#r' 
 
 -f >.„-»- 
 
 Bandelier, A. b' 
 
 T< 
 
 basin of Lake Titics.c£ 
 
 FROLi: 
 
 American Gsofrraph; cal society Bullet' 
 
 '7 no. 8. 
 
 L9C5 
 
 np. 
 
 449-460, 
 
Cornell University 
 Library 
 
 The original of this book is in 
 the Cornell University Library. 
 
 There are no known copyright restrictions in 
 the United States on the use of the text. 
 
 http://www.archive.org/details/cu31924101546442 
 
BULLETIN 
 
 : OF THE 
 
 American Geographical Society. 
 
 Vol. XXXVII 1905. No. 8 
 
 THE BASIN OF LAKE TITICACA. 
 
 BY 
 
 A. F. BANDELIER. 
 
 In the heart of the western part of South America, between the 
 i6th and 17th degrees of south latitude and the 68th and 70th of 
 longitude west of Greenwich, lies the extensive water-sheet of Lake 
 Titicaca, at an altitude of 12,466 feet above the level of the Pacific. 
 By air-line it is about 300 miles from the western, and at least 2,000 
 miles from the Atlantic, shore. Peru claims two-thirds of its sur- 
 face, and Bolivia the southeastern third; but, as yet, the boundary- 
 line is rather indefinite. 
 
 Approach to the lake from the west and southwest is quite 
 abrupt as far as Arequipa, where the peaks of the slumbering 
 volcano Misti (19,250 feet), and its northern companion Charchani 
 (20,300), rise like towering monuments. They are, for some dis- 
 tance, the most southerly pillars (Pichu-Pichu, about 18,000 feet 
 high, forming but an elongated crest) of the Peruvian coast Cor- 
 dillera. North of Arequipa a number of summits, part of which 
 are mentioned in dim Indian tradition as formerly active volcanoes, 
 rise to still greater elevations, like the Koro-Puna, the altitude of 
 which is said to exceed 23,000 feet, which would make it the highest 
 mountain on the American continent, so far as known. Crossing, 
 between Misti and Charchani, a wild labyrinth of volcanic debris, 
 a high table-land is reached, on which the railroad has to climb to 
 14,660 feet at the station of Cruces, near Vincocaya. This plateau 
 is cold, bleak, and correspondingly dismal. The vicuna still roams 
 ■over it in small flocks, and human abodes are few and far between. 
 From near the culminating point the column of smoke issuing from 
 the crater of the volcano of Ubinas rises on the southern horizon, 
 
 449 
 
450 The Basin of Lake Titicaca. 
 
 and the dark mass of Ornate, the top of which was blown to atoms 
 during the terrible eruption of 1600 (like that of Coseguina in 
 Central America, and Mont Pele in Martinique), rises up in the 
 neighbourhood of Ubinas with its four prongs, indicating the rums 
 of its crater. 
 
 Descent to Lake Titicaca is gradual. The beautiful lagoons of 
 Sar-acocha and Parinacocha, picturesque, but with very bleak shores, 
 nestle at the base of the snow-clad range of Lampa; then a deep 
 furrow is entered that has some arborescent vegetation, and a 
 number of Indian settlements, haciendas, and the mining estab- 
 lishment of Maravillas. This furrow leads to the marshy shores 
 of the Rio Ramis at Juliaca, a village, the site of which was occu- 
 pied by an Indian settlement at the time of the Conquest. The 
 region now reached belonged to what in primitive times was termed 
 the Collao, and held by Indians of Aymari stock. Since then the 
 Quichua have, under Spanish rule, slowly encroached, and they 
 occupy the territory as far as the northern half of the town of 
 Puno. 
 
 Lake Titicaca is reached by the railroad near the former prin- 
 cipal settlements of the Aymard, Paucar-Colla, and Hatun-Colla, 
 now reduced to insignificant hamlets. Neither were they ever 
 very populous. The idea of a former large population belongs to 
 romance. Statistics prove that the present Indian population of 
 these regions has not diminished; and the evidence adduced in 
 favour of a supposed dense primitive settlement of the land, based 
 upon the countless artificial terraces that streak nearly every slope,. 
 is by no means conclusive. The extent of levels is small, and, 
 where they lie along streams and are therefore susceptible of irri- 
 gation, they are also exposed to overflow; and a large portion of 
 them is occupied by marshes, which, along the Ramis, are often 
 alkaline. For this reason the Indian had to cultivate the slopes in 
 preference, converting them into narrow terraced garden-beds. 
 This system obtains to-day, and many of the andenes (as the 
 Quichuas call them) are modern. Furthermore, the Indian had no 
 fertilizers; neither will he use manure to-day, but prefers to let the 
 soil restore its productiveness through rest. That rest lasts, accord- 
 ing to local conditions, from one to ten and more years,' so that 
 only.a limited extent of the numberless artificial terraces was tilled 
 at the same time, and these are, consequently, no criterion as to- 
 former numbers of the people. 
 
 Already from the heights overlooking the track along the lake 
 shore a glimpse is obtained of the magnificent chain of snowy 
 
The Basin of Lake Titicaca. 
 
 451 
 
 peaks extending along the eastern side of Lake Titicaca from 
 southeastern Peru into central Bolivia, and known, in its Bolivian 
 portion, as the Andes of Bolivia, or Cordillera Real. Along the 
 base of these magnificent Alps of South America, conspicuous 
 through the strong individualization of each mountain and the 
 
 Note. — Ckucuiio is an older name of Lake Titicaca. 
 
 abundance of characteristic glaciers, Lake Titicaca extends from 
 northwest to southeast like a trough, abrupt and very deep along 
 the northeastern, and gently sloping from the western, shore, as 
 far as the main body of the lake extends. 
 
462 The Basin of Lake Titicaca. 
 
 Puno existed, as an Indian village, certainly in 1548; hence, it 
 is likely that the site was occupied already before the Conquest. 
 It now has about 5,500 inhabitants, the majority of whom are 
 mixed with Indian blood. The Spainish settlement was officially 
 recognized as such in 1668 or 1669. It is the Peruvian port for 
 the navigation of the lake to the Bolivian shores. 
 
 Lake Titicaca 'fills a trough about 120 miles long, at the south- 
 western base of the Bolivian Andes. It consists of the main lake, 
 41 miles wide at its greatest expansion, and several large bays, 
 like the lagoon of Chucuito, on which Puno is located, the most 
 northern bay in front of Taraco and Vilque on Peruvian territory, 
 and the extreme southeastern lagoons of Chililaya and Uinamarca. 
 While in the main lake accurate determinations of depths on the 
 west give no figures much in excess of 609 feet, on the eastern or 
 Bolivian side we find 828 and 841 feet. Near the little island of 
 Koa 1,312 feet are recorded. It is certain that about the Island 
 of Titicaca, the promontory of Copacavana, and as far as the 
 mouth of the straits of Tiquina, the lake is very deep; but in these 
 straits, by which the main lake communicates with its southeastern 
 annexes, depths diminish almost suddenly to 230 feet. It looks, 
 therefore, as if Lake Titicaca had been imperceptibly falling during 
 a period of unknown length. The lagoon of Umayo (near which 
 stand the Inca buildings of Sillustani) was formerly connected, 
 if tradition is right, with the lake; now the water has receded 
 from the channels that established connection. 
 
 The only outflow of Lake Titicaca is the Desaguadero, a chan- 
 nel of sluggish waters draining into the lagoon of Poopo, in the 
 southwest, which, in turn, has no visible outlet. There is tradi- 
 tional indication to the effect that the Desaguadero was opened 
 (through some unknown process) at a very remote period, still 
 within the scope of dim memory. It is also possible that, not 
 many centuries previous to the Columbian Era, the lagoon of Uina- 
 marca bathed a part of what are now the ruins of Tiahuanaco. 
 
 The water of Lake Titicaca is brackish, but not unwholesome. 
 We had to drink it for several weeks almost exclusively, on the 
 island of Koati, and felt no bad effects. It is always cold. At a 
 depth of 99 feet Agassiz found its temperature to be -1^° centigrade, 
 at 450 feet 10°. 6. Nobody has, as yet, been able to withstand its 
 numbing effects longer than about a quarter of an hour. 
 
 There is, as indicated, a single outlet to the lake, the Desa- 
 guadero, and the feeders are few, on the surface. The Ramis and 
 Suchez enter from the north, from the east the Rio de Achacache. 
 
The Basin of Lake Titicacd. 453 
 
 None of these rivers is more than an ordinary stream, not large 
 enough for navigation. 
 
 During winter the sky is mostly of an intense blue, the air 
 chilling, while the sun's rays scorch and burn face and hands. 
 Still, thunderstorms occur every month, and snowfalls are not 
 uncommon. In summer a lowering sky often covers the moun- 
 tain ranges, wrhich are the only redeeming feature of the landscape; 
 thunderstorms are of almost daily occurrence, thunderbolts very 
 frequent, and waterspouts not rare. We saw two together, in the 
 middle of the lake, and reliable informers state that as many as 
 five have been observed at the same time. During tempestuous 
 nights St. Elmo fires gleam on the steamers' masts. And, yet, rare 
 is the evening when, for a few hours at least, the Bolivian Cordil- 
 lera does not shine out, even if thin vapour rises before it from the 
 deep gorges at its foot, and seldom is the whole chain, from the 
 Carabaya range, in the north, to Illimani, in the extreme south, 
 completely shrouded. In August, when winter is at its height and 
 the skies cloudless, the Bolivian Andes sometimes display an alpine 
 glow of unequalled splendour. 
 
 Our observations of temperature on Titica.ca Island, carried on 
 for three and a half months in succession, and later ones on the 
 same island and on its neighbour, Koati, indicate a very equable 
 climate. The mean for January, 1895, "^^^ 54- 9°; for February, 
 55.2°; in March, 54.8°. In winter of the same year temperatures 
 were, of course, lower, but the extremes of temperature from 
 January first to July first, 1895, were only 65° and ■^■7,° . I do not 
 lose sight of the fact that such local observations, carried on for a 
 short period only, have no value except for their time and the 
 place where they were taken; but meteorological observations at 
 La Paz extend over a period of many years, and establish the 
 equableness of the climate beyond any doubt. That climate has, 
 in fact, but two seasons — winter, which is cold and less moist, and 
 summer, which is wet and equally cold. One is as disagreeable as 
 the other, though not unhealthy. 
 
 To the coldness of the climate (first of all a consequence of the 
 great altitude) the proximity of the snowy ranges contributes. In 
 the west the few summits that rise above the snow-line between 
 Puno and Moquegua, the Kaualluni, and the Uilcaconga, are neither 
 high nor striking in appearance. The lofty ranges between Cuzco 
 and Juliaca, towering Kunnu-Rona (Santa Rosa), and bold Vilcata, 
 that rises above the source of the Amazon River, at La Raya, are 
 too distant. But the eastern range approaches the lake as near 
 
454 The Basin of Lake Titicaca. 
 
 as twenty-five miles in its two culminating peaks of the Sorata 
 group, the Hilampi, or Hanko-Kunnu, and the Hanko-Uma, rising 
 respectively to 21,300 and 21,490 feet, according to latest determi- 
 nations. The Sorata group terminates the "royal range" {Cordil- 
 lera real) in the north. South of it that alpine chain of unsurpassed 
 beauty in outline and in abundance of formidable glaciers, but 
 without the softening feature of meadows and forests — grand, 
 solemn, and chilling— extends to its southern pillar, Illimani, the 
 altitude of which was determined by Sir Martin Conway, on 
 his first (and thus far ofi/y) ascent, in September, 1898, at 21,190 
 feet. Between Sorata and Illimani the strikingly individualized 
 peaks exceed, as far as known, twenty thousand feet at only two 
 or three places. The Karka-Jake (erroneously called Huayna 
 Potosi) reaches 20,300, and the Chachacomani 20,600 feet. 
 
 Glaciers abound on the Pacific slopes of the chain. The Atlantic 
 declivity is very steep; hence shows few of these truly alpine 
 features. At present, and for a number of years past, the glaciers 
 of the Bolivian Andes have been in slow retrocession. The deep 
 clefts, meandering from the fronts of actual glaciers towards the 
 upper Puna, or tableland, bear every mark of having been eroded 
 by glacial action at a very remote period. 
 
 North of the Sorata group there is a break in the continuity of 
 the snowy range. The Andes of Carabaya lie farther from the 
 lake basin than those of Bolivia, but are visible from the western 
 shores and from Puno. Their relative altitudes are as yet but 
 imperfectly known. Their most important and most northerly 
 group, Ananea and Palomani, may yet prove higher than believed, 
 for both snmmits rise out of enormous glaciers and snow fields. 
 South of them the chain is made up of picturesque clusters of sharp 
 peaks and cones; but the snow-line is not unbroken, as farther 
 south. The eastern slope is, as everywhere else in that region, 
 very steep. 
 
 This extensive barrier, separating Lake Titicaca from the wooded 
 lowlands of the Amazonian basin, acts to intensify the chill in the 
 climate of Titicaca. The Indian, who looks at nature only from 
 the standpoint of the dangerous or the useful — who feels, above all, 
 his own helplessness in presence of incomprehensible phenomena 
 — sees in the stately peaks the abodes of spirits, and worships 
 them as such. The taller the mountain, the more powerful the 
 Achachila (literally, "grandfather ") supposed to dwell in it. To 
 the Indians of the main islands in the lake the principal Nevados 
 appeal mostly as originators of angry blasts and damaging frost. 
 
The Basin of Lake Titicaca. 455 
 
 The region of the lake, as well as the mountains, has as yet 
 been but partially investigated from a geological point of view. 
 The crest of the Bolivian Andes and the upper slopes are Silurian, 
 but here and there traversed by upheavals of syenite, as, for 
 instance, at the so-called Huayna Potosi. Lower down and as 
 far as the middle of the table-land, or Puna, Devonian strata 
 appear. On the west the Permian formation prevails. The basin 
 proper is also crossed by streaks of carboniferous rocks, containing 
 stone-coal of a fair quality, not seldom accompanied by trachytic 
 (andesitic) dykes. The Puna south of the lake is said to be 
 Devonian in the main. Explorations conducted by Prof. A. 
 Dereims, the French geologist, will at last furnish us a true picture 
 of the geology of the region. 
 
 The mineral resources of the surroundings of Lake Titicaca are 
 not inconsiderable. Coal does not, as yet, play any r61e in pro- 
 duction, which is the more regrettable that its quality is good and 
 the seams are of easy acdess. Copper in its native state, as large 
 chunks or blocks, but chiefly disseminated in minute particles 
 through a Permian sand-rock, in proportions varying from less than 
 ten to seventy-five per cent., crops out at Corocoro, near the Desa- 
 guadero, in enormous veins tilted at an angle of about forty-five 
 degrees. Tin is worked in the Cordillera not far south of La Paz; 
 and it is also found at Carabuco, near the lake shore. Silver ores 
 abound- in the district of Sicasica, and gold is now washed on a 
 large scale at the very doors of the City of La Paz. In the mineral 
 wealth lies the future of the lake region and in the industries 
 therefrom resulting; neither the vegetable nor the animal kingdom 
 offers any inducements. 
 
 At an elevation of twelve thousand feet, vegetation, even under 
 the tropics, is reduced to a minimum. Not many spots on the 
 lake-shore are sheltered enough to permit the growth of trees and 
 of low, bushy maize. The surroundings of the sanctuary of Copa- 
 cavana offer a succession of favoured nooks, where-, besides some 
 trees imported by the Spaniards, like the eucalyptus, the wild 
 olive tree {Buddleya coriaced), the kenua {Polylepsis racemosa'), and 
 Sambucus Peruviana, grow in stunted specimens. This is also the 
 case on the southern promontory of Titicaca Island, at Kenuani, 
 and especially in two gardens on the island erroneously ascribed to 
 the Inca while they are both Spanish and of the eighteenth cen- 
 tury if not later. On these two lovely spots an abundance of 
 European flowers and some vegetables, also strawberries, still 
 thrive under the shade of imported and indigenous trees. On the 
 
456 The Basin of Lake Titicaca. 
 
 peninsula of Huata similar spots are found; elsewhere vegetation 
 is limited to the dismal Puna grass {Stipa Ichu) and smaller shrubs, 
 some of which are used by the Indians in medicine and witchcraft. 
 
 Alimentary plants may be divided into two classes — indigenous 
 and imported. Among the iormer the potato occupies the first rank. 
 It is the leading staple of the mountain region in general. But 
 cold and moisture prevent its preservation in the natural state. 
 Combustible to dry the potato there is not enough; hence the only 
 means to keep it is to let it freeze, then press out as fully as possi- 
 ble every trace of moisture, let it freeze again, and thus, reducing 
 the tuber to a mere cork-like, insipid mass, render it invulnerable 
 against cold and humidity. The chuna, as this unpalatable meta- 
 morphosis of the (otherwise splendid) potato is called, is the main 
 food on these highlands, and of all classes of society, without 
 exception. Next to the potato comes the oCa {Oxalis tuberosa), the 
 quinoa (Chenopodium quinoa), maize, and the so-called papa liza, a 
 most indigestible tuber of the potato kind. Imported- alimentary 
 plants are limited to barley and to a coarse, large bean, the haba. 
 Vegetables would grow very well on the islands. We have seen, on 
 Koati, plants that were almost arboriferous ; but the Indian is far 
 too conservative to change his diet. He prefers his primitive food, 
 heavy and yet not nutritious, to wholesome aliments unknown to 
 his ancestors. 
 
 Neither hill nor vale is without flowers. Along the lake and in 
 sunny coves the kantuka — carmine, red, yellow, and white — droops 
 from the branches among the vivid green foliage of a tall shrub. 
 On humid expanses the panti-panti displays its white or pink blos- 
 soms. Verbenas cover the slopes in patches. Many an insignificant 
 bush has a modest but pleasing little flower, and the quinoa pre- 
 sents, when in bloom, entire fields draped in the national colours of 
 Bolivia — red, yellow, and green. On the other hand, miles and 
 miles offer nothing to the eye but, along the shores of the lake, 
 lonely gray or dull green, with rocks intervening, or, in the inte- 
 rior, interminable levels of coarse Puna grass, dotted by the adobe 
 or stone huts of the natives, their tiny fields, and an occasional 
 drove of llamas. 
 
 Animal life is irregularly distributed. The middle of the lake 
 sees occasionally a gull {Larus serranus) following the steamer. Of. 
 the six kinds of fishes thus far known few specimens appear on the 
 surface. It is on the shores of bays like the large one of Acha- 
 cache, on the waters of the picturesque harbours of the island of 
 Titicaca and the peninsula of Huata, also on the roadsteads of 
 
The Basin of Lake Titicaca. 457 
 
 Chililaya and others, mostly where the tatora or lake reed {Mala- 
 cochaete Tatora) thrives in 'shallow places, that aquatic birds of the 
 lake are mostly seen. The Q.)xo\:z.\Fulica gigantea), a stately 
 water-hen of black metallic plumage, with bright-coloured head and 
 bill, glides quietly over the little ruffled surface of inner bays. 
 Turbulent divers {Podiceps, Tachybaptus, and Centropelma) furrow 
 the water around them like diminutive tug-boats, crossing and 
 chasing each other with amazing rapidity. Along the mainland 
 the flamingo and the rose-coloured spoon-bill are not rare, and a 
 dark-green ibis stalks over marshy expanses. An ashy gray night 
 heron is sometimes seen erect on rocky points, and the black cor- 
 morant peoples smaller islands, affording the Indians, who visit 
 them in frail balsas, a repulsive aliment in young birds and eggs 
 with greenish yolks, both equally disgusting through their fishy 
 smell and taste. A gray eagle soars over the rocks skirting the 
 shores, and that handsome scavenger, the alkamari, stalks in pairs 
 over fields and bare slopes. I forbear mentioning enigmatical 
 beasts like the supposed seal, although there is enough in reports 
 about the animal to excite legitimate curiosity. Neither do I dwell 
 on the innocent toads, which the Indians gather and expose on dry 
 rocks as often as they fear drouth. I also pass by the small scor- 
 pions, centipedes, and spiders, often disturbed by the opening of 
 ancient sepulchres. But I cannot overlook the godfather of the 
 Island of Titicaca, the species of wild-cat to which the name is 
 due. Titi means the lynx-like feline haunting the shores of the 
 Lake, from which, probably, it sometimes found ways and means 
 of crossing over to the island. Karka (whence Kaka) signifies a 
 rock, or cluster of rocks. The strangely-situated cliff on the 
 island, which was a shrine since time immemorial, is called Titi- 
 Kala, or stone of the cat, since there are on its vertical face some 
 concavities recalling heads of cats, on which the Indians looked 
 with superstitious awe. Hence the wild-cat (called Mulu-Mulu 
 along the slopes of Illimani) has given its name to the main island, 
 whence, after the sixteenth century, it became generally applied to 
 the whole lake. To-day still, the Aymard call the water-sheet 
 simply kota—\ak& or sea. The Titi pays disagreeable attention to 
 domestic fowl. During the seventeenth century its occasional 
 appearance on the island gave rise to the well-worn tale of the car- 
 buncle a cat with fiery eyes and a luminous stone on its forehead. 
 In former times the vicuna and even the guanaco were occasional 
 visitors of the lake-shore, and they roamed through the districts 
 west and south of the lake. The eastern shore, as well as the 
 
458 The Basin of Lake Titicaca. 
 
 Andes, were and are, especially the latter, frequented by the vicuna 
 only. What we were told' about the occurrence of the guanaco is 
 problematic. I would also allude here to the not uninteresting fact 
 that the condor, that huge bird of prey and carrion, while roost- 
 ing in the Bolivian Andes about lUimani, is hardly ever a visitor of 
 the lake or the plateau, whereas the easily-domesticated Puna 
 goose {Benicla Melanopterd) lives on one of the haciendas as tame 
 as any ordinary member of the "civilized " species, and we had the 
 daily visit of a pair of these handsome birds on the sandy shore of 
 Challa, on the island. 
 
 Both the llama and the paca are kept in large numbers around 
 the lake — the former as a beast of burthen, the latter for its pre- 
 cious wool. Along the base of the northern range, on the dismal 
 and cold levels that skirt the snowy peaks of Suchez, of Katantika, 
 Kololo and Altarani and Ca, the paca pastures by thousands. They 
 do not, however, go up near the mountain passes, where, as all 
 along the snowy range, emerald-green lagunes, fed from the gla- 
 ciers, offer a far better drinking water than any on the Puna. The 
 vizcacha accompanies the vicufia to the highest crests, but it also 
 makes its home on the Puna, wherever clusters of rocks protrude. 
 So does the chinchilla, on the Puna nearer the Chilian frontier. 
 
 Since the Spanish occupation the Indian has additional domes- 
 tic animals. He has the ass, the mule, the horse, and cattle and 
 sheep. But he takes no care of them; hence the breed is poor and 
 weakly. The llama and its kind, however, he fondles, for it is a 
 time-honoured associate. 
 
 Outside of the city of Puno and a few unimportant villages 
 along the lake, the white population hardly exists. Indians — 
 Quichua in the north and northeast, Aymard to the south, and 
 Mestizos — are the overwhelming majority. Intercourse with them 
 — protracted contact — is far from agreeable, especially with the 
 Aymari, who are a retrograde, stubborn, and naturally blood- 
 thirsty stock. Distribution of the settlements is unequal. After 
 sallying from the lagoon of Chucuito, where the village of the same 
 name and that of Acora are the chief places, we pass successively Juli, 
 Have, Pomata, and Yunguyu, leaving Zepita in the distance. All 
 these villages together, and their surroundings, have to-day, accord- 
 ing to Peruvian statistics, as many Indians as at the time of the Con- 
 quest. The Peruvian coast is low and monotonous; it fronts the 
 widest part of the lake called Pampa de Have, where storms are 
 much feared. Already there, the largest island of the lake, Titicaca, 
 appears, seemingly connected with the peninsula of Copacavana 
 
The Basin of Lake Titicaca. 459 
 
 facing it. As we approach, Titicaca detaches itself, assuming its 
 true insular shape. It is about seven miles long, and has the form 
 of an elongated toad. The lake makes not less than seventeen well- 
 marked indentations into its shores, some of which are most pictur- 
 esque bays. The backbone of the island is a long ridge, but the 
 highest points lie outside of it— northwest and southeast. The 
 former, Chullun-Cayani, overlooks the lake directly ; the other, Pal- 
 Jasa, is off the shore, and both are 800 feet above the lake. The 
 jsland is one of the most romantic spots imaginable, notwithstand- 
 ing lack of vegetation. We lived on it for over five months in all, 
 and, with all the inclemency of climate and the evil disposition of 
 the eight hundred Indians who live on the two haciendas into which 
 the island is divided — with all their constant disobedience of most 
 positive orders from the owners to respect and attend us as if we 
 had been the proprietors themselves — this marvellous island, with 
 its coves, its promontories, its unparalleled panoramas and outlooks, 
 its modestly picturesque ruins, remains, and will always remain, the 
 ideal of our dreams. 
 
 It is as if a refiex of its beauty had been cast on the peninsula of 
 Copacavana, from which it is separated by the narrow channel of 
 Yampupata. The sanctuary of Our Lady of Copacavana, famous in 
 Bolivia since 1586 (although the once beautiful church was only 
 begun in about 1640,), stands on a lovely site ; but, unfortunately, the 
 village of Copacavana is a dreary hamlet, and only alive when the 
 festivals of the Church gather thousands of boisterous Indians to 
 dance in the plaza. Such dances are survivals to which the Indian 
 clings with great tenacity, while their original meaning is only 
 known to a small number of the aborigines. These celebrations, 
 while attractive at first sight through quaintness and colour, become 
 disgusting very soon through the carousals which form, and always 
 formed, an indispensable corollary. 
 
 Copacavana has many natural charms. The triangle which the 
 peninsula forms is as diversified in topography as odd in outline. 
 Its Indian population is comparatively dense, of evil inclinations, 
 and does not appear to have suffered any diminution since 1538. 
 Tillage of the soil on a small scale and some cattle-raising give 
 them occupation when their time is not taken up by feasting and 
 carousing; and so it is all over Bolivia and the mountainous section 
 
 of Peru. 
 
 Few lakes in the world have a greater number of islands and 
 islets than Titicaca. Entrance to the bay of Chucuito is guarded 
 by two large ones — Amantani and Taquili, of which the latter has 
 
460 The Basin of Lake Titicaca. 
 
 served repeatedly as a place where political prisoners were left to 
 pine. Beyond the main island of Titicaca, towards the Bolivian 
 end of the lake, lies Koati— green, almost treeless, but notorious 
 for its Inca ruins, not extensive nor architecturally remarkable, but 
 well preserved and on a spot of unusual beauty, selected by the 
 Indians for its warmth and because it faces the majesMc summits of 
 the Sorata chain, which are, to the Indian, very "big medicine." 
 
 Around Titicaca Island are a number of small isles, that appear 
 like remnants of a former connection between the peninsula of Co- 
 pacavana and that of Huata. The straits of Tiquina, very narrow, 
 but not deep, and with steep, rocky shores, separate the two pro- 
 montories. They also separate the main lake from the southern, 
 comparatively very shallow, basins of Chililaya or Huarina, and of 
 Uinamarca, at the southwestern end of which opens the Desagua- 
 dero channel. Between the two basins lie a number of not unpre- 
 possessing islands, mostly inhabited. There are about thirty of 
 them, many of which bear ruins of primitive settlements. South of 
 the lake expands the bleak Puna, between twelve and' fourteen 
 thousand feet above the Pacific, and, at a distance of forty miles 
 south of Chililaya, the cleft of the La Paz River opens, through 
 which that stream, hugging the mountain notch of Illimani, mean- 
 ders to the basin of the Amazon. 
 
 Of the pre-Columbian condition of the inhabitants of Titicaca 
 basin, many ruins, promiscuously scattered, bear testimony. Aside 
 from those of Tiahuanaco, the past of which is still shrouded in 
 mystery, ruins of Inca origin in the shape of more or less elaborate 
 storehouses of stone and adobe, and ceremonial edifices of modest 
 size on the Island of Titicaca, on Koati, and around Copacavana — 
 finally the many vestiges of Aymara occupation in the so-called 
 Chullpas or dwellings with house-burial, on the Puna as well as in 
 the mountains — show that, in pre-Spanish times, the lake region 
 was occupied almost exclusively by Aymara Indians. These had, 
 to a limited extent, been overcome by the Inca tribe of Cuzco, but 
 never incorporated into a national or political union, of which no 
 aggressive Indian tribe in America had any conception. These ruins 
 indicate a population not in excess of the present number of Indiansy 
 but of a shifting disposition, to which Spanish rule put an end. The 
 influence of the Incas was limited to levying tribute, and their 
 architectural remains in the lake basin indicate permanence only 
 in very few places.