HomeBiographyWritingsPhotosContactAncient Maya Settlement Patterns and Environment at Tikal, Guatemala: Implications for Subsistence Models Dennis Edward Puleston A Dissertation in Anthropology, University of Pennsylvania, 1973
 

Chapter Six: Estimates of Ancient Populations

 
 

The Population of Late Classic Tikal
Discussion of the North and South Survey Strips in Terms of Cultivable Land
The Population of Early Classic Tikal
The Population of Late Preclassic Tikal
Postclassic Settlement in the Tikal Area
The Settlement of Intersite Areas
Total Structures in the Tikal National Park

 

The following sections dealing with population estimate based on structure counts (Appendix 6, Figures 7, 8, 9, 10, and 11) and the results of the ceramic test-pitting survey (Appendix 7, Figures 12-13) are dealt with in the order of their  reliability. Consideration of the Late Classic period comes first and is followed by the Early Classic and Preclassic. The Postclassic period is placed last, not because it is least reliable, but because to the special problems and questions it poses.

The question of what the population of Tikal was at any time rests largely on the definition of Tikal as a site. Such definition was one of the original aims of the present study. Haviland (1965:19), in his original estimates of Tikal's population, used the outermost boundaries of the Tikal site map. In this area of 16 km2 he calculated a Late Classic population of 10,809 persons. With the mapping of the South Brecha Survey Strip in 1965, the discovery of a significant drop in structure density of 6.5 kilometers placed the problem in a new light. Tikal did seem to have limits which could be defined in terms of settlement density. The discovery of the North Earthworks on the North Brecha Survey Strip at approximately 4.5 kilometers confirmed our emerging impression of the magnitude of the site (see Fig. 1 in Puleston and Callender 1967:41).On the basis of a preliminary report on  the present survey, Haviland (1969:430) published a new, rounded-off estimate of 49,000 persons for Late Classic Tikal. Because of better information the location of the south earthworks, this figure was reduced to 45,000 a year later (Haviland 1970:193). More recently Haviland (1972:138) has reduced his figure again, this time to 40,000 persons, on the basis of new information relevant to the calculation of the average size of Maya nuclear family. It is to be understood that this is a "rock bottom" estimate (Haviland 1972, personal communication) which it surely is.

The present analysis confirms his original estimates. In the following discussion only the Late Classic data will be referred to, though Preclassic and Early Classic data are also presented in the tables. Evidence relevant to these time periods will be discussed individually in separate sections that follow. The Late Classic data are presented first and by themselves because they are comparatively much more reliable.

The Population of Late Classic Tikal

Calculations based on the North and South Survey Strips

The first level of information used to estimate Tikal's Late Classic population comes from the ceramically tested North and South Survey Strips. Material derived from Appendix 3 and presented in Table IV shows that for the sections of "Residential Tikal" (only Epicentral Tikal is excluded) between 5.0 km north and 6.5 km south, there is a total of 660 residential structures in use in Late Classic times. In a total area of five km2, this means there is a density of approximately 132 residential structures/km2. The variation in settlement density within these areas, as noted by Fry (1969), will be discussed subsequently though it is worth remarking now while our attentions on Table IV, the great disparity between the residential structure densities on the North and South Strips. These are 105 and 150 structures km2 respectively. The number of occupants per house, estimated above on the basis of floor areas, is 5.4 persons. When this figure is applied to the density figure of 132 structures/km2, we find that "Residential Tikal" may have sustained a population density of 713 people/km2 . This is not an inflated figure as we shall see; in fact, it is a :total area" figure for "Residential Tikal" as revealed on the North and south Survey Strips. Because the covered portions of the strips include more than three quarters of a square kilometer of escoba bajo and follow a deep arroyo on the north it is strips had been placed half a kilometer to the east or west.

Haviland (1969:430) refers to an "earlier estimate: of 600-700 persons/km2 for the overall population density of Tikal [in fact Haviland's (1966:35) figures of 10,000-11,000 in 76 km2 yield a range of 625-688 persons/km2]. This estimate was for the central 16 km2 of Tikal and included all bajo areas; it is not far off the figure of 713 people/km2 for residential Tikal arrived at above. The fact that Haviland's figure is slightly lower is attributable to two factors. First, a significant proportion of the central 16 km2 of Tikal is taken up by Epicentral Tikal where apart from the Central Acropolis the density of residential structures is low. Second, the peripheral five km2 which makes up nearly one third of Haviland's sample was not thoroughly mapped and missed many structures.

TABLE IV.    Presentation of data on "Residential Tikal" and "Intersite Areas" for the North and South Survey Strips, including data on areas and the total number of residential structures in use (from Appendix 3), followed by calculations of structure and population densities.

TableIV

Calculations Based on all Six Survey Strips

Turning now to Table V, the reader will observe that information from portions of all six of the survey strips have been brought together. Since we are now dealing with ceramically untested settlement remains from four of the strips, consideration of the least reliable temporal category, the Preclassic is deferred.

Percentage occupation figures are deduced from the combined North and South Survey Strip ceramic data and used here because the newly mapped portions of the East and West Survey Strips were not test-pitted. Data for the excavated and test-pitted groups on the portions of the East and West Strips which fell within the central 9 km2 of Tikal were not used because: 1) it was felt they would not be as representative of settlement further from the center of Tikal as the North and South Survey Strip data; and 2) they lacked the advantage of having been randomly selected and seemed to have locational biases towards  areas closer to the center of Tikal and the Tikal Project camp headquarters. Spatially Table VI is divided horizontally into consideration of "Residential Tikal" and the "Intersite" Areas which extend to the edge of what appears to be "Residential Uaxactun." Vertically the table is divided into two blocks, "Cultural Data" and "Environmental Data" brought together from Appendix 2. The latter could have been arranged on a separate table but since the same divisions are used for both bodies of data it seemed easier to follow if presented on a single table.

Percentage occupation figures are deduced from the combined North and South Survey Strip ceramic data and used here because the newly mapped portions of the East and West Survey Strips were not test-pitted. Data for the excavated and test-pitted groups on the portions of the East and West Strips which fell within the central 9 km2 of Tikal were not used because: 1) it was felt they would not be as representative of settlement further from the center of Tikal as the North and South Survey Strip data; and 2) they lacked the advantage of having been randomly selected and seemed to have locational biases towards  areas closer to the center of Tikal and the Tikal Project camp headquarters. Spatially Table VI is divided horizontally into consideration of "Residential Tikal" and the "Intersite" Areas which extend to the edge of what appears to be "Residential Uaxactun." Vertically the table is divided into two blocks, "Cultural Data" and "Environmental Data" brought together from Appendix 2. The latter could have been arranged on a separate table but since the same divisions are used for both bodies of data it seemed easier to follow if presented on a single table.

TABLE V.  The calculation of total residential structures in use in "Residential Tikal" and "Intersite Areas" during Early and late Classic times for all six Survey Strips. Also presented are total area for each strip segment, as well as calculations of the area occupied by uplands, escoba bajo, and tintal bajo. These data are used in the calculations of settlement and population density presented in Table VI. Structure counts are taken from Appendix 4, dpp values from Table I and environmental data from Appendix 2.

TableV

Cultivable Land and Population Density for all Six Strips

Table VI presents a total of 48 separate final calculations relating to settlement and population densities and their relationship to the land. We will deal here only with those calculations pertinent to "Residential Tikal" in Late Classic times. At the bottom of the third column of figures to the right, under "Residential Tikal," we find that the density of Late Classic residential structures, on a total of 11.25 km2 is 99.6 or approximately 100 structures/km2.

We now have the choice of considering Late Classic settlement density in terms of a number of different categories of "available cultivable land," thereby removing the distortion in settlement density figures created by the variable of bajo and construction area. For instance, if we have reason to believe that settlement density and distribution within "Residential Tikal" is dependent only on the availability of well-drained uplands, we can note at the top of the list that the effective settlement density is 170.5 structures/km2. If escobal is to be included as cultivable lands this figure drops to 121.7. If we want to include tintal as an immediate influence on settlement density, although it probably was not and exclude only those areas actually covered by plaza surfaces and construction, effective density drops to 104.5 structures/km2. For description and information on the variation bajo soils and vegetation, see the corresponding sections in Chapter VIII on the environment.

TABLE VI.  The calculation of structure and population densities in "Residential Tikal" and "Intersite Areas" during Early and Late Classic times. The figures in the first column are derived from the last three columns of Table V. The second and third column figures are produced by dividing total residential structures (1074.0, 1120.2, 490.6, and 4.5.5 from Table V), by "cultivable land" values in the first column. The fourth and fifth column figures are derived by dividing the first column figures by total residential structures (1074.9, 1120.2, 490.6, and 415.5, again from Table V). The sixth and seventh column figures are obtained by multiplying the second and third column figures by 5.4. Though fully excavated structures in intersite areas tended to e smaller than 54 m2, it is felt that the excavated sample is biased towards smaller structures also so the standard figure is used.

TableVI

Moving on to the fifth column in Table VI, the reader will find a series of four sets of figures which provide calculations of the amount of land that would have been available to each residential structure if 99% of them were contemporaneously occupied in Late Classic times, as has been suggested by ceramic test-pit and excavation data discussed above. Starting at the bottom of the column, we will see that if we take the total area of 11.25 km2, each of 1,120 structures would have had an average of one hectare (2.5 acres) for use in whatever way the occupants saw fit. Removing the area taken up by plazas and structures, we find that they actually would have had .96 hectares (2.37 acres) available for kitchen-garden food production. This calculation does not take into account space taken up by trails, quarries, and other non-productive areas which might have further reduced the space available for food-producing.

If in addition we want to remove tintal bajo from the category of "cultivable land," a subtraction which seems eminently reasonable for the reasons discussed in the section on soils, the amount of land available per structure drops to .82 hectares (2.03 acres).

Finally we come to consideration of escoba bajo, which at best must be considered marginal, not because of any lack in the soils but because such lands are occasionally subject to flooding. Escoba bajo for this reason appears to pose significant limitations on the cultivation of maize (Reina 1967:13), root crops (Cowgill 1971:54), and perhaps even the ramon, which does not seem to do particularly well on poorly drained soils (Puleston 1968, Figure 8, p. 40). If escoba bajo is removed completely from the category of "cultivable land," each structure could not have had more than an average of .59 hectares (1.45 acres) at its disposal in Late Classic times. Since the bajo was probably used in some way, perhaps as a "high-risk area" for the cultivation of maize, the effective "land available" figure probably lies somewhere between .59 and .82 hectares for an average structure in an average year. If maize was cultivated in escoba bajos, this assumes that these lands (about .5 acres per house) and their products would have been available to the average householder, which of course they may not have been.

In the seventh column of Table VI, the calculations derived in column 5 are simply multiplied by 5.4 to obtain an estimate of population density/km2 during Late Classic times. For "Residential Tikal" an overall average density for the sample of terrain covered by the four survey strips comes to about 538 people/km2. If we with to consider only the uplands, to which in fact settlement is restricted with only rare exceptions (Strs. 30-11/15 in the central nine km2 of Tikal and Strs. SE(E)-1 and SE(S)-374 on the survey strips), densities within the earthworks may have typically fallen within the range of 921 people/km2. Intermediate values, depending on whether or not we wish to include escoba and tintal bajo, are also given. These calculations become useful when we with to estimate population densities on the basis of large scale maps or aerial photographs, as will be demonstrated below.

Originally, I found myself, at this point, subtracting the small portions of the survey strip sections considered on Tables IV, V, and VI that extend into the central nine km2 of Tikal.  For the purpose of obtaining a representative sample of "Residential Tikal" settlement, I no longer feel this is warranted for three reasons. First, areas of high settlement density within 1.0-1.5 km of the center of Tikal are not unrepresentative of other portions of "Residential Tikal." are not unrepresentative of other portions of "Residential Tikal." From the small portions of ridge-top settlement covered in the vicinity of Chikin Tikal (between 4.0 and 4.5 km on the West Brecha Survey Strip), at the end of the East Brecha Survey Strip (between 11.0 and 12.0 km), and on the South Brecha Survey Strip near Bobal (between 4.5 and 5.0 km), I would judge that blocks of settlement as dense as anywhere in the central nine km2 of Tikal would be found in many places in a complete survey.

Second, ridge-top settlement is under-represented on the survey strips. As a result of chance more than anything else, the survey misses the major ridge-top areas withinwhat appears to be "Residential Tiakl." The East Strip skirts the southern edge of the Santa Fe Bajo and only crosses ridge-tops notheast of Corosal and at its eastern terminus. The South Strip, by chance, as can be seen in Figure 6, passes through a trough in the southwest-northeast running Üolantun ridge, missing the major high points, particularly towards the northeast. TheWest Strip, apart from its brush between Chikin Tikal and Canmul, completely by-passes the major area of high ground which lies between thee sites and Tikal, a little to the south. Finally, the North Strip bypasses high ground, most significantly just within the earthworks because it is almost exactly aligned with a long, twisting arroyo that drains into the Bajo de Joventud. Low settlement density in this area was noted by Fry (1969) and, I think, can best be attributed to the large areas taken up by the arroyo bottom, which is: 1) subject to flooding; and 2) bordered by the long steep slopes that follow it. The situation seems to be analogous to that found along the steep arroyo that runs southeast from the Aguada Subin, perhaps even to the degree that structures near the earthworks were comandeered for fill, just as structures in the vicintiy of Temple IV may have been.

Finally, residential settlement within the central nine km2 is an important pat of the sample. In sum, I do not feel that inclusion of the data from within the central nine km 2 unfairly weights the sample, particularly when such a small portion actually falls within what I have designated as Central Tikal on the basis of high settlement density.

The Population Within the Earthworks

With settlement and population densities calculated, we can now go about estimating the Late Classic population that may have occupied all of Tikal within the earthworks and the site epicenter of approximately 1.5 km2 will not be separated out here. To make this estimate, we must determine the area of Tikal within the earthworks, ??????? on the North Brecha and to the Southest between the East and South Brechas (Figure 20).

Haviland (1969:430), assuming Tikal to be roughly circular, arrived at a figure of 162.78 km2 for the total area of Late Classic Tikal. This area, however, includes a good del of tintal bajo which was certainly not inhabited and probably not cultivated. With a better idea as to the location of the southeast earthworks, he reduced this figure to 123 km2 (Haviland 1970:190). In Figure 20 a minimal area for Tikal is indicated by combining data from the observed locations of both the North and Southeast Eathworks, with settlement density data, and the limits of tintal bajo. The area enclosed is approximately 122 km2, including the epicentral area. If a small pocket of tintal in the SE, amounting to approximately 1.5 km2 is subtracted, we are left with 120.5 km2 of uplands and escoba bajo. Table VI indicates a density figure of 657 people/km2 for the combination of uplands and escoba bajo (Af+Ab1) in residential Tiakl. Multiplying this figure  and 120.5 km2 we obtain an estimated population of 79,168 people, or just under 80,000. This is a median figure; neither "rock bottom," nor maximal. It is considerably higher than Haviland's (1972:138) estimate of 40,000 for a number of reasons. First, and perhaps least significant, is my use of 5.4 people/house versus his lates figure of 5. Even if he were to concede the use of 5.4 people/house his estimate would rise to no more than 43,200 people. Of considerably greater significance to the difference, is his miscalculation of the number of people and structues per km 2 in a central zone of approaches approximately 60 km2

Figure 20. Probable boundaries of Classic Tikal, as defined by a combination of settlement density, earthworks, terrain, and locations of swamps.

Figure20

In the original definition of this zone, incorrect information he had received from Fry on the location of the South Earthworks led him to place the southern limits of Tikal at 10 km, giving Tikal a diameter of 14.5 km and a total area (circular) "on the order of 162.78 km2. It was enclosed by "peripheral zone" of 60 km2. With a total of 123 km2 and a north south axis of 12.5 km. I would guess that Haviland now conceived of Tikal as having a rectangular shape with an east-west axis of about ten km. He did not change the estimated population density (100 people/km2) of his "peripheral zone" which , if it followed the presumed rectangular shape, amounted to a peripheral trim 1.5 km. wide. As in the case of Teotihuacan (Millon 1970), Tikal was not circular or rectangular in shape but more amoeba-like, spreading out onto terrain favorable for settlement and cultivation and avoiding swampy areas where such activities would have been difficult Figure 22). Haviland (1970:186) appears to have avoided full consideration of settlement density in defining the limits of Tikal, as evidenced by his selection of 8.0 km rather than 6.5 km for the southern limit of the site.

Here we come to the crucial difference between Haviland's and my estimates of Tikal's population. Within the amoeba-like form (120 km2) settlement density is high almost by definition; within the circular or rectangular form (123 km2) large areas of logwood bajo and even uplands with low settlement density have been included.  In Haviland's periphery he apparently calculates an average of 20 Late Classic structues/km2. For an equivalent zone of 1.5 km on the more irregular shape, let us say between 5.0 and 6.5 km on the South Survey Strip there is an average of more than 112 Late Classic structures/km2 (see Appendix 3); a figure five times greater than Haviland's. Thus though his total area for Tikal is essentially identical to mine the results are quite different when he assumes that nearly half that total area had only 20 structures and less on it per square kilometer. As shown on the last line of Table V Haviland's density figure for his peripheral zone inside the earthworks is far too low even for settlement outside the earthworks. With 596 structures mapped on intersite areas totaling 15.12 km2 the average of 39.4 structures/km2 is double Haviland's figure of 20 structure/km2.

I would like to point out that 112 structures/km2 is not a high figure; the area between 5.0 and 6.5 km on the South Survey Strip includes substantial areas of corosal and low ground that is transitional into escoba bajo on which settlement does not occur.

Haviland claims his figure is "rock bottom" but it is my opinion that he has already penetrated a rather soft bottom. At this point, I doubt if there are any "rock bottoms" in this business. A "low" figure on the basis of my approach, is that which results from an overall density figure based only on those portions of the four survey strips that fall outside the central nine km2 but within the site limits I have defined (Figure 22). As I have pointed out above, however, I feel that this procedure creates a less representative sample because of the disproportionately large areas of escoba bajo included on the sections of the East and West Strips where they are close to Tikal, which are in included in the sample and the "uplands escobal" density figures (Table VI) for residential Tikal. As these calculations are fairly easy to run through, I will calculate this figure now for the sake of the completeness of the completeness of the argument. The    cut-off points, half a kilometer inside the double lines which mark the boundaries of the 6 km2 map (Carr and Hazard 1961) are as follows: North Strip - 1.64 km, South Strip - 1.36 km, West Strip - 1.32 km. East Strip - 1.66 km. The resulting sample area combined from all four strips is reduced from 11.25 km2 to 9.76 km2. This figure is further reduced to 8.25 km2 after 1.52 km2 of tintal bajo on the east strip is subtracted out. On this area of uplands and escoba bajo, then, a total of 960 structures have been located (derived from Appendix 6 and the maps). This yields an average density of 116.5 structures/km2. If we assume then that 99% of these were occupied in Late Classic times, as suggested by the average of the dpp values for Fry's evaluations summarized in the NBSS I and SBSS I, II columns presented in Table I, we are left with 115.3 structures km2, presumed to have been occupied contemporaneously in Late Classic times. Further reducing this figure by 16%, to eliminate non-residential structures leaves 96.9 structures. Finally, multiplying this figure by 5.4 persons/residential structure we obtain a population density figure of 523 people/km2 for the sample.

Haviland's (1965:19) population calculations for the central nine km2 can now be used here. though for the sake of consistency the calculated figure of 5.4 people/structure will be substituted for his 5.6 This results in a total of 7,803 people occupying the 1.445 structures counted by Haviland as contemporaneously occupied Late Classic residences. It should be remarked that this figure does not include what has been defined as the site "epicenter," though careful study of major "palace" structures suggests that they probably functioned as elite residences, as well as centers for administrative and ceremonial activity (Harrison 1968:172). Harrison (1970) estimates that the Central Acropolis area alone was occupied by 200 people.

Now we may calculate the total population for Tikal within the earthworks in Late Classic times. If we subtract nine km2 from the total calculated area of "the site of Tikal" (120.5 km2) we are left with 111.5 km2, we obtain a figure of 58,314 people, which, when added to central nine km2 figure, yields a total of 66, 177 people or just over 65,000. As I have stated above, I fell that the sample used to arrive at this figure is disproportionately high in escoba bajo because of the way the east and west strips skirt the tintal bajos on either side of Tikal. Confirmation or rejection of my ideas in this regard will have to await further mapping. In the meantime we have what is perhaps a more useful "ranging" estimate of 65,000-80,000 people for "the site of Tikal" in Late Classic times.

Discussion of the North and South Survey strips in terms of "Cultivable Land"

A glance down the Late Classic (L.C1.) column of Appendix 4 reveals that upland structure densities are fairly consistent within certain sections of the survey strip maps. These areas, which will be designated as "Sections," can be defined by breaks in the consistency of upland settlement as calculated in the appendix after all mapped bajo areas have been subtracted. The purpose of this exercise is to help define Late Classic Tikal as a "site" and also to demonstrate how cultural and environmental influences on settlement density can be separated. Since these designations depend on ceramic as well as mapping data, only the North and South Brecha Survey Strips will be considered.

Section 1, 8.0-12.- km (North Strip)

Starting at the north end of the North Survey Strip we note that upland settlement destiny, as reflected in the calculated "Hectares of Good Land per Residential Str." is high. Land available runs from .36 to 1.34 hectares per structure. This zone of denser settlement may be associated with the site of Jimbal, in which case it may fit the definition of "zone" as described by Bullard (1960:367), though at the indicated size it would probably consist of more than 50 to 100 plaza units. The lateness of the carved monuments might be used to argue that Jimbal was not more than a satellite to Tikal or Uaxactun until very ate in Late Classic times. If the above settlement is included within the Jimbal site area, it would have a radius of almost five km, which seems too great. Two possible explanations come to mind: 1) Jimbal has an elongated shape that follows the east side of the long arroyo that goes into the Bajo de Joventud; and 2) a satellite may lie to the east or west of the Survey Strip between 9.0 and 10.0 km. I have been strongly tempted to mark off this section at 11.0 km for two very different reasons: 1) the distance to Jimbal and the size of the site, which would suggest a smaller radius of dense settlement if Jimbal is the only site in the vicinity; and 2) the poor representation of Imix ceramics in the test-pit material, in contrast to the 11.0-12.0 km portion of this section and Section 2 (Figure 12). This poor representation of Imix makes the 8.0-11.0 km portion of this section rather comparable to Section 7 on the South Survey Strip, which will be discussed below. For the sake  of discussion, these two "sub-sections" will be designated Section 1-A (11.0-12.0 km) and Section 1-B (8.0-11.0 km). I have not made a major section break here because of the consistently high level of settlement density on all but one patch of low ground between 8.5 and 9.0 km. Many structures on this section have the appearance of being crowded right down to the edge of escoba bajo (Strs. NW(N)-256/266, NE(N)-119/131).

Section 2, 5.0-8.0 km (North Strip)

The "Land per Structure" figures here are all over two hectares, ranging from 2.05 to 15.32 hectares. This wide variation seems to be a reflection of comparatively low settlement density. We are still in the vicinity of the arroyo, but unlike the area just inside the earthworks there is plenty of level ground. IN spite of the presence of a few large groups, including the unusual Strs. NE(N)-102/105, this section presents no comparison to the crowded appearance of either portion of Section 1. The possibility that this section, along with Section 10B and Section 8, were areas of slash-and-burn agriculture in Late Classic times will be discussed below.

Section 3, 1.5-5.0 km (North Strip)

Denny and Dee Dee Green preparing to send drums out Navajuelal for storing water from drying aguada

This section is characterized by "land available" figures that range from .46 to 2.51 hectres per reidential structure. Excluding the latter figure, which appears to be an exception, the highest figure is 1.15. The high value of the 4.0-4.5 km block, as suggested above, may be due to the steepness of the sides of the arroyo which takes up most of this block. The possibility also exists that some sort of vacant zone that followed the inside of the earthworks may have been defined for cultural reasons by the Maya. However, and explanation would then be required for the presence of Strs. NW (N)-178/181, NE (N)-74, 75, and 389 in this zone. The location of the large, tightly-clustered group consisting of Strs. NW(N)-182/194 outside the earthworks perhaps deserves comment here, since it was not discussed in the direct review of the maps. It seems curious that such a prominent group would not receive the protection that must have been offered by the apparently defensive earthworks system.

Why is the group neither enclosed by a deviation in the earthworks nor located within them? Fry's excavation of a test-pit in the group produced possible Preclassic, definite Early Classic, definite IK, definite Imix, and possible Eznab ceramics. These sherds suggest that the group was fairly continuously occupied during the time that the earthworks were constructed and through their main period of use. The earthworks do bear a slight relationship to the group as they change direction where they come closest to it . Perhaps the group was already constructed at the time the earthworks were constructed and simply did not fit into the master plan. Local landholders may have had very little to say about exactly where the earthworks passed. The fact that Section 3 extends to 5.0 km rather than 4.0 km is largely a function of the presence of this enigmatic group. Since most of these structures were presumably residence, however, I do not feel that the section limits should be changed.

Within this section, two subsections will be define, 3-A and 3-B. The first, 3-A, extends from 2.5 to 5.0 km, with "land available per structure" figures hovering around one hectare. The second, 3-B, would encompass blocks 68 and 69, extending from 1.5 to 2.5 km, with about .5 hectares of uplands available to each structure. The next major break, which involves a drop in "land available" figures to a value approximately one-third of even the second sub-section described above, occurs at 1.5 km.

Section 4, 1.0-1.5 km (North Strip)

Within this section, settlement is very compact with .16 hectares (less than .5 acres) of uplands available for each structure. It is part of the residential portion of Central Tikal (see Plate 4b). The crowding of structures down to the edge of escoba bajo on the hill slopes north of 1.0 k, and along the north side of the Maudslay Causeway is quite evident.

Though excavations on this portion of the survey strip were not selected randomly, the Evacuation that has been done suggests that the occupation of structures on Section 4 is chronologically equivalent to 3-B. Extending the sample to include the residential settlement that follows the north edge of the Maudslay Causeway to the southwest, we find that of the eight apparently residential groups that have been excavated or tested, seven produced Early Classic ceramics and eight produced Late Classic ceramics.

Section 5, 1.0(N)-0.40(S) km (North and South Strips)

This section covers what has been defined as Epicentral Tikal. The available land figures calculated in Appendix 4 have little meaning here, since it probably provided only a small portion if any at all, of the food consumed by the elite presumed to occupy many of the range type structures here. The land that is available between the Maler, Tozzer, and Maudslay Causeway may have been used for food production though slopes for the most part are fairly steep. As described in the section on soils, evidence of rapid erosion was found in the Causeway Reservoir.

Section 6, 0.40-1.0 km (South Strip)

The calculations from blocks 74 and 75 indicate a return to the situation found in Section 4, with something in the range of .16-.20 hectares per residential structure. The figure for block 74 is slightly inflated since it includes part of the settlement vacuum south of Temple V, which for sacred reasons may not have been available for secular kitchen gardening. A similar vacuum occurs around the site of Bobal on the South Brecha Survey Strip. Chronologically this section again seems to match Sections 4 and 3-B with approximately equal representation of Early and Late Classic ceramics in excavated areas. Out of the 12 groups tested (testing was not random) and excavated, ten produced Early Classic ceramics and all produced Late Classic ceramics. The two that did not produce identifiable Early Classic material were single test pits (Ops. 35 m and 36 A) that produced very few sherds anyway, so the possibility that Early Classic material might be revealed by further excavation remains high.

Section 6 clearly represents more of the residential portion of Central Tikal.

Section 7, 1.0-6.5 km (South Strip)

On this section "land available" figures seem to hover at slightly more than .5 hectres (l.5 acres) per structure. This is roughly comparable to Sections 1 and 3 on the North Survey Strip. The figures calculated for blocks 79 and 81 are above this average but as in block 64, just inside the earthworks, an arroyo with steep slopes has presented flooding dangers that seem to have been avoided wherever possible. Because steep slopes are included as uplands in the calculations, they ten to distort the figures where they occur. This is interesting and noteworhty, for it suggests that arroyo slopes were not preferred for kitchen-gardening. If they were, settlements would simply be more concentrated along their upper edges, resulting in density figures that would be roughly the same as on level ground. Except for the area close to Tikal (the ridge on the south edge of Square 5G provides a good example), on the outer five kilometers of the East Strip, SW of Bobal on the South Strip, and in the vicinity of Chickin Tikal on the West Strip such concentrations do not occur.

The high figure for block 86 (6.0-6.5 km) can perhaps be attributed to poor drainage north and northwest of the Laguna Verde Reservoir.

Section 8, 6.5-10.0 km (South Strip)

"Land available" figures do not seem so variable here as they are in Section 2, to which this section seems to be roughly comparable. The average figure seems to remain pretty steadily in the vicinity of five to six hectares (roughly 15 acres) per residential structure. As I have suggested elsewhere (Puleston 1968), this may have been an area devoted to slash-and-burn maize cultivation and controlled by the elite of Epicentral Tikal. A rather specialized sample of eight small, square (or nearly square) structures were fairly well excavated in this section (Strs. SW(S)-157, 158, 332, SE(S)-373, 452, 382, 393). As I have reported in another part of this paper, all of them show evidence of having been occupied in Early Classic times and only five show evidence of probable or even possible Late Classic occupation. A similar pattern is reflected in the results of the test-pitting survey, identifiable Imix ceramics showing up in only one of the tested groups (Figure 13).

View from one of the Temples of the jungle and the airstrip at Tikal

Section 9, 10.0-10.5 km (South Strip)

The great drop in the "land available" figure for this single block can be attributed to the inclusion of the small settlement which appears to be associated with Navajuelal. The influence of the bajo is also evident in the southerly location of the settlement. Perhaps there is an aguada at the edge of the bajo somewhere to the west.

Section 10, 10-5-12.0 km (South Strip)

In this small section the "land available" figures would seem to suggest a return to the conditions found in Sections 2 and 8, but here I have reason to doubt these calculations. First of all, informal reconnaissance by one of my reliable workmen who camped here suggests that there are numerous structures on the to of the knoll where Strs. SW(S)-200, and 201 are located. Second, the two structures included on the strip are located on the side of a fairly steep slope, not far from the bajo, a pattern characteristic of "crowded" conditions, as we have seen above. Third, the sample size is actually very small.

In conclusion, this short tour down the North and South Survey Strips with the aid of Appendix 4 has shown, in my estimation, the explanatory potential of a careful and methodical integration of ceramic, settlement, and ecological data.

The Population of Early Classic Tikal

As indicated by the various data presented here, one of the more interesting tentative conclusions to be drawn from this study is the small change in settlement density between Early and Late Classic times.

Problems in Calculation

This conclusion cannot be accepted at face value, however. Three major problems must first receive consideration. These are: 1) the much greater chance that Early Classic sherds in contrast to Late Classic sherds, have been spread around by fill transfers carried out by the Maya and in this way "contaminated" plaza groups which were not occupied in Early Classic times; 2) the certainty that plaza groups that started out with one or two structures in Early Classic times had three or four by Late Classic times; 3) the possibility that the number of occupants in an average residence, as reflected in floor areas, changed significantly from Early Classic to Late Classic times; 4) the possibility that Early Classic occupation tended to be less permanent than Late Classic occupation.

The first of these problems provides an example of one of the limitations inherent in a test-pit survey. Complete excavation of a structure platform should reveal fairly clearly whether or not Early Classic ceramics, such as might turn up in a test-pit, are actually associated with Early Classic construction and occupation. Test-pits cannot do this with nearly the same level of certainty., In spite o this difficulty I do not feel the problem to be a serious one, mainly because I do not see any reason why fills should have been transported from one structure group to another, Such transportation is possible, but, with all the basic materials near at hand, it seems to have been unnecessary. Exceptions no doubt occur, and perhaps more frequently so in the vicinity of areas of major construction, such as satellite sites. Haviland (personal communication) reports that in peripheral parts of the central nine km2 there is evidence of older houses having been "scraped away," though he doubts the materials were taken very far.

The second problem is more serious. There is good evidence from excavations in the central nine km2 of Tikal (Haviland 1963), and also from the survey strip excavations for that matter (Strs. SW(S)-159, 332), that the number of structures associated with a plaza increased with time. Assuming a minimal transfer of ceramics within a group, however, a random test-pit survey, theoretically, should side-step this problem for the following reason. I will start with an assumed universe of plaza groups, each with four structures. Beneath two structures in each group there are Early Classic construction. I will also assume for the sake of simplicity that occupation debris associated with these Early Classic structures was incorporated as fill exclusively into the later structures that covered them. A test-pit survey that revealed the collapse and latest midden debris behind these structures a thousand years  later would produce Early Classic material mixed with Late Classic material behind exactly half of these house ruins. The other half would produce only Late Classic material. Thus a completely random test-pit survey would reveal the true nature of this doubling in the number of structures even if only a percentage of them were excavated. Half the time, test-pits would produce only Late Classic material, the other half, they would produce a mixture of Early and Late Classic. With only one test-pit in each group the thoughtful evacuator would not be able to know if it were the number of plaza groups or structures per plaza group that had doubled, but he would have a basis for calculating the correct increment in settlement density and possibly population.

Workmen loading mules to supply satellite excavation in Navajuelal where Dee Dee Green and Lilita Bergs spent 6 days a week

It is at this point that we run into trouble when we consider the test-pit survey that was carried out on the Tikal survey strips, and it is largely the fault of the post-holing technique descibed above. The technique, which brings the excavato to the ceramically most productive spot in the group, would more often than not esult in a test-pit behind or off the end of the sturcture that contained the most productive fill; the on e with both Early and Late Classic ceramics. Since this might also be one of the larger structures (because it was built on top of an earlier one) it might also have been a more important one in Maya times and thus have produced a larger midde, incresing the chance of bias even further. Haviland (personal communication) says that in his experience, the largest structures often produce the smallest middens and that a correlation between structure size and midden size is unlikely.

The essential point here, however, is that chronological representation as well as the ceramic sample was crucial and the solution to the problem would have been to use individual structures rather than plaza groups as the basic unit in the survey. This procedure was considered by Fry (1969:49) but rejected, because "all excavated plazuela units show a relative contemporaneity of structures." This is generally true for Late Classic times in the cental nine km2 of Tikal as revealed by Haviland's excavations, but it is not so true for earlier periods at the same locus, also as revealed by Haviland's (1963) excavations. Another way around this problem would have been to use the same basic plan that was used but further to stratify the sample by randomizing the choice of structures on a selected plaza unit. Then the post-hole diggers could have been used to locate the best spot for a test-pit, on the selected structure, without biasing the sample towards the continuously occupied structures of every group. I strongly urged Fry to use the plaza group as a basic unit in the survey, but I did not forsee the problem I have outlined above. Why have I spent so much time in discussing this problem? the reason is that it is possible that this bias may have contributed, along with the likelihood of fill transfer within a particular group, to a disporportionate upward shift in the representation of Early Classic in our test-pit samples.

The third problem is rather one of interest than one of concern. Its solution will have to await the appearance of information on the size of typical Early Classic house platfoms.

The fourth problem is raised by Haviland's (personal communication) observation that excavated Early Classic structures do not seem to show evidence for as many sequent rebuildings as occur in Late Classic structures. This might be attributed to 1) short occupation and abandoment of Early Classic structures or 2) long occupation with infrequent rebuilding. In support of the former argument Haviland (personal communication) has also observed that in some Early Classic structures there is evidence that they were abandoned early in this period, others appear to have been built late in Early Classic times. Assuming that this was a general pattern population estimates for the Early Classic presented here may be too high. Since the Early Classic ceamics produced by the test-pit survey could not be plaed in a finer temporal category than that spanned by the use of Manik ceramics, about 300 years (i.e. the entire Erly Classic period), the finer divisions recognized by Haviland could not be dealt with. For my purposes here, I have assumed contemporaneous and continuous occupation for all plaza groups which produced Early Classic ceramics. Perhaps if Haviland's data can be quantified and the sample is large enough a correction factor can be appied to figures I use here to determine what proportion of Early Classic structures are likely to have been occupied early during this period as opposed to later.

Depsite the structures Haviland refers to I suspect that many, if not most, Early Classic structures were occupied fairly continuously perhaps with comparatiely less frequent rebuildings than Late Classic occupation sites. The existence of Early Classic chultuns and a spacing of residential groups tht is similar to that of Late Classic times suggest that permanet kitchen gardens were of importance. If this was so it is unlikely that occupation sites would have been abandoned since the land would not have been fallowed. Perhaps a greater frequency of rebuilding in Late Classic times can be related to a socio-economic change of some kind: greater emphasis might have been placed on rebuilding for reasons of prestige, or perhaps as the population increased greater amounts of labor or time could have been diverted towards the rebuilding of private residences.

The Calculations

In the face of these difficulties, a calculation of Early Classic population size at Tikal will be made on the same basis that the Late Classic population estimate was made. comments will follow.

Table IV shows an overalll residential structure density for Early Classic times that is 6.5 strutures less/km2 than that for Late Classic Times. This reduction results in a population density that is 45.1 persons/km2 less, or 678 people/km2. Assuming that the random test-pitting of residential areas in the North and South Survey Strip have produced a chronologically representative picture, we can draw in the raw settlement data from the East and West Survey Strips. For the residential areas this sample includes all the structures between 1.0 and 5.5 km on the West Strip and 0.5 and 8.5 km on the East Strip. When we include them, the density figure drops from 678 to 516 people/km2 (see Table VI). If we consider only uplands and escoba, the figure is 631 people/km2. This is only 26.5 people/km2 less than the Late Classic figure. On a total area of 120.5 km2, Early classic Tikal could thus have had a population of 76,000 people. If "hidden" structures, which in all cases to date have turned out to be occupied in Early Classic times were added, it might even be possible to argue that Early Classic Tikal sustained a larger population than Late Classic Tikal. In light of Haviland's evidence that some Early Classic structures were not continuously occupied, however, it is more likely that this figure should be reduced by some unknown percentage as I have already indicated. We have no way of knowing what  this percentage should be until 1) a more truly random survey is undertaken, 2) more estensive excavation reveals a reliable picture of the proportion of Early to Late Classic structures on a typical pluzuela unit, and 3) we have a good idea of how continuous Early Classic occupation of typical houe sites really was.

The Population of Late Preclassic Tiakl

2.         Our idea of how many structures and/or groups were occupied during the Preclassic period and what their floor areas were is even less certain than it was for Early Classic times. Despite the problems, enough data are available to give us at least a rough idea of the demography of Late Preclassic Tikal. The data from Appendix 3, as presented in Figure 21, indicate that Late Preclassic Tikal was a good deal smaller than Classic Tikal, which is as might be expected. Its "site" radius may have been between 1.0 and 1.5 km, extending  to .25 km on the North Strip and 1.25 km on the south. With a total area of perhaps 3.8 km2, its residential nucleus would have fit within the original 16 km2 site map (Carr and Hazard 1961).  Utilizing again the assumption that group structure counts and floor areas in Late Preclassic times were equivalent to those of the Classic period, with the same doubts and queries that applied to the Early classic data, we may arrive at a structure density estimate of 215 structures/km2 for Late Preclassic Tikal (Table III). This figure is disproportionately high for two reasons, because 1) the very small sample is located on an unusually densely settled section of Tikal, and 2) a conversion factor of 60% was used to convert the number of mapped structures into the number of contemporaneously occupied Late Preclassic structues. This figure was clculated on the basis of excavations in ceremonial and residential structures within all portions of the nine km2 map covered by the North and South Brecha Survey Strips; a step made necessary by the small size of the Preclassic core of Tikal which does not overlaponto the randomly test-pitted protions of the Strips. On the section of the South Brecha Survey Strip between .5 and l.5 km, 16 groups have been excavated, most by a single test pit. Of these five (31%) produced definite Preclasic ceramics, six (37%) produced at least probably Preclassic ceramics, and ten (62%) produced at least possible Preclassic ceramics. On another part of the site, Square 6E, where many groups have been tested, seven out of 17 groups (41%) produced definite Preclassic ceramics. Haviland (personal communication) reports that for the 13 groups he controls in Square 6E, Preclassic representation might be as low as 38% or as high as 46%.

Figure 21. Possible area of the core of Preclassic Tikal, as suggested by the distribution of Preclassic ceramic material found at the site.

Figure21

These data from residential areas, despite sampling limitations, suggest that the 60% conversion factor is too hgih. If we use the lowest of the 6E figures (38%) a much lower and perhaps more realistic structure density for Late Preclassic times is obtained. With a gross total of 2,120 residential structures on the central nine km2 of Tikal, an average of 235 structures/km2 can be derived; and 38% of this is 89.3 structures/km2. On an area of 3.8 km2, this density produces a total of approximately 340 structures which, with an assumed 5.4 people per house, could mean a population of something on the order of 1,800 people. In contrast to the Early Classic figures this seems low. It seems low in terms of the already magnificent architectural and artistic achievements of the Late Preclassic Maya in Tikal's North Acropolis (Coe 1965:17) as well as in terms of even greatly reduced versions of Early Classic population estimates deiscussed above.

Figure 22 seems to show that a ring of lower density settlement extended to 4.0 km on the norht and about 6.5 km on the south. Id this ring is considered to be part of Late Preclassic Tikal, the site may have had an aea of 32 km2. Appendix 3 provides a total of 910.9 people for the four km2 sample of this ring (starting at 1.0 and 1.5 km on the North and South Stips respectively). This yields an average of about 228 people/km2. Thus something on the order of 6.430 people for the extra 28.2 km2 might be added to the core to give us a total of 8,230 people for Late Preclassic Tikal.

Figure 22. Structure profiles for the major time periods on the North and South Survey Strip. The data are derived from the "Structures in Use columns of Appendix 3.

Figure22

1.         Sites like Uolantun which fall within this ring, however, must have had considerable autonomy in omparison to the situation in later times. Excavation at the latter site suggests a considerable degree of relifious independence and wealth in Preclassic and Early Classic times, as reflected in a series of nine major superimposed ceremonial structures, which show a change from a pattern which employed squat but massive lateral extensions to create architectual impressiveness to a pattern similar to that seen in the "Great Temples" of Tikal. All this construction activity came to a halt in Late Classic times. The assumption here is that sequent and innovative architectural developments at such a distance from Tikal are evidence of local control and independence.Similarly the cessation of such investments  of wealth and energy indicate loss of control as a result of gradual centralization of power and influence at Tikal's epicenter. We appear to have here an example of lineraization as defined by Flannery (1972), one of the mechanisms basic to the rise of state societies in which lower-order controls are bypassed by higher-order controls.

Postclassic Settlement in the Tikal Area

Certainly the most dramatic transititon in Fig. 22 is that between Late Classic and Postclassic times. Following a rapid decline in population during Eznab times, Tikal populations drop to near zero. The Postclassic population was so meagre and scattered that the survey strip sample is not really representative. On the portions of the survey strips that fall within the central nine km2 of Tikal, one possible Caban sherd was found in an unsealed excavation unit at the locus of Str. 3D-126. Another possible Caban sherd was found in the vicinity of the monumnets i front of Str. 4D-32. Though no other even possible Caban material turned up on this portion of the survey strips, there is good evidence  for the existence of a small settlement southeast of the Tikal Reservoir in Caban sherds that show up at four loci there (Strs. 5F-17, 18; 4G-7; Ch. 5G-7; and Ch. 5G-8) as well as in the tikal Reservoir itself. I suspect that one or two dozen persons may have occupied this ara in Postclassic times. Another farily substantial deposit was found in Group 4F-2 (Haviland, personal communication),. A scatter of Caban sherds has also turned up in excavations in the vicinities of the Temple of the Inscriptions, and Strs. 6E-143 and 146 as well as in other isolated localities in the central nine km2. In total these finds do not suggest the occupation of more than a hundred or so people and this number may even have been spread out over a considerable time period. Occupation of masonry sturctures, as well as poli and thatch structures, is indicated.

The ceramic survey carried out on the North and South Survey Strips revealed only one other Caban occupation. This material, from Strs. SE(%)-373 and 452, was located in a group southeast of an aguada, as was the largest concentration of Caban material from the central nine km2 of Tikal. The absence of more Caban material in the periopheral areas clearly eliminates the slight possibility which has occasionally been voiced, that in Postclassic times peasant populations may have continued to survive in fairly decent numbers in peripheral areas, even after the ceremonial and administrative center was totally abandoned. A similar postion was implictly taken by Thompson (1931: 230) when he proposed that the collapse of Classic civilization resulted from the overthrow of a small ruling elite, while the rest of the population, "perhaps reduced in nubmers by warfare, disease, or disruption of the economic pattern, evolgved into small isolated ocmmunities under humble leaders without any political or religious cohesion, perhaps a return to Middle Formative status." The use of a piquant metaphorical presentation for the alternative reveals his hand in more recent publications.

It has been further supposed that the great core of  the Central area embracing most of the geat ceremonial centers, misnamed cities, reverted to forest and remained virtually uninhabited for a millenium until awakened from sleep, like some sleeping beauty of the tropics, by the machetes of chewing-gum gatherers and the kisses of archaeological Prince Charmings (thompson 1966:23, 1970:50).

In fact, for the Tikal region at least, this characterization does not appear to be far off the mark.

The Settlement of Intersite Areas

Perhaps one of the most significant accomplishments of the present survey is the production of data relevant to the problem of calculatin settlement and population densities in intersite areas. On the North and South Brecha Survey Strips wher ceramic test-pitting has provided us with some degree of temporal control, a total of 6.25 km2 of what can be considered "intersite settlement area" has been mapped (see Table IV). For all six survey strips the total is 15.12 km2 (Table V), from which 2.56 km2 of tintal bajo and 0.43 km2 of tintal bajo and 0.43 km2 of what is certainly Classic construction can be usefully subtracted, for a total of 12.13 km2  of available uplands and escoba bajo. The total structure count for the intersite portions of all six strips is 596 structures (Table VI), which provides us with an average density of 49.1 structures/km2. The Tikal National Park, which covers an area of 576 km2, can be divided into the categories of 1) "the site of Tikal"--122 km2 (21%), 2) tintal bajo--100 km2 (17%), and 3) combination upland and escoba bajo "intersite" areas, outside the apparent limits of both the North and South earthworks, totalling 354 km2 (62%). On the basis of the data presented in Table for Late Classic times, this intersite area with 34.3 residential structures/km2 may contain as many as 12,142 residential structures. Moving across this table, we can see that this suggests and availability of 2.92 hectares (7.21 acres) or almost three hectares of escoba and uplands per structure, if we assume that these structures were occupied contemporaneously. Here there seems to be some basis for question, particularly in Sections 8 and 1-B. In Section 8 (6.5-10.0 km) on the south Strip, nine out of ten randomly selected and test-pitted groups produced only very slight or no evidence of Late Classic occupation (Figure 13). This is in strong contrast to other sections on the North and South Survey Strips which all have good representation of Late Classic ceramics, including Sections 1-A and 9 in "intersite areas" and Sections 2,  3, 6, and 7 in "Central Tikal," suggesting (though not demonstating conclusively) farily ocntinuous occupation. The one other exception, as we have noted in the description of the sections, is Section 1-B. Here, in spite of high settlement density, the representation of Imix ceramics is surprisingly low (Figure 12), and in this respect Section 1-B is comparable to Section 8. Section 2 also bears a certain similarity to these two with respect to its low setlement density, but Late Classic occupations seem to be farily well represented.

All this seems to suggest something unique exists about the Late Classic occupations of Sections 1-B on the North Stip and Section 8 on the South Strip. This uniquenes does not appear to extend back into Early classic times. Definitely identifiable Early Classicceramics showed up in eight of the ten test-pitted groups, and all eight of the more extensively excavated structures in Section 8. They were also definitelyidentified in all the test-pitted groups in Section 1-B.

What happened to settlement in Sections 1-B and 8 in Late Classic times? Evidence suggests to me that it declined in density for some reason and perhaps became less stable or permanent. As an explanation I would like to refer to an hypothesis I made five years ago (Puleston 1968:114) to the effect that certain outlying areas may have been reserved for slash-and-burn cultivation of maize for the utilization of the elite of Tikal. The evidence for severe limitation on land presented here lends support to the hypothesis that the products of slash-and-burn agriculture would have been the prerogative of the upper classes. The evidence further suggests that as a pattern this trend became particularly significant in Late Classic times, perhaps to the point tha traditionally occupied house sites and gardens  were taken over to make way for the gastronomic inclinations of an increasingly imperious elite.

Total Structures in the Tikal National Park

An estimate of the total number of structures in the Tikal National Park is perhaps worth calculating. Table V shows us that the sample of the "site of Tikal" has 1347 structures on 11.25 km2 for an average density of 119.7 structures/km2. On a total upland and escoba bajo area of 120 km2 this produces a possible total of 13,364 structures.

Table V also shows that on 15.12 km2 of "intersite area" 596 structures were mapped for a density of 39.4 structures/km2. On a total area of 354 km2 this produces a possible total of 13,948 structures. If these two figures are added together, we obtain an estimated total of 27,312, or 25-30,000 surface structures of all kinds for the entire Tikal National Park.

Furthermore, the overall similarity of settlement patterns in the northeast Peten and the ubiquitous distribution of plaza group clusters described by Bullard (1960) suggest that the generalities derived for the Tikal area may well apply to the entire region. In other words, while major sites probably had population densities of 500-1,000 persons/km 2 in their immediate icinity, populations in outlying areas do not seem to have dropped below a range of 150-200 persons/km2. Referring back to Late Classic population densities (Table VI), one can say that major sites probably had upland population densities in a range of 600-1,000 people/km2 and populations in outlying areas of the same type may have ranged from 200-300 people/km2, except of course in those areas which might have been converted to maize cultivation as suggested above.

 

Go Back to Top of Page

PagesNotesPrefaceIntroductionCh 1Ch 2Ch 3Ch 4Ch 5Ch 6Ch 7Ch 8Ch 9ContactVoivode Multimedia