It is time now to talk of my own role in this story. I have been fascinated by the Olmec since I was a graduate student in anthropology, particularly after reading Covarrubias’s exciting book, Mexico South. As I learned more about this controversial civilization, a conviction grew that Covarrubias, Caso, and Stirling had been right about it all along. For three subsequent field seasons I “labored in the vineyards,” digging relatively simple Formative-stage sites on the Pacific Coast of Guatemala and in northwestern Costa Rica, until I was fairly certain about what pottery and other materials from this remote epoch should look like. Then in 1964, I decided to concentrate on San Lorenzo.

San Lorenzo is not one, but a group of three related archeological sites. To reach the area, one travels to the grimy oil town of Minatitlán, situated on the Coatzacoalcos River about twenty-five miles above its mouth. Here one must search out and hire a dilapidated diesel boat that are the main form of transport on the river. The Coatzacoalcos and its tributaries drain all of the northern half of the Isthmus of Tehuantepec; at one point in its course, the river splits and flows about a large, swampy island called Tacamichapa, reputed by the local people to have belonged to Doña Marina, the famous Indian mistress of Cortés. The west branch, which skirts Tacamichapa, is the Chiquito River, and the boat enters it after two hours of travel. After another three to four hours, passing between banks lined with fields of tall corn alternating with patches of green jungle, the boat touches at its destination, the village of Tenochtitlán. This was our home for three seasons of excavations (1966-1968).

Walking up from the river edge through the village, one immediately notices that all of the native houses, built of poles or boards and thatched with palm fronds, are placed on artificial mounds, some of them quite high and long. Tenochtitlán is one of the three ancient sites in the area. It was so-named by a local schoolteacher, who, noting the size and number of ancient mounds there, decided that it must have been a great ancient city and so erroneously named it after the Aztec capital.

San Lorenzo is another of the archeological sites in the area, located one and a half miles south-southwest of Tenochtitlán. The third is the little, palm-shaded village of Potrero Nuevo, one and three-quarter miles east-southeast of San Lorenzo. They are known collectively to archeologists as San Lorenzo Tenochtitlán.

Again, it was Stirling who discovered the San Lorenzo Tenochtitlán group in 1945, having heard rumors in Coatzacoalcos that there had been found some large carved stones upriver in that zone. By spring of the next year, Stirling, accompanied by his wife, Marion Illig, and Philip Drucker, had uncovered fifteen of the finest Olmec basalt monuments ever seen, including five of the largest and most beautiful Colossal Heads known to date. While the 1946 project was devoted to excavations at the most important site, San Lorenzo (where they made their camp), the collections had never been studied, and the mystery of the age of the Olmec monuments there remained unsolved.

I decided to get some money for a three-year project at San Lorenzo Tenochtitlán, but before I did so, I visited the area in December 1964. In my innocence, I had not realized that I was walking into a hornet’s nest. The last archeologist to visit the zone, Alfonso Medellín Zenil, had almost been lynched there by the natives and left the village of Tenochtitlán under the protection of soldiers with submachine guns. Apparently the local populace had not taken kindly to the wholesale removal of their Olmec monuments to the Museum of Anthropology in Jalapa, Veracruz, without some form of compensation by the government, preferably in the form of a new school. The situation was still unresolved when I first went there. It has taken some patient diplomacy over several years to convince the local populace of our good intentions. In the end, we succeeded, in large part because of the good will left by the Stirling group twenty years before.

Finally, word was received that the National Science Foundation in Washington, an independent fund-giving organization entirely supported by federal money, had decided to back my archeological project over a period of three years. The first task was to build a camp, no simple job considering the spotty transportation along the river. For the first few months, we lived in tents, while the rain came down day and night.

Next, of course, was to begin digging. Priority was devoted, in our first season, to discovering some sort of cultural sequence for the area. Here the principles of stratigraphy had to be applied and we were presented with a made-to-order situation along the riverbank just below the village of Tenochtitlán. The swirling waters that rush downstream during each rainy season had cut away remains of a clearly stratified series of villages: Visible in the bank, beside numerous fragments of broken pottery, were charcoal-rich hearths, providing ample opportunity for radiocarbon determinations. At the same time, we began excavating some recently discovered monuments at San Lorenzo. The major culture exposed at the riverbank and at San Lorenzo turned out to have pottery similar to the Early Formative ceramics that I had been finding in my digging among village sites on the Pacific Coast of Guatemala, along with numerous Olmec figures. Most significantly, this culture, which we called San Lorenzo, proved to be definitely associated with the great Olmec monuments.

Naturally, I was anxious about the samples of charcoal from the 1966 season, which I had submitted to Dr. Minze Stuiver, the Dutch–born physicist who headed the Yale Radiocarbon Laboratory. When the results came in, they were, indeed, gratifying: The San Lorenzo culture, and therefore Olmec civilization in this area, dated to 1200–900 BCE. This was older by several centuries than even La Venta.

Building a sequence of prehistoric cultures is usually a complex business. After three seasons of stratigraphic work in the San Lorenzo zone, we could reconstruct some sort of prehistoric picture in which the San Lorenzo phase appears just as one of a series of intermittent occupations. It is believed that the area was first settled well before 1300 BCE by pottery-using farmers who began altering the San Lorenzo plateau. They were followed, in turn, by two other groups descended culturally from them. Then, after 1200 BCE, a band of outsiders with a vastly superior civilization took over; these were the Olmec of the San Lorenzo phase. They, in turn, declined around 900 BCE and disappeared, and perhaps San Lorenzo Tenochtitlán was completely abandoned to the the jungle. Yet another group came in the Middle Formative stage, possibly from La Venta itself, in what we are calling the Palangana phase, between 600 and 300 BCE. Again, the area is deserted, until as late as 900 years after Christ, when a great wave of pioneering agriculturalists once again populated the entire district; some of these were undoubtedly living in the Coatzacoalcos drainage when the first Spaniards came through, and our workmen from Tenochtitlán may be their descendants.

The next task was to completely map San Lorenzo itself, the place where Stirling had discovered most of the great stones. No previous map existed, so Ray Krotser, a retired civil engineer from California, was in charge of this project, which turned out to be no easy job. The basic equipment was relatively simple: a plane table, on which the plastic film used for the map was affixed; an alidade, basically nothing more than a telescope mounted on a ruler and which sits on the plane table; and a stadia rod, a collapsible pole marked off in meters and centimeters. The rodman would place the stadia rod over a point, and it would be sighted by Krotser through his alidade. Parallel cross hairs within the telescope enclose a certain interval on the rod; this figure multiplied by 100 produces the horizontal distance of the point over which the rod is held in meters. Other simple calculations provide the elevation of that point. Points of equal elevation are then connected by a contour line. This, in capsule, is how a contour map is made. But one has to take into consideration the heavy bush, the ticks (which were present by the millions), the mosquitoes, and, not least of all, deadly snakes like the fer-de-lance (we killed seven of these creatures during the 1967 season at San Lorenzo). Krotser discovered that only by sending large crews of machete wielders to cut down all the second-growth forest covering large parts of San Lorenzo could all the mounds be mapped, right down to the smallest ones on which undoubtedly had sat the thatch-roofed dwellings of the ancients.

After many months of hard work, spread over two field trips, a map at a scale of 1:1000 was at last ready. It shows a strange picture, indeed. We had known San Lorenzo as a plateau rising about 150 feet above the surrounding savanna-covered plains; we had also known that it was deeply cut into by ravines on its north, west, and south sides, and that it was within or on the edge of these deep gullies that Stirling had found the Colossal Heads and other Olmec monuments. However, it was not until we had seen the final map - and also done quite a bit of deep digging at San Lorenzo - that we came to a previously unsuspected conclusion: The ravines, the ridges that enclose them, and, in fact, the entire site as we now see it, represent a gigantic artifact, the result of human labor on a stupendous scale. Not only that, but the long, flat-topped ridges are obviously planned, for what purpose we cannot even guess. On the west, the Group C and Group D ridges, each about 100 feet long, are mirror images of each other: Every feature on one is matched by the identical feature on the other. The same thing is true of the pair formed by the much longer Southeast and Southwest ridges. Our deepest cuts in the San Lorenzo ridges reached culture-bearing layers down to twenty-five feet below ground level. There must he thousands upon thousands of tons of fill and debris in these finger-like constructions, all brought in basketloads on the backs of sweating Olmec.

The map also showed some other odd features of the site. There are over twenty depressions of various sizes and shapes dotting the surface of San Lorenzo. We called these lagunas, as they contain water except at the height of the dry season, but if they were cleaned out, they would probably provide water throughout the year. The lagunas are artificial, as can be seen not only from a trench we put into one of them (it had been lined with blocks of consolidated volcanic ash hauled up from the deepest ravines), but also from the geometric shape that two of them have: They are six-sided. We have no real idea what their purpose was, but according to our present evidence the lagunas we now see were another product of Olmec civilization, that is, of its San Lorenzo phase.

San Lorenzo is about three-quarters of a mile long (north-south), and its surface is dotted with several hundred earthen mounds. The central group is clearly planned, with long, La Venta-like mounds enclosing courts. However, far more common are the undistinguished little bumps that archeologists recognize as house mounds: From their distribution, we know that houses were often arranged on two or three sides of tiny family plazas, like extended-family dwellings. If there are about 200 of such house platforms, and five persons on an average dwelt within each house, then the resident population during the last years of the San Lorenzo phase might have been approximately 1,000. We dug into some of these house mounds and found broken metates for grinding corn still on the floor: Those that we investigated on the Northwest ridge seem to have specialized in the manufacture of stone beads and ear ornaments, for a number of stone tools used to fashion these items were recovered.

Stirling made some remarkably good guesses in his archeological career and was proved wrong in only one: In his 1955 article on San Lorenzo Tenochtitlán, he surmised that most of the great carved stones which he had found on the edge of slopes or bottoms of the deep ravines at San Lorenzo had been pushed there by an invading people. Solving the mystery of what happened to the monuments, and why Stirling found them as he did, provided some of our most exciting moments of discovery.

In our first season there, my assistant Richard Diehl, then a graduate student at Pennsylvania State University, was taken by local peasants to two stones projecting just above the ground, at different parts of San Lorenzo. Both being of basalt, Diehl immediately concluded that they were buried monuments and began digging. Both proved to be intimately associated with strata of San Lorenzo debris - potsherds, broken pottery figurines, and the like. Monument 21 (we number every monument in the order of its discovery) turned out to be a relief representation of a running animal, perhaps a dog or coyote. More important than its carving, however, was the fact that it had been placed face down over an offering of serpentine axes and blanks that had been used to make these tools. In other words, it had not been simply destroyed (it did show evidence of attempts at defacement as well), but had been carefully abandoned and buried with some show of ceremony. This offered us the possibility of finding other such buried monuments under the surface of San Lorenzo and associating them with our cultural sequence. The other stone, an enormous and brutally defaced altar (Monument 20), was similarly tied in with the San Lorenzo culture after careful excavations by Diehl and an archeologist from the University of Veracruz, Francisco Beverido.

But the archeological “payoff” came in March of 1967. We had known of an upright, stone stela sticking above the southern slope of our Group D ridge, on the western side of the site. In fact, Stilling had written about it in his field notes of 1946. A hunch led me to try excavating this undistinguished slab, Monument 23, to see if I could relate it to cultural layers and also, perhaps, to find offerings at its base. I never did find any offering, but this is of small moment compared to what was found. March 2 started out as gray and drizzly, typical of what we began to think of as “Olmec weather.” I laid out a rectangular cut and had a laborer start digging, stripping off the deposits according to “natural” stratigraphy, that is, following differences in color, texture, and so forth.

The deposits proved to have few sherds, and I began to wonder whether the ridge itself was not mainly sterile, a natural formation (we had yet to map this area); I was very much mistaken. Since there was little room for Pedro Camaño, the workman whom I had set to this task, we laid out another square (Cut 2) to the north, in order to approach the stela from the side. This was the luckiest thing I have ever done, as my journal for March 8 records:

The main event of this day was the discovery of Mon. 34, a magnificent life-sized kneeling figure encountered in Cut 2. . . . It is just north of the stela, facing east. . . . There are no arms to the figure: rather, in their place are two ratcheted disks, perforated, which must have held movable stone arms – the largest jointed figure known! The right leg is tucked under the body.

It was time for a little reflection. Here was Monument 34, headless and therefore purposefully destroyed, but placed on a red gravel floor and covered up with a special fill containing a great deal of limestone and fragments of bentonitic rock brought up in basketloads from the deepest ravines, an event that we know took place late in the San Lorenzo phase. The stela had been set in the same floor and likewise covered up. My guess was that there would be more monuments similarly positioned, in a direct line heading north.

I turned out to be right. In the next month or so we hit upon, in this imaginary line, Monument 38, a fragment broken off some great altar; Monument 37, a crouching, headless figure of a gigantic jaguar that had been equipped with almost walrus-like tusks; Monument 40, a loose drain stone; Monument 41, an enormous, four-sided column with a brutal, low-relief figure of an Olmec deity; and the tiny Monument 43, perhaps the most diminutive Olmec “monument” ever found, representing a fantastic eight-legged creature, probably a spider. In place of eyes, it has the five-dot symbol known to be the Mesoamerican sign for jade.

Farther west on the same Group D ridge, I had been shown another stela similar to Monument 23 but much more broken. Careful excavations showed this also to be tied into the San Lorenzo phase deposits. One face was carved in a relief with a snarling, profile creature, somewhat resembling the old Olmec were-jaguar, or Rain God, but which I think (on the evidence of the dragon-like body and the crossed bands in the eyes) to have been the great god known as the Feathered Serpent. With our experience of the north-south line to the east, I began trenching away from this stela (Monument 30). Sure enough, more buried monuments began appearing in a line running west along the edge of the ridge. The last picked up was a magnificent sculpture of a caped personage holding in his hands the head of the deadly fer-de-lance, still so common in the area and probably an object of awe to the Olmec. Stylized wings appear in back of the snake’s head, suggesting that this is another representation of the Feathered Serpent. Unfortunately, time had run out on the 1967 season; the thunder clouds of the approaching rains were already upon us so we could not determine how far west the line ran.

The whole series of monumental discoveries threw an entirely new light on San Lorenzo. We concluded from our work that the stones recovered by Stirling from the ravines had not been pushed there, but had fallen into them from the destructive forces of several millennia of erosion on the ridges above; that all of the San Lorenzo monuments had been carefully buried on a special floor on the top of the ridges or elsewhere on the site, then covered up with a specially selected fill; and that this had taken place near the end of the San Lorenzo phase. The natural question is; Who did it, and why? The fact that the same kinds of pottery and figurines typical of the San Lorenzo phase continued to be made for a while thereafter led us to think that some kind of internal Olmec revolt took place around 900 BCE at San Lorenzo, as there is no evidence of an outside invasion.

The amount of pent-up hatred and fury represented by this enormous act of destruction must have been awesome, indeed. These monuments are large, and basalt is a hard stone. Wherever possible, heads were smashed from bodies, “altars” were smashed to pieces, and strange, dimpled depressions and slots were cut into Colossal Heads. There are no signs that wedges or the fire-and-water treatment were used to break up the larger stones: I suspect that they built huge tripods over monuments, hoisted other monuments over these, and let them drop from great heights. Why was this done? Because the Olmec monuments must have stood for the class of leaders that held the tributary, populace in such a firm grip, forcing from them incredible expenditures of labor. These stones must have been the symbols of all that had held them in thrall, and they destroyed these symbols with fervor. But the Olmec must also have feared their power after the act, for by burying them with such care, they removed the hated objects from their sight without incurring their posthumous wrath.

The drain stone that was found in the north-south line of buried monuments is U-shaped in cross section and pecked out of a single piece of basalt. Other such stones were seen by Stirling lying in a ravine south of the Group C ridge, along with flat pieces of basalt. He suggested that the U-shaped stones had once been placed end-to-end and fitted with covers to make up a drain system, though he never did discover one so arranged. But fortunately, we did, and its excavation consumed the belter part of the 1968 season at San Lorenzo.

One day in April of 1967, my workman Agustin Camaño (the brother of Pedro Camaño who became, in effect, an archeologist during this dig) mentioned that he had seen a section of a real drain system with stones just like those encountered by Stirling protruding from the slope of the same ravine, and that water gushed forth from it during the rainy season. I climbed with him down into the jungle-filled gully, and there it was, crying out for excavation. This was obviously a job for Ray Krotser, with his years of experience in civil engineering. While his wife, archeologist Paula Homberger Krotser, concentrated on trenching several pyramids elsewhere at the site, Krotser, after completion of the map, worked full time on the drain. The trouble with our drain was that in some places, it was buried up to fifteen feet deep, hardly presenting an easy task to follow it out.

By April of the following year, Krotser was fairly sure he had found all of the system as it was, although part could have been destroyed in ancient times. The U-shaped stones had been placed end-to-end exactly as Stirling had guessed, within a shallow trench, and then covered over with the capstones. Finally, the entire system was buried in fill. There is a principal drain line running west at a gentle degree of slope from the San Lorenzo plateau, and three subsidiary drain lines that tie into it from the south at a steep angle. The sophistication of the joints between the branch and main lines prove the Olmec to have been accomplished engineers.

The size of the operation is particularly impressive (nothing that the Olmec did was on a small scale, or easy). The main line as we excavated it is 558 feet long, with ninety-eight feet of subsidiary lines. There are no less than thirty tons of basalt in the system; all brought in from the outside, and all worked with great precision and without the benefit of metal tools. What was the system made for? We wish we knew. Several of the U-shaped stones at the head of the main line are perforated, as though these were the entrances for the water. If this was the start of the system, then what had it been draining? Mention has been made of the ponds or lagunas on the surface of San Lorenzo. We know from the geometric shape of some of these and from a cut made into one laguna that they are man-made. Several lie to the north and south of the drain head; Ray Krotser calculates that these lagunas could once have been part of a super-laguna, the approximate center of which would have been above this point. This still does not tell us why they would have built the lagunas, and less why one of them should be drained with such an elaborate and time-consuming system. One thinks of the sacred pools of the temples and palaces of India and Ceylon, to which priests and the faithful repair for holy ablutions during festivals. Quite possibly we are confronted with the same thing among the Olmec of 3,000 years ago. At any rate, whether ritual or not, we have here the first system of water control yet known for the New World.

Every archeologist has sometimes wished he had some instrument in his hands that could see what lies underground and direct him to the best places to dig. Since World War II, the physical scientists have come close to realizing this dream. Among the instruments that have been developed in the United States and abroad are the resistivity meter, which detects objects or constructions by passing an electric current through the soil: seismic instruments, which perform roughly the same task by means of shock waves; and periscopes, which can be lowered through holes drilled through the roof of underground tombs. None of these, however, has quite fulfilled the promise that they seemed to hold out to archeology: There are just too many “bugs” in these and other systems. Ground water, for instance, can seriously affect the readings of a resistivity meter.

In a different class altogether is the magnetometer. This was developed to detect differences in the intensity of the magnetic field over any part of the earth at a given moment. Such variations are called anomalies, and they are sometimes caused by buried objects containing a fairly large amount of iron. Magnetometers have been used to detect mineral and oil deposits, locate avalanche victims in the Alps, and have even been successfully used in the search for the lost submarine Thresher. In 1964, Varian Associates of Palo Alto, California, leaders in the magnetometry field, began work in locating the ancient Greek city of Sybaris, deeply buried under layers of silt since medieval times, but known to have existed as a flourishing (and sybaritic) settlement in southern Italy. Working in cooperation with Dr. Froelich Rainey of the University Museum (University of Pennsylvania), long an advocate of the use of physical detection methods in archeology, the magnetometry specialists located walls and buildings of the famous city.

In the 1968 season, I was contacted by Dr. Rainey and the Varian people. They wanted to try out an extremely sensitive model, the cesium magnetometer, on San Lorenzo: Here, the high magnetic intensity of the iron-rich basalt monuments and the low background “noise” of the surrounding deposits should pay off spectacularly. It sounded too good to be true, but we decided to give it a try. The Varian team, headed by Sheldon Breiner, arrived on March 27, bringing with them an instrument that looked like a tin can mounted on a rod. The “tin can” contained cesium gas, which produces a radio frequency signal proportional to the intensity of the anomaly passed over: It is so sensitive that it can measure variations of one part in 5 million of the earth’s magnetic field. The signal is picked up by a small receiver worn on the belt, and transformed into clicks and numbers.

The next day, Breiner and his assistants began walking over a part of San Lorenzo, watched by curious but skeptical archeologists and workmen. All at once the magnetometer began clicking furiously. “Dig here,” said Breiner. Eventually, the workman’s pick grazed over a monument, buried at exactly the depth predicted by Breiner. Not far away, the instrument began to record another anomaly, and again a monument was excavated in that spot. Within a week’s time, we had had the opportunity to dig in four of the places shown as magnetic anomalies, and all proved to be broken basalt monuments. Three were but fragments broken from much larger stones, but one (Monument 52) turned out to be one of the most beautiful and complete Olmec statues ever found at San Lorenzo: a standing were-jaguar with snarling mouth and cleft head, the great Olmec Rain God himself – a fitting close to our archeological work at San Lorenzo. The astonishment of the non-physicists standing around to watch was complete.

In a survey made over a fraction of the site, Breiner and his assistants were able to place on our map about thirty-five other anomalies. We regretted that we did not have this “magic wand” with us from the beginning of the project.

We answered many questions at San Lorenzo over three seasons but left several unanswered and raised others that had not even been predicted. For example, the rulers of the area, the men who created the great Olmec civilization of the San Lorenzo phase, came from elsewhere - but no one knows where. At the time of their arrival, the place was already inhabited, and it may well be that three of the crudest relief sculptures found were the work of pre-San Lorenzo people. This elite class surely must have come from a region where they already knew how to work and transport great basalt boulders. The perfection of their carving suggests a long prior period of artistic development. I have suggested a homeland in the Tuxtla Mountains, but their origins remain a mystery.

I also feel that high culture came to an end at San Lorenzo through a violent internal revolt. After 900 BCE, when San Lorenzo began returning to the jungle, the torch of Olmec civilization must have been passed to La Venta, the island capital in the Tonalá swamps fifty-five miles away.

The most important questions that can be asked are these, however: What were the factors that allowed San Lorenzo to be the first civilized center of Mesoamerica and probably of the New World? How did they manage to maintain this high level for three centuries? What were their population, their social and political structure, and their economy? And why might revolution have finished them off? There are no written documents for this epoch, as distant from us as the Trojan War, so we can never be exactly sure of what went on. But there are other ways to the truth, which we will examine shortly.