I. TRANSPORT OF STONE MONUMENTS TO THE LA VENTA AND SAN LORENZO SITES
Joseph S. Velson and Thomas C. Clark
1
I. TRANSPORT OF STONE MONUMENTS TO THE LA VENTA AND SAN LORENZO SITES
Joseph S. Velson and Thomas C. Clark
INTRODUCTION
Because monumental stone buildings and sculptures are often the
most impressive, readily accessable and best preserved artifacts of a
culture, archaeologists have long been interested in the technology
associated with the quarrying and transportation of large stones.
Working with the problem of a "heavy" lithic technology need
not offer only information about ancient engineering practices, but
also a point from which some demographic and socio-political implica-
tions may be explored.
The complexities of an ancient heavy transport operation are
more readily seen by breaking the operation down into its component
parts. It then becomes possible to assign quantitative values such
as man-power and man-hours to each of these components: quarrying
operations, construction of transportation aids, land and/or water
travel, maintenance and supplies, etc.
The total amount of manpower involved can then be estimated.
If we can determine how many individuals were engaged in a project,
we may be able to estimate the size of a minimum support population.
It may also be possible to say that a certain number of these indiv-
iduals were full-time non-agricultural specialists, and that a certain
number of other individuals were part-time workers normally engaged
in other presumably subsistence related activities.(l)
Because it appears that so much can be learned from such
studies, we have selected one particular New World culture on which
to focus our attentions. Specifically, we are considering the various
stone monuments--including the altars, stelae and colossal heads--
that are characteristic of the Olmec culture of southern Veracruz,
Mexico.
Geography and Climate. The Olmec culture area was concentrated
in that part of the tropical lowland Gulf Coast plain bordered by the
Papaloapan River on the west, the Tonala River on the east, the highland
area to the south, and the Gulf of Mexico to the north (Drucker 1947).
(1)    Parenthetical numbers refer to Notes at end of text.
2
This area covers between 6,200 and 7,000 square miles (Coe 1962:86;
Bernal 1969:17), although the area surrounded by the four major sites
comes to not more than 2,500 square miles (Heizer 1968).
With the exception of the intrusive volcanic highland of the
Tuxtla Mountains, the Olmec coastal lowland area has been described
as a plain which "varies between slightly undulating in most places
to gently rolling as the elevation limit is approached" (Poleman 1964:
31). The coastal plain is composed mainly of the alluvial deposits
from the river deltas and flood plains of three of the largest river
systems in Mesoamerica--the Papaloapan, the Coatzacoalcos, and the
Tonala-Blasillo. The Tuxtlas are a small isolated range whose highest
peak does not exceed 6,000 feet. Here lies the source of much of the
stone material used by the Olmec (Williams and Heizer 1965). Much of
this coastal plain is flooded during the rainy season with the excep-
tion of some uplifted Pleistocene domes overlying salt and oil deposits.
All, or nearly all, of these islands are known to have had Olmec sites
on them (Heizer, personal communication).
The heavy tropical jungle growth was the most prevalent vege-
tation formation in pre-Columbian times, but has been severely restricted
by intensive logging operations and extensive use of grazing lands.
The rain forest consists of essentially three stories of trees which
range up to 50 meters in height. The dense canopy forest with the
taller trees includes laurel, tinco, Ceiba, and mahogany. The inter-
mediate and lower stories consist of palms, magnolia, fig, rubber
tree, immature members of the upper story, numerous bushes and shrubs.
The forest is characterized by many climbing plants, some of which
achieve lengths of fifty meters or more. There are several other
vegetative and faunal adaptive zones in the area--savannah, beach,
mangrove forest, and river levee forest to name a few--but none are
nearly as extensive as the tropical rain forest. (West, Thom, Psuty
1969; Wagner 1965).
The climatic variation in the area is due largely to its
geographical position between the middle and lower latitudes of the
northern hemisphere. The land below 1,000 meters elevation is often
called "tierra caliente" and has average yearly temperatures between
20 and 30 degrees Centigrade. The relative humidity is well over 80%
for most of the year. The rainy season runs from late May to January
with a peak period in June and a maximal peak in September. The
entire area receives over 2,000 mm. of rainfall annually, and some
portions of the Tuxtlas receive over 5,000 mm. During this time
many of the rivers over-run their banks and flood the surrounding
countryside for many miles, making almost impossible any kind of
prolonged construction project. The area is blanketed by a dense net-
work of seasonal streams which make overland travel during the rainy
season very difficult.   The dry season, which extends from early February
to late May, is not really "dry" as the rains never actually stop
completely. This is the time of nortes, storms characterized by high
3
winds and intense rains which are caused by large masses of polar air
which have penetrated far south.(2)
La Venta and San Lorenzo. The Olmec culture is known essentially
from four sites:  La Venta, San Lorenzo, Tres Zapotes, and Laguna de los
Cerros; and is dated to the first millenium B.C. Only La Venta and San
Lorenzo have been excavated in sufficient detail to aid our consideration
of ancient heavy transport, so this inquiry will be limited to these
two sites.
The La Venta site is located on an island to the east of the
Tonala River about 12 miles from its mouth. The island itself is an
emergent salt dome with a dry land surface area of about 2.1 square
miles. The mounds of La Venta are oriented in a bilaterally symmetrical
manner along a center-line which runs 8 degrees west of north. Most of
the investigations at the site have taken place in the northern-most
areas--centering around the 100 foot fluted pyramid of Complex C and
the area to the north of that designated as Complex A (Drucker, Heizer,
and Squier 1959). Excavations have revealed a large number of stone
monuments (ibid; Heizer, Graham, and Napton 1968) as well as extensive
evidence of large massive offerings (Drucker, et.al. 1959). The site
appears to have been continually maintained and modified over a period
of 400 years, from about 1000 to 600 B.C. (Berger, Graham, and Heizer
1967). The constructions are apparently of a religious nature as no
occupation or trash debris has been found within the site area
(Heizer 1961:45). The massive offerings, the fact that much of the
materials for sculpting and for refurbishing the site were transported
from great distances (Williams and Heizer 1965), and the large amount
of manpower that would have been required for general maintenance and
construction indicate a large, well organized labor force which was
continually involved in the up-keep of the site (cf. Drucker 1947; 1961;
1952; Heizer 1959; 1962).
San Lorenzo is one of three sites clustered along the banks of
the Rio Chiquito, a branch of the Coatzacoalcos River about 50 miles
upriver from the modern town of Coatzacoalcos (Stirling 1955). The
site, which is only 1.2 km. long, is situated on a plateau which rises
above the surrounding savannas.   The florescent Olmec occupation at the
site is known from the San Lorenzo Phase, which has been placed within
the 1200-900 B.C. time span (Coe, Diehl, and Stuiver 1967). The site
is also characterized by a large number of stone monuments in the Olmec
style (Coe 1968:69; de la Fuente 1973; Clewlow 1974), as well as major
earthworks. The entire plateau has been modified and tremendous amounts
of earth moved to construct the artificial ridges on which the site
rests. The presence of what appears to be a complex drainage system
represents another example of the tremendous expenditure of labor at
the site (Coe 1968:57). The site has been extensively mapped and photo-
graphed by Coe and unlike La Venta, it appears to have had a resident
population of not more than 1,000 individuals who were probably
supported by the farmers of the surrounding area (Coe 1967; 1968; 1969).
4
Stone Monuments. The number of stone monuments known from San
Lorenzo and La Venta is quite large. Recently Clewlow (1974) has sub-
jected over 200 of these to a stylistic and chronological analysis.
These range in size from sculptures weighing only a few pounds to La
Venta Altar #1, the largest of known Olmec pieces, which weighs 36.5
tons. The maximum weight which could have been lifted by the ancient
Olmec is uncertain. Monument 34 at San Lorenzo, which weighs 1,000
pounds, was carried by 17 workers on a litter made out of poles (Coe
1965:79). Heizer (1966: fig. 6) shows 35 men carrying the weight of
a 1.5 ton andesite column at La Venta. Perhaps a weight of 3 tons
could be lifted, but the number of men involved would, due to sheer
nlmbers, be so far away from the weight being raised that the entire
litter might collapse due to the concentration of the great weight in
the center. However, the majority of Olmec stone monuments do fall
within this "portable" category, i.e., less than three or four tons.
We have arbitrarily limited our inquiry to consider only
stones which weigh over five tons. With stones of this size there is
little doubt that they would have been dragged. By working with the
idea of dragging a large stone we hope to better visualize the planning
and logistics that would have characterized such an operation.
The large monuments moved by the San Lorenzo and La Venta Olmec
include eleven colossal heads, three stelae, and over a dozen altars
and other stone monuments (See Appendix A.).
Source of Stone. Basalt was the most common material used by
the Olmec for their monumental artistic endeavors, although andesite
and schist were also known. Although the basaltic lavas used could
have come from several sources, the place of origin of most of the
material was in the area of Cerro Cintepec, an extinct Plio-Pleistocene
volcano located along the southern flank of the Tuxtla mountains a few
kilometers southeast of Lake Catemaco (Williams and Heizer 1965).
Along the slopes of the Cerro Cintepec, the basalt used by the Olmec
occurs as naturally formed boulders already detached from the rock. (3)
Other Olmec monuments had their origins at other places in the
Tuxtlas. Cerro el Vigia, a volcano located about 4 kilometers west
of Santiago Tuxtla, appears to be the source of much of the basalt
used at the site of Tres Zapotes (Heizer, Smith, and Williams 1965).
Although a number of columnar basalt exposures have been examined in
the Tuxtla Mountains, none of these prove to have been the source for
those used at the La Venta site. Other lithic materials were obtained
from Volcan La Uni6n, over 100 kilometers southeast of the La Venta site.
Limestone slabs from Chinameca, approximately 60 kilometers west of
La Venta were transported to the site (Williams and Heizer 1965:6-8).
The Olmec thus appear to have acquired their lithic materials
from several locations, some of them over 100 kilometers from the site
to which they were transported. For the purposes of this paper, it is
not possible to deal with each site and each lithic source known. We
have decided to limit our inquiry to the large stone monuments of San
w
0
L)
w
a-
LL~
0
0
0
0
0 )- . _
ors
C.() i;
0 10
0 0
0) a
400
(D
P--
c
0.
0)
m
01)
4-
wk-o
CH
av
0
U)
a)
C)
0
U,)
-C
"-4
FX4
6
might have been established.   These would have doubtlessly served a
multiple function in the practical development of trade and communica-
tion routes.
This overland route would have been the most economical in
terms of distance, 65 kilometers in all. However, San Lorenzo is not
situated on the level alluvial plain but on a plateau which rises to
about 340 feet above sea level (Coe 1967). This means that a heavily
laden sledge would have had to be drawn up steep irregular inclines
to reach the site.
Route 2 is suggested by M. Coe who believes that the stones
"must have been floated down on rafts to the Gulf of Mexico and along
the coast to the mouth of the Rio Coatzacoalcos, and dragged from the
river up to the San Lorenzo Plateau with ropes" (Coe, Diehl, Stuiver
1967). It is not clear what is meant by "floated down on rafts to
the Gulf." If this means drawing a heavily burdened sledge over a
ridged volcanic landscape to the San Juan-Papaloapan River system,
the route is so difficult to negotiate that it may be readily discarded.
Traversing the mountains to the east to gain access to the Gulf is
likewise improbable. The only really feasible route utilizing the
Gulf of Mexico would be to descend from the Tuxtlas in a southeasterly
direction and then turn due east. The mountainous terrain would be
avoided, and though longer in distance the route may have seemed
attractive because less energy would be spent dragging the stone on
level ground or transporting it on water. One difficulty presented by
this model is that of approaching the Gulf from the Veracruz coastal
plain. Surrounding Laguna Ostion are expansive swamps abetted by the
drainage of the Rio Chalcalapa. Though this obstacle might have been
eliminated by draining the marsh or crossing with an earthern causeway,
this would involve a large labor expenditure the fruits of which would
not have lasted through the rainy season. Drucker (personal communica-
tion) suggests that during the rainy season the Rio Calzadas, which
enters the Coatzacoalcos near its mouth, would have enough water to
serve as a possible avenue of transport. Coe's concept of water
transport is useful though this particular route presents some
difficulties.
Transport on water is advantageous for several reasons. The
surface of a large river (or the sea when calm) has fewer obstacles.
Less manpower is required to move the cargo. A river may not always be
the most direct route but is often the only possible route. Bernal notes
that "the rest of the (Olmec) area with the exception of a few humid
plains and the swamps, was and is still covered with tall vegetation
which in reality is an impenetrable jungle whose only open spaces are
those cut by the rivers that form the only possible means of communica-
tion." (Bernal 1969:18-99).
Route 3 appears to be the most feasible. This involves leaving
the Tuxtla highlands following the easiest contours in a southeast
direction, crossing the coastal plain in the same direction to arrive
7
at the banks of the Rio Coatzacoalcos perhaps upstream of Minatitlan.
From there the monument would be transported up the river to the site.
This route is approximately 114 kilometers in length. Though it is 49
kilometers longer than the first route, the journey on land is more
consistantly level and the advantages of water travel are exploited.
The first part of this route, the overland trail from Cerro
Cintepec to the Rio Coatzacoalcos (58 km.), might have been cleared
specially in anticipation of the stone movement. This assumes that
Cerro Cintepec basalt may have been quarried at irregular, perhaps
quite lengthy, intervals. Because it seems that the stone was removed
at periodic intervals from its source (Heizer 1961), it appears more
probable that the route from the mountains to the river was a well
established "road".
Cerro Cintepec to La Venta. We have examined two transport
routes from Cerro Cintepec to La Venta.
Route 4 is the least probable, consisting simply of a straight
line overland route. This route is complicated by having to cross two
major rivers, numerous smaller ones, as well as the watery swamp
surrounding the isLand of La Venta. The rivers are too wide to be
bridged, so rafts would have to have been constructed to ferry the
stones from one side to the other and finally up to La Venta. We
concur with Drucker who discredits this route: "There are some swampy
sections that couldn't have been traversed by heavy loads, and some
rather rough (though not terribly high) terrain between the Coatzacoalcos
and the Tonala rivers that would be difficult to roll or drag big stones
across," (personal communication).
Route 5 takes full advantage of the two large rivers and the
Gulf Coast waters. It appears to be the easiest and quickest route.
The monument would have been dragged from the Cerro Cintepec to the
Coatzacoalcos, perhaps along the same path described for route 3. At
that point it would have been transferred to some water-transport
structure to be paddled or towed to the Gulf. The route then follows
the coastline to the mouth of the Tonald where it turns inland to the
La Venta site. The distance from quarry to site would be approximately
139 kilometers, only 33 kilometers longer than the direct overland
route.
There are certain conditions which are most favorable for
aquatic transport. During the months of rainy season the rivers are
flooded because of the heavy rainfall. Today as much as 122 inches
of rain annually falls on MinatitlAn (Bernal 1969). During this season
river travel as well as travel on land is difficult. The rivers are
not just swollen and raging, but encumbered with tree trunks and,
debris. The rainy season overlaps with the nortes which intermitantly
hinder travel on land and particularly on water. The time which appears
pptimal for transport of heavy monuments is dry season. Navigation
on the open sea would have been most likely when weather hazards were
8
least likely to have occured, perhaps just after a norte. Depending on
the stability of Olmec marine craft there may have been comparatively
few brief periods in which sea travel could have been negociated.
Mode of Land Transport. An important part of the problem of
transportation of monoliths is the mechanism used to carry the heavy
stones on the land. Not having access to wheeled vehicles or domes-
ticated beasts of burden that might have facilitated carriage, Olmec
engineers had to rely on resources locally available to move stones
over the routes previously discussed. Of course men can double as
draft animals and must have done so in the movement of Olmec sculpture.
It is possible that monuments were dragged overland without any sort
of helping device, but this is hard drudgery. Conducting a many ton
boulder which rests directly on the earth by hitching many teams of
men to it would have been wasteful of time and energy, due to the
problems that would have been incurred by friction, irregular terrain,
and possible damage to the stone itself.
Sledges. The simplest perhistoric device to move large
monuments on land was the sledge. Atkinson suggests the use of sledges
in Neolithic England and documents the use in modern times, despite the
advent of the wheel, in the British Isles and among the Naga people of
Assam (1961). Heyerdahl observed the use of miro manga erua (a sledge
device) on Easter Island and concludes that a similar idea was used
anciently in the dragging of the monolithic sculptures (Heyerdahl 1958:
149). Sledges were also employed by Assyrians and Egyptians in their
transport of monumental sculpture (see plates 1-3).
Heyerdahl (ibid.) describes the sledge on Easter Island as a
Y-shaped figure with crosspieces. This seems nearly identical to the
ancient model described by Atkinson in which "the strongest form would
have been one whose main longitudinal members were formed of a natural
pair of forked tree trunks, in the shape of a modern tuning fork, joined
by nature at their bases and braced transversely by cross-members
morticed and pegged" (1961).
While this might have been a useful design for some areas it
may not have been for the Olmecs. The difficulty in locating forked
trunks of the desired thickness, length, and shape far outweigh the
advantages of strength. Transporting such a "natural" sledge to the
quarry site would involve considerable difficulties if the wood were
not available locally.
A sledge made up of two parallel beams reinforced with a series
of cross-members could be transported to the quarry and assembled there
with much less effort.
In the case of La Venta Altar 1 (which we have estimated to
weigh 38 tons before final sculpting), three runners might have been
used to provide additional support.
9
We have developed a hypothetical model (Figures 2 and 3) of a
sledge similar to the Egyptian example described by Solver (1940).
The sledge measures 12 ft. X 11 ft., while the base measurements of
Altar 1 are 8-1/2 ft. X 9 ft. The sledge is slightly longer and wider
to leave room for wedges which may have been provided to prevent the
monument from slipping. The idea of a long, narrow sledge seems
unfeasible due to tne danger of uneven support beneath the runners.
Should the two ends of the sledge become suspended on high points along
the track, the midsection might collapse'due to inadequate support.
The ends of the runners are tapered upward to facilitate the introduction
of sleepers beneath the sledge while dragging is in progress. The wood
had to be strong and hard. For our calculations we have used the data
for mahogany.
To be carried successfully, the stone would have to have at
least one flat surface--this would be the side that would rest on the
sledge.  Some boulders may have h'ad a naturally flattened surface (this
could have been a factor in selection of potential monument material)
but if they did not, one had to be prepared.
Using levers and pulling with ropes, the stone could have been
turned on its flat side (if it wasn't already) and dragged onto the
sledge, which might have been approached by a slight ramp of earth.
An interesting alternative may have been easier: assuming the flat
side-was vertical to the ground at some point after being planed, the
sledge would be assembled onto that surface affording a custom fit.
Then with levers and hauling with gangs and ropes the monolith was
rolled over so that it sat securely on the horizontal sled. Turning
the monument on its level side could have been aided by jamming stones
under the side being jacked up as the Easter Islanders did in the
erection of their monoliths (Heyerdahl 1958).
Once the monument was laid in its most stable position shims
might have been slipped into the interfacing area of rock and sledge.
The monolith would have been securely lashed to the sled with fiber
ropes, sinew, or vines. These bindings (like the ropes for pulling)
would have been replaced as strain and wear broke them. Once the
monolith was mounted on the sledge it probably remained there until it
was unloaded at the destination, which in the case of Altar 1 was La
Venta.
Sledge Dragging and Manpower. Sledge dragging no doubt involved
a large number of men; to make an estimate as to the manpower required
we have looked at other similar examples of stone transport.
In experimenting with a copy of the bluestone weighing 1.5 tons
and placed on a simple wooden sledge, Atkinson (1959) found that 32
schoolboys could just pull the sledge and its load up a four degree
grade. This works out to 109 pounds of pull per man. Heyerdahl (1958)
describes an experiment in which 180 men drag a 12-ton Easter Island
monument without a sledge. This required 133 pounds per man. An account
10
of the construction, transport, and erection of the Seringapatam obelisk
in the early 19th Century tells that "the number employed at one time,
on the drag-ropes...was about 600 men."   (Wilkes quoted in Kennedy 1821:
312). This stone weighed about 35 tons, thus requiring 116 pounds pull
per man (cf. Barber 1900). A 132-ton Egyptian obelisk and sledge
required 5,585 men to transport it across the desert. Adding a minimum
of 5 or 6 tons for the weight of the sledge we see that each man had
to pull 49 pounds. King Mehthuhotpe IV of the XI Dynasty sent an
expedition to the Wady Hammamat quarries, numbering 10,000 men, to quarry
stone for a large sarcophagus. The lid was dragged to the Nile by
3,000 sailors. Although sources differ as to the weight and size of
the lid, the range is such that each worker would have had to exert
between 12 and 20 pounds of pull (Erman 1894; Clark and Engelbach 1930:
32; Breasted:vol.1:448). Wilson (1888:584) calculates the weight of
the Menhir of Lochmariaquer at 347 tons, and Salmon believes that 4,500
men each pulling 165 pounds were needed to drag it. Grant cites an
example involving the removal of a two ton statue from Eleusis using
150 men to drag it (27 pounds/man), but also describes the transport
of an 11-ton marble sculpture found by Newton near Cnidus. There, 100
sailors were said to have moved this weight, which would have required
a tremendous pull of 200 pounds per man (Grant 1966:131,180). Wilson
(1882:226) describes a 7.5 ton stone transported by 92 men at 113 pounds
per man; and Kida (1912:5) recounts how, in 1908, a 155-ton Japanese
megalith on a sleigh was dragged by 550 men using iron chains. This
required 58 pounds of pull per worker. An example from the New World
is given by Howells (1960:161) who quotes Cowgill (1957) as saying that
the 8-ton Stela 18 at Uaxactun was dragged by 160 men at 100 pounds per
man.
We have presented here a broad range of figures, the discre-
pancies in which may be attributed to a variety of factors ranging from
the incorrect recording of data and poor observation technique to the
failure to mention the use of mechanical devices or the physical
condition of the workers. Most of these examples refer to short range
transport: a few hundred yards to a mile or two at the most. Maschet
made a study of the continuous potential pulling power of man and
concluded that "for steady pulling at the rate of 1-1/2 miles per hour
for 8 hours per day, it falls as low as 30 lbs," (Maschet in Barber
1900:41). The figures obtained from the Egyptian examples would seem
to be more indicative of the manpower necessary for long distance
hauling (i.e., 12-50 lbs./man). For this reason we will use the figure
of 50 lbs./man as the maximum for long distance continuous dragging of
heavy stone monuments.
If we accept the figure of 50 pounds per man, then 1,626 men
would be required to pull the sledge and stone; but this cannot be
accepted as a final figure for the totality of men involved in the land
transport of the monolith. There are other factors to consider, some
of which add to this number of men and some of which will reduce it.
A careful study of the Egyptian and Assyrian depictions of land transport
scenes offers some clues as to the complexity of the problem (see Plates
1-3).
11
Plate 1 is an Egyptian representation of the dragging of the
60-ton alabaster statue of Djehutihetep, from a 12th-Dynasty tomb
painting at El Bersheh (from Wilkinson 1842).
One hundred and seventy-two men, in four rows,
of fourty-three each, pull the ropes attached
to the front of the sledge; and a liquid,
probably grease, is poured from a vase, by a
person standing on the pedestal of the statue,
in order to facilitate its progress as it
slides over the ground; which was probably
covered with a bed of planks, though they are
not indicated in the painting. (Wilkinson
1842:325-326).
The lubricant question is especially interesting (cf. Takahashi
1937). There is little doubt that the Egyptians had knowledge of
sleepers, but none are shown (cf. Layard 1853:115). If rollers were
used, then a lubricant would not be necessary. Hence, Wilkinson's
interpretation would seem to be suited to the evidence at hand. He
goes on to describe the scene surrounding the dragging of the stone:
Some of the persons employed in this laborious
duty appear to be Egyptians, the others are
foreign slaves, who are clad in the costume of
their country; and behind are four rows of men,
who, though only twelve in number, may be
intended to represent the set which relieved
the others when fatigued.
Below are persons carrying vases of the liquid
or perhaps water, for use of the workmen, and
some implements connected with the transport of
the statue, followed by taskmasters with their
wands of office.  On the knee of the figure
stands a man who claps his hands, to measure
cadence of a song, to mark the time and ensure
their simultaneous draught; for it is evident
that, in order that the whole power might be
applied at the same instant, a sign of this kind
was necessary (ibid.).
Plate 2 is an Assyrian relief from Nineveh showing the King
supervising the transportation of a winged-bull statue. This particular
monument probably weighed about 30 tons. In addition to the men on the
ropes there is also shown the use of the lever, of sleepers, and the
presence of supervisory personnel. The sculpture itself is on its
side, roughly blocked out with blocks of wood placed beneath the statue
to keep its weight evenly distributed on the sledge.
Plate 3 is another scene in the same series from Nineveh. This
12
shows the statue now in the final stages of transport. It is held
upright by "beams, held together by cross bars and wedges" in addition
"to blocks of stone, or wood, piled up under the body." Note also the
"cables (which) appear to be of great length and thickness, and ropes
of various dimensions" (Layard 1853:113-114).
In these three scenes there is a wealth of evidence on the
many facets of ancient stone transport. In addition to the men necessary
to drag the stone there are men to work the levers, move the sleepers,
supervise the operation, provide replacement for fatigued and injured
workers, and the supplying of food. From this we can see that there
must also be men to maintain and replace broken rope and cable, as
well as manufacture and repair levers and sleepers. There must be
workers to grade the roadway so that it is as level as possible, and
there must be those who are continually supplying the lubricant and
applying it as the sledge progresses.
The transmission of power involved in these transport scenes
can be broken down into essentially three categories: ropes, reduction
of friction, and simple mechanical devices (Atkinson 1961).
Ropes of leather, animal hair, or vegetable fiber were probably
known in the Old World (Atkinson 1956; 1961). According to one informant,
the natives of Easter Island used to make "thick ropes from the tough
bark of the hau-hau tree" (Heyerdahl 1958:149). The Olmecs apparently
had rope, as shown on Altar No. 4 of La Venta and Monument 14 from San
Lorenzo (Stirling 1955). Plant fibers would have been available in
abundant quantities for use by the Olmec, and it strikes us as logical
to assume that the nulmerous vines and crawlers that literally tie the
rain forest together could have been braided into ropes of sufficient
strength for such a project.
Ropes were probably attached directly to the front parts of
the sledge and perhaps around the stone itself and then run out to the
towing crews. Though there is no evidence of it, it is possible that
towing bars were attached to the ends of the rope to afford the draggers
a good hold. However the representations of Assyrians and Egyptians
show men pulling at every point of the tow ropes. This last idea is
more fruitful, because more pull can be exerted per foot of rope by
men being stationed along the length of the rope than by a few men
hauling at a tow bar.
Reduction of friction would certainly reduce the number of men
necessary to pull the sledge, and it would seem that this would have
been a matter of prime concern to the ancient stone movers. Mulloy
(1970) notes that totora reeds or dry grass could have been used to
reduce friction in the transport of the Easter Island statutes. He
cites one source which recounts one tradition telling of a "paste of
taro and sweet potatoes" that was used to reduce friction (Metraux
1940, cited in Mulloy 1970:12). Speaking of the transport of the
Egyptian obelisks, Barber believed that "with wood, well lubricated with
oil, and operating upon fine sand, a reasonable traction was obtained"
13
(1900:91). Even wood on wood reduces the friction--for example, a
sledge running on sleepers as opposed to running directly over the
ground.
It seems probable that the Olmec did consider the question of
friction reducing agents, though we cannot know for certain whether
any were employed. Use of the wheel and rollers has been well
documented for the Old World, but there is nothing in the New World
to suggest that this development ever extended beyond use in toys or
games.
Any lubricant which would have reduced the number of men
pulling a heavily laden sledge would have been a great help. Even
running over wooden sleepers would reduce the number of men from 1,626
to 618. Wood on wood has a coefficient of friction of approximately
.38, though there is a range depending on the type of wood used. With
soap, the figure would be reduced to 244 men, and with tallow it drops
to 114 men. Even if these substances were known, difficulties would
probably have arisen in obtaining sufficient quantities to last for
several days, weeks, or months of hauling.
It is tempting to assume that mud, clay, or perhaps leaves
were used in this capacity with a subsequent reduction in the manpower
requirement to 450-550 men. If the assumption that some sort of friction
reducing device was used is valid, then it is possible to suggest
a range of from 500 to 1,000 men as being necessary to drag the sledge.
One thousand men will be used as a maximum figure here.
In addition to the possible use of sleepers, other simple
mechanical devices may have been known to the Olmec. These might
include the simple folcrum lever and perhaps the tourniquet (cf.
Atkinson 1961; Plate 1). It is difficult to gauge the effect of the
lever on the long distance transport of the Olmec stone monuments.
A lever might have been used in conjunction with each "heave" to ease
the burden on the dragging crew, and was probably employed in helping
to move the sledge up and down steep grades.
Another interesting question regarding dragging large stones
is the degree to which a regular cadence will increase the effective
pulling power of the workers. We know of no way to calculate the effect
of such an organized effort, though Barber (1900:94) points out that
"when hauling a weight...a 'one, two, three and a surge' will produce
a momentary force represented by nearly the weight of the whole mass
of men, or several times their ordinary pulling force." If this is
true, then a 150 pound man would have the potential to pull 150 pounds
of weight. The 40.6 ton sledge and monument would then have required
only 206 men to pull it without lubricant. However, it seems unlikely
that such a maximum effort could be exerted continuously over a period
of several hours.
We have from 500 to 1,000 men to drag the sledge, but more are
necessary for other related tasks. If the sledge were run on sleepers,
14
then there would be men required to constantly replace these if they
broke and to move them regularly to the front of the sledge. If the
sleepers were each 12 feet by 6 inches by 6 inches and made of
uahogan-y, each wo%A weigh96 %Ois.    11 oe were Leceseary fox every
foot of sledge runner, then 12 sleepers would be under the sledge at
a tine.  It seems reasonable to say that 12 o-r 13 additional slee-pers
were always on hand if some broke or splintered. In addition to the
pulling crew, about 25 men would be required for the functioning of
the sleeper system. (The sleepers under the sledge do not require
men, therefore a total of 25 sleepers requires 25 men).
Men would also be required to level the road in front of the
sledge, to be constantly providing mud, clay, or whatever lubricant
was used and to be gathering and braiding ropes to replace frayed or
broken ones. All totaled it seems that these tasks could easily have
required an additional 100 men, and probably at least 200. Add to this
a dozen or so supervisory personnel and the entire procession would
consist of between 700 and 1,200 men. If we assume that the workers
actually pulling the sledge would have been pe riodically relieved by
a member of the road gang or the sleeper crew, then no more men would
be needed. If however, a separate crew of perhaps 50 men were kept
rested and then periodically rotated into the hauling schedule, then
we would have between 750 and 1,250 men. This would then be the total
number of men actually accompanying the sledge along its overland
journey, but many more persons would be indirectly involved in the land
transport (those working in food production). This will be discussed
later.
The rate of progress of this procession was probably fairly
slow. We have only one example of dragging a monument on a sledge
which gives some idea as to the time involved. Kida (1912:5) in
describing the transport of a 159-ton Japanese megalith using a
sleigh and iron chains, notes that it took seven days to travel
slightly less than two miles (3,318 yards). If we convert to metric,
this works out to 431 meters/day.
Here we must digress momentarily and consider the probable
length of the work day. Erasmus (1955:330), in his study of Mayo
work patterns determined that the "total working time of...males...
was between eight and nine hours a day, of which between six and one-
half and seven hours were dedicated to what we have called 'economic'
pursuits." It should be noted that this study took place in the
summer time when the days are longer. If the Olmec moved stone
monuments during the dry season the days would not have been quite as
long. If the total work time was reduced one hour due to the length
of the winter day, this would leave us with a seven to eight hour day.
Assuming a maximum eight hour day, we must make adjustments for delays
that no doubt would have occurred regularly--fatigued workers,
broken ropes, readjusting the position of the monument on the sledge,
etc. A more accurate estimate of the amount of time actually spent
in dragging might be closer to 6 hours per day.
15
There is great danger here of anthropocentrically imposing our
own 8-hour day on the Olmec; just as it must be realized that the Mayo
observed by Erasmus may have already been victims of such Western work
schedules. Had the Olmec been highly motivated in their stone working
endeavors, it is not inconceivable that they may have worked for ten
or twelve hours a day, or at least as long as there was daylight. In
light of this it seems best to suggest a range of between 6 and 12 hours
of work per day. Using the lowest figure (which gives the maximum
travel time involved) in conjunction with Kida's data, the rate of
transport works out to 71.8 meters/hour.
The distance from the Ce'rro Cintepec to the Coatzacoalcos
River is 58 kilometers. With our rate of 431 meters per day, the
journey would have taken 13.5 days. This does not include the time
necessary for loading the sledge--only the actual time on the road.
Water Transport Mechanism.   Clinton Edwards (1965) in his
exhaustive investigation of aboriginal watercraft on the Pacific coast
of South America documents the use of the following in aquatic trans-
portation: reed bundle floats, hide floats, sewn bark canoes, dugouts,
gourd rafts, and log rafts. As tantalizing as they may seem we will
discard the potential use of floats of reeds and hides, bark canoes,
or gourd rafts from our study. There is no ethnographic evidence that
these mechanisms were employed by peoples in the Veracruz-Tabasco area
anciently or recently, and the presence of the necessary materials such
as Scirpus totora for the bundle rafts, sealskin for the hide floats,
and the large Lagenaria gourds are lacking in the natural environment.
Even if the natural resources for construction were on hand, as possibly
in the case of bark canoes, the feeble little crafts were not sturdy
enough to carry more than a couple men, much less several tons of
basalt. Canoes and logs bound for rafting, however, are potential
means of heavy transport on the water and will be considered each in
its turn.
Log Rafts.
For the sea journey the raft has some marked advantages
over the boat, in that it is unsinkable and cannot be
swamped. On the other hand a raft to support a given
weight is very much larger and heavier than a boat, or
composite of several boats lashed together, to carry
the same burden, and is therefore more maneuverable in
an emergency. Moreover, while it is very doubtful if
they would be practical for the inland part of the
journey (Atkinson 1956:111).
Both forms of water transport must be considered as a method
employed by the Olmec to move their large stone monuments. It is
possible that both rafts and canoes with support structures were used,
one for the sea part of the voyage and one for the river transport; but
the effort involved in transferring the monument from one to the other
makes it seem improbable.
16
We assume that for a raft a light wood which was easily available
to the Olmec would have been selected. Ceiba saurauma is a very light
wood found today throughout the tropical rain forests of the Olmec
heartland. The introduction of cattle has greatly increased the amount
of grazing lands, but as a result the forest lands are much smaller
today than they were in prehistoric times. Ceiba was probably much
more common then than it is today. This tree grows up to heights of
150 feet and can be found up to ten feet in diameter. It has a
specific gravity of 0.089 (dry) and weighs 23 pounds per cubic foot.
With water weighing 60 pounds per cubic foot, one cubic foot of Ceiba
will support a maximum of 37 pounds without sinking below the waterline.
Using Altar 1 from La Venta as a sample, we can make a
theoretical model of a raft with varying dimensions. Assuming that
Altar 1 weighed 38 tons before being carved into final form, the Olmecs
would have required a minimum of 2,054 cubic feet of ceiba to support
the stone. 2,000 cubic feet of Ceiba can take many forms from an
almost square 25 ft. X 27 ft. X 3 ft. to an elongated 50 ft. X 8-1/4 ft.
X 5 ft.  The possibilities are limited only by the size of trees
available.
The raft would have probably been made of two layers of Ceiba
lashed at right angles to each other. For lashings numerous vines and
crawlers from the forest were always available, although other materials
may have been used for rope. The logs may have been slightly notched
so that they rested in place more securely.
A raft with a five foot draft might be too much for some of
the shallower parts of the slow moving Gulf Coast rivers. Perhaps
three feet would be more reasonable. The raft would actually have
to be larger than 2,054 cubic feet for several reasons: (1) to
allow for an increase in weight of some men on board to attend to the
lashings and help with steering (and perhaps even paddling); (2) to
allow for an increase in weight due to prolonged exposure to the
water; and (3) to allow for some freeboard above the waterline. With
these considerations in mind, we can construct a model raft.
Starting with a raft 35 feet long by 20 feet wide by 3 feet
deep, we see that it contains 2,100 cubic feet of Ceiba, enough to
support 77,700 pounds. With the addition of the stone, the sledge,
and ten men, this raft will sink below the surface of the water. To
support this weight and allow a one foot clearance above the waterline,
a raft 35 feet long, 20 feet wide and 4.2 feet deep would be required.
A total weight of the raft, plus the monument, plus the sledge,
sleepers and ten men would be 75.3 tons (see Appendix C for calcula-
tions).
This is the largest raft that would have to be built for any
single monument, and if individual rafts were built for each stone,
then the smaller monuments would require proportionately smaller
craft. It is possible that a large raft such as the one described
17
was constructed and then reused for other monuments; but again,
considering the effort that would be involved in returning the raft
to the point of debarkation, drying it out, replacing worn lashings,
etc., this also seems rather unlikely.  Perhaps such a raft--if ever
built at all-was used only for the very largest monuments, a canoe
structure being more suitable for the smaller ones.
Canoe Rafts. The most feasible alternative to a log raft
is the dugout canoe. The use of the dugout seems almost universal.
Atkinson (1956) mentions the use of 35-55 foot oak dugouts in his
reconstruction of the transport of the bluestones from their source
to Stonehenge. Joseph Ames, a shipping foreman with experience in
Micronesia, observed a 60 foot dugout war canoe rowed by 40 men with
a carrying capacity of 5,000 pounds.of cargo in the Fijis (personal
communication). Pizarro noted the use of three large dugouts with
sixty paddlers in the Gulf of Darien (Edwards 1965). Columbus
encountered a large Mayan seagoing dugout used for trade near the
Bay Islands. He described it to be the length of a galley, eight
feet wide (beam), and manned by 40 men. Thompson (1954) remarks that
canoes are depicted in murals dating from 1150 A.D. in Chichen Itza,
and adds that the Chontal Maya used forty-man canoes. At Tikal mytho-
logical beings ride in a large canoe incised in bone.
As anciently, the dugout canoe has continued to be used in
river basins of the Coatzacoalcos and the Tonala, though they are now
more frequently powered by outboard motors than by oars. Bernal
believes that one jade carving from Cerro de la Mesas is a representa-
tion of an Olmec canoe (1969:plate 68a).
Drucker reports that most dugout canoes made today are
fashioned from either Ceiba or mahogany. Though Ceiba is convenient
the wood is not particularly strong. It cracks and weathers rapidly
unless treated with resin as a preservative and sealer. Mahogany
(Swietania macrophylla) is much preferred because it is strong, durable,
hard (yet carves well), and it is decay resistant.
Atkinson proposed the joining of several canoes side by side
to carry stones to Stonehenge from distant quarries, and even conducted
a replicative experiment by building a raft consisting of three canoes
(1960). The stone rested on the pole superstructure which also served
to link the elm canoes.
This idea also has the support of Heizer who suggests that
the canoe structure is the only logical alternative not only in terms
of manuverability but also in ease of construction (Heizer in Bernal
1969:52).
Drucker feels that such a canoe "barge" is unfeasible because
of decreased manuverability and the time-consuming nature of canoe
carving (personal communication).
The time and labor-intensive nature of canoe construction may
not have been a problem. Instead of engaging in a major canoe
18
construction project in conjunction with large stone transport,
Heizer has suggested that the canoes used in everyday life were also
used in heavy transport (personal communication). Workers would travel
in their canoes to the staging area where the dragged'stone would
eventually be brought. The raft of canoes could then have been
assembled on the spot.  At journey's end, the super-structure assembly
would be removed and the workers would return to their homes in their
individual canoes.
In our reconstruction of a canoe-raft, the assumption is made
that mahogany (Swietenia macrophylla) was used. S. macrophylla has
been used extensively for canoes throughout the riverine communities
of Veracruz and Tabasco in the recent past. It is admirably suited
to canoe construction because of its hardness, strength, shock resis-
tance and relatively high dimensional stability (Lamb 1966). The tree
itself is usually over 100 ft. in height with a straight cylindrical
bole before branching of 40 - 60 ft. (ibid.). Canoes carved from a
mahogany log could have been 50 ft. long.
Canoes observed in Veracruz and Tabasco today exhibit a wide
range of sizes. Heizer (personal communication) had an opportunity
to measure four canoes at Villahermosa ranging in size from 42 ft. in
length to 18 ft.
Length         Interior Depth        Beam
42'               24"                45"
30'               22"                29"
26'               24"                46"
18'               12"                20"
A canoe measured by Drucker was 26-1/2 ft. long, 1-1/2 ft.
deep, and 2-1/2 ft. wide.  It had a carrying capacity of 3/4 ton with
three inches of freeboard (personal communication).
Though big canoes 50 feet or more in length could be made
today, Drucker (ibid..) explains that shorter canoes are more manageable;
the larger forms may not have been constructed except for special pur-
poses. Also, the availability of mahogany has been reduced in coastal
areas due to intensive logging operations. Large trees needed for
larger canoes are becoming increasingly unavailable.
If canoes similar to that noted by Drucker were joined in a
raft, more than 51 of them would be needed to carry the monument alone.
Doubling the length and width of a canoe roughly quadruples its
capacity, so a dugout 50 ft. long with a 5 ft. beam would probably
have a capacity of about 3 tons. Fourteen canoes of these dimensions
would be needed to carry a 38-ton monument (see Fig. 3).
A mahogany superstructure to join the canoes and bear the
monument would weigh about two tons. The weight of as many as 120 men
19
for rowing (and perhaps bailing) would add an additional 8.4 tons
to the load. The monument still rests on its sledge and sleepers,
another 3.2 tons. These additional items could add as much as 13.6
tons to the total load, or 52 tons in all (see Appendix D).
In these calculations it has been assumed that the depth
of the canoe has been a constant variable, but in all likelihood as
the dugouts were made longer and wider they would have been made
deeper. We are not certain how much deeper they may have been, but
if the depth was doubled the capacity of the canoe would have been
more than doubled. Perhaps for a 50 foot dugout the depth would be
2-1/2 feet. A combination of 14 canoes 50 ft. by 5 ft. by 2-1/2 ft.
deep could easily carry Altar 1 and all its attendant men and
equipment. This model approximates the minimum number of dugouts
needed and a possible design of the canoe raft. Smaller monuments
would have required fewer canoes and less men, but we are interested
here in the greatest numbers of men and equipment needed.
To construct a canoe-raft, the individual canoes would have met
at the termination of the land transport route somewhere along the shores
of the Coatzacoalcos River and there been fitted together to form a raft.
The sledge carrying the monument may then have been pulled and
pushed across some sort of ramp onto the canoe-raft and lashed into
place. The details of the loading operation can not be known, but it
is of interest to note Barber's description of Egyptian obelisk
loading. The embarkation was probably effected as follows:
A dry dock was dug out at a short distance
from the river bank at Assouan, in a position at
right angles across the road along which the obelisks
were to be dragged from the quarry. In this dock
the lighter was built, and was afterward floated
so that its deck was the exact height of the road-
way. The obelisk was then hauled from the quarry
and turned half around just before reaching the
lighter, and launching skids were led to the deck.
In this position forty drag ropes could be made
fast to the obelisk and led over the ship's deck
to the roadway beyond, where one-hundred fourty
men could be harnessed to each, and the obelisk
dragged on board. The dike separating the dock
from the Nile was then entirely cut away and the
lighter floated into the river. When the boat
reached Thebes the operation was reversed.
(Barber 1900:94).
It is questionable that this complex procedure was used by the
Olmec. Possibly the raft was loaded while stranded an low tide or while
the river was lowest during the dry season in anticipation of the rising
tide or the swelling of the river during the rainy season. It is not
known whether our theoretical embarkation point was effected by tides,
20
nor can we say that the rainy season was best for water transport. Coe,
et.al. (1967) consider riverine transport to San Lorenzo as being
possible only during the rainy season, but it must be noted that rain-
swelled rivers are particularly hazardous to negociate because of float-
ing debris.
The task of moving the canoe-raft on the river was probably
accomplished by a number of towing canoes which were connected to the
raft by towlines of fiber or sinew rope. The dugouts intended for
towing might have been brought from surrounding areas as the large
50 foot canoes were. They were probably not of a certain specified
dimension and were not made specially for use in monument transport
but were the dugouts used in daily life conscripted for the project.
These canoes probably varied in size and therefore in the number of
paddlers in each one. The idea of towing is utilized in ancient
Egypt to tow obelisk barges. Solver (1940) discusses three lines of
towing crafts--the center being chiefly concerned with pulling and the
outer two had the primary objective of steering the barge and keeping
it midstream. This was surely the case with the Olmecs; some of the
towing canoes were for drawing the raft through the water while others
were intended to conduct the raft past obstacles and through bends in
the river. It is not known if a rudder was used. It is likely that
men on board the raft could help propel it by paddling from the
canoes on the perimeter. The idea of towing from the shore may be
rejected as the margins of the river are either too steep or swampy
and densely overgrown in either case. We assume that the same mechanism
and manner of propulsion and steering was used in marine as well as
riverine navigation.
Rate of Progress in Aquatic Transport. Twelve hundred men
might have been able to drag the sledge carrying Altar 1 from the
highlands to the Coatzacoalcos River in 14 days, using the Japanese
monolith moving data as a basis for calculations. Meanwhile a smaller
crew of perhaps some 100 workers could have been assembling the canoe-
raft.  It is possible that the men involved with dragging also
assembled the raft upon arrival. One way or the other, probably not
more than one or two days were spent in construction. Another day would
have been required to load the monument with its sledge onto the raft
and to secure it for the voyage.
It has been our assumption that the laborers involved in the
whole project of developing a ceremonial center such as La Venta
or San Lorenzo, importing stone from the Tuxtlas and other sources,
as well as collecting of other materials were furnished by the
population living in the support area adj acent to that center. Thus
quarrying, dragging, towing, carving, and supervising intended for
the edification of La Venta would have been done by the men who used
that center, or those who were benefited by it. The same is true
for San Lorenzo.
Most of the 1,200 men involved in dragging Altar 1 to the
Coatzacoalcos River would have then returned to the support area from
21
whence they came. Some might have remained to aid in the water
transport, but the manpower requirements of this phase would have been
much less than those of land transport.
There could have been as many as 120 persons on board the
canoe-raft itself (the maximum number of individuals that could fit
comfortably in rowing or paddling positions). This leaves over
1,000 additional workers.  Some of these undoubtedly continued on in
the towing canoes, but many were no longer needed and must have
returned to their homes.
Heizer and Drucker (personal communications) believe that
the Coatzacoalcos river flows at a rate of 1-2 miles per hour. The
La Venta water transport route (Fig. 1, route 5) involves 32 km. of
travel down the Coatzacoalcos with the current, and 49 km. against
the current along the coast and up the Tonala. The canoe structure
would require 72 men to move it at a rate of 1.5 knots while bucking
a current of 1.5 mph. (4) Since there is ample room for this number of
men on the canoe structure itself, there would be little need for towing
canoes other than for steering purposes. The canoe structure would
require only directional control while moving down the Coatzacoalcos
because the current would move it along at the desired rate. At the
rate of 1.5 knots (which is approximately--for our purposes--1.5 mph.,
1 knot equals 1 nautical mile per hour, which is slightly more than
one land mile) the canoe-raft with its load would reach La Venta in
about six days (still assuming a six hour work day). Going up the
river in San Lorenzo the manpower requirements are the same, but the
time necessary would be only 4 days. The total manpower requirements
for the Cerro Cintepec to La Venta transport of Altar 1 would be
110,592 man-hours. The San Lorenzo route would involve a little less,
something on the order of 107,328 man-hours. Should additional
canoes have been used for steering and extra towing, we might add.
120 man-hours per canoe (20 men per canoe at 6 hours each) per day.
Ten canoes for 6 days of water transport would increase our La Venta
figure to 117,792 man-hours. So as not to underestimate, we will use
this as our figure for the Cerro Cintepec to La Venta route, and 114,528
man-hours for the trip to San Lorenzo.
There is always a danger that the ocean part of the journey
might have been plagued by rough seas, nortes, or other unexpected
hazards, but this would not greatly effect our overall calculation.
For every additional day on the water, only 432 man-hours would be
added for those on the raft. If 10 towing canoes were being used,
the total daily figure would be 1,632 man-hours.
Conclusion. We have tried to make a reconstruction of the
most probable means of transport of a large stone monument in ancient
Olmec times. We have tried to incorporate into our calculations a
maximum of ethnographic and archaeological data and a minimum of
speculative reconstructions. If our model is a valid one in light
of what we know of Olmec organization, we might ask what more can
we infer about the Olmec.
22
It appears that something on the order of 117,792 man-hours
would have been required to transport the single largest stone
monument from the quarry to the site. This seems to be divided
roughly into 1,200 men working for 14 days and 472 men working for
6 days. These figures represent the maximum amount of time and
labor involved. We assume that the Olmec workers were normally
engaged in agricultural pursuits, as there is no evidence to suggest
a permanent working laborer class.(5)
There is the question of whether there was sufficient time
available for these farmers to work on something besides their fields.
The requirements of slash-and-burn agriculture are such that the
milpero may find himself with up to 100 continuous days of free time
in a year (Heizer 1961). This 100 days follows the harvesting of
the old milpa and the cutting of the new one, but preceeds the burning
of the fields. It coincides with the dry season, from late January
to mid-May. Using La Venta as an example, we can see that there was
apparently a large agricultural population occupying the lowland area
between the Tonala and Coatzacoalcos Rivers. Heizer (1960:219;1961;
1962) has calculated that the area probably contained about 18,000
individuals. Using the figure of 4.5 persons per average family
(Cook and Borah 1960) we see that there were probably 4,000 family
heads available for this work. To move a stone such as we have
described, the manpower and time were apparently available; but how
does this fit in with the total scheme of Olmec construction projects
and the overall time involved?
Heizer (1968:23) has postulated that the origins of the Olmec
culture might have come from a mutually beneficial arrangement between
the local farmers of the southern Veracruz-Tabasco lowland and an
organization of ritualists. He suggests that ritualists would provide
religious information relating to the proper time to burn the fields,
plant the crops, etc.(6) While the farmers would contribute to the
support of the administrative priesthood by supplying food and goods
as well as labor and men to be trained as craft specialists.
From this it is possible that a "highly innovative" and
perhaps "dynastic" group of small numbers arose and directed the
energies of the farmers towards the construction of the sites and the
transportation of the stone monuments (ibid.), perhaps as a token of
devotion and in remuneration for the ritualists religious-calendric
advice.
Such a social structure would probably have sufficient control
to engage in periodic massive public work projects, but not enough
control to require this type of service regularly.
The Olmec farmer, in return for the religious ceremonies and
blessings, could have repaid the hierarchial class regularly with food
and perhaps small amounts of labor at the sites; but at the time of a
great religious event--the end of a cycle, for instance--the importance
23
of starting and continuing large religious projects, along with
encouragement from the hierarchy, could have been enough to motivate
the farmers to perform such tasks.  These endeavors would, of course,
be in their own self-interest as well because they wanted to guarantee
the continued support and advice of the priests.
Such a ruling class would be able to plan in advance the major
undertakings that would be required at each major religious ceremony,
and could thus inform the agriculturalists well in advance that some
of them would have to work at the La Venta site, or that some would
be required to quarry limestone at Chinameca for a massive offering.
The mechanical aspects of this part of the social organization really
cannot ever be known. Some sort of voluntary community service is
expected of the farmers in many modern communities in the area, and
this type of community obligation in conjunction with a corvee labor
operation perhaps similar to the Inca 'mit' might have facilitated
the procurement of the necessary labor force.
Heizer (1961) has estimated a total of 2,000,000 man-days
as being necessary for the total construction of the site of La Venta.
This estimate was made before the exploration of the Stirling
Acropolis (Heizer, Graham and Napton 1968), and we feel that it would
not seem excessive to add an additional 1,000,000 or more man-hours
for the construction of this complex alone. (It should be kept in
mind that additional mounds may yet lie undiscovered, so additional
manpower may have been required). Of course, if all family heads
worked for 100 days for several years in succession, they could
have built the entire site in 7 or 8 years! This is totally
irrelevant, for we know that the site was constructed in four major
efforts spread out over 400 years (Drucker, Heizer and Squier 1959),
but it helps to emphasize the point that it is not what could have
been done, but what is normally done that is important to know
(Sanders and Price 1968:55).
The Olmec "invention" of the earliest calendric system in
the New World has been suggested (Coe 1957).   As it appears to be the
forerunner of the better-known Maya calendric system with its 52-year
cycle, Heizer (1960:218+; 1961:47) has argued for the four major
construction periods at La Venta being spaced at 104 year intervals
throughout the 400-year occupation of the site. If the La Venta
fluted pyramid required 800,000 man-days of labor to build, this
could have been accomplished in eight work periods--one every 52
years. This works out to 1,000 men working 100 days every 52 years
(Heizer 1960:220). If the general mound building at the site
required 300,000 man-days, this could have been separated into four
100-day work periods, one every 104 years, involving 750 men each
time (ibid.). If we add to this the 1,000,000 man-days for the
Stirling Acropolis, this would require the addition of 2,500 more men
every 104 years. This brings the total manpower requirement for
every 104-year interval to a bit more than the maximum capacity of
the support population. Of course, there is nothing to prevent work
24
in "off" years, and perhaps every 52 years the entirety of the work was
evenly divided instead of having the emphasis placed on the 104-year
interval. If all of these constructions were being attempted at once,
we would have 4,250 men at work during one season. It seems more
probable then, that the work was somewhat spread out, with peak
production periods probably correlated with the cycles of the relig-
ious calendar.  If 118,000 man-hours were required to move the largest
stone monuments, it would seem that this undertaking, too, would have
been planned for a propitious religious moment. Stirling (1955) has
suggested that the massive heads represent chieftans and were carved
perhaps only once a generation (or once every 52 years?), which might
coincide with the calendric cycle.
As nearly as can be calculated, the total weight of all La
Venta stone monuments appears to be approximately 325 tons (see
Appendix A). At 118,000 man-hours per 38 tons, all of the Olmec
sculpture brought to the site would have required c. 1,015,000 man-
hours in transportation time alone.   If all monuments were transported
at such cyclical intervals, this would be another factor which would
overload the 4,000 men per year capacity. It seems likely that smaller
stones could have been transported at almost any time; but that only
the largest monuments were commissioned for a 52-year cycle ending.
It would seem that the La Venta Olmec support population was, at
least every 52 years, working at or near its maximum capacity.
We might ask here what they did the remaining 51 years when
there was little work going on at the site.  There was apparently a
large amount of "free" time available, but as Kaplan points out with
reference to the Maya culture:
To show that the Maya only had to farm two or
three months of the year and had plenty of spare
time for community development, given social
differentiation and increasing power at the top,
presents a picture of bored aboriginals wandering
aimlessly through the brush in search of a power
structure to put them to work (Kaplan 1965:280).
Thus, although there was ample free time available, it is
not enough to just say that the farmers got together and decided to
build a ceremonial center. The underlying forces that motivated the
work force at the end of each 52-year period to engage in such major
undertakings did not suddenly appear at these intervals and then
disappear. They were always present, though perhaps not in sufficient
quantity or to a sufficient degree except when they worked in
conjuaction with the religious-calendric cycle.
This is not to say that this is the way it was, but when
dealing with a group of individuals with the capacity for 160 million
man-hours of work over 400 years, there must be some reason why only
a fraction of this capacity was ever used. Because any 400-year
effort requires continued control over the factors involved, the
25
ritual leadership probably planned the entire venture in advance. If
the degree of control were such that the dynastic rulers couldn't
demand a certain output except on special occasions, this might
explain how such a great work capacity existed without being fully
utilized. A point may have been reached:
beyond which the common people refused to go
to meet what they considered intolerably
heavy demands for their services, goods, and
time.  A bare suggestion of this in the La
Venta site is the occurrance in its final
phase, or large tombs and sarcophagi which
may be interpreted as the material expression
of the ultimate development of class differences
in the form of burials of high priests within
the ceremonial precinct (Heizer 1961:54).
It would be interesting to speculate as to the manpower
requirements of the San Lorenzo site, but unfortunately no support
population estimates have been published. Using our figure of
118,000 man-hours per 38 tons; plus the estimate of the weight of
all San Lorenzo monuments as being between 200 and 250 tons, we
can suggest that a labor force of at least this capacity must have
been available. Coe (1968) describes a survey of a 75 sq. kilometer
region surrounding San Lorenzo in which he says not more than 2,500
individuals could have supported themselves, but beyond this we have
no information regarding whatever limitations--geographical or other-
wise--that there might have been on the San Lorenzo support area.
We would like to assume a similar type of ruling class and work
pattern, but until we have some concrete evidence as to what went on
at San Lorenzo, we will have to continue to work with the La Venta
materials for much of our inferences.
Returning to the problem of transporting the stone monuments,
it appears that the only limiting factor (in terms of the quantity
of stone transported) was the availability of labor. Stone was always
available in abundant quantities, and apparently there was enough time
to move even the largest monument in one dry season. The hierarchial
social structure did not inhibit the transport of smaller stones, and
was a valuable support mechanism in the transport of the larger ones
because of its value in planning ahead and controling large numbers
of people. The only limitation we can see was in the availability
of labor during peak periods. If we take Heizer's estimate regarding
the population of the La Venta support area as correct, there were
only 4,000 individuals available at any one time. If all the work
was actually done at 52-year intervals, then something on the order
of 4,500 workers, or a total population of 20,250 individuals was
required.
Our reconstruction of Olmec heavy stone transport is in many
ways highly speculative. In applying historical and ethnographic
examples and the results of modern-day replicative experiments to the
26
Olmec we have tried to use caution, but the inherent uncertainties
of analagous reasoning leave many unanswered questions.
It has been said that we know now perhaps 90% of all that
we will ever know about the Olmec. The La Venta site alone has been
almost completely destroyed by the expansion of the town of La Venta
and the nearby oil refinery.
Though much of the material presented here can neither be
confirmed nor denied, the uncertain future of Olmec studies may
soon pressure many scholars into similar methods of analysis.
We hope that with carefully formulated models students of
archaeology and history will continue to investigate the mysteries
of ancient heavy transport as one tool in the reconstruction of
ancient civilizations.
27
NOTES
(1)  The number and distribution of such large stones in sculpture
and architecture throughout the world is quite large and diverse.
A description of each monument, its location and condition,
would not be properly discussed in a paper such as this. There
is abundant information concerning almost all phases of large
stone transport which will be referred to when applicable. For
a general discussion of the evidence for ancient heavy transport
see Heizer (1966).
(2) For more information on climate see West, et.al. 1969; Poleman
1964; Vivo Escoto 1965.
(3) Just as Monument 19 at La Venta shows how the Olmec artisan
adapted to the medium with which he worked, it is interesting to
speculate that the shape of some of the colossal heads might
have been predetermined by the configuration of the boulder as
found at its source. A similar observation has been made by
Hawkes and Woolley (1963) for the Old World:
For important sculptures the Sumerians imported
diorite (and sometimes trachyte), a hard stone
capable of taking a fine polish... It would
appear that this was not quarried but came in
the form of boulders, and the size and shape of
such could not but influence the sculptor's work.
... it is curious how frequently a seated Sumerian
statue suggests the shape of the natural boulder
from which it was economically carved (Hawkes
and Woolley 1963:775).
(4) Our special thanks to Mr. Charles L. Wickers, Jr., Chief
Engineer of the Port of San Francisco, for these calculations.
Our thanks also to Mr. Mark Rasmussen for his work on the maps
(Figures 1 and 5).
(5) Were there evidence for such a permanent working class, the
question arises as to whether a large non-producing population
could be supported by the Olmec agriculturalists. The La Venta
support area contains about 900 sq. kilometers (Heizer 1961;
Drucker 1961). With a population density of 20 persons per
square kilometer (Sanders 1953), this comes out to 18,000
individuals. Slash-and-burn agriculture demands that approx-
imately 5/6ths of the land be in fallow at any one time. Still,
the potential land under cultivation in any one year would be
1/6th of 900, or 166 sq. kilometers. If a family of 5 can live
on 1.5 hectares under cultivation (Drucker and Heizer 1960),
then this area has the potential to support well over 45,000
individuals...well beyond the 18,000 we get using Sander's
population density figure. An additional fraction of a hectare
28
per farmer would provide the surplus necessary to feed 4,000-
5,000 additional individuals. Thus, if the need arose, a large
non-agricultural class could be supported by the primary
producers (see also Heizer 1960;1962).
If the workers used for transport were normally farmers, they
would have already produced enough food to feed themselves
during the course of their normal agricultural year. The
problem would be in transporting the food to the place of
work. Some could have been brought by the workers who went
to the work area by canoe, but overland transport of as
much as 75 kg per man would have been difficult.   (753 g/day
X 100 days = 75.3 kilos. See also Sanders 1953). This, then,
would have been a factor taken into consideration by the
priestly class when the work was commissioned, and perhaps the
agricultural producers along the transport route were told to
plant more corm the previous year to feed the additional people
who would be passing through their area in the next dry season.
(6) Ritual plays an important role in the agricultural practices of
modem inhabitants of the southern Veracruz-Tabasco area today,
and may have its origins in these earliest ceremonies (cf.
Foster 1942).
29
Appendix A: HEAVY BASALT MONUMENTS FOUND AT LA VENTA AND SAN LORENZO
La Venta Monuments                         Weight in short tons
Colossal Head #1                                    24.0
*Colossal Head #2                                    11.8
Colossal Head #3                                    12.3
Colossal Head #4                                    19.8
Stela #1                                             5.5
Stela #2                                            10.5
Stela #3                                            25.5
*Altar #1                                            36.5
Altar #2                                             5.5
*Altar #3                                            13.7
*Altar #4                                            33.7
*Altar #5                                            18.6
*Altar #6                                             2.7
Altar #7                                             4.3
Monument #8                                          9.9
Monument #68                                         8.3
242.6
Note: A rough calculation of the total weight of all the Olmec
monuments known from La Venta based on dimensions given in
Escultura Monumental Olmeca (de la Fuente 1973) is 325 tons.
Not all of the monument dimensions are given, so calculations
of their weights is not possible. For the sake of calculation,
all monuments were assumed to be made of basalt with a weight
of 180 pounds per cubic foot (3.183 short tons per cubic meter).
This is only a rough estimate, but the range of 300 to 350 tons
is probably fairly accurate.
San Lorenzo Monuments                     Weighttin short tons
*Colossal Head #1                                    25.3
*Colossal Head #2                                    11.8
*Colossal Head #3                                     9.4
*Colossal Head #4                                     6.0
*Colossal Head #5                                     9.9
Colossal Head #6 (Mon. 17)                           9.0 (8-10)
Monument #14                                        35.0 (30-40)
Monument #20                                        16.3
Monument #51                                        12.3
Monument #60                                         6.2
141.2
Note: The total weight of all the San Lorenzo monuments listed in
Escultura Monumental Olmeca was 181 tons.
30
Appendix A (continued):
An unknown percentage of the Olmec monuments from this site may yet
be undiscovered so an arbitrary addition of 40 or 50 tons would not
seem unreasonable. Thus the range for San Lorenzo is probably between
200 and 250 tons. The possibility of many more as yet undiscovered
monuments does exist (Coe 1968:55).
* monuments made of basalt from Cerro Cintepec.
Some of the weights above were calculated by the authors using the
dimensions given in the following sources:
Clewlow et. al. 1967
Clewlow 1974
de la Fuente 1973
Williams and Heizer 1965
31
Appendix B: SLEDGE MECHANISM (see Fig. 2-3).
Rurmers: 2 runners for smaller monuments
3 runners for heavier monuments (LV Altar 1)
12' X 1-1/2' X 1-1/2'            3
therefore each runner has 27 ft. , and all 3 runners have
81 ft.3.                     3
if mahogany weighs 32 lb./ft. , then weight of 3 runners
is 2,592 lbs.
Crosspieces:  6 crosspieces (1 for every two feet runner)
11' X 1' X 1'                  3
therefore each beam has 11 ft. , and all 6 beams have
66 ft.3, and weigh 32 X 66 = 2,112 lbs.
Bracing blocks: 2 bracing blocks
9' X  'xl                     3                            3
therefore each block has 9 ft.  and all blocks have 18 ft.3
and weigh 32 X 18 = 576 lbs.
Complete sledge weighs 5,280 lbs. or 2.6 tons
Sleepers: 25 sleepers (12 always under sledge)
12' X 1/2' X 1/2'
therefore each sleeper has 3 ft.3 and weighs 96 lb. each.
Weight of complete sledge plus LV Altar 1 = 2.6 + 38 = 40.6
32
Appendix C: RAFT CONSTRUCTION
Weight of unfinished LV Altar 1
Weight of sledge and sleepers
Weight of 10 or 12 men
Total weight that need be supported
76,000 lb.
5,280 lb.
1,700 lb.
82,980 lbs.
1 ft.3 of dry Ceiba can support 37 lbs.
82,980/37 = ft.3 needed to support sledge, etc. = 2,242.7 ft.3
Volume of raft + 700 ft.3 for freeboard = total vol. of raft
2,242.7 + 700 = 2,942.7 ft.3 of ceiba
2,942.7/(35' X 20') = 4.2 feet of draft
2,942.7 X 23 (weight of ft.3 of ceiba) = 67,682.2 lbs.
Total wt. of raft
Final
Final
Final
To tal
dimensions of raft: 35' X 20' X 4.2'
weight of load:    41.5 tons (82,980 lbs .)
weight of raft:    33.8 tons (67,682 lbs.)
weight             75.3 tons (150,662 lbs.)
The raft will support 41.5 tons with a 3 ft. draft and 1.2 ft. of
freeboard.
0          3          6
Feet
Figure 2: Egyptian Sledge (Solver 1940)
0          3         6
Feet
Figure 3:   Hypothetical Sledge for Olmec Transport
33
Appendix D: CANOE-RAFT CONSTRUCTION (see Fig. 4).
Weight of LV Altar 1: 38 tons (unfinished)
Dimensions of Drucker's Canoe: 26-1/2' X 2-1/2' X 1-1/2'
Capacity of Drucker's Canoe: 3/4 ton
Number of canoes the same as Drucker's to carry LV Altar 1 = 51
Possible canoe length: 40-60 ft.
Capacity of 50 ft. canoe (with beam of 5 ft.) = 2 X 2 X 3/4 = 3 tons
Capacity of Canoe-Raft made of 14 50' canoes = 3 X 14 = 52 tons
(The capacity is greater because canoe depth increased).
Superstructure: all poles are 6" X 6" thick and are 50', 20', or
30' long (each pole may not be that long; but poles are lashed
to span the raft.)
total weight of superstructure: 3,750 lbs. or 1.9 = 2 tons
Total length of full raft: 150'
Total width of raft: 30'
Depth: 2-2-1/2'
Anticipated freeboard loaded with stone and men (120) = 6+"
to
0
c]-
4H
0
0
CH
Xi                            1
01)                          0)
'ii                          0
Pr4                           A
cvi
H
c.)
a)
0
0
0
0
E
0  i
0
a  i
_  .  I
4-
a)V0
0:0
V~~~~c
I .I
. IK
I       I
X - cm It)
0-
0
0
0
E
0
4-O
0
0
.ci
aJ
I0 *
00I-e  ..-
000  . . .I      0   000
00' , .    I &-0U1
w
C-)
w
a.
0
C)
LL
0
m
-
0
0.
.2
0,
. LO
0
-j--
0
0
0
C'0,
W
0)
4,-
0
Lo
Q
U)
0
a)
4-.
0
a)
0.
. oo  o
-, 09
N
0 4
L-
C
*  0
(I)
u OS1
.000
0
. 0 -Ooll
0,
-
tC            t-                00             C
-              I
-1:1
yz _
a1)
-P
H , -
1
l o
0
%t  .
.   co
ii-
2 0.
ci
1J
0
Q   co
Q A
0 *
r-4
I
p
pp
c
p
6
C.,
Cq
ci
II
7
I
-
1-
,:;. 
-. L MY
.
46
: ^'
1-
1-1
f
m

r
m
1

I
A
I
C

r:--
GI,
30
ot
I
r-4
- --,>   m                     A
'N
4
II. f- r-
c l 11--Z -  -1
l-- - -Tt
llll

/
I,;,p
I
I.
1-4
/
C-
._  ,,c:-  M-.
_ AS  a
I-_
!it
I
*4.
if-
U
0
0
0
V.
-
-or'
0
E.
0
U
0.
0
"10
H.
bD
I
4
i
I
I
_    M--
tV-
4,
-i
aa
0
pb
5
rz
d
C>
es
no
cS4
. .
34
BIBLIOGRAPHY
Atkinson, R.J.C.'
1956
1959
1961
Atkinson, R.J.C.
Stonehenge. London: Hamish Hamilton.
Stonehenge and Avebury. H.M. Stationery Office.
London.
Neolithic Engineering. Antiquity 25:140.
and B.S. Browne
Notes on 'Neolithic Engineering'. Antiquity 37:146.
Barber, F.M.
1900        The Mechanical Triumphs of the Ancient Egyptians.
London: Kegan Paul, Trench, Trubner.
Berger, R., J.A. Graham and R.F. Heizer
1967        A Reconsideration of the Age of the La Venta Site.
Contributions of the University of California
Archaeological Research Facility 3:1-24.
Bernal, I.
1969        The Olmec World.  Berkeley and Los Angeles:
University of California Press.
Breasted, J.
n.d.        Ancient Records.   5 vols.  Cambridge University Press.
Childe, V.G.
1950        The Urban Revolution.  Town Planning Review 21:3-17.
Clewlow, C.W., Jr.
1974        A Stylistic and Chronological Study of Olmec
Monumental Sculpture. Contributions of the
University of California Archaeological Research
Facility 19.
Clewlow, C.W., Jr., R.A. Cowan, J.F. O'Connell and C. Benemann
1967         Colossal Heads of the Olmec Culture. Contributions
of the University of California Archaeological
Research Facility 4.
Clark S. and R. Engelbach
1930        Ancient Egyptian Masonry. Oxford University Press.
Coe, Michael D.
1957
1962
Cycle 7 Monuments in Middle America: A Reconsideration.
American Anthropologist 59:597-611.
Mexico. London: Thames and Hudson.
35
Coe, M.D.
1965a
1965b
1965c
1967
Archaeological Synthesis of Southern Veracruz and
Tabasco. Handbook of Middle American Indians 3:
679-715. Austin: University of Texas Press.
The Olmec Style and its Distributions. Handbook of
Middle American Indians 3:739-775. Austin:
University of Texas Press.
The Jaguar's Children: Pre-Classic Central Mexico.
New York: Museum of Primitive Art. 126 p.
Map of San Lorenzo. New Haven: Yale University
Press.
San Lorenzo and the Olmec Civilization.
Oaks Conference on the Olmec, p. 41-78.
Dumbarton
1969        Photogrammetry and the Ecology of Olmec Civilization.
Paper read at the Working Conference on Aerial
Photography and Anthropology, Cambridge, Mass. May
10-12.
Coe, M.D., R.A. Diehl and M. Stuiver
1967        Olmec Civilization, Veracruz, Mexico: Dating of
the San Lorenzo Phase. Science 155:1399-1401.
Comision Intersectorial Coordinadora de Levantamientos de la Carta
Geografica de la Republica Mexicaria
1957        Estados Unidos Mexicanos. (1:500,000). mapa.
Cook, S.F.
1960
and W. Borah
The Indian Population of Central Mexico,
Ibero-Americana, No. 44. Berkeley.
1531-1610.
de la Fuente, B.
1973
Drucker, P.
1947
1952
1961
Escultura Monumental Olmeca. Mexico.
Some Implications of the Ceramic Complex at La Venta.
Smithsonian Miscellaneous Collection: 107.
La Venta, Tabasco: A Study of Olmec Ceramics and
Art. Bureau of American Ethnography--Bulletin 153.
The La Venta Olmec Support Area. Kroeber Anthropo-
logical Society Papers 25:59-72.
Drucker, P.,
1959
R.F. Heizer and R.J. Squier
Excavations at La Venta, Tabasco, 1955.
American Ethnography--Bulletin 170.
Bureau of
Drucker, P. and
1960
R.F. Heizer
A Study of the Milpa System of La Venta Island and its
Archaeological Implications. Southwestern Journal
of Anthropology 16:36-45.
1968
Edwards, C.E.
1965
Erasmus, C.
1955
1965
Erman, A.
1894
Aboriginal Watercraft on the Pacific Coast of South
America. Ibero-Americana 47. Berkeley.
Work Patterns in a Mayo Village. American Anthro-
pologist 57:2.
Monument Building, Some Field Experiments.
Southwestern Journal of Anthropology 21:4.
Life in Ancient Egypt. London.
Foster, George M.
1942        A Primitive Mexican Economy. Monographs of the
American Ethnological Society V. 115 p.
Grant, J.
1966        A Pillage of Art.  Roy Publishers, N.Y.
Hawkes, J. and Sir. L. Woolley
1963        History of Mankind, Cultural and Scientific Development
Vol. I. Published for the International Commission
for a History of the Scientific and Cultural
Development of Mankind, London: George Allen and
UNWIN Limited. 873 p.
Heizer, Robert F.
1959        Scientific and General Characteristics of Olmec
Culture. Actos del XXXIII Congreso Internacional
de Americanistas II:178-182. San Jose.
1960        Agriculture and the Theocratic State in Lowland
Southeastern Mexico. American Antiquity 26:215-222.
1961        Inferences on the Nature of Olmec Society Based upon
Data from the La Venta Site. Kroeber Anthropological
Society Papers 25:43-57.
1962        The Possible Sociopolitical Structure of the La
Venta Olmecs. Acten des 34th Internationalen
Amerikanistenkongresses, Wein, 1960:310-317.
1966        Ancient Heavy Transport, Methods and Achievements.
Science 153:821-830.
1968        New Observations on La Venta. Dumbarton Oaks
Conference on the Olmec. p. 9-40.
Heizer, R.F., J.A. Graham and L.K. Napton
1968        The 1968 Investigations at La Venta. University of
California Archaeological Research Facility,
Contributions 5:127-154.
36
37
Heizer, R.F.,
1965
Heyerdahl. T.
1958
Howells, W.W.
1960
T. Smith and H. Williams
Notes on Colossal Head No. 2 from Tres Zapotes.
American Antiquity 31:102-104.
Aku-Aku. New York: Allen and Unwin.
Estimating Population Numbers through Architectural
and Skeletal Remains. In Applications of Quantitative
Methods in Archaeology. Viki-ng Fund Publication #28:
158-176.
Kaplan. D.
1963
Men, Monuments, and Political Systems.
Journal of Anthropology 19:397-407.
Southwestern
Kennedy, A.
1821
Kida, S.
1912
Lamb, F.B.
1966
Layard, A.H.
1853
Metraux, A.
1957
Mulloy, W.
1970
Poleman, T.
1964
Account of the Erection of a Granite Obelisk, of a
Single Stone, about seventy feet high, at Seringapatam.
Transactions of the Royal Society of Edinburough,
Vol. IX, p. 307-316.
Dokushi Hyakuwa. Tokyo: Sanseido Shoten. pp. 190-
193. Manuscript translated by Richard W. Howell and
Masaaki Takane.
Mahogany of Tropical America. Ann Arbor: University
of Michigan Press.
Nineveh and Babylon, Second Expedition. New York:
G.P. Putnam & Co.
Easter Island, a Stone-Age Civilization of the
Pacific. Translated from the French by Michael
Bullock. New York: Oxford University Press. 248 p.
A Speculative Reconstruction of Techniques of Carving,
Transporting and Erecting Easter Island Statutes.
Archaeology and Physical Anthropology in Oceania 5:1.
The Papaloapan Project: Agricultural Development in
the Mexican Tropics. Stanford: Stanford University
Press. 167 p.
Sahlins, M.D.
1955
Sanders, W.T.
1953
Esoteric Efflorescence in Easter Island.
Anthropologist. Menasha, Vol. 57, 1955.
1052.
American
p. 1045-
The Anthropogeography of Central Veracruz. In
Huastecos, Totonacos, y sus Vecinos. Revista
Mexicana de Estudios Anthropologicos. Vol. 13,
Nos. 2-3. pp. 27-78. Mexico.
Sanders, W.T. and B.J.Price
1968        Mesoamerica: The Evolution of a Civilization.   New
York: Random House, 264 p.
Secretaria de Agricultura y Gauaderia de Mexico
1952        Mapa del Golfo de Mexico--Coatzacoalcos.   (1:100,000).
Secretaria del Estado
1889-1911   Carta de la Republica Mexicana (1:100,000).
Comision Geografica de Guerra y Fomento. Mexico.
Solver, C.V.
1940        The Egyptian Obelisk-Ships. The Mariner's Mirror.
Vol. XXVI, No. 3. pp. 237-256.
Standley, P.C.
1920        Trees and Shrubs of Mexico.   U.S. Government
Printing Office.
Stirling, M.W.
1955        Stone Monuments of the Rio Chiquito, Veracruz,
Mexico. Bureau of American Ethnography--Bulletin
157:5-24.
Takahashi, Itsuo
1937        Megalith Transport and Construction of Megaliths
of the Ishibutai. In Report Upon Archaeological
Research in the Department of Literature, Kyoto
Imperial University, Vol. XIV, 1937. Tokyo.
Tamayo, J.A.
1964        The Hydrography of Middle America.   Handbook of
Middle American Indians. 1:84-121. Austin:
University of Texas Press.
38
39
Vivo E. Jorge A.
1964        Weather and Climate of Mexico and Central America.
Handbook of Middle American Indians 1:187-215.
Austin: University of Texas Press.
Wagner, P.L.
1964        Natural Vegetation of Middle America. Handbook of
Middle American Indians 1:216-265. Austin:
University of Texas Press.
West, R.C., N. P. Psuty and B.G. Thom
1969        The Tabasco Lowlands of Southeastern Mexico.
Louisiana State University Coastal Studies Series
#27. Baton Rouge: LSU Press. 193 p.
West, R.C.
1964        Surface Configuration and Associated Geology of
Middle America. Handbook of Middle American Indians
1:33-83. Austin: University of Texas Press.
Williams, H. and R.F. Heizer
1965        Sources of Rock Used in Olmec Monuments. Contributions
of the University of California Archaeological Research
Facility 1: 1-39 .
Wilkinson, J.G.
1842
Wilson, T.
1888
Manners and Customs of the Ancient Egyptians.
I, Vol. iii. London. Second edition.
Series
Megalithic Monuments of Brittany. American Naturalist
22:573-589.