yet it is hard to suppose them meant merely for ornament. This edifice, originally in two stories, is a mass of chambers, mostly small, which are connected by narrow passages. The large walled court which adjoins it is adorned by stuccoed niches. The walls are well preserved, but all the ceilings and roofs have gone. There are so few apertures for light that it is hard, as in most of the ancient Peruvian houses to understand how light was admitted. Probably light was sacrificed for the sake of warmth, for the nights are extremely cold, even in summer. Doorways are covered sometimes by a single slab, sometimes by flat stones projecting each beyond the other, so as to have the effect of an arch, but no true arch ever seems to have been found in Peru or anywhere else in the Western Hemisphere. Sacrificial objects, dug up in front of the building, confirm the legend that the place was a shrine of the Moon Mother, but the name by which it has been known is the Palace of the Virgins of the Sun. There may, therefore, have been in conjunction with the shrine one of the numerous establishments in which the Incas kept the women who were sent up to them as a tribute from the provinces, and who, among other things, wove fine fabrics and made various articles needed for worship. The early Spanish writers, with their heads full of Christian nuns and Roman Vestals, called them Virgins of the Sun, but the name was altogether inappropriate, for many of them were kept as concubines for the reigning Inca.
Four miles from Koati and two from the mainland, lies the larger and more sacred Island of the Sun. It is ten miles long, nowhere more than a mile wide, and very irregular in shape, being deeply indented by bays. A ridge of hills, rising in places to one thousand feet or more, traverses it from end to end, and much of the surface is too steep and rocky for tillage. There are many groups of ruins on it, the origin and character of some among which have given rise to controversies into which I need not enter, proposing to describe two only. One of these is the so-called Fountain, or Bath and Garden of the Inca. Two buildings stand on the shore, evidently of a date anterior to the Conquest, and one was probably a royal residence. The most recent and most competent investigators divide them into two classes: those which the Indians call Chulpas, and are the work of an earlier race or races, and those which they ascribe to the Incas, the latter being larger and better built, and accompanied by pottery, weapons, and other relics, indicating a more advanced culture. Hard by a flight of low steps, rising from the water through a grove of trees, leads up to a spot where a rivulet, led in a channel from the hill above, pours itself into a receptacle hewed out of one piece of stone, whence it pursues its course in a murmuring rill to the lake below. The terraced garden on each side is planted with flowers, most of which are the same as those in European or North American gardens; but the brilliant red blossoms of the shrub called the Flor del Inca give a true local colour, and the view over the lake to the distant snows is unlike anything else in the world. How much of the beauty we now see was planned by the unknown monarch, who first made these terraces, and did the spot commend itself to him by the wonderful prospect it commands? Most of the so-called palaces of these isles occupy sites that look across the lake to the great snowy range, but a learned archæologist suggests that this was due not to admiration of their grandeur, but to veneration for them as potent deities so that they might be more readily and frequently adored.
On this majestic range our eyes had been fixed all day long. Its northernmost summit, Illampu, stands more than twenty miles back from the eastern shore of the lake, and more than thirty miles from the Island of the Sun. Thence the chain trends southward, ending one hundred miles away in the gigantic Illimani, which looks down upon La Paz. All day long we had watched the white clouds rise and gather, and swathe the great peaks and rest in the glacier hollows between them, and seem to dissolve or move away, leaving some top clear for a moment, and then settle down again, just as one sees the vapours that rise from the Lombard plain form into clouds that float round and enwrap Monte Rosa during the heats of a summer day. Evening was beginning to fall when our vessel, after coasting along the island, anchored in the secluded bay of Challa, where, behind a rocky cape, there is an Indian hamlet and a garden and stone tank like that at the Bath of the Inca. We landed and rambled through it, finding its thick trees and rustling shade specially charming in this bare land. Just as we emerged from them and regained the lake shore, the sun was setting, and as the air cooled, the clouds that draped the mountains thinned and scattered and suddenly vanished, and the majestic line of pinnacles stood out, glowing rosy red in the level sunlight, and then turned in a few moments to a ghostly white, doubly ghostly against a deep blue-grey sky, as swift black night began to descend.
Early next morning we set off on foot along the track, well beaten by the feet of many generations of worshippers, which leads along the rocky slopes from Challa to the Sanctuary of the Rock. Here are no houses, for this end of the isle is rough and bare, giving only scanty pasture and a few aromatic flowers, but the little bays where the green water ripples on the sands, and the picturesque cliffs, and the vast stretch of lake beyond, made every step delightful. To our surprise we passed a spot where some enterprising stranger had bored for coal and found a bed, but not worth working. One could hardly be sorry, for though fuel is badly needed here, a colliery and its chimney would fit neither the landscape nor the associations. Less than three miles' walking brought us to a place where the remains of a wall cross the island, here scarcely a mile wide, and seem to mark off the sacred part which in Inca days was entered only for the purposes of worship. A little farther, two marks in the rock, resembling giant footprints, are, according to Indian tradition, the footprints of the Sun God and the Moon Goddess, when they appeared here. The marks are obviously natural and due to the form in which a softer bit of the sandstone rock has scaled off and left a whitish surface, while the harder part, probably containing a little more iron, as it is browner in hue, has been less affected by the elements. Then, after ascending a few low steps which seem to be ancient, we came out on a level space of grass in front of a ridge of rock about twenty-five feet high. This is Titi Kala, the Sacred Rock, the centre of the most ancient mythology of South America. Its face, which looks southwest over this space of grass, apparently artificially levelled, is on that side precipitous, presenting a not quite smooth face in which veins of slightly different colours of brown and yellowish grey are seen. At one point these veins so run as to present something like the head of a wild cat or puma; and as Titi means a wild cat in Aymará, and Kala, or Kaka a rock, this is supposed to be the origin of the name Titi Kala, which has been extended from the rock to the island and from the island to the lake.28
The rock is composed of a light yellowish brown rather hard sandstone of carboniferous age, with a slaty cleavage. The back of the ridge is convex, and is easily climbed. From it the ground falls rapidly to the lake, about three hundred feet below. Except for what may possibly be an artificial incision at the top, the rock appears to be entirely in its natural state, the cave-like hollow at its base shewing no sign of man's handiwork. Neither does any existing building touch it. There are, however, traces of walls enclosing the space in front of it, especially on the north side, where there seems to have been a walled-in enclosure; and there are other ancient remains hard by. The only one of these sufficiently preserved to enable us to conjecture its purpose is a somewhat perplexing two-storied edifice, resembling, though less large and handsome, that which I have described as existing on the island of Koati. It is called the Chingana, or Labyrinth, and doubtless dates from Inca times, as it contains niches and other features characteristic of the architecture of that period. The numerous rooms are small, scantily lighted, and connected by narrow passages. A few flowers had rooted on the top of the walls, and I found tufts of maidenhair fern nestling in the moist, dark corners within. All the roofs have perished. There is nothing to suggest a place of worship, so probably the building contained the quarters provided for the various attendants on the religious rites performed here, and perhaps also for the women who were kept near many sanctuaries and palaces for the service of the Sun and the Incas. None of the other ruins is identifiable as a temple, so we are left in doubt whether any temple that may have existed was destroyed by the zeal of the Spanish Conquerors, or whether the worship of the Sun and the local spirits was conducted in the open air in front of the Rock, whose surface was, according to some rather doubtful authorities, covered with plates of gold and silver. In front of the Rock there lies a flat stone which it has been conjectured may have been used for sacrifices. All our authorities agree that the place was most sacred. Some say no one was allowed to touch it; and at it oracles were delivered, which the Spaniards accepted
Lake Titicaca was originally, it would seem, called the lake of Chucuito, from an ancient town on its western shore.
