Yearbearers in the Central Mexican Calendar

The 365-day year in central Mexico, known as the xihuitl, incorporated one complete 260-day divination calendar (tonalpohualli), using 20 day signs and 13 numbers, and 105 days of the tonalpohualli were repeated in the last part of the year. The 365-day year had 18 veintenas and an added five-day nemontemi (18 x 20 + 5 days) at yearend.Since the veintenas were counted using the 20 day signs and numbers of the tonalpohualli calendar, there were actually two occurrences of the numbered day sign used to name a specific year, and the second occurrence was usually the «yearbearer.» Based on Aztec models, it seems likely that the year was named for a day sign that coordinated with the last 20-day festival in the year. Each yearbearer was combined with a sequence of numbers running from 1 to 13 to designate one of the 52 years in the Calendar Round. Known as the xiuhmolpilli or «tying of the years,» this cycle of 52 years is similar to our century, but only half as long. Throughout the course of the 52-year xiuhmolpilli, no single yearbearer date is repeated, and because of the way the tonalpohualli calendar integrated with the 365-day year, only four day signs could name the year. With the nemontemi period at yearend, the day sign used to name the year shifted five days later in the tonalpohualli sequence each year, but in the fifth year, the cycle went back to the first yearbearer, but the numeral coefficient increased by four. For example, 1 Reed is followed by 2 Flint, the yearbearer in the next year, then 3 House, 4 Rabbit, and 5 Reed, and the count continued on using all 13 numbers in combination with the four yearbearers to form a cycle of 52-years. The concept of cyclical time is clearly apparent in calendar wheels made in the colonial period to explain the Aztec calendar to Europeans.

Although most show the 260-day calendar or the yearbearer cycle of 52 years, some record the annual festival cycle. The Aztecs ended the 52-year cycle with a festival to bury the «bundle» of years that represent the xiuhmolpilli (^{Milbrath 1997, 202}). Generally the year bundle was buried in Tititl, the 17^th veintena festival in most Aztec calendars (^{Caso 1967, Table 10}). Nonetheless, a number of festival calendars and calendar wheels provide evidence that different veintena festivals could end the year.¹ Veytia Calendar Wheel 5, a Tlaxcalan calendar, shows the year ending in Panquetzaliztli, like the Mixtec system (^{Anders, Jansen & Reyes García 1993}), but the Relaciones Geográf icas from Tlaxcala places the year ending in Izcalli, like the most common Aztec pattern (^{Acuña 1984, 228}; ^{Milbrath 2013, Fig. 1.3}).

Veintena festival names in Tlaxcala and the Valley of Mexico are similar and their placement in the year is roughly the same (^{Milbrath 2013, Tables 2.1, 2.2}). These central Mexican festival calendars share many elements, including the ceremonies and deities honored in specific festival periods. The veintena festivals are closely related to seasonal cycles, but they sometimes incorporate reenactments of «mytho-historical» events. Aztec festivals clearly dramatized events that we consider mythological, such as the birth of the Aztec solar god Huitzilopochtli and the migration legend reenacted in the Aztec Panquetzaliztli festival (^{Schwaller 2019, 104}).²

Seasonal Adjustments to the Festival Calendar

The yearbearers cycle of 52 years is equivalent to 73 tonalpohualli cycles, integrating the two cycles in a continuous sequence (73 x 260 = 365 x 52). No seasonal adjustment was possible in this form of dating because it depends on an unbroken sequence of 260 days. Nonetheless, since the veintena festivals are not included in the central Mexican xiuhmolpilli inscriptions, these could be shifted to follow the seasonal cycle.

A subject of great debate is whether the dates of the festivals were adjusted to account for leap years. Publications by ^{Michel Graulich (1990, 316-24)}; (²⁰⁰²); (²⁰⁰⁸) argue against the inclusion of a leap year, but some form of calendar adjustment is implied by the link between the festival cycle and solar events, as discussed by ^{Anthony Aveni, Calnek & Hartung (1988, 289-90)}, ^{Johanna Broda de Casas (1969)}; (1982, 93), ^{Carmen Aguilera (1989)}, and others ( ^{Jansen & Peréz Jiménez 2017, 17}; ^{Šprajc 2000}; ^{Tena 1987, 68-69}). In 1519, the majority of Aztec calendars began the year in February, and possibly the February date was memorialized in Aztec architectural orientations, such as the alignment at Malinalco to the sun’s horizon position on February 12 and again on October 29, (^{Broda de Casas 2015, 222-23}; ^{Galindo 1994, 129-30}; ^{Šprajc 2001, 355-56}). This type of orientation divides the year into segments of 260 and 105 days, and perhaps could have been used to adjust the position of the first month of the year.

A number of ethnohistorical sources maintain a leap year adjustment existed and the festivals themselves seem to be correlated with seasonal events.^{3 Sahagún (1950-1982, 2: 35; 4: 144)} noted that an extra day was added every four years by extending the nemontemi for six days rather than five. He also suggested that a leap year adjustment took place during a special ceremony held every four years in Izcalli, the last festival of the year (^{Castillo Farreras 1971, 84}). If any such intercalation was made, it apparently was not in 1520, a bissextile year in the European calendar, and no apparent adjustment was made in 1519 or 1521, based on the correlations compiled in ^{Alfonso Caso’s (1967)} extensive study of the calendar. If there was some form of adjustment to the festival dates, it did not impact the tonalpohualli and the associated yearbearers, which ran on continuously. It is not clear how an intercalation could be accomplished without altering the pattern of yearbearers naming the year, but it is possible these shifts were made by changing the yearbearers, as may be the case in the Maya area (^{Milbrath in press}). Perhaps this explains why we see different yearbearer sets chosen at different points in time in Mesoamerican history (^{Caso 1967}; ^{Edmonson 1988}).

Alternatively, the beginning of the year could have changed over time, which may explain why some calendars begin on different months in central Mexico (^{Caso 1967, Table 10}).

For the yearbearer to maintain a relationship with a specific veintena festival, these 20-day periods could not be easily shifted. ^{Caso (1971)} concluded that each year was named for the last tonalpohualli day in the last veintena festival of the year, arguing that under these circumstances, no intercalation was possible because any adjustments to the position of festivals would have affected the continuous tonalpohualli count and the yearbearer dates derived from this calendar (^{Milbrath 2013, 113-14, n.17}). Nonetheless, it is clear that the veintenas show some relationship to the seasonal cycle, and perhaps periodical adjustments kept the festivals in synch with the seasons. The Aztecs could have made such an adjustment during Panquetzaliztli, for ^{Sahagún (1950-82, 4: 143-44; 7: 27)} notes the New Fire Ceremony was timed by the midnight zenith of the Pleiades, which would provide a guide to readjusting the veintenas to the seasonal cycle every 52 years. Adding 13 days to the length of the festival calendar every 52-years would shift the position of the yearbearer, so instead of the yearbearer falling on 1 Reed, as in 1519, in 52 years the yearbearer would be 13 days later in the tonalpohualli, falling on 2 Deer, which is not one of the Aztec yearbearers. Clearly the yearbearers did not change every 52 years, and the yearbearer must have been allowed to drift in relation to the veintenas, so that it would not always be positioned in the same place relative to the veintenas. It should be noted that the yearbearer could shift forward 26 days through the last veintena and the nemontemi and still remain in the same year.

Integrating the Festival Cycle and the Yearbearers

There is considerable disagreement about the positioning of the yearbearers in the year. The Relaciones Geográficas states that the Tlaxcalan year was named for the first day of the year, which is like the pattern recorded in the Codex Magliabecchiano (^{Acuña 1984, 222}; ^{Batalla Rosado 2002}; ^{Boone 1983}). On the other hand, ^{Alfonso Caso (1971, 346, Table 4)} noted that the Aztec yearbearer had to fall in a «significant position,» which he argued was the 360th day of the year, just prior to the five-day nemontemi. In order to align the yearbearer with the last day of the last veintena, his model ended the Aztec year in Tititl, even though his tables compiling actual records of the veintena cycle most often show the year ending in Izcalli (^{Caso 1967, Table 10}). The Codex Telleriano-Remensis, Codex Vaticanus A, and the calendars collected by Durán and Sahagún all name Izcalli as the last veintena of the year (^{Batalla 2002, 225-29}; ^{Quiñones Keber 1995, 150-51}). ^{H. B. Nicholson’s (1971, Table 4)} synthesis of the festivals indicates that Izcalli ended the year, thus the yearbearer was positioned very near the end of the year, but was not the last day of the last veintena festival.Although Nicholson never explicitly discussed the placement of the yearbearer, it would be the last day of Tititl, which is the 340^th day of the year in his alignment of the festivals (Table 1).

I have proposed that around the time of the conquest the yearbearers all coordinated with the first day of Izcalli, the 341^st day of the year, because the veintena festivals actually began a half-day earlier than the tonalpohualli days (Table 2; ^{Milbrath 2013, Table 1.5} ). Positioning the veintena festivals a half day earlier than the tonalpohualli date would place the yearbearer on the first day of Izcalli, if the festival began on the previous tonalpohualli day (compare Tables 1 and 2). Such a pattern would not alter the otherwise sound principles of Caso’s correlation with the year dates, which have withstood «the test of time.»^{4 Caso (1971, 345)} pointed out that a difference of one day in the positioning of the veintenas relative to the European calendar should not be considered significant because, even though the «day» started at midnight for the Europeans, the Aztecs began the day at either dawn or noon. Caso noted that there is some uncertainty about when the day began, but he argued that the day began at noon, based on the Codex Telleriano-Remensis 48v, which says «ellos cuentan el día, desde medio día hasta otro día a medio día» (^{Caso 1967, 53}).⁵ On the other hand, the Dominican friar ^{Diego Durán (1971, 134)} tells us that dawn was considered to be the «birth of the day.» His Franciscan counterpart, friar ^{Bernardino de Sahagún (1950-1982, 2: 141, 143)} also noted that the tonalpohualli day began at dawn, but the festival period for Panquetzaliztli began at midnight and the feast day itself began at dawn of the 20^th day, implying that the veintena began at a different time than the tonalpohualli day. With the tonalpohualli day beginning at dawn or noon, and the veintena festivals at midnight, it seems there could be some «wiggle room» to adjust the festivals by shifting the beginning of the veintena by one day every four years or six hours every year, leaving the yearbearer cycle and tonalpohualli running continuously without adjustment.

It seems that Izcalli, the last veintena festival in the year, actually began up to a half day earlier than the tonalpohualli day, so Izcalli incorporated the day sign and number used to designate the yearbearer as the first day in the veintena (Table 2). ^{Sahagún’s (1959-1982, 2: 159-66)} description of Izcalli in the Florentine Codex indicates that at midnight the priest began to drill the fire. If the Izcalli festival began at midnight it would already be in progress by the time the tonalpohualli day shifted, six or twelve hours later at dawn or noon. This means the yearbearer day sign would correspond to the first day of Izcalli. In Maya dating, there are also some calendar records in which the veintena dates in the year (Haab) are a day earlier than the Tzolkin day, so this patterning has precedents (^{Tokovinine 2010, 17-18}).

Codex Borgia Yearbearer Pages 49-52

We now turn to the Codex Borgia pages 49-52, where the yearbearer festival clearly coordinates with Izcalli, the last veintena of the year. These four pages show the yearbearers 4 House, 4 Rabbit, 4 Reed and 4 Flint, all marking the second occurrence of these dates in the year (Figs. 1a, 1b, 1c, 1d). The iconography here suggests yearbearer festivals were actually yearend ceremonies, with fire drilling rituals like those performed in Aztec accounts of Izcalli, the veintena ending the year (^{Sahagún 1950-1982, 2: 170}). The yearbearer pages (49-52) appear with dates that coordinate with Izcalli, a festival characterized by a fire-drilling ceremony that is shown on these pages (^{Fig. 1, lower right}). The central trees and the skybearers holding up the heavens on the upper right of the Codex Borgia yearbearer pages evoke concepts of world renewal associated with the trees that helped raise the sky after the great flood in Aztec accounts (Historia de los Mexicanos por sus pinturas, ^{Garibay 1979, 32}). Yearbearer ceremonies showing trees and skybearers symbolically reestablish world order at yearend, not unlike the Uayeb ceremonies in Maya codices (^{Bricker & Bricker 2011, 120-42}; ^{Vail and Looper 2015, 202-04}).

Figure 1a Codex Borgia page 49, with scenes labeled and year 4 House and day 5 Movement (scenes 6 and 7) referring to 17 February 1458 during Izcalli, the last veintena of the year (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}). 

Figure 1b Codex Borgia page 50, with the year 4 Rabbit and day 5 Wind referring to 14 February 1471 during Izcalli, the last veintena of the year (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, Anders, Jansen & Reyes García). 

Figure 1c Codex Borgia page 51, with the year 4 Reed and day 5 Deer referring to 11 February 1484 during Izcalli, the last veintena of the year(Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}). 

Figure 1d Codex Borgia page 52, with detail of the year 4 Flint and day 5 Grass, referring to 7 February 1497 during Izcalli, the last veintena of the year (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}) 

Although there are a number of different calendars in central Mexico, most include the same month names and the majority end the year in Izcalli (^{Caso 1967, Table 10}), but the dates for the months in different calendars may vary up to 20 days. For example, the Relaciones Geográf icas records month dates that are 16 days later than those in Nicholson’s synthesis (^{Acuña 1984}; ^{Nicholson 1971, Table 4}; ^{Milbrath 2013, Tables 2.1, 2.2}). In order to discuss calendar dates in the Codex Borgia, I have had to work with one calendar format so I have chosen the festival dates compiled for the Handbook of Middle American Indians. These have been adjusted for their position in near the end of the 15^th century in the year 4 Flint, the year shown as the last yearbearer on Codex Borgia page 52 (Fig. 1d). The dates used here for the festivals in the Codex Borgia employ no intercalation, and because the events are close to the time of the conquest, the differences between these festival dates and the Aztec dates in 1519 are minimal, with or without an intercalation. Fortunately, the yearbearers and xiuhmolpilli dates in the Codex Borgia are not affected by the issue of an intercalation, because they are based on dates drawn from the repeating tonalpohualli cycle of 260 days.

The four yearbearer pages on Codex Borgia 49-52 is followed by page 53 that includes images referencing the central direction, so that on Codex Borgia pages 49-53 are most often described as an almanac representing the five cardinal directions (^{Anders et al. 1993, 261-84}; ^{Batalla Rosado 2008, 445-54}; ^{Boone 2007, 121-32}; ^{Nowotny 2005, 112-21}). Here, the yearbearers record dates with year signs spaced at 13-year intervals, naming the years as 4 House, 4 Rabbit, 4 Reed, and 4 Flint, and taken together they refer to the entire 52-year cycle.⁶ The late date of the codex (^{Aveni 1999}; ^{Boone 2007}; ^{Bricker 2001}; ^{Milbrath 2013, 15-16}) suggests these yearbearers refer to the 15^th century or early 16^th century. Standing alone, the yearbearers mark both the tonalpohualli day naming the year and the year itself, with the date on Codex Borgia page 52b referencing the day 4 Flint in the year 4 Flint, corresponding to January 24, 1497 in the Julian calendar employed until the 1582 Gregorian reform (Fig. 1d; Table 3; add 9 days for the Gregorian calendar equivalent). In a festival calendar with the year ending in Izcalli (XVIII), the day 4 Flint in the year 4 Flint on Codex Borgia page 52b would be positioned as the yearbearer on the last day in Tititl (XVII), as in Table 3. But if the veintena began a half day earlier than the tonalpohualli day, as I have proposed, the yearbearer would fall on the first day of Izcalli, the last 20-day veintena festival in the calendar, as in Table 4.

On the lower left of each page, alongside day signs with the number five, seated deities represent the Macuiltonaleque, easily recognized by the white hand painted across the chin. These tonalpohualli notations do not seem to be the names of the Macuiltonaleque and probably represent Calendar Round dates when paired with the year dates (see detail boxes in Fig. 1). The idea that these day signs with the number five represent dates was first suggested by Anton Nowotny in the 1961 German edition of Tlacuilolli and restated in the English translation (^{Nowotny 2005, 251}). ^{Christine Hernández (2004, 336-56, Table 11.11)} proposed that the year signs with numbers can be paired with day signs bearing the number five, forming xiuhmolpilli dates that refer to specific dates in the 52-year cycle. ^{Boone (2007, 129-30)} lends support by pointing out that these tonalpohualli dates are not the usual names for the Macuiltonaleque, as seen on Codex Borgia pages 47a-48a.

The tonalpohualli dates (5 Earthquake, 5 Wind, 5 Deer, and 5 Grass) all have two occurrences in the 365-day year, but paired with the yearbearers they mark the second occurrence in the year, falling in Tititl.⁷ When the year signs on pages 49-52 are paired with the tonalpohualli date alongside, the date falls midway through the month Izcalli. These tonalpohualli dates consistently fall 14 or 15 days after the yearbearer dates on these pages. For example, on Codex Borgia page 52, the yearbearer date 4 Flint marks both the day and year 4 Flint, and counting forward to the tonalpohualli date 5 Grass leads to the 14th day of Izcalli (Table 3). However, if we presume that Izcalli began at midnight, half a day earlier, 5 Grass would be the 15th day of Izcalli (Table 4). In any case, all the tonalpohualli dates maintain the same relationship to the dates in the European calendar. Using the Julian calendar, these xiuhmolpilli dates are as follows: 17 February 1458 (4 House 5 Motion, 49b); 14 February 1471 (4 Rabbit 5 Wind, 50b); 11 February 1484 (4 Reed 5 Deer, 51b); and 7 February 1497 (4 Flint 5 Grass, 52b), dates that are included in a large set of xiuhmolpilli notations in the Codex Borgia (^{Milbrath 2013, 12, Table 3.1}).

Another set of numbers is represented on the far left, in a vertical column on Codex Borgia pages 49-52. Here twelve red dots mark intervals between day signs in the cells moving from one page to the next, as on Codex Borgia page 49, where 1 Crocodile + 12 days leads to 1 Jaguar on Codex Borgia page 50. This is the standard interpretation of the dots as intervals, counting the number of days between the trecenas (13-day periods) when moving from page to page, but they may also play a second role. In the context of the yearend festivals, they can be used to count forward from dates represented by the day signs with the number five to the yearend and on to the first day of the New Year (^{Milbrath 2013, 35}). For the year 4 Flint, adding an interval of twelve days to the date 5 Grass carries the count to through the yearend and on to the veintena day beginning the next year on 4 Lizard, as in Table 3, or 3 House in the layout that shifts the veintenas a half day earlier, as in Table 4.

The xiuhmolpilli dates on Codex Borgia pages 27-28 and the yearbearers almanacs on Codex Borgia pages 49-52 record dates ranging from 1467 to 1497, and many dates reference significant astronomical events (Fig. 2; ^{Milbrath 2013, Table 3.1}).⁸ The xiuhmolpilli dates on Codex Borgia page 28 span from May 1467 to January 1472, and here images of Tlaloc are associated with dates that correspond to veintena festivals honoring Tlaloc (^{Hernández 2004}; ^{Milbrath 2013, 33-35, Table 2.6}). This helps link veintena iconography directly with the recorded dates, as in the pattern we see on Codex Borgia pages 4952, where the dates are linked to Izcalli, portrayed by fire rituals shown in the lower right of each page (^{Milbrath 2013}, ²⁰²⁰).

Figure 2 Codex Borgia page 28, with year dates that begin on the lower right with 1 Reed, followed by 2 Knife in the upper right, and 3 House in the upper left, and 4 Rabbit in the lower right and 5 Reed in the center (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}). 

On Codex Borgia page 27, the 52-year cycle begins with the day 1 Crocodile in the year 1 Reed, on the lower right, referring to March 26, 1467 and the same tonalpohualli date 260 days later on the winter solstice in 1467 ( ^{December 11; Milbrath 2013, Table 3.1}). This is followed by a xiuhmolpilli date 13 years later in 1 Flint (1480) on the upper right, then another date to the left corresponding to 1 House (1493), and then a date of 1 Rabbit (1506) with the implied cycle projecting forward to the initial 1 Reed year date, but now 52 years later in 1519.

Even though the narrative on Codex Borgia pages 29-46 lacks any year dates, it falls in the same 52-year cycle laid out on Codex Borgia page 27. The yearbearer 4 Flint on Borgia 52 represents the year 1496, the same year associated with the narrative on pages 29-46. This is the year of a total solar eclipse depicted on page 40, one seen all across Mexico (^{Milbrath 2013, 44-45}).⁹ Furthermore, in the year 4 Flint, Venus phases coordinated with important solar dates, transitioning from Evening Star to Morning Star just after the winter solstice, reaching maximum altitude around the spring equinox, and then the Morning Star disappeared around the fall equinox and reappeared as the Evening Star just before the winter solstice (^{Milbrath 2013, Plates 1, 2, 5, 13, 18}). This seasonal pattern would repeat every eight years over the course of a century, but the year 4 Flint had special importance because it was the year of a total solar eclipse, the only one recorded in Aztec annals (^{Aveni y Calnek 1999}).

A review of the most notable interpretations of the Codex Borgia pages 2946 will provide a background for understanding the evolution of ideas related to this unique narrative and my identification of veintena festivals in the sequence. At the beginning of the 20^th century, ^{Eduard Seler [1904-1909] 1963, 2, 9-10, 20, 50, 61}) interpreted the narrative on pages 29-46 as a Venus sequence that started with the disappearance of Venus as the Morning Star and a 77-day interval in Superior Conjunction, when Venus was invisible behind the sun (pages 29-31). Then a 252-day Evening Star phase (pages 32-39) was followed by a 12day Inferior Conjunction phase on pages 40-44 that ended with the rise of the Morning Star on pages 45-46. Although he attempted to work out these intervals based on some of the day signs, iconographic analysis rather than calendrics underpins his study of the Venus phases. Seler’s astronomical interpretation has been largely rejected, even though scholars still rely on many of his deity identifications (^{Boone 2007}; ^{Márquez Huitzil 2017}).

^{Elizabeth Boone’s (2007, 178-210)} discussion of the narrative assigned Venus only a minor role, and she divided the sequence on Codex Borgia pages 29-46 into eight episodes that are steeped in legends of creation cosmology, featuring the birth of the gods and the calendar. Her emphasis on mythological underpinnings for the narrative certainly has merit, but I would argue that the representations of the veintenas in the narrative may also allude to mythological events. This can be seen in the festival of Quecholli, which refers to the Chichimec past, the mythical time of peregrinations, and the origins of cosmic warfare (^{Broda de Casas 1991, 103-05}). As will be seen, Codex Borgia page 45 represents Quecholli, marking the beginning of the dry season, a time for warfare in central Mexico (see Fig. 6; ^{Milbrath 2013, 77, 96, 98, 106}).

^{Juan José Batalla (2008, 413-37)} also described an eight-part division of Codex Borgia pages 29-46, but he emphasized ritual events. Pages 29-32 are interpreted as images of a ritual in the Temple of Cihuacoatl, followed by the second ritual on pages 33-38 in the Temple of the Heavens and the Sacred Bundle, and the third rite involves the sacrifice of the new sun on pages 39-40, while pages 41-42 refer to a fourth rite showing ritual sacrifice offered to Itztlacoliuhqui. The last four pages (43-46) are recognized as separate rituals, but only the last page is related in any specific way to Aztec rituals. Page 46 is interpreted as the New Fire Ceremony, known only from Aztec sources, but this seems unlikely because the Codex Borgia is from the Puebla-Tlaxcala Valley, an area where there is no known record of such a ceremony (see Fig. 7). Indeed, the New Fire Ceremony is a distinctly Aztec ceremony featuring the Aztec tribal god Huitzilopochtli, and he does not appear in any of the images of the Codex Borgia and would not be expected to be present given the origin point of the Codex Borgia in a neighboring valley that was hostile to the Aztecs (^{Milbrath 2013, 118n34}).

Karl ^{Nowotny (2005, 26-34, 267-75)} proposed that the scenes on Codex Borgia pages 29-46 are a series of temple rituals with attendant priests, a proposition first made in 1961 German edition of Tlacuilolli. He argued that the images represent a unified cycle of festivals and rituals, and the days recorded may represent the same position in the 365-day year over a sequence of four years, but he noted that the relationship to the 18 annual festivals could not be determined. Contrary to Nowotny, I think it is possible to recognize veintena festivals in the images on Codex Borgia pages 29-46, but these festivals are positioned at key moments in the seasonal cycle, as will be discussed further in the next section.

Gordon ^{Brotherston (2003)} proposed a specific sequence of 18 festivals is shown on the 18 pages of the Codex Borgia narrative and he compared certain details in the images to the Florentine Codex, arguing that this provides a key to the festival represented on each page. ^{Brotherston (2003, 85)} says:

At precisely the right intervals, there is clear evidence of the Feasts we identified as a template for the cycle as whole.Threads and broom, albeit rudimentary, mark Ochpaniztli (F1.9; p. 33-34); the birds, the arrow, the bag and even the nose-bone of the hunter in Quecholli (F2.3) emerge on page 36 and are carried down the left margin to page 38; falling water poured by Tlaloc at Atemoztli (F2.5) runs off the very bottom of the following page, the last of the obverse (p. 38); the conical cap is worn at Tlacaxipehualiztli (F2.9; p. 42); Tezcatlipoca and the feathered warriors of Toxcatl (Fl.3) dominate page 45; and on the next and final page, the sequence culminates in the pozole pot of Etzalcualiztli (F1.4), at the summer solstice (p. 46).

His evidence seems very tenuous, and often simply wrong. For example, he says a pozole pot on Codex Borgia page 46 is linked to Etzalcualiztli, but the pot is boiling the body of Quetzalcoatl and is clearly not a bean stew consumed in Etzalcualiztli (see Fig. 6). And, he also misses some key elements, such as the image of Xipe Totec on the stone of sacrifice during Tlacaxipehualiztli on page 33 (Fig. 3; ^{Milbrath 2013, 26}).

Recently, Maarten ^{Jansen and Gabina Aurora Peréz Jiménez (2017, 474, 479,523)} have abandoned the lyrical reading of the narrative and the nine-part division they presented in their Codex Borgia commentary (^{Anders, Jansen & Reyes García 1993, 175-245}), and they now link Codex Borgia pages 29-46 directly with the Aztec festival calendar, but they also miss some obvious connections. Even though they recognize Xipe’s sacrifice on page 33 (Fig. 3), they say the Tlacaxipehualiztli festival dedicated to this god is represented on the previous page by a personified sacrificial knife used in that ceremony. In their summary of the festival alignments, they claim to have identified elements of all 18 festivals on pages 29-46 ( ^{Jansen & Peréz Jiménez 2017, 523}). Although they most often assign one festival per page, in some cases two festivals or even four are assigned to a single page using questionable logic. They say page 29 is «introductory,» and then go on to say that pages 30-31 represent a single festival beginning the year (Cuahuitlehua), and page 32 depicts Tlacaxipehualiztli, while pages 33-34 represent Tozoztonltli and Huey Tozoztli. Then pages 35-38 are related to Toxcatl and Etzalcualiztli (Figs. 4, 5). Pages 39-40 are said to depict Tecuilhuitontli and Huey Tecuilhuitl, and pages 41-42 represent Micccailhuitontli and Huey Miccailhuitl, and all the remaining pages similarly represent one festival per page except page 45, which they link to four sequential festivals, all lumped together as Quecholli-Panquetzaliztli-Atemoztli-Tititl (see Fig. 6).

Figure 3 Codex Borgia page 33, showing Xipe’s sacrifice in front of the temple during Tlacaxipehualiztli, the veintena of the spring equinox (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}). 

Figure 4 Codex Borgia page 36, featuring Tezcatlipoca the god honored in Toxcatl, on a bundle with the winds emerging at the beginning of the rainy season (1898 Loubat edition, Wikimedia Commons). 

Figure 5 Codex Borgia page 37, showing Tlaloc the god honored in Etzalcualiztli, standing on rain clouds, while a solar god plays music on the summer solstice (1898 Loubat edition, Wikimedia Commons). 

Figure 6 Codex Borgia page 45, depicting the god of the hunt, Mixcoatl-Camaxtli, honored in Quecholli at the beginning of the dry season (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}). 

Figure 7 Codex Borgia page 46, showing the fire serpents and fire ceremony in Panquetzaliztli, the veintena of the winter solstice (Restoration by Ian Breheny of photographic facsimile by Fondo de Cultura Económica, ^{Anders, Jansen & Reyes García 1993}). 

An analysis of their interpretation of Codex Borgia page 45 shows the flaws in their approach. They describe the central figure on page 45 as Tlahuizcalpantecuhtli merged with Mixcoatl, identifying this god as the hunting god honored in Quecholli, agreeing with an interpretation I published in 2013, but omitting that I also noted a link with Camaxtli, the Tlaxcalan god honored in Quecholli (^{Milbrath 2013, 23, 30}). Then they go on to relate the skull rack just below to the Tititl festival, even though the skull rack also plays an important part in the Quecholli festival (^{Durán 1971, 147-48}; ^{Milbrath 2019}; ^{Olivier 2015, 360, 382, 459, 579}; ^{Sahagún 1950-1982, 2: 138, 180, 185-86}). The banners are said to a sign of Panquetzaliztli, but similar banners appear page 29, which I have interpreted as war banners representing insignia that mark the dry season as the season of warfare. War banners are appropriate here on page 45 because this page correlates with the beginning of the dry season (^{Milbrath 2013, 31, 77, 96, Plates 1, 18}). Their evidence for Atemoztli on page 45 is even weaker, for the only connection they note is that the descending eagle above a river may be related to the 16^th month Atemoztli («Water Descending»; ^{Jansen & Peréz Jiménez 2017, 518}). There is no know precedent for a descending eagle representing descending water and the basin of water probably symbolizes a lake rather than a river (compare the rivers represented on pages 13b, 65b, 67b). Also, they seem to ignore that the eagle god is extracting Tlahuizcalpantecuhtli’s heart, which seems to be the key element in this image (see Fig. 6).

Also on shaky ground, ^{Jansen & Peréz Jiménez (2017, 451, 521-523n123)} relate Codex Borgia page 46 to the new fire ritual in Izcalli, a festival they also identify on pages 49-52 (Figs. 1, 7). Although I have also linked pages 49-52 to Izcalli (^{Milbrath 2013, 42, 105}), the format of these pages differs considerably from page 46. The fire drilling scenes on pages 46 and 49-52 are somewhat similar, but Jansen and Peréz Jiménez do not explain the prominence of the fire serpents on page 46, which are not noted in ^{Sahagún’s (1950-1982, 2: 159-172)} account of the new fire drilled during Izcalli. In fact, the fire serpents correspond to his description of Panquetzaliztli, when the flaming fire serpent was brought down from the pyramid temple as a symbol of warfare (^{Milbrath 2013, 30}; ^{Sahagún 1950-1982, 2: 141-150}).

There are some points of agreement in our interpretations, for ^{Jansen and Peréz Jiménez (2017, 488, 491, 515, 528)} seem to follow my alignment of certain festivals. As in my 2013 book, they link page 36 to Tezcatlipoca’s festival in Toxcatl at the beginning of the rainy season, and they relate page 37 to Tlaloc’s festival of Etzalcualiztli, and they also share my view that page 44 corresponds to Xochiquetzal’s festival in Hueypachtli (^{Milbrath 2013, 29, 86}). But we disagree fundamentally on the organization of the narrative, for they assign festivals in an irregular pattern, sometimes even identifying four veintenas on a single page, whereas I propose that that each page represents a 20-day period, providing a chronological framework for the 18 pages, and I argue that the veintena festivals are only represented a points of seasonal transition and in conjunction with major solar events.

The narrative is not a festival calendar but it makes references to festivals at key points in the solar year to provide a chronological framework using 20-day intervals.¹⁰ At least six festivals are referenced in the scenes, but the images more generally focus on events involving Venus and the seasonal cycle over the course of a year (^{Milbrath 2013, 106-07}). It is noteworthy that the sequence begins on page 29 with the new moon and the disappearance of Venus as the Evening Star after the winter solstice (^{Milbrath 2013, 77, Plate 1}). Veintena festivals are referenced primarily on pages associated with important solar events, such as the spring equinox focusing on Xipe Totec in Tlacaxipehualiztli on page 33, the summer solstice showing both Tlaloc and a solar god in Etzalcualiztli on page 37, and the winter solstice in Panquetzaliztli featuring the fire serpents on page 46 (Figs. 3, 5, 7; ^{Milbrath 2013, 26-31, Plates 5, 9, 18}). Representations of veintena festivals appear in the narrative sequence on pages 29-46 in an abbreviated form because the focus of these 18 pages is on a sequence of astronomical events. Festivals similar to those pictured in colonial Aztec sources are not easy to recognize in this context, but there are scenes that can be related to specific festivals.

Codex Borgia Veintena Festivals

Codex Borgia pages 29-46 are also loosely integrated with the seasonal cycle of festivals, beginning on page 29 with Atemoztli, just following the winter solstice, and closing on page 46 during Panquetzaliztli, the veintena of the winter solstice when the sun’s light was rekindled in a fire ceremony (^{Milbrath 2013, 24, 76}). The festivals represented in the narrative on pages 29-46 are spaced at appropriate intervals, presuming each page represents a 20-day period, as seems to be indicated by the sets of day signs marking 20-day intervals on six of the 18 pages (^{Milbrath 2013, 79, 103-104}). In the sequence represented on pages 29-46, six seasonal festivals can be recognized as a focal event in the central image (pages 33, 36, 37, 44, 45 and 46). Although the 20-day periods of the festival calendar provide a chronological framework, the narrative is not a representation of the festival calendar because it includes references to the annual veintena rituals only at pivotal points in the solar cycle.

In the sequence of festivals on Codex Borgia pages 29-46, the first clear veintena festival appears on page 33, depicting Xipe Totec’s sacrifice in Tlacaxipehualiztli, a spring equinox ceremony celebrated in March (Fig. 3, ^{Durán 1971, 142-48}; ^{Milbrath 2013, 21, 26}; ^{Sahagún 1950-1982, 2: 57-60}). Here Xipe is represented with closed eyes, lying face up on a round stone of sacrifice. Page 33 also shows a sun disk on the roof top, probably representing the sun’s position rising above the temple on the spring equinox in March (^{Milbrath 2013, 26, Plate 5}).

The narrative also refers to seasonal transitions at the solar zenith, marking the beginning of the rainy season during Toxcatl on page 36 (Fig. 4; ^{Milbrath 2013, Plate 8}). Here, Tezcatlipoca appears with a bundle at the center of the page wearing the buccal mask of Ehecatl, the wind god. This seems to be a representation of Toxcatl, the festival of Tezcatlipoca in May, at the beginning of the rainy season when the winds brought rainfall (^{Durán 1971, 426-29}; ^{Milbrath 2013, 22, Plate 8}; ^{Sahagún 1950-1982, 2: 66-77}). The period assigned to Toxcatl on page 36 in my analysis (May 10-May 29, 1496) probably overlaps with the solar zenith when the sun’s apparent declination is equal to the observer’s location in degrees of latitude. In the Aztec area at the time of the conquest, Toxcatl was dated May 5-24, corresponding to the first solar zenith (May 8 in the Julian calendar or May 18 in the Gregorian calendar at the latitude of 19^o 28 in central Mexico).¹¹

The following page features the summer solstice festival of Etzalcualiztli honoring the rain god Tlaloc, pictured with his Tlaloc vessel in the center of page 37, standing on clouds that brought rain during Etzalcualiztli (^{Durán 1971, 430-33}; ^{Milbrath 2013, 29}; ^{Sahagún 1950-1982, 2: 78-90}). On the upper left of the page, a solstice ritual may be represented by a solar god playing music in a flowered temple (Fig. 5; Milbrath 2013, 29, Plate 9).¹²

Codex Borgia page 44 shows a link with the October festival of Hueypachtli, dedicated to Xochiquetzal and Tlaloc, here represented in combination at the center of the page (^{Durán 1971}, 452-56; ^{Milbrath 2013, 29-30}; ^{Sahagún 1950-1982, 2: 23, 131-133}). This scene also incorporates images of the Atamalcualiztli ceremony performed every eight years during October (^{Milbrath 2000; 2013, 30}). In Sahagún’s Primeros Memoriales (f. 154r) this ceremony pictures Xochiquetzal, Tlaloc, and a flowering tree, all elements represented on page 44 ( ^{Jiménez Moreno 1940}; ^{Sahagún 1993}).

And on page 45, we see an image related to Quecholli honoring the god Mixcoatl-Camaxtli, seen here carrying a net bag for hunting, and an atlatl and darts, but the skeletal face and headdress indicate Mixcoatl-Camaxtli is now fused with the god of the Morning Star, Tlahuizcalpantecuhtli (Fig. 6; ^{Milbrath 2013, 96, 108, Plate 17}). The festival of Quecholli in the Valley of Mexico was dedicated to Mixcoatl at the beginning of the dry season in November, but Tlaxcala honored his counterpart, Camaxtli, a closely related god of the hunt (^{Durán 1971}, 142-49, 455-56; ^{Olivier 2015}; ^{Sahagún 1950-1982, 2: 25, 134-40}).

On Codex Borgia page 46, we see clear references to Panquetzaliztli, the winter solstice festival, which featured the fire serpent and the fire god, Xiuhtecuhtli, shown here encased in the fire serpent at the bottom of the page (Fig. 7; ^{Milbrath 2013, 30, 107, Plate 18}). The fire serpent was associated with fire ceremonies performed in Panquetzaliztli in late November through mid-December (^{Durán 1971, 457-60}; ^{Milbrath 2013, 30}; ^{Sahagún 1950-1982, 2: 141-150}).

Page 46 represents the fire serpent six times, evoking the Panquetzaliztli festival, which featured the fire serpent descending the pyramid to be burned in the cuauxicalli (eagle vessel), here represented on the lower left of page 46 by two descending eagles carrying bowls on their backs. This annual ceremony should not be confused with the Aztec New Fire Ceremony, which was performed every 52 years in either Quecholli or Panquetzaliztli (^{Sahagún 1950-82, 4: 143-144; 7: 27}). And the fire serpent ceremony in Panquetzaliztli is also quite different from the annual new fire ceremony performed during the Izcalli, because the fire serpent does not play a prominent role in Izcalli, though it may have been present as a costuming element worn by the Fire God, Xiuhtecuhtli, who was honored during Izcalli (^{Nicholson 1971, Fig. 10}; Table 4).

The 18-page narrative in the Codex Borgia features an embedded festival calendar associated with changing seasonal images over the course of a year (^{Milbrath 2013, 106-07}; ²⁰¹⁵). Imagery of fire ceremonies and fire deities appear on pages linked with the dry season (Figs. 3, 7), in contrast to scenes on pages 36-44 that show images related to the rainy season, such as bees, butterflies, hummingbirds, and flowers (Figs. 4, 5). Also, images representing maize show a seasonal transformation with the alternation of the rainy and dry seasons. Maize fields and maize consumption are represented during the rainy season, as on Codex Borgia page 43 which shows a mature maize field and deities eating maize, whereas death deities appear to be carrying maize into the underworld during the dry season on Codex Borgia page 31.

Concluding Remarks

The seasonal veintena festivals represented in the Codex Borgia provide a chronological structure for the narrative sequence, but the principal focus is on a series of astronomical events. The analysis of «real-time» astronomical events in relation to veintena rituals represents a complete departure from interpretations published by other scholars. In my opinion, the narrative is not a representation of the central Mexican festival calendar, but it records specific festivals that coordinate with solar events and seasonal changes in the year. The narrative begins with the new moon following the winter solstice, setting up the context for astronomical events in the year. Veintena festivals in the narrative appear primarily on pages associated with important solar events, such as the spring equinox in Tlacaxipehualiztli on page 33, the summer solstice in Etzalcualiztli on page 37, and the winter solstice in Panquetzaliztli on page 46. The end of the year in Izcalli receives no special emphasis in the narrative because it is not actually intended to be a representation of the festival calendar, and instead it begins with a significant astronomical event and ends with a veintena that marks the winter solstice. The veintena rituals at yearend are instead found in the yearbearer almanac on Codex Borgia pages 49-52, which shows the new fire drilled in Izcalli and dates that reference the yearend in Izcalli, as well as a count that leads to the New Year, setting the veintena cycle in the context of the solar year and the 52-year xiuhmolpilli.

The seasonal coordinates in the Aztec festival calendar suggest that the veintenas were adjusted periodically to keep in tune with the seasons. This may have been done by adding an extra day every four years during Izcalli or adding 13 days to the length of the festival calendar every 52-years during Panquetzaliztli. In the Aztec calendar, Izcalli and Panquetzaliztli began at midnight, at least six hours before the tonalpohualli shifted to the next day, so that the yearbearer corresponded to the first day of Izcalli, and the same may be true in the Codex Borgia. Comparative evidence from Tlaxcalan and Aztec calendars helps to confirm the placement of Izcalli as the last veintena festival of the year.

The last yearbearer on Codex Borgia page 52 is 4 Flint, the same year represented by the sequence of festivals on pages 29-46. The year 4 Flint is very significant in terms of astronomy, for it is the year of the only total eclipse recorded in Aztec annals, and it was also a year that Venus transitioned from the Evening Star to the Morning Star just after the winter solstice, and transitioned back to the Evening Star just before the winter solstice in 1496. This explains why the narrative focuses on this series of astronomical events. Aztec colonial annals depicted important events in each year that include a number of astronomical references, but the Codex Borgia narrative highlights events in much greater detail, focusing on a single year to show specific astronomical cycles coordinating with veintenas festivals in the year of a total eclipse seen all across Mexico.