3

The matter of per capita caloric intake of the peasantry is more difficult, even in terms of countrywide generalizations. We have

[26] See the discussion in Borah and Cook, The Aboriginal Population of Central Mexico on the Eve of the Spanish Conquest , pp. 6–21, 60–71, et passim .

[27] For example, Relación de Nexapa, 12 September 1579; Relación de Piaztla, 2 January 1581; Relación de Atitalaquia y su partido, 22 February 1580–in PNE , IV, p. 42; V, p. 79; VI, p. 207. See the very explicit statement in Gómara, II, pp. 400 and 147.

no firm basis for estimating overall production and dividing it by the probable number of people. However, certain aspects of the problem can yield insight under discussion. One of these relates to the energy requirement of the Aztecs and kindred tribes in the conditions of their existence. A present-day European or American man of average size, doing moderate work, is thought to require 3,000 kilocalories a day. Women and children need correspondingly less, such that the mean may be near 2,500 kilocalories. It is very doubtful whether the aboriginal peoples of central Mexico got anywhere near this amount. There is, indeed, an increasing body of evidence suggesting that, except in unusually favorable circumstances, few people even in Europe enjoyed average daily caloric intakes that would correspond to present-day standards of nutrition.^[28]

In order to pursue further our examination of the dietary and metabolic status of the central Mexican Indians prior to the coming of the Europeans, we must first determine certain anatomical and physiological constants. These may be obtained with relative ease for a living population, but can be secured only by indirection and difficulty for a long-extinct people. The first of these magnitudes is simple body size, in its most elementary terms defined as height and weight. From these factors we may derive the probable basal metabolic rate and the caloric ration necessary to support it. Additional data are available concerning physical effort and expendable labor.

Estimates of body size for the pre-Conquest inhabitants of central Mexico must be drawn from three sources. The first is contemporary statements, which are surprisingly scanty. The earliest mention, the report of the city council of Veracruz to Charles V, describes the natives as of medium stature with well proportioned bodies.^[29] Gómara, presumably summarizing the testimony he gathered from Hernán Cortés and other conquerors in Spain after central Mexico was under Spanish control, makes almost the same statement: "of medium stature, but robust."^[30] The Anonymous Conqueror's statement is also close: "well-shaped, rather tall than short."^[31] The implications are

[28] Braudel, Civilisation matérielle et capitalisme , pp. 97–99; Ashtor, "Essai sur l'alimentation des diverses classes sociales dans l'Orient médiéval," pp. 1043–1053.

[29] Veracruz, 10 July 1519, in Cortés, Cartas y documentos , p. 23.

[30] Gómara, II, 398.

[31] [Anonymous Conqueror], Relación de algunas cosas de la Nueva España  . . . , p. 41.

that relative to the Spaniards of the time, the Indians were of average size and weight. Since by our standards the Spaniards of that time were small men, with an average height of perhaps 160 centimeters (5 feet 3 inches) and an average weight of perhaps 60 kilos (132 lbs.), the Indians can have been no taller and no heavier.

The second source of information on body size is the measurement of skeletons of the fifteenth and sixteenth centuries found in excavations in central Mexico. Here too our information is surprisingly scanty, but what there is of it yields reconstructed statures for males ranging from 1.59 to 1.64 meters.^[32]

A third method of estimate is by comparison with present-day conditions. It is possible that some change has occurred in the more than four centuries since the Conquest, particularly through nutritional betterment. However, in rural areas, among the Indians, dietary conditions are much the same as they were in the sixteenth century. Moreover, except in the North of Mexico (which lies outside our area of study) and among the mestizos, no increase in size has been noted, and certainly no decrease. We are therefore justified in thinking that body size among the rural indigenous population has not altered significantly since the time of Cortés.

There have been literally scores of investigations which have included measurements of the height and weight of groups of people all over the earth. It is manifestly impossible and also unnecessary to cite most of these. However, it is worthwhile to mention some of the data which concern the aboriginal inhabitants of both North and South America, with the emphasis upon Mexico. These figures have been compiled and are presented in condensed form in Table 2.1.

For height, the eighteen averages of adult males in Indian groups living north of central Mexico have a range of 161.1 to 174.9 and a mean of 166.5 centimeters. These values are as good as can be obtained with present-day anthropometric methods. The samples are adequate and the techniques satisfactory. For the set of twenty averages obtained for Indians in central Mexico, the analogous mean is 158.5 and the range from 154.2 to 163.4 centimeters. If the individual averages for each area, central Mexico and north of central Mexico, are compared

[32] Genovés T., "Anthropometry of Late Prehistoric Remains," in HMAI , IX, table 3 between pp. 40–41.

numerically, the value of t is 6.39. A very significant difference in height between the Indians of the two areas is therefore indicated.

There are in Table 2.1 sixteen averages for Maya groups in Yucatán and Guatemala. For them the mean is 156.0 centimeters. If t is calculated for the Maya against the central Mexican samples, it is found to be 2.83, a figure just at the 1% level of probability and hence definitely significant statistically. However, the Mesoamerican and South American samples may be compared by making an analysis of variance for the three aggregates, representing central Mexico, Yucatán, and South America. The value of F equals 1.30, a totally nonsignificant figure. Therefore, these reports as a whole show heights within a very narrow range, probably within the margin of error of sampling and measurement.

For the central Mexican group, in which we are particularly interested, the mean is 158.5 centimeters, but the variation is considerable. The Tarascans and some of the Nahuas are taller than the Otomí and the Oaxaca Indians. Nevertheless, the mean of the entire twenty samples must represent substantially the true condition. This mean may be taken as 159 centimeters, especially in view of the fact that the Nahua average is 159.8 for seven samples. In English units, 159 centimeters equals 5 feet 2 1/2 inches. The Nahuas would average one-half inch taller than the general mean.

With one exception, the averages for body weight of Mexican Indians shown in Table 2.1 are taken from the compilation by Marshall Newman.^[33] Those for central Mexico give a mean of 53.6 kilos, and those for the Maya 53.8 kilos. In order to check the consistency of this mean, Newman's formula may be employed: the ratio of the stature to the cube root of the weight equals a value which lies between 38.0 and 43.0. There are other formulas, but none are perfect, and Newman's is close enough for ordinary purposes. For central Mexico plus Yucatán, we get 42.1, a value which lies well within the prescribed range. Our final values for central Mexico are: height 159 centimeters, weight 53.6 kilograms.

These values relate to adult males. The other half of the adult population, the female, is universally smaller. The male/female

[33] Marshall T. Newman, "Adaptations in the Physique of American Aborigines to Nutritional Factors."

ratio of size can be approximated from some of the data given by various investigators, such as are presented in Table 2.2 for height alone. Clearly there is variation because of both sampling error and differences in body build. However, the average female/male ratio is very close to 92.5 (the female/male ratio is more convenient for use than its reciprocal). A very large sample of white Americans shows 92.7;^[34] Goldstein's series of Mexicans in several age groups over nineteen years averages 92.5.^[35] In view of the uniformity found in the published accounts, the pre-Conquest Mexicans can be regarded as differing little in this respect from those of the present day. Then, if the ratio is 92.5, the height of pre-Conquest females was 159 × .925, or 147 centimeters. The approximate average weight may be obtained by taking 92.5 percent of the average male weight: 53.6 × .925, or 49.6 kilograms.

In order to estimate the basal metabolism and hence the caloric requirements of a person having the dimensions just specified, we should know the surface area and the number of calories required per square meter of surface. It is true that much present-day research stresses the importance of body composition and utilizes fat-free body weight as readily as it does surface area. However, when we are dealing with an extinct population, and since we know absolutely nothing about the body composition of the pre-Conquest Indians, surface area provides a better criterion. We shall try both methods.

Without considering surface area, it is possible to calculate the probable basal metabolism directly from weight if one follows the prescription of Max Kleiber. Kleiber asserts that the interspecific mean intrinsic metabolic rate is 70 kilocalories per day per weight in kilograms to the three-quarters power.^[36] If this formula is applied to the average Aztec or Mixtec, as may be, the result is 70 × (53.6)^3/4 , or 1,387 kilocalories per day, or 57.8 kilocalories per hour.

The original formulation of the relation between surface area and height and weight was worked out many years ago by Dubois and is shown graphically in numerous standard texts of physiology and biochemistry. It represents a broad average for

[34] Howard W. Stoudt et al., "Heights and Weights of White Americans."

[35] Reproduced in Gabriel Ward Lasker, "The Age Factor in Bodily Measurements of Adult Male and Female Mexicans," p. 57, table 2.

[36] Max Kleiber, "Body Size and Metabolic Rate," p. 512.

most of humanity, and its validity for man has never been seriously attacked. Therefore we may use the relationship thus set forth with reasonable confidence that it applied to prehistoric as well as to living populations. A chart after Dubois is shown by Ruch and Patton^[37] and by Philip Bard.^[38] The closest it can be read for 159 centimeters and 53.6 kilos is 1.54 square meters, plus or minus 0.02 square meter. Since this error is less than 1% it may be ignored.

A tabulation by Bard, based on data from Benedict and from Rubner, indicates that 41 men with an average weight of 53.4 kilos produce 914 kilocalories per square meter of body surface in 24 hours; the total heat would be 1,408 kilocalories per day.^[39] Another formulation shows, according to the Aub and Dubois standards, that a man between 18 and 30 years of age produces about 40 kilocalories per square meter of body surface per hour.^[40] Hence 40 × 24 × 1.54 = 1,478 kilocalories per day. Since these data have never been controverted in principle, we may conclude that if the pre-Conquest Mexican Indian adult men of 14 to 40 years averaged 159 centimeters and 53.6 kilos, then their mean basal metabolic rate was approximately 1,425 kilocalories per day (1,425 is the mean of 1,387, 1,408, and 1,478).

It should be emphasized that the values just given relate only to the average adult man under basal conditions. If we are concerned with populations, we must know the comparable figures for other components, specifically women and children. For adult women our estimate of body size indicates a height of 147 centimeters and a weight of 49.6 kilos. The closest reading of the Dubois graph gives a surface area of 1.41 square meters. Thus the metabolic rate of women is close to 36.5 kilocalories per square meter per hour, and the daily rate is 1,235 kilocalories.

The basal metabolism of children collectively is much more difficult to assess, for several reasons. First, the number of children varies in different populations; second, the intrinsic metabolic rate varies with age, i.e., the heat production per unit

[37] Theodore C. Ruch and Harry D. Patton, Physiology and Biophysics , p. 1045, fig. 12.

[38] Philip Bard, ed., Medical Physiology , p. 482, fig. 171.

[39] Ibid. , p. 480, table 37.

[40] Ibid. , p. 478, table 31.

weight decreases after late infancy; third, the total metabolism increases with increase in size.

With respect to the proportion of children in the population, we do not, of course, have any direct information concerning the situation prior to the Conquest. However, there is a considerable body of evidence for the colonial period, particularly the late eighteenth century. Furthermore, we know that the demographic status of the central Mexican population was much the same in both the fifteenth and the eighteenth centuries. Both must have been characterized by high birth and death rates. Both were under strong pressure toward increase. Hence the proportion of children in both must have been more or less similar. In a previous essay, our examination of data for the late eighteenth century made it clear that, in spite of numerous ethnic and geographical differences, the number of children in Mexico under the age of sixteen years approached 45% of the total population.^[41] In the absence of any contrary evidence, the same proportion reasonably may be assumed for the population in, let us say, 1518.

Since body size and metabolism both vary enormously with age, and since we have little refined data pertaining to the distribution of these magnitudes in the youthful population, some type of short-cut is mandatory. The simplest procedure is to take the features characteristic of the median age in the range 0–16 years, and let them stand as an average, although a very crude one, of all children. In a primitive population of this sort, if children, metabolically speaking, are all those under 16 years of age, the median can be estimated from the late-eighteenth-century data.^[42] The value lies somewhere near the age of 6 years.

The size and the intrinsic metabolic rate of 6-year-old children can be calculated only by the use of data derived from present-day American or European sources. When applied to pre-Conquest Mexican Indian children, the result is likely to contain a relatively large error. Hence we can arrive at only a crude approximation, although even an approximation is preferable to nothing. According to Wohl and Goodhart, 6-year-olds of the smallest size, both boys and girls, are 42.8 inches tall and have an average weight of 38.75 pounds. We select the smallest

[41] Cook and Borah, Essays , I, pp. 201–299, esp. p. 255 et seq .

[42] Ibid. , pp. 257–259.

size (5th centile) in order to approach as closely as possible the probable size of Mexican children.^[43] Then, converting units to the metric system, we have 108 centimeters in height and 17.6 kilos in weight.

We employ next the Dubois graph, with extrapolation where necessary, to estimate the body surface at 0.7 square meter. Closer estimate is not feasible. Finally, the metabolic rate is found by reference to the table in Ruch and Patton, according to which for 6-year-old boys and girls a mean of 51.7 kilocalories is produced per square meter per hour.^[44] Then the metabolic rate per child becomes 869 kilocalories per day, say 870 kilocalories. An error of plus or minus 10–20% will have to be allowed, but a working mean of 870 kilocalories per day is reasonable.

We are now in a position to make an approximate estimate of the metabolism of the population. An adult male produces per day 1,425, an adult female 1,235, and a child 870 kilocalories. According to our data already cited, there are 55% adults (27.5% adult males and the same percentage of females) and 45% children. Therefore, 1,000 people produce 275 × 1,425 plus 275 × 1,235 plus 450 × 870 kilocalories. The total is 1,123,000 kilocalories per day, or 1,123 per person. This would be the basal value, the minimal amount needed to maintain organic integrity, with no excess even for the easiest kind of physical activity or for the digestion of normal food.

In order to evaluate the caloric requirement in daily life of the pre-Conquest central Mexican Indians, we have to know something about the level of their physical effort. In turn, it is necessary to equate this activity with the increase caused thereby in the metabolic rate over the basal value. We shall examine the second phase of the problem first, and do so by considering data derived from study of present-day populations.

There have been a great many investigations of the effect of various types of physical exertion upon the energy flow through the human organism. These studies have one aspect in common: they all pertain to present-day subjects who are of European-American physique and who subsist according to what are considered nutritionally adequate standards. For the central

[43] Michael G. Wohl and Robert S. Goodhart, eds., Modern Nutrition in Health and Disease: Dietotherapy , p. 10, tables 1.6 and 1.7.

[44] Ruch and Patton, p. 1045, table 3.

Mexican Indians we can make some adjustments for body size, but the nutritional aspect remains a separate problem.

We select a few of the most accessible and at the same time representative figures. The activity data refer to moderately large, well-conditioned North Americans or Europeans. The basis of at least some estimates is a 70-kilo man who is approximately 170 centimeters tall (154 pounds and 5 feet 7 inches). Such an individual has a surface area of 1.8 square meters. His basal metabolism would be 40 × 1.8 × 24 = 1,728 kilocalories a day, or in a round figure 1,730. If we use the hourly rate, we have a basal value of 72 kilocalories.

We now present two simple lists. Each shows the extra energy expended, presumably by a 70-kilo man, in performing certain tasks. The values per square meter have been converted to those for the whole person by using the multiplication factor 1.8. The units shown are kilocalories per hour. The lists are modified from the sources indicated, the figures in parentheses representing a further modification to arrive at values for an average central Mexican Indian adult male.

Sherman also shows a balance sheet in which is set forth a hypothetical 24-hour day, divided into portions according to activity.^[47] The basal metabolism is included in these figures:

[45] Henry C. Sherman, Chemistry of Food and Nutrition , p. 189, table 23.

[46] Ruch and Patton, p. 1045, table 2.

[47] Sherman, p. 190.

The activity load of the pre-Conquest central Mexican Indians is difficult to estimate. Initially, however, there must be established a reduction factor. A group which averages small size is sure to expend less energy than one of large dimensions. While we may know the comparative sizes, we are not sure of the exact relationship between the two groups in energy expenditure. On the other hand, some sensible assumption is required, and the simplest is that of direct proportionality. Therefore, if the mean basal rates in the two populations are in the proportion of 1,780 to 1,425, the values for the activity levels given above should be reduced to 82.4 percent. In the lists, the values thus adjusted for the central Mexican Indians are placed in parentheses. For example, the kilocalories used per day are reduced from 3,380 to 2,785.

The labor schedule of the overwhelming majority of pre-Conquest Mexican Indians was based upon farming, an activity conducted without aid from domesticated animals, by means of the digging stick as the principal implement. The work must be regarded as moderate, although numerous individual tasks were performed which might be classed as heavy. Relatively little work was done as strenuous as our really intensive manual labor displayed in both agriculture and industry. The women, of course, assumed all domestic duties, and their level of labor would have resembled what Ruch and Patton had in mind when they mentioned housework. Labor at home and in public service for the men, associated with occupations such as farming or building, would have been more onerous, but hardly could have surpassed the demands of steady bicycling for an equal amount of time. If so, the extra energy cost above the basal level would not have exceeded 155–160 kilocalories per hour. The same approximate value can be put on the energy cost of carrying loads, both domestically and over long distances, for the human cargador performed the function of the beast of burden used in the Old World. These cargadores carried a load of 50 pounds

each and traveled approximately one league an hour, or 2.5 to 3.0 miles depending upon the terrain. This activity, disregarding the loads, corresponds quite closely to Sherman's slow walk, which consumes an extra 106 kilocalories an hour. With the load, the value would certainly rise to 155–160 kilocalories.

Thus we formulate a workload estimate of 155–160 kilocalories per hour per man (for the sake of convenience, we shall use the value 156.7), and 70 kilocalories per day per woman. For children the value would be much smaller, but even the youngest children did a little work beyond their normal activity in helping their parents support the domestic establishment. On the other hand, it is extremely difficult to segregate work from play, so that it is preferable to reconsider children after a discussion of some other matters.

Another factor of importance is the allocation and distribution of time. In the instance of cargadores, we have the explicit testimony of Bernal Diaz, reporting on the situation in 1519 before the Europeans had had a chance to alter it, that a day's march was five leagues, or five hours' travel under load.^[48] The workday for the bearers must have included the time needed to make and break camp and to mount and dismantle the loads, perhaps an additional hour in all. The working day, then, was six hours. This is just about what the present-day Mexican farmer puts in, if one counts sustained work. Indeed, the six-hour day is prevalent throughout Latin America. It is a response to dietary limitations and also an adaptation to the climate. The serious working day extends from 6:00 or 7:00 a.m. to 12:00 or 1:00 p.m., at which time the heat in the tropical areas becomes unbearable. In many instances, although not always, a little more work may be done in the late afternoon, but approaching darkness precludes any extended effort. The rest of the time is spent eating, sleeping, and relaxing. This regimen is of very ancient origin and conforms to the exigencies of the region.^[49]

Let us now reconstruct a hypothetical day for an Aztec

[48] See the explicit statements of Bernal Díaz del Castillo (I, pp. 146 and 177): " . . . indios de carga, que en aquellas partes llaman tamemes , que llevan dos arrobas de peso a cuestas y caminan con ellas cinco leguas." The league was one hour's walk, the distance traveled varying with the difficulty of the terrain.

[49] If one allows for coffee breaks, visits to the toilet, and other interruptions in sustained work, the regime may not be much different from that in industrialized societies.

macehual , farmer, or cargador. He gets about eight hours sleep at night and one hour at noon. He works reasonably hard for six hours. He divides the remaining nine hours such that he indulges in light exercise for three hours and sits quietly the other six hours. Alternately, we may say that during these nine hours he engages in light activity. In either case, we may allow him an average extra expenditure of 37 kilocalories per hour. Then his formula is as follows:

Further adjustments are in order, the chief one of which relates to the number of days annually during which a full workload was carried. Throughout the year a proportion of days would be devoted to rest from all types of labor possible; that they came in our frequency of one in seven (now two in seven) seems unlikely; rather, they undoubtedly came in a relation to the Mesoamerican calendar with its 20-day units. Further, the accounts of Sahagún and those of other writers, native and Spanish, make it clear that the pre-Conquest Mexican Indians enjoyed a great many religious celebrations and festivals throughout the year.^[50] Some of these were confined to the ruling classes, priests, and nobility, but the common people participated in many of them. Merely as an estimate, let us say that days of rest and days of celebration and festival for the peasants gave them about 75 work-free days. If the work is deleted from the schedule shown above for 75 days in the year, and light activity is substituted, the value 940 is changed to 222 for the 75 days. Hence on these days the energy output was 2,067 kilocalories instead of 2,785. The annual output was reduced from 2,785 × 365 = 1,016,525 kilocalories to 75 × 2,067 + 290 × 2,785 = 962,675 kilocalories per year. This is equivalent to 2,637 kilocalories per day, say 2,635.

Let us turn now to the women. We know that they assisted

[50] See Sahagún, I, pp. 93–281; Motolinía, Historia , pp. 35–71.

to a significant extent in the farm work in addition to their work in the home. The extent of this extra labor cannot be precisely ascertained. However, we know that it could not have equaled that of the men, because so much time had to be spent in housekeeping, child care, and other necessary duties. On the other hand, the amount of farm labor must have been appreciable. It will be a moderate supposition to allocate to the adult female two hours of the six hours of labor on 290 days of the year. In kilocalories, the female balance sheet for a working day therefore looks as follows, where the cost of all types of work and activity is taken as having the same values as for men:

If we allow the same 75 days of rest and fiesta as for the men, we get per year for a female adult 2,155 × 290 + 1,915 × 75 = 768,575 kilocalories. This means 2,106 kilocalories per day, say 2,105.

We may now revert to the problem of the children. These young individuals cannot be said to have undergone labor in the strict sense. They performed many tasks, some of them onerous, but on the whole their activity must be classed as light. On the other hand, particularly with very young children, the hours of sleep must have been extended. In order to take these variables into account to some degree, although strict accuracy is impossible, we shall consider that the children slept ten hours (including a siesta) and spent their waking hours in light activity.

The basal metabolism of a child at the median age of 6 years was estimated to be 870 kilocalories per day, or 36.3 per hour. The sleeping metabolism is roughly 10% below the basal, or close to 33 kilocalories per hour and 330 per day. The basal value for 14 hours of waking is 508 kilocalories, rounded off to 510. The cost of digestion, because of the smaller size, may be

reduced to 80 kilocalories. The extra energy cost for 14 hours light activity in a child may be considered as lying between one-half and three-quarters of the value for an adult, or approximately 25 kilocalories per hour. This is admittedly a very crude estimate, but in view of the great range in age and size must represent an approach to the actual value. Thus we add to the other items 14 × 25 kilocalories per day, or 350, making a total energy output of 1,270 kilocalories per day. Since childhood activities will continue on nearly the same level, regardless of social events, no deduction need be made for days of rest or fiesta.

We now have a general average of 2,635 kilocalories for men, 2,105 for women, and 1,270 for children. If we apply the same proportions of these three components in the population as previously (27.5, 27.5, and 45.0% respectively) the average energy production per person per day becomes 1,875 kilocalories, a value we round off to 1,900 kilocalories. This value takes into account body size, age, and physical activity. It also assumes a diet adequate to supply the necessary calories. It does not take into account protein, accessory substances, or other factors which may affect the energy yield of the food eaten. Perhaps the most important point for our discussion is that it does not take into account circumstances that might reduce the caloric allowance available to the population, which might live not at a level of adequate nutrition but rather at one of semistarvation. This matter of possible semistarvation requires further exploration in our discussion.

The data concerning metabolism we have discussed up to now are based upon contemporary European and American dietary standards. They assume, unless otherwise indicated, that what is today considered an adequate diet was available to all persons, including Mexican and South American inhabitants, and that both the basal and the activity rates were determined only by such factors as age and body size. There is room, and indeed evidence, for doubting that this assumption is valid and warranted.

The level of nutrition today among the various groups in the Mexican population, both urban and rural, is the subject of acrimonious comment by many medical authorities and sociologists. They contend that a large segment of the population

suffers from chronic undernutrition.^[51] If this is true at present, and there is considerable evidence for the view, then it has been true throughout the history of the region. Consequently, the estimates offered here for mean energy production are too high and do not correspond to a human aggregate which was underfed. At this point we are interested primarily in the effect of inadequate caloric intake upon the mean output. By inadequate, we have in mind any level between that which is agreed upon as appropriate to maintain a population at full working efficiency and that which produces severe, manifest clinical symptoms such as are associated with outright chronic starvation.

The classical study involving undernutrition in great masses of people was carried out by Alexis Ivanovsky in Russia and made available to the western world in 1923.^[52] Ivanovsky followed the condition of over 2,000 adults for three years during a rigorous famine, and found an average reduction in height of 4.7 centimeters in males and 3.5 centimeters in females. The weight diminished to a varying degree, but the range was approximately from 20 to 60% of the initial value. He also found that among those who were exposed to inanition for long periods, the loss in size took place in the first year. Thereafter there was relatively little change.

A more rigidly controlled experiment was that of Ancel Keys and his colleagues at the University of Minnesota in the late 1940's, after the experiences of the Second World War had stimulated interest in the matter.^[53] In the Minnesota studies, a rapid initial loss of weight was observed; it was followed by a slower reduction until equilibrium was reached. The basal metabolism of the subjects also fell to a new low level. The changes are described by Keys and Francisco Grande:

Equally important is the fact that, given time, the body weight tends to

[51] See, for example, Richmond K. Anderson et al., "A Study of the Nutritional Status and Food Habits of Otomí Indians in the Mezquital Valley of Mexico"; Mendizábal, "Evolución económica y social," Obras Completas , VI, tables on food consumption between pp. 192–193; Salvador Zubirán and Adolfo Chávez V., "Algunos datos sobre la situación nutricional en México"; Carlos Pérez Hidalgo et al., "Recopilación sobre el consumo de nutritivos en diferentes zonas de México"; Ana María Flores, La magnitud del hambre en México, passim , esp. pp. 9–25; and Whetten, Rural Mexico , pp. 304–316.

[52] Alexis Ivanovsky, "Physical Modifications of the Population of Russia Under Famine."

[53] Ancel Keys and his colleagues have given a full report in the two volumes of The Biology of Human Starvation .

reach a steady state and calorie expenditure tends to balance calorie intake, no matter what the level of the latter may be. When we changed the diet of young men from 3,500 to 1,500 Calories daily, the weight loss was rapid at first and decreased exponentially with time until calorie equilibrium was achieved with the body weight being 25% less than it had been.^[54]

A similar result had been obtained in the well-known experiment of Benedict, Miles, Roth, and Smith (1919). These investigators kept a squad of young men on a low-calorie diet, an average of 1,930 kilocalories per day per man, after they had become accustomed to a diet consisting of 3,000 kilocalories per day. During the last three days of the restricted ration, the average net kilocalories produced per day was 2,245 and there was marked reduction in both the basal rate and the rate during moderate exercise. The subjects, however, felt no ill effects and were able to perform their tasks with facility. Sherman described the state of all these subjects as follows:

When conditions are otherwise favorable, healthy young men can adjust themselves to lowered energy intakes through reduced body weight and lowered BMR so that they can get along with 1/3 less food calories and still feel and act normal.^[55]

It is thus generally conceded that the food intake can be diminished from the high level associated with what are considered adequate diets with no permanent ill effects. There will be reduced weight, even reduced height, and at the same time lowered energy costs for all phases of life. If the reduction of intake is not too severe, the individual goes into and remains in what has been called the compensated phase of undernutrition. No clinical changes are observed except those mentioned; the person can continue a quite normal daily life.^[56]

We suggest that such a condition was present in the population of central Mexico on the eve of the Conquest, the cause being a low caloric intake. Such testimony as has come down to us all points in this one direction. The Relaciones Geográficas in many instances emphasize the frugality of diet for commoners

[54] Ancel Keys and Francisco Grande, "Body Weight, Body Composition, and Calorie Status," in Wohl and Goodhart, p. 24.

[55] Sherman, Chemistry of Food and Nutrition , p. 195.

[56] In addition to works already cited, see Francisco Gomez Mont, "Under-nutrition," in Wohl and Goodhart, pp. 984–995, esp. p. 988.

before the coming of the Spaniards. In the report for Cuilapan, written by Fray Agustín de Salazar, the comment is especially striking:

El comer de ellos es grima y espanto porque con unas tortillas de maiz y poco de agi y otras cosillas se contentan . . .

(Their food is a matter of disgust and horror, for with a few maize tortillas, a little chile, and other trifles, they are satisfied . . . )^[57]

Fray Agustín de Salazar was writing about his own time in 1581, but the Indian diet he described was that of the pre-Conquest as well. In Cuilapan it had not changed, much less improved.

In the history of Gonzalo Fernández de Oviedo y Valdés, who was never himself in central Mexico but questioned Spaniards who had been, including men of Cortés' army, is a summary of early testimony. According to Oviedo, the Indians of New Spain were the poorest of the many peoples who up to his time had been encountered in the New World. Their diet for the most part was maize and vegetables flavored with chile. The quantity was little, as he explains carefully, "not because they would not eat more if they had the food," but because the upper classes carefully assessed harvests and left them the bare minimum to sustain life and work until the next harvest.^[58] Oviedo's description of poverty is corroborated by Motolinía, who spent many years in New Spain:

Estos Indios cuasi no tienen estorbo que les impida para ganar el cielo . . . porque su vida se contenta con muy poco, y tan poco, que apenas tienen con qué se vestir y alimentar. Su comida es muy paupérrima, y lo mismo es el vestido . . .

(These Indians have almost no hindrance that might keep them from earning entrance to Heaven . . . for in life they are satisfied with very little, so little that they scarcely have the wherewithal to clothe and feed themselves. They eat most poorly indeed and clothe themselves in equal poverty . . . )^[59]

The Anonymous Conqueror summed up the matter tersely and elegantly:

 . . . es gente que se mantiene con poco alimento. (They live on little food.)^[60]

[57] In Tlalocan , II, p. 25.

[58] Oviedo y Valdés, Historia general , IV, pp. 248–250. See also [Anonymous Conqueror], Relación , p. 50.

[59] Motolinía, Historia , p. 85.

[60] [Anonymous Conqueror], pp. 41 and 50.

There is no certainty, however, that the level of food intake fell much below the compensated phase of undernutrition, for, irrespective of areas of seriously deficient diet today, no one has adduced evidence to show widespread clinical manifestations of starvation among the natives before Cortés. If the farmers and laborers, who constituted the overwhelming bulk of the population, were just below or even definitely below the present-day standard of adequate nutrition, then the mean daily rate of energy production fell below the 1,900 kilocalories we have calculated above. How far cannot be determined with any exactness. Sherman's level of a one-third reduction, which would mean here a reduction to 1,265 kilocalories, seems extreme. We shall be safer if we estimate the level as lying between 1,400 and 1,800 kilocalories. Clearly the marginal diet of the commoners would support a considerable amount of physical effort, but distinctly less than that possible to the better-fed and larger European or American laborer of today. The subliminal energy intake could be compensated by reducing physical effort. That there was malnutrition as against undernutrition seems unlikely, because the wide diversity of foodstuffs that were eaten would supply needed elements. The requirement for protein and amino acids would be met largely from maize, beans, and other plant components in the diet. Any requirement for protein of animal origin would be small, on the order of perhaps 15 to 30 grams daily for an adult male, and was easily met by eating insects, grubs, snakes, amphibians, birds, small mammals, and indeed "any living thing."^[61]

Our discussion has now arrived at the point at which we may make some examination of the probable amount of maize

[61] Anthony A. Albanese and Louise A. Orto, "The Proteins and Amino Acids," in Wohl and Goodhart, pp. 95–155, esp. pp. 97, 108, and 113–116; National Academy of Sciences—National Research Council, Evaluation of Protein Nutrition, passim . Anderson et al., in their "Study of the Nutritional Status and Food Habits of Otomí Indians in the Mezquital Valley of Mexico," reach the conclusion that: "The region is arid and barren and, economically and culturally, one of the most depressed in the country. The inhabitants eat very few of the foods which are commonly considered as essential to a good nutrition pattern. Their consumption of meat, dairy products, fruits, and vegetables is exceedingly low. However, through the eating of tortillas, the drinking of pulque (the fermented unfiltered juice of the century plant), and the eating of every conceivably edible plant available, a fairly good diet is maintained." Guillermo Bonfil Batalla, in Diagnóstico sobre el hambre en Sudzal, Yucatán (pp. 133–135), indicates a daily consumption as typical not merely of Sudzal but also of "more than twenty communities all over the country" that would come to approximately 1,808 kilocalories daily, 1,466 coming from the ingestion of 419 grams of maize, approximately 1/2 liter.

consumed daily by an Indian peasant and of probable production in pre-Conquest central Mexico. Obviously anything we indicate here must be highly tentative. We have now indicated that the overwhelming majority of the Indians lived at a level of compensated undernutrition. Accordingly, our previously published estimate of probable daily consumption is too high. That estimate was based upon the allowance prescribed by the viceroy in 1555 for Indians conscripted for repair of the dikes of the lakes of the Valley of Mexico, and presumably reflected a customary ration. The allowance was one cuartillo of maize a day.^[62] At 48 cuartillos in a fanega of 100 Castilian pounds (of 460 grams each, as against our pound of 453.6 grams), one cuartillo would be 958 grams of maize with a caloric value of approximately 3,350 kilocalories.^[63] That is a generous ration, and when supplemented by other foodstuffs provided by the Indians themselves might reach a value of 3,800 or 4,000 kilocalories.

It is possible that the ration, even at that relatively early date after the Conquest, already corresponded to European ideas, which were more generous; but what evidence we have on European rations for men engaged in fairly strenuous physical activity would not come to so high a value.^[64] Since the maize would be ground to meal and baked into tortillas for consumption, a service performed by women, we are inclined to suggest that the ration was for a man and his wife. Divided between them, it would come to the caloric values that we have indicated as probable. We shall return to these questions later in this essay. For the moment it is enough to suggest that the more

[62] Borah and Cook, The Aboriginal Population of Central Mexico on the Eve of the Spanish Conquest , p. 90; but see the entire discussion, pp. 89–92.

[63] We calculate 100 grams of maize as having 350 kilocalories. The value is a rough average of those given for various kinds of maize in Mercedes Hernández et al., Valor nutritivo de los alimentos mexicanos; Tablas de uso práctico , p. 6, and Juan Roca and Roberto Llamas, "Régimen alimenticio de los habitantes de la región de Izúcar de Matamoros (Puebla)," p. 584. For the maize of the sixteenth century, Roca and Llamas are probably nearer the true value at 341.81 kilocalories, since Hernández et al. deal with foods already greatly affected by the improvements of the Green Revolution—but the range from 341.81 to 366 is not great.

[64] See the references in note 6; additionally, Cesáreo Fernández Duro, La armada invencible , I, pp. 248–320, esp. pp. 274–278; and Maria Cristina Silveira and Carlos Silveira, "A alimentação na 'Armada Invencível.'" The Silveiras base their calculations on a manuscript prepared by the Marqués de Santa Cruz, proposing supplies and rations for the Great Armada, the copy they consulted being in the Biblioteca Nacional, Lisbon. Fernández Duro publishes the manuscript.

normal allowance for an adult male commoner was closer to a range from one-third to one-half of a cuartillo of maize (319–479 grams), say from 1,120 to 1,676 kilocalories, the remainder of the day's consumption of foodstuffs and perhaps a far greater proportion of vitamins, proteins, and trace elements coming from fruits, nuts, frijoles, chile, pulque, small amounts of animal substance, etc. The amount of maize annually consumed by a peasant family might range from 10 to 20 fanegas, as suggested for the post-Conquest period in the Valley of Mexico by Charles Gibson.^[65] For a family of average size before the Conquest, the likelier range would be 10 to 15 fanegas except in unusually favorable circumstances. The maize would be supplemented by a remarkable range of other foodstuffs.

The productivity of the pre-Conquest agriculture that provided this food is also difficult to estimate, since we have merely two sets of clues at this time and they are not easily reconciled. One set lies in the determinations of yields by Spanish tribute assessors in arriving at amounts and value of grain for tribute requirements set in plantings. Since such determinations occurred in the decades immediately succeeding the Conquest, we may assume that they reflect conditions substantially unchanged from aboriginal conditions. On an average, the determinations show a yield of 6.47 fanegas of maize for a planting of 1,000 square brazas,^[66] or a harvest on the order of 1,060 kilograms of maize per hectare.^[67] That average, which includes crops from irrigated and unirrigated land and from elevations from coast to plateau, is distinctly higher than the reported averages of Mexican agriculture just before the advent of the Green Revolution. We cannot exclude the possibility, therefore, that tribute maize was grown on the best land available to each town and was not representative of other yields.

[65] Gibson, The Aztecs Under Spanish Rule , p. 311.

[66] Cook and Borah, The Indian Population of Central Mexico, 1531–1610 , p. 19.

[67] Conversion of fanegas to kilograms is simple, since 1 fanega equals 46 kilograms. The braza is calculated at 1.6718 meters, the value given by Spain, Real Academia Española, Diccionario de la lengua española , from which we have taken the value for fanega. Sixteenth-century maize at this weight converts to dry measure at .83 kilograms per liter. These yields, which would work out to 14–16 bushels per acre, fall easily within the normal ranges indicated by Charles Gibson (The Aztecs Under Spanish Rule , pp. 309–310) for the Valley of Mexico in colonial times.

Our second set of clues on the probable yields of pre-Conquest agriculture is, of course, Mexican experience as recorded just before the remarkable changes that began in the 1940's and 1950's. In the first agricultural census of 1930, recording the agricultural year 1929–30, the range of yield for the states of central Mexico is enormous, from 472 to 1,237 kilograms of maize per hectare, and the returns subject to all the suspicions hovering around a new kind of statistical inquiry and one dependent upon thousands of respondents. The average for the states within the sedentary area of central Mexico as it was in 1520 is 633 kilograms per hectare, as against a countrywide average of 522. The countrywide average is for ejidos alone, 585, and for the private sector alone, 512. Productive units of one hectare and under were not asked to report.^[68] Since one would normally expect the private sector to be more productive, we are inclined to suspect underreporting on a considerable scale. The 1940 agricultural census, reporting the 1939–40 agricultural year, has more favorable results: The area of central Mexico gave average maize yields of 679 kilograms per hectare, the returns being divided between productive units of more than 5 hectares, with an average of 665 kilograms, and those of 5 hectares and less, with an average of 699 kilograms per hectare.^[69] So the two censuses, whatever the defects in them, indicate yields of maize for central Mexico ranging from 630 to 700 kilograms per hectare.^[70]

[68] Mexico, Dirección General de Estadística, Primer censo agrícola-ganadero, 1930. Resumen general, cuadro VIIIA, pp. 70–73. We have omitted Nayarit, Jalisco, Guanajuato, and Querétaro in our reconstitution of what was central Mexico in the sixteenth century, since they were then either sparsely settled or under the control of nomadic Indians. The 1930 census returns show an even greater range than the overall state averages indicate, since the average for ejidos in Puebla is 412 kilograms of maize per hectare, as against 1,424 kilograms for the Federal District.

[69] Calculated from Mexico, Dirección General de Estadística, Segundo censo agrícola-ganadero, 1940. Resumen general, cuadros 2 and 4, pp. 54–186 and 201–242.

[70] See also the annual averages of yields of maize per hectare from 1925 to 1937 in Mexico, Secretaría de Agricultura y Fomento, Memoria , 1937–1938, I, pp. 302–303, which show considerable fluctuation. If one turns to the 1950 agricultural census, the average yield per hectare for all of Mexico, after deduction of the crop of hybrid maize, was 786 kilograms. Land sowed to hybrid maize yielded on average 1,621 kilograms per hectare, but in that year so little was so sown that the average yield of maize of all types for all of the country was 790 kilograms per hectare. (Mexico, Dirección General de Estadística, Censos agropecuarios. 1. Totales comparativos en 1930, 1940 y 1950. 2. Por entidades y distritos economico-agricolas en 1950 , p. 15.)

These returns may be low because of underreporting—but may be too low, further, for comparison with pre-Conquest agriculture, because of the havoc of four and a half centuries of erosion, continued cultivation of land without adequate preservation of fertility, and the adoption of the plow as against the digging stick. The plow gives higher yields per man-hour spent on agriculture, whereas the digging stick usually gives higher yields per unit of land.^[71] Accordingly, we come to a range of probability: yields of maize for aboriginal agriculture may have been somewhere from 700 to 1,200 kilograms per hectare. That range of yield would have maintained the population we have postulated at the levels of nutrition postulated through cultivation of from 10 to 15% of the land.