Morro 
 Velho, see Burton, Vol. I. Chap. XXVI. 
 
 Geographicas, &c., de Minas Geraes, 1863, p. 33. 
 
THE GOLD-MINES OF BRAZIL. 
 
 539 
 
 
 1849. 
 
 1850. 
 
 1851. 
 
 1852. 
 
 1853. 
 
 1854. 
 
 Stone raised, tons, .... 
 Stone and Ore stamped, tons, 
 Gold produced, pounds Troy, 
 
 69,336 
 69,004 
 2,583 
 
 67,106 
 64,313 
 2,517 
 
 79,810 
 81,629 
 3,057 
 
 82,642 
 81,236 
 3,323 
 
 85,698 
 86,866 
 3,623 
 
 86,048 
 86,433 
 3,464 
 
 Net profit, ... 
 
 38,136 
 
 35,880 
 
 51,586 
 
 55,391 
 
 49,273 
 
 44,740 
 
 
 1856 
 
 1856. 
 
 1857. 
 
 1858. 
 
 1859. 
 
 1860. 
 
 Stone raised, tons, .... 
 Stone and Ore stamped, tons, 
 Gold produced, pounds Troy, 
 
 87,297 
 86,848 
 3,325 
 
 89,877 
 87,424 
 2,992 
 
 86,407 
 86,335 
 2,539 
 
 88,901 
 87,270 
 2,733 
 
 88,968 
 82,880 
 3,294 
 
 91,361 
 74,528 
 3,974 
 
 Net profit . ... 
 
 34,460 
 
 32,233 
 
 787 
 
 8,545 
 
 38,058 
 
 60,460 
 
 
 
 
 
 
 
 
 
 1861. 
 
 1862. 
 
 1863. 
 
 1864. 
 
 1865. 
 
 1866. 
 
 Stone raised, tons, .... 
 Stone and Ore stamped, tons, 
 Gold produced, pounds Troy, 
 
 96,612 
 71,902 
 5,051 
 
 90,896 
 67,508 
 5,182 
 
 84,758 
 65,697 
 4,713 
 
 65,435 
 62,147 
 2,852 
 
 78,883 
 59,607 
 4,153 
 
 Ss 
 as 
 
 o . 
 
 -3 
 
 e o 
 o bo 
 
 >"H 
 
 Net profit . . . . . 
 
 96,769 
 
 87,531 
 
 63,285 
 
 
 80,438 
 
 
 
 S 
 
 
 
 
 
 14,629 
 
 
 % 
 
 
 
 
 
 
 
 
 13 to 85 palmos, and it had in March, 1861, a perpendicular depth 
 of 1,190 palmos, and in February, 1862, of 1,480 palmos. The 
 mine of Bahu averages the same dimensions. Both possess six 
 inclined planes for the transport of the ore. In the year 1859 
 there were occupied in the service of the mine 274 free natives 
 and strangers and 407 slaves ; in all, 681 men ; among them 242 
 broqueiros [blasters], who extracted in 311 working-days 89,000 
 tons of ore, of which 6,119 were rejected as poor, and the remain 
 ing 82,881 tons were stamped in 6 mills with 134 hands. 
 With the breaking of the stone, in the service of the mills 
 and the amalgamating establishment, were occupied in this year 
 9 Europeans, 21 native men, and 24 native women, 79 men-slaves, 
 
540 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 and 254 women-slaves; in all, 387 persons. The mean yield per 
 ton was 3.9 oitavas.* The amalgamation has been made at the rate 
 of 70 pounds of mercury to the ton of stamped ore, and the loss 
 of mercury was 0.58 ounces for each cubic foot amalgamated, 
 which makes about 6%. The total product of the establishment 
 was in the year referred to 342,885 oitavas of gold, with whose 
 extraction was incurred an expense of 115:808 $067 Reis [about 
 $58.000 more or less], i. e. 357 Reis [less than 20 cents] per 
 oitava. The state of health of the establishment is flattering ; 
 the mortality in the whole of its population amounted during 
 1859 to only 2.76%, and, excluding accidental deaths, only to 2.14%, 
 a circumstance very noteworthy for a population composed for 
 the most part of miners, which proves the solicitude the super- 
 intendency feels for the welfare of its employees. The capital of 
 the company subscribed since 1830 is 1 28,400 sterling, which 
 has during thirty years, up to 1860, produced a net income of 
 466,874 6s. 1 d., the possessions of the establishment being worth, 
 beside this, 100,000. The last dividend of the company was 
 2 for share of 15." 
 
 The Morro Velho mine is an example of successful vein- 
 mining to a great depth. There is no appearance of a dimi 
 nution in richness of the ore in descending. 
 
 Phillips states, on the authority of Mr. Hockin, the 
 managing director of the mines, that 
 
 "The rock treated at the Morro Velho mine is principally a 
 mixture of magnetic, arsenical, and common iron pyrites, finely 
 disseminated in, and intimately mixed with, a quartzose gangue. 
 The composition of what is called pure ore may be taken at about 
 43 per cent of silica, and 57 per cent of pyritous matter. Of 
 these minerals, arsenical pyrites is usually the most auriferous, 
 though it does not occur in large quantities. Pure specimens of 
 
 * An oitava is 7.343 grains Troy. 
 
THE GOLD-MINES OF BRAZIL. 541 
 
 this substance afford by assay from four to six ounces of gold per 
 ton, and wherever crystals of this mineral make their appearance 
 the yield of the precious metal is large. Cubical pyrites is of more 
 frequent occurrence, but is far less rich in gold ; solid specimens 
 of this substance, but slightly mixed with quartz, yield about an 
 ounce and a half of gold per ton by assay. 
 
 " Magnetic pyrites constitutes the largest proportion of the sul 
 phides found, but this is very slightly auriferous since pure speci 
 mens generally yield rather less than four pennyweights per ton. 
 Branches of clay-slate are often found in the principal veins, and 
 this rock, by assay, affords from five to seven and a half penny 
 weights of gold per torj. Quartz without any admixture of sul 
 phides has never been found to be auriferous, and it is a remark 
 able fact that the smallest speck of gold is rarely seen previous to 
 concentration in any of the ores of the mine."* 
 
 The gold-mines of Gongo Soco lie about twenty miles 
 east of Morro Velho, on the opposite side of the valley of 
 the Rio das Velhas. They were once very productive, and 
 became famous ; Weddell says that the old miners once 
 took out one hundred pounds in three hours ! but through 
 bad management the company that worked them failed. 
 The income of the company during the thirty or more 
 years of its existence was 1,388,416, of which 375,163 
 was profit. f 
 
 Gardner says, on the authority of Helmreichen, that at 
 the Gongo Soco mines there is the following succession of 
 rocks : a bed of itacolumite, underneath which is a bed of 
 auriferous jacutinga fifty fathoms thick ; $ then a thick layer 
 
 * For the details of the method of treatment of these ores at Morro Velho 
 see Phillips, op. cit., to which there is a frontispiece representing the Morro 
 Velho establishment. 
 
 f Lieutenant Moraes quoted by Burton. 
 
 J This section does not agree very well with Castelnau s description of the 
 
542 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 of ferruginous itacolumite, with a dip of 45, lying on clay- 
 slate, containing great masses of ironstone. Underlying the 
 slates is granite. Castelnau says that these beds dip to the 
 south. The gold occurred free in the jacutinga, and was 
 separated by washing. M. Weddell* describes the Gongo 
 Soco jacutinga as black and friable as coal ; it was said to 
 be very soft, so as to admit of being worked with a pick ; 
 blasting was unnecessary. When the gold was not visible, 
 the ore was stamped and washed without using mercury. 
 When the precious metal was visible it was treated first in 
 a mortar, and then washed in a bateia or wooden washing- 
 bowl. The Gongo Soco gold is said to contain palla 
 dium,! and is deep yellow in color. Burton says that he 
 has seen specimens " of a bright brassy tint, and sometimes 
 dingy red, like worked unpolished copper." 
 
 The Rossa Grande Gold-mining Company owns a tract 
 of land twenty-one square miles in area, not far from the 
 mines of Morro Velho. The gold occurs in a mixture 
 of quartz sometimes associated with iron-ore, at others 
 with arsenical pyrites, or ferric oxide. Some of these 
 ores are said to be very rich, producing even as much as 
 fifty oitavas to the ton. Besides the vein-rock, gold oc 
 curs in jacutinga, and in alluvial washings. Burton visited 
 the mine, and reports it as looking very much like a 
 failure. 
 
 The Morro de Santa Anna, where the Dom Pedro North 
 
 mines : Castelnau says that the jacutinga is ordinarily only sixteen centimetres 
 in thickness. 
 
 * Castelnau, Hist, du Voyage, Tom. I. p 243 (bis). 
 
 t The gold of Brazil is always alloyed with silver, and occasionally with 
 platina ; sometimes it contains a considerable percentage of iron, when it is 
 very dark in color. Iridium and irid-osmiuin occur in the gold-washings of 
 Minas. 
 
THE GOLD-MINES OF BRAZIL. 543 
 
 d El-Rci Company was established, is a mountain about four 
 thousand feet in height above the sea, and some two thou 
 sand feet above the neighboring valley. 
 
 I give Phillips s description of the geological structure of 
 this mountain, but I must confess that it is somewhat 
 enigmatical : * 
 
 " The face of the mountain is covered with canga, an iron con 
 glomerate, about four feet thick; this is auriferous, and will 
 probably pay for stamping. Beneath the canga is the first jacu- 
 tinga formation, about sixty feet in thickness, containing veins rich 
 in the precious metal ; the jacutinga partakes more of the charac 
 ter of mica-slate than of iron-sand, and the auriferous veins in it 
 are more like quartz than ironstone. This rests on a stratum of 
 hard ironstone about three feet in thickness, which is the second 
 jacutinga formation, but quartz is the predominating constituent, 
 and rock is, according to Captain Treloar, a more correct name for 
 it than jacutinga. This lode averages about four feet wide ; it 
 opens and contracts, and where it expands it is generally found 
 most productive. Subjacent to the layer is a layer of hard clay 
 and mica-slate, of about five feet in thickness, and then comes the 
 rock formation which has yielded the chief returns of gold. In 
 the present workings it is about ten feet wide ; but in its longi 
 tudinal course it so expands and contracts as to become in some 
 
 places extinct The general direction of the lodes at Morro 
 
 de Santa Anna may be said to be easterly and westerly, and their 
 underlie northerly ; but both vary, owing to the lodes hugging the 
 contour of the mountain." 
 
 Near Santa Anna is another mine called Maquine , 
 worked by the company owning the Morro de Santa Anna, 
 which has been abandoned. Out of all the gold-mines of 
 Brazil only two have paid, Morro Vclho and Maquind. 
 
 * Phillips, op. cit., p. 85. 
 
544 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 The rest have failed, some of them after a more or less 
 prosperous career, and notoriously in most cases from bad 
 management and an imprudent outlay of funds. 
 
 At Congonhas do Campo, Burton describes the gold as 
 occurring " in the pores and cavities of friable or rotten 
 quartz injected into greenstone," * and states that " Mr. 
 Luccock detected dust-gold among schist-clays and the 
 other component parts of the ground, and the latter con 
 tained the ore with equal certainty, and nearly equal quan 
 tity, whether of the prevailing red hue, or any of the shades 
 of brown or yellow. 
 
 The gold-mine of Sao Vicente belongs to the East d El- 
 Rei Company, which is, we understand, about to be recon 
 structed. The gold is found in a quartz vein that strikes 
 east- west with a dip of 28, and whose rock character we have 
 already described. Burton says that " failure is its actual 
 state," but that " the little lode may pay if worked safely, 
 that is to say, scientifically and economically." A specimen 
 of the Sao Vicente gold was examined for me by Mr. Clarke 
 of the Cornell Chemical Laboratory for palladium. There 
 was not even a trace. 
 
 The Cata Branca mines belong to the Morro Velho Com 
 pany, and are situated two miles east of the village of 
 Corrego Secco. The following notes on the mines are from 
 Burton f : 
 
 " The serra of Cata Branca trends where mined from east of 
 north to west of south. The containing rock proved to be mica 
 ceous granular quartz with visible gold, as in California. The strike 
 was N. 15 W., and the dip 80 to 85 ; in some places the stratifi 
 cation was nearly vertical, in others it was bent to the slope of the 
 
 * Captain Burton, Highlands of the Brazil, Vol. I. p. 174. 
 t Vol. I. p. 182. 
 
THE GOLD-MINES OF BRAZIL. 545 
 
 mountain, and generally it was irregular. The lode, narrow at the 
 surface, widened below from six to eighteen feet, and the greatest 
 depth attained was thirty-two fathoms." 
 
 He describes the vein-rock as varying granular to compact, 
 and states that canga and jacutinga occur here. 
 
 " The lode, which could not be called a constant productive, 
 abounds in vuglis, or vein cavities, tubes, pipes, and branches, called 
 by the Brazilian miner Olhos, eyes, surrounded by a soft ma 
 terial, mainly running vertically, and richer in free gold than the 
 average. Near these pockets, but not disseminated through the 
 vein, was a small quantity of auriferous pyrites, iron and arsenical. 
 A little fine yellow dust, oxide of bismuth, ran down the middle of 
 the lode and gave granular gold. The best specimen averaged 
 from 21.75 to 22 carats, our standard gold. 
 
 " The Santa Antonio lode lay parallel with and east of the Cata 
 Branca. The Aredes mine, eight miles to the southwest, was be 
 yond the peak ; here the serra is covered with boulders of hard 
 quartz, very numerous at the base of the great vein. They rest 
 on the common, soft, variously colored clays of the country, and 
 are intersected with lines of sugary quartz, which gave a very little 
 very fine gold. This formation extends far to south and west of 
 Itabira ; openings were made in it, and one, the Sumidouro, was 
 successful. Aredes showed also a small formation of jacotinga 
 containing red gold, sometimes alloyed with palladium and accom 
 panied with oxide of manganese." 
 
 In 1843 M. d Osery, geologist of the expedition of Count 
 Castelnau, visited and examined the Cata Branca mine. He 
 reported* that at the locality the rocks consist of itacolumite 
 and clay slates alternating, and in strata almost perpendic 
 ular. He described the auriferous vein as running nearly 
 
 * Castelnau, Expedition dans I Ame rique de Sud, Hist, du Voyage, Tome I. 
 p. 244. By a blunder in pagination, pp. 241 - 256 are repeated in this volume. 
 
 1 1 
 
546 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 north-south and traversed by fissures or faults in which 
 the gold occurred. It was also found in the fissures for a 
 distance of two or three palms on each side of the line of 
 the faults, together with bismuth. Sometimes gold was 
 found where the vein quartz came in contact with the en 
 closing rock ; the interior of the vein was very barren. 
 D Osery thought that originally the vein was composed of 
 pure quartz, which was afterwards disturbed and the gold 
 and bismuth introduced by sublimation. 
 
 The mine proved a failure because of bad working and a 
 want of economy. 
 
 Gold-washings occur, as already described, in almost every 
 province in the empire. 
 
 In Maranhao is a mine belonging to a company called the 
 Montes Aureos Gold-Mining Company (limited). I know 
 nothing of it further than that it is said so far to have 
 been worked with but little success. 
 
 The generally received opinion that the gold-mines of 
 Brazil are exhausted is a very great mistake. There are 
 still surface deposits of great extent which, with modern 
 appliances, could be successfully worked. The underground 
 wealth of the country is almost untouched,* and if the 
 mining public of America knew Brazil better, I am per 
 suaded that the gold-fields of that country would not be 
 neglected by American capitalists. 
 
 * In this belief I am supported by Burton, and Liais, in treating of the head 
 of the basin of the Sao Francisco, says : " Quant aux filons pyriteux qui abon- 
 dent dans les re gions montagncuses circonscrivant le bassin du San Francisco, 
 et oil ses divers affluents prennent leur source, ils sont c te a peine attaques. 
 C est la cependant quo reside la grande richesse aurif ere de la province de Minas- 
 Geraes. Carc est de la surface de composee de ces filons pyriteux qu etait pro- 
 venu 1 or qui fut jadis retire des depots meubles " L? Sun-Francisco ait Bresil, 
 par M. Liais, Bull, de la Socieltf Geoyrajihif., 5 rae Serie, E. 2, p. 309. 
 
RESUME OF THE GEOLOGY OF BRAZIL. 547 
 
 CHAPTER XIX. 
 
 RESUME OF THE GEOLOGY OF BRAZIL. 
 
 Eozoic Rocks and their Distribution in Brazil. Absence of Limestones. 
 The Silurian Age in Brazil. The Auriferous Clay-Slates of Minas proba 
 bly Lower Silurian. Note on the Silurian of the Andes. The Distribu 
 tion of Marine Animals in the Palaeozoic. The Devonian Age in Brazil 
 and South America. The Carboniferous of Brazil and Bolivia. The New 
 Bed Sandstone. The Cretaceous, its Distribution in Brazil and South 
 America. Several distinct Periods represented. Tertiary Rocks. Drift. 
 The Glacial Phenomena of Patagonia. Tapanhoacanga. The Drift 
 of Rio and of the Region of Decomposition. The Drift of the Dry 
 Region of Bahia, Sergipe, and Alagoas. Examination into the Merits of 
 the various Theories proposed to account for the Formation of the Brazilian 
 Drift. The Theory of Subaerial Decomposition. Wave Action during a 
 Subsidence. Wave Action during Elevation. All these Theories unsat 
 isfactory; The Glacial Hypothesis the most reasonable. 
 
 Eozoic. The gneiss of the province of Rio de Janeiro is 
 an orthoclase variety, varying from schistose to coarse 
 grained and porphyritic, or homogeneous and granitic ; so 
 far as I have been able to observe, it is everywhere stratified, 
 and consists of metamorphic sedimentary deposits. Though 
 much of the rock would be described as granite if seen in 
 the hand specimen or in a single quarry, I have never failed 
 to find the large masses stratified, so that in this work I 
 have included all the varieties under the general head of 
 gneiss. These rocks in the province of Rio are of great 
 thickness, and the Serra do Mar and the Serra da Manti- 
 qucira are wholly composed of them. According to Pissis 
 they are divided into two groups, an upper and lower, sub 
 divided as follows : 
 
548 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Gneiss, with numerous heavy beds of quartzite, and an 
 Upper. { abundance of mineral veins which are not found in 
 the lower division. 
 
 1. Fine-grained gneiss, without garnets and with some 
 
 subordinate beds of quartzite in coarse grains with a 
 
 little mica. 
 Lower. -{ 
 
 2. Gneiss fine grained, very rich in mica, and with an 
 
 abundance of garnets. 
 
 3. Porphyritic gneiss. 
 
 On crossing the Mucury district I found the gneiss, which 
 was at first coarse and porphyritic, becoming finer on going 
 westward, and finally giving way to heavy beds of mica^ 
 slate or mica-schistose gneiss with bands of quartz. The 
 same succession seems to obtain elsewhere in Brazil, the 
 gneisses proper being overlaid by mica-slates, the older 
 rocks along the coast generally lying to the eastward. This 
 corresponds very well with what D Orbigny says of the suc 
 cession of similar strata in Bolivia and the Andes, where the 
 gneisses are immediately overlaid by mica-slates. Elie de 
 Beaumont and other geologists of note have long since sig^ 
 nalized the gneisses of the Serra do Mar as among the very 
 oldest stratified rocks of the globe. The system of upheaval 
 of the gneiss of the Serra do Mar D Orbigny calls the Bra 
 zilian ; and Elie de Beaumont, in his report on the " Conside 
 rations g-enerales sur la G-eologie de I* Amerique meridio 
 nals" by M. D Orbigny,* says that this system is one of the 
 oldest known, and that perhaps it preceded the soulcvement 
 of the most ancient system of mountains hitherto described 
 in Europe. It is certainly the oldest of the rock formations 
 of the Brazilian plateau. When we come to compare the 
 Brazilian gneisses with the Laurentian rocks of Canada and 
 
 * Comptes Rendns, 28 May, 1843. 
 
RSUM OF THE GEOLOGY OF BRAZIL. 549 
 
 Europe we find such strong resemblance in lithological 
 character, and in the system of the upheaval, I can see no 
 reason why we should not refer them to the eozoic. The 
 axis of upheaval is the same as that of the Laurentides. 
 
 In North America heavy beds of limestone are interstrat- 
 ified with the gneiss of the Laurentian. In the Serra do 
 Mar beds of limestone are very rare, and the thin bed I ex 
 amined at Pirahy is the only one I have seen in the Serra 
 do Mar. This contained only faint streaks of serpentine. 
 Limestones appear to occur interbedded with the gneiss at 
 Cantagallo. Some of the limestones of the interior of 
 Bahia may belong to the same series. 
 
 The absence of clay-slates among the gneisses in Brazil 
 recalls also the Laurentian of North America. There can be 
 little doubt that the great mass of the gneisses of the coast 
 provinces north and south of Rio are eozoic, but these rocks 
 have in the northern provinces been so slightly examined 
 that it is impossible to describe them with any detail, and 
 some of the mica-schists associated with them may be Lower 
 Silurian or Cambrian. Along the coast of the Province of 
 Bahia there are dioritic gneisses in the series, and on the 
 Sao Francisco and elsewhere we find syenites. The study 
 of these old rocks in the southern provinces is attended with 
 immense difficulty, owing to the forests, the decomposition 
 of the surface, and the thickness of the drift. But in the 
 drier northern provinces, where the rocks are more exposed 
 and less affected by decomposition, they may be well exam 
 ined. In the preceding chapters I have shown that gneiss 
 is found in every province of the empire. Not only does it 
 form the great coast belt extending from Maranhao to the 
 mouth of the Rio de la Plata, but it sends off a band from 
 southern Minas Geraes into Goyaz, and the Monies Pyre- 
 
550 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ncos and a considerable part of the mountainous region of 
 Central Goyaz are composed of it. The same rock shows 
 itself in the cataracts of the Tocantins, the Xingu, the Tapa- 
 jos, the Arinos, and the Madeira, showing that the table 
 land of Brazil is everywhere underlaid by it. 
 
 In the present state of our knowledge of the stratigraphy 
 of these rocks it is quite impossible to do more than guess 
 at their thickness, for as in Canada and elsewhere there are 
 numerous reversed folds, and one may travel for miles 
 over the surface of the Brazilian gneisses, finding them 
 always highly inclined, and all dipping in the same way. 
 The Serra do Mar, where crossed by the Dom Pedro II. Rail 
 road, is a monoclinal ridge, but it must be composed of 
 several reversed folds, else the thickness would be enor 
 mous. 
 
 The highlands of Venezuela and Guiana are largely 
 composed of gneiss similar to that of Brazil, and disturbed 
 by the same system of upheaval as has been remarked by 
 Humboldt, D Orbigny, Agassiz, and others, and this gneiss 
 area, bounding the Amazonian valley on the north, was 
 doubtless an island at the opening of the palaeozoic time. 
 The highlands of Brazil formed another island, while the 
 Chiquitos gneiss region to the southwestward was probably 
 another. 
 
 Since the foregoing was written and sent to the printer, 
 I have been honored by a visit from Dr. T. Sterry Hunt, 
 who has examined with care the large suite of metamorphic 
 rocks I brought home from Brazil. Dr. Hunt has kindly 
 furnished me with the following note for publication : 
 
 "The gneissic rocks . of Rio de Janeiro and the Serra do Mar 
 present the characteristic types of the Laurentian of North Amer 
 ica, including as they do coarse granitoid and porphyritic varieties 
 
RESUME OF THE GEOLOGY OF BRAZIL. 551 
 
 with red orthoclase and fine-grained gray and white banded gneisses, 
 often hornblendic. The white crystalline limestone with pale green 
 serpentine which occurs with these Brazilian gneisses is not distin 
 guishable from that of the North American Laurentian. The 
 fine-grained, tender micaceous and hornblendic schists, which in 
 Brazil succeed the gneisses, are very like the similar rocks which 
 in some parts of New England and Acadia appear to follow the 
 Laurentian, and are associated with staurotide, cyanite, and chias- 
 tolite slates ; while the auriferous argillites and quartzites which 
 follow these schists in Brazil strikingly resemble those which in 
 Nova Scotia occupy a similar stratigraphical position. This triple 
 parallelism in lithe-logical and mineralogical character in the rocks 
 of regions so widely separated is in itself a strong argument in 
 favor of their geological parallelism." 
 
 Silurian. Notwithstanding all that has been published 
 by the various geologists who have studied the gold region 
 of Minas Geraes, the exact succession of the different mem 
 bers of the metamorphic series lying just inside of the gneiss 
 belt has never been satisfactorily worked out. The clay 
 and talcose schists, the itacolumite,* itabirite, and other as 
 sociated metamorphic rocks of this region appear to be lower 
 palasozoic in age. I have called attention to the striking 
 resemblance borne by the clay-slates and associated quartz 
 ites to the gold-bearing rocks of Nova Scotia, and I have 
 suggested that they may be the equivalents of the Quebec 
 group of North America. The gold-bearing rocks in Minas 
 Geraes resemble the similar auriferous series of the south 
 ern Atlantic States in which itacolumite occurs. 
 
 * On page 149 I have spoken of the occurrence on the Rio Gavata of a schis 
 tose quartz rock resembling a sandstone. I had not a good opportunity of ex 
 amining the locality, and I doubted whether it was a metamorphic rock. A 
 further examination of a specimen of the rock in company with Dr. T. S. 
 Hunt has proved it to be a true itacolumite. 
 
552 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Clay-slates with auriferous veins occur in other parts of 
 Brazil besides Minas, as, for instance, in Goyaz, and in the 
 vicinity of Cuiaba in Matto Grosso. 
 
 These rocks are everywhere so metamorphosed, that all 
 trace of fossils has been completely obliterated.* 
 
 * The Silurian rocks of the Andes of Bolivia and Peru have been examined 
 by the English geologist, Mr. Forebs (Quart. Jour. Geol. Soc. Vol. XVII. p. 53), 
 who thus describes their distribution : " The rocks which I have grouped to 
 gether as pertaining to the Silurian epoch show themselves continuously, or very 
 nearly so, over an area from northwest to southeast of more than seven hundred 
 miles ; and the area occupied by them cannot be estimated at less than 80,000 to 
 100,000 square miles. They form the mountain-chain of the high Andes, rising 
 to an absolute height of 25,000 feet above the seu, and, in the part of South 
 America more particularly the subject of this memoir, continuous through Peru 
 from the north of Cusco over the snowy ranges of Carabaya and Apollobamba, 
 across the provinces of Munecas, Larecaja, La Paz, Yungas, Sica-Sica, Inqui- 
 sivi, Ayopaya, Cochabamba, Cliza, Misquc, Chayanta, Yamparez, Porco, To- 
 mini, and Cinti throwing off spurs along the eastern side of the main chain, 
 right through the province of Caupolican down to the river Beni in Mojos, into 
 Yuracorcs, Valle Grande, Santa Cruz and Chuquisaca, and to the east into the 
 provinces of Oruro, Potosi, and Chichas." The rocks consist of clay-slate, 
 shales, and graywackes, and, according to Mr. Forbes, they probably repre 
 sent the whole Silurian from top to bottom. D Orbigny had already de 
 scribed ten species of Silurian fossils from the Central Andes ; to this list Mr. 
 Forbes has added nineteen new species, described by Mr. Salter in his paper 
 following that of Mr. Forbes. The genera represented in the Andean Silurian 
 are Cruzeana, Lingula, Orthis, Graptolithus, Phacops, Asaphus, Boliviana, Pa 
 tella, Bellerophon, Area?, Ctenodonta, Cucullella, Strophomena, Tentaculites, 
 Beyrichia, Homalonotus. Two species of Phacops described by D Orbigny are 
 doubtful, being probably Devonian. If we subtract these we have left only twen 
 ty-seven species of fossils known for the Andean Silurian. Jn the course of his 
 paper Mr. Salter makes an interesting remark that speck s in the Silurian had 
 a very limited range, those of India, Australia, and Europe being entirely dif 
 ferent. On the other hand, the species of the Devonian, especially of the upper 
 part, had a very wide range, while the carboniferous types are almost cos 
 mopolitan. I must confess that, after a careful study of the carboniferous 
 Brachiopoda of Nova Scotia, I am hardly prepared to go quite so far as to 
 admit that the Producti from Bolivia, Nova Scotia, Ireland, and Belgium, 
 thrown together under the name of Cora, are all the same species, and I may 
 
RSUM OF THE GEOLOGY OF BRAZIL. 553 
 
 This series offers a large number of crystallized minerals, 
 among which are the topaz, enclaze, <fcc. The topazes of 
 Minas appear to be found in the cascalho formed from the 
 debris of these rocks. 
 
 Devonian. Some of the metamorphic rocks of Minas 
 Geracs or Bahia may be devonian, but I have seen no rocks 
 referable to that age on the coast, unless it be that the slate 
 conglomerates, sandstone, and shale, with fossil plants found 
 on the Rio Pardo, may belong to it.* 
 
 Carboniferous. There can be no uncertainty about the 
 existence of true carboniferous strata in Brazil, for besides 
 the coal we have an abundance of fossil plants of carbonif 
 erous genera. The coal-basins lie just south of the tropics, 
 but within the range of the palm, and they are a coast for 
 mation, corresponding in this respect to the coal-basins of 
 Acadia, Massachusetts, and Rhode Island. I know of no 
 carboniferous strata north of Rio on the coast. It would 
 seem as though the depression of the coast which allowed 
 the accumulation of the coal-beds of the southern provinces 
 had not extended to the north. The very slight disturb 
 ance of the coal-beds is noteworthy, as is also their bitumi 
 nous character.! 
 
 say the same of the other Nova Scotian Brachiopoda referred by the distin 
 guished Mr. Davidson to European forms. But whether the species were or 
 were not absolutely cosmopolitan during the carboniferous, the resemblance 
 of the marine animals was much greater during the carboniferous than before 
 that time. 
 
 * Messrs. D Orbigny, Salter, and Forbes refer to the devonian certain fos- 
 siliferous rocks of the eastern plateau of Bolivia. See Forbes, op. cit., p. 51 ; 
 and Salter, op. cit., p. 63. The Falkland Islands, described by Darwin, are 
 composed of rocks probably belonging to the Lower devonian. (Quarterly 
 Journal Geological Society, London, Vol. II. p. 267, 1846.) 
 
 t Carboniferous rocks are found in the Rio Guapore , one of the branches of 
 the Madeira. The carboniferous rocks of Bolivia have been studied by D Or- 
 VOL. I. 24 
 
554 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Triassic.* I have referred to the triassic a thick series 
 of red sandstones, lithologically identical with the Connecti 
 cut River and New Jersey new red sandstone, apparently 
 barren of fossil remains, and which occupy a large area in 
 the Province of Sergipe, underlying the cretaceous. These 
 rocks are more or less inclined. I know of no trap asso 
 ciated with them. 
 
 Jurassic.^ I have seen no rocks on the Brazilian coast 
 referable to this age. I can explain their absence only by 
 
 bigny and Forbes. I extract a short account of these last by D Archiac, Ge o- 
 logie et Pal&ntologie, p. 499 : 
 
 " Les roches carboniferes de la Bolivie, situees a 1 ouest des Andes, se ren- 
 contrent, par places, comme de petits bassins allonge s gene ralemcnt du S. O. ati 
 N. E. situes au milieu de la grande plaine quaternaire qui entoure le lac de 
 Titicaca, se montrant aussi au nord du lac, et plus au sud dans les provinces 
 d Arque et d Oruro. Le point le plus has oil on les observe est a 3,800 metres 
 d altitude et on pent les suivre jusqn a 4,000 et 4,500 metres. A 1 ouest du lac, 
 entre Tiquina et la Guardia, M. Forbes donne une ide e complete de la serie des 
 assises disposees en bnssin renverse et pre sentant successivement des plus an- 
 ciennes aux plus re centes qui occupent le milieu du plissement, des gres blancs, 
 des conglomcrats et des gres rouges, des argiles blanches panache es, des cal- 
 caires en banes epais, bleus et jaunes, des argiles schisteuses panachees, un cal- 
 caire bleu puissant, enfin des gres jaunes et blancs." And in a note he says : 
 " Les fossiles de ces assises, etudie s par M. Salter, sont ; Productus Semireticu- 
 latus (P. Inca d Orb.) P. longispinus (Capacii d Orb.) Spirifr.r Condor, S. Bolivi- 
 ensis, Athi/ris subtil ita, Orthis resupinata, 0. Andii, Ttfiynchonella nov. sp. Euom- 
 phaltts (Phanerotinus ?) , Bellerophon voisin de B. Urn, des polypiers et des 
 crinoides inde termines. Des provinces d Arque et d Oruro ont e te obtenus 
 les Spirifcr Condor et lineatus, les Productus Cora, semireticukitiis, Bolivientit et 
 I Orthis Andii." 
 
 * Mr. M. D. Forbes refers to the Triassic or Permian a series of red and 
 yellow sandstones, saliferous and gypseous marls, clays, gypsums, cupriferous 
 sandstones, and red conglomerate found in the Andes, but which contain no 
 determinable organic remains. (Quart. Jour. Geol. Soc., Vol. XVII. p. 36.) 
 
 t Jurassic rocks containing Ammonites, Terebratulae, Spirifers, and other 
 fossil characteristics of that epoch have been found in the Andes by MM. Cros- 
 nier, Maycn, D Orbigny, Darwin, Domeyko, Forbes, &c. They extend from 
 Chili to Peru. 
 
RESUME OF THE GEOLOGY OF BRAZIL. 555 
 
 supposing that during the Jurassic the coast stood higher 
 than at present. In this respect the Brazilian coast would 
 resemble that of Eastern North America. 
 
 Cretaceous. The cretaceous rocks of Brazil are un 
 known on the coast south of the Abrolhos, which islands I 
 believe to be outliers of this formation. Properly speaking 
 the cretaceous deposits begin a few miles south of the Bay 
 of Bahia, and occur at intervals along the coast northward, 
 occupying, at least in several instances, separate basins, some 
 of which are fresh-water. We find cretaceous rocks in 
 Bahia, Sergipe, Alagoas, Pernambuco, Parahyba do Norte, 
 Ceara, and Piauhy. It is difficult to estimate their exact 
 extent, because they are largely covered up by tertiary 
 beds. It is very probable that marine cretaceous beds un 
 derlie the tertiary deposits throughout the whole valley of 
 the Amazonas, but the only place where they show them 
 selves, so far as I can ascertain, is on the Aquiry, an affluent 
 of the Rio Purus, where they have been examined by M. 
 Chandler, as is stated by Professor Agassiz.* I am not 
 aware that they are exposed anywhere to the eastward on 
 either side of the valley. 
 
 Among the cretaceous rocks of Brazil several periods are 
 represented. 
 
 The fossil mollusks of the fresh-water beds of the Bahia 
 Basin have a very strong wealden look, but they are asso 
 ciated with teleostian fishes and other remains, which are 
 certainly cretaceous. They evidently belong low down in 
 the series, and they may represent the Neocomien. 
 
 The compact limestones at Maroim affording Ammonites, 
 Ceratites, Natica, &c., are probably middle cretaceous. Over 
 these are the flaggy white and grayish limestones with Ino- 
 
 * Journey in Brazil. 
 
556 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 ceramus, Ammonites, fish, &c., apparently representing the 
 white chalk, SSnonien. For the fresh-water beds at Bahia 
 I would propose the name Bahian group ; for the Maroim 
 limestones, that of Sergipian group ; for the flaggy limestone 
 beds near Aracaju, the Cotinguiban group, and for the 
 Aquiry beds the Amazonian group. The cretaceous of Bra 
 zil would then be divided as follows : 
 
 Amazonian group (Aquiry) with Mosasaurus, Mcestrichtien ? 
 Cotinguiban group with Inoceramus, Ammonites, &c., Senonien ? 
 Sergipian group with Ammonites and Ceratites, middle cretaceous ? 
 Bahian aroup, Crocodilus, Pisodus, species of Melania and other 
 fresh-water shells, cyprids, &c., Neocomien ? 
 
 The sandstones, shales, and limestones of the Abrolhos 
 and the lower Sao Francisco I believe to be cretaceous, 
 but I have no fossils by which to determine their exact 
 age. They may correspond in part to the Sergipian and 
 Cotinguiban groups. Those of the Sao Francisco are mud 
 and sand deposits instead of limestone, which accumu 
 lated along the coast elsewhere in clearer water. It will 
 be remembered that I have described the limestones near 
 Proprid as sandy and even pebbly. 
 
 The cretaceous rocks nowhere form very high hills. 
 They appear to have been deposited in a shallow sea, which 
 was not deep enough to penetrate into the Sao Francis 
 can valley above the falls. The sandstones above the 
 falls, described by Burton as cretaceous, I am persuaded 
 will be found to be tertiary. The cretaceous rocks have 
 suffered slight disturbance, and at the Abrolhos it is 
 worthy of note that they are associated with volcanic de 
 posits. 
 
 At the time of the deposition of the cretaceous, the north- 
 
RESUME OF THE GEOLOGY OF BRAZIL. 557 
 
 ern part of South America was depressed more than at pres 
 ent, while the coast of the southern provinces of Brazil 
 seems to have been higher than now. 
 
 In speaking of the cretaceous of South America, M. 
 D Archiac, Geologie et Paleontologie, p. 624, says : 
 
 "Nous avons fait voir que, d apres les recherches de M. H. 
 Karsten, confirmees depuis par celles de M. Wall, on pouvait 
 presumer que Fetage inferieur de la craie tuffeau, le gault et une 
 partie du groupe neocomien etaient representes dans le Venezuela, 
 particuliei ement dans les cordilleres de Merida et Truxillo. Tous 
 les calcaires cretaces des chaines de ce pays, comme ceux du meme 
 age, que Ton suit jusqu au Chili, sont d ailleurs entierement noirs, 
 bitumineux et semblables & ceux des grandes montagnes de 
 TEurope." 
 
 Tertiary. The clays and ferruginous sandstone forming 
 the coast plains outside the cordilheira are undisturbed, 
 and overlie the cretaceous unconformably. They are over 
 laid by the drift-clays, which descend from the cordilheira 
 and cover their glaciated surfaces, so that, though I have 
 nowhere found fossils in them, I have felt justified in 
 referring them to the tertiary. The horizontal beds of 
 clays, sandstone, &c., of the Jequitinhonha and Sao Fran 
 cisco valley are everywhere undisturbed, even where they 
 closely approach the coast where the cretaceous rocks have 
 suffered upheaval. They resemble the coast beds, except 
 that they are thicker, stand at a very much higher level, 
 and in some cases form beds of pure sandstone and con 
 glomerate with limestone and iron ore. They, too, are cov 
 ered by the drift-clays. I suppose that they are also ter 
 tiary, but older than the coast clays. To the same group 
 evidently belong the horizontal deposits of the plateau of 
 Sao Paulo, similar strata occupying the upper part of the 
 
GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 valley of the Parahyba do Sul, and the clays and sand 
 stones of the elevated plains of the north. These beds 
 must have been deposited when the continent stood at 
 a level full 3,000 feet lower than at present. The mate 
 rial was evidently derived from the wearing away of the 
 decomposed gneissose rocks, and it appears to have been 
 deposited rapidly in a muddy sea, not favorable for the 
 existence of life. After these beds were deposited the 
 coast rose very uniformly, and they suffered very exten 
 sive denudation. Along the coast outside the cordilheira 
 there were deposited, probably in a large part made up of 
 the results of the older beds, the coast sandstones and clays. 
 According to my own observation the upper level of the 
 coast clays south of Bahia is always much below the level 
 of the lowest beds of the older beds. I have never seen 
 them tie in with one another, but I strongly suspect that in 
 the vicinity of Monte Pascoal an outlier of the older beds 
 lies surrounded by the newer. 
 
 The stratified and loose sands and clays of the Taboleiros 
 at Alagoinhas appear to be older than the drift. They are 
 certainly newer than the Coast Tertiary group. They need 
 much more study, and I must confess that there are some 
 puzzling points in connection with them.* 
 
 Drift. In South America from Tierra del Fuego north 
 ward, to at least 41 S., glacial phenomena have been 
 observed and reported by Darwin and others, and these 
 phenomena appear to be identical with those so well studied 
 in the northern hemisphere. Drift occurs in the Falkland 
 Islands (Darwin), Australia, and New Zealand. The An 
 tarctic Continent is buried in ice and snow. No doubt can 
 exist that a drift period prevailed over the southern part of 
 
 * Tertiary rocks arc found over large areas both north and south of Brazil. 
 
RESUME OF THE GEOLOGY OF BRAZIL. 559 
 
 the southern hemisphere. D Archiac* has already called 
 attention to the fact that no mention of either striae, fur 
 rows, or polished surfaces has been made by those who have 
 studied the drift of South America, which seems very re 
 markable. He suggests that it may be perhaps owing to 
 a want of attention on the part of the travellers. 
 
 It is not to be wondered at that, when Professor Agassiz 
 claimed in 1865 to have found glacial drift in the vicinity 
 of Rio, scientific men were astonished and doubted the cor 
 rectness of the Professor s deductions ; and when from under 
 the equator he reported the discovery of glacial moraines 
 the statement seemed past belief. 
 
 In the preceding pages, in connection with a careful 
 description of the Brazilian coast, I have noted with much 
 detail the occurrence of certain surface deposits northward 
 to Pernambuco, at least, which deposits I have claimed to 
 be glacial drift. I propose now in this chapter to bring 
 together as concisely as possible all the facts bearing upon 
 this subject, and then to discuss them for the purpose of 
 showing that no other hypothesis than that of the glaciation 
 of the coast is sufficient to account for them. 
 
 Yon Eschwege describes a formation which is known in 
 Minas Geraes as Tapanoaeanga.^ It consists of angular 
 
 * G&oloyie et Pafcontologie, p. 719. 
 
 t Von Eschwege, in his Georjnostisches Gemalde von Brasilien, p. 30, gives so 
 interesting an account of the Tapanhoacanga that I translate it almost entire. 
 He says : " This rock is composed of sharp-cornered, angular, rarely slightly 
 rounded fragments of micaceous iron (eiseny!imrner), specular iron, and mag 
 netic oxide of iron, held together by a red, yellow, or brown ochreous cement. 
 These fragments are from several lines to eight inches in diameter. It is 
 often very auriferous, and contains sometimes scales of talc, chlorite, and 
 here and there fragments of itacolumite. The cement becomes in some places 
 
560 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 or rounded fragments of micaceous iron and other rocks 
 cemented together by an ochreous paste, which sometimes 
 exists without the gravel. This formation, with a thickness 
 of .from six to nine feet, more or less, he states, wraps the 
 highest mountains round about like a mantle. The same 
 material is found elsewhere. Mawe, in his description of 
 the mines- of Jaragua in Sao Paulo, speaks of the gold as 
 
 so abundant that the embedded pieces are not visible ; this then forms dis 
 tinct deposits of red ironstone in thin layers, containing ordinarily many 
 little flakes of mica. This rock is not only found in the valleys and on 
 the slopes of the mountains, but it covers their most elevated ridges and 
 flanks like a sort of mantle from half a toise to a toise and a half in thick 
 ness [a toise equals 6,395 feet] ; it is in general superimposed upon the fer 
 ruginous schist and clay-slate. The most important foreign mineral depos 
 its found in it are brown haematite and wavellite, which occur in consider 
 able masses near Villa Rica The Serra do Tapanhoacanga near Con- 
 
 gonhas do Campo (Province of Minas), whose summit rises to a height of 
 4,800 feet, is completely covered, over an area of several miles ; all the flank 
 of the mountain where Villa Rica is situated is incrusted with it ; the surface 
 is overturned by the mining works. The Campo de Saramenha, vis-a-vis, is as 
 if paved with it. It is abundant along the route from Villa Rica to Serro do 
 Frio ; it is probably also found in the Province of Goyaz. Tapanhoacanga sig 
 nifies in an African idiom * negro s head ; the miners have given this name to 
 the rock in question because of its uneven, knotty surface, which appears con 
 cretionary like an haematite. It is difficult to explain the origin of this con 
 glomerate. The angular fragments, the irregularity with which they are piled 
 up one on the other, the manner in which this rock covers, like a coat or glaz 
 ing, the top and flanks of the mountains, tend to make one believe that it is not 
 the result of the rapid degradation of the ferruginous mountains which formed 
 only the most elevated points in the country, and of which the peak of Itabira, 
 the Serra da Piedade, and others are the remains, but that it is due to the ex 
 tremely prompt drying up of the liquid, which sojourned formerly on the moun 
 tains, and has brought there the fragments which to-day cover them. These 
 could not follow the liquid to the bottom of the valleys, and were arrested, like 
 solidified lavas, on the midst of the slopes. The disorder with which they are 
 piled up proves sufficiently that the deposit was not gradually made." Castel- 
 nau says that the canga is certainly of phi tonic origin ! 
 
 * This is not African, but Tupi, Tapanhitna meaning negro, and acanga head. 
 
RESUME OF THE GEOLOGY OF BRAZIL. 561 
 
 occurring in a layer of cascalho, he writes it incorrectly 
 cascalhao, or gravel of rounded pebbles, principally quartz, 
 which wraps the hills round about, and is covered by a 
 sheet of soil. This surface deposit rests on gneiss. At 
 Minas Novas the gold, as we have already described, has 
 been mined from a similar gravel composed of rounded 
 quartz pebbles, &c., with a ferruginous cement, and over 
 laid by a similar bed of clay, the whole resting on clay-slate 
 decomposed in place. 
 
 At Rio the rounded surface of the decomposed gneiss is 
 covered by the same sheet of quartz pebbles and overlying 
 clays, and all the province, except the flat alluvial plains, 
 such as border the coast, and whose elevation is usually less 
 than twenty feet, is covered with the same deposit to the 
 tops of the highest hills I have examined. The pebble 
 sheet, it is true, varies in thickness, and in some localities is 
 absent, especially over areas in which quartz veins are not 
 abundant. In some places the pebbles are coarse, in others 
 fine, and occasionally we find intermingled with them frag 
 ments of gneiss, trap or tertiary sandstone. We find these 
 surface deposits everywhere lying immediately over a 
 rounded surface of gneiss, albeit the rock may be decom 
 posed to a great depth. The pebbles and rock fragments 
 are not confined to the pebble sheet alone, but, sometimes 
 rounded, sometimes angular, they are frequently found in 
 the overlying clay. This last may vary greatly in thickness 
 and color, but the general composition is very uniformly 
 the same. The whole deposit is everywhere without struc 
 ture, presenting no trace of stratification. The same layers 
 extend over the provinces south of Rio. To the northward 
 they are found all over Minas. I have seen them covering 
 uniformly the hills of Espirito Santo, and the coast of 
 
 24* jj 
 
562 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 Bahia, Sergipe, Alag6as, and Pernambuco. Northward I 
 have myself not seen these deposits, but Professor Agassiz 
 reports their existence in various localities on the coast 
 north of Pernambuco, and in the valley of the Amazonas 
 westward to the confines of Peru. 
 
 At Rio, as described by Professor Agassiz in the " Journey 
 in Brazil," and by myself in the chapter on Rio de Janeiro, 
 there are, in the valley of Tijuca, near Rio, and elsewhere, 
 deposits of immense boulders of trap, gneiss, <fcc., which 
 are evidently morainic and the work of local glaciers ; and 
 the Professor has described similar moraines as existing in 
 the Province of Ccara. The pebble and clay sheet covers a 
 large portion of the province of Minas, and is found not 
 only on the hills, but on the campos. In the provinces of 
 Bahia, Sergipe, Alagoas, and of the north there is, as I have 
 described, a zone of dry country, lying just behind the coast 
 forest belt, and largely composed of gneiss, mica-slate, and 
 the like, over which the surface deposits consist of boulders 
 of rock of all sizes, rounded and angular, scattered over the 
 surface, and sometimes piled up in confusion, with very 
 little soil, the rock frequently being bare : with these occur 
 rounded quartz pebbles. The surface of the gneiss country 
 in this dry zone in Bahia, as observed by Messrs. Allen 
 and Nicolay, and in Sergipe and Alagoas by myself, is re 
 markably even, and over large tracts forms a plain. The 
 topography is remarkable for shallow depressions without 
 outlets, forming ponds during the rainy season. Mr. Allen 
 describes having seen immense pot-holes worn in the rocks 
 by the action of falling water on some of the highest swell 
 ings and elevations of these plains, now far away from any 
 obstacle over which water could be precipitated. 
 
 Such, in brief, are the characteristic features of the sur- 
 
RESUME OF THE GEOLOGY OF BRAZIL. 563 
 
 face deposits of the Brazilian coast, to which I have applied 
 the name of drift. I have many times called attention to 
 the rounded surface on which the drift rests, though I have 
 nowhere seen either polished or striated rocks, which is 
 not at all wonderful, as the surface of the rock, wherever 
 I have examined it, even where the decomposition was least, 
 as on the Rio de Sao Francisco below the Falls of Paulo 
 AfFonso, was always more or less decomposed. 
 
 Nowhere over the whole region covered by the drifts do 
 we find other water deposits than those clearly referable to 
 the action of rivers or lakes. Above the old sea level of 
 Rio, Victoria, Bahia, &c. there are neither raised beaches 
 nor any other testimony of the action of the sea. 
 
 All this immense sheet of structureless clays, gravels, and 
 boulder deposits stretching along the whole coast, and cov 
 ering alike the coast tertiary plains, the elevated campos, 
 and the serras from bottom to top, belongs to the same for 
 mation, and is referable to the work of the same geological 
 agent. "We have claimed with Professor Agassiz, to whom 
 belongs the honor of the first announcement of the occur 
 rence of drift in Brazil, that that agent was glacier ice. 
 This hypothesis has been much disputed, and many other 
 ways of accounting for the formation of the sheet of detritus 
 have been proposed. Among them the most important are 
 the following, the respective merits of each of which we 
 propose to examine in detail : 
 
 I. Sub-aerial decomposition. 
 
 II. Wave action acting over the surface of the country 
 during a slow subsidence of the coast. 
 
 III. Wave action extending over the surface during a 
 slow rise of the land. 
 
 I. Decomposition. We have seen how decomposition 
 
564 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 may, as at Rio in the case of trap-dikes, at Victoria, and 
 on the islands of the Abrolhos, produce, with the aid of 
 rains, not only a soil, but boulders of decomposition, which 
 may be rounded or angular, and resemble drift boulders 
 so closely as to make it exceedingly difficult to distin 
 guish them from erratics. One can easily conceive how, 
 in the gradual decomposition of a bare surface of rock, 
 gneiss, for example, as the rock wasted, the resulting 
 clay and sand may be washed away, and spread over the 
 surface of the soil on the lower grounds. Such is indeed 
 the case, and one sees at the foot of the gneiss precipices, 
 not only at Rio, but elsewhere, a soil of this kind. It 
 closely resembles the drift, but is more washed, the sandy 
 portion remaining near the foot of the precipice, while the 
 muddy part is carried farther off. 
 
 Where the surface of a hill is very uneven, unfurnished 
 with soil, and strewn with blocks of rock, as in the dry zone 
 of Bahia, and the Sao Francisco, one may readily see how by 
 decomposition a structureless soil might be formed covering 
 the surface ; but there are a few facts which make this whole 
 hypothesis of the formation of the drift of no value. In 
 bluffs, natural or otherwise, in the vicinity of Rio, as well as 
 in the cuttings on the Dom Pedro Segundo and Cantagallo 
 railways, and on the Uniao e Industria road, and in the 
 Minas Novas region, one may see a great thickness of the 
 surface deposit lying on rock decomposed in situ, and lying 
 undisturbed on the solid rock. Though in old excavations 
 it is very difficult, in any fresh cutting it is the easiest thing 
 in the world to point out the line of separation between the 
 surface detritus and the decomposed rock,* which, by the 
 by, may have only the very thinnest coating of clay, or may 
 
 * This line is shown in the engraving 1 on page 508. 
 
KESUMti OF THE GEOLOGY OF BRAZIL. 565 
 
 be bare. The distinction between this drift and the de 
 composed rock is of the sharpest kind. The surface deposit 
 is without structure, and has the same appearance that the 
 decomposed rock below would have if it were ground up, 
 and intimately mixed together without washing, while in a 
 mass of the decomposed rock one sees the relative arrange 
 ment of the materials preserved undisturbed, with the quartz 
 veins, &c., in place. The veins invariably terminate ab 
 ruptly at the line separating the decomposed rock from the 
 overlying deposit. 
 
 One never sees a quartz vein traceable through the clays, 
 as it certainly would be if these had resulted from decom 
 position alone. The clay is usually remarkably free from 
 quartz pebbles or boulders, and one rarely sees even a peb 
 ble on the surface of the ground in the gneiss regions near 
 Rio, which would certainly not be the case if it were simply 
 a product of decomposition. The greatest objection to 
 the theory under discussion is presented by the sheet of 
 rounded and angular quartz pebbles, for that could never 
 have originated through decomposition. It is evidently the 
 result of mechanical action of some kind, and I am con 
 vinced that we must refer the overlying clay-sheet to the 
 same cause. It is evident that the agent, whatever it was, 
 that rounded the pebbles and ground up the clays must 
 have had some part to play in the moulding of the country, 
 though it is to erosion and decomposition that I should 
 attribute the broader topographical features of the coast, 
 and I would refer to the agent that formed the drift the 
 moulding of the actual surface on which the superficial 
 deposits now rest. 
 
 Let us now discuss the merits of the hypothesis that the 
 surface detritus has been the result of water action, and 
 
566 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 examine the two theories of wave action brought to bear 
 over the country during a gradual rise of the land from 
 the sea, and of similar action exerted in like manner over 
 the country during a subsidence. 
 
 II. Wave Action during" a Rise of the Land. It has been 
 suggested that, in a rise of the land, wave action brought to 
 bear over the surface might leave a coating of loose mate 
 rial similar to the drift strewn over the whole country, but 
 I must confess that, even for a single isolated hill, I can 
 not see how this theory would have the slightest weight, 
 for it could never produce an arrangement of the materials 
 such as actually exists. Suppose, for instance, that we have 
 
 a hill a few hundred feet in height, and which is rising. 
 At a certain time the sea level is found at a, so that 
 the waves will wash the hill-top. No one who has been 
 on the sea -shore will doubt that the effect will be to 
 remove the finer materials, and carry them off to be de 
 posited somewhere in quiet water ; and this is the nat 
 ural effect of the sea where it breaks against a slope or 
 over a ridge. It may be that on the sides of the moun 
 tain these materials, if in sufficient abundance, may be 
 deposited, and we will suppose that they are so depos 
 ited at e, but on steep slopes we should not expect to find 
 them. Gravels and sands would be washed by the waves 
 into deeper water, but could rest only on gentle slopes. If 
 the slope from c to e were twenty or more degrees, and 
 
RESUME OF THE GEOLOGY OF BRALIL. 567 
 
 the land should rise gradually so as to bring wave action to 
 bear successively over that slope, we should expect to find 
 the rock surface swept quite clean of loose materials, only 
 sand and gravel being thrown beyond the reach of the waves 
 into hollows or flat places on the rocks ; and it seems to me 
 that the result of such a rise would be, even if there were 
 already a coating of decomposed rock on the surface, only 
 to wash it over and gully it out on the hill tops and sides, 
 if it were deep, and cover it with sand and gravel, or to re 
 move it entirely, leaving the hill tops and sides bare, the 
 loose materials being transported into the valleys, and there 
 left as stratified deposits ; and not only in the valleys, but 
 upon flat places on the hill-sides, should we expect to find 
 such deposits. Now we find nothing of the sort. Asso 
 ciated with the drift are neither sands nor stratified deposits 
 of any kind. Besides, this hypothesis does not touch the 
 question of the formation of the great angular and rounded 
 boulders unassociated with sands strewn over the dry zone, 
 nor does it explain the transport of boulders over an irreg 
 ular surface. The tendency would have been to sweep the 
 hill-tops and the steeper and seaward slopes bare, which 
 is not the case. 
 
 III. Wave Action during a Subsidence. The hypothesis 
 of the action of waves over the country during a slow 
 subsidence of the land is equally untenable. On a gentle 
 slope sands and gravels would be formed, and perhaps 
 deposited immediately upon the rock, and this sheet one 
 might suppose drawn up like a curtain over the hills as 
 the land sank. The lighter materials settling in deeper 
 water might, at the same time, form a sheet drawn up 
 over the first, so that we might have the surface covered 
 by a sheet of sand and pebbles and over this a sheet of clay. 
 
568 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 But we have the pebble sheet unaccompanied by washed 
 sands, which is something incomprehensible under this 
 hypothesis, and, what is of more importance, we find them 
 lying on slopes so steep that it would be impossible for them 
 to remain unless supported by the weight of the overlying 
 clay. It is incomprehensible how water could have formed 
 these deposits without at the same time laying down strati 
 fied beds of sand, gravel, &c. in the valleys, or of leaving 
 sand deposits on the plains. It is impossible that the sea 
 should have been without currents and without tides, and the 
 inevitable effect of their action would be to sweep away the 
 lighter material and deposit it along the shore. In the ter 
 tiary beds, not only of the coast, but of the interior, we have 
 an example of a sandy deposit rapidly thrown down in a 
 muddy sea, and which, besides sand, contains a very large 
 percentage of clay, passing even into pure white clay ; but 
 nowhere is it difficult to distinguish these tertiary deposits 
 from the drift. The tertiary clays were the products of 
 the destruction of gneiss and other metamorphic rocks, 
 and so was the drift, but in the former the material 
 has been washed, though rarely ever arranged. The mica 
 has been completely destroyed, and one sees nothing of 
 it, while in the drift-earth it is constantly to be found. 
 One would expect that, if the Brazilian drift were a sedi 
 mentary deposit, the clays would resemble those of the ter 
 tiary, which is nowhere the case. They are everywhere 
 just such a material as would result from the mechanical 
 trituration of the rocks, and are wholly without stratifica 
 tion or signs of having been deposited by the action of 
 water. There has been within recent times a slow eleva 
 tion of the Brazilian coast. In sheltered as well as exposed 
 situations the deposits which it has brought above the sur- 
 
RESUME OF THE GEOLOGY OF BRAZIL. 569 
 
 face consist of sand ; gravels and clays are exceedingly 
 rare along the exposed coast. Nowhere in river, lake, or 
 sea deposits have I ever seen on the Brazilian coast any 
 thing resembling the drift-clays. I have studied with care 
 the effect of the action of the sea on the solid rocks of 
 all kinds along the coast, with the view to ascertain, if 
 possible, whether the peculiar evenly moulded surface cov 
 ered by the drift could in any way be due to water action 
 joined to the effects of decomposition. Where the rock 
 is gneiss and very homogeneous in structure, and not well 
 bedded, as is the case in the range of hills skirting the sea 
 shore on both sides of the entrance to the bay of Rio, the 
 rocks swept by the waves may have a very smooth and reg 
 ular outline, but where, as on the shores near Boa Yiagem 
 at Rio, Os Busos, Ilhdos, or Bahia, the rock is well bedded 
 and the strata are very highly inclined, the softer beds give 
 way first and leave the harder projecting, and the rocks 
 within reach of the waves are worn in the most irregular 
 manner. Now I do not see how we can resist the conclu 
 sion that, if the surface clays and gravels were the products 
 of wave action, we ought to find the surface of the rock on 
 which they rest showing some signs of that action in the 
 wearing away of the softer beds, leaving the harder stand 
 ing up ; but this is never the case. There can be no transi 
 tion more abrupt than that from a wave-washed, rock-bound 
 shore, and the smooth, even outlines of the hills above the 
 line of wave action. Take, for instance, the coast between 
 the lighthouse at Bahia and the Morro do Conselho, which 
 is to a large extent rocky, and examine the moulding of the 
 rocks washed by the Atlantic surf, and then compare it with 
 the moulding of the rock on the seaward side of any of the 
 exposed hills, where, if washed by the sea, the rock would 
 
570 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 have been subjected to the pounding of the same surf, and 
 you will be convinced that the moulding of the drift-covered 
 rock-surface was due to an agent that did not respect so 
 thoroughly the difference in hardness between the beds as 
 water does. It is useless to suppose that the sea might 
 have washed against a decomposed surface which might 
 have been rounded down by the surf. The wash of the 
 waves would have removed it entirely. The rounded wave- 
 washed rocks on the shores of Rio are bare, and it is incon 
 ceivable how, in the face of the tremendous Atlantic surf, 
 they could ever be covered by detritus by wave action, as 
 all the hills lying along the coast of Bahia and Rio inva 
 riably are. Along the coast where decomposition prevails, 
 from the wetness of the climate, the clays are very abun 
 dant, and it is a rare thing to find boulders of any other 
 rock than quartz ; but in measure, as one goes inland and 
 approaches the dry zone, as is beautifully seen on the Sao 
 Francisco, the clays grow less abundant while the pebble- 
 layer gradually passes into a sheet of boulders of rock scat 
 tered over the surface with little admixture of earth. The 
 rock being of the same general character over large areas, 
 it is usually a difficult matter to decide whether a boulder 
 is travelled or not ; * but I have seen at Piranhas syenite 
 boulders lying on gneiss, though evidently coming from not 
 far away, and I have seen gneiss and quartz boulders lying 
 in the clays on the tertiary plain on the Mucury. I have 
 already called attention to the intermixture of greenstone 
 and gneiss boulders of immense size in the valley of Tijuca, 
 occupying situations into which water could not have 
 brought them, and into which they could not have fallen. 
 
 * This is an important point to bear in mind. The geology of Brazil is so 
 very simple that we find the same kind of rock over immense areas. 
 
RESUME OF THE GEOLOGY OF BRAZIL. 571 
 
 No one seeing the boulder-scattered surface of Bahia and 
 the Sao Francisco, where the decomposition is exceedingly 
 slight, would ever, I am persuaded, seek for an explanation 
 of the distribution of these masses over the surface in run 
 ning water or wave action, which last would have been pow 
 erless over so uniformly level a surface. I must insist upon 
 the fact that the unarranged materials are precisely like our 
 unmodified drift in the north, and that the surface of the 
 rock on which they lie has the moulding of the surfaces on 
 which our northern drift lies, and that if we refer the north 
 ern drift to the action of glacier ice, we must do the same 
 thing for the Brazilian surface detritus, contrary as it is to 
 all our preconceived opinions of the distribution of drift. 
 The fact that neither Professor Agassiz nor myself, nor 
 any one else of our expedition, has been able to discover 
 glacial stria) in Brazil is of very secondary importance. The 
 drift itself exists all over the country, and it cannot be 
 explained away. I have looked carefully for stria?, but 
 there has been everywhere enough decomposition of the sur 
 face of the rock as well as of the boulders scattered over it 
 to have destroyed all trace of them. Once I thought I had 
 found striae. On the Dom Pedro Segundo Railroad, near 
 Mendes, while engaged in making an examination of the 
 cuttings, I found one in which the drift-clay had been 
 removed from over the decomposed gneiss, exposing the gla 
 ciated surface. This appeared to be quite fresh, and to my 
 surprise was deeply furrowed with parallel strise. I took 
 pains to inquire of the engineers of the road, and learned 
 that the drift had slid off from the upper part of the cutting, 
 which was a sufficient explanation. I speak of this only in 
 order to put other observers on their guard against being 
 deceived by any similarly striated surfaces. I can offer but 
 
572 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 one explanation of the formation of the pot-holes observed 
 by Mr. Allen, and that is that they were formed by glacial 
 cascades in the same way as the pot-holes seen so often 
 on the surface of ridges in the north have been formed 
 during the drift ; for, according to the testimony of Mr. 
 Allen, the pot-holes of the Province of Bahia occur on 
 the gneiss plains, far away from any present obstacle over 
 which the water may have flowed. Mr. Allen describes 
 them as being exceedingly well preserved, and having 
 smooth sides. 
 
 The drift is, as above stated, removed everywhere down 
 to the limit of wave action before the later elevation of 
 the coast ; but the occurrence of the drift on some out 
 lying hillocks of the tertiary clays on the line of the exten 
 sion of the Cantagallo Railroad, between Porto Novo and 
 Porto das Caixas, which are now surrounded by recent 
 sands, made me suspect that the clay was once continuous 
 below the present sea level between them and the main 
 land. The fact, too, that it extended uniformly down to 
 the same level everywhere was almost sufficient proof that 
 it formerly extended to a much lower level. At Bahia, as 
 already described, recent sands blown or washed over the 
 drift have been cemented and have protected it from the 
 action of the sea, so that it may be seen extending beneath 
 them down nearly to low tide. This fact seems to prove 
 satisfactorily that formerly the land stood at a higher level 
 even than now.* Drift occurs on some of the islands off the 
 coast. 
 
 * From the observations of Darwin and others we know that this recent 
 uprise has been much greater in the south than in the north, and it seems to 
 increase in going south from Eio to the Straits of Magellan. It would seem 
 that the great movements just antecedent and posterior to the drift period in 
 
RESUME OF THE GEOLOGY OF BRAZIL. 573 
 
 I believe that during the time of the drift the country 
 stood at a much higher level than at present, and that it 
 was covered by a general glacier. Over the coast region, 
 where decomposition of the rocks had largely obtained, and 
 where the surface of the rock, rendered even by this agent, 
 had been covered by a thick layer of loose material, the 
 glacier reworked this loose material, and when it disappeared 
 left it as a paste, in which the harder materials, such as 
 fragments from quartz veins, &c., more or less rounded, were 
 embedded. The layer of quartz pebbles underlying the paste 
 appears to have consisted of coarser fragments borne along 
 by the bottom of the glacier, while the paste seems to have 
 been more or less distributed through the body of the 
 glacier. A glacier moving over the gneiss regions of Rio 
 or Espirito Santo to-day would find few loose rocks to trans 
 port, for the precipices are smooth and unbroken, and little 
 falls from them, so that one could not expect to see mo 
 raines of coarse materials formed by the glaciers of that 
 region, and if the ancient glaciers moved over a country 
 whose surface was decomposed, it is not wonderful that the 
 drift consists of paste with but few boulders. On the con 
 trary, over the dry zone the cliifs are ragged and broken, 
 and the rock surface is apt to be broken up, and we should 
 expect to find over such a region drift of a different char 
 acter from that which obtains over the moist coast region, 
 and resembling more closely the drift of North America. 
 
 In the drift-paste I have never seen the slightest trace of 
 organic remains of any kind. 
 
 Post-Tertiary. To this epoch belong the cavern depos- 
 
 South America have corresponded with those of North America during the 
 same period. In North America the oscillation of level was greater in the north 
 than in the south ; in South America it was just the reverse. 
 
574 GEOLOGY AND PHYSICAL GEOGRAPHY. 
 
 its in Minas Geraes, affording the remains of Mastodon, 
 Megatherium, &c., and the lagoa deposits on the borders of 
 the Rio de Sao Francisco already described. 
 
 Recent. To recent times belong the sands containing 
 recent shells, <fcc., exposed by the late uprise of the coast, 
 the solidified beaches, rock reefs of Pernambuco and else 
 where, the coral reefs, the peat deposits, and the alluvial 
 beds of the rivers and lakes. 
 
APPENDIX. 
 
OX THE BOTOCUDOS. 
 
 Origin of the Name Botocudo. Stature. Physical Form and Characteristics. 
 
 Manner of Wearing Hair. Lip and Ear Ornaments. Professor Wy man s 
 Description of Skull of Botocudo from Sao Matheos. Comparison with 
 other described Botocudo Skulls. Color of Botocudo. Manner of Paint 
 ing the Body. Dislike to being clothed. Bows and Arrows described. 
 Gerber s Enumeration of the Tribes. Von Tschudi s Description of the 
 Tribes. Ranches and Huts. Food. Mode of procuring Fire. Manu 
 factures. Marriage Customs. The Botocudos cruel Husbands. Fa 
 cility with which Wounds heal. Treatment of Children. Religious Ideas. 
 
 Belief in the Bad Spirit, Janchon. No Belief in a Supreme God. 
 Burial Customs. War Customs. Cannibalism. Dance. The Boto 
 cudos fast disappearing. Botocudo Character. Geographical Distribution 
 of the Botocudos. Peculiarities of their Language, Pronunciation, Gram 
 matical Structure, &c. Botocudo Vocabularies. 
 
 No Indian tribe of Brazil save the Tupis has been more cele 
 brated than that known as the Aimores, Aimbores, or Botocudos, 
 the latter being the name by which it is known in Brazil, as well as 
 in most recent works on the country. They call themselves Enge- 
 rackmung* a word which I cannot translate. Miing, in Botocudo, 
 means to go, and the termination, which is more likely to be a 
 separate word in the proper name, probably has the same meaning, 
 but I have not been able to find a definition for the remainder of 
 
 * I give Prince Neuwied s orthography. Gcrber, also a German, in his 
 Nocoes Geograjicos, &c., p. 24, spells the word Endgerelammg, and he says that 
 Guido Marliere gives it Cracknnm. The name Botocudo is spelled by different 
 authors in all possible ways, as, for instance, Botokoudy, Botokude, Bootoo- 
 cudy, &c. 
 
 VOL. i. 25 KK 
 
578 APPENDIX. 
 
 the word. The name Botocudo was without doubt applied to the 
 tribe by the Portuguese, because of the custom of piercing the 
 under lip and the ears, and inserting therein round, flat pieces of 
 wood, like barrel-corks, or botoques, as they are called in Portuguese. 
 The termination udo in Portuguese has the signification of fur 
 nished with, as in the words cabelludo, velludo. It is true that 
 Bodoque means a pellet of clay, such as is thrown from a sort of 
 bow in use among the Indians of Brazil, and that the same word 
 means also a kind of stone or earth, employed by the Indians 
 to ornament the body. Some have thought that the name of the 
 tribe was derived from this word, but the derivation I have above 
 given is without doubt the correct one. The Corografia Brasili- 
 ca * gives the same origin for the word, and so do Neuwied and 
 Von Tschudi. 
 
 In Espirito Santo and in the Mucury region they are com 
 monly called Bugres, a name which Von Tschudi derives from the 
 French. In Sao Matheos and on the Doce I heard them called 
 Tapuyos, a Tupi word applied to savages generally. 
 
 Judging from the Botocudos I have seen, I should, with Von 
 Tschudi and M. Serres, describe the race as of middling height. 
 I have seen many who were five feet ten inches in height, and 
 I remember especially one powerful fellow, who could not have 
 measxired less than five feet eleven inches. D Orbigny makes the 
 
 * "As ontras nacres convizinhas, ao mcnos algumas chamam-lhos Aymbores, 
 e os conquistadorcs por corrupyao Aymore s; mas de muitos tempos por ca 
 quasi nao tcm outro nome entrc os Ohristaos scnao o de Botocudos pelo extra- 
 vagante e ridicule costume de furarem as orelhas e os bei^os e dilatarem-nos 
 notavelmente com rodellas de pan, pareccndo-lhes quo fleam assim mais gentis 
 e airozos." Corografia BrasiUca, Tome II. p. 72. 
 
 The Dlccionario Geographico derives the name from Boto and codea, " because 
 the Indians of this nation were roJhos [short and thick], and went with the body 
 covered with a coating of gum-copal, with which they were accustomed to paint 
 themselves, to preserve them from the stings of mosquitoes and other insects." 
 This is a custom which at present, at least, does not seem to be in use among 
 the Indians. 
 
ON THE BOTOCUDOS. 579 
 
 mean height of the male Botocudo 1.620 metres, and his extreme 
 height only 1.000 metre, but this must be a typographical error.* 
 According to M. Porte, the height of the male Botocudo varies 
 from 1.85 metres to 1.18 metres, and that of the women from 
 1.35 metres to 1.16 metre.f 
 
 The limbs and body of the Botocudo, though exceedingly strong, 
 look soft and effeminate, and the muscles have not the same prom 
 inence and knottiness seen in the muscularly developed white or 
 negro. 
 
 They are generally broad-shouldered and large-bodied, but their 
 arms, and especially their legs, are apt to be thin, though very 
 muscular, and the latter strike one as being disproportionately 
 small, when compared with those of the negro and white man, the 
 calf being but slightly developed. Von Tschudi calls attention to 
 this, and Agassiz speaks of the small size of the legs of the Indian 
 in comparison with his square, heavily built trunk, but I have seen 
 Botocudos as well proportioned as the whites. J In all the males 
 
 * L llomme Americain, Tome I. p. 102. 
 
 t Comptes Rendus, Tome XXI. p. 5. 
 
 J Specimens of Naknenuks were carried to France by M. Porte and were 
 examined by M. Serres, who published in the Comptes Rendus (Tome XXI. 
 p. 7) a description of them. I am not aware that any other scientist has made 
 a more detailed study of the Botocudo than he, and I quote a paragraph or two 
 from him relating more especially to the configuration of the trunk. M. Serres 
 says : 
 
 " La poitrine e tait bien conformee chez I liomme ; un peu aplatie snrlede- 
 vant elle paraissait d une seule venue et ne pre sentait pas 1 espece de voussure quc 
 Ton rcmarque au niveau du grand pectoral chez les hommes de la race caucasique 
 dcveloppes au mume degre ; voussure qu offniicnt d une maniere marque e les 
 Americains loways, comme on le remarque chez les hommes les plus forts de 
 la race caucasique. En revanche elle paraissait plus allongee chez le Botocude 
 ct plus large quo 1 ordinaire a la region infe rieure. La poitrine de la femme 
 e tait, en arricre plus arquce, que celle de rhommc; en avant, elle s inclinait en 
 bas d une maniere si marque e, qu il m a fallu la mesurer plusieurs fois pour 
 m assurer qu il n y avait rien d exage re dans le portrait qu en a fait notre peintre 
 si distingue du Museum M. Verner. De cette inclinaison de la poitrine re sul- 
 tait 1 abaissement du sein, abaissement qui rappelait celui dcs femmes ethio- 
 
580 APPENDIX. 
 
 the pelvis seemed extraordinarily narrow, and the hinder parts very 
 small. The hands are well formed but small. Those of the women 
 are particularly so, as M. Serres has remarked. The feet are 
 smaller than in the Caucasian race. 
 
 The physiognomy of the Botocudos varies so extraordinarily that 
 it is exceedingly difficult to describe its peculiarities. Of a dozen 
 or more Botocudos in the fazeuda of Capitao Grande no two looked 
 alike.* There were two young men who were partially civilized, 
 and spoke Portuguese, and I should never have taken them for 
 anything else than very light-colored mulattoes. They all have 
 low foreheads, as Von Tschudi, Neuwied, and M. Serres have re 
 marked. Their eyes are black, usually small and full of life. 
 Neuwied says that blue eyes sometimes occur. M. Serres says 
 that those of the women he saw were more open than those of the 
 men. The exterior angle of the eye is sometimes a little oblique. 
 As for the nose, it is usually rather short : and in the four profiles 
 of Botocudos at Sao Matheos in my note-book it is represented as 
 having a concave outline, the extremity being large, while the ala> 
 are rather wide ; but I have seen examples of narrow and arched 
 noses. I give two of these profiles in the following wood-cut, f 
 
 piques, ct qni pourrait devenir nn caractcre de grandc importance s il n y avnit 
 rien d individuel dans cette disposition. 
 
 " Comme celui do I hoinme, le thorax de la fcmmc etait tres-elargi infcrieure- 
 ment ; cet c largissementne parait avoir sa cause dans 1 abaissement dc la volume 
 du foie, qui je rcconnus par la percussion dans Ics limites infc rieures quo n atteint 
 jamais cct organe dans son etat nature! chez la femme caucasiqne. 
 
 " Avec cct abaissement du foie co incidait un abaissement de 1 ombilic, et h 
 celui-ci re pondait un abaissement du pnbis, que je rcconnus avec peine, a cause 
 de la saillie graisscuse du mont de Venus. L abaissement dc 1 ombilic faisait 
 saillir 1 abdompn en bas et surles cotes et celui du pubis inclinait en bas et en 
 arriere le bassin ; de la resultait amplcur de la region fessiere, deja moins de- 
 vclope e que cbez la fcmme caucasique." 
 
 * Nothing can be more false than the oft-quoted and sweeping assertion of 
 Ulloa : "Vistoun Indio dc qualquer region, se puede decir que sc han visto 
 todos en quanto el color y Contestura." Noticias Americanos, p. 252. 
 
 t I do not offer these sketches as accurate portraits. They were drawn from 
 
ON TliE BOTOCUDOS. 
 
 581 
 
 BOTOCUDO MAN AND WOMAN. 
 
 M. Serres describes the nose of the men as straight, and those 
 of the women as slightly arched at the base. In both, according 
 to the same author, the alee are large, but more so in the men 
 than the women. 
 
 Neuwied s plates of the Botocudos are well drawn, but they 
 give one absolutely no idea of the race. The chief Krengnatmuck, 
 barring his abominable head, has the figure of a Caucasian, while 
 his wife might have posed for a Venus. Neuwied s figures were evi 
 dently drawn from Caucasian models. The Botocudos as a race are 
 very ugly, but some of the young girls might, by a very liberal 
 construction of the word, be called pretty. As a general rule, the 
 women have the abdomen very large, the breasts flabby and pen 
 dent, and not infrequently they are bow-legged. The children, like 
 all Brazilian children, are apt to fall into the habit of dirt-eating, 
 and are very often stunted, swollen, and sickly.* 
 
 nature, and the outlines are not far from correct. I have introduced them also 
 for the purpose of showing the appearance of the pierced car and lip. The 
 woman carries a bag on her back. 
 
 * I saw in Brazil a large number of children and adults who Avere addicted 
 to this habit. In most cases the clay is not eaten because of want of nourish 
 ment, but from a morbid appetite. Among some nations, however, as the 
 
582 APPENDIX. 
 
 The cheek-bones are generally quite high, though not so much 
 so as among the Tupi descendants. They appear especially promi 
 nent in the emaciated old women, who are wretchedly ugly in ap 
 pearance. The face is somewhat flattened. The mouth is always 
 very large, and the lips are quite thick. 
 
 The hair is black, coarse, and straight. The beard is of the same 
 character, and very sparse.* They generally pull it out ; but I 
 have frequently seen men with a very sparse beard,t as, for in 
 stance, the one whose profile I have given. They sometimes cut 
 off or pull out the eyebrows, and the women at least suffer hair to 
 grow on no part of the body except the head. The hair is always 
 worn short, and falls over the forehead. Sometimes it is shaved 
 away for two or three finger-breadths all round, with a razor made 
 from bamboo ; but this custom is not general, and none of those I 
 saw at Sao Matheos were shaven in this way. The women usually 
 wear in their perforated lips and ears round disks of wood (60- 
 toques), like the cork to a large, wide-mouthed bottle. Of the many 
 Botocudos I have seen on the Rio Doce, at Sao Matheos, at Colo- 
 nia Leopoldina, Urucu, and Philadelphia, only the adult women 
 had both ears and lips pierced. The old men invariably had the 
 ears perforated, but I do not remember ever having seen a male 
 with a hole in his lip, and I never saw a child with either ear or lip 
 pei forated, which leads me to suppose that the custom is going 
 out of use. The piercing of the lip and ear is performed, according 
 to Xeuwied, when the child is seven or eight years of age. Neu- 
 wied says that it is done with a sharp piece of wood ; other writers 
 
 Ottomacs on the Orinoco, large quantities of clay are eaten in times of great 
 scarcity of food. Humboldt has investigated this subject, in his usual exhaust 
 ive manner. Sec his Travels, Bohn s Edition, Vol. II. p. 495. 
 
 * M. Serres says (loc. cit.) : " Lcurs cheveux etaient noirs, epais, courts, lisses 
 et limitc s en demi-cercle sur le front. Ceux de 1 homme e taient plus rudes que 
 ceux de la femme." 
 
 t Dr. Karl August Tiilsner describes the Botocndos of the Colonie Leo 
 poldina as wearing a sparse beard. (Die Colonie Leopoldina in Brasilien, Gottin- 
 gen, 1860, p. 65.) 
 
ON THE BOTOCUDOS. 583 
 
 say that the sharp spine of the Airi palm is used. The openings 
 once made, small pieces of wood are inserted to distend them, 
 afterwards larger and larger ones being used until the opening of 
 the ear may be, according to Neuwied, even four inches in diam 
 eter ! I have never seen a lip-plug in use more than two inches 
 in diameter. The ear-plug is much larger. The lip or ear orna 
 ment consists of a thin section of the stem of a Barrigudo tree 
 (Chorisia), which furnishes a wood quite as light, if not lighter 
 than cork, and of a white color.* The lip-plug is usually about 
 three fourths of an inch thick. The lip surrounds it like a thick 
 red cord of flesh. It is usually worn the most of the time, but 
 may be, and is from time to time, removed. The lip then hangs 
 of course against the chin, a hideous loop of flesh, comparable 
 more to a great worm than anything else, displaying the teeth 
 with a horrible grin. The pressure of the plug against the lower 
 incisors in front pushes them out of place, and even causes them 
 to fall out, so that an old woman with the lip ornament always 
 wants the lower front teeth, and not infrequently the upper. 
 Neuwied describes and figures the jaw of a Botocudo in which 
 the alveolae of the front incisors had completely disappeared, leav 
 ing the bone as sharp as a knife. 
 
 Neuwied, in the Atlas to his Travels, on Plate 17, represents four 
 heads, three profiles and one full face. These figures are really of 
 very little value, as they have evidently not been drawn from na 
 ture. In Figure I. the ear-plugs are represented twice as thick as 
 they ought to be, and the under lip is repi esented as touching the 
 upper, which is absolutely impossible. The position of the plug 
 is better repi esented in Figure IV. It is usually carried, in the re 
 pose of the features, nearly horizontally. In a smile it is inclined 
 upwards, and often touches the nose. In eating it may be taken out, 
 bxit none of the Botocudo women I have seen eating removed it. 
 A more comical sight than an old woman sucking a stick of sugar- 
 
 * Fletcher says that these plugs are made from the wood of the aloe, which 
 is incorrect. Ewbank speaks of the plugs as made of pito wood, doubtless mean 
 ing the same thing. 
 
584 APPENDIX. 
 
 cane can scarcely be imagined. In quarrels the perforated ears and 
 lip are apt to suffer, and it is no uncommon thing to see them 
 broken. In this case the ornament is not necessarily discarded. 
 The two ends are then tied together with a bit of bark, or some 
 thing of the sort, and the plug is replaced. At Urucu I saw a rather 
 young woman whose lip had been torn and tied up. Usually the 
 ear-plug is not worn, and the loop of flesh is left dangling, some 
 times reaching to the shoulder. When the plug is removed the 
 opening generally appears very irregular, as in the man whose pro 
 file I have given. In travelling through the forest this loop would 
 be likely to be caught against limbs of trees and be torn, so it is 
 very often turned up and laid over the ear, which shortens the 
 organ in the first place, and produces a horrible deformity. I 
 observed that two old women at Sao Matheos wore the ear-flap in 
 this way even in camp. 
 
 Neuwied* cites a number of examples of nations that pierce the 
 ear and lip. The Aguitequedichagas, Lengoas, and Charruas of 
 Paraguay wore large blocks of wood in their ears and lips, but the 
 lip-plug was smaller than that of the Botocudo. The Gamellas of 
 Maranhao used immense wooden lip-plugs, and Major 0. C. James 
 informs me that the Bugres of Sao Paulo have the same custom, 
 though it is, however, now going out of use as the Indians become 
 civilized. Major James says that the civilized Indians close up the 
 opening in the lip with wax. The Muras, on the Amazonas, used 
 to pierce the lip, but the custom is now abandoned. The Tupi- 
 nambas wore ornaments of nephrite stone in the lip. Mr. George 
 Gibbs has called my attention to the fact, that the Koloshians 
 of Alaska pierce the lip and wear a plug. It is very interesting to 
 know that this custom obtains among savages so widely separated. 
 Wood describes a nation in Africa that pierces and distends the 
 upper lip by inserting a ring, a custom more hideous than that of 
 the Botocudo. 
 
 I give figures of the skull of a male Botocudo I obtained for 
 
 * Reise nach Brasilien, Band II. seite 7. 
 
ON THE BOTOCUDOS. 
 
 585 
 
 the Museum of Comparative Zoology at Sao Matheos. The 
 man s name was Kupara ck, or the Onga. He had died of disease, 
 and had been buried in the vicinity of the fazenda, but the rains 
 had uncovered the body, which his relatives had left to rot in a 
 swamp. A half-civilized Indian led me to the spot, and himself pro 
 cured for me the skull, which we 
 earned to the house and placed 
 on a table. When the Indians 
 came in to supper the Botocudos 
 gathered around and made sport 
 of it, thrusting their fingers into 
 the eyeless sockets and laughing 
 at it, though at the same time 
 they knew that it belonged to 
 one of their near relations. This 
 skull I placed in the hands of 
 Professor Jeffries Wyman, of 
 Cambridge, Massachusetts, who 
 
 has kindly furnished me with the following interesting and val 
 uable notes upon it. " From the references in the Thesaurus 
 Craniorum of Dr. J. Barnard Davis, page 235, it appears that only 
 
 a few crania of Botocudos 
 have been described, not 
 more than five in all ; and 
 of these but one has been 
 measured, and this very im 
 perfectly by Dr. Davis, as 
 he had only a cast, the origi 
 nal being in Stockholm. 
 The specimen from Sao Ma 
 theos is, therefore, a valua 
 ble addition to the previous 
 collections. It is that of a man somewhat advanced in life, the 
 teeth gone and the alveoli largely absorbed ; the sagittal and 
 25* 
 
586 
 
 APPENDIX. 
 
 lambdoidal sutures are closed in those portions where the two 
 join. The side walls of the head are vertical and the top some 
 what roof-shaped. The foramen magnum has about the same po 
 sition as in the American aborigines generally, its index being 40.6, 
 while in these it is 40.9. The breadth across the malar bones, 
 
 together with the roof- 
 shaped top, give to the 
 whole, when viewed in 
 front, a somewhat py 
 ramidal form compared 
 with that of the other 
 barbarous tribes gener 
 ally. The size of the 
 cranium is large, its 
 length being 510 milli 
 metres, and its capacity 
 
 1,435 centimetres, or 88 cubic inches; while theirs is only 1,376 
 centimetres, or 84 cubic inches. The length of the skull being 
 taken as 100, its breadth is 72.8, and it is, therefore, decidedly 
 elongated or dolichocephalic. The whole is massive and heavy, 
 and, at the hinder part especially, quite thick. 
 
 "Though somewhat smaller, this skull, as will be seen by the 
 following table, agrees very nearly in its proportions with that de 
 scribed by Dr. Davis in his Thesaurus. 
 
 
 Circum 
 ference. 
 
 Length. 
 
 Breadth. 
 
 Height. 
 
 Length. 
 
 Sao Matheos .... 
 
 in. ni. 
 510 
 
 III.]!]. 
 
 187 
 
 m.m. 
 136 
 
 m.m. 
 138 
 
 m.m. 
 380 
 
 Davis s collection . . . 
 
 525 
 
 190.5 
 
 139.5 
 
 144.5 
 
 401 
 
 Blumenbach s collection 
 
 
 165 
 
 
 139.5 
 
 
 " When compared with Blumenbach s specimen there is a wide 
 difference. The one figured by him in his Decades Craniorum, 
 
ON THE BOTOCUDOS. 587 
 
 Plate LVIIL, was brought from Brazil by the Prince Wied Neu- 
 wied, and is also figured in Morton s Crania Americana, Plate XV. 
 In the first case the view is oblique, and in the second exactly in 
 profile, this view being taken from a drawing furnished by the 
 Prince. 
 
 " In its thickness and massiveness it agrees with that of Sao 
 Matheos. As far as one can judge from the figures of Blumen- 
 bach and Morton, the one described by them is short and high, 
 and, according to Blumenbach s description, remarkable for its bru 
 tality, or, to use his own words, If you disregard for a moment 
 the under jaw and interval of the orbits, the projecting nasal 
 spine, and the other particulars peculiar to man, the general aspect 
 approaches nearer to that of the orang-outang than that of any 
 other skull from a barbarous nation to be seen in my collection. 
 His figure seems to justify his words, and represents the jaws more 
 projecting and simious than does that of Dr. Morton. 
 
 " Of the original shape of the jaw of the cranium of Sao Ma 
 theos it is now hardly possible to form a correct idea, since it is 
 somewhat broken. The teeth are gone, and the alveoli partly 
 absorbed. There is nothing, however, in what remains to in 
 dicate great size or forward projection. The whole cranium will 
 compare favorably with the crania of other barbarous tribes of 
 America. Certainly there is nothing indicative of extreme degra 
 dation." 
 
 In a letter accompanying the above notes Professor Wyman 
 says : " It is quite curious to see what an entirely different look 
 ing thing the skull of Sao Matheos is when compared with the 
 famous one described and figured by Blumenbach, and which has 
 hitherto given the key-note to all that has been written about the 
 skulls of Botocudos. If there were only your Sao Matheos skull 
 and his, and they had fallen into different hands for description, 
 one would have given us the connecting link of man with the apes, 
 while the other would have given us a highly respectable American 
 savage." 
 
588 APPENDIX. 
 
 The skull described by Blumcnbach* is figured as a vignette to 
 the second volume of Prince Max. zu Neuwied s Reise nach Bra- 
 silien. It seems to me to be most extraordinarily short and small 
 compared with the size of the jaw. Among all the Botocudos I 
 saw in Brazil there was not one with so ape-like a head and such 
 projecting jaws. 
 
 Von Tschudif figures the skull of a Botocudo chief, named Poro- 
 kum, from the Mucury. He gives a front but somewhat oblique 
 view of it, which, however, shows that it agrees with my specimen 
 in the perpendicularity of the sides and in the roof-shaped top. 
 
 The color of the Botocudo is a light yellowish-brown, like that 
 of a very white mulatto, or perhaps more correctly speaking that 
 of a white man somewhat tanned, not burned, by exposure to 
 the sun. Neuwied says that they are of a reddish-brown color, 
 Von Tschudi a dirty, nearly bronzed brown. $ The color varies 
 much. I should say that the bodies of the Botocudos I have seen 
 were, on the average, much lighter in color than those of the white 
 canoe-men of the Jequitinhonha, accustomed to work almost naked 
 in the sun, and that as a race they were much whiter than the 
 Tupi descendants along the coast, who differ most markedly in 
 their whole physiognomy, stature, &c., from their uncivilized 
 neighbors. 
 
 D Orbigny has called attention to the yellowish skin-tint of the 
 Brasilio-Guaram races. It is interesting to observe that the Boto- 
 
 * Prince Neuwied quotes at the end of the chapter on Botoctidos, Vol. II. 
 p. 70, a few words from Bluraenbach relative to this skull. Blumenbach says : 
 " Der Botocude womit Ew. meine ethnologische Sammlung bereichert haben, 
 und der ebcn so sehr zu den merkwiirdigsten als zu den seltensten Stiicken der- 
 selben gehort, ahnelt in seiner Totalform (doch ohne den Unterkiefer) dem vom 
 Ourangutang mchr, als einem der acht Negerschadel die ich besitze, wcnn gleich 
 bey manchen von diesen die Oberkiefer starker als an dem brasilianischen 
 Cannibalen prominircn." 
 
 t Reisen durch Siid-Amerika, Zweiter Band, 328 Seite. 
 
 J M. Serres describes the color of the Botocudos he examined as " brun rou- 
 geatre, un pcu plus rosee que celle des loways." 
 
ON THE BOTOCUDOS. 589 
 
 cudos, a nation fitted for a life in damp shady forests, and unable 
 to bear the sun in the open campos of the interior, are of a very 
 pale color, and very much less dark than those races which live on 
 the plains or on open grounds. Southey* thought that their pale 
 color was the result of their life in the shady forests, and D Or- 
 bignyt held the same belief, stating in confirmation of it that, 
 while the Guaranis of the plains of Corrientes and the Gran Chaco 
 are dark colored, the Guarayos and Sirionos, though belonging to 
 the same race of the Guaranis, who for about four hundred years 
 have lived in the damp and shady forests, are very light colored. 
 
 Gumilla says, that the people living in the forests of the Orinoco 
 are almost white, while those of the plains are dark. $ I am told 
 that the Botocudos are capable of blushing. I never detected any 
 thing of the sort, and I doubt whether it is possible in any except 
 those considerably civilized. 
 
 Among the Indians I saw on the coast only one young woman 
 had her cheeks painted, though the custom seems, at least for 
 merly, to have been quite common with the Indians in the forest 
 not only to paint the face but the body. This young woman had 
 a bright red spot on each cheek, painted with a tint prepared from 
 the seeds of the Urucu (Bixa Orellana Linn.), a fruit common in 
 the Brazilian forests, and from which anatto, or cheese-coloring, is 
 prepared. This same color is also used by other tribes in Brazil 
 to paint their bodies. A deep blue-black of greater durability 
 is extracted from the fruit of the Genipapo (Genipa), and is also 
 used for ornamenting the body. The style of ornamentation ap 
 pears not to be fixed, but to vary according to the caprice of the 
 individual. Neuwied describes three principal styles. In one 
 the face from the mouth upwards is painted red with urucu. In 
 
 * History of Brazil, Chap. XIII. 
 
 t L Homme Americain, Tome I. p. 79. 
 
 t Hist, de VOrtnoque, Trad., Avignon, 1752. Tome I. p. 108. Cited by 
 D Orbigny. See Humboldt, Travels, Bonn s edition, Vol. II. p. 463. 
 
 Henderson, in his usually inaccurate style, says that the Botocudos paint 
 themselves green or yellow. 
 
590 APPENDIX. 
 
 another the body, upper arms, and thighs to just below the 
 knees are stained black, the colored portion being separated from 
 the unpainted by a red stripe. Sometimes one half the body was 
 painted black, the rest being left uncolored. Occasionally a black 
 line like a mustache was drawn across the upper lip, and ex 
 tended to the ears, the rest of the face being red, and Neuwied 
 says that sometimes the sides of the body are blackened from the 
 shoulders to the feet, the middle of the body being without color. 
 The colors are usually prepared in the upper shell of a turtle, and 
 are carried in a joint of bamboo. 
 
 As Von Tschudi has well remarked, a naked Botocudo warrior, 
 with his black painted body, red face, and lip and ear ornaments, 
 must present a most demoniacal appearance. In the forest the 
 Botocudos go naked. Not a single Indian I saw wore any orna 
 ment on the head or body, unless it were a simple string of beads. 
 When at work on the fazendas they go as nearly naked as possible ; 
 the men often tie a shirt by the arms around the waist, letting the 
 body hang down in front; the women wear only a tattered skirt. 
 These garments are immediately discarded as soon as they return 
 to the forest, and one may see little bands entirely naked in the 
 vicinity of Philadelphia and other settlements in the Botocudo 
 region. Like other savage nations, the Botocudo shows no symp 
 tom of shame on exposing the person in the presence of those who 
 are dressed. 
 
 The ornaments of the wild Indian woman consist of a band 
 sometimes worn about the head, from which on one side depend 
 a number of long strings to the ends of which are attached hoofs 
 of capibaras ; of collars made of hard berry-like fruits strung on 
 threads, together with the teeth of monkeys, &c., or of strings of 
 the hoofs of wild pigs ; and of armlets of beads and teeth. The 
 chiefs sometimes ornament themselves with feathers, but this is 
 rare. 
 
 I never saw a savage Indian woman wearing a flower, though 
 the civilized Indians are very fond of them, as Mrs. Agassiz tells us 
 in the "Journey in Brazil." 
 
ON THE BOTOCUDOS. 591 
 
 The arms of the Botocudo consist of the bow and arrow ; clubs 
 are rarely used. The bow in ordinary use is about five feet in 
 length, and is made from the wood of the Airi-palm. This wood 
 is exceedingly hard, heavy and strong, and of a very dark reddish- 
 brown color. The bow is thickest in the middle, where it. is round, 
 and it tapers regularly to each end. So difficult are these bows 
 to bend, tha-t no one but an Indian can use them. Mr. Copeland 
 brought one from the Guandu with him, which not even our 
 strong canoe-men could bend.* These bows vary somewhat in 
 size, some being even seven feet in length. 
 
 The arrows are usually about six feet in length, and are made 
 of the Ubd, Cannachuba, &c., which furnish light and strong reed- 
 like stems. 
 
 The arrow used in war is tipped with a javelin-like head, five or 
 six inches long, and sometimes two inches broad, which is made of 
 a piece from the side of a joint of Bamboo, and is conseqiiently 
 convex on one side and concave on the other. This is cut into a 
 sort of elliptical shape, and sharpened to a long acute point. It 
 is then hardened in fire, and the arrow-head is prepared with 
 an edge as sharp as a knife. The head is Abound into the arrow- 
 shaft with bark. This arrow, which is employed not only in war, 
 but in the chase of the Tapir, is, like the other kinds in use among 
 these Indians, tipped with the feathers of some large bird, a single 
 feather being bound on each side. It makes a terrible wound, and 
 one particularly dangerous, because of the concave shape of the 
 arrow, which facilitates bleeding. 
 
 Another arrow in use sometimes in war, but usually in the 
 chase, is furnished with a thin, narrow head, about a foot long, with 
 backward projecting points cut on one side, a terrible weapon. 
 
 For birds and small game an arrow is used whose tip is made 
 
 * Speaking of the strength of the Indians of the vicinity of Rio, and of the 
 stiffness of their bows, Lery says, Cap. XIII. : " Si longitudine et crassitu- 
 dine nostros adeo superant, ut cos nee lentare nee adducere ullus nostrum 
 possit ; quin potius immo totis viribus puerorum decem annorum arcubus cur- 
 vandis opus esse." 
 
592 APPENDIX. 
 
 from a stem cut at the node where several little branches have 
 their origin in a circle ; the stem is fashioned into a blunt point just 
 above the node and the branches are cut off short. An arrow of 
 this kind, of course, does not tear but only bruises. I once en 
 countered several Indians near Urucu returning to their camp 
 from hunting small lizards with these blunt arrows. Guns are 
 not much in use among these savages, though they soon learn to 
 use them very expertly.* 
 
 The Botocudos use a sort of speaking-trumpet made of the skin 
 of the tail of the great armadillo (Dasypus gigas) to call one 
 another in the forest. 
 
 Before the discovery of America the Indians of Brazil, both 
 Tupi and Botocud o, used cutting instruments of stone of various 
 shapes, and at Linhares on the Rio Doce the subdelegado pre 
 sented me with a stone instrument, probably of Tupi origin, of 
 the shape of a saddler s cutting-knife. It was made from a very 
 hard gray stone. I unfortunately lost the specimen, so that I 
 am unable to figure it. I have never seen any of the stone imple 
 ments of the Botocudos. Their principal cutting instrument, be 
 sides their bamboo razors, consists of a common stout knife, like 
 a butcher-knife, though they make knives from hoop-iron, or any 
 thing else that will serve the purpose. This knife they always 
 carry slung over the back by a cord around the neck. 
 
 The nation of the Botocudos is divided into a number of little 
 tribus, as they are called, or collections of a larger or smaller 
 number of families, each tribe inhabiting a certain region in 
 which they have their Aldeamento or head-quarters. Each one of 
 these tribes is governed by a chief usually selected for his strength 
 and bravery, and the tribe frequently takes its name from the 
 leader. Thus a tribe in the Mucury region, headed by a chief of 
 Herculean strength named Pojichd^ goes by his name. Henrique 
 
 * Von Tschudi says that the Indians not only shoot more successfully, but 
 to a greater distance with their arrows, than the European can with his gun. 
 
 t A son of Pojicha was a servant in the house of Signor Gazinelli, in Santae 
 Clara. 
 
ON THE BOTOCUDOS. 593 
 
 Gerber * says that " they divide themselves into several tribes, 
 of which some are domesticated and gathered together into vil 
 lages ; others, still wild, wander through the forests of the valleys 
 of the Mucury, Doce, Tambacury, Urupuca, &c. All of them, how 
 ever, distinguish themselves disadvantageously from the Macha- 
 lalis and Malalis by the inferior degree of their intellectual facul 
 ties. The principal domesticated tribes are : 
 
 a. The Naknenuks (dwellers in the Serra), a confederation of 
 various tribes, who occupy the valleys of the upper Todos os 
 Santos, Pote, and Mucury, in the Aldeamentos of the Capitao 
 Felippe, in the forest of Sao Joao, of the Captain Pote, on the 
 margin of the brook Pote, of the Captain Timotheo, on the head 
 waters of the Todos os Santos, <fec. 
 
 b. The tribes of Pojicha, encamped three leagues below Phila 
 delphia. 
 
 c. The tribes of Giporok, on the margins of the Urucu and 
 Lower Mucury. 
 
 d. The Bakues, on the left margin of the Mucury. 
 
 e. The Aranaus on the margins of the Surubim and Sassuhy. 
 Von Tschudi f has given the distribution of the tribes with so 
 
 much detail and apparent precision that I quote what he has to 
 say on the subject : 
 
 " The nation of the Botocudos is broken up into a multitude of 
 tribes, of which the most divide themselves again into independent 
 hordes. On the head-waters of the Mucury and Todos os Santos 
 live the Naknenuks. From my researches into the meaning of 
 this word two entirely opposite explanations have presented them 
 selves to me. According to one version, the name should mean 
 Lords of the land, according to the other, Not from this land. 
 I am not in the position to determine which translation is the more 
 correct. | To the Naknenuks must be reckoned as belonging the 
 
 * Noqoes Geograficas, &c., pp. 24, 25. 
 
 t Reisen dureh Siid-Amerika, Vol. II. p. 264. 
 
 t See Gerber, quoted above. Von Martius translates the name " homines terrce." 
 
 LL 
 
594 APPENDIX. 
 
 Americanos d Agoa Branca on the Rio Preto, belonging to the 
 basin of the Jequitinhonha, where they possess a considerable 
 aldea. 
 
 " The Naknenuks of the Mucury consist of the following known 
 hordes, which are called after their chiefs : The horde of Pole, 
 probably the strongest of all, only two leagues distant from Phila 
 delphia, that of Cracatan, Braz, Poton, Timotheo, Inhome, Felipe, 
 Ninkate, and Nortete. The last, which formerly numbered over a 
 hundred bows fit for war, has lately melted away to only a few 
 families 
 
 " South of the Serra Mapmap-crak, which separates the basin, 
 live the Aranatis, the bitterest enemies of the Naknenuks, on the 
 Rio Aranau. From the northern tributary of the Mucury, the Rio 
 Panpan [Pampam or Pampao] to Santa Clara stretch the Bakues, 
 and west from these to near the shore the tribe of Urufu. On 
 the source of the Rio Preto are the bands of Joao Ima, Casimiro, 
 Maciel, and other subordinate chiefs, Jumerai, Capord, Ampaquejti. 
 In the southern basin of the Mucury we meet at the Rtbeirao de 
 Saudade with the tribe of the redoutable Captain Poschischd* some 
 what farther east of the Riberao das Lages, the tribus of Mekmek, 
 Shiporok, and Potik, and still eastward toward the coast, in the 
 basin of the Sao Matheos, the hordes of Pokorun, Batata, and 
 others. On the Rio Urucu, or the largest southern affluent of the 
 Mucury, the Shiporoks and the chiefs Juquirana and Maron. 
 
 " On the subject of the name Shiporok I have no more light 
 than on that of Naknenuk. According to some, Shiporok means 
 enemy, and with this name the Indians commonly designate their 
 adversaries. The tribe, however, does not name itself so. By 
 what name it designates itself I could not learn. An Indian sol 
 dier, and one excellently acquainted with the language of this 
 race, and of whom I made numerous inquiries concerning it, 
 
 * Von Tschudi, like a German, has mistaken a sound very near the French 
 j for an sh. Shiporok is pronounced zhlpffrff k, the sound zh representing a 
 sound somewhat intermediate between the Frencli j and the Spanish ch. 
 
ON THE BOTOCUDOS. 595 
 
 assured me that Shiporok meant from this side of the mountain? 
 or from behind the mountain, 1 and that this was the only name of 
 the race. In the language of the Botocudos, Shiporak means 
 brother, and Shiporok, arm. In the year 1816 Prince Max. zu 
 Neuwied met with a brave Botocudo chief, called Jeparak, at the 
 Quartel dos Arcos on the river Belmonte. It is not unlikely that 
 his baud went later southward, and settled in the basin of the Mu- 
 cury. About thirty years later the Shiporoks were at the Lagoa 
 d Arara, on the north bank of the Mucury, and about eight years 
 after we find them on the Rio Urucu, a southern tributary of this 
 stream. By the indistinct, often suppressed sounds of the vowels 
 in many words of the language of the Botocudos, I believe that 
 the mode of writing the word on the part of Prince Maximilian 
 is no hindrance to the opinion that the Indian race of the Jepa- 
 raks which he met with at Belmonte are identical with the Shi 
 poroks at Urucu. 
 
 " All the Indian bands on the basin of the Mucury, with the 
 exception of the Malalis, Machacalis, and perhaps the Aranaus, 
 belong to the race of the Botocudos or Engerakmung, as they call 
 themselves. Some bands have settled down in permanent dwell 
 ing-places, Aldeamentos or Aldeas, and we have especially to 
 notice the Aldeamento do Poton, the Aldo. do Pote, Aldo. do Cra- 
 catan, Aldo. de Curien9a (das Cursiumas), Aldo. do Nortete, Aldo. 
 de Sao Joao, Aldo. d Agoa Boa, Aldo. dos Aranaus. In the year 
 1817, out of these eight Aldeamentos one hundred and four indi 
 viduals were converted to Christianity (of course only in name) ; 
 among them the three chiefs, Pote, Poton, and Cracatan. By 
 the inhabitants of some of these villages some cxiltivation of the 
 ground is earned on, but it is confined chiefly to the cultivation of 
 maize and mandioca. The number of individuals in these bands 
 varies much. Some count several hundreds, others scarcely eighty 
 to a hundred, with only about twenty efficient fighting men. 
 From the number of the bow-bearing warriors one can with surety 
 estimate the number of souls -in a band, because this last, on the 
 
596 APPENDIX. 
 
 average, amounts to four times the number of warriors. I believe 
 that I am not far from the truth when I estimate the total num 
 ber of Indians in the basin of the Mucury at from twenty-eight 
 hundred to three thousand souls. 
 
 The Botocudos, when travelling in the forest, build for them 
 selves shelters of palm-leaves, which they stick in the ground in 
 a half-circle, the tips of the fronds arching together forming a 
 sort of roof. In travelling through the forest between the Mucury 
 and Peruhype, I saw great numbers of deserted ranches of this 
 kind in the forest. 
 
 Where they encamp long in one place they make their ranch 
 more substantial with a better roof, and often of sufficient size 
 to hold several families. The whole furnishing of the cabin of 
 a Botocudo is of the simplest possible kind. The fire is made 
 in the middle. Rarely ever are earthen pots used for cooking. 
 They make use of gourds and the cup-like receptacle of the 
 Sapucaia (Lecythis) for drinking purposes and for the preparation 
 of their food. Water they carry in the joints of the Taquara-assu, 
 in which they also keep their painting -materials. Beds are made 
 of estopa or bast-fibre. 
 
 The food of the Botocudos consists of sapucaia nuts, palmito 
 buds, and the fruits of the Inga, Jaboticaba, Ara9a or Goyaba, 
 Maracuja (passion-flower), &c., with the roots of Cipos and other 
 plants. They are fond of Indian corn, bananas, and mandioca, 
 which they steal whenever they can from the plantations. 
 
 They hunt game of all kinds, but they are particularly fond of 
 monkeys, whose flesh, as I can myself testify, is exceedingly savory. 
 They even eat the onga and other carnivores, the ant-eater, alli 
 gators and lizards, and the boa-constrictor. 
 
 Among birds they are particularly fond of the Mutum, the 
 Jacupemba, &c., and they also eat their eggs. Fish are usually 
 shot with small bows, which are used with great dexterity. Some 
 times they employ a poisonous root, which, put into the water 
 of a pool, kills the fish. 
 
ON THE BOTOCUDOS. 597 
 
 They are very fond of the great, fat larvse of certain insects 
 which burrow in decaying wood. Among these, according to Neu- 
 wied, is the larva of the Prionus cervicornis, which, with other 
 species, live in the trunks of the Bombax, or Barrigudo. Numbers 
 of these disgusting grubs are impaled on a sharp stick and toasted 
 at the fire. 
 
 Usually all animal food is cooked in this way. Bananas, pota 
 toes, &c., they sometimes bake in the hot ashes. Ants also are 
 eaten. 
 
 They are fond of honey, and formerly they used to cut down 
 hollow trees with stone axes to obtain it. To-day steel axes and 
 hatchets are occasionally to be seen among the Indians. 
 
 Fire is to the Botocudo an object of much care, because if it is 
 lost it is only to be rekindled with great difficulty. In order to 
 obtain it the Indian procures a stick of some light, dry wood, and 
 makes a small hollow in it. This stick he places on the ground and 
 holds securely with his foot. He then takes a long dry stick, one 
 end of which is somewhat blunted, and places it in the hollow 
 above mentioned. The other end is taken between the two palms, 
 the stick is held vertically, and by a rapid motion of the hands 
 it is caused to twirl until the friction of the lower end in the 
 hollow of the other stick has caused it to take fire, when estopa 
 or bast is ignited, and a fire is speedily made. This method of 
 procuring fire is also employed by some of the aborigines of North 
 America. We find the same custom in Africa among the Bush 
 men and Caffres,* and in the Aleutian and Caroline Islands. 
 
 The only things manufactured by the Botocudos consist of bows 
 and arrows, a few little ornaments, and bags made of the bast-fibre 
 of different plants. These last they barter with the whites for 
 food, &c. They bring in to the fazendas the wax of wild bees, 
 ipecacuanha, skins, &c., but this barter is conducted only on the 
 very smallest scale. 
 
 * Alberti, Descrip. Phys. et Hist, des Caffres, p. 36. Campbell, Reise in Sud- 
 Afrika, p. 37. 
 
598 APPENDIX. 
 
 The Botocudos take usually but one wife. Von Tschudi says 
 that when a man has chosen a woman for his wife, he agrees with 
 the father as to a certain tribute which he shall pay in game or 
 something else, when the woman is handed over to him and 
 with no further ceremony is thereafter his wife. 
 
 Neuwied says that a man may have as many wives as he can 
 take care of. Adultery is rare, and is visited with heavy punish 
 ment on the woman. The husbands are very cruel and unkind 
 to their wives. The husband, when angry with his spouse, beats 
 her unmercifully, and cuts her with his knife. I never saw a mar 
 ried woman who was not covered with scars, on her face, back, 
 breast, and arms ; it is the commonest thing to see them six 
 inches or more in length, and one woman may bear the marks of 
 many terrible wounds, which it seems marvellous she should have 
 survived. 
 
 The good health of savages, and the facility with which they 
 recover from injuries which would have proved fatal to an ordinary 
 civilized man, have been often commented upon by many authors. 
 Numerous instances are on record of Negroes, Malays, Pacific 
 Islanders, and American Indians who have survived horrible 
 wounds and mutilations, showing that Nature s power of healing is 
 greater among savage than among civilized nations.* 
 
 The woman is really the slave of the tmsband, and all hard work 
 falls to her lot. On the march she carries the family goods, or the 
 larger share, packed in a bag, which is slung on the back by a band 
 which passes over the forehead. The mother carries her child 
 on her back sitting in a loop of bark which passes over her 
 forehead, the child clasping her neck. 
 
 The children are kindly treated, at least when young, but the 
 tie between parent and child is not strong, f At Sao Matheos 
 
 * Waltz, Introduction to Anthropology, p. 126. 
 
 t The women carry their children on their backs sitting in a loop of bark 
 which passes over the mother s forehead, the child clasping the mother about 
 the neck. 
 
ON THE BOTOCUDOS. 599 
 
 there was on the fazenda a young woman who had two children ; 
 one, a boy several years old, was sick from dirt-eating, and was 
 stunted in growth, yellow, and swollen ; the other was a babe at 
 the breast. The mother was anxious to sell the elder, and I could 
 have bought him for a trifle. One day the babe suddenly died. 
 The mother immediately dug a grave for it in the floor of the 
 rancho, and went pleasantly about her work as usual, the only 
 effect on her being to make her determine not to sell the boy. 
 Children are frequently bartered away to the fazendeiros, who in 
 reality hold them as slaves. 
 
 Of the religious ideas of the Botocudos we have not so much 
 information as we could wish. Most writers agree with Neuwied 
 that the Botocudos believe in and fear a bad spirit called Janchon, 
 and Neuwied says that they recognize many of them, which they 
 distinguish as great and small. The great devil comes in the 
 guise of a black man visiting the camps ; sometimes he sleeps 
 awhile by the fire and then goes away, but all who see him die. 
 This same devil is accused of beating the dogs to death with sticks, 
 and of killing children. 
 
 The Tupi race seems to have some idea of a God, and they 
 called Him Tupa or Tupan, which name is derived from the 
 word tuba, father. It is a difficult matter to-day to arrive at 
 any clear knowledge of the primitive religious belief of the 
 Botocudos, because they have derived so many ideas through 
 intercourse with the Portuguese. I was unable to learn that 
 the Botocudos had any idea of a God. The moon, which they 
 call Tanru, is an object of fear, the Indians believing that occa 
 sionally it falls upon the earth, destroying men, and that it 
 sends storm and famine. No worship is offered to any of these 
 things, and they have neither priests nor medicine-men like the 
 
 Tu P is Bancroft Libiwy 
 
 The dead are buried in the immediate vicinity of the camp, 
 
 or even in the wigwam, and this last appears to be usually the 
 case. The child that died at Sao Matheos was buried in the 
 
600 APPENDIX. 
 
 earth forming the floor of the ranch in which the Indians were 
 quartered. When a death occurs they usually desert the camping- 
 ground for another, but I feel sure that I saw the Indians occupying 
 the same ranch in which the dead child was buried. The relatives 
 gather together and howl for one day after the death, the women 
 taking a specially prominent part in the ceremony ; but the next 
 day they all go about their work as if nothing had happened. 
 In some places the hands of the defunct are bound together before 
 burial. The corpse is buried in a horizontal position, and a fire is 
 lit to keep away the devil, for they believe that if this evil spirit 
 should find no fire at a grave he would dig out the body. Deserted 
 wigwams in the vicinity of graves are a common sight in the forest. 
 Sometimes a shelter of palm-leaves is built over the burial-place, 
 as Neuwied has remarked. Nothing is put in the grave with 
 the dead body.* 
 
 The different tribes or Aldeamentos are frequently at war with 
 one another. At the time of my visit to the Mucury there was a 
 skirmish between two bands near Cannas Brabas on the Phi^ 
 delphia, the particulars of which I was unable to learn. I had 
 sent my baggage on ahead of me to Philadelphia in care of 
 Signor Battista, who was conducting a train of ox-carts with 
 salt, &c. Near Cannas Brabas he was attacked by the Botocudos 
 and shot at, but escaped by plunging into the forest, from which 
 he emerged nearly naked. On arriving at Cannas Brabas we 
 found the place nearly deserted, but we passed the place of danger 
 in the night without seeing an Indian. 
 
 Their mode of warfare is the attack by night or from ambush. 
 A victory is celebrated by song and dance. I found the belief 
 everywhere current that they sometimes ate the bodies of the 
 slain, and from all the information I have received, I think 
 there can be no doubt that cannibalism is one of the customs 
 
 * The fcitor at Capitilo Grande told me that when a little child died it was the 
 custom to put by the grave a bottle of milk drawn from the mother s breast, 
 together with the bones of some wild animals. 
 
ON THE BOTOCUDOS. 601 
 
 of the Botocudos. At present the Indians are very peaceful, but 
 from time to time they have committed outrages on the -whites, 
 murdering, and burning houses on the Mucury, at Sao Matheos, 
 and on the Doce. 
 
 When at the Fazenda do Capitao Grande I witnessed one even 
 ing the dance of the Botocudos. A bottle of cachaga, or native 
 rum, had been given them, over which they were very merry. 
 Four of them took their position naked in the yard in front of 
 the house, and formed a square, facing one another, then all 
 four placed their left feet together, the right leg remaining 
 stretched out. All bowed their heads together and placed their 
 arms on each other s shoulders. Then they began a monotonous 
 song,* Caldul -d-a, Calani -a-ha , to which they all kept time by 
 hitching a step forward with the right foot, keeping the left quiet. 
 This dance they kept up for some time. They are apparently 
 very fond of it, and the overseer of the fazenda informed me 
 that they sometimes kept it up until, exhausted by fatigue, they 
 fell to the ground. 
 
 These Indians used to be very numerous at Sao Matheos, living 
 in the forests even quite close to the sea ; but they have been so 
 killed off that at present none are found in the vicinity except on 
 the head-waters of the river. They were hunted down by the Por 
 tuguese settlers like wild beasts, and one gentleman told me at -, 
 
 that during his life he had. either with his own hand or at his 
 command, been the means of putting to death by knife and gun 
 and poison over a thousand of these poor creatures ! The injuries 
 committed by the Botocudos on the whites are as nothing com 
 pared with the wrongs inflicted upon them by those who have 
 dispossessed them of their home, and have almost destroyed the 
 race. 
 
 The Indians have learned the use of rum and tobacco, of both of 
 which they are very fond, and which are rapidly working their ruin. 
 In the Mucury, Colonia Leopoldina, Sao Matheos, and on the Doce, 
 
 * St. Hilaire compares the song of the Botocudo to that of the Chinese. 
 VOL. I. 26 
 
602 APPENDIX. 
 
 the Indians come into the settlements to beg, and they not un- 
 frequently are employed to work on the fazendas, their service 
 being voluntary. At Sao Matheos they are paid in victuals and 
 rum ; but on the Mucury they have learned the use of money, 
 which they call pataca. At Sao Matheos I saw both men and 
 women at work with the negroes. They are not much to be de 
 pended upon, usually remaining but a few days on the fazeuda, and 
 then returning to their wild life in the forest. They are very lazy, 
 and half a dozen are scarcely worth an able-bodied negro. They 
 seemed very docile and good-natured ; indeed, I was particularly 
 struck with this last feature in their character. At their work they 
 laughed and played jokes on one another, and in the house at their 
 meals were quite as merry as the negroes. They have nothing of 
 the gravity, stolidity, and want of curiosity of our Northern In 
 dians. They have no idea of meum et tuum, and they are particu 
 larly addicted to stealing bananas, corn, or anything else they hap 
 pen to take a fancy to. They often come almost stark naked into 
 Philadelphia and Urucu to beg. 
 
 One may occasionally find a civilized Botocudo on a fazenda, 
 but the children sold by the parents and employed as servants on 
 .the plantations scarcely ever grow up. 
 
 At present the Botocudos are confined to the virgin forest be 
 tween the Rio Doce and the Rio Pardo. They are very rarely seen 
 near the coast, and never frequent the campos of the interior. The 
 race is fast diminishing, and in a few years will pass out of exist 
 ence. The Indians of the coast tribes of the Tupi race, thanks to 
 the labor of the Jesuits, have become civilized and converted to 
 Christianity, and now form an integral part of the Brazilian popu 
 lation, but the Botocudos resist civilization and the influence of 
 Christianity, and are sunk in the lowest barbarism. 
 
 The language of the Botocudos is entirely different from the 
 Tupi, and from that of the other coast tribes, as the Patachos, 
 Machalalis, &c. Although spoken by all the Botocudos, there are 
 dialectic differences observable in each band, different words 
 
ON THE BOTOCUDOS. 603 
 
 being sometimes used by different bands to distinguish the same 
 object. The language of the Naknenuks, Jiporoks, &c. is one 
 and the same. Latham s division of the Botocudo into the Boto- 
 cudo proper and Naknenuk is incorrect, and the differences ob 
 servable between his vocabularies are traceable to the collectors of 
 these vocabularies, who have spelled the words incorrectly, as they 
 caught the sound from the lip of the native. 
 
 At present we know very little of the grammatical structure of 
 the language, the vocabulary being very incomplete. 
 
 It is a very simple language, with very few or no inflexions. 
 The first words the stranger is likely to hear uttered by a native 
 namely, gling-gling bear one of the most distinguishing features 
 of the language. It is extremely rich in reduplicated words. 
 Thus we have Ton-ton = bad ; Kiacu-kack-kack = a butterfly ; 
 Ong-ong = to sing; Naak-naak = a gull; Encarang-cuong-cuong- 
 gipakiu = the great boa, &c., &c. 
 
 This reduplication appears rarely ever to extend itself to words 
 of two syllables. Instead of doubling a word of two syllables, only 
 the last syllable is reduplicated, as in the sentence, min-yan-yan- 
 rl-ma-kd-um = I am thirsty. These constantly recurring doubled 
 syllables give to the language a stuttering character. The Tupi- 
 Guarani is also rich in reduplicated words, but not so much so as 
 the Botocudo. 
 
 The principal points in the structure of the language thus far 
 noted by Neuwied and others are these : 
 
 There is but one gender, namely, neuter. There are two num 
 bers, a singular and a plural, and perhaps a dual. The plural is 
 formed by adding uruhti or ruM (many) to the singular, thus : 
 Kjiem = house ; Kjiem-iiruhu =. houses, also village ; Tyon = tree ; 
 Tyon-uruhu = trees or forest. In writing my vocabulary at Sao 
 Matheos, my interpreter, a native Botocudo who spoke Portu 
 guese quite well, gave me what appeared to be a dual form for 
 several words, and these I subjoin exactly as I noted them from 
 his lips. 
 
604 APPENDIX. 
 
 Man cuah-hah 
 
 Two men ni t-chS-vS 
 
 Woman p5-chl k 
 
 Two women c nit-cho-vo -d"n 
 
 Eye ki-t<5 m 
 
 Two eyes nlk-i-to m-ch5-vo" 
 
 Ear n5n-ho n 
 
 Two ears iion-hon-cho -vo" 
 
 Arm yl-mun 
 
 Two arms yi-mun-chS-vS 
 
 Hand i p-a 
 
 Two hands i p-a -cho -vo 
 
 Leg ip-mak 
 
 Two legs i p-mak-cho -vo 
 
 The word for man in this vocabulary corresponds with that 
 given by Neuwied, but in no vocabulary can I find the word po- 
 chik, meaning woman. The dual forms of both are strange, yet 
 my Botocudo insisted that nl t-cho-vo meant dous hometis, and nit- 
 cho-vo-Q n, duas mulheres. There may possibly be some mistake 
 here, but I have given the words in the hope that some one else 
 may be able to explain them. With the other words the dual was 
 formed by adding the termination -cho-vo, but in the case of Kitom 
 a prefix nl was added. 
 
 One thing is certain, the Botocudos cannot count. Their only 
 numeral adjectives being mokenam, which means, rather, single, and 
 uruhu, many.* The Botocudos at Capitao Grande kept an account 
 of their days work on their fingers and toes, and I was assured 
 that the largest number they could reckon was twenty. At the 
 expiration of ten days work, for instance, an Indian who wished to 
 settle with the feitor would go to him and count off templdn, day, 
 ten times on the fingers. In the examples above given there is 
 
 * The Tupis count only to three ; for higher numbers they use the Portu 
 guese. 
 
ON THE BOTOCUDOS. 605 
 
 a termination which signifies two, and which answers to a dual. 
 The subject is one of much interest, and I regret much that during 
 my stay in the Mucury I was unable to give it more study. Von 
 Tschudi gives numerals up to ten, which he obtained from a Nak- 
 nenuk through the aid of an Indian soldier ; but he himself ex 
 presses a doubt as to the veracity of his interpreter. 
 
 There are in the Botocudo language two cases, a nominative 
 and an objective. The latter is marked by the syllable te (ti or 
 de) between two substantives coming together and governing one 
 another (Gottling quoted by Neuwied) ; thus, Taru means moon, 
 but also sun, sky, and time ; Taru-ti-po, literally, sky-runner. 
 
 The adjective always follows the substantive, as cuidn-cudgi, the 
 small ant-eater. Neuwied says that the comparative is formed 
 by adding the termination uruhu, and the superlative by the addi 
 tion of the adverb gicaram, as cuang-mah, the stomach is empty, 
 cuang-mah-gicaram.* The only pronouns we know are KjicJc = I, 
 and Hd = he, she, or it. We have KjucTc for my, which, by the 
 way, may be used before the noun it qualifies, as Kjick-kjuck 
 magndn-joop, but Neuwied says that his Botocudo Quack used 
 Kjick as well as KjucJc. So far as the verb is concerned it seems 
 to be very simple in its construction, and to have only two forms ; 
 namely, infinitive and participle. The third person singular pres 
 ent is formed by prefixing to the verb he, het, or simply a, which 
 appear to be only different forms of the third person singular 
 present of the verb to be. We at present know too little of the 
 language to speak at all positively about other points in its con 
 struction. 
 
 Prince Max. zu Neuwied gives in the second volume of his 
 Seise a German-Botocudo vocabulary of several hundred words. 
 Latham (Elements of Comparative Philology, p. 509) gives four 
 short vocabularies, and Von Tschudi (Reisen durch Brasilien, Vol. 
 II. p. 288) has another short vocabulary. Other vocabularies have 
 
 * I should hardly call this a superlative. 
 
606 APFEXDIX. 
 
 been collated by Guido Marliere, Von Eschwege, Jomard, Renault, 
 Von Martins, St. Hilaire, D Orbigny, and others. Neuwied s is 
 prepared with great care, and he had the advantage of long inter 
 course with a lad, whom he retained in his employ. Von Tschudi s 
 was hastily written with the aid of an Indian interpreter, and the 
 sounds are very imperfectly represented, a soft ch or zli sound 
 being represented by the German sch, &c. I do not know where 
 Latham s vocabularies were collected. They are very inaccurate. 
 While at Sao Matheos I spent a long time with a young Botocudo, 
 who spoke Portuguese, and collected quite a vocabulary, using a 
 phonetic alphabet of my own, by which, with an ear accustomed 
 to the pronunciation of many languages, I think I have been able 
 to express very nearly the true pronunciation of the words. The 
 pronunciation is exceedingly indistinct, and the words are very 
 hard to catch. The stupidity of my pundit was discouraging, and 
 the labor of getting together the vocabulary was very great. 
 
 This vocabulary is too voluminous to be inserted in this volume. 
 I hope to publish it elsewhere. 
 
 The language is usually spoken on a high key, but in a weak 
 tone and rather rapidly. It is particularly rich in nasals, but has 
 neither gutturals nor sibilants. 
 
 The sounds of s and z do not occur in it, nor those of / and x. 
 I have observed the sound of v only in the dual termination 
 -cho-vo. In many cases it is impossible to distinguish between 
 I and r. 
 
INDEX. 
 
 A. 
 
 Abacaxi pine-apple, 431. 
 
 Abrolhos Islands, derivation ot name ; danger 
 in passing, 174 ; situation of, 175 ; geology 
 of, 175 - 179 ; depth of water in vicinity of, 
 192 ; coral reefs of, 192 - 11)9 ; navigable chan 
 nel inside of, 201. 
 
 Acantkastrrp.a Braziliensis Verr., 62 ; described, 
 195, 205. 
 
 Aqaricia Agaricites Edw. and Ilaime, 62, 195. 
 
 Agassiz, Prof. Louis, on glacier of Pacatuba, 
 469 ; on the structure of the Organ Moun 
 tains, 15 ; on the Amazonian sandstones, 487. 
 
 Agassiz, Madame, 469. 
 
 Agassiz, Mr. A. E. , on the former connection 
 between the Atlantic and Pacific across the 
 Isthmus of Darien, 393. 
 
 Agate, 531. 
 
 Agoas Marinhas, 151. 
 
 Agrestes, name given to Western Sergipe, 379. 
 
 Aimore s or Aimbore s, 57". 
 
 Airi palm (Astrocaryum), 94, 216, 298 
 
 Alabaster, 298. 
 
 Alagoas, general description of province of, 
 422. 
 
 Alcobaca, 224. 
 
 Aldea Velha, 85. 
 
 Allen, Mr. J. A., 275, 295, 301 ; notes on the 
 geological character of the country between 
 Chique-Chique and Bahia, 309-318 ; on salt 
 deposit of Sao Francisco Valley, 331. 
 
 Alluvial lands on Kio Pardo, fertility of, 243 ; 
 not proper for coffee, 245. 
 
 Alto dos Bois, excursion to, 142. 
 
 Alum, 471. 
 
 Amazonas, derivation of name, 484. 
 
 Amazonian formation described by Professor 
 Agassiz and Major Coutinho, 488 ; Prof. Ag- 
 assiz s theory of the, 490 ; probably marine 
 and tertiary, 491 ; of Para, 491. 
 
 Amazonian group, cretaceous, 556. 
 
 Amethysts, 471, 530. 
 
 Amianthus, 471. 
 
 Ammonites, 384. 
 
 Ammonites acutocarinatits Shumard, 389; A. 
 Gibbonianus Lea, 389, A. Hallii Meek nnd 
 Hayden, 388 ; A. Peruvians Von Buch, 389 ; 
 A. semistriatus D Orb., 389. 
 
 Awpullaria, eggs of, 47. 
 
 Amygdaloid, 530. 
 
 Anacardiitm occirientale, 465. 
 
 Ananchytes, occurrence of, on Isthmus of Pan 
 ama, 392. 
 
 Anchieta, letter on Manati, 75. 
 
 Angico, bark of, used in tanning, 411. 
 
 Aninga, 395. 
 
 Anta (Tapirvs Americanus), 94. 
 
 Antedon Dubenii or A. Braziliensis, 62. 
 
 Ants a pest on the Una do Governador, 5. 
 
 Ant-hills, 82, 375. 
 
 Antimony, 448. 
 
 Aplysia Argo D Orb, 204, 214, 230. 
 
 Aplysia Braziliensis Sander Bang, 204. 
 
 Aquitequediehagas, custom of piercing the 
 
 lip and ears among. 584. 
 Aracaju (city), 381, 383. 
 Aracare, fossiliferous shales, &c. of, 396. 
 Aracaty, 460 ; country in vicinity of, 461. 
 Aranaiis, tribe of, 593. 
 Aranha caranguejeira (Mygale) of Abrolhos, 
 
 180. 
 
 Arara (Psittacus macoa), 95. 
 Araraitaguaba ( Arara macaw, ita stone, guaba 
 
 eat, locality so called from a stone which the 
 
 parrots are said to eat), 516. 
 Aratii land crab, 239. 
 Ankles gold-mine, 545. 
 Aricuri palm ( Cocos schizophylla), 237. 
 Armacjio, 40. 
 Armacilo, locality, 40 ; a trying-house, 182, 183, 
 
 184. 
 
 Arms of Botocudos, 590. 
 Aroeira (Sc/iinus terebinthifolius), 423. 
 Arraial da Chapada, gold-washings of, 497 ; 
 
 their present condition, 161 ; da Conceicjao, 
 
 498 ; da Conquista, 255 ; d ltinga ( Yg or ky 
 
 water, and tinga white), 165 ; das Queima- 
 
 das, want of rain, 323 ; de Sao Oon^alo, 475. 
 Arroio dos Ratos, coal-mines of, 520. 
 Arrows of Botocudos, 591. 
 Arsenical pyrites, 540. 
 Article, definite, use of before names of places 
 
 in Brazil, ix ; use of before name Brazil, viii. 
 Artorarpus Braziliensis Gom., 245. 
 As Azeites, 115. 
 
 Asplflorfiynr.hus Comptoni, 467, 468. 
 Assahi (Euterpe oleracea), 425. 
 Assii (town), 452. 
 Astrocaryum Airi Mart., 216. 
 As Trovoadas, 144. 
 Ateles, 286. 
 
 Atfles hypnxanthus, 95. 
 Attalea palm, 144 
 Augite, 297. 
 Ave-Lallemant, Dr., on coal-mines of Rio 
 
 Grande do Sul, 520. 
 Avicennia, social plant, 256. 
 
 1!. 
 
 Bahia, description of point on which city is 
 built, 333 ; description of city, 334 ; climate 
 of, 337 ; population, 335 ; bay and harbor of, 
 335 ; exports of, 335, 336 ; steamship con 
 nections, 337 ; Bahia de Todos os Santos, 
 267 ; compared with Bay of Rio, 272. 
 
608 
 
 INDEX. 
 
 Bahia Steam Navigation Co., 337, 421. 
 
 Bahia and Sao Francisco Railroad, 338 ; geol 
 ogy of, 353. 
 
 Bahia de Camamu described, 262. 
 
 Bahia de Espirito Santo described, 65 ; depth 
 of, 72 ; littoral fauna of, 73 
 
 Bahia de Nova Almeida, 85. 
 
 Bahia de Santa Cruz, 85. 
 
 Bahian group, 550. 
 
 Buhii, isolated hills or gneiss plain at, 303. 
 
 Bakues, tribe of, 593. 
 
 Bamboo, 84, 141, 216. 
 
 Barauua (Melauoxylon), ashes used in tan 
 ning, 411. 
 
 Barbacena, 4. 
 
 Barbacena, Visconde de, 519. 
 
 Barlwus, 432 ; on mouth of Rio Parahyba do 
 Norte, 442. 
 
 Barnacles, 73. 
 
 Barra do jardim, 465. 
 
 Barra do Commandatuba, colony at, 246. 
 
 Barra do Rio de Contas, 260. 
 
 Barra do Patipe, 241. 
 
 Barra do Poxim, 241. 
 
 Barra do Rio Grande, width of Sao Francisco 
 valley below, 332. 
 
 Barra do Rio de Sao Francisco, 393. 
 
 Barra da Vareda, country in vicinity of, 252. 
 
 Barra Secca, 106 ; lands at, 107 ; consolidated 
 beach, 107. 
 
 Barreiras do Siri, tertiary beds of, 56. 
 
 Barrel quartz at Minas Novas, 157. 
 
 Barrigudo trees, 91, 166, 251. 
 
 Barro vermelho, 514. 
 
 Basin of Rio Parnahyba one sided, 474. 
 
 Bateia, wooden pan for washing gold and dia 
 monds, 160. 
 
 Bates s "The Naturalist on the Amazonas," 
 494. 
 
 Beach ridges, 113, 114, 115, 123 ; formation of, 
 due sometimes to storms, 220. 
 
 Beach, raised, at Santos, 507. 
 
 Beaches, solidification of, at Bahia, 342 ; at 
 Rio Vermelho, 344 ; at the Abrolhos, 179. 
 
 Beaches solidified, 189. 
 
 Baturite, 460. 
 
 Bats, bones of, in caves at Lagoa Santa, 285. 
 
 Bauhinia, 250. 
 
 Bay of Rio, description of, 6 ; depth, 7 ; shell 
 deposits of, 7 ; corals of, 7 ; tide of, 7, 8 ; 
 saltness of, 8 ; entrance to, 8 ; once filled 
 with tertiary beds, 22. 
 
 Beans, 120, 454. 
 
 Beira-mar, 459. 
 
 Belmonte, 173 ; sand plain near, how formed, 
 235. 
 
 Bemdego meteorolite, 325. 
 
 Benevente, harbor of, &c., 60. 
 
 Bignonia Braziliensis, 93 ; B. Tecoma, 408 ; 
 species of, social plant, 256. 
 
 Biotite, 447 
 
 Birds, bones of, in caves at Lagoa Santa, 285. 
 
 Bismuth, occurrence of in auriferous vein at 
 Cata Branca, 545. 
 
 Black water streams, 122, 217, 227. 
 
 Blumenbach on skull of Botocudo, 587. 
 
 Boassica, cretaceous sandstone at, derivation 
 of name, 398. 
 
 Boa Vista, height of pass of, 12, 26. 
 
 Bolerio, 141. 
 
 Bom Fim, cretaceous strata of, 346, 
 
 Bone of whale valuable for manure, 184 
 
 Bones, fossil, in caves, mode of occurrence of, 
 284 ; immense number of recent species, 285. 
 
 Bone caverns, extent of, how formed, stalac 
 tites and clay deposits in, 283 ; number of, 
 284 : of Rio das Velhas, 280. 
 
 Bodoijue, 578. 
 
 Bore or Pororoca, 486. 
 
 skulls of, 585 ; color of, 588 ; dislike of cloth- 
 
 religious ideas of, 599 ; funeral customs, 599 ; 
 want of numerals, 604 ; warfare of, 600; lan 
 guage of, 602 ; Botocjue, or lip and ear orna 
 ment, 578, 583. 
 
 Boulders, scarcity of in drift, 24 ; occurrence 
 of, at Tjjuca, 28 - 30 ; paucity of, owing to 
 decomposition, 573 ; boulders of decomposi 
 tion, 594 ; at Victoria, how formed, 69. 
 
 Bows of Botocudo, 591. 
 
 Bra(jo do Norte of River Sao Matheos, 117. 
 
 BKHJO do Sul of Rio Sao Matheos, 117 
 
 Brachiopods, palaeozoic, at first cachoeira of 
 the Rio Tapajos, 488- 
 
 Brackish streams, 255. 
 
 Bradypus trirlactylun, 94 
 
 Brailyjnts torrjuatus, 94. 
 
 Braganza, waste of land on bay of, 491. 
 
 Brazil, derivation of name, x; use of article 
 before the word, viii. 
 
 Bromeliaceous plants, 249, 250 
 
 Brunet, Dr., 272, 421 
 
 Bugres, 578. 
 
 liuriti palm (Mauritia rinifera), 277. 
 
 Burning of forest, injurious effects of, 78 ; 
 change of flora caused by, 423. 
 
 Burton, 276 ; on use of definite article before 
 word Brazil, viii; on Jacutinga, 535: on 
 occurrence of cocoa palm at Brejo do Saiga- 
 do, 119. 
 
 Byrsonema, 465. 
 
 Cabbage palm, 94. 
 
 Cabral, discovery of Brazil, 226. 
 
 Cacao ( Theobroma Cacao), 120, 236, 244. 259, 
 
 260, 261. 
 
 Cacaoeiros, cacao plantations, 244. 
 Cachoeira (Town), 272 
 
 Cachoeira (Bahia), county in vicinity of, 322. 
 Cachoeiras, 21. 
 Caco, sugary quartz, 534. 
 Cacti, 237 ; arborescent, 152. 
 Cngel6 whale, 182. 
 Caiauh6 ( Elce.is Guineensis L.), 270 
 Caititu, 94. 
 
 Cajii or Cashew, 459, 465. 
 Cajueira (Anac.ardium occidental), 116. 
 Caladium, 249. 
 Catamites, 525 
 
 Calamopleurus cylindricus, 468. 
 Caldeiroes, 314 ; of probable glacial origin, 315 ; 
 
 mastodon remains at, 325 ; occurrence of 
 
 diamonds in, 307. 
 Calhao, 152 
 Callit/irix, 95, 286. 
 Camamu, geology of, 262. 
 Camassari , geology of vicinity of, 361. 
 
INDEX. 
 
 609 
 
 Campinas, plains of, 506V 
 
 Canipo, section at, 365. 
 
 Campos, city of, 46 ; vicinity of, 47. 
 
 Campos, character of their fiorae largely due 
 to fires, 320. 
 
 Cauipos of Alto dos Bois, 146 
 
 Campos, Liais s picture of, 290. 
 
 Canal uniting Macah6 with Campos, 43 ; be 
 tween Cannavieiras and Commandatuba, 
 246 ; between Rio Doce and Victoria im 
 practicable, 80 ; connecting Kio Jecii with 
 Victoria, 64 ; projected along coast south of 
 Caravellas, 223. 
 
 Candles of Caruahuba wax, 453. 
 
 Candona Candida Miill., 348. 
 
 Canga, gold in, 536, 543. 
 
 Cannibalism among Botocudos, 600. 
 
 Cansanijao (Jatropka urens), 250. 
 
 Cantagallo, old gold-mines at, 50. 
 
 Canon of the Jequitinhonha, 163. 
 
 Cape Frio, 39 ; height of, 3 J. 
 
 Capibara (Hydrocharus Capabara). 94. 
 
 Capibara, fossil (Hydrochtzrus sulcidens), 287. 
 
 Capoes, 146, 147. 
 
 Carapato (Ixodes ricinus), 153, 155, 257 
 
 Carapina, 82. 
 
 Caravellas, 223. 
 
 Carboniferous in Brazil, 553 ; fossils, occur 
 rence of, on Guapor6 and Momorg, 488. 
 
 Cariocar, 291, 465. 
 
 Carnadeiras, fever of Upper Sao Francisco, 292. 
 
 Caruahuba palm (Copernicia cerifera), de 
 scription of, 452, 446, 459, 460, 401 ; wax of, 
 453. 
 
 Carnivora more abundant in Post Tertiary 
 than at present, 288. 
 
 Carnelian, 531 
 
 Carrasco, 147, 253, 429. 
 
 Cascalho, meaning of word, 51 ; auriferous, 
 of Minas Nova region, 159 ; auriferous, at 
 Jaragua, 512. 
 
 Caseate Grande, 27. 
 
 Cassis Marlagascarifnsis (C. Cameo), at Os 
 Busies, 39, 198, 203 
 
 Castelnau, 496, 501, 504 ; on chapada diaman- 
 tina, 307 ; on whale fishery at Baliia, 185. 
 
 Cata Branca gold-mines, 544 ; auriferous vein 
 at , 545. 
 
 Catinga de Porco (tree), ashes used in tanning, 
 411. 
 
 Catiirgas, 144, 151, 250, 251, 296, 322, 452. 
 
 Cattle and cattle fazendas in Province of Rio 
 Grande do Norte, 454. 
 
 Cavia ( Ccr/ngenys rupestrit), 257 ; bones of, in 
 caves, 285. 
 
 Caximba Nova, 446. 
 
 Caxoeira do Angelim, 168 ; do Campo, bone 
 caves at, 281, 285 ; de Dorma, 167 ; da Fa- 
 rinha, 168 ; do Inferno, 88, 169. 
 
 Caxoeirinha do Jequitinhonha, 171. 
 
 Cachoeira de Paulo Affonso, distance from the 
 sea, llnlfeld s description of, 414 ; Liais on. 
 415, 418 ; compared with those of Salto 
 Grande on the Jequitinhonha, 417 ; Burton 
 on, 418 ; compared with Niagara, 418. 
 
 Caxoeirinha do Rio Pardo, 242 
 
 Caxoeira de Santa Anna, 169. 
 
 Cazal, 226, 430, 486 , on mastodon remains, 
 261 ; on fossil bones, 280 
 
 Cebus, 95. 
 
 Cecropin, 94 : social plant, 256. 
 
 Cedro cedrela, 94, 261, 459. 
 
 26* 
 
 Cemeterio of the Abrolhos, 180. 
 
 Centipede, 375. 
 
 Ceratites Harttii, 386 ; C. (Amm.) Pierdenalis 
 Von Buch, 388. 
 
 Cereus, 250, 296, 407, 423. 
 
 Cerithium, 229. 
 
 Cervus, 95. 
 
 Cereus rufus. fossil with megatherium, &c , 
 286. 
 
 Cervus simplicicornis found with megatheri 
 um, &c., 286. 
 
 Cesalpina echiitata, 94 
 
 Chalcedony, 531. 
 
 C/tama, 229 
 
 Chandless, 500, 503; discovery of Mosasauri- 
 au and turtle remains on the Purus, 494 
 
 Chapada, meaning of term, 132; at Agua da 
 Nova, 149, 150 ; magnificent view from edge 
 of, 150. 
 
 Chapada Diamantina, geology of, 295 ; height 
 of, 301 ; da Mangabeira, 497 ; at Santa Rita, 
 144 ; do Sincori, 306 ; between Sucuriu and 
 Agua Suja, 153. 
 
 Chapadas of Piauhy, 474 ; of Rio Jequitin 
 honha, 165, 168 ; of Rio de Sao Francisco, 
 277, 291 ; between the basins of the Toean- 
 tins and Sao Francisco, 499. 
 
 Chapadao de Santa Maria, 499 , do Urucuia. 
 499. 
 
 Chapeiroes, 211, 212, 213 ; dangerous to navi 
 gation, 201 ; distribution of, 211 ; mode of 
 growth of, 199; coalescence to form reefs, 
 200 ; appearance of, 200 
 
 Charruas, custom of piercing lip and ears, 584 
 
 Chelonia Mi/das, 112, 113. 
 
 Chiefs, Botocudo, 592. 
 
 Chique-Cbique, cactus, 408: town, origin of 
 name, 309 ; occurrence of diamonds near, 
 307. 
 C. ilamydotherium , 286. 
 
 Chlorite, 328. 
 
 Chloiite slate, 299. 
 
 Chrome localities in Goyaz, 503. 
 
 Ci ilar is, 386. 
 
 C ladocyclus Garrlneri, 468. 
 
 Claussen, Mr., 281. 
 
 Clays of Amazonian formation, 488. 
 
 Clays of Tertiary, 57, 124, 225. 
 
 Clay-eating among Botocudos and Brazilians, 
 581. 
 
 Clay-slate, 151, 298, 328, 464, 500 ; auriferous, 
 distribution of, 552. 
 
 Clay-slates, auriferous, of Brazil, resemblance 
 borne to the gold-bearing rocks of Nova 
 Scotia, 551 ; probably Lower Silurian, 551 ; 
 auriferous veins in, 533 ; absence of, in eozoic 
 of Brazil, 549. 
 
 Climate of campos, 253 . of Ceara, 458 ; moist- 
 ness of, dependent on forests, 321 ; of Co- 
 Ionia Leopoldina, unhealthiness of, 217; 
 of Fernando de Noronha, 439; of Maran- 
 hao, 485 ; of Rio Grande do Norte, 454 ; 
 of Sao Francisco below falls, 420 ; of Per- 
 nambuco, 429, of Rio Doce, 98; of Rio 
 Grande do Norte, 441 ; of Lower Sao Fran 
 cisco, 420. 
 
 Coal, Candiota, report on, by Mr. Ginty, 628 , 
 facilities for mining in Rio Grande do Sul, 
 526 ; cost of, 527 ; said to occur on Rio 
 Piauhy, 380 
 
 Coal Basin on Rio Jacuahy, 530 ; on Rio Sao 
 Sepe, 529 ; of Kio Tubarao, 519 
 
G10 
 
 INDEX. 
 
 Coal-mines of Rio Grande do Sul, discovery 
 of, 619. 
 
 Cochineal, 408, 454. 
 
 Cocoa palm ( Cocos nucifera), distribution of, 
 &c., 118, ^38, 271, 431, 437 ; oil of, 238. 
 
 Cocos coronata Mart., 298 ; C.Jiexuosa, 146 ; 
 C. nucifera, distribution of, &c., 118; C. 
 schizophylla, 237 , 298. 
 
 Cotfee, 79, 101, 123, 244, 261, 265, 459 ; export 
 of, from B.ihia for year 1864-65, 336 ; plan 
 tation of Campinas, 514 ; of the Colouia Leo- 
 poldina, excellence of, 217. 
 
 Colonia do Guandii, 90; Leopoldina, 217: do 
 Mucury, history and present state of, 131 ; 
 de Santa Leopoldina, 78 ; poverty of soil, 
 79 : de Santa Isabel, 65 ; de Santa Maria, 76 ; 
 de fanta Theresa, 518 ; do Uruc-ii, produc 
 tions, climate, scenery, &c., 133. 
 
 Colony, American, on Jeuuitinhonha, 172 ; on 
 Rio Doce, 105. 
 
 Color of skin lighter in forest-dwelling tribes, 
 588. 
 
 Comatula, 62. 
 
 Commercio, 306. 
 
 Compass, miner s. 530. 
 
 Conglomerate, 242, 301, 462 ; cretaceous, 354 ; 
 tertiary at Agoa da Nova, 150. 
 
 Congonhas do Campo, gold at, 544. 
 
 Conocarpus, social plant, 256. 
 
 Consolidated beach at As Pedras, 113 : at Bar- 
 ra Secca, 107 ; at mouth of Rio Parahyba 
 do Norte, 442 ; at Guarapary, 62 ; of Per- 
 nambuco, 434 ; at the islands of the Abrol 
 hos, 179. 
 
 Cn/iaifera offleinoKs. 93. 
 
 Copeland, Mr., 142,242. 
 
 Copocabana, gneiss of, 10. 
 
 Copper, 472 ; boulder, found near Cachoeira, 
 300 ; localities for, in Brazil, 300 ; sulphate 
 and other salts of, 488. 
 
 Corals at Itamaracu, 437 ; dead, near Rio Sant 
 Antonio, 234: distribution of on submerged 
 border of Recife do Lixo, 210 ; used for man 
 ufacturing into lime, 213. 
 
 Coral banks near Ilha Itaparica, 269 ; in Bay 
 of Caniamii,265 ; dead, in Lagoauear llheos, 
 259. 
 
 Coral islands, 211, 212. 
 
 Coral reefs, height of edge of, 209 ; in vicinity 
 of Pernambuco, 214; near Camamii, 213; 
 around Quieppe Island, 213; of Abrolhos, 
 187 ; of Pernambuco, 188 ; ofMaceio, 188,213, 
 425 ; near Cape Sao Roque, 188 : of Roccas, 
 214; off Periperi,213; of Santa Barbara dos 
 Abrolhos, 192 ; species that contribute most 
 to build up, 214 ; raised border to, composed 
 of barnacles, serpulse, &c., 205 ; submerged 
 border of, 205. 
 
 Corcovado, description of, 9 ; magnificence of 
 view from, 11. 
 
 Corn, 152. 
 
 Coroa Vermelha, 211. 
 Cnronula, 112. 
 
 Corrego de Santo Ignacio, diamonds in, 307. 
 
 Cotigipe, cretaceous rocks at, 359. 
 Cotinguiban group 556. 
 
 Cotton, 225, 336. 441, 442, 443, 459, 486; of 
 Calhao, excellent quality of, 152 ; of Per 
 nambuco, quality of, 431 ; of Rio Grande do 
 Norte, 452 ; value of, exported from Bahia 
 during year 1864 - 65, 336 ; factory, 206. 
 Crabs, burrows made by, in shales, 353. 
 
 Cratheus, 457. 
 Crato, 460, 463. 
 Crux, 95. 
 
 Cretaceous, of Abrolhos, 175 ; of Monserrate 
 and vicinity of Bahia, 346 ; beds at Bahia 
 
 IIBUUO ui v^ctira, 
 
 Resume, 555. 
 Crocodilus Harttii Marsh. 355. 
 Crustaceans, abundance of, on reefs ; list of 
 
 species occurring on the Abrolhos reefs, 
 
 203 ; list of species occurring elsewhere in 
 
 Brazil, 203. 
 Crijsoberyls, 151, 164. 
 Cuiaba, 500 ; diamonds at, 504. 
 Curimataes (fish), 399. 
 Curitiba, 517- 
 Cuscuta, 375. 
 Cutia (Dasyprocta), 94. 
 Cypra-a exanthema, 40. 
 Cypris(!)AUportianaJonef. 348 ; C. concukata 
 
 Jones, 348 j C. (?) Monserratensis Jones, 348. 
 
 1). 
 
 Dance of Botocudos, 601. 
 
 Darwin on phosphate incrustation of Abrol 
 hos, 178 ; on corals of Brazil, 187 ; on Island 
 of 1 ernaudo de Noronha, 437. 
 
 Date palm, 425. 
 
 Davis, Dr. J. Barnard, on skull of Botocudo, 
 585. 
 
 D Archiac, carboniferous rocks in Bolivia, 554. 
 
 De Beaumont, Elie, on the gneiss of the Serra 
 do Mar, 548. 
 
 Decomposition of gneiss, 24 ; Darwin, Agassiz, 
 Pissis, Heusser, and Claraz on, 24 ; cause 
 of, 25, 26. 
 
 Decomposition of rocks in India and North 
 America, 26. 
 
 Decomposition, drift not referable to, 563. 
 
 Decomposed rock separate between drift and 
 easily distinguished, 564. 
 
 Deer of Campos, 147. 
 
 Dende (Elrris Guineensis), 270, 425- 
 
 Denudation, subaerial, effects of, 33 
 
 D Orbigny on gneisses of Bolivia, 548. 
 
 D Oery, on Cata Branca mine, 546. 
 
 Desterro, 519. 
 
 Devonian rocks in Brazil, 553. 
 
 Diamantiferous rock from chapada, 304. 
 
 Diamantino, 500, 504. 
 
 Diauiantina, diamonds at, 156. 
 
 Diamonds, discovery of, at Diamantina, 156 ; 
 mode of occurrence of, at Chapada Diaman 
 tina, 302 -.309; yield of, at, 308; of Sin- 
 cora, 306 ; of Pe tanga, Bahia, mode of oc 
 currence of and method of washing, 369 ; 
 probable source of, 370 ; localities for in 
 Goyaz, 503; occurrence of in Province of 
 Parana, 518. 
 
 Dirtelphys, 94. 
 
 Dinosaurian remains, cretaceous, Bahia, 356 
 
 Diorite, 296. 
 
 Diplorta, absence of, in Brazil, 214. 
 
 Diplntliemium tnaritimum, 237. 
 
 Disthene, 325. 
 
INDEX. 
 
 Gil 
 
 Dolphin, species of, found in the Araguaya, 
 502. 
 
 Doni Pedro II., the Emperor, geological obser 
 vations of, 35. 
 
 Dom Pedro II. Kailroad, geological observa 
 tions on, 14 
 
 Dom Pedro, North d El-Rei Company, 542. 
 
 Drawings, Indian, 297, 326. 
 
 Drift on Cautugallo Railroad, 20 ; of Rio, 23, 
 561 ; at Tyuca, 27 ; of Victoria, 69 ; on Phila 
 delphia road, 132 ; at Alto dosBois, 145 ; near 
 Fazenda da Lagoa, 148 ; near Calhio, 151 ; 
 of Miuas Nova and vicinity, mode of occur 
 rence of gold in, 160 ; at Bahia, 341 ; once 
 continuous down to tbe surface of the sea, 
 342 ; on Bahia Railroad, 366 ; on Lower Sao 
 Francisco. 419 ; of C eara, 469 ; on Sao Paulo 
 Railroad, 508 ; of Patagonia, 558 ; Resume 
 of arguments to prove existence of in Brazil, 
 558 ; distribution of in Brazil, 561 ; clays of, 
 totally unlike those of the tertiary, 568 ; for 
 mer extension down to level of sea, 572. 
 
 Drought in Ceara, 459. 
 
 Dry season, abundance of life during, 155. 
 
 Dual form in Botocudo language, 604. 
 
 Dunes, 124, 332, 345, 380, 382, 394 ; in Province 
 of Rio Grande do Norte, 455 ; Ceara, 458. 
 
 E. 
 
 Ears, perforation of, by Potocudos, how per 
 formed, 582. 
 
 Ear-plug, 583. 
 
 Echinaster crassispina, 62. 
 
 Ecliinometra Mic/ielini Desor, 62, 203, 214 ; 
 nests of, in rock at island of Maric.is, 36. 
 
 Edwards, Prof. A. II., on turba of Camamu, 
 263. 
 
 Eggs of loggerhead turtle, 110. 
 
 Elephants in Post Tertiary in Brazil, 288. 
 
 Elevation of coast within recent times, evidence 
 of, at Rio, 35. 
 
 Embaliba ( Cecropia), 94. 
 
 Ema (Rkea Americana), 146, 254. 
 
 Encope emarginatus, occurrence of, at Rio 
 Sant Antonio, 235. 
 
 Engeraokmung, 577. 
 
 Entomnstraca, cretaceous, 347, 348. 
 
 Entre Montes (village), 411. 
 
 Eozoic rocks, occurrence of gold in, 532 ; dis 
 tribution of, in Brazil, 547 ; of Venezuela 
 and Guiana, 550. 
 
 Eretmochelys imbricata, notes on, 112. 
 
 Eriocaulon, 423. 
 
 Erosion, topographical features produced by, 
 in moist and wooded region, 318. 
 
 Erratics, 489. 
 
 Escadinhas, Cachoeira das, 89. 
 
 Estreito on .Joquitinhonha, 166. 
 
 Estaneia, Geology of vicinity of, 379. 
 
 Estherians, 445. 
 
 Eugenia, 249. 
 
 Eunieea humilis Edw. and Haime, 62, 74, 
 196, 229. 
 
 Exports of city of Bahia during year 1864-65, 
 336, 337. 
 
 F. 
 
 Falls of Paulo Affonso, 414 : of Jequitinhonha, 
 170. 
 
 Farinha, 120 ; exportation of, from Rio de Con- 
 
 tas, 261. 
 Farrancho, 167. 
 
 Faults, apparent absence of, 448. 
 Fauna, marine, of Brazil, resemblance to that 
 
 of West Indies, 198. 
 Faunas, comparison of Brazilian radiate fauna 
 
 with the West Indian, 214. 
 Favia conferta Verr.,194; F. gravida Verr., 
 
 194 ; F. leptophylla, 207. 
 Feira da Conceicao, couutry in vicinity of, 322. 
 Felis concolor, 95, 286 ; F. macroura, 95 ; F. 
 
 oMfa, 95 ; F. pardalis, 95 ; jF. mitis, 286 ; 
 
 F. proto/iunther, 287. 
 
 Ferns, fossil, of Caudiota coal-mines, 522. 
 Fazenda do Capitao Grande, 117 ; do Tenente 
 
 Ilonorio Ottoni, drift in vicinity of, 148 ; at 
 
 Escravania, immense number of bones in 
 
 cave at, 285. 
 Feijao (beaus), 79. 
 Feijo on structure of Serra da Ybiapaba, 457 ; 
 
 on gold of Ceara, 472. 
 Fevers of Rio Doce, t-8 ; of Sao Matheos, 118 ; 
 
 on the Jequitijihouha, cause of, 169. 
 Fiddle crabs, 230. 
 
 Fires, effect of, on vegetation, 83, 320, 375. 
 Fish, excellence of Brazilian, 186; fossil, 347, 
 
 353, 354, 355, 359, 371, 384, 445, 466, 477. 
 Fisheries of Esyirito Santo, richness of, 81. 
 Fishery of Garoupa in Abrolhos region, 185. 
 Fleming! tes Pedroanus Carruthers, 524. 
 Fletcher, Rev. J. C.,514. 
 Folds, reversed, in Eozoic, 550. 
 Forbes on Siluria of Bolivia and Peru, 552. 
 Forest, luxuriance of, on Rio Doce, 93 ; on Rio 
 
 Pardo, 245. 
 Forests, belt of, coincidence with belt of rains, 
 
 319 ; losing their hold, giving way to cam- 
 
 pos, 319; effect of fires, 320 ; decomposition 
 
 dependent on forests, 320. 
 Fortaleza (city), location and population of, 
 
 460. 
 Fossils (foreign), carried in ballast, necessity 
 
 of caution, 269. 
 
 Fossils of Siluria in Bolivia and Peru, 552. 
 Fossil bones, 280, 311; plants of Abrolhos, 
 
 176 ; plants of carboniferous, 523 ; shells of 
 
 recent species in reef rock at Porto Seguro, 
 
 229; shells, reptilian remains, &c., in cre 
 taceous of Monserrate, 345; of cretaceous 
 
 of Maro i m, 385 ; shells near Propria, 404. 
 Frade de Macahe, height of, 42 ; de Itape ine- 
 
 rini, height of, 58 ; de Sao Leopardo, 68. 
 Francylvania, location, history of, &c., 95. 
 Freshet (annual of Sao Francisco, 420. 
 Frigate-bird (Tachypetes aquihna), cemetery 
 
 of the, 180. 
 
 Frogs, bones of, in caves, 275. 
 Fruita pao (Artocarpus incisa , 245. 
 Funeral customs of Botocudos, 599. 
 
 G. 
 
 Galena, 448, 472. 
 
 Gamella (bateia), 512. 
 
 Gamellas, Indian tribe, custom of piercing lip 
 and ear, 584. 
 
 Gardner on mythical reef, 189 ; note on Pi 
 ranha, 400 ; Paper on the Interior of the 
 Province of Ceani, 460 ; probably wrong in 
 referring the sandstone of Piauhy and Ceara 
 to the cretaceous, 476. 
 
612 
 
 INDEX. 
 
 Garnets, 13, 21, 49, 50, 151. 
 
 Garrafao, near Rio Itabapuana, height of, 53. 
 
 Gavia, height of, 10. 
 
 Gazzinelli, Signoru, 129. 
 
 Gelasimus Marac.oani Latreille, 229 ; G. palus- 
 
 tris Edwards, 230. 
 
 Genipapo, black color, furnished by, 589. 
 Geoffroya, 408. 
 Gerber, 3. 
 
 German colonists at Santa Clara, 132. 
 Gibbs, Hon. George, 584. 
 Glaciated surfaces, character of, 24. 
 Glacier of Pacatuba, 469 ; of the Amazonas, 
 
 490. 
 
 Glacial phenomena of Patagonia, 558. 
 Glyptodon. 286. 
 
 _2, 323, 324, 338, 340, SOU, 372, 4U5, 445, 
 449, 458, 461, 402, 403, 474, 500, 501, 549, 
 650. 
 
 God, Botocudos, destitution of belief in, 599. 
 
 Goitre in Brazil, 330. 
 
 Gold, 297, 448, 471, 497, 503, 513 ; at Canta- 
 gallo, 50 ; on the Rio Mangarahy, 77 ; on 
 Rio do Castello, 59 ; in drift at Sucuriu, 
 153 ; of vicinity of Miuas Novas, ancient 
 method of working, 160 ; abundance of, 
 in Minas Nova region, not yet exhausted, 
 156, 161 ; probable existence of, in tertiary 
 deposits iii Brazil, 162 ; mode of occurr 
 ence of, in Parahyba do Norte, 448 ; at A illa 
 das Lavras da Maugabeira, 463 ; mines of 
 Jaragua, 513; of Brazil, age of, 632; dis 
 tribution of, in Brazil, 632; mode of oc 
 currence at Morro Velho, 537 ; of Morro 
 Velho invisible in the ore, 541 ; mode of oc 
 currence of, at Rossa Grande, 542 ; red, 542, 
 545 ; of Gongo Soco, 641 ; color of, 642. 
 (See Chap. XVIII.) 
 
 Gorgonia gracilis Verrill, 209. 
 
 Goyabeira, 47. 
 
 Granite, 295, 500. 
 
 Granja, 460. 
 
 Graphite, 448, 458, 472, 525. 
 
 Grazina ( Phai ton), 180. 
 
 Greenstone, 328. 
 
 Green turtle, 112. 
 
 Grisolitas, 151. 
 
 Guarapary, harbor of, 63. 
 
 Guarapuava, gold at, 513. 
 
 Guayana, Tertiary rocks of, 493. 
 
 Guayamu, or Guainumu (Cardiosoma Guan- 
 
 Gullies formed by rains, 159. 
 
 Guriri (Diplntliemhnn maritimum Mart.), 
 
 237 ; a social plant, 256. 
 Gypsum, 471. 
 
 II. 
 
 Halfcld, 414 ; on mastodon remains found near 
 
 Paulo Alfonso, 418. 
 Harbor of Ilheos, 258; of Maceio, 425; of 
 
 Pernambuco, 435. 
 Hares, 95. 
 
 Hehastra-a aperta Verr., 207, 213, 269. 
 Heliconia, 94, 249. 
 Helmreichen, Dr. Virgilio, section across the 
 
 Jequitinhonha valley, 138 ; on the Serra da 
 
 Chapada, 306. 
 
 Ileusser and Claraz, 534. 
 
 "Highlands of Brazil," 276. 
 
 Hills or Serras, isolated, on limestone plain, 
 
 310. 
 
 Holothurians, 62. 
 Hoplophorutj 286. 
 
 Hornstone, hills of, near Volta da Serra, 312. 
 Houses of Botocudos, 596. 
 Humboldt on name Brazil, x; on distribution 
 
 of cocoa palm, 120 ; on social plants, 255. 
 Humpback whales, 181. 
 Hunt, Dr. T. Sterry, on Laurentian of Brazil, 
 
 650. 
 Hyatt, Alpheus, on cretaceous fossils from 
 
 MaroYm, 385. 
 Hymenogorgia (Gorgonia) quercifolia, 62, 74, 
 
 196, 229. 
 
 I. 
 
 Ico, 460. 
 
 Iguape ( Yg water, gua of varied color, p6 way, 
 Mart.), fertility of vicinity of, 271. 
 
 Ilex paraguayensis, 517. 
 
 Ilha do Paqueti, das Cobras, do Governador, 
 Euxada, 6 ; do Lima, 45; Escalvada, Raza, 
 63; doBoi,decomposinggneiss,70 ; Baleciro, 
 corals of 74 ; de Santa Barbara dos Abrol- 
 hos, 175 Redonda dos Abrolhos, 178; Ita- 
 parica, 267, coral banks of, 213 ; Grande, 
 258; Pequena, 258; Boyapeba, 265; dos 
 Fradres,267; Itamaraca, 437 ; de Fernando 
 de Noronha, 437; do Bananal, 602; de 
 Quioppe, 523. 
 
 Ilhas de Maricas, evidence of recent rise of, 35, 
 36 ; de Santa Anna, 41 ; dos Paeotes, 71. 
 
 Ilheos, 258- 
 
 Imbuzeiro (Spondias tuberosa Arr. ), 323. 
 
 Implements, human, in bone caverns, 286; 
 of Botocudos, 592. 
 
 Indai& palm, 144 ; (Attalea compta Mart.), 
 144, 166. 
 
 Inga, 94, 116, 249 : gneiss at, 446. 
 
 Inoceramus in cretaceous near Aracaju, 384. 
 
 Ipe (Tecoma), 94. 
 
 Ipecacuanha, 459. 
 
 Ipomcea littnralis, 237. 
 
 Iron ore, 58, 299, 301, 407, 447, 503, 515, 525. 
 
 Iron mines and works of Ypanema. 515. 
 
 Island of Victoria, 66 ; height of, 67. 
 
 Itabapuana (Hy water, taba village, apuan 
 round?) 
 
 Itabirite and jacutinga auriferous, 633. 
 
 Itabirite, 534. 
 
 Itacolumi, height of, 3 ; (Maranhao), 484. 
 
 Itacolumitc, 541, 545 ; the term loosely em 
 ployed, 332 ; gold in, 533. 
 
 Itaipins reef, 257. 
 
 Itambe (ltd stone, and faimbe rough), 
 
 Itanhaem (Ita stone, and nhetng speak, echo 
 ing rock). 
 
 Itapagipe, 334. 
 
 Itapitinga (reef), 258. 
 
 Itapuan (ita rock, apuan round), 346. 
 
 Itatiaiossu, Pico de, 2 ; height of, 3. 
 
 Itii (hit or Ytu a fall), gold at, 510. 
 
 .1. 
 
 JacarandA (Signonia Braziliensis), 93. 228, 
 
 246, 261, 459. 
 Jacar6, salt deposits of, 329, 
 
INDEX. 
 
 613 
 
 Jacehus (Hapale) leucocephalus, 95. 
 Jaguaripe, 268. 
 
 Jack tree (Artocarpus Braziliensis Gom.), 245. 
 Jacobina, character of country in vicinity of, 
 
 312 ; climate of, 317. 
 Jacupemba ( Penelope marail), 95. 
 Jacutinga, derivation of name, 633 ; de 
 scribed, 534, 535 ; gold in, free, 535 ; at 
 
 Gongo Soco, 541. 
 James, Major 0. 0., 507, 514, 584. 
 Janahuba, 374. 
 
 Janchon, bad spirit of Botocudos, 599. 
 Jangadas, number of, in Province of Pernam- 
 
 buco, 432. 
 
 Januaria, salt licks of, 320. 
 Jaragua, port of Maceio, 425 ; gold-mines of, 
 
 Mawe s account of, 511. 
 Jatropha, 249. 
 
 Jequii, tertiary bluffs of, 422. 
 Jiporock, tribe of, 593 ; Botocudo chief, 594. 
 Joazeiro, 327, 408. 
 Jundiahy, 513. 
 Jurassic rocks not yet known in Brazil, 554 ; 
 
 of Andes, 554. 
 Jurupencem, head of navigation of Araguaya, 
 
 501 
 
 K. 
 
 Kielmeyera, 291. 
 
 Kjoekkemnoeddings on the Ula do Governa- 
 
 dor, 6 ; at Santos, 6 ; at Santa Cruz, 6, 85. 
 Knobs, isolated, on lake plain, 315. 
 Koloshians of Alaska, custom of piercing the 
 
 lip, 584. 
 
 Kuparack, skull of, described, 584. 
 Kyanite, 145. 
 
 Lacerda, Dr. de, 337, 421. 
 
 Lagoa de Freitas, 11 ; de Marica, 37; Saquare- 
 ma, 37 ; Araruama, 38 ; Feia, 44 ; do Cam- 
 pello, 45; Maroba, 56; Jacun6, 83; Jupa- 
 ranaa, its depth, color of water, tertiary 
 beds, 100; do Aviso, 103; de Monserras, 
 106; Tapada, 107; Mariricii, 115; Gavata, 
 227 ; do Braco, near Belmoute, 236; Santa, 
 281 ; at Jacobina on the Sao Francisco, 409 ; 
 da Pedra, mastodon remains in, 418 ; do 
 Norte, Alagoas, 422 ; do Sul, Alagoas, 422 ; 
 Paranagui, 474. 
 
 Lagoons along coast, formation of, 44. 
 
 Lake Plain of Bahia, 314 ; lakes of, 314. 
 
 Lameirao, 76. 
 
 Lands of Rio das Velhas, 290. 
 
 Language of Botocudos, 602 ; simplicity of 
 grammatical structure of, 603. 
 
 Lapa Vermelha, 282. 
 
 Laurentian rocks, 447, 449 ; resemblance be 
 tween Brazilian gneisses and Laurentian of 
 Canada, 549. 
 
 Laurus, 249. 
 
 Lead, sulphate of, chromate of, sulphide of, 
 carbonate of, chloro-phosphate of, 448. 
 
 Leather-back turtle, 112. 
 
 Leaves, shedding of, by catinga woods in dry 
 season, 322 ; rapid growth of, on approach 
 of wet season, 322. 
 
 Lecythis,9,?A9. 
 
 Leite, Dr. Franca, his colony on the Rio 
 Doce, 95. 
 
 Lengoas custom of piercing lip and ears, 584. 
 
 Lenj6es, diamond washings at, 307 
 
 Linhares, fertility of lands near, 103. 
 
 Lizards, bones of, in caverns, 285. 
 
 Lepieiotus, 347- 
 
 Lepidatm temnums, 468. 
 
 Llianas, 250. L. JuparanSa (Jui, a frog, para- 
 nd, sea or lake), described, 99 ; compared 
 with lakes of Alagoas, 422. 
 
 Liais on height of Piedade, 3 ; on Sao Francis 
 co, 275 ; on the gold-mines of Brazil, 546. 
 
 Legs of Botocudo, thinness of, 579. 
 
 Licuri palm, 256, 423. 
 
 Lignite, 471. 
 
 Lime manufactured from corals, 213. 
 
 Limestone at Ypiranga, 15; at Cantagallo, 51 ; 
 of coral reefs, 198 ; of Sao Francisco, 279 ; 
 of Rio das Velhas, 281; at Malhada, 295 ; 
 east of Chapada Diamantina ; plain iu 
 Western Bahia, 310 ; aridity of, 317 ; at 
 Estancia, 380 ; at ParShyba gold-mines, 
 447 ; near Crato, 464 ; almost entire absence 
 of, from Laurentian of Brazil, 551. 
 
 Linhares, Rio Doce at, 97 ; inassape soils, 97 ; 
 situation of, 99 ; tertiary beds at, 99. 
 
 Linckia ornitfiopus LUtken, 214. 
 
 Lip, perforation of, by Botocudos, how per 
 formed, 582 ; ornament, 583. 
 
 Littorina. 229. 
 
 Lixo, Recife do, 202. 
 
 Lizards, abundance of, on Abrolhos, 180. 
 
 Lode, Descubridora, Parahyba, 448 ; Luna, 
 447 ; Boa Espera^a, 448. 
 
 Loggerhead turtle ( Tlialassochelys cauana), 
 abundance of, at Barra Secca, notes on, 108. 
 
 Lund, Dr. P. AV., on bone caverns of Brazil, 
 281. 
 
 H. 
 
 Macahg and Campos Canal, 43. 
 
 Ma<jainbamba, beach of, 39. 
 
 Macambira, 407, 441. 
 
 Maccape, locality for fossil fishes in Ceara,466. 
 
 Maceio Reef, 213 : topography and geology of 
 the vicinity, 422 ; city and harbor, 425. 
 
 Machacalis, village of, 167. 
 
 niarrauchenia, 286. 
 
 Miicuj6, discovery of diamonds at, 306. 
 
 Madrepora, absence of, from Brazilian polyp 
 fauna, 214. 
 
 Madrijos, female whales with young calves, 
 184. 
 
 Mcpantlrina, absence of, from Brazilian polyp 
 fauna, 214. 
 
 Mefstric/itien, 494. 
 
 Magalhaes, Dr. Conto de, 502. 
 
 Magnesia, 471. 
 
 Maquin(?, gold-mine of, 543. 
 
 Maize, 79, 142. 
 
 Malaphyr, 531. 
 
 Malhada, 265. 
 
 Mammals, number of fossil species discovered 
 by Claussen and Lund in the bone caverns 
 of Brazil, 285 ; remains of recent species of, 
 found occurring with extinct species, 286. 
 
 Mammalian fauna richer in the post tertiary 
 than now, 286 ; of true South American 
 type, 287 ; not richer in genera formerly, 
 287 ; richness in giant forms, 288. 
 
 Maniao (Papaya), 217. 
 
 Man, remains of, in bone caverns, 286. 
 
614 
 
 INDEX. 
 
 Manati ( Manatus Americanus), at Victoria, 75 , 
 at Sao Matheos, 122 ; in Rio Peruhype, 224. 
 
 Mandioca, 79, 99, 120, 123, 454, 458. 
 
 Mandubi, derivation of uame, &c., 409. 
 
 Mangaba, or Mangabeira (Hancornia speciosa 
 Gom.), 149, 374. 459, 405. 
 
 Manganese, 535, 536, 545. 
 
 Manguinhos, tertiary beds at, 55. 
 
 Mangrove swamps, mud of, 55 ; formation of, 
 220 ; unhealthiuess of, 240 ; swamp at mouth 
 of Kio Jequitmhouha, 239. 
 
 Mangroves, red (Lagunciilaria racemosa) and 
 white (Avicennia tomentosa), 122 ; as agents 
 in the silting up of swamps, 222 ; social 
 plants, 256. 
 
 Manicina, absence of, from Brazilian polyp 
 fauna, 214. 
 
 Maracassumt5, gold-mines of, 485. 
 
 Maragogipe, 271. 
 
 Marauguape, 460. 
 
 Maranhao, population of province and city, 
 486 ; gold-mines, 546. 
 
 Maroim, species of fly, noted for biting just at 
 nightfall, 240 ; city of, 384 ; cretaceous of, 
 384. 
 
 Marsh, Prof. 0. C., description of new croco- 
 dilians, 355 ; on occurrence of inosasau- 
 rians in Europe and America, 494. 
 
 MassapcS soils, 99, 118, 395. 
 
 Mastodon, remains of, 261,286,298,324,325, 
 418, 419, 471, 573. 
 
 Mate, Paraguayan tea, 517. 
 
 Matta de Sao Joao, 368. 
 
 Mattas, 379, 429. 
 
 Mauritia vinifera, 290. 
 
 Mawe, Travels in Brazil, 51 ; on gold-mines of 
 Jaragui, 511. 
 
 McGrath, Dr. t 437. 
 
 Megaptere, 181. 
 
 Megatherium, 286, 471. 
 
 Melania terebriformis Morris, 348, 350 ; M. 
 Nicolayana Ilartt, 350. 
 
 Melaphyr, 530. 
 
 Melastoma, 249, 291, 375 ; M. Hookerianus, 
 408, 423. 
 
 Me.rte.nsia rtichotoma, 256. 
 
 Messalia Ortoni. Gabb, 493. 
 
 Mestre Alvaro, 83. 
 
 Meteorolite of Bemdego, 325. 
 
 Meteorolites, shower of, in Ceari, 472. 
 
 Metamorphism no criterion of age, 280 ; signs 
 of, in eozoic rock not disappearing toward 
 the western part of plateau of Brazil, 500. 
 
 Miahype, 61. 
 
 Mica-slate, 135. 142, 148, 164, 252, 296, 298, 
 328, 445, 446, 458, 525,551. 
 
 Mice, occurrence of bones of, in caverns of 
 Lagoa Santa, 285. 
 
 Mi&epora aleieornis Linn, 40, 206; vnr. M. 
 di^itata, 206; far. M. tellu/.nsa,2Q6; var. 
 M. feneatrata Verr., 206; M. Eraziliensis 
 Verr., 195 ; M. Nitiiia Verr., 205. 
 
 Millepores, stinging properties of, 207. 
 
 Mimosa, 249,250 408. 
 
 Minas da Caxoeria, 447. 
 
 Minas Geraes, a land-locked province, necessi 
 ty of roads to the coast, 130. 
 
 Minas Novas, cause of failure of mines, 157 ; 
 geology of vicinity of, 157. 
 
 Mining companies. Companhia Metallurgica 
 do Assurua, 308 ; Tasso Brazilian Gold-Min 
 ing Company (limited), 449 ; Sao Joao d El- 
 
 Rei Mining Company, 536; Rossa Grande 
 Gold-Mining Company, 542 ; Doin Pedro 
 North d El-Rei Company, 542; East d El- 
 Rei Company, 544; Monies Aureos Gold- 
 Mining Company (limited), 546. 
 
 Mispickel, the principal gold-bearer, 534, 540. 
 
 Moisture, existence of forests, dependence on. 
 321. 
 
 Molybdate of lead, 472. 
 
 Monkeys of Rio Doce, 95. 
 
 Mouserrat, geology of, 334. 
 
 Monte Moreno, 65; Jutuquara, height, &c., 
 68 ; Parcoal, 226 ; Sauto, 325. 
 
 Montes Pyreneos, height of, 501. 
 
 Moon, superstitions regarding, among Botocu- 
 dos, 599. 
 
 Moraine, gigantic, of Amazonian glacier, 490. 
 
 Moraines of Pacatuba, 469. 
 
 Morainic deposits of Tijuca, 562. 
 
 Morro do Supateiro, 48; de Mestre Alvaro, 81 ; 
 de Agah, 59 ; de Nossa Seuhora da Peuha, 
 66; da Serra, 81, 83, 84 ; do Padre, 92 ; da 
 Terra Alta, 97 ; do Kupan, 133; do Ariao, 
 164 ; de Sao Paulo, 266 ; Redondo, 297 ; do 
 Caldeirao-assii, pot-hole on, 316; do Consel- 
 ho, sand-plains of vicinity of, 345 ; do Chaves, 
 geology of, 404 ; dos Cabellos Brancos, 513 ; 
 Velho gold-mine, 536 ; mode of extracting 
 gold, 538; profits, 539, 540; depth of mines, 
 539 ; de .^auta Anna, 542. 
 
 Mouchez, 226. 
 
 Moulding of drift-covered surfaces wholly dif 
 ferent from that of surfaces exposed to the 
 action of waves, 569. 
 
 Mountain ranges badly represented on maps, 
 278. 
 
 Mucury region, resemblance in topography, 
 soils, &c., to the coffee regions near Rio, 
 133 ; topographical features due to moist 
 climate and forests, 319. 
 
 Mud bottom bordering coral reef, 209. 
 
 Mundo Novo (Bahia), mastodon remains, 325 J 
 locality for fossil fishes in Ceari, 466. 
 
 Muras, custom of piercing lip, 584. 
 
 Murici (Byrsonhna), 374. 
 
 Musm Harttii Verr., 73, 196, 229. 
 
 Mussels, 73. 
 
 Mutum ( Crax), 95. 
 
 Mygale, 180. 
 
 Mycetes Ursinus, 95. 
 
 Mylodon, 286. 
 
 Mystiqa whale, 182. 
 
 Mythical reef, belief in accounted for, 189. 
 
 N. 
 
 Naegeli, Dr. II., 6. 
 
 Naknenuks, 679, 593. 
 
 Natal, 454. 
 
 Natica pra>,lnne;a Leymerie, 385. 
 
 Natividade, 497- 
 
 Navigation by steam of Sao Francisco, 399, 
 
 421. 
 
 Navigation of the Araguaya, 501. 
 New Red Sandstone of Estancia, 379. 
 Nazareth, 268. 
 Neocomen, 555. 
 
 Neritina, 348 ; N. pupa Gabb, 493. 
 Nests of sea-urchins in rock, 36. 
 Nest of loggerhead turtle, mode of excavating, 
 
 109. 
 
INDEX. 
 
 615 
 
 Xeuwicd, Prince Maximilian zu. 122 ; on 
 journey from Ilheos to Bahia, 249 ; on coun 
 try between Urubii and Faz. da Caxoeira, 
 257. See Appendix. 
 
 Nicolay, Rev. Mr., on turba deposits of Cama- 
 niii, 264 ; on geology of Chapada Uiamanti- 
 na, 301 ; geology of interior of Bakia, 301. 
 
 Na.ggera.thia obtusa Carruthers, 525. 
 
 Noruega whale, 182. 
 
 Nova Almeida, 85. 
 
 Nuggets of gold, 297. 
 
 Numerals wanting in language of Botocudos, 
 604. 
 
 0. 
 
 Octopods, 63, 204. 
 
 Oculina, absence of. from Brazilian polyp fau 
 na, 214. 
 
 Odontopteris Ptantiana Carruthers, 524. 
 
 Oiteiros, 362. 
 
 " Olhos," 545. 
 
 Olivenca, 248. 
 
 Oliveira, Sr., 437. 
 
 Onca, 95. 
 
 Oolitic rock at Aracare", 396. 
 
 Opal, common, 327. 
 
 Opluactis Krebsii Lutken, 197. 
 
 Ophiolepis paucispina Mull, and Troschel, 197. 
 
 Ophiomyxajlaccida Lutken, 197. 
 
 Op/tiontreis reticulata Lutken, 197, 203. 
 
 Ophiotkrix violacea Miill. and Troschel, 197, 
 203. 
 
 Ophiura cinerea Lyman, 62, 197, 203. 
 
 Opossums, 94 ; great abundance of bones of, 
 in bone caverns, 285. 
 
 Opuntia, 408. 
 
 Oreaster gigas, 19T. 
 
 Organic remains, none in drift, 573. 
 
 Orgfios, Serra dos, height of, 3, 7. 
 
 Ornament of Botocudo, 590. 
 
 Orthoclase, 9, 446. 
 
 Orton, 493. 
 
 Os Busies, 40. 
 
 Of Po??oes (Poroes), 255. 
 
 Ostrich, American, 146. 
 
 Ottoni, Senator Theophilo B., his project to 
 settle the Mucury and open a wagon-road 
 from Santa Clara to Minas Novas, 131 : min 
 ing right of in Minas Novas district, 162. 
 
 Ouro Preto, vicinity of auriferous, 536. 
 
 Ovuliim gibbosum, 196. 
 
 Owls, instrumental in accumulating bones in 
 caverns, 285. 
 
 Oysters, 73, 229. 
 
 P. 
 
 Paca (Ccrlogeny.t Paca), 94. 
 
 Pachyderms, more abundant in post tertiary 
 than at present, 287. 
 
 Pacific and Atlantic connected during the cre 
 taceous, 391. 
 
 Palladium, 545: in gold of Gongo Soco, 542. 
 
 Palms of Bay of Rio, 7 ; of Rio Doce 94 ; 
 cocoa ( Cocos nuciffra), 118 ; Nnyd, 118 ; 
 Mim, 118; Timbore, 118; Indaia, 144; 
 Carnahuba, 452. 
 
 Palm oil, 270. 
 
 Palmetto (Euterpe"), 94; (Euterpe erfwtoMart.), 
 216. 
 
 Palythoa, 40, 62, 192, 193. 
 
 Pao d Arco (Bignonia), 94, 459. 
 
 Pao de Assucar (Rio), 8, 9 ; Victoria, 66 ; Rio 
 de Sao Francisco, 410. 
 
 Pao Brazil ( Cesalpinia echinata), 94, 228. 
 
 Parafuso, sand-beds at, 360. 
 
 Parahyba (city), population, commerce, &c. 
 of, 442. 
 
 Parahyba do Norte, geology of gold-region of, 
 445. 
 
 Parana, province of, 517. 
 
 Paranagui, 518. 
 
 Parcel dos Abrolhos, 199 ; das Paredes. 201. 
 
 Passiflora, 249. 
 
 Patagonia, drift of, 558. 
 
 Paulo Affonso, falls of, 414 : compared with 
 Niagara and the Salto Grande do Jequi- 
 tiuhonha, 418. 
 
 Peanut (Arackis hypogtza), 409. 
 
 Peat, 365, 509. 
 
 Pebble sheet of drift, 24. 
 
 Pectinia, 74 ; P. Braziliensis Edw. and Hahne, 
 208. 
 
 Pedra Bonita, 32 ; Formosa, 53 ; Lisa, 54. 
 
 Penedo, geology of vicinity of, 397 ; popula 
 tion, commerce, &c., 398. 
 
 Penn, Staff Commander, on the reefs near 
 Cape Sao Roque, 188 , on recife, 188. 
 
 Perigot, Dr., discoverer of the Brazilian coal 
 fields. 520. 
 
 Periperi, cretaceous rocks at, 359. 
 
 Pernambuco, rivers of, 430 ; convenience of, 
 as a port, 432; situation of city, 433; har 
 bor of, 435 ; population of city and prov 
 ince, 432 ; derivation of name, 432. 
 
 Philadelphia, 131 ; height above sea, 134 ; 
 country in vicinity of, 134 ; richness of soil, 
 fitness for colonization, 134 ; road, present 
 state of, 135. 
 
 Phonolite of Fernando de Noronha, 438. 
 
 Physiognomy of Botocudos, 580. 
 
 Piassaba (Attalea J unifera Mart.), 237, 247, 
 271, 336, 373. 
 
 Pico do Enchadao, 169; do Enchadinho, 169; 
 de Itambe, 137: d ltabira, 535; Piedade, 
 Alto da Surra da, height of, 3. 
 
 Pine-apples, 431. 
 
 Pittgoas d agua, 164. 
 
 Piqui (caryocar Srasiliensibus), 149. 
 
 Piranha (Pygocentnis), derivation of name, 
 399 ; description of habits, 399 ; note on, by 
 Gardner, 400 ; voracity of, 401, 402, 502. 
 
 Pirrtmpeba (fish), 403 
 
 Piso on Brazilian reef, 187. 
 
 Pisorlus in cretaceous at Bahia, 355. 
 
 Pissi? on gneiss of Corcovada and Copocabana, 
 10; on eozoic gneisses of Brazil, 548. 
 
 Pistacte, 297, 324, 327, 485. 
 
 Pitanga, section at, 368 ; diamond mines, 369. 
 
 Pitangueira (Evgenia), 19, 59. 
 
 Plain of tertiary at Carapina, 82. 
 
 Plains of sand at Victoria, 82; coast, forma- 
 tior of. 221 ; at mouth of Jequitinhonha, 
 221. 
 
 Planorbis Monserratensis Hartt, 351. 
 
 Plants, fossil, on Rio Pardo, 243 ; Abrolhos, 176 ; 
 of carbonifera of Rio Grande do Sul, 524. 
 
 Plant, Nathaniel, Esq., report on coal-fields of 
 Rio Jaguarao. 521. 
 
 Plataforma, cretaceous beds of, 347. 
 
 Plateau of Province of Bahia, 310 : middle, of 
 Bahia, 313; third or coast, of Bahia, 316; 
 second, of Bahia, vegetation, climate of, 317 ; 
 of Appodi, 451 ; of Brazil in Sao Paulo and 
 
616 
 
 INDEX. 
 
 Parani, 505 ; of Brazil, everywhere under 
 laid by gneiss, 550. 
 Platinum, 448 ; occurrence of in gold-mines 
 
 of Brazil, 542. 
 
 Plexaurella dickotoma Roll, 62, 197. 
 Po-assii (Yg-apo-assii, running water or big 
 
 swamp), 237 ; probable results of widening, 
 
 240. 
 
 Pocjoes, 255. 
 Pohl, 496. 
 
 Pojicha, chief and tribe of Botocudos, 592, 594. 
 Pojuea tunnel, 371. 
 Polyps, distribution of, along Brazilian coast, 
 
 191. 
 
 Ponipeo, 428. 
 Ponds on reefs, 210. 
 Ponta de Jecii, 64 ; do Tubarao, 65 ; Caixa de 
 
 Pregos, 267 ; Garcia, 267. 
 Pontal at mouth of Rio Arassuahy, 163 ; da 
 
 Barra do Sao Francisco, dunes at, 395 
 Porcupines, bones of in caves at Lagda Santa, 
 
 285. 
 
 Porites, in Bay of Rio, 192. 
 Porites soliila Verrill, 203, 208. 
 Porphyry, 446, 449. 
 Porokum, skull of, 588. 
 Pororoca or Bore, 487. 
 Porto de Souza, geology of vicinity, 90 ; das 
 
 Caixas, 6, 21 ; da Villa Nova, recent deposits 
 
 at, 18 ; da Pedra, 76 ; de Souza, 89 ; richness 
 
 of soil at, 91. 
 Porto Alegre (KioMucury), 128; (Kio Grande 
 
 do Sul), 531. 
 
 Porto Seguro, 228 ; harbor and reef of, 228. 
 Porto das Piranhas, 412. 
 Post tertiary, 573. 
 Pot-holes, occurrence of diamonds in, 307 ; of 
 
 Lake Plain of probable glacial origin, 315, 
 
 562, 571. 
 Prado, 225. 
 
 Praia Grande, geological section at, 13. 
 Precipices, character of, near Rio, 31. 
 Prionus cervicornis, larvas of, eaten by Botocu 
 dos, 597. 
 
 Pristis, species of, in Rio Doce, 91. 
 Propria, 404. 
 Pratnpit/ieciis, 286. 
 Przewodowski, Sir., 272. 
 Psitlium Guiava, 47. 
 
 Ptfris caudala, 144, 252; social plant, 256. 
 Plfrngorifia ( Gorgonia) gracilis Verr., 81. 
 Pumice, 179, 248. 
 Pyaocentrus (fish), 403. 
 Pyrites, magnetic iron, gold of, 534, 541 ; iron, 
 
 associated with gold at Morro Velho mine, 
 
 537. 
 
 Q. 
 
 Quartz, auriferous, 448 ; reins, 448, 533. 
 Quartzite, 152, 296, 298 ; auriferous, 534. 
 Quadersandstein, 298. 
 Quixeramobim, 460. 
 Quina, 19, 459. 
 Quiricare, Rio, 117. 
 
 R. 
 
 Railroads of Brazil, 293 ; Cantagallo, 6, 19 ; 
 Dom Pedro Segundo, geology of, 14 ; Per-, j 
 nambuco and Sao Francisco, 436; Santos j 
 and Sao Paulo, 506. 
 
 Rains on Lower Sao Francisco, 420 ; of Ma- 
 ran hao, 485. 
 
 Recife, meaning of word, 191, 432; city of, 
 432 ; do Lixo, 202 ; do Leste, 211 ; da Pedra 
 Grande, 211 ; Itanhaein, 212 ; dos Itacolu- 
 mis, 212, 227 ; das Timbebas, 212 ; do Porto 
 Seguro, 212. 
 
 Reconcavo, 272. 
 
 Reefs, Coral, see Coral Reefs ; stone, or con 
 solidated beaches, of Guarapary, 62 ; at 
 Barra Secca, 107, 113; Porto Seguro, 229 ; 
 Santa Cruz, 232; Bahia, 342; Rio Ver- 
 melho, 344 ; Peruarnbuco, 434 ; Parahyba 
 do Norte, 442, 454. 
 
 Reiuhardt, Prof. G., 279; general conclusions 
 with reference to the fossil fauna of bone 
 caverns, 287 ; on fossil iiiau of Brazil, 287. 
 
 Renilla Dance, 192. 
 
 Rkacolepis buccalis, 468 ; -R. olfersii, 468 ; R. 
 latus, 468. 
 
 Rhea JJarwini, 147. 
 
 Rhexia, species of, social plants, 256. 
 
 Hiacho, meaning of word, 123 ; dos Porcos, 
 457 ; do Mundo Novo, 457 ; das Ostras, 124. 
 
 Uibeirao Diamantino, 149 ; d Agoa Nova, 149 ; 
 da Issara, 250 ; do Meio, 154. 
 
 Rice, 116, 261, 409, 410, 454. 
 
 Riedel, 281. 
 
 Rio, Bay of, 6 ; d Agua Limpa, 154 ; d Agua 
 Suja, 137, 153 ; das Americauas, 126 ; An- 
 darahy (Andira bat, ky river), diamond 
 washings on, 306; Apiapitanga, 85; Ap- 
 podi, 452; Araguaya, 501 ; Arassuahy (Ara- 
 assu-ky great parrot water, or Coaracy-cy 
 water of the sun, Mart.), 137, 163; gold 
 in sands of, 163 ; d Areia, that part of 
 a river whose bed is composed of sand, 
 so called in contrast with the part whose 
 banks and bed are rocky, 126, 171, 242 ; 
 da Barra Secca, 106 ; Beuevente, 60 ; Bom 
 
 of, 521 ; Canind6, 474 ; Capanema (cda 
 leaf or wood, panemo sterile), 271 ; Capi- 
 baribe, 430 ; Capivary, 137, 145 ; Carahype, 
 83 ; Carapina, 76 ; Caravellas, 219 ; Caravel- 
 las, mouth of, 223 ; Carimatau , 452 ; Carun- 
 haiiha ( Caryra to run, anhe enough, " Flu- 
 vius sat rapidus," Martius), 295 ; do Cas- 
 tello, gold-washings on, 59; da Caxoeira, 
 227 ; Ceara-inerim, 451 ; Chopot i, 87 ; Com- 
 mamlatuba( Comanrla feijao,and tyba place), 
 242 ; de Contas, 260 ; Corrientes, 88 ; Cotin- 
 guiba ( Cntitc to wash and iba tree, River of 
 the Sapindus or soap tree?), 379 ; Cotingui- 
 ba or Cotindiba, 381 ; bar of, 382 ; Craminu- 
 an, 227 ; Crubixa, 75 ; Cuitii, 89 ; Curipe, 
 76 : Doce, description of Basin of, 86 ; width 
 at Porto de Souza, 90 ; luxuriance of forest 
 in, 93 ; freshets of, 98 ; productions of, 98 ; 
 Bar of, 102, 104 ; below Linhares, 103 ; Com 
 pany of, 104 ; American colony on, 105 ; salt 
 trade of, difficulty of carrying on, 104, 105 ; 
 da Fabrica, 219 ; Fanado, 137, 145 ; do Frade, 
 227 : do Fundao, 248 ; Giboya, 250 ; Grande 
 do Norte, 451 ; Itanhaein, 224; ImbiiQahi, 
 364 ; Ipojuca, 430 ; Irapirang ( Yra honey 
 and piranga red), 381 ; Iriritiba, 60 ; Gra- 
 vatu, 137, 148; Gualaeho, 88 ; Guandu, 89 ; 
 Guarapary, 63 ; Guaxindiba, 122; Gurgueia, 
 474 ; Iguassii, 44 ; Itabapuana, 62 ; Itacam- 
 
INDEX. 
 
 617 
 
 birussu, 136; Itahunas, 122; Itahype(ito 
 stone, ky water, pe way, river among stones), 
 259; Itamarandiba, 137; Itapicuru, 323; 
 Itapemeriin, 58; Jacoruna or Jacuruna, 
 (Jacit species of Penelope, una black), 268 ; 
 Jacuahy, coal-basin on, 530; Jacuhy, 76; 
 Jaguaribe, 458 ; Jaguaripe (river of the On- 
 5,1), 267 ; Japaratuba, 379 ; Jecu, 64 ; Jequi6, 
 
 J.1UU, tO 5 *J UUOOtJllltt, A4t j UULLllllUCO, <JUV , 
 
 Jucurucii, 224 ; Juparauaa, 99 ; Jussiape, 
 259 ; Macabii, 44 ; Macucii, 18 ; da Mae 
 d Agua, 154 ; Mauianguape, 443 ; Mangara- 
 hyba, 76 ; Mangamhy, 77 ; Manhuassii, 90 ; 
 Mariricu, 107,223; Meariiu or Meary, bore 
 of, 486 ; do Meio (Medio), 77 ; Mucury, 125 ; 
 Mugiquisaba, 235 ; Muriahe, 4, 47 ; Mutum, 
 91 ; Pampao, 126 ; Panca, 95 : Para, of Mi- 
 nas, 288 ; Paraguassii, 269, 303 ; Paraguay, 
 valley of, worn in tertiary sandstones, 504 ; 
 Parahyba do Sul, 3, 45 ; Parahyba do Xorte, 
 441 ; Parahybuna, 4 ; Parana, 518 ; Parana- 
 
 tuuauiiu, t , riauiij , j.u^, ot<7 . IT iia*. i*jauu, 
 
 3 ; Pirahy, 3 ; Pirahytinga, name given to 
 upper part of the Parahyba do Sul, 3 ; Pi 
 ranhas, 451 ; das Pirangas, 87 ; Pirapama, 
 430 ; Pirapitinga, 4 ; Piriqui-assu , 85 ; Piri- 
 qui-mirim, 85; do Peixe, 323; Piuma, 59; 
 Pomba, 4 ; Potengi, 452 ; Poty, 474 ; Poxim 
 (poxim, ugly). 241 ; Preto, 4, 125 ; Purus, 
 494 ; Real, 379, 380 ; dos Keis Magos, 85 ; 
 de Salitre, 327; Salgado (Ceari), 457; da 
 Salsa, 237 ; Santa Maria, 65 ; de Sao Fran 
 cisco, 276, geology of, 276, explorers of, 
 275, rank of, among rivers, 288, courses 
 described, 288 ; Upper, unhealthiness of, 
 292, navigability of, cost of removing ob 
 structions from, 292, opening of, to steam 
 
 Antonio, mastodon remains found in vicinity 
 
 ^uhi-grande, 88 ; Setubal, 137 ; country in 
 vicinity of, 140; Serenhacm, 430: Serigi, 
 272 ; da Serra, 65 ; Setubinho, 140, 147 ; 
 Sipo, 241 ; Soledade. 137 ; Taipe, 227 ; Tap- 
 ajos, granite on, 600 ; sandstones of, 504 ; 
 TauhS, 76 ; Tibaje, 513 ; Tiete. 510 ; Tocan- 
 tins, 501; Todos os Santos, 126; Trahiry, 
 451 ; Una (black river) of Espirito Santo, 
 south of Victoria, 64 ; west of Victoria, 76 ; 
 of Bahia, 246 ; of Pernambuco, 430 ; Upan- 
 
 Liais s picture of the scenery on, 291, Verde, 
 
 328; Xingu, 503. 
 Rise, gradual, of coast within recent times, 
 
 probably still in progress, 36 ; observation 
 
 on, at Victoria, 72. 
 Roads, scarcity of wagon-roads in Minas 
 
 Geraes, 130 ; wagon-road from Sauta Clara 
 
 to Philadelphia, 131. 
 Rock crystals, 531. 
 
 Rock salt said to occur on Rio Iluallaga, 329. 
 Roccas, 214. 
 Rosewood, 93, 102. 
 
 Ruminants mure abundant in post-tertiary 
 than at present in Brazil, 287. 
 
 Sabar-i, 290. 
 
 St. John, Mr. 0. H., 275, 295, 475. 
 
 Saline streams, 304. 
 
 Saline deposits of Valley of Sao Francisco, 
 328 ; method of extracting the salt, 329. 
 
 Salines of Araruama, 38 ; of Maijambamba, 39. 
 
 Salt, cost of, in the interior, 130 ; commerce 
 in, on Rio Doce, 104; of Fazenda Aldea, 
 impurities of, 328 ; of Sao Francisco valley, 
 note by Mr. Allen, 329 ; in CearA, 471 ; 
 trade in, of Mucury, 128; of Rio Graude 
 do Norte, 455. 
 
 Salter on fossils of Silurian of Andes, 552. 
 
 Salto Grande do Jequitinhonha (village and 
 falls), 170 ; do Parana, 518. 
 
 Saltpetre in bone caves at Lagoa Santa, 284 ; 
 of Rio de Salitre, 330. 
 
 Samambaia, 144 ; a pest in Brazil, growth of, 
 aided by fires, 256. 
 
 Sambahiba, 374. 
 
 Sand-beaches, formation of, 219. 
 
 Sand, blowing, on Sao Francisco, 332. 
 
 Sands blown over drift at Bahia, 346 ; of Ta- 
 boleiros, age of, 377. 
 
 Sandstone, tertiary, 47, 48, 55, 56, 57, 60, 100, 
 101, 113, 123, 124, 150, 243, 270, 277, 301, 
 305, 311, 332, 371, 379, 458, 464, 474, 475, 
 484, 489, 501 ; cretaceous, 175, 346, 347, 348, 
 397, 398 ; paleozoic, 243 ; triassic, 379. 
 
 Santa Catharina, corals of, 192 ; S. Clara, 128 ; 
 S Cruz, 85 ; S. Euzebia, 310 ; S. Luzia, 452 ; 
 S. Maria, 76. 
 
 Santo Amaro, 272. 
 
 Santos, 506. 
 
 Silo Christorao (city), 381; S. Fidelis, 49; S. 
 Joao da Barra, 45; S. Joao d El-Kei, 536; 
 S. Joao d Ipanema, 515 ; S. Matheos, 80, 
 121 ; S. Miguel, 167 ; S. Paulo (city), 506. 
 510 ; S. Vicente, 534. 
 
 Sapucahy, cretaceous limestones at, 383. 
 
 Sapucaia (Lecythis), 94. 
 
 Saurians, remains of, at Monsenute, 347. 
 
 Sawfish in Rio Doce, 91. 
 
 Schieber, Mr. George. 126, 129. 
 
 Sc/iinus terebinthi folia, 19, 423. 
 
 Schists, hornblenilic, 446. 
 
 Schlobacli, engineer, on height of Philadel 
 phia, 134. 
 
 School, 151, 327. 
 
 Sciurus (fstuans, 286. 
 
 Scorpions, 375. 
 
 Sea-anemones, 62. 
 
 Sea breeze at Penedo, 420. 
 
 Sea-level, ancient, at Victoria, 71. 
 
 Sea-urchins, nests excavated in rocks by, 38. 
 
 Sebastiao Leme do Prado, discoverer of gold 
 mines of Minas Novas, 156. 
 
 Sellow, 530. 
 
 Senonien, 556. 
 
 Sergipe d El-Rei, 381. 
 
 Sergipian group, 556. 
 
 Seriema (Dicholophus cristatus), 254. 
 
 Serpentine, occurrence of, in eozoic limestone 
 at Pirahy, 549. 
 
 Serra dos Akuore s, 84 ; das Almas, 254 ; das 
 
618 
 
 INDEX. 
 
 Araras, 499 ; de Araripe (Arari-ipe, locality 
 of), 278, 457, 464 ; Arassoiava or Guaracjoi- 
 ava, 516 ; de Aratanha, glacial phenomena 
 of, 469 ; de Arere, 461 ; do Assurua, 308 ; 
 da Borborema, 440, 451 ; do Boquerao, 304 ; 
 de Buenos Ayres, 291 ; dos Cairiris Novos, 
 451 ; dos Cairiris Velhos, 428, 440 ; do Cal- 
 deirao da On^a, 303 ; de Caytel6, 296 ; dp Con- 
 durii, 263; de Curumatahy (Ciiriinatdfish, 
 499, and ky water or river), 291 ; dos Dous 
 Irmaos, 278, 476; do Espinha^o, 136, 277 ; i 
 do Feijoal, 168 ; do Oado Brabo, 327 ; de ! 
 Garauhuns, cotton of, 431 ; da GarQa, 291 ; 
 Geral. 498 ; Grande, 456 ; do Grao Mogor, 
 136, 139 ; das Ibiturunas. 87 ; d ltabayana, 
 379, 381 ; d ltacambira, 136 ; de Itaparica, 
 419 ; de Itaraca (Ita stone, and aca horned), 
 247 ; de Itaubira (Ita stone, berab flaming), 
 298, ; de Jaragua, 511 ; das Lages, 298 ; da 
 Lapa, 331 ; da Lua Chela, 168, 169 ; near 
 Macah6, height of, 42 ; do Macaco, 162 ; 
 Maugabeira, 304 ; da Mantiqueira, general 
 description of, geological structure. &c., 2 ; 
 da Mantiqueira, 86 ; do Mar, 1, 14, 505, 547 ; 
 da Mata da Corda, 291 ; do Matuca, 52 ; do 
 Mocambo, 303; dos Monies Altos, 296; do 
 Morro Queimado, 21 ; d Olho d Agua. sand 
 stone at, 414 ; da Onca, 48 ; dos Orgaos, 
 topographical and geological description of, 
 15; das Panellas, 168 ; do Paranan, 499; do 
 Paraiina (Para river, una black. Black- 
 river), 291; das Pedras d Agua, 312; das 
 Pedras Brancas, 303 ; de Pereira, 462 ; da 
 Piedade, 289, 535 ; do Pintor, 308 ; do Rio 
 do Peixe, 323 ; de Sao Fidelia, 49 ; de Sao 
 Joao, 41 ; de Sao Romao e Santa Paz, 52 ; 
 da Saude, 303 ; do Sincora, 299, 306 ; Sobra- 
 do, 137 ; da Sussuarana (Suassii deer, rana 
 false, Suassuarana puma, or Felis con- 
 color), 250: da Tabatinga, 473 ; da Teixeira, 
 440, 446 ; da Terra Dura, 316 ; de Tiuba, 
 324 ; do tombador, 316 ; da Topa, 475 ; dos 
 Vertentes, 501 ; daVigia, 168 ; da Villa Vel- 
 ha, 296 ; da Ybiapaba, derivation of name, 
 456 ; structure of, 457. 
 
 Serrado, 147. 
 
 Sertao, 429, 459. 
 
 Shales, cretaceous, of Monserrate, 349 ; with 
 fish remains near Proprii, 404. 
 
 Sidfrastrrra stellata Verr., 62, 74, 193, 203, 
 207, 214 ; var. conferta Verr., 194. 
 
 Sienite, 324. 
 
 Sincori, diamond-washings at, 307. 
 
 Silurian rocks of Brazil, 551 ; limited range of 
 animals during, 552 ; of Bolivar and Peru, 
 552 ; fossils at, 552. 
 
 Silver always found in Brazilian gold, 542. 
 
 Slate, conglomerate, 242. 
 
 Slates on Rio da Mae d Agua, 154 ; (auriferos) 
 of Minor Novas and Chapada, 157 ; decom 
 position of, 158. 
 
 Slides of rocks on Bahia Railroad, 373. 
 
 Sloths, 94. 
 
 Smilorion neogfTMX, 286. 
 
 Smith, Mr. S. I., on Brazilian Crustacean, 203. 
 
 Social plants, variety of, in Brazil, 255. 
 
 Soils of the Rio Doce region, fertility of, 96- 
 103 ; thinness of on Lake I lain of Bahia, 314 ; 
 of Rio Sao Matheos, 117, 118 ; of tertiary 
 lands of Mucury, 133; of Urucii, 133; of 
 vicinity of Calhao, 152. 
 
 Solanaceee, 250. 
 
 Spar, double-refracting, 525 
 
 Sphargis coriacea, 112. 
 
 Sphenopteris, 525. 
 
 Specular iron, 472. 
 
 Spiders of Abrolhos, prey on lizards and birds, 
 180 
 
 Spix and Martius on vegetation of sand plain 
 near Rio, 19 ; on journey from Malhada to 
 Cachoeira, 295 ; on dead corals in lake near 
 Ilheos, 259 ; on benidego meteorolite, 325. 
 
 Squirrel, ^)5. 
 
 Staurotide, 252. 
 
 Steamboat lines from Bahia, 338. 
 
 Stone implements of Botocudos, 592; reefs, 
 see Reefs. 
 
 Sturgeons, none in South America, 403. 
 
 Stratified deposits, absence of, in connection 
 with drift, 566. 
 
 Strire, glacial , reason for not finding in Brazil, 
 57. 
 
 Striated surfaces, absence of, in Brazil, 562 ; 
 not reported in Patagonia and Chili. 
 
 Structure, want of, in drift, 564. 
 
 Stunthn (e. g. in der 7th Stunde), 530. 
 
 Submerged border of coral reefs of Lixo, life 
 of, 205. 
 
 Subterranean streams, 498 
 
 Sugar, 120, 261, 272, 380, 393, 431, 441, 442 ; 
 method of preparing at Muriahe, 48 ; cane, 
 99, 452, 458, 459 ; plantations, 45, 48, 52, 
 271, 452. 
 
 Sugary quartz, 545. 
 
 Sulphate of magnesia, 330. 
 
 Sulphide of antimony, 472. 
 
 Sulphides occurring with gold, 534. 
 
 Sulphur said to occur in the Province of Rio 
 Grand do Norte, 455. 
 
 Sumidouro cavern, remains of man in, 286. 
 
 Swamps of Rio Maricu, 116; between the 
 Rios Peruhype and Caravellas, 219 ; between 
 Rios Jequitinhonha and Pardo, 238 : jour 
 ney in, 239. 
 
 Sympkyllia Harttii, 196. 
 
 T 
 
 Tabatinga clay, 363. 
 
 Table land between Sao Francisco and Tocam- 
 tins basins, 277 ; of Sao Paulo, 509. 
 
 Table-topped hills or Serras of the Sao Fran 
 cisco valley, 331 ; of Erer6, Obydos, Cupati, 
 Almeyrim, 489, 490. 
 
 Taboleiros, 303, 304, 312, 362, 365, 373, 464 
 
 Talcose slates, auriferous, probably Silurian, 
 551. 
 
 Talhadao, 417. 
 
 Tamanduatahy, peat-bogs near, 509 
 
 Tapanhoacanga, 536, 559. 
 
 Tapcra daCima, deposits of iron ore near, 332. 
 
 Tapir ( Tapirus Americanus), 94. 
 
 Tapuyos, 578. 
 
 Taquara (Bamboo), 216. 
 
 Taquara-assu requires humidity and consid 
 erable elevation, 84 ; social plant, 256. 
 
 Taquara lisa, 141. 
 
 Tea, Paraguayan, 517 ; Chinese, culture of, in 
 Brazil, 517. 
 
 Teiii lizard ( Teius monitor), 111. 
 
 Tellina Amazoniensis Gabb. , 493. 
 
 Termo do Jardim, 457 
 
 Terra roxa, 514. 
 
INDEX. 
 
 619 
 
 Terrace form of the province of Bahia, 301. 
 
 Tertiary, Cantagallo Railroad, 20 ; between Os 
 Buzios and M:icah, 41; Barreiras do Siri, 
 56 ; Linhares, 99 ; Lagoa Juparanaa, 100 : 
 Sao Matheos, 117 ; Itahuna, 123 : Santa 
 Clara, 129 ; Jequitinhonha Valley, 139 ; be 
 tween Peruhype and Mucury, 215 ; near 
 Porto Seguro, 225 : between Porto Seguro 
 and Santa Cruz, 231 ; Rio Pardo, 243 ; fos 
 sils of, reported from Bay of Bahia, 269 ; 
 Reconcavo, 272 ; vicinity of Camassari, 361 ; 
 lands of Bahia, 373 ; Alagoinhas, 375 ; Ala- 
 goas, 422 ; Macei.), 424 ; Pernambuco, 429 ; 
 Parahyba do Norte, 443; Piauhy, 474 ; Ma- 
 ranhrto, 484 ; Amazonas, 492 ; Pebas, 493 ; 
 western part of Amazonas-Paraguay water 
 shed, 503 ; of Brazil, Resume, 557. 
 
 Thalassochdys cauana, abundance of, at Barra 
 Seeca, mode of depositing eggs, &c., 108. 
 
 Thoracosaitrus Bahiensis Marsh, 357. 
 
 Tide pools on the reefs, life of, 197. 
 
 Tutsskrift Lutken s, 281. 
 
 Tijuca, peak of, height, 12 ; drift of valley of, 
 26. 
 
 TMandsia. 21, 251. 
 
 Tobacco, 336, 441, 454. 
 
 Topaz, 151, 553. 
 
 Topography of gneiss region in the interior of 
 Bahia, Sergipe", and Alagoas, 318 ; of ter 
 tiary, near Porto Seguro, 225 ; of tertiary 
 near Pojuca tunnel, 371- 
 
 Torgjusen, Jacob, 202. 
 
 Tourmaline, 135, 145. 
 
 Tortoise, cretaceous, 358. 
 
 Toxoclon, 286. 
 
 Traipii, geology of vicinity, 406. 
 
 Trancozo, 227. 
 
 Tramroad, steam, of the Paraguassu, 270, 338. 
 
 Trap of Abrolhos, 176 ; decomposition of. 177 ; 
 hills, range of, in Rio Grande do Sul, 530. 
 
 Travelled boulders, 570. 
 
 Tres Irmaos, 10. 
 
 Triassic sandstones of Estancia, 379, 554. 
 
 Tripnemtes, absence of, from Brazilian radi 
 ate fauna, 198. 
 
 Trumpet used by Botocudos 592. 
 
 Turba deposits at Camamu, 262. 
 
 Turbonel/a minweula Gabb., 493. 
 
 Tucum (Astror.aryum tucuma), 116. 
 
 Tupi names in Brazil, ix. 
 
 Tupinambas, lip ornaments, 584. 
 
 Turi, gold-mines of, 485 
 
 Turtles (sea), 108. 
 
 " Turtles, fossil or Septaria, at Maroim, 393. 
 
 Turner, Engineer, 371. 
 
 IT. 
 
 Uba grass (Gyneriiim parvifolium Sees), 45, 
 94, 395 ; social plant, 256. 
 
 Uca una, 204. 
 
 Ulloa, on the Island of Fernando de Noronha, 
 439. 
 
 Unio, 101. 
 
 Unio (Anndon?) Totittm- Sanctorum, Hartt, 
 348, 351. 
 
 Urubii, 257; change in geological structure, 
 climate, and vegetation below, on Sao Fran 
 cisco, asi. 
 
 Urucu (Bixa OreUana Linn.) used as paint 
 by Botocudos ; colony on Mucury, 133. 
 
 v. 
 
 Vaccinium, 118. 
 
 Vadelli, 301. 
 
 Valenoa, 266. 
 
 Valley of Sao Simao, 168; of the Calhao 
 Arassuahy, 150. 
 
 Valleys of vicinity of Alto dos Bois, 156 ; 
 without outlets, 303. 
 
 Vanilla, 517. 
 
 Vegetation of sand-plains near Victoria, 62, 
 65 ; of Swamp of Rio Mariricu, 116 ; on 
 the Rio Sao Matheos below the city, 121 ; 
 of chapada near Santa Rita, 144 ; of Alto 
 dos Bois, 146 ; of Chapada of Minius No 
 ras, 147 ; of vicinity of CalhJo, 152 ; of 
 vicinity of Sucuriu , 154 ; of Jequitinhonha 
 less luxuriant than that of the Doce, 172 ; 
 of the Abrolhos, 180 ; of country between 
 Mucury and Peruhype, 216 ; of sand-plain 
 at Belmonte, 237 ; "on Rio Pardo, 244 ; of 
 Pardo, 245; of Campos, 290; of Serra 
 da Villa Velha, 296: of interior of Bahia, 
 317 ; of Campos of Alagoinhas, 375 ; near 
 mouth of Sao Francisco, 395 ; of Rio Grande 
 do Norte, 452. 
 
 Veins in decomposed rock never traceable into 
 drift, 565 ; of granite, 164 ; (quartz), in Mi- 
 nas Novas region, 157 ; auriferous, at Cha 
 pada, 158 ; auriferous, of Parahyba, 448. 
 
 Venus flexuosa in sands on the Cantagallo 
 Railroad Extension, 19. 
 
 Verrill, on resemblance between marine fauna 
 of West Indies and Brazil, 198. 
 
 Victoria, harbor of, 66, 80 ; (Conquista), 255. 
 
 Villa da Barra do Sao Matheos, 122 ; da Barra 
 do Ttabapuana, 54 ; Vicosa, 218 : da Barra 
 do Jardim, 465 ; do Crato, 464 ; das Lavras 
 da Mangabeira, 463 ; Nova da Rainha, 324 ; 
 Nova near Penedo, 397 ; do Rio de Contas, 
 298; Sao Bernardo, 461; da Serra, 84; do 
 Sucuriu, 153; Nova, Sao Francisco, sand 
 stones at, 397. 
 
 Vinhatico (Acacia), 94, 261. 
 
 Vii-ipara (Paludina) Lacen1<K Hartt, 350; 
 V. (Pahidinn) Williamsii Hartt, 351. 
 
 Vocabularies, Botocudo, 605. 
 
 Valuta, 203, 214. 
 
 Von Eschwege, 501, 515, 535, 559 
 
 Von Tschudi on Botocudo tribes, 592 : on dis 
 tribution of cocoa-palm in Brazil, 119 ; on 
 the Colony of Santa Leopoldina, 77. 
 
 W. 
 
 Ward, Mr. Thomas, 277, 279, 501. 
 
 Waste of land at mouth of the Amazonas, 491. 
 
 Water, scarcity of, in interior of Bahia, 256. 
 
 Wax of Carnahuba, 453. 
 
 Wave action, contrast between the rock sur 
 faces produced by wave action and those 
 formed by the glaciers, 342 ; during sub 
 sidence, drift not referable to, 563, 567; 
 during rise of land, drift not referable to, 
 563, 565. 
 
 Weddel,541. 
 
 Widmannstadtian figures shown by Bemdego 
 meteorolite, 326. 
 
 Weiss, 530. 
 
 Wet season, few animals seen, 155. 
 
 Whale, flesh of, used for food, 185. 
 
G20 
 
 INDEX. 
 
 Whale fishery of Brazil and of the Abrolhos, 
 181 ; duration of fishery, 182; Amia^oes or 
 trying hours, 183 ; launches and boats used 
 in, number of men employed, wages, diffi 
 culties of fishing, imperfect methods, 183 ; 
 at liahia, 185. 
 
 Whalebone, 184. 
 
 Wheat, 142. 
 
 White water rivers, 223. 
 
 White ants, hills of, 257. 
 
 Winds of campos, 254 ; on Sao Francisco be 
 low falls, 421. 
 
 Wilson, Mr Hugh, 269, 338. 
 
 Williams, Mr. C. II., on Indian rock drawings, 
 326. 
 
 Williamson, Mr. E., on the geology and gold 
 mines of Purahyba do Norte, 443 
 
 Wounds, facility with which they heal among 
 
 Botocudos, 598. 
 Wyman, Prof. Jeffries, on Botocudo skull, 585. 
 
 T. 
 
 Yellow fever, 337. 
 Ypiranga, limestone at, 15. 
 
 Z. 
 
 Zinc, sulphide of, 448, 472. 
 Zinebra, 172. 
 Zizyphus Joazeiro, 408. 
 Zoanthus, 62, 192. 
 
 THE END. 
 
 Cambridge : Electrotyped and Printed by Welch, Bigelow, & Co. 
 
RETURN TO the circulation desk of any 
 University of California Library 
 
 or to the 
 
 NORTHERN REGIONAL LIBRARY FACILITY 
 Bldg. 400, Richmond Field Station 
 University of California 
 Richmond, CA 94804-4698 
 
 ALL BOOKS MAY BE RECALLED AFTER 7 DAYS 
 
 2-month loans may be renewed by calling 
 (510)642-6753 
 
 1-year loans may be recharged by bringing 
 books to NRLF 
 
 Renewals and recharges may be made 4 
 days prior to due date. 
 
 DUE AS STAMPED BELOW 
 
 MftY 1 4 2001 
 
 12,000(11/95)