Chapter 10

Dawn in the Desert: The Rise of the Toltecs        


            Mexico's geographic location within the trade wind belt means that most of the country's moisture is derived from the Gulf of Mexico to the east. Were it not for the fact that the folded ridges of the Sierra Madre Oriental rise to heights of over 3000 m (10,000 ft) within little more than 100 km (60 mi) of the coast, the greater part of the country might be expected to have been lush and green. However, because of that "accident" of topography, most of Mexico lies in a "rain-shadow" behind the mountains, with the result that water is in critically short supply through much of the year.

            Nowhere is that deficiency more in evidence than in the northern interior of the country, because as the latitude of 30º is neared, the winds become more southerly, paralleling the Mexican coast and bringing their moisture ashore in Texas and the Gulf coastal areas of the United States instead. Along Mexico's east coast, the last great river flowing out of the interior is the Río Pánuco, which reaches the sea where the bustling oil and industrial port of Tampico now stands. For all intents and purposes, the Pánuco marked the northernmost boundary of the Mesoamerican cultural realm because beyond it, maize -- the indigenous staff of life -- could no longer be grown with dependable regularity.

            Once the verdant crests of the Sierra Madre Oriental with their stands of oak and pine forest are passed, the landscape quickly changes into one of creosote bush, yucca, and cacti. With all but the highest mountain uplifts now cut off from the trade winds, the vast interior of northern Mexico is a region almost totally dependent for its water supply on the scattered and sporadic monsoonal downpours of summer. As temperatures over the high plateaus of northern Mexico and the southwestern United States build up with the northerly advance of the sun, an upper-level flow of moist air starts moving in off the Pacific, triggering thundershowers that are as spotty in their distribution as they are short and violent in nature. These monsoonal rains begin earlier and last longer in the south of the Mexican plateau than they do in the north, and how far they penetrate or how much precipitation they produce can vary markedly from one year to another. As in Monsoon Asia, a very careful timing of the agricultural cycle might produce a successful crop, but in just as many if not more instances a total disaster could result as well. Small wonder then, that the local inhabitants found little opportunity and even less incentive to attempt to gain their livelihood in other than the most rudimentary manner, that is, through collecting (gathering and hunting).

            To be sure, during the Pleistocene ice age things were rather different. That was the time of the "pluvials," or heavy rains, which were occasioned by the equatorward displacement of the mid-latitude storm tracks. At that time large parts of the Mexican plateau supported a dense grass cover which enabled a rich fauna of Prehistoric grazing animals to inhabit the region. But as the climate warmed and the continental ice sheets retreated, so did the storm belts shift northward, nudging the Mexican plateau toward the semiarid and arid climate which it experiences today. While some of the great herds of deer, elk, and bison survived by moving northward, many species -- such as the native camels and horses, as well as the great lumbering mammoths and mastodons -- became extinct as the temperatures rose, the rains ceased, and the grasslands degenerated into desert and scrub. What had been a Paleolithic hunter's paradise had turned into a harsh and niggardly expanse inhabited chiefly by small rodents and reptiles, and the occasional larger predator. Indeed, one of the most productive local environments continued to be the waters of the temporary playa lakes which occupied the floors of many of the mountain basins, for there both fish and waterfowl could be found, at least seasonally.

Figure 54.

The climatic station of Zacatecas is representative of a large part of the northern Mexican plateau and as such typifies the environment in which the nomadic Chichimecs were found. Although the warmest month is May, the water need (and hence the temperature) is moderate throughout die year, due both to the place's northerly latitude and to its high elevation. Frosts can occur in the low-sun (winter) months, but the greatest drawback to agriculture is the deficiency of moisture. A monsoonal distribution of precipitation is still apparent, but barely meets the moisture requirements of plants for more than a couple of months. As a result, the prevailing vegetation of the region surrounding Zacatecas is of a short-grass steppe or semidesert variety.


            The higher crests of the Sierra Madre Occidental on the western edges of the Mexican plateau formed something of a green oasis in the drab brownness of the northern desert. Here a slightly greater annual rainfall coupled with cooler temperatures made the moisture effective enough to support extensive forests of pine, but once the descent of the western slopes was made, so too was there a return to semiarid conditions, only now in company with tropical temperatures. The aridity of the Pacific versant of the Sierra Madre Occidental is relieved only here and there by the waters of some larger river snaking its way to the coast. Northward beyond the great delta of the Río Grande de Santiago in what today is Nayarit state, the patches of green become smaller and more scattered the farther up the coast one goes. It was in these little alluvial plains along such rivers as the Sinaloa and the Fuerte that the last outposts of the Mesoamerican culture realm were to be found, clinging to their existence as tiny islands of corn cultivation in a vast surrounding sea of nomadism.

            Although the boundaries between the settled agriculturalist and the nomadic hunter-gatherer were fairly rigidly drawn by nature itself, this did not prevent the two culture-worlds from interacting with one another. Perhaps much of this interaction was the peaceful interchange of goods and ideas, because we do know that trade items from Mesoamerica reached well up into the American Southwest and that some especially prized commodities, such as native copper, found their way from as far away as the shores of Lake Superior to Central Mexico. In the same way, certain cultural traits like the ritual ball game and sacrifices to the morning star spread well beyond the limits of Mesoamerica into adjacent regions of North America. But inevitably some of this cross-cultural interaction was also of a more violent nature, and raids by hungry nomads into frontier areas of agricultural production must have been both frequent and repeated. Thus, depending on a given nomadic tribe's geography, they might have been drawn into the expanding web of civilization either early or late, either to a great degree or perhaps not at all. In any case, for at least a half dozen centuries following the founding of Teotihuacán, the line of contact between city dweller and barbarian must have fluctuated uneasily through the barren desertlands of the Mexican plateau just a few score kilometers beyond the horizon from the Pyramid of the Sun.


            Following their successful onslaught on Teotihuacán, the Toltecs must have experienced a certain numbing realization of what they had done. In place of the large, thriving metropolis that had tempted their incursion in the first instance now lay the glowing embers of a dead and vacant city -- the humble homes of its people in ruins, its vast, sprawling marketplaces silent, its artisan quarters abandoned, its religious and ruling elite gone. Once its stores of foodstuffs had been ransacked and its objects of art had been pilfered, there was nothing left but the hulking masses of its great pyramids standing as mute witnesses to the death of a great civilization. How long it took for Teotihuacán's terrified and impoverished populace to shrink back into the countryside is difficult to imagine; certainly, the exodus must have been a rapid one because with the supplies of food and water cut off, the city's inhabitants would have had to flee quickly to other, more congenial areas of subsistence or perish in the attempt. The pandemonium, panic, and sheer human suffering which engulfed the survivors must have been catastrophic. The loss in life was no doubt staggering; the loss in cultural and intellectual terms was almost irreparable.

            Ironically, it was probably the latter which troubled the Toltecs the most. The total fabric of the city's life was gone. Its social structure was in tatters. Its religious being was in question. Existence had suddenly become meaningless. It was as though the gods had forsaken them, for without the priestly elite, there was no interpretation of their will, no ordered timetable by which to commemorate them. Time itself had lost its meaning, for without the calendar one day was the same as any other.

            Surely, the mysterious malaise which settled over the survivors of Teotihuacán could not have long gone unnoticed by the Toltec conquerors. What was this "calendar" of which everyone seemed to speak? How had it ordered the lives of the people, and why was there such an empty void without it? Was there something unseen -- invisible in the city itself but nevertheless real and powerful -- through which the priests had communicated with the gods? No nomads, lurking on the desert fringes of the great metropolis, would have suspected that in addition to the pyramids and palaces and marketplaces and workshops -- which they could see -- was a mystical force that drove the entire engine of civilization, but which they could not see. It was something of which they would only belatedly become aware once the engine had stopped.

            In their quest to find out what this "calendar" was all about, the Toltecs probably had to rely primarily on untutored, secondary sources, for the priests who were the custodians of such privileged information had long since fled. Indeed, the native sources tell us that at the first sign of threat and turmoil, the priests of Teotihuacán had assembled their books and religious paraphernalia and abandoned the city for a land in the east. Nevertheless, the Toltecs must have come into possession of enough of the priestly records to recreate the calendrical system almost exactly, though at the same time they clearly availed themselves of the opportunity to make some distinctive modifications and embellishments of their own.


            The basic structure of the Toltec calendrical system was unmistakably patterned on that of its Teotihuacano predecessor, for 13 numerals alternated with 20 day-names to form the basis of a "sacred almanac" whose primary function was to divine the fortunes of individuals and schedule major religious festivities. Another count containing 18 groups of 20 days followed by 5 additional days which were considered "unlucky" measured the length of a "year" -- the latter being a concept with which the Toltecs must have had only a most rudimentary acquaintance. The two counts ran simultaneously and only after 52 years had passed would the numbers and names of the days once again be the same as when the count began. This point, in particular, must have impressed the Toltecs, because it was as though it proved that history repeated itself. To the primitive mind, it may also have suggested the necessity for humans to do something special -- to make the appropriate offerings to the gods -- so as to ensure that the world would continue for another 52 years, for such a provision was built into the Toltec version of the calendars from the very start.

            To identify specific years within any given 52-year "bundle," the Toltecs also imitated the practice of their Teotihuacano forebears by naming each year for the day of the sacred almanac on which it ended.  Because the least common divisor of both calendars was 5, this meant that the secular year could only end on one of the same four day-names (20 / 5), though the numeral would vary from 1 through 13 within each 52-year "bundle" of years. Thus, in effect the days which became the "year bearers" of the Nahua peoples were those named "Rabbit," "Reed," "Flint-Knife," and "House,' or Tochtli, Acatl, Tecpatl, and Calli in their own language.

            Another characteristic of the Toltec calendar which was derived from earlier Mesoamerican practice was the calibration of the beginning of the secular year with the southward zenithal passage of the sun. At the latitude of Tula (20º N), the Toltec capital city, this event occurred on July 24. Yet, to these newly acculturated nomads, this event must have seemed far less auspicious than the zenithal passage of the Pleiades (the distinctive asterism of seven visibly clustered stars which has intrigued peoples throughout the world), for they made this latter passage the occasion to mark the end of each of their 52-year cycles with a special celebration called the "binding of the years."

            Perhaps geography explains much of the Toltecs' fascination with the Pleiades, for no other native people of Mesoamerica who came under the influence of the calendar lived so directly beneath this star group. Indeed, at the latitude of Tula, the Pleiades transited the zenith each year at midnight on the evening of October 9. Although the position of all stars in the heavens changes with time due to precession (the very slow wobble of Earth on its axis), throughout the entire duration of the Toltec Empire this star cluster remained within 2º of the zenith. Even by the end of Aztec times it was within 2º.5 of the zenith, and at the present time the Pleiades are little more than 4º from the zenith as measured from Central Mexico.

            Although the Toltec 52-year cycle began with "1 Rabbit," thanks to Nahua mythology the "binding of the years" ceremony took place in the second year of the cycle, "2 Reed." This stemmed from the Toltecs' understanding that there had been four previous creations of the world, or "suns," -- each of them associated with one of the cardinal points, each lasting for 676 years (13 multiples of 52), and each having been terminated by such disasters as being devoured by jaguars, consumed by fire, destroyed by wind, or submerged by flood.

            After this last cataclysm it supposedly took a year to "raise the heavens again," and in the second year a feast was prepared for the gods (Krickeberg, 1980, 24). Since by tradition fire was first acquired on this occasion, the ceremony itself involved putting out all the fires of the past era and rekindling new ones. Thus, in the Nahua mind the fateful moment in which the Pleiades passed through the zenith -- in relative proximity to Orion's Belt, which they visualized as a "fire drill" -- provided an appropriate opportunity for demonstrating humankind's gratitude to the gods for giving them fire. To the Toltecs -- and later the Aztecs -- this meant cutting out the heart of a sacrificial victim and kindling the first new fire in his chest cavity. Indeed, some form of human sacrifice was to become an integral part of the celebration of the end of each 20-day "month," as well as on the occasion of an ominous day called "4 Movement." The Nahua peoples believed that the present "world," or the "fifth sun," had been created on a day with that name and that it would end on another day named "4 Movement" when devastating temblors would destroy the earth.

            Any researcher who would attempt to reconstruct the calendrical system of the Toltecs must necessarily take into account the facts and fictions by which they lived. Certainly the most exhaustive study that has been made of the Nahua calendar was that carried out by the Mexican scholar Alfonso Caso (1967), who based his correlation on the correspondence of certain key dates of the Spanish conquest and native Indian sources. The events which he chose to demonstrate this correspondence included Cortés's arrival in Tenochtitlán (November 9, 1519), his forced retreat from the city known as La Noche Triste (June 30, 1520), and his ultimate destruction of the Aztec capital and capture of the emperor Cuauhtémoc (August 13, 1521). On the basis of these correspondences, Caso concluded that the Nahua calendar had to have begun with the "month" of Atlcahualo, although of 42 sources which he surveyed ranging from the sixteenth through the twentieth century, he found a wide difference in opinion. (Remember that what we term a "month" in the Mesoamerican calendar is in fact an interval of 20 days, and had nothing to do with the length of the period of the moon's revolution.) Like Caso, 14 of the sources opted for the "month" of Atlcahualo, another 14 preferred Tlacaxipehualiztli, 7 cited Izcalli, 3 chose Tititl, 2 Atemoztli, and 1 each Panquetzaliztli and Toxcatl. Thus, no fewer than 7 of the 18"months" have been suggested as the starting point of the Aztec year. Even the two earliest chroniclers, the Spanish clerics Sahagún and Durán, disagreed: The first cited a beginning Julian calendar date of February 2; the second, March 1.

            By devising a computer program which permitted me the flexibility of testing all of the "months" of the Aztec year as potential starting dates, I was able to establish that any of the first 8 "months" would have produced the same three correspondences as Caso's correlation. However, if the calendar had begun with the 9th month, La Noche Triste would have fallen during the nemontemi, or 5-day unlucky period, whereas from the 10th through the 18th "months" it would have occurred on the day 3 Huey Tecuilhuitl rather than on 18 Tecuilhuitontli as Caso argued. From "month" 13 on, the most critical of the three dates -- the fall of Tenochtitlán -- would have occurred on 7 Xocotl Huetzi rather than 2 Xocotl Huetzi, thereby rendering the last 6 of the Aztec "months" as impossible candidates for the beginning of their year. Thus, while any of the first 8 Aztec "months" produced totally as accurate results as Caso claimed for AtIcahualo, when they were tested against a fourth well known date from the Conquest -- namely, that of the great Cholula massacre carried out by Alvarado on May 23, 1520 -- both the 7th and 8th "months" had to be dismissed as well.

            Having now narrowed the field to the first 6 "months" of the Aztec year, I proceeded to test each of them to determine on which day of the secular calendar the final day of the sacred almanac fell. This was important because Caso argued that the name of the year took its name from that day; and since Caso had selected Atlcahualo as the first "month" of the year, he concluded that the critical day would have been 20 Tititl.

            Choosing the final celebration of the "binding of the years" in 1507 as my test case, I found that all 6 of the "months" produced exactly the same results: the day "1 Rabbit" of the sacred almanac fell on 20 Tititl of the secular calendar (corresponding to January 22, 1507). One round earlier in the sacred almanac, however -- i.e., on May 7, 1506 -- the secular calendar which began with "month" 1, lzcalli, would again have reached the date 20 Tititl, whereas those versions of the secular calendar which began in "months" 2 through 5 (including Caso's choice, Atlcahualo) would all have recorded 20 Huey Tozoztli. On the other hand, the version of the secular calendar which began with "month" 6 would have reached the date 5 Nemontemi, the last of the so-called 5 unlucky days. Thus, through this test I had further narrowed the field of possible candidates for the secular year's starting "month" to the only two with truly distinctive characteristics -- "month" 1, Izcalli, and "month" 6,' Toxcatl.

            Caso claimed that the Toltecs would never have chosen the name of one of the 5 unlucky days as the name of the year, and therefore had ruled out the possibility of Toxcatl having served as the beginning of the secular year. Nonetheless, several earlier writers including Bernardino Sahagún had commented on how similar the celebrations of its festival -- the greatest of the entire year -- were to those of the "binding of the years," so I was not about to discount it quite yet. A further correlation was required to narrow the choices between Izcalli and Toxcatl.

            All researchers examining the Mesoamerican calendars have been struck by how closely the Nahua version mirrors that of the Maya. Indeed, a comparison of the day-numbers of the sacred almanac in each of the Nahua and Maya versions reveals that they are only two days apart! For example, what would have been the day 12 Ik in the Maya almanac corresponded to the day "1 Rabbit" in the Aztec count. (Remember that after the numeral 13 is passed in the sacred almanac the day-numbers begin with 1 again.) One of the days in this discrepancy could easily have resulted from a lapse in the count occasioned by the chaos attending Teotihuacán's fall, while the other might well derive from a difference in counting techniques -- the Olmecs and the Maya starting from sunset and the Toltecs and the Aztecs from sunrise.

            However, when the day- names of these two dates are combined with their respective numbers, we find that the relative positions of these dates in the sacred almanac are actually 106 days apart. This is because 12 Ik is day 142 of the Maya count whereas "1 Rabbit" is day 248 of the Aztec count. Thus, from 142 to 248 represents a shift of 106 days. In other words, when we correlate individual days of the Maya sacred almanac with their counterparts in the Nahua count, we find that the advance of the latter over the former amounts to five 20-day "months," an additional 5 days for the unlucky nemontemi period, and an extra day because the Toltecs and their heirs did not use zeros with which to count. (To illustrate this point, a day which would have been numbered 0 Pop by the Maya would have been 1 Pop to the Aztecs. It seems likely that the pragmatic Nahua saw no value in a number which meant "in progress" or "not yet complete." By the same token, their failure to grasp the notion of zero probably likewise accounts for their having abandoned sunset as the time from which to tabulate their count of the days and substituted sunrise instead.) Of course, this 106-day interval can also be thought of as a restatement of the remainder between one round of the 260-day sacred almanac and one round of the 365-day "Vague Year" expressed according to the Nahua counting mode.

            Naturally, with a displacement of 106 days between corresponding dates in the two calendars, we can no longer argue for some casual lapse having taken place between them but must recognize instead that a carefully calculated and conscious shift had been made of the starting point of the secular calendar. This recognition therefore led me to test the date of zenithal sun passage over Tula as it might have been recorded in the unaltered Maya secular calendar and then compare that to the date as it would have been recorded in a version of the Aztec calendar based on Caso's correlations. (For this test, I used the example I cited above in which the day of the Maya almanac equated to 12 Ik and that of the Aztec to "1 Rabbit.") According to the unaltered Maya secular calendar the day was 0 Yax, while in the Aztec version it was 1 Izcalli. This meant that I had now established a correlation between the "months" of the Maya and Aztec secular calendars, for if Yax corresponded to Izcalli, then Zac would correspond with Atlcahualo, and so on.

            However, to get the position of 12 Ik in the Maya almanac to match that of the position of "1 Rabbit" in the Aztec almanac I would have to advance the count of the latter by five "months" and six days, as I described above. When I did this, I found that the date on which "1 Rabbit" was reached was on 1 Toxcatl, but only if the nemontemi had also been inserted somewhere ahead of 1 Toxcatl. (Of course, it doesn't matter where the nemontemi is inserted between Izcalli and Toxcatl to obtain such a result, but all early accounts of the structure of the calendar suggest that the 5-day unlucky period immediately preceded the beginning of the new cycle -- i.e., it was positioned just ahead of the "month" of Toxcatl. Only with such a configuration will the Aztec secular calendar mesh exactly with the Maya secular calendar.


            Once a convincing correlation had been established between these two versions of the Mesoamerican calendrical system, I reasoned that the chronology of the Nahua time-count should now be as secure and reliable as that of the Long Count. It should, therefore, be possible to devise a computer program to run the Nahua secular calendar "backward" in the same manner that I had done with the Maya calendar in order to establish when it had actually come into being. I was well aware that, because the "Vague Year" of the Nahua "slipped" forward one day in every four years just as the Maya secular calendar did, the "month" of Toxcatl would have "migrated" widely through the year. Thus, in 1521, when Cortés conquered the Aztecs, the first day of Toxcatl fell on May 4; in 1403, on June 3; in 1091, on August 20; and so on. Therefore, it was necessary to insert a "flag" in my program which would search out the closest correspondence between the beginning of Toxcatl and the "binding of the years" ceremony, which in our calendar would have occurred in the month of October. Concurrently, I would be looking for a "suitable" beginning date for the southward zenithal passage of the sun -- in other words, one which would have made "good sense" to a group of pragmatic nomads -- in short, a straightforward beginning of the secular year with a day numbered "1" in both the sacred and secular counts. (While the Toltecs would have been no more likely to tamper with the religiously sanctified sacred almanac than any other Mesoamerican people ever were, they had good reason to shift around the secular calendar to suit their needs, just as the Maya had done back in A.D. 48.)

            As I ran the computer program "backward," I discovered that the closest correspondence between the beginning of Toxcatl and the "binding of the years" occurred in the year 883 ("2 Reed") when they were just two days apart. However, if the secular calendar had begun in the previous year ("1 Rabbit"), as tradition indicated it had, then the July 24 zenithal passage of the sun would have occurred on the date 7 Atlcahualo -- certainly not a rational day of "beginnings." Going back a further 52-year cycle, I found that the beginning of Toxcatl fell on October 24 in the year 831 (some 15 days after the zenithal passage of the Pleiades), whereas the southward passage of the zenithal sun took place on July 24, 830. In the secular calendar, this happened on 14 lzcalli -- again hardly an auspicious day on which to begin a calendar. If, however, we go back yet one more interval of 52 years, we find that Toxcatl would have commenced on the day November 6 (a full 27 days following the zenithal passage of the Pleiades), but that the southward passage of the zenithal sun on July 24, 778, would have coincided with the day 1 Cuetzpallin ("Lizard") of the sacred almanac and 1 Izcalli of the secular count. Although there is every reason to believe that a considerable period of planning and anticipation had to have preceded this date, the internal structure of the calendar itself argues for its having been inaugurated on July 24, 778. Indeed, this conclusion finds strong reinforcement in the research of Krickeberg (1980, 209), who states that the Nahua believed that the present "sun," or "world," came into being in the equivalent of our year 726. When we realize that this notion places the creation of the "Toltec world" exactly 52 years before the starting date of their calendar, we can appreciate how quickly these recent nomads from the desert had come to embrace one of the most fundamental concepts of Mesoamerican civilization -- to wit, the cyclical repetition of history.

            By selecting 1 Cuetzpallin 1 Izcalli as the beginning date of their secular calendar, the Toltecs also assured that their first celebration of the "binding of the years" -- on October 9, 779 -- fell on the auspicious day of "1 Death" in the sacred almanac. By the same token, if they had retained an awareness of the 52nd sunset after the summer solstice  -- i.e., August 13 -- they could not have failed to appreciate that it fell on 8 Cuetzpallin 1 Atlcahualo, exactly one "month" after the southward passage of the zenithal sun. And similarly, they would have found that the autumnal equinox (September 22) occurred on 9 Cuetzpallin 1 Tozoztontli, so they must have felt quite satisfied in having established such an impressive measure of consonance with the rhythms of the heavens. It is also interesting that by inaugurating their secular count on July 24, 778, the Toltecs had, in fact, made Izcalli the first "month" of their year. However, unlike the Maya and the other Mesoamericans whose 5-day unlucky period fell just ahead of the beginning of the secular year, the Toltecs had chosen to shift it so that it more closely approximated what they considered to be a celestial event of yet greater significance -- the midnight zenithal passage of the Pleiades --hence its insertion just ahead of the "month" of Toxcatl.

            In addition to the astronomically fixed celebrations of the zenithal passages of the sun and the Pleiades, there were the equinoxes and solstices that occupied defined positions in the ritual year as well. Of these, it would appear that it was the winter solstice that became the occasion for the most important celebration, because the sun was then at its farthest remove south and there was need to ensure its return to more northerly latitudes. One of the lasting legacies of the Nahua calendar is the annual celebration which now serves to mark the feast day of the patron saint of Mexico, Our Lady of Guadalupe. In 1531 a newly converted Indian peasant by the name of Juan Diego, trudging across the hillside at Tepeyac just north of Tenochtiitlán, claims to have seen the Virgin. Although the Spanish bishop was at first dubious, a shrine was erected there which has since become the most important pilgrimage site in Catholic Mexico. Sahagún, in writing about the place in the 1580's, was already "suspicious" about both its name and its function, for by then it had become known as "Our Lady of Guadalupe-Tonantzin," the latter appelation being that of the Aztec sun-goddess. He also observed that the temple of the sun goddess had stood on the same hill in pre-Conquest times, and he was at pains to explain why, of all the churches dedicated to the Virgin, this was the only one at which the Indians worshiped (Sahagún, 1981, 329). Had he looked more deeply into the matter, he would likewise have been intrigued by the fact that the annual feast day falls on December 12, which in the Julian calendar marked the winter solstice!

            In the same vein, one cannot help wondering how, in the wake of the Spanish conquest, the various individual Indian villages of Mexico and Guatemala came to choose their local patron saints from among the array which the Catholic church offered them, and whether the choice of any given "saint's day" might not have been simply a thinly veiled artifice for continuing a time-honored religious celebration from the pre-Columbian past.

            In addition to the fixed religious festivals of the Nahua year were those which "migrated" through the seasons with the slippage of the secular calendar. The final day of each 20-day "month" provided the occasion for a celebration which almost invariably involved some form of human sacrifice, as did the occurrence of the feared day "4 Movement." By Aztec times, these ritualized killings had reached such a peak that the birth of each new day was greeted by the offering of a human heart.


            Because Itzcóatl, the fourth emperor of the Aztecs, had all the history books of earler civilizations burned, the fragmentary and contradictory records that we have of the Toltecs derive chiefly from two sources: lxtlilxóchitl's Historia Tolteca-Chichemeca and the Anales de Cuauhtitlán, both of which are post-Conquest documents that purport to list the Toltec kings and the lengths of their respective reigns. Although each of them lists 10 rulers, only three names are common to both lists and only one of their reigns overlaps the same period of time. Indeed, the Historia's list is particularly suspect, because six of the rulers are therein said to have reigned for exactly 52 years and the last -- Quetzalcóatl -- supposedly reigned for 74.

            Moreover, because the Toltecs lacked the Long Count, considerable debate has arisen among scholars as to the particular 52-year cycle in which any given event may have taken place. Contributing to the confusion has been the contention of some researchers that the Toltecs used a modified version of the Mixtec calendar, which caused their dates to be 12 years at variance with (i.e., earlier than) the time-count inherited from the Olmecs (Davies, 1977, 441-466). As Caso has pointed out, however, such an argument is totally at odds with what we know about the precision of Mesoamerican timekeeping; indeed, it does nothing to explain the striking correspondences between the later Aztec calendar and the earlier Maya version of the same time-count which we discussed earlier. Therefore, the chronology given below is that based on the version of the Aztec calendar in use at the time of the arrival of the Spaniards.

            If we disregard for the moment the more legendary accounts of the earliest nomad chieftains (for example, the earliest chieftain, Mixcóatl, or "Cloud Serpent," supposedly laid the groundwork for the first Toltec capital at Ixtapalapa, to the southeast of modern Mexico City), then perhaps one of the first events recorded in Toltec history to which we can assign a reasonably reliable date was the birth in the year Ce Acatl ("1 Reed") of the chieftain-king who took the name "Quetzalcóatl" and who later was deified by his subjects. (One should keep in mind, however, that the original deity named Quetzalcóatl was a culture hero who supposedly first acquainted the peoples of Mesoamerica with agriculture and the calendar and whose representation as a feathered serpent is repeatedly juxtaposed with a second deity -- Tlaloc, the rain-god -- on the façade of one of the principal pyramids at Teotihuacán. It is likewise interesting that in this representation, Quetzalcóatl is clearly associated with seashells, implying a link with the distant sea.) The son of another chieftain named Mixcóatl and his wife Chimalmán, the boy was named "Ce Acatl Topiltzin," the latter appelation meaning "our prince." The year which most convincingly coincides with this event was A.D. 947. In the year "2 Rabbit" (974) his father established the Toltec capital at Tulancingo, but when Quetzalcóatl ascended to the throne three years later (in the year "5 House"), he moved his capital to Tula, some 100 km (60 mi) to the west. (Sanders believes that the Toltecs had occupied the region surrounding Tula as early as A.D. 700 [Adams, 1991, 220], while Krickeberg dates the actual founding of the city to the year 856 [1980, 424]. Although the terrain surrounding Tula would have made a gridded master plan difficult to carry out, at least a part of the city is laid out with an orientation based on the August 13 alignments found at Teotihuacán [Adams, 1991, 23 11.) This move appears to have been a response to pressure exerted by the so-called historic Olmec, or "Olmec Uixtotin," who at that time controlled both the valleys of Mexico and Puebla. Despite having enjoyed a relatively long reign of peace and prosperity in his new capital, Quetzalcóatl was forced by a power struggle within the Toltec hierarchy to flee with his retinue to the Gulf coast in the next year named "1 Reed" (999). Although the deposed king died on reaching the coast, legend has it that he was subsequently reincarnated as the morning star (the planet Venus), having vowed to return one day rightfully to reclaim his throne. His retinue continued to the east, arriving on the coast of the Yucatán, and then moved inland to subjugate the by then almost moribund Maya city-state of Chichén Itzá.

            Meanwhile, back in the Toltec capital of Tula, new chieftain-kings came and went. Matlacxóchitl is reported as having died in the year "10 Rabbit" (1034), whereas Nauhyotzin, his successor, passed away in the year "12 House" (1049). Matlaccoatzin then assumed the throne and ruled until his death in the year "1 House" (1077). He was followed on the throne by Tlicohuatzin, who reigned until the year "9 Rabbit" (1098). The subsequent events of Toltec history all took place within the reign of a king called Huémac.

            Little else comes down to us in the way of Toltec history apart from certain "innovations" which they developed in carrying out their calendrical rituals. In the year "7 Rabbit" (1122), for instance, we are told that the Toltecs initiated the practice of sacrificing children to the rain-god, and in the year "8 Rabbit" (1162), as a consequence of certain "auguries," they note that they began sacrificing prisoners by shooting them with arrows, a practice they claim to have learned from the Huastecs. When war broke out five years later (in the year "13 Reed," or 1167), the Toltecs report that they began the sacrifice of prisoners by flaying them. Ironically, though most "civilized" societies would have found such practices totally repugnant, for the Toltecs they represented evidence of their advance from nomadic hunting and gathering to a life as settled cultivators, for these innovations in human sacrifice must be seen as attempts to propitiate newly acquired agricultural deities, such as Tlaloc and Xipe. None of these "refinements" seem to have stayed the hand of destiny, however, for in the following year, " 1 Flint-Knife," Chichimecs swarmed in from the northern desert and laid waste Tula, causing Huémac to flee southward with his court. It was at this time that Huémac relocated the Toltec capital to Chapultepec, on the western outskirts of present day Mexico City. His kingdom lost and his subjects scattered over the countryside, a despondent King Huémac committed suicide in the year "7 Rabbit" (1174). Although bountiful harvests were recorded in the year "2 Reed" (1195) -- when the Toltecs celebrated what was to be their final "binding of the years" ceremony to ensure the world's survival for another 52 years -- this belated stroke of good fortune was not enough to rekindle their vitality as a people, and in the year "1 Flint-Knife" (1220) they disappeared as an organized force from the pages of pre-Columbian history.



            Although what passes for Toltec history is a mixture of legend and fact, it does provide us with a temporal framework against which we can assess certain of the key events whose commemoration by the Aztecs speaks to the continuance of a lengthy tradition. For the purposes of studying how the Toltecs actually used the calendrical information to which they fell heir, we should keep in mind that their domination of the Mexican plateau lasted roughly 500 years -- i.e., from the late eighth to the early thirteenth century.

            Within that time span, there were nine celebrations of the "binding of the years," namely, in 779, 831, 883, 935, 987, 1039, 1091, 1143, and 1195. Within the same period, there were four times when the Venusian double-cycle coincided with such an event. These occasions were in 831, 935, 1039, and 1143. Therefore, if any correspondence is to be found between the calendars as handed down to the Aztecs and the Venusian events, it must focus on these four years, in all of which a superior conjunction took place.

            We have already remarked how the Nahua peoples of the Mexican plateau defined the four phases of Venus in the Codex Borgia, assigning a 243-day period to its existence as a morning star, followed by a 77-day period of invisibility, then a 252-day period as an evening star, and finally another 12-day disappearance into the underworld before it returned once more as a morning star.

            We have also commented on the difficulty which the Mesoamericans had in accurately defining the planet's movements, due to their failure to understand the nature of its "disappearances." While modern astronomers can mathematically pinpoint the moment of Venus's conjunctions with the sun, the Mesoamericans had to rely on their first visual sightings of the planet, either as an evening star just after sunset or as a morning star just before sunrise.

            In one of the more valuable papers to emanate from the 1984 Symposium on Archaeoastronomy held at the National Autonomous University of Mexico, J. Daniel Flores Gutiérrez presented a detailed computer analysis of the trajectories of Venus as they occurred over more than 300 years of Mesoamerican history. In addition to making the point that the "binding of the years" festivals of the Toltecs and Aztecs were of two alternating types --Venus as an evening star in the years 1195, 1299, 1403, and 1507, and as a morning star in the years 1247, 1351, and 1455 -- his study revealed that, regardless of the type of festival, no Venusian event (e.g., the planet's disappearance, reappearance, or elongation -- the latter being its angular distance from the sun)  correlated with any key date of the solar year whatsoever. The author was thus left expressing the hope that someone might yet find one or more of the planet's periods -- i.e., of 584 days, of 8 years, of 104 years, or of 260 years -- in the structure of the codices (Flores Gutiérrez, 1991, 353).

            However, because we are unable to make any sense out of the motions of Venus against the background of our own calendar, before we abandon all hope of finding a meaningful temporal pattern to its movements, we should examine them in reference to the indigenous Mesoamerican calendar. In Table 6, all of the conjunctions of Venus, superior as well as inferior, that occurred in the years "2 Reed" -- i.e., at the time of the "binding of the years" -- are listed according to the dates of their occurrence in both Julian and Nahua calendars.

            It is readily apparent from Table 6 that with each successive celebration of the Venusian 104-year "double-cycle," the conjunction of the planet with the sun moved forward in the year about a month, according to the Julian calendar. However, because during this same period the Mesoamerican calendar had "slipped" forward some 26 days (i.e., 13 days during each 52-year "bundle" of years), according to the indigenous time-count the date of any given pair of conjunctions never varied by more than 2 days --  the greatest discrepancy having taken place in the twelfth century. On the other hand, between any two successive pairs of conjunctions a 4-, 5-, or 6-day hiatus did, in fact, occur. However, whereas a forward shift of more than seven months took place according to the Julian calendar, there was no more than a 33-day forward movement in the date of the Mesoamerican sacred almanac -- i.e., from 12 Ollin to 5 Cuetzpallin. Moreover, as can be seen, with the exception of the first four occurrences and two of the last three occurrences, all the conjunctions of Venus took place within the Nahua "month" of Tozoztontli.


Table 6 - Dates of Conjunctions of Venus, A.D. 727-1507


Julian Date


Day of Toltec / Aztec



November 2


12 Ollin 9 Huey Tozontontli


October 20


12 Ollin 9 Huey Tozontontli


October 1


6 Ozomatli 3 Huey Tozontontli


September 18


6 Ozomatli 3 Huey Tozontontli


August 31


1 Miquiztli 18 Tozontontli


August 17


13 Coatl 17 Tozontontli


July 31


9 Cipactli 13 Tozontontli


July 17


8 Xóchitl 12 Tozontontli


July 2


6 Tecpatl 10 Tozontontli


June 17


4 Cozcacuauhtli 8 Tozontontli


June 4


4 Cozcacuauhtli 8 Tozontontli


May 18


13 Malinalli 4 Tozontontli


May 6


1 Acatl 5 Tozontontli


April 18


9 Tochtli 20 Tlacaxipehualiztli


April 6


10 Atl 1 Tozontontli


March 19


5 Cuetzpallin 16 Tlacaxipehualitztli

*S - Superior; I - Inferior.


            As mentioned earlier, it was not, however, the conjunctions of Venus and the sun which the Mesoamericans were defining but rather the planet's mysterious disappearances into the "underworld." Because they never understood the celestial mechanics behind its disappearances, they could never really pinpoint the planet's nearest actual approach to the sun. Instead, all they could do is record its last visual sighting in the east (in the case of a superior conjunction) and its first sighting in the west. Therefore, all they were really aware of was that Venus -- on its longest visit to the underworld -- usually disappeared during the "month" of Atlcahualo and usually reappeared during the "month" of Toxcatl. A greater precision than this was impossible to obtain, but how this correlated to the reconvergence of the sacred almanac and the secular calendar is not readily apparent, for in each of these years the initial date of the year, "1 Rabbit," fell anywhere from 2 to 3 Julian months ahead of conjunction and always about a Julian month ahead of the planet's disappearance.

            It is, therefore, quite apparent that, when calibrated against the Mesoamerican calendar, the movements of Venus showed a remarkable cyclicity, suggesting that, on average, the planet's disappearances and reappearances could often be "predicted" within a day or two. Indeed, never would Venus fail to reach its expected place in the heavens for more than six days. Perhaps by modern scientific standards a prediction this approximate would not constitute "precision," but it must nevertheless have sufficed to reinforce the confidence of the masses in the priestly hierarchy, which was probably a consideration of some importance in itself.

            However, to me, the lack of a precise correspondence between one of these Venusian events and the solar year strongly suggests that scholars have been looking in the wrong place to calibrate the planet's movements. Instead of trying to match either the last day that Venus is visible before a conjunction or the first day it is visible after one with the beginning of one 104-year period or the end of another, perhaps we should be attempting to calibrate the planet's cycle against the solar year at a time when we are certain that it will be visible against some fixed horizon marker. This might be Venus's extreme northerly or southerly rising or setting position -- as delineated by such an alignment as that discovered by Horst Hartung from Uxmal to the pyramid of Nohpat in the flat expanse of the Yucatán -- or by its rise over some commanding topographic feature in a region of more rugged terrain.

            Inasmuch as it was Venus's heliacal rise as the morning star that most concerned the Mesoamericans, it seems very probable that the "binding, of the years" festival with which it is supposed to have coincided had to have been one of the "morning star" variety -- i.e., one that occurred in the 104-year cycle associated with the years 1247, 1351, and 1455. (Note that each of these dates falls within the Aztec era rather than in Toltec times. Furthermore, it is well established that a Venusian period did not coincide with the Aztecs' final celebration of the ceremony in 1507.) It is also known that, although the ceremony itself was held in a year named "2 Reed," such as 1455, the new 52-year cycle actually began in the previous year which was "1 Rabbit."

            Among the Nahuatl-speaking peoples of the Mexican plateau, one of the principal centers for the worship of Venus was Cholula on the western edge of the Puebla basin. (Others were Teotitlán del Camino and Tehuacán in the valley between Cholula and Oaxaca.) Indeed, Cholula was known to have been both an important market center and also the primary pilgrimage site of highland Mexico. After the fall from grace of the Toltec king Quetzalcóatl and his banishment from Tula, he stopped first at Cholula on his way into exile in the east.

            For anyone surveying the eastern horizon from Cholula, there is no question but that the most impressive feature within view is the great snow-capped cone of Orizaba, Mexico's highest mountain (5701 m, 18,700 ft). It is interesting and no doubt significant that in Nahuatl the peak is known as Citlaltépetl, or "the mountain of the star." One has good reason to believe that the "star" in question might well have been Venus, because throughout Toltec and Aztec times, the brightest star to rise in the vicinity of Orizaba was Spica (magnitude 0.97), but even it rose 6º away to the south. It is true, however, than in Teotihuacano times Spica rose near Orizaba for about a century, but it seems very unlikely that the Nahuatl name for the peak would have commemorated an event which had not been observed for nearly a millennium. (Even stranger is the explanation given for the mountain's name by the Spanish writer Clavijero, who claims that it received its appelation in colonial times after the peak had experienced a 20-year series of eruptions beginning in 1545 [Macazaga Ordoño, 1979, 451 ].)

            From the clues at hand -- namely, that Cholula was a major pilgrimage center dedicated to the worship of Quetzalcóatl (i.e., the planet Venus), that the most striking physical feature within view of Cholula is the cone of Orizaba, that the native place-name for the feature was "mountain of the star," and that no bright star (other than possibly Venus) rose anywhere near the peak -- the hypothesis which remains to be tested is this: If Venus does indeed rise over the mountain, when does it do so, and does this in any way coincide with the beginning or ending of the solar year as known to the Nahuas?

            Using a computer, a detailed study of the trajectory of Venus as seen from Cholula was made for the entire period of Nahua ascendancy over the Mexican plateau -- i.e., from the eighth through the fifteenth century. This allowed seven full 104-year Venusian cycles to be examined, beginning in 830 and ending in 1454. One of the results of this exercise was the discovery that Venus was found to rise six times over Orizaba in any given eight-year period. This cyclical pattern is sketched out in Table 7.

            Thus, every eight years these Venusian events are repeated at the same place in the sky within a day or two of the time they were last witnessed. In the second and seventh years of the cycle Venus puts in no appearance over the mountain at all, whereas in the third, fifth, and eighth years it rises over Orizaba in the spring (March and April). However, in the first, fourth, and sixth years of the cycle the heliacal rise of Venus over Orizaba will be seen to occur in the autumn months of October and November. It is on these occasions, of course, that its fateful appearance most closely coincides with the midnight zenithal passage of the Pleiades. Even so, from the early twelfth century to the mid-fifteenth century, the date of Venus's rise over Orizaba advanced about 10 days (from November 11 to November 1), so a truly convincing case for a horizon-based calibration of the planet's movements with the "binding of the years" ritual still remains somewhat in doubt.


Table 7 - Dates of Venus's Rise over Orizaba

Year of Cycle

Approximate Dates


Late October - Early November




Late April (19 -21)


Mid - November (10 -12)


Late March ( 25 - 28)


Mid - October ( 15 - 18)




Early April ( 8 - 10)



            At the peak of the Toltecs' power their capital city of Tula probably numbered between 30,000 and 40,000 inhabitants, and its trading network extended from as far south as the Nicoya Peninsula of Costa Rica (from which it imported polychrome ceramics), eastward into the Yucatán Peninsula, and as far northward as the present-day southwestern United States (where the architectural influences of both Tlaloc, the rain-god, and Quetzalcóatl, the feathered serpent, may be identified). Interestingly, although Plumbate pottery from Soconusco shows up in Tula, no metal has ever been found there, despite the fact that the Toltec period is known to have coincided with the first appearance of metallurgy in Mesoamerica. (Antagonism with the Purépecha, the first practitioners of the new craft, may well be the explanation.) Nonetheless, mining colonies had been established in the northern desert -- probably as early as Teotihuacano times -- to supply semiprecious stones to the civilizations of Central Mexico, but about the year A.D. 600 a southward retreat had begun which had all but abandoned the Chichimec "outback" to the nomads by 850. A Toltec re-expansion into this region appears to have taken place about 900, and Casas Grandes in northwestern Chihuahua seems to have been a thriving Toltec trading outpost by about 1050. Although Central Mexican influences had begun to show up in western Mexico at such sites as Ixtépete near modem Guadalajara and at Ixtlán del Rio in the borderlands of Nayarit from about A.D. 300 onward, these contacts were considerably strengthened during Toltec times. Contact with the Maya area of the Yucatán was not limited to commerce as we have seen, for about the year 1000, Toltec warriors moved in to establish a new militaristic regime in the old ceremonial center of Chichén Itzá, not only giving that place a new lease on life but materially altering its architecture, art, and religion as well. Not the least of these cultural influences was the introduction of the cult of Quetzalcóatl, who became known to the Maya by the name of Kukulkan.

            Geographically closer at hand were such ceremonial centers as Cholula and Xochicalco which, despite their proximity to the Toltec heartland, managed to retain a fairly strong Teotihuacano character. It would seem that the neighbors nearest to the Toltecs who received the most cultural impact were the Mixtecs. Although they had originally received the calendar by way of the Zapotecs, together with the hieroglyphic system of dots and bars to record numerals, under Toltec influence they abandoned the use of bars. As a result, we find that all of the beautifully colored Mixtec codices which have been preserved define calendar dates only with dots.


            Situated on a mountaintop about 20 km (12 mi) south-southwest of Cuernavaca (latitude 18º.8 N) is the fortified Late Classic ceremonial center of Xochicalco. Its architecture shows the strong influence of Teotihuacán, and it has been argued by some scholars that the hilltop city represented one of the refuges of the fleeing elite of that great metropolis as it came under attack from the desert nomads. Quetzalcóatl finds vivid representation there along the sides of the main pyramid, and the site also boasts one of the largest ball courts of all Mesoarnerica. Astronomically, the city's chief claim to fame is its zenith tube and observation chamber hollowed out of the living rock. Having an aperture of 40 cm (16 in.), the vertical tube opens into a subterranean cavern some 8.5 m (27.9 ft) beneath it and was obviously used to calibrate the zenithal passage of the sun (Broda, 1986, 92). At Xochicalco's latitude, this is an event which takes place on May 15 and July 29 each year. It would appear that the inspiration for this device was the similar feature found in Mound P at Monte Albán, which heralded the heliacal rise of the star Capella (magnitude 0.08) on the morning of the first zenithal sun passage over that site. Interestingly, in the seventh century A.D., when the city was probably founded, the heliacal rise of the Pleiades took place on May 15 over Popocatépetl, as seen from Xochicalco. Since no evidence has ever been presented to argue for a separate and independent calendar beginning on May 15, it must be presumed that the zenith tube served the function of calibrating a locally observable phenomenon within the existing Mesoamerican calendar.

            The German geographer Franz Tichy has argued that Pyramids C and D at Xochicalco served to mark key astronomical alignments as viewed from the altar of glyphs, which is located directly between them. Looking eastward, he finds that the northern corner of Pyramid C marks the sunrise azimuth on the summer solstice, while the southern corner of the same structure defines the sunrise azimuth at the winter solstice. Turning toward the west, he finds that the northern corner of Pyramid D is aligned to the sunset azimuth on the days of the zenithal sun passage -- i.e., May 15 and July 29. A line drawn through the middle of both pyramids and over the stela on the intermediate altar divides the solar year exactly in half, he states (Broda, 1986, 85). It seems clear, therefore, that although Teotihuacán had itself fallen victim to the barbarians, in places like Xochicalco, the torch of astronomical knowledge had still not been extinguished in Late Classic times.


            Nestled in the karstic hills of northern Veracruz is the architecturally elaborate and intriguing site of El Tajín. Although evidence of settlement in the area goes back as far as 1500 B.C. (Adams, 1991, 232), life at the village level appears to have emerged by Late Preclassic times but seemingly without noticeable Olmec influence. The region surrounding El Tajín appears to have been solidly within the Totonac cultural sphere then just as it is today. During its earlier stage of city growth (A.D. 100-500), El Tajín was strongly influenced by Teotihuacán and may well have served as a commercial outlier for the plateau metropolis in the supply of lowland tropical products. However, its greatest development architecturally and influentially occurred as Teotihuacán's power waned and finally disappeared (A.D. 500-1100), and therefore the construction of most of its spectacular urban core occurred about the same time as the emergence of the Toltecs.

            Although no fewer than 11 ball courts have been found at the site, testifying to the great ritual importance of the game, the crowning architectural achievement of El Tajín is the so-called Pyramid of the Niches. It received this name because each of its four sides is decorated by boxlike niches whose number is invariably described in the literature as totaling 365. This in turn has prompted the suggestion that the structure commemorated the length of the solar year and that small stucco figurines representing the prevailing deity of each day may have occupied the niches.

            However, it is readily apparent to anyone taking the time to actually count the niches that their number does not total 365. The reason for this is that the front, or east, side of the pyramid is surmounted by a staircase which disrupts the basic symmetry of the structure, requiring some of the niches to have been telescoped to about half of their standard size. Even so, one can count only 26 such features on each side of the staircase, yielding a total for the east face of 52. Inset into the median of the staircase itself are five panels, the four lowest of which are each composed of three smaller niches. It is uncertain whether the topmost fifth panel, which is noticeably damaged, likewise contained three niches or whether this contained only one central niche, yielding a total of 13 rather than 15 such embellishments in the staircase as a whole. It is additionally uncertain whether these 13 or 15 smaller niches are to be reckoned into the total for the structure as a whole. If not, the front face contains 52 niches; when they are included, it contains either 65 or 67 niches.

Figure 55.

The front, or east, face of the Pyramid of the Niches at El Tajín in northern Veracruz state. A major ceremonial center of the Totonac people (one of whom is seen in traditional dress in the foreground) at least since A.D. 600, El Tajín shows the influence of both Teotihuacán and the Toltecs. As explained in the text, there is some question as to how the structure shown above could have functioned calendrically. because it contains considerably fewer than the 365 niches so often stated in the literature.


            Despite the destruction of most of the pyramid's top tier, it appears that each of its four sides could have contained 5 niches, yielding a total of 20. However, if the top tier of the pyramid actually served as an open temple, its front, or east, side may have had a doorway in place of 3 of the niches and still preserved its structural symmetry. Thus, we must add either 17 or 20 additional niches at the top of the pyramid, bringing the total for the east side and topmost tier to a minimum of 69 (if we disregard the small staircase niches and allow for an open temple at the top) or a maximum of 87 (if we count all the staircase niches but allow for no open temple at the summit).

            The remaining three sides -- i.e., the south, west, and north -- are symmetrical in that each of them has a total of 87 niches. Thus, together these three faces yield a total of 261 niches, which when added to those on the east side and top tier of the pyramid come to either 330 or 348, depending on how you wish to count the latter. (However, so pervasive has the contention been that El Tajín's architectural jewel contains 365 niches that the scale model of the pyramid which was constructed for display in the Museum of Anthropology in Mexico City was distorted so that it would total this number!) When I attempted to point out in a scholarly article the discrepancy between the actual number of niches which can be counted and the theoretical figure of 365, my observation was dismissed with the comment that "the remaining [17 to 35] niches were buried beneath the staircase." If true, such an argument pointedly begs the question of how the pyramid could then have functioned as a counter for the length of the solar year.

            If, on the other hand, we recognize the futility of attempting to argue that the niches served to record a 365-day count, especially when anywhere from 17 to 35 of the niches were hidden beneath the staircase, quite another reading of the pyramid's purpose is possible. Suppose we take the 52 niches on the east face as one count, the 13 (or 15?) smaller niches in the staircase as a second count, and the 20 niches (or were there only 17?) in the top tier as a third count, we find three of the key multiples of the Mesoamerican calendar neatly defined in three separate architectural components of the structure. If the remaining three sides of the structure are conceived as forming a fourth count (totaling 261 niches), we have as close an accommodation to the length of the sacred almanac as architectural symmetry permits. It is thus possible that the Pyramid of the Niches may have functioned as a calendrical counter, but if so, it was more likely to have done so for the 260 days of the divinatory almanac than for the 365 days of the solar year.

            The dating of El Tajín's florescence and the overwhelming emphasis of its sculptural detail on human sacrifice (Kampen, 1972) strongly suggest that the site received considerable Toltec influence. Hieroglyphics repeatedly refer to a figure identified as "13 Rabbit," who may have been one of the city's leading rulers. Depictions of "Eagle Knights" also strengthen the impression that warriors and militarism played key roles in the city's development. But, aside from the possible calendrical associations of the Pyramid of the Niches, El Tajín demonstrates no real involvement in matters astronomical. Ensconced amongst a myriad of low limestone hills, it has a situation which all but precludes any kind of long-distance orientation, be it solsticial or otherwise. Located on the lower slopes of the Sierra Madre Oriental, it experiences a climate which is dominated by overcast skies and light drizzle through much of the year. Indeed, its very name commemorates the storm-god, and El Tajín has itself fallen victim to the destructive force of hurricane-driven rains. Commemorating astronomical cycles in stone appears to have been El Tajín's sole, but very distinctive, contribution to the evolution of the Mesoamerican calendar.


            Unlike the sequestered ceremonial center of El Tajín, Zempoala, the Totonac capital, was situated in a far more open setting near the coast of the Gulf of Mexico, about 40 km (25 mi) north of present-day Veracruz. No doubt because of its accessibility, it had very early been reached by Olmec influences from the south, for as we have already noted, it was one of the first ceremonial centers in the central Veracruz region to demonstrate a solsticial orientation -- in this instance, to Orizaba, or Citlaltépetl, on the winter solstice sunset. Its proximity to Cortés's toehold in Veracruz also made it the first Mesoamerican city of any size to be visited by the Spanish conquistadores.

            In the central plaza of Zempoala, just beneath the massive pyramids that frame its northeastern corner, are three intriguing rings of stone, each fashioned of rounded beach cobbles cemented together to form a series of small, stepped pillars. The largest of the rings contains 40 of the stepped pillars, the middle-sized ring has 28 such features, and the smallest ring numbers 13 stepped pillars around its circumference. It would appear that the three rings were used to calibrate different astronomical cycles, possibly by moving a marker or an idol from one stepped pillar to the next with each passing day (in somewhat the same way that has been suggested for recording the passage of time at the Pyramid of the Niches).

Figure 56.

The three stone rings of Zempoala as viewed from the top of the main pyramid. Inasmuch as the three rings are surmounted by 13, 28, and 40 steplike pillars, respectively, it appears that they were used by the Totonac priests as counting devices to keep track of eclipse cycles.


            Deciphering the smallest, 13-pillar ring is probably the easiest and most straightforward task, for one cycle of this ring very likely constituted a so-called trecena, or 13-day interval, of the sacred calendar. Used in conjunction with the middle-sized, 28-pillar ring, it could very well have defined a 364-day "year," much like that found by our "New World Hipparchus" at the outset of his calendrical experimentation. Or it might have defined the number of lunations in a year.

            But what was the base from which the 28-pillar ring was derived? Was it simply the number of trecenas within a given year? If so, it certainly would have been possible to distinguish between the first 20 and the last 8, so as to define the length of the sacred almanac, if that was desired. Or was the 28 derived from an approximation of the number of days between full moons? Or from the number of "lunar mansions" in which the moon "rested" on its never-ending celestial journey? (The latter concept is familiar in early Middle Eastern and Asian astronomy.) Whether it was used alone or in conjunction with the 13-pillar ring, a lunar association seems very likely.

            The largest ring is the most enigmatic, for no cycle based on 40 is known from Mesoamerica, although it obviously could have served to define two cycles of 20. Naturally, if it had been used as one component in defining a "year," then we might have expected to find some means of recording nine full circuits of the ring -i.e., 9 x 40 = 360 -but no such "device" is present. If it had been used in conjunction with the middle-sized ring, it would, of course, define an interval of 1120 days (40 x 28), which bears no relationship to either the sacred or secular calendar. However, had it been used together with the smallest ring of 13 pillars, it could have calibrated two full cycles of the sacred almanac, or 40 x 13 = 520 days. The latter, known as a double tzolkin in Mayan terminology, equates to three eclipse half years, and thereby provides a useful interval in predicting eclipses. (An eclipse year is the length of time it takes for the sun to move from one of its intersections with the path of the moon, or node, until it returns to the same intersection, or node. It measures 346.62 days in length. Hence, an eclipse half-year totals 173.31 days, and three eclipse half years add up to 519.93 days. In Mesoamerican terms, this value would be rounded to 520 days, or the equivalent of two rounds of the 260-day sacred almanac.) It is, therefore, quite possible that by using the three rings together, the Totonac priests were able to calibrate the movements of the moon closely enough so as to know when it might next be "devoured." In any event, there is every reason to believe that the three stone rings of Zempoala afford yet another bit of evidence testifying to the intellectual curiosity and architectural ingenuity of the early Mesoamericans.

(Note: The author has published a revised version of the interpretation of the stone rings at Zempoala in a later paper.  It may be accessed from the Home Page of this web site.)

(Return to Table of Contents) (Continue to Chapter 11)