Cusco

The firm part is that the Inca themselves used the puma as an image for their city. Sarmiento records it, and the place name Pumachupan has survived.⁹ Rowe's reading drew on Betanzos's and Sarmiento's naming of the city's parts.¹⁷

Several things counsel caution. Sarmiento says Pachacuti "called" the city the lion city. That reads at least as easily as a figure of speech as it does a surveyed outline. In the same passage the head is missing until the next ruler adds Sacsayhuamán, so the "puma" was not drawn in one stroke.⁹ Archaeological work also points to slower, more organic growth of the city, starting before Pachacuti.² The astronomer Giulio Magli, who discusses the idea, presents it as "a long-standing tradition."¹⁸ Other writers have proposed different animals, such as a llama, for the same plan (see Sources to check below).

The images of a cat laid over a map of Cusco that appear in guidebooks are modern drawings. A fair summary is that the puma was an Inca way of speaking about the city, and perhaps of thinking about it. Whether the streets were laid out to trace a puma's body is unproven.

What is known of the Qorikancha's interior comes from three kinds of source, and they deserve different amounts of trust. The earliest are Spanish accounts from the 1530s to 1550s, by men who saw the temple or spoke with those who did. The fullest is the Royal Commentaries of Garcilaso de la Vega, published in 1609 by a writer born in Cusco in 1539 to an Inca noblewoman and a Spanish captain, working in Spain from boyhood memory. The third is the archaeology of what survives, which is less than the walls suggest.

The plan. Pedro Cieza de León, who was in Cusco around 1550, describes an enclosure "more than four hundred paces in circuit," surrounded by a strong wall of finely fitted stone, with a band of gold running round it halfway up.²⁴ Inside, buildings with thatched roofs stood around a central courtyard. Garcilaso puts the gold band at the top of the wall, as a cornice more than a yard wide. He adds that the Spaniards later replaced it with a band of white plaster "in memory of the former one," which he saw.²⁵ The western end of the enclosure is the curved wall that now carries the church. It stands higher than the rest because it is built against the slope of the hill, like a balcony over the ground below.²⁶ Below it, terraces step down to a sloping lawn above Avenida El Sol. This was the friars' vegetable garden in Garcilaso's day.²⁵

12

Water ran through the complex. Garcilaso writes that there were five fountains with gold pipes, used to wash offerings. Only one still worked in his youth, watering the convent garden, until a flood in 1558 broke its buried pipe.²⁵ In one of the surviving rooms a block has three holes cut through it, often read as a drain for rain or poured offerings of chicha (maize beer).¹⁹

The shrines. Garcilaso gives the room-by-room account that guides still follow. The main temple held a gold image of the Sun, a face with rays made from a single thick plate, wide enough to fill one wall. The mummies of past rulers sat on gold seats on either side, with Huayna Capac facing the image. Around a cloister stood five separate halls with pyramid-shaped roofs. The hall of the Moon, Mama Killa, was lined with silver and held the mummies of the queens. A second silver hall belonged to Venus, which he calls Chasca, "the one with long and curly hair," and to the Pleiades and the other stars, with a ceiling strewn with stars. A gold-lined hall honored Illapa, a single word for thunder, lightning and the thunderbolt. Another gold hall had a rainbow painted across its plates. The fifth was a meeting hall for the high priest, the Villac Umu, and his assistants.²⁵

The earlier writers are less tidy. Cieza counts four houses, with the Sun's figure in the richest one.²⁴ Garcilaso insists the Sun was the only image in the temple. Sarmiento de Gamboa, writing in 1572, says Pachacuti also made gold images of the creator Viracocha and of lightning, Chuqui Illa, and set them beside the Sun.⁹

Archaeology can confirm the courtyard plan but not the dedications. Ian Farrington's 2013 synthesis counts nine buildings, eight of them opening onto the patio. Only two survive well enough to read their layout. Two more keep partial walls, and five are known only from foundations.²⁶ The names now posted on the rooms, Moon, Stars, Thunder and Rainbow, follow Garcilaso's list. No inscription or find ties a particular room to a particular deity. Mariusz Ziółkowski and Jacek Kościuk, who surveyed the site with a laser scanner, use plain numbers for the rooms and write of the "so-called" Enclosure of the Stars.²⁶

The golden garden. Cieza describes a garden whose soil was made of lumps of gold, planted with gold maize, stalks, leaves and cobs. He adds more than twenty gold llamas with their young and the herders who watched them.²⁴ Garcilaso goes much further: trees, flowers, snakes, lizards, butterflies, birds, a field of maize and quinoa, life-size men, women and children, and gold spades and hoes for the gardeners.²⁵ Neither man saw the garden. Cieza relied on Spaniards who had seen the temple in 1533, and Garcilaso was born six years later. Some modern summaries add jaguars, guinea pigs and monkeys, or say that a single gold stalk survives. These details do not appear in either early account.²⁷²⁸ Some kind of gold garden is plausible, but its size and contents rest on these two writers.

The Punchao. A second and better-attested Sun image was the Punchao, "day." Juan de Betanzos describes it as the size of a one-year-old child.²⁶ When Spanish troops took the last Inca stronghold at Vilcabamba in 1572, they captured it along with the young ruler Túpac Amaru. Viceroy Francisco de Toledo described it that year as cast gold, hollow, holding powder made from the hearts of earlier rulers, and ringed with small gold plates that dazzled anyone looking at it.²⁶ The historian Catherine Julien argued that Toledo carried it to Spain in 1581, and that a gold idol taken from his estate after his death may be the Punchao, kept in the royal treasure house in Madrid into the early 1600s.²⁹ After that it disappears. Garcilaso's Sun image, wide enough to fill a wall, seems to be a different object. He says it fell to the conquistador Mancio Serra de Leguizamo, who lost it at cards in a single night.²⁵

Ruin and rebuilding. The stripping began in 1533, when three Spaniards came to Cusco to collect gold for Atahualpa's ransom. Pedro Sancho, Pizarro's secretary, recorded "five hundred-odd plates of gold torn from some house-walls in Cuzco," the smallest weighing four or five pounds. They also brought a gold seat and a gold fountain.²⁰ The figure of 700 plates, common in modern accounts, comes from later retellings.³⁰²⁸ The building stayed standing: in 1537 the Almagro faction used it as a prison for two of the Pizarro brothers.²⁶

12

Francisco Pizarro's brother Juan received the temple and handed it to the Dominican order. Sources disagree on the date. Smarthistory gives 1534.²² The Dominicans' own history gives 1538.²³ Juan Pizarro died in 1536, which makes the later date hard to fit.³¹ The friars built their church on the Inca walls and consecrated it in 1633. The earthquake of 1650 brought much of it down, and it was rebuilt.²³¹⁹

The earthquake of 21 May 1950 collapsed the bell tower and cracked the apse. It also exposed Inca masonry that the convent had hidden for four centuries.²³⁵ A UNESCO mission led by George Kubler came in June 1951 to advise on repairs. One of its members was the Peruvian architect Oscar Ladrón de Guevara.⁵ He then directed the restoration of the Qorikancha and Santo Domingo for years, and his plans are dated 1960 and 1963.²⁶ Colonial walls that covered Inca rooms were removed. Missing parts of the Inca walls were rebuilt by reassembling original stones, a method called anastylosis. The tower was rebuilt around a hidden concrete core, and the main work ended in 1976.²⁶²³³² Photographs of one niche from the 1940s, set beside its present state, show how much was put back.²⁶

Digging has continued in the garden below the church. A 1981 photograph shows excavations there.³³ A small underground site museum now sits beneath the lawn. Guidebooks say its objects come from excavations in the early 1990s.³⁴

Where it stands. The general layout is genuine. So are the curved wall and long stretches of the enclosure and room walls. The rest is a mix of Inca stone, colonial building and twentieth-century reconstruction, and in places the joins are hard to see.²⁶ The gold is gone, and its descriptions rest on a handful of Spanish writers. The room names come from Garcilaso, writing from memory in Spain. The Punchao, the one sacred object traced past the conquest, was last recorded in Madrid.

Everything known about the ceques passes through one document. Cobo's 1653 list of the shrines was copied from an older manuscript that no longer survives.³⁶ Cobo built most of his account of the Inca from earlier writers, often without naming them, and the four chapters on the shrines of Cusco read like one such borrowed block.³⁷ The translator Roland Hamilton found Cobo's own handwritten manuscript in Seville in 1974. Rowe, working from it, judged that the odd spellings of shrine names came from Cobo's source, a poor copy by someone who knew little Quechua.³⁷ Hamilton later translated the religious books of Cobo's history into English as Inca Religion and Customs (1990), the version most English readers use.³⁸

Who wrote the lost original is disputed. Zuidema credits a text of 1561 by the jurist Juan Polo de Ondegardo, or written at his instigation, and argues that Polo translated what Inca specialists read off the cords of their quipus, the knotted-string records of the Inca state.³⁹ In his 1964 book he had suggested a different author, the priest Cristóbal de Molina.⁴⁰ The archaeologist John Rowe, who had once proposed Molina himself, rejected both names in 1981. The list mentions the parish "of the Martyrs," which existed only between 1559 and 1572, so it was written in those years. Polo said he had found more than 400 shrines and had drawn up a chart of them, while the list is a written narrative with about 332 entries. Polo also wrote in 1571 that he had never found the Creator's temple in Cusco, and the list gives its location.³⁷ Bauer, following Rowe, also sets the Polo attribution aside and writes that the author and date of the original manuscript have not been clearly identified.³⁶

The list is organized in three steps. It names a suyu, or quarter, then each ceque in that quarter, then each huaca on the ceque in order, from the one nearest the Qorikancha to the farthest.³⁷ Rowe gave every shrine a short code: the first two letters of the quarter, the line number and the shrine number. Ch-1:2 is the second huaca on the first line of Chinchaysuyu, and Cu-13:3 the third on the thirteenth line of Kuntisuyu. Bauer and most later writers use these codes, and Bauer's book prints a table setting them beside Zuidema's older scheme, which labeled the quarters I to IV and each line by group and rank.³⁷⁴¹ Rowe also noted that this kind of ordered list is well suited to a quipu, a point on which he and Zuidema agreed.³⁷

Rowe showed that the list is incomplete. A partial list of Chinchaysuyu shrines by another priest, Cristóbal de Albornoz, has 12 of its 35 shrines missing from Cobo. Some lines begin far out in the valley, so their first shrines near the temple were probably lost. Two shrines have no names, and at the end four Chinchaysuyu shrines are added because they had not been recorded on their lines.³⁷ The familiar total of 328 is therefore a count of a damaged record, and the number of lines, 41 or 42, also depends on how the sources are read.⁴⁸

The shrines range from the small to the very great. Many were springs, boulders, caves and canals.⁸ Michosamaro (Ch-1:1) was a stone said to be a companion of the first Inca, Manco Capac, who was killed by one of the women of their party and turned to stone. Guaracince (Ch-2:1), in the temple square, was where earthquakes were thought to form. Racramirpay (Ch-2:2) was a stone set in a window, said to be a warrior who appeared in the air to help Inca Yupanqui, the ruler later called Pachacuti, win a battle. Fourteen entries are pururaucas, stones that the Inca said had risen as soldiers to fight the Chanca, a rival people who attacked Cusco. Three shrines on the ninth Chinchaysuyu line mark places in that war, and Cusicancha (Ch-5:1) was the house where Inca Yupanqui was born.³⁷

Royal places were shrines too. Aucaypata (Ch-5:4), the main square, received the "universal sacrifice" for the Sun and the other huacas. Sabacurinca (Ch-5:6) was a finely carved seat where the Incas sat, and through it the whole fortress of Sacsayhuamán was worshiped. Tambomachay (An-1:9) was a house where Inca Yupanqui lodged when hunting, flanked on its line by two springs.³⁷ An-5:1 was a stone called usno in the lower part of the plaza, the first huaca to which young nobles made offerings at their coming of age. An ushnu was a ritual platform or pillar used for drinking to the Sun, and Rowe notes that the great ushnu of the main plaza is itself missing from the list.³⁷

At the other end of the scale stands Huanacauri (Co-6:7), a mountain of 4,089 meters southeast of the city and one of the most important huacas of the Inca. The chronicler Guaman Poma, drawing around 1615, set its name beside Inti, the Sun, among the "idols" of the Inca.⁴² Cobo's list calls it the oldest shrine of the Inca after the cave of origin at Pacaritambo, where one of the first Inca's brothers turned to stone. The Spaniards took its gold but ignored the plain stone, which the Inca hid and kept.³⁷

The first big theory was social. Zuidema's The Ceque System of Cuzco (1964) treated the list as a map of the social organization of the capital.³⁵ The four quarters fell into two moieties, or halves of the community: Hanan (upper) Cusco held Chinchaysuyu and Antisuyu, and Hurin (lower) Cusco held Collasuyu and Kuntisuyu.⁴⁰⁸ In each quarter the lines came in groups of three, ranked as collana, payan and cayao. Collana means "first" or "most prominent"; payan and cayao were the middle and lowest ranks.⁴⁰⁴³ The list says which kin group served each line. Some were panacas, the descent groups of past rulers, formed by each king's descendants except the heir; others were ayllus, the ordinary Andean kin and land-holding groups.⁴⁰ Zuidema read the pattern of ranks and groups as a model of Inca kinship and politics. He also argued that the line a group served did not show where it lived.⁴⁰ Zuidema returned to the theme in Inca Civilization in Cuzco (1990), which covers the valley's districts, irrigation, age classes and calendar.⁴⁴

The second theory was about time and the sky. Zuidema proposed that the 328 huacas stood for the days of a ritual year.⁸ 328 is twelve sidereal months of 27⅓ days, the time the Moon takes to return to the same place among the stars, as opposed to the synodic month of about 29½ days from one full moon to the next. He counts eight days on average per line (41 × 8 = 328) and the remaining 37 days of the solar year as a period outside the count.³⁹ He and others link those 37 days to the time when the Pleiades, a star cluster closely watched in the Andes, are hidden in the Sun's glare.⁴⁵ With the astronomer Anthony Aveni he measured horizon alignments around Cusco in 1976 and 1980. They argued that some lines ran straight to points on the horizon and worked as sight lines from the Qorikancha.⁴⁶⁴⁷ The list itself gives support for a few cases. Sucanca (Ch-8:7) was a hill with two markers, and when the Sun reached them it was time to plant maize. Chinchincalla (Cu-13:3) had two more.³⁷ Catachillay (Ch-8:10), a spring on the same line as Sucanca, has a name also used for the Pleiades.⁴⁷

Bauer and the astronomer David Dearborn, in Astronomy and Empire in the Ancient Andes (1995), reached conclusions that differ in several points from Zuidema's and from Aveni's.⁴⁸ They accept that the Inca watched the Sun and took a real interest in the Pleiades. They found less evidence for sight lines along the ceques and argued that the shrines on many lines stray too far from a straight path to guide the eye. They also noted that the Pleiades vanish for 40 to 50 days in practice, which loosens the link to 37.⁴⁷ Zuidema reviewed the book critically in 1996.⁴⁹ Zuidema had argued that a ceque could be a sight line with its huacas lying a little to either side, and the list does say one spring (Ch-9:4) stood "somewhat apart" from its line.⁴⁷³⁷ In a 2010 book he wrote that astronomical observation "did not depend on the directions of the ceques."⁴⁶ A Polish team's 3D scan of the Qorikancha, published in 2018, found that a corridor Zuidema and Aveni used points about 22 minutes of arc away from their figure. The interval it marks becomes 57 or 58 days instead of the 55 that his calendar needed.⁴⁶

A third reading came from water. The anthropologist Jeanette Sherbondy went to Cusco in 1975 to study irrigation. She mapped a canal that starts at Tambomachay and runs past the village of Yuncaypata, whose people may not touch a drop, to fields at San Sebastián held by descendants of two royal panacas. She then noticed that Tambomachay sits on the collana line of Antisuyu, served by one of those same panacas.⁴³ Matching her canal map to the list, she argued that the ceques recorded who held which canals and fields. A line served by a kin group included the spring that fed its canal, and neighboring lines marked the edges of its irrigation district. The canals of Chacán, for example, were held by Iñaca panaca, whose line ran through their source, while two other lines marked the district's east and west edges.⁴³ In Andean thought the right to worship a spring went with the right to its water. Sherbondy concluded that the system was a register of water and land titles, which is why the lines are unevenly spaced.⁴³ She also counted about a third of all the huacas as water sources.⁵⁰ Her work was a direct challenge to Zuidema's view that lines and residence were unrelated, though she saw it as confirming other parts of his model.⁴³

The fourth test was the ground itself. From 1990 to 1995 Bauer led a field project to find the huacas.³ His aim, in a 1993 grant summary, was to use archives in Peru and Spain and then field survey to place the shrines in real space, since "many remain unknown."⁵¹ In his first report on the fieldwork, Bauer identified likely locations for 85 shrines and probable routes for nine ceques in Collasuyu, the southeast quarter of the valley.⁵² He found internal inconsistencies, and possibly errors, in the list, and concluded that the routes varied far more than the models allowed.⁵² The Sacred Landscape of the Inca (1998) works through all four quarters, with Rowe's English translation of the list and an index of shrine names by Margot Beyersdorff.⁴¹⁵³ The pattern that came out of the survey is that the shrines set the course of each line, not the other way round. In Bauer's words, "it is the specific locations of the huacas that define the course of the lines and not vice versa."⁴⁷ The lines bend to reach them and generally do not cross one another, and four of them follow the four main Inca roads out of the city.⁸

The four readings do not cancel out. The ceques plainly carried social meaning, since each line belonged to a kin group.⁸ Some huacas plainly had calendar meaning, and the list itself names the planting markers.⁴³⁷ Sherbondy's link between lines and water rights rests on her canal maps, and the survey bears out that many lines track springs and canals.⁸ What remains open is how far the whole system was designed as a calendar, or as a set of straight sight lines, and whether a damaged colonial copy can carry a count as exact as 328.³⁷⁴⁶ Worshippers walked each line to make offerings, sometimes on Inca roads and sometimes straight across country.⁴ That ritual use of the lines is the one point every reading shares.

Key books:

• Bernabé Cobo, Inca Religion and Customs, translated by Roland Hamilton, 1990. The English text of the shrine list and Cobo's account of Inca religion.³⁸
• R. Tom Zuidema, The Ceque System of Cuzco, 1964. The founding structural study of the lines, ranks and kin groups.³⁵
• R. Tom Zuidema, Inca Civilization in Cuzco, 1990. His later synthesis, including the 328-day calendar.⁴⁴
• Brian S. Bauer and David S. P. Dearborn, Astronomy and Empire in the Ancient Andes, 1995. The main critique of the ceque sight-line and calendar arguments.⁴⁸
• Brian S. Bauer, The Sacred Landscape of the Inca, 1998. The field survey of the shrines, with Rowe's translation of the list.³
• Brian S. Bauer, Ancient Cuzco: Heartland of the Inca, 2004. The valley's archaeology from the first hunters to 1532, with a section on the Qorikancha and the ceque shrines.⁵⁴
• Terence N. D'Altroy, The Incas, 2nd edition, 2014. A general history of the empire that sets the ceques among Inca religion and politics.⁵⁵
• John H. Rowe, "Una relación de los adoratorios del antiguo Cuzco," 1981. The Spanish text of the list with Rowe's codes and his case on its date and author.³⁷
• Jeanette E. Sherbondy, "Los ceques: código de canales en el Cusco incaico," 1986. The water reading in article form.⁴³

For centuries the story ran that a tunnel, a chinkana, linked the Qorikancha to Sacsayhuamán. An anonymous Jesuit account of 1594 describes a main passage from the temple, under the bishop's house behind the cathedral, to the fortress about a mile away.⁵⁷ Garcilaso de la Vega and Martín de Murúa describe similar passages.⁵⁸ According to the archaeologists now working on it, the entrances were sealed in the early 1840s on the orders of the prefect Miguel de San Román, after accidents underground.⁵⁹⁵⁸

In January 2025 the Peruvian archaeologists Jorge Calero Flores and Mildred Fernández Palomino, of the Chinkana-Sacsayhuamán Project, announced the results of a survey with ground-penetrating radar and acoustic prospecting, guided by the colonial accounts. They reported a main tunnel more than 1,750 meters long and up to 2.5 meters deep, running from the Qorikancha toward Sacsayhuamán, with three branches toward Callispuquio, the church of San Cristóbal and the Muyucmarca sector of the fortress.⁶⁰⁵⁷

Excavation began on 27 May 2025 under a permit from Peru's Ministry of Culture, at what the team takes to be the original entrance on the Sacsayhuamán side.⁶¹ After 45 days and five excavation units, the team reported radar evidence of cavities up to 2 meters high, and a crack it believes leads to another section.⁵⁹ By 2026 it reported walls and carved stone corridors, five passages linking the two sites, and possible galleries 4 to 9 meters underground, with deeper cavities on the radar.⁵⁸ The team has also spoken of a network of about 8 kilometers.⁵⁹ Calero Flores has said that confirming the network's extent "is not easy," because the passages are deep and hard to detect.⁵⁸

Where it stands, as of 2026: the tunnel has moved from legend to active archaeology. Excavation has found built underground structures at Sacsayhuamán, and the radar shows voids along the line the colonial accounts describe. The full route to the Qorikancha rests so far on geophysics and documents rather than on a passage anyone has walked. No peer-reviewed report has appeared yet, and the published accounts come from the team itself, so the larger figures should be read as the team's working claims.⁵⁸⁶¹

Some popular claims about Inca masonry are true in part, and some are folklore.

• "No mortar." True for the fine masonry of temples and palaces. Most Inca walls used clay mortar.²¹ One early Spanish account even speaks of "a sort of bitumen" in the Qorikancha's walls, though the stones fit so well that no joint could be seen.¹⁹
• "You cannot fit a knife blade between the stones." Spanish writers said much the same: Cieza that a coin could not be put between two stones, Pedro Pizarro that the point of a pin could not.³⁰ It holds for the best walls and not for the many rougher ones.
• "The Inca melted or softened stone." There is a local story that Inca masons, like a small bird, could soften rock with a plant's juice, and a few writers have proposed an acid etching method.²¹ Neither has been shown. Protzen's experiments show that hammerstones and patient fitting are enough.⁶²
• "Someone older built the giant walls." The colonial writers record the Inca and their labor levies building Sacsayhuamán.⁹³⁰ There may be Killke work on the hill,¹⁵ but the great zigzag walls are attributed to the Inca.
• "The walls are earthquake-proof." They have survived better than colonial buildings, but they also carry the scars of at least two strong quakes.⁶⁴

Dean also looks at how the Spanish and, later, the tourism industry laid their own meanings onto Inca stone.⁷