Reports of the

UNIVERSITY OF CALIFORNIA
ARCHAEOLOGICAL SURVEY

No.64

THE QUANTITATIVE APPROACH TO THE RELATION BETWEEN POPULATION

AND SETTLEMENT SIZE

S. F. Cook and Robert F. Heizer

University of California Archaeological Resea;rch Facility

Department of Anthropology

Berkeley

March 1965

REIPORTS OF THE

UNIVERSITY OF CALIFORNIA ARCHAEOLOGICAL SURVEY

NO. 64

THE QUANTITATIVE

APPROACH TO THE RELATION BETWEEN POPULATION

AND SETTLENENT SIZE

S. F. Cook and Robert F. Heizer

UNIVERSITY OF

CALIFORNIA ARCHAEOLOGICAL RESEARCH FACILITY

Department of Anthropology

Berkeley

March 1965

ACKNOWLEDGMENT

The authors wish to acknowledge support for research
and preparation of these papers by the National Science

Foundation (Grant G-23734). This monograph is in part a
continuation of research supported earlier by the Wenner-

Gren Foundation for Anthropological Research and the Miller
Institute for Basic Research in Science, and to these two

organizations we express appreciation for their assistance.

The files of archaeological site records of the

University of California Archaeological Research Facility
have been a main source of new data.

Dr. A. B. Elsasser, Eugene Prince, Alan Westcott, Jay
von Werlhof, and Martin Heicksen have provided welcome
assistance in collecting data.

- iii -

CONTENTS

INTRODUCTION .     .     .     .    .     a     .     .        1

I. ON THE NUMERICAL EXPRESSION OF RELATIONSHIPS

AMONG ARCHAEOLOGICAL VARIABLES     .     .     .       1
II.  A RESTUDY OF T. T. WATERMAN'S YUROK DATA  .     .        8

Estimation of number of houses           .     .       9
Calculation of settlement size                         9
III.  CONSIDERATION OF S. K.. LOTHROP'S DATA

FOR TIERRA DEL FUEGO         .     .     .            14
IV. SETTLEMENT AREA AS A CLUE TO REGIONAL

POPULATION .     .     .     .     .     .     .      17
V. HOUSE AND VILLAGE SIZES IN THE ABORIGINAL

CENTRAL VALLEY OF CALIFORNIA .     .     .     .      21
VI.  HOUSE SIZE, SETTLEMENT AREA, AND POPULATION

THROUGHOUT NORTHERN AND CENTRAL CALIFORNIA

Part 1                                              32
Part 2                                              35
Part 3                                              39

TABLES

1.  Values for Waterman's Yurok Data     .                  42
2.  Values for Lothrop's Three Groups of Sites .    .       43

Notes for Tables 3, 4, and 5   .     .     .     .      44
3.  "Canaliino" Culture Sites Mentioned by Rogers   .       45
4.  Data Concerning Villages Mentioned by Explorers  .      47

Notes to Table 4 .     .     .     .     .     .      49
5.  Summary of Data for Sixteen Channel Sites  .     .      51
6. House Sizes in Central Valley and Adjacent

Foothills  .     .     .     .     .                  52
7.  Data from Documentary Sources        .     .    .       55
8.  Detailed Data from Documentary Sources     .     .      56

References and Notes to Table 8    .     .     .      60
9. Final Estimates for Thirty Tribes     .     .    .       70

FIGURES      .     .     .     .     .     .     .     .      72

BIBLIOGRAPHY .

92

1

INTRODUCTION

The following group of short papers deal with specific topics
bearing on the broad problem of the quantitative expression of rela-

tionships between population and settlement size. The subjects treated
are diverse in detail, and no attempt is made to achieve any comprehen-

sive or unified synthesis or interpretation. At the same time the several
articles have one characteristic in common: they each require the close
examination and discussion of numerical data which therefore must be

presented as completely as possible. Otherwise conclusions and formula-
tions cannot be appreciated or justified.

I. ON THE NUMERICAL EXPRESSION OF RELATIONSHIPS

AMONG ARCHAEOLOGICAL VARIABLES

During recent years numerous investigators have explored the

possibility of demonstrating the existence of quantitative relationships
among various types of cultural attributes. In the realm of the archae-
ology and ethnology of primitive peoples attention has been directed to
the association between spatial elements, such as site area or other

ecological factors on the one hand, and population on the other. It has
become evident that discussion is desirable in order to clarify some of

the elementary principles governing the formulation of the many relation-
ships which undoubtedly exist.

If more than the vaguest affinity is present, if two parameters or
two tangible material variables are causally or even fortuitously con-
nected with each other, then it should be possible to secure the data

which would permit a clear algebraic formulation of the association. It
is not enough to assert verbally that one entity depends upon another;

the manner, direction, and extent of the dependence should be set forth
in the rigid language of numbers.

As an illustration, and in no spirit of criticism, we may refer to
the exposition of culture and population given by Hester (1962) in his

paper on the typology of New World cultures.   He shows (ibid., p. 1003) a
graph depicting cultural intensity on the ordinate and population density
on the abscissa. The curve is in the form of a distribution skewed

slightly to the right. The quantities are represented by '"low" to t`hight'
for cultural intensity, and "few" to "many" for population; The shape of

2

the curve is deduced from an hypothesis which may be, but is not proved
to be, valid. There are no actual values given to either variable so as
to demonstrate that this curve ever appears in reality. How much more
satisfying it would be if some real figures could be provided and the
hypothesis validated for at least one case.

The most cursory inspection of Hester's (1962) Figure 1 indicates
that this two-dimensional graph represents more than two variables.

Cultural intensity is stated to depend upon food supply, as does also

population. Indeed Hester describes this as an "inter-acting system."'

Furthermore he attributes the skewing of the curve to the "time element,"
thus introducing a new and wholly independent variable which, strictly
speaking, should require another dimension for its representation. In

short, Hester's Figure 1 is the visual picture of an idea, or the diagram-
matic expression of an hypothesis. It is not the mathematical formulation
of the quantitative relationship between two parameters capable of being
couched in numbers.

Another type of approach is illustrated in recent papers by Hack
(1942) and Schwartz (1956, 1963), and there are of course many others.

In each of these three papers the author plotted the relative population
of some specific archaeological area in the American Southwest against
absolute time in years. In computing the population, each author made
the assumption that the number of people at a given moment was directly

proportional to another, and measurable, parameter; for example, to total
site area, number of villages, or aggregate potsherd count. Here the
missing step is the mathematical demonstration that the relationship

between population and, say, village number actually may be expressed as
a linear function. The assumption is a quite legitimate first approxima-
tion, but the final proof would have to rest upon a test with the appro-
priate numerical data.

However, when we attempt to assemble real evidence for such tests
and such proofs, we encounter certain difficulties. These difficulties

are infrequently fully appreciated, and certainly are rarely resolved with
precision.  We shall discuss some of them, restricting the field of inter-
est primarily to that of settlement patterns, and in particular to the
relations between space and population.

1. In equating two variables one must be treated as independent,
the other as dependent upon it. Frequently there is no difficulty in
making a selection.   If we trace the trend of population in time, the

latter is independent; but if we relate population to space, which deter-
mines the other? In many situations, especially those involving floor

3

space, site size, etc., space occupied is clearly a function of the

number of occupants. Population will then be x and space y. However
contingencies may arise when an increasing number of people will be
contained by the available space and will be restricted in their

expansion. Then the ordinates would have to be reversed. The pri-

mary requisite is that the reasonable and logical formulation be adop-
ted for each set of data.  Interdependencies are to be avoided, such
as Hester's (1962) cultural intensity and population density, each of

which can be at the same time dependent upon the other. A graph relat-

ing them would have little meaning unless each could be paired separately
with a third and quite independent variable.   Thus with a random popula-
tion, if one plots for each person height against weight, a close associ-
ation will be found. But does weight depend upon height or vice versa?

There is no solution unless perhaps one plots height and weight separate-
ly against age. Then the time course may be used to study both types of
size change.

2. If any two parameters are quantitatively related, not only

must one be a function of the other, but this function, theoretically at
least, must be capable of expression by an equation, the visible presen-
tation of which is a plot or graph. To discover what is the true nature
of the function and by what equation it should be expressed, is one of

the most troublesome problems confronting any natural scientist.   In the
present context we are admittedly operating upon a very elementary,

unsophisticated level, for the data of archaeology and human ecology will
not as yet justify advanced or intricate treatment.

If we ignore, as being beyond the scope of this inquiry, trigono-
metric and hyperbolic functions, as well as those involving the second

and higher powers of x, there remain several possible types of equation
which may be applicable to area- population data.

The first is the linear function which, when the values for the
two variables are plotted directly on rectangular coordinates, gives a
straight line. The general equation is: y = a + bx. Here b is a pro-
portionality constant representing geometrically the slope of the line

produced by plotting the appropriate figures and statistically the coef-
ficient of the regression of y upon x. The constant a denotes the value
of y when x equals zero, or, on the graph, the point of intersection of
the regression line with the y-axis. It should also be noted that the
exponent of x is always 1.0.

All other functions, when the numerical values are plotted directly,
result in curves, not straight lines. If the data are to be handled by

4

even the most rudimentary statistical methods, it is desirable that they

be converted to a linear form. With biological and sociological processes
this conversion may very often be accomplished effectively by the use of
the logarithms of the numerical values instead of those values directly.

There are several modes of formulation utilizing logarithms, of which two
are of immediate interest in connection with settlement patterns.

A. We plot the logarithms of y against the direct values of x. If
the graph is then linear, the relationship may be expressed as y = abX.

Here x itself is an exponent. Using ordinary logarithms we may also write
the equation log y = log a + x(log b), or, with natural logarithms,

1lge Y = loge a + bx. The latter expression is often put in the form
y = a ebx. This is the exponential function seen and used widely in
following growth processes in organisms or, with a negative exponent,
radioactive and other types of decay.

B. We plot the logarithm of y against the logarithm of x. If the
data thus rectified show a straight line in the graph, then y = a xb, or,

in logarithmic form, log y = log a + b log x.   Here b, not x, is the expo-
nent, and the type of function is better designated logarithmic than
exponential.

The extent to which the student will pursue his inquiry into the

nature of the relationship between his variables will depend upon (1) how
exhaustive he wishes to make his treatment, and (2) how far his data will
support extreme accuracy of analysis. The first step would normally be

to plot the points directly, exponentially and logarithmically. If visible
linearity is produced exclusively by any one of these methods, his problem

is solved. Otherwise he must resort to more and more sophisticated methods.
Very frequently, however, the intrinsic variability of the data is such
that the more elaborate methods are inappropriate.

As an illustration of the material with which it is often necessary
to deal-, the data which have been published by Naroll (1961) may be cited
to show the relationship between the population and the total floor space

in primitive villages and towns.   (For simplicity we have omitted the large
Inca city of Cuzco.)  In Figures 1, 2, and 3 respectively the numerical
values, with population as x and floor space as y, have been plotted

directly, exponentially (semi-logarithmically) a;id logarithmically. Certain
features are immediately evident.

In the linear (direct) plot the points are unevenly dispersed, with
nearly half being bunched so near the intersection of the ordinates that

they cannot be individually spotted with the scale employed. It would be

5

very difficult to tell whether there is a straight-line function.   Plot-
ting log y against x produces a better vertical distribution of points

but the trend is clearly curvilinear, concave downward. Plotting log y
against log x (Fig. 3) results in a distribution which is approximately,
if not exactly, linear. The third method, then, yields the best repre-
sentation of the data, as Naroll has demonstrated in some detail.

When it is difficult to reach an unequivocal conclusion by con-
structing graphs, or when it is desired to make comparisons of several

to many sets of data derived from different sources or distinguished by
variation in some important contributing factor, it is probably best to
select a uniform mode of procedure for equating the variables. This is
true even though the method selected may not produce in every case the

most exact approach to the real function relating the variables. There
are a few simple principles which are likely to prove helpful.

When the phenomenon under consideration is a process during the

course of which some parameter changes in size or amount-that is to say
a rate-then time is the independent variable and the exponential rela-
tionship should be investigated. The increase or decrease in such

entities as number of settlements, population, or food supply frequently
produces a straight line with a semi-logarithmic plot. On the other

hand, when there is no change;' no rate, involved, and interest centers
in the association between the absolute or relative magnitudes of two

variables at a single point in time, as, for example between number of
settlements and mean food consumption per settlement, it is probable

that on the whole the logarithmic formulation is to be preferred. There
are several reasons, both theoretical and procedural, why this choice
might be made.

A. From the equations log y = log a + b log x and y = a xb

it is seen that b as the exponent of x defines the slope of the line in
the log-log plot.  It is also the coefficient of the regression of y
upon x. As such it constitutes an immediate and practical empirical

index to the relation between y and x which may be employed for compar-
ative and descriptive purposes.

B.  If the value of b is 1.0, then y = a x, and the relationship
is linear. In other words, the direct or linear function is a special

case of the logarithmic function and is included in it. In dealing with
an unknown relationship the more general solution should be sought first,
and then if the result warrants, the restricted solution investigated.

C.  The quantitative treatment of settlement patterns, populations,

6

influences of the physical environment, and social or economic attributes
of a culture almost invariably involve the assembling of an array of data
which must be subjected to some sort of statistical analysis. A basic
desideratum will be a reasonably symmetrical distribution of points for

each parameter or attribute. Nevertheless we very often encounter skewed
distributions. When this happens considerable improvement may usually be
accomplished if the data are recast as the logarithms. A good illustration
of this effect is provided by the data from Naroll (1961) shown in Figures
1 and 3.  In Figure 1 the points for both area and population are heavily
concentrated toward the region of small values, whereas in Figure 3 the
logarithms are relatively evenly distributed over the graph.

D. A corollary to the normalization of the distribution curve is
the reduction of undue weight on a few extreme points. Human societies

are likely to produce an extensive group of units all more or less similar
in magnitude, plus a few of extraordinary size.    In taking the arithmetical
mean of these the few very large units will throw the average beyond what
is really characteristic of the group. However, if we use the logarithms,
we may take their average and then reconvert this to the antilogarithm.

The geometrical mean is thereby yielded, which, with skewed distributions,
tends to represent more truly the characteristic magnitude of the feature,

For instance, let us consider Naroll's data for the population of
18 settlements.  The range is from 75 to 200,000 persons if we include

Cuzco (this town was deleted in Figs. 1-3 in order to simplify the graphs),
Of the 18 places, 16 had a population of 5, 000 or less; one had 15, 000;
and, as mentioned, Cuzco had 200,000. Omitting Cuzco, the arithmetical
mean is 2,023; including this city, the mean is 13,021, a figure which
gives a distorted picture of the general run of settlements. If we use
the logarithms and determine the geometrical means, we get respectively
populations of 655 and 900 persons. These numbers much more closely

approach the region of densest concentration of village populations, that
is the mode, than do those obtained by arithmetical average, and for many
purposes are far more satisfactory.

E. It has been mentioned that the regression coefficient b is
useful as an expression of the quantitative relationship between two
variables. The same may be said of certain other derived magnitudes,
such as the standard deviation of the mean and the standard error of

estimate.  To be sure, the numerical value of these will depend upon the
units in which the variables are presented, provided that the direct,

linear equation is employed. But in the logarithmic form the units have
no influence, and the slope of the regression line will be the same

whether, for example, we plot area in tens of square meters or thousands
of square feet.

7

3. Once the nature of the relationship is established, or indeed
before it is established, the data at the disposal of the student of

settlements and their populations almost invariably require some type of
test with respect to the degree or strength of association between the
two variables under consideration.  As a rule, when plotted the points

derived from the observed numerical values will not fall perfectly along
a line or a curve, but will be scattered more or less widely, perhaps as
a band, but frequently as a cluster or constellation.   It is necessary

then, if either a linear or curvilinear function is to be established, to
evaluate the closeness with which the points will fit any presumed line,

and to express the final judgment by means of some numerical constant, in
terms of probability. Two simple devices are available for this purpose.

The first method is to determine the conventional coefficient of

correlation, r. Here is an immediate measure of the strength of associa-
tion as estimated by the probability level corresponding to the magnitude
of r and to the number of degrees of freedom provided by the original
data.

The second device is to compute the standard error of estimate, or
the standard deviation of the values of y from the regression line. This

term is a measure of the vertical dispersion of the points, and diminishes
in magnitude the closer the points fit to the line, until, at the limit, a
perfect coincidence, it approaches zero. With archaeologic and ethno-

graphic data, however, the standard error of estimate usually assumes a
very appreciable value.

When it is attempted to compare different sets of data with respect
to goodness of fit, certain difficulties may arise. The standard error of
estimate must be expressed in the same units as are employed for plotting
y. Hence no comparison is possible unless both graphs or tabulations are
based upon identical units. For instance, the standard error of estimate
obtained with direct figures cannot be compared with that obtained from

the corresponding logarithms. There is, however, a compensating procedure.
We may calculate for a specific set of data, first the standard deviation
of the points from the mean of y, and then the standard error of estimate
(i.e. the standard deviation from the regression of y upon x). Both these
terms are expressed in the same units; that is, those employed for expres-
sing y. Consequently the ratio of the standard error of estimate to the

standard deviation from the mean eliminates the absolute units and becomes
a statement of the relative degree of dispersion of points. Since the
standard deviation from the mean must always be greater than zero, the

smaller the ratio the better the fit of the points. In this way any method
of formulating the data may be compared with any other.

8

In brief summary of the foregoing discussion, the following

precepts may be offered. If there is to be a truly quantitative anal-
ysis of space- population or other similar relationship, all pertinent
variables must be cast in absolute numbers. The reasonable and appro-
priate variable must be designated as independent. The data must be
graphed and studied with respect to the governing type of function.

The semi-log formulation may be employed when rates are involved, and
the log-log formulation for static attributes unless otherwise clearly
indicated. Strength of association may be tested by the use of the
correlation coefficient and the standard error of estimate. If any

particular situation requires exhaustive analysis, then more sophisti-
cated mathematical and statistical methods should be considered.

II. A RESTUDY OF T. T. WATERMAN'S YUROK DATA

In 1920, T. T. Waterman published a monograph entitled "Yurok
Geography" which included a series of seventeen maps of villages, of

which fifteen lay on the Klamath River and two on the outer coast. All
were drawn to scale and showed every known structure and residual pit
in their correct relative positions. Thirty years later Cook and

Treganza (1950) made an analysis of the data for site area and population
of the Yurok by arriving at a rough estimate of the areas shown on

Waterman's maps as covered by houses, and considering the population to
be represented by the total number of house pits plus standing houses.

A high correlation was demonstrated when the logarithms of these magni-
tudes were plotted, and a logarithmic function appeared to exist. More
recently results of other investigators (in particular Naroll, 1961)

seem to have confirmed the presence of a logarithmic relationship between
population and living area, and have opened up a new field of exploration
in the quantitative association of house and site size with respect to
number of occupants. In view of these developments a new and closer
study of the Yurok data will be of value.

Waterman represented the houses on his maps by rectangles placed

according to the location of structures existing in 1920 and of pits left by
former houses.  He also included sweat houses, together with the principal
features of physiographic importance such as rivers, creeks, and contour
lines. It is thus relatively easy to mark off desired areas and measure

them.  The houses on the maps, of course, may be counted, and in addition
Waterman (1920:206-207) lists his estimate, derived from informants, of
the number of persons each village contained aboriginally. Since the

Yurok dwellings were uniformly inhabited by a single family, the population

9

may be computed by multiplying the number of dwellings by the probable
family or household size.

The number of houses may be estimated by three methods (shown in
Table 1 as methods 1, 2, and 3 respectively). The first is to count all
dwellings shown on the maps, whether inhabited in 1920 or present only
as a pit. The second is to utilize the pits only, on two grounds:

(a) that the modern structures may not have been in existence prior to
1850; and (b) that since most of the data from archaeological sites in
California record house pits exclusively, consistency in this respect
may be desirable. The third method is to employ Waterman's estimates
(2p. cit.) for aboriginal house counts as given in the table in his
monograph.

Settlement size may be calculated in several different ways, in
large measure depending upon the assumption one wishes to make concern-
ing just what constituted the space under consideration. It is clear
that we are not dealing with floor space, or aggregate space under a
roof, for this would have to be a linear function of the number of

houses, but with what we may call site space or living space. This is
the total area of the settlement, including the aggregate floor space.
The demarcation of site space is necessarily somewhat arbitrary. How-
ever, for the purpose of testing the Yurok data, six assumptions have
been made; that is six methods have been devised for computing area,

none of which perhaps will satisfy all possible ethnographic criteria.

1. A rough approximation is reached by enclosing the cluster of
houses on each map by the most appropriate geometrical figure, such as
a triangle, rectangle, ellipse, or a combination of these. From the

linear scale of feet always provided the area may be derived by formula.
This is the method used by Cook and Treganza (1950) in their original
study.

2. A minimum living space would be represented by the houses and
the land enclosed by them. Consequently the external corner or wall of
each house on the periphery of a group, including sweat houses, is con-
nected by a straight line with the one next to it, thus forming an

irregular figure which closely embraces the group of houses as a whole.

Some slight latitude has to be allowed occasionally in determining which
are the peripheral houses, but a reasonably good fit may be secured. The
area outlined is then traced off and measured with a polar planimeter.

3. On the assumption that in reality no occupant of a peripheral
house is going to be confined within the limits established by the

10

preceding method, we extend the border of the settlement. Again this
must be arbitrary; hence we take a line paralleling the first but 25
feet outward from it.

4. Using the same assumption, we extend the limit to a line

drawn 100 feet outward. This distance might perhaps satisfy most day

to day local, domestic requirements. Nevertheless more distant activi-
ties might need to be accounted for.   In general, to estimate the space
demanded by these may be not only difficult but impossible. In this

particular context, however, it is feasible to make a partial adjustment.
The fact has been mentioned that fifteen of the Yurok villages front on
the Klamath River and two on the ocean. In all instances the village is
oriented strictly toward the river bank or ocean shore, due of course to
the heavy dependence upon aquatic sources of food. Consequently it is

reasonable to extend the site area from the house cluster to the water's
edge.

5. The site is extended as suggested above, by projecting the

area produced by method 3 to the shore.    Parallel lines are drawn tangent
to the outer limits of the figure and approximately perpendicular to the
margin of the river or ocean. Thus to the space included by method 3 is
added the area lying between it and the water.

6. The site is extended as in method 5, but the figure produced
by method 4 (a line 100 ft. beyond the houses) is used as the point of
departure.

In order to study the relationship between population (houses)
and living space (area) we may plot graphs with the two parameters as

coordinates (see Fig. 4). Using the logarithms of the individual values,
we find that the points uniformly group themselves in a linear fashion,
thus confirming the earlier findings of Cook and Treganza (1950). How-

ever it is probably more enlightening to derive certain numerical constants
which are subject to interpretation and can be used for making comparisons.

If the log plot of population and area is linear, then the function
follows the equation:

log y = log a - b log x     or    y   a xb

Here are two constants, a and , which can have significance for an ethno-
graphic or archaeologic study. We may add a third: the correlation
coefficient r (see p. 7).

Considering the plotted figure as a statistical expression, it is

11

evident that we have the regression of the log of one variable upon the
log of the other. Then b is simply the regression coefficient which

characterizes the slope of the line, and a is the reading on the y-axis
of the intersection with the regression line.

The importance of r is that it gives us an estimate of the strength,
or degree of association, between the two variables-population and living
space.  If r is small, then the association is poor, and even if present is
obscured by disturbing factors.  If r is large, then the association is

good, and its validity can be tested by the probability level of r for the
appropriate number of degrees of freedom.   Thus, to use an illustration

unrelated to the Yurok, Naroll (1961) tabulated floor space against popu-
lation for eighteen widely separated settlements and obtained an r of
+0.88. This he considered, correctly in our opinion, established the
interrelationship firmly.

One further problem requires comment. Since we have two variables,
population and area, one of them must be treated as dependent upon the
other; that is, one must be represented by x and the other by y. The
correlation will not be affected by the decision, but the values of j

the regression coefficient, and of a, will be quite different, depending
upon which parameter is selected as the independent variable. In human
terms, does living space or settlement area depend upon population, or

does the population depend upon the area? Mathematically either could be
expressed as a function of the other. Cook and Treganza (1950:232, fig.
1) used area as x. Naroll (1961) used population as x. With small

settlement societies such as that of the Yurok and most other California
tribes, the latter is probably the more satisfactory solution, since we

may prefer to believe that the number of people determines the space they
occupy rather than the converse. On the other hand, in very densely

populated regions many authorities feel that living space does influence
the number of people. In the present study we may by-pass this contro-
versy and show the data according to both methods of formulation.

Table 1 gives the values of r, b, and a for Waterman's Yurok

villages in pursuance of the three methods described for estimating house
number and the six methods for computing area. There is added one esti-
mate, using population method I and area method 2, but omitting the two
seacoast villages. Moreover., in each case the constants are given when
using population and area alternately as the independent variable. The
findings may be discussed briefly.

1. All nine correlations show high values for r. The lowest is

+0.751. For 16 degrees of freedom a value for r of +0.590 is at the one

12

per cent level of probability. Hence any greater value is beyond the
one per cent level and must be considered highly significant.

2. Varying neither the methods of counting the houses nor of

calculating site area appears to alter the validity of the association
between the two parameters. This result implies a very fundamental

relationship. It also suggests that an investigator has considerable

scope in selecting his method for expressing and computing both popula-

tion and area as long as he is internally consistent. Theory as to what
constitutes living space becomes a secondary consideration, provided
that some rational hypothesis be adopted and adhered to.

3. Although the presence of a high correlation is extremely

stable, the value of the regression coefficient b is very labile.   We
may disregard the systematic difference introduced by reversing the

ordinates with each combination of population and area method, and con-
sider the values for those calculations only where population is the
independent variable.

Changing the method of estimating house number shifts the value
of b between +1.393 and +1.023. The significance of the change is not
readily apparent. Changing the method of estimating area moves b from
+1.593 to a minimum of +0.771. Here it is to be noted that the five

methods reported (2 to 6 inclusive) embrace different average areas for
the villages.

No. of       Area       Mean area     Value of
villages     method      (sq. ft.)         b

17          2           25,450       +1.593
17          3          47,400        +1.167
17          4          131,200       +0.771
17          5          103,000       +1.117
17          6          187,600       +0.850
15          2           28,400       +1.197

The figures are not notably precise, yet there appears an unequiv-
ocal tendency for b to diminish as average area increases. It must

therefore be concluded that the extent of the area per settlement which
is yielded by whatever method of estimate is selected will have a marked
influence upon the slope of the regression line; that is to say, the

numerical value of b. This conclusion also means that the way in which
settlement area changes to correspond with changing population depends

upon the specifications according to which the nature of the area itself
is established.

13

The regression constant is of course also sensitive to individual
points in the log plot. This feature is seen when two villages are

deleted, leaving fifteen instead of seventeen (see Table 1). One of the
omitted sites is Tsurai, the point for which falls close to the center

group on the graph (see Fig. 4); the other is Omen which has the smallest
area of any of the villages. Removing this latter single place radically
shifts the trend of the points and therefore the value of b.

We must conclude that the constant b should be regarded with a

great deal of caution and not be employed for comparisons of regions or

groups of settlements unless all known sources of variation are carefully
controlled.  On the other hand, with the accumulation of knowledge, we

may eventually be able to utilize the magnitude of the regression coeffi-
cient as a test for the theoretical interpretation of empirical relation-
ships.

4. The constant a represents the point where the regression line
crosses the y or vertical axis; that is to say the value of y when x

equals zero. Otherwise stated, this means that when the population is
the independent variable a describes the limiting, minimal size of the
living or settlement area-the smallest area on which there can be any

inhabitants at all. This value, as seen in Table 1, is always positive,
and might give some clue to living conditions in the region. When the
area is taken as the independent variable, then a would represent the
population when there was no area, a manifest absurdity. Here a is
always negative, and may be 'disregarded.

5. The constant a, as shown in the seventh column of Table 1, is
a logarithm. This may be converted into square feet as such (see eighth

column of table) and may give an idea of the smallest possible settlement.
An additional characteristic of the system is that when a, as the loga-
rithm, is plotted against the logarithm of the mean area for the entire
group of settlements, as computed by six methods, a straight line is

obtained. Here again is a demonstration of logarithmic function (see
Fig. 5). The result is reasonable in that one would expect the calcu-

lated minimal area which could support a population to be related to the

average area occupied by actual populations under the existing conditions.
Furthermore, it is possible that if we knew both b and a for a particular
region, we might be able to estimate the magnitude of individual habita-
tion groups with some degree of success, for these two constants define
the position of the regression line and hence permit interpolation of

specific values. Naturally the reliability and accuracy of the interpo-

lation is measured by the probability level of the corresponding correla-
tion coefficient, together with the standard deviation of the individual
points from the regression.

14

6. The absolute values of the limiting areas expressed as square
feet are of interest. With the first three cases, where the area method
is the same (method 1) and the house counts do not differ greatly (199
with method 1; 165 with method 2; 177 with method 3), the difference in

a  ranging from 1,940 to 5,575 square feet, is due mainly to a correspond-
ing variation in b. The steeper the slope of the regression line, the

lower the point of its intersection with the y-axis.    In the second group,
where the house number is constant and the area method varies, it is

instructive to compare the values derived from a with probable minimum
values estimated from other considerations.

a in sq. ft.      Area in sq. ft. estimated as stated

650        Area of a house with a 12 ft. radius:    452

3,220        Area covered by a circle of 25 ft. radius:    1,965

22,200        Area covered by a circle of 100 ft. radius.    31,419

The correspondence is far from perfect, but the correct order of
magnitude is obtained. With regard to settlement areas which include
projections to water fronts, no statements can be made since the total
space depends upon the unpredictable linear distance to the shore from
the village.

III. CONSIDERATION OF S. K. LOTHROP'S DATA FOR TIERRA DEL FUEGO

The observations and facts which have made possible a quantitative
examination of area- population relationships in primitive settlements

are derived in large measure from California. The principal exception is
the formulation by Narroll (1961) of floor areas and populations in selec-
ted villages throughout the world. It is desirable to secure as many

cases as possible of local regions within which an independent test can be
made. An instance of this sort is provided by the maps and figures pub-

lished by S. K. Lothrop (1928) concerning the tribes of Tierra del Fuego.

In order to ascertain whether an exponential or a logarithmic

function exists relating living area and population, it is necessary to

have both site size measurements and population counts or indices for an
adequate series of localities. Lothrop describes three groups of sites,
all in the general vicinity of Harberton, Tierra del Fuego, and all per-

taining to the Yahgan tribe.   He also provides three maps (op. cit., Figs.

15

98, 100, 101) on which the sites are plotted. The number of house pits

is likewise given in corresponding tables in the text (oj. cit., pp. 184,
187, 188 respectively). A printed scale of distance in feet or meters is
supplied with each map.

By placing the printed sheet of the book under a binocular dissect-
ing microscope and laying the edge of a millimeter rule over the line on
the map, the equivalent of one millimeter on the rule may be closely

estimated in terms of meters or feet on the printed scale. Thus in Figures
98 and 100, one millimeter equals approximately 30.5 meters, and in Figure
101 it equals 28.6 feet. Each site is depicted on the maps as a mass of

fine parallel lines. The area of this mass can now be determined by direct
measurement of appropriate dimensions. If the outline of the site can be

regarded as a rectangle, the long and short axes are multiplied. If it is

closest to a circle, then the area is taken as ?r2; if it is oval, then the
two axes are measured and the formula for an ellipse (rab) is applied.    In
certain cases where the form is highly irregular, it is2possible to sub-

divide the diagram into portions, each of which can be handled separately.

There are., however, two obvious sources of error. The first is in
the measurement of the site representations which can never be completely
accurate, but this error is random and amounts to no more than a few per
cent. The second error is in the original drafting of the maps. Do the
areas shown on the maps really reflect the actual size and shape of the

sites? From a careful examination of the published result it appears to
us that the author did as faithful a job as could be expected, and that

certainly, in a broad manner, the relative sizes of the sites are adequately
depicted. It is true that we do have only a gross representation, and that
we have to allow for considerable latitude in the values for site area; on
the other hand the range of calculated square footages for the 65 sites

measured runs from 2,600 up to 785,000; in other words a factor of 300.    It
is very doubtful if inaccuracy in drawing outlines on a map or in actually
surveying the sites would reach this order of magnitude. We may therefore
accept the areas taken from Lothrop's maps as being an acceptable first
approximation.

Population is indirectly determined from the house pits counted.

For the purpose of the present study it is not necessary to know the total

number of inhabitants, since the point at issue is the relationship between
site sizes and site populations without reference to the absolute sum of

either. Indeed if we were concerned with totals, we would have to determine
whether all the sites and all the houses were simultaneously inhabited,

questions of extreme difficulty and answerable only from analysis of archae-
ological data. But the number of house pits may be considered as directly

proportional to the number of people, and therefore a valid index to popula-
tion.

16

The numerical values are plotted in Figure 6 as the logarithm of

the square feet of site area against the logarithm of the house pit count.
The scatter diagram for each of the three groups of sites shows a general
trend toward a straight line. A further analysis is given in Table 2.

The strength of association can be gauged by the values of , the corre-
lation coefficient. The first group shows a value of +0.799, which is

well beyond the one per cent level of probability and is highly signifi-
cant. The second and third groups, with only ten and five cases, give

values which lie respectively at about the two per cent and the five per
cent levels of probability; in other words moderately but not highly

significant. The consolidated data give a value of +0.689 for x which

for 65 cases is highly significant. There can be little doubt, therefore,
that the Tierra del Fuego region shows the same type of site area -popula-
tion function as does the Yurok on the Klamath River.

The slope of the regression line b differs slightly among the
groups of sites shown on Lothrop's three maps. This result is to be

anticipated since there are inherent distinctions between any two sets of
settlements, and since there is unavoidable variation in the measurement
of parameters and calculation of data.

The limiting area a is computed from the regression formula, but

graphically represents the backward extrapolation of the regression line
to its intersection with the x-axis. It may be interpreted as indicating
the minimum site area which could accommodate any house pits or dwellings.
Its numerical value depends upon the slope of the regression line (1) and
the average of the (logarithms of) the site areas plotted on the ordinate.
We note in Table 2 that the second group of sites (from Lothrop 1929, Fig.
100 and table on p. 187) shows the smallest value for b and a very large
limiting area in square feet as derived from a. It is also demonstrated
in our Figure 6 that the points for the ten sites in this group are dis-
placed upward, implying, as is true, that the average area of these ten
sites is greater than that for either of the other two groups. This

probably means simply that Lothrop, in selecting his groups of sites for

illustration, picked upon one in which the sites were larger than most of

those in the region. Otherwise we would have to search for some intrinsic
characteristic setting apart the local area in which the sites of the
second group were located.

17

IV. SETTLENENT AREA AS A CLUE TO REGIONAL POPULATION

There are a great many localities, inhabited in ancient or recent
times, of which the areas are known or can be measured, but for which

the populations cannot be determined by enumeration or by indirect docu-
mentary evidence. Yet, if the principle of exponential or logarithmic

relationship set forth in the previous discussions has general validity,

-there should exist in every instance a numerical constant which expresses
this function for the area and population of the particular community

type and ecological province concerned.  It follows, therefore, that if
we could learn or calculate the value of that factor, we could convert

size of settlement areas into the corresponding magnitudes of population.

One possibility exists when, for a homogeneous, relatively compact
region, data with respect to even a few of many separate settlements are
to be found either in the field or in the literature.  If in addition

these data supply figures for area or size on the one hand and for popu-
lation or some directly dependent parameter on the other, we can both

examine the validity of the logarithmic relationship and derive a probable
value for the regression coefficient. It may then be possible to apply

this information to the entire region, and, after a series of adjustments
depending upon the individual case, to emerge with a rational estimate of
the total population. This paper attempts to explore a single illustra-
tive example and to ascertain whether the logarithmic formula can be
applied to a specific region in the manner described.

We have selected a portion of the area of the tribe known as the
Chumash, who lived on the mainland and islands of Santa Barbara County,
California, at the time of Spanish settlement in 1769. The Chumash

settlements actually extended along the coast and on the offshore islands
from just above Los Angeles to San Luis Obispo, and into the hinterland
as far as the southwestern corner of the San Joaquin Valley. We are

interested here in the tightly-packed line of villages which reached from
near Ventura westward along the shore and narrow coastal plain to Point
Concepcion, a distance of approximately 100 kilometers.

The sources consist of the reports written by Palou (Bolton 1926),
Crespi (Bolton 1927), Portola (Smith and Teggart 1909), and Costanso

(van Hemert-Engert and Teggart 1910) covering the Portola expedition of

1769; the later account by Fages (Priestley 1937); the ethnographic survey
by Kroeber (1925); and the archaeological investigations of D. B. Rogers
(1929).

There are eight sites along the Santa Barbara Channel which were

18

examined by Rogers about thirty-five years ago and for which counts of
occupied houses were made by members of the Portol expedition who

visited the villages in 1769. Their names are set forth in capital

letters in Table 3. Outline drawings of the sites are given on maps
by Rogers (1929). Since the scale of linear units is also given on
each map, it is not difficult to measure very accurately the area of
each site separately by means of a polar planimeter. It is probable

that the representation of the site itself, as drawn on the map, is a
reasonably close approximation to the actual area on the ground.

The house counts were made as carefully as circumstances permitted
by Fray Juan Crespi, an unbiased and literate explorer and missionary,
and by Gaspar de Portola, who was in command of the expedition.    Their
figures were quoted by other participants and by later writers such as
Costanso, Palou, and Fages.   Crespi and Portola agreed in most cases
(see detailed data in Table 4). Where a discrepancy occurs we have

either accepted Crespi's figure or, if the difference is considerable,
have averaged the two estimates. For some villages the expression

"imore than"' is used. As a partial adjustment in these instances, the
stated figure has been increased by ten per cent.

When the logarithms of the values for the eight sites are plotted,
the points, despite some scattering, fall approximately along a straight
line (Fig. 7). The correlation coefficient between log area and log

house count is +0.762. This value, for seven degrees of freedom, lies

close to the two per cent level of probability, and the relationship must
be regarded as definitely significant. The regression coefficient has
the value of +1.317. We now have a credible numerical factor on the
basis of which we may expand our calculation in order to evaluate the
total population of the region.

In his extensive account Rogers (1929) lists and describes in the

text a total of 56 sites, including the eight mentioned above, and refers
to the native culture period which ended with the Spanish invasion as the
"Canaliino."  These 56 sites are listed in Table 3 according to the desig-

nations given by Rogers. The localities all fall within somewhat narrower
limits than those previously indicated. Roger's eastern border was the
mouth of Carpinteria Creek and his western border the mouth of Gaviota

Creek. Hence the territory embraced by the 56 sites was smaller than that
covered by Portola's explorations, and smaller than that actually occupied
by the Chumash tribe (cf. Kroeber 1925:526). There is no certainty that

all the sites were actively inhabited by the Indians in 1769. This point
requires further examination.

Of the 56 sites described by Rogers, there are 26, in addition to

19

the eight previously utilized for computing the regression coefficient,
which are drawn on Rogers' maps, and the areas of these may be deter-

mined by planimeter measurement. These names are underlined in Table 3.
We have no house counts for these 26 sites; however counts may be esti-
mated by interpolation of the area values on the logarithmic graph
already established. Thus one finds for each area the point on the

regression line where the appropriate ordinate intersects and reads off
.the corresponding house count value on the other axis. The individual
estimates may be moderately inaccurate (the standard error of estimate

is approximately ten per cent of the mean), but with a reasonably large
number of cases the probability is small that any serious systematic
error is introduced.

There now remain 22 sites described by Rogers as being essen-

tially contemporary with the Spanish expedition of 1769, for which there
are neither measurements of area nor house counts. These are given by
name in Table 3 but are otherwise not identified. For these sites the

only recourse left to us, but one which cannot be seriously at fault, is
to use the average count for the 34 sites already considered. This

average turns out to be 37.1 houses or, for convenience, 37. We cannot,
of course, argue that these 22 sites actually contained an average of 37
houses, but assume this in order to be able to treat the full site count,
If we are in error, it is probably in the direction of overestimation.
For the entire 56 sites the total number of houses would be 2,072.

The next step is to convert number of houses to population numbers.
If areas are disregarded, it is found that the explorers of 1769 recorded
house counts, together with estimated number of inhabitants, for 16

villages. The detailed data for these is given in Table 4 and summarized
in Table 5. The average number of persons per house, by direct division,
is 9.52. This figure must be subjected to further scrutiny with respect

to two factors:  first, the size of the individual house; and second, the
number of persons ascribed to each village.

The houses on the Santa Barbara coast were large as Indian dwell-
ings go. The chief items of direct testimony are statements by Costanso

(van Hemert-Engert and Teggart 1910:132) that the houses are "hasta veinte
varas de diametro" and "contiene cada casa tres, 6 quatro familias."    This
applies to one village. Crespi (Bolton 1927:163), referring to another
village, says, "some of them so large that they house many families."

Note that he says "some of them." On the other hand it must be pointed
out that in the Villages west of Gaviota the houses are clearly single
family units.  Fages uses the word "fuegos" instead of "casas," thus
clearly implying small structures.

20

The over-all average of nine or ten persons per house appears to
reflect a mixture of dwelling types. Such an average is definitely too

large for uniformly single family units, yet it does not justify an assump-
tion of four or even three families in all the houses. If in the big

central villages there were a fair proportion of three to four family dwell-
ings together with rather frequent single family houses, and if at the west-
ern end of the Channel all the houses were single family, then we could
arrive easily at a minimum of ten persons per house for the entire area.

Turning now to the village populations cited by Crespi and others,

we encounter the perennial problem of the reliability of contemporary wit-
nesses. In the present instance it must be emphasized that these were

exceptionally responsible and intelligent men with a background of profound
experience, each of whom made a very serious effort to achieve an accurate

account of his journey, one of great importance to both his church and state.
Consequently it is most unlikely that there was falsification or deliberate
exaggeration. The only unsettled question is whether they saw all of the
natives who inhabited the villages that were visited. Here it must be

argued that the reports did not include the total number of souls. In the
first place, many of the infants, very small children, and some of the

women would not be in the crowds that greeted the white men at each stopping
point. In the second place it is repeatedly stated by the diarists that

numerous men were absent fishing or visiting the islands, because many of
the deep-sea canoes were not in their home ports. In the third place we

have the surprising statement of Pedro Fages (Priestley 1937:32), referring
to the stretch of villages westward from Carpinteria to Point Concepcion,
that they average about "six hundred men capable of bearing arms." Men

capable of bearing arms would approximate one-third of the entire population.
Hence one can interpret Fages' remark as indicating that the villages aver-
aged about 1,800 inhabitants. Although we need not accept Fages' assertion
without reservation, in view of the facts as a whole, we may with propriety
increase the Crespi-Portola figures by twenty-five per cent. A correspond-
ing increase from 9.52 to 11.90 would be required in the estimate of the

average number of persons per house; let us say, for simplicity, 12 persons.
Based on a normal six members per family, this means an average of two

families per house, a value which seems to conform to our other evidence

concerning house sizes among the Chumash. If we now include all the 56 late
culture sites of Rogers, the population of these villages alone would amount
to 24, 800 persons.

The final adjustment relates to the 56 sites. How many villages
along the Santa Barbara Channel coast were actually inhabited in 1769?

Crespi mentions 16 villages personally visited or observed by the members
of the expedition; undoubtedly many others were missed or ignored in the
written account. The first European to see the Santa Barbara Channel was

21

the Spanish maritime explorer Rodriguez Cabrillo who sailed along this

coast in 1542-1543. As quoted by Kroeber (1925), Cabrillo, through his
pilot, Ferrel, listed in his diary 41 native villages.   Kroeber points
out numerous spelling variants and duplications which would reduce the

number from 41 to 30.  Cabrillo, however, did not describe the sector of
the coast extending from Carpinteria eastward to the vicinity of Ventura.
If we turn to Kroeber's (1925, pl. 48) map of the Chumash territory, we

find that he shows as authenticated 29 settlements from Carpinteria west-
ward to Gaviota, inclusive; 3 west of Gaviota to Point Concepcion; and 8
from Carpinteria eastward to Ventura; making 40 in all. Thus Kroeber

admits 29 on that stretch of coast for which Cabrillo claimed about 30

and where Rogers found 56 recent culture sites. Conceding readily that

all of Rogers' sites probably were not simultaneously occupied either in
1542 or 1769, Kroeber nevertheless tends to be ultraconservative both in
his interpretation of Cabrillo and in his allocation of villages accord-
ing to the imprecise ethnographic data. There will be no great error on
the side of excessive optimism if we increase Kroeber's estimate and

decrease that of Rogers so as to allow 35 villages for the central por-
tion of the coast, 8 for the sector east of Carpinteria, and 3 for the
sector west of Gaviota. The total for the entire strip would then be
46 villages.

Finally we revise the earlier figure based upon Rogers' 56 sites.

Using a total of 46 villages, an average of 37 houses per village, and 12

persons per house, we get a probable aboriginal population of approximately
20,400.  The members of the Portola expedition thought there was a minimum
of 10,000 persons between Ventura and Point Concepcion, but they probably
did not see all the villages and underestimated the number of inhabitants
of those they did see. Kroeber's guess of 10,000 for the entire Chumash

tribe must be regarded as much too low.   A range of 18,000 to 22,000 souls
*for the mainland coast alone is more likely to approach the truth.

V. HOUSE AND VILLAGE SIZES IN THE ABORIGINAL CENTRAL VALLEY

OF CALIFORNIA

The size and dimensions of aboriginal dwelling houses are important

for the description of settlement patterns and the estimation of population
in primitive communities. We deal here with the problem as it relates to
the Central Valley of California and its peripheral uplands.

As with other parts of California, two distinct and significant

sources of information exist-ethnographic or historical documentation and

22

the archaeological site records on file at the University of California
Archaeological Research Facility at Berkeley. The former sets forth the
results of observation by visitors more than a century ago or of tribal
knowledge derived from informants, the latter consist of counts or

measurements of house pits taken by students or faculty during field
trips or surveys.

The data which are to be found in the site reports on file at the
Archaeological Research Facility have been consolidated in Table 6. Here
are shown, first, the conventional site designations, and then the tribe
concerned and their habitat. Based upon tribal and ecological character-
istics, four regions may be distinguished within the territory as a whole.
The first consists of the foothill areas occupied by the Wintun at the

northern end of the valley and extending down the west side as far as Yolo
County. The second embraces the Sierra Nevada foothills in the possession
of the Maidu and the Miwok as far south as Madera County. The third

includes the foothills of the Coast Ranges in Stanislaus County, believed
to have been occupied by the Yokuts, together with the Sierra foothills of
Fresno and Tulare counties which were held by the Monache and Yokuts. The
fourth region covers the valley floor or flat land from Red Bluff to
Bakersfield, irrespective of tribal affiliation.

Next in the table appear measurements of house pits made at the
inspection of a site and recorded in the files by field workers. Such

measurements have been expressed almost invariably in feet diameter, occa-
sionally as length and width. For area the former measurement may be

converted by the ordinary formula gr2  and the latter by the formula for
the area of an ellipse. Each site has been treated as a unit. Where

several pits were seen at a site and a range of size is mentioned, the
midpoint of the range should be used for the calculation of area. The

table presents the original data in feet diameter unless otherwise indi-
cated. Finally, notes are added as they appear in the Facility site

records. For comparative purposes the values obtained for floor area at
each site are averaged within each region and the result placed in the
summary to the table.

Before accepting the mean values for floor space some discussion
is necessary. Let us begin with the area which includes the lower foot-

hills surrounding the Sacramento Valley, in possession of the Wintun, from
north or northeast of Red Bluff southward to the Sacramento Delta. There
are seven sites from which pits (or houses) with diameters ranging from 4

to 14 feet are reported. These are clearly small, standard, single family
units such as are seen throughout large sections of Central California,
In addition there are two sites, Teh-95 and Gle-10, which show a single

pit, respective sizes being 30 by 30 feet, and 40 feet in diameter. These

23

are recorded as a "sweat house" and a "dance house."   There is no reason
to question the diagnosis, and hence it is evident that the size of the
pit may segregate dwelling from ceremonial structures. Moreover it

should be emphasized that the maximum diameter reported for an ordinary
dwelling house is 14 feet.

In the second region, occupied by the Maidu and Miwok in the

Sierra Nevada foothills, there are eight sites, from which are reported

pits from 10 to 22 feet in longest dimension, again clearly single family
dwellings.  In addition, however, there are on record eight sites at

which much larger structures were found. These measured from 12 to 52
feet in diameter. Three of them are designated "dance house" (But-65,

But-94, Mrp-2), one a "former sweat house" (Mad-29), and one a "present
sweat house" (Mad-37). Concerning the last there can be no doubt what-
ever; concerning the others very little doubt that they are correctly

identified by the field observers as pits for ceremonial structures. At

each of three other sites (But-66, But-84, Eld-73) only one pit was found
and the observers assumed that it was the remains of a dance or sweat

house. Important villages which were tribelet centers or "capitals" in
which the tribelet chief lived usually contained a ceremonial house
(Kroeber 1932:257-259; 1962).

At this point particular attention should be called to sites

Mrp-2 and Mad-37, for here the two types of pit occur together. At Mrp-2
there were 15 house pits ranging in diameter from 14 to 22 feet, plus a

dance house pit of 52 feet in diameter. At Mad-37 there was at least one
house pit 18 feet in diameter, plus the present sweat house of 33 feet in
diameter.  There can be little question, therefore, that in this area the
dwellings did not exceed 22 feet in diameter and that the large pits
represent ceremonial structures.

The third region includes sites on both sides of the valley, all
of which were probably occupied by Yokuts and tribes of the Mono stock.

To the west there are two sites, Sta-129 and Sta-130. At the former was

found one pit 33 feet in diameter, with no further description; at Sta-130
there was recorded one pit 47 feet in diameter, designated as a "large
dance house," and another pit 35 feet in diameter described as a sweat
house. Additionally, there were two pits, each 14 feet in diameter,

reported as being dwelling houses. The diagnosis here again may be accepted
as correct.

To the east, in the foothills of Fresno and Tulare counties, there
are twenty sites reported which show house pits ranging from 6 to 20 feet
in their longest dimension; no larger pits were recorded. Consequently

there can be no alternative but to consider all the listed pits as being
those of dwellings of the single family type.

24

For the fourth region, the flat valley bottom, the data unfortu-
nately are very scanty. The native groups were wiped out as social

entities by 1820 and for more than a hundred years their homesites have
been converted to farm land. As a result there are only eight sites at
which house pits of any description are on record at the Archaeological
Research Facility, although there are a few additional ones mentioned
by Schenck and Dawson (1929).

The Facility files begin with Teh-34, Teh-53, and Col-10, at

which sites the pits range from 7 to 15 feet in largest dimension, clearly
single family dwellings. To these may be added Mer-43 with several pits
18 to 20 feet in diameter. It is these four sites which provide the
floor space figure in the summary to Table 6.

There are four additional sites which require comment. Sac-ll and
Sac-117 showed a single pit, in each case 25 feet in diameter.    There is
nothing in the size of the pits or in the site descriptions to indicate
the nature of the original structures. At Mer-58 there was one "dance

house pit" of 43 feet in diameter, plus twenty others for which no dimen-
sions are given but which presumably were small. Hence the single large
pit probably was indeed that of a dance house. At Mer-89 there were two
large pits, 35 and 45 feet in diameter respectively, plus others of

doubtful number and size, presumably smaller.   The large pits may or may
not have been those of dance or sweat houses.

One conclusion emerges from the preceding analysis: in the Central
Valley and adjacent hills there were two clearly distinct sizes of build-
ings. The smaller may be regarded without much doubt as single family
dwellings; they are overwhelmingly the dominant type in the hill areas

regardless of tribe.  The larger., in certain instances, must be considered
to represent sweat or dance houses, particularly where a single one is
found in close propinquity to a cluster of smaller pits. Where one is

found alone, especially on the floor of the valley, there is a distinct

possibility that we are dealing with a large., multifamily or communal unit.
To what extent this is true cannot be determined by the existing-or

probable future-field data secured at the sites concerned. In the mean-
time we may state with some assurance that the normal house type, even in
the lower foothills, was the single family structure.

Verbal and documentary evidence comes to us from two sources. The

first consists of several accounts written by explorers, missionaries, and
soldiers in the early nineteenth century. Many of these have been trans-
lated (e.g. Cook 1955b, 1960, 1962) and may be utilized in the English

version. A number of them contain information of value concerning village
and house organization.

25

1. The diary of Father Jose Maria de Zalvidea: In 1806 this

missionary visited the villages on the northern slopes of the Tehachapi
Mountains directly facing the southern San Joaquin Valley and also the
villages on the shore of Buenavista Lake. One of the places seen was
Cuia, in the range of the Yokya Chumash, which had nine houses and 41

persons, or 4.56 persons per house.   Another was Sisupistu, a village of
the tribelet Hometwoli, at the eastern end of the lake. Here Zalvidea

counted 28 houses and 50 to 60 men. From enumeration ratios it has been
estimated (Cook 1955b:54) that this number indicates a total population
of 180 persons, or 6.43 per house.   It would appear, therefore, that in
this area there were small, single family dwellings.

2. A letter written by Father Juan Martiln (Cook 1960:243): This
priest visited Tulare Lake in 1804. He went to the village of Bubal,
the principal residence of the Wowol. At the end of his third day of
travel Father Martin arrived at the first "suburb" of the village and
named it La Salve. The heathen concealed their women in "some little

huts.," but later allowed them to come out.  In the evening the party was
visited by people from the "main village" which was some distance away.
It is clear that La Salve was a small village, peripheral to Bubal, and
contained small dwellings, probably single family ones.

3.  The diary of Father Ramon Abella:   In 1811 this missionary
traversed the delta region by boat and ascended the Sacramento River

nearly to the city of that name. On his return Father Abella encountered
the territory of the Tauquimenes (Cook 1955b:57). The main village had 60
houses and 1,000 people.  The houses "teinen en circuito 28 o 30 varas,
con su orcon en medio." A vara is approximately 33 inches; hence the

houses were about 990 inches or 82.5 feet in circumference. This means
about 26 feet in diameter and 408 square feet in area. If there were
1,000 inhabitants, there would have been nearly 17 persons per house.

This figure may seem excessive, yet Abella writes that the first or main
village "habra como unas 1,000 (almas) segun lo grande que por aqui son

las casas.1' The indication is therefore that the figures cited represent
a considered judgment.

At one league downstream Abella saw another village of 14 houses,
and farther on villages of two or three houses. There was evidently a
system of main plus suburban villages.

4. The diary of Father Narciso Duran: In 1817 this priest visited
the delta and reported on the houses in two villages. One of these, of

the Ochejamnes, had 40 houses; the other, of the Chucumnes, had 35 (Cook
1960:274-275). With regard to the latter Duran says explicitly, ofWe

counted 35 houses, some 20, 40, or 50 paces in circumference, a fact which

26

indicates a large number of people."   Note that Duran here says 'tpasos"

(paces), not yrvaras.hy The Academy Dictionary lists the "rpaso geometrico"
as being 5 feet, 1,393 millimeters, or 57.6 inches long.   Hence these

houses would have been from 16 to 40 feet radius and 803 to 4,520 square

feet in area.  These are truly enormous dimensions; if the pasos of Duran
are taken as varas, then the radii are 8.8 and 21.9 feet, and the areas

243 and 1,506 square feet respectively.   However, at this point consider-
ation should be given the statements of Estudillo.

5. The diary of Lieutenant Jose Maria Estudillo: In 1819 this

officer conducted a military expedition to the San Joaquin Valley, during
the course of which he made a careful examination of the village of

Chischa in the Sierra foothills, probably on the middle Kaweah River

(Cook 1955b:47). The population was estimated at 1,250 souls, the area
at 80,000 square varas or 608,000 square feet.   However, the point of

greatest interest in the present connection is that Estudillo designated
the value of the paso as it was used in Spanish California. He gave

dimensions for Chischa as 358 pasos or 626.5 varas long, and 432 pasos

or 756 varas across. Hence one paso must equal 1.75 varas or 57.6 inches.

6.  The diary of Lufs Arguello (1821):   This is the account of his
famous expedition from San Francisco up the Sacramento Valley, westward

to the coast, and back down to San Francisco. Between Suisun Bay and Red
Bluff he saw a dozen or more villages which he stated had a population of
500 to 1,000 or more. These all appear to have been located on the banks
of the Sacramento River or at the mouths of its tributaries.

Apart from the Spanish-Mexican explorers, there are statements by
Americans who early entered the valley. These people repeatedly mention

villages of anywhere from 100 to 1,500 inhabitants. In fact the unanimity
of opinion on this point is so overwhelming that, in spite of assertions
by modern ethnographers, it cannot be seriously doubted that up to 1830
numerous villages of this size were to be seen along the courses of the

main rivers from north to south. Concurrently the house number is gener-
ally put at 20, 50, or 100. Hence we must think in terms of 5 to 15 or
more persons per house.

One of the earlier visitors was John Work, who came down from the
Columbia River in 1832. He says (1945:20) that north of Butte Creek

"ithe Indians...dwell in holes in the ground of a conical form."   Among

the Maidu on the lower Feather River in one village Work (1945:25) counted
28 houses, sunk in the earth so as to resemble "a round hillock." Near
Oroville he saw (j. cit     p. 32) four villages of 40 to 50 houses each.
These all evidently were single family dwellings.

27

In the years 1832 and 1833 American explorers entered the Central
Valley, primarily as beaver trappers.  Among them was J. J. Warner, who
later settled in Southern California.  He is quoted in An Illustrated
History of San Joaquin County, California (1890:28) as saying, "In the
fall of 1832 there were a number of Indian villages on Kings River,

between its mouth and the mountains; also on the San Joaquin River from
the base of the mountains down to, and some distance below, the great

slough. On the Merced River from the mountains to its junction with the
San Joaquin there were no Indian villages; but from about this point on
the San Joaquin, as well as on all of its principal tributaries, the

Indian villages were numerous; and many of these villages contained from
50 to 100 dwellings, all of which were built with poles and thatched with
rushes."

The Hayes Mission Books at the Bancroft Library in Berkeley, which
were assembled by Benjamin I. Hayes of San Diego in the 1860's, contain a
newspaper clipping signed by "Trapper" the pen name of someone who was

on Ewing Young's expedition of 1832, and who very likely was J. J. Warner.
He stated., "The banks of the Sacramento and San Joaquin, and the numerous
tributaries of these rivers, and the Tule Lake, were at this time studded
with Indian villages of from one to twelve hundred inhabitants each. The
population of this extensive valley was so great that it caused surprise,

and required a close investigation into the nature of a country that with-
out cultivation could afford the means of subsistence to so great a
communi ty. "

Later visitors commented repeatedly upon the great loss in popula-
tion caused by the epidemics of 1833 and subsequent years, and referred
frequently to the great size of the former villages by way of contrast

with the pitifully small remnant found after 1835. It is very clear that
by 1849 the riparian tribes had in effect been obliterated. (For a more
extended discussion of these matters see Cook 1955a, 1955b.) One result

of this extreme disintegration has been that informants in the early years

of the twentieth century were unable to convey to ethnographers an adequate
quantitative picture of native culture along the big rivers of the Central
Valley. Nevertheless considerable information of value has been recorded.

In the northern valley house sizes have been described by Dixon,

Kroeber, and McKern. Dixon (1905), writing of the Northern Maidu, states
that he found two types of permanent houses among the groups in and

bordering on the valley. Type 1 consisted of a circular or conical house
6 to 12 meters in diameter which would be occupied by "several families,"'
and was most common in the valley proper. Type 2 was found in the foot-
hills, and was 2? to 5 meters in diameter. Most villages contained from
one to 20 houses of type 1.  The area range for type 1 is 301 to 1,216

28

square feet; for type 2 it is 53 to 211 square feet. The latter type is
undoubtedly a single family dwelling, and can scarcely have been much
less than 150 or 200 square feet in area.

Kroeber (1929) described the Southern Maidu of the Feather River
Basin and stated that the houses were of the single family type and
that "a village might have six or seven houses." This is a long way

from the assertions of Work and other contemporaries, that there were
many villages in this region of 50 to 100 houses. In a later paper

Kroeber (1932) said that the Wintun had houses on the style of the Patwin,
but that they seemed to have been smaller. The Patwin houses have been

the subject of a paper by McKern (1923), who bases his text on the testi-
mony of an informant from near Colusa, and describes in detail the con-

struction of houses but gives the sizes of only two. These had respective
dimensions of 18 by 22 feet and 23 by 28 feet, or 396 and 644 square feet
in area. These values lie within the lower portion of the range of

Dixon's type 1 houses of the Maidu. The maximum diameter, according to
McKern, averaged 30 feet, or 706 square feet. It is very difficult to

see how more than two families or one extended family could live in such
a house.

Concerning the delta and the lower San Joaquin River, there is

hardly a whisper of ethnographic evidence. The Indians acquainted with
this region were so nearly destroyed prior to 1850, the earliest birth

date for a reliable modern informant, that we quite lack any substantive
cultural data. For the upper San Joaquin and Tulare Lake area we have
the significant work of Gayton and of Latta, together with a discussion
by Kroeber.

To summarize Gayton's (1948) data: the Chunut and Tachi on Tulare
Lake had an oval house, stated to be the most common type in the valley.

This was a communal edifice, housing several families. The Bankalachi of
upper Deer Creek in the foothills had a conical, single family house.

The W-ukchumni had a conical house which varied with the size of the family.
The Choinimni in the lower foothills had a house which on the average held
a family of six; no dimensions are given. The Kechayi in the foothills of
the San Joaquin River had a one-room, circular house, 10 feet in diameter,
which held a single family. The floor area was 78.5 square feet. The
Chukchansi in the Fresno River foothills had a conical house, 12 to 14

feet in diameter, or 133 square feet in area, which held one family. The

Michahai and Waksachi of the Kings and Kaweah river foothills had a conical
house 9 to 12 feet in diameter, or 87 feet in area, which contained a
single family.

F. F. Latta has written two books, of which the more important here

29

is the Handbook of the Yokuts Indians (1949). Latta states that the
Yokuts of the valley, as opposed to those of the foothills, built at

least five types of dwellings. The permanent houses varied from 10 to
12 feet in diameter to a much larger size. There was a communal house
which might reach a length of 300 feet, and might be used by many fami-
lies.  These structures were found among the Wowol, Chunut, and Tachi,
also probably among the Tulamni, Tuhohi, and Hometwoli-all valley

tribelets.  It is suggested by Latta's discussion that these big houses
were destroyed by the Spaniards in the early years of the nineteenth
century and that they were replaced by smaller dwellings.

The Yauelmani on the lower Kern River built similar, but smaller,
houses. Thus a description by T. A. Baker (Latta 1949) of houses in

1863 near Bakersfield shows that the Indians were living mostly in oval
shaped shelters, 10 by 15 feet in dimension.   It has also been a general
observation that the Yauelmani, together with the Wechihit and Apiachi,

built a small tule house of oblong shape, probably not larger than 10 by
18 feet. The Yaudanchi built conical winter houses of 12 to 15 feet in
diameter, and similarly the Chukchansi and Gashowu. Latta (1949) also
gives a description by T. J. Mayfield of the typical house of the

Choinimnion upper Kings River. There was one circular room, 10 to 20
feet in diameter, with a smoke hole in the top.

Kroeber (1925:521) lists five types of Yokuts houses. The first
is the cormunal house characteristic of the Tulamni., Hometwoli, Wowol,
Chunut, and Tachi: "These houses sometimes ran to a size where they

accommodated a little more than 10 families."   The second was a modifi-

cation of the first style, found among the Yauelmani, which "approached"
the first style in size. The third type is seen among the valley tribes
Wechihit and Tachi, who "today build small tule houses" which are ellip-
tical in form. The fourth was built by the Yaudanchi, Chukchansi, and
Gashowu, hill tribes. It was a conical tule dwelling of small size.
The fifth was a bark structure similar to the preceding.

The work of these three investigators makes it clear that as far
as living memory is concerned, the big communal house never existed be-

yond the limits of the flat valley, and in the valley itself disappeared
at or about the time of the Caucasian invasion of 1800 to 1835.

Considering now the written reports as a whole, certain conclusions
appear valid. There were two primary aboriginal house types extant in the
Central Valley and neighboring foothills: a large, communal structure

which we may designate, following Dixon, as type 1; and a small, single
family unit, type II.

30

Type I is reported from the delta by Abella (1811) and Duran
(1817), both of whom took measurements. By inference from village

populations and house numbers, the communal house was seen by Arguello
(1821) in the Sacramento Valley and by Warner (1890) and "Trapper."

These dwellings were also reported by the informants of Dixon, McKern,
Gayton, Latta, and Kroeber for riparian villages from the Northern

Maidu to the Southern Yokuts. But nowhere is there a trace of type I
houses in the foothills, even at the margin of the valley floor. On

the other hand, type II is described from hill areas by Zalvidea (1806)
and other Ibero-American explorers, together with nearly all the modern
ethnographers. The two types therefore conformed to the terrain very
closely, in that type 1 was restricted rigidly to the river and lake
bottom lands whereas type II was universal in the uplands.

The data for size indicate that the communal house could vary

within wide limits, apparently in accordance with population and space
requirements. Hence we have measurements and estimates of from 400 to
4,500 square feet floor area, or even larger if we allow for Latta's

300 foot building.  On the other hand, the house of type II is reported
very consistently as being of from 100 to 400 square feet.

If we concede the aboriginal distribution of types as outlined,

then we have to account for the almost complete disappearance of type I
by the 1850's and its replacement by type II, even on the valley floor.
The reason is rather clear. The construction and maintenance of the
large, multifamily dwelling required, first, level land such as was

provided by the broad river flats, and second, a stable society. The
second criterion was satisfied by the large, well ordered village and

its satellites as they originally existed, backed by a copious, reliable
food supply such as could be derived from the rivers, lakes, and marshes.
On the other hand, the equilibrium was delicate, and there was little
reserve economic strength in the culture. When the Spaniards or the

Americans struck, they disrupted the going concern, destroyed the food
supply, devastated the villages, dispersed the families by missioniza-
tion, slavery, disease, and massacre.   Hence the survivors simply did

not have the material, or probably the emotional, resources wherewith to
erect and occupy the big houses. They were forced, if they remained in
the old habitat, to adopt the flexible, mobile, inexpensive style of
existence provided by the small, single family dwelling, a manner of
living to which the hill tribes were already well adapted.

The archaeological site records, as shown in Table 6, now find a

logical explanation. The universal occurrence of the single family house
in the upland areas merely reflects ancient usage and is to be expected.
The occasional appearance of a solitary large house pit undoubtedly

31

indicates the former presence of a sweat or dance house, whether or not
so designated by the field observer. At those very few sites in the
flat valley where pits of any kind have been discovered, the majority

are of the small house type and probably demonstrate the tendency mani-
fested in the past hundred and fifty years to adopt this style as a
substitute for the communal dwelling.

The single large pits found at sites Sac-111, Sac-117, Mer-58,

and Mer-89 might be regarded as representing communal houses. However,
one factor militates against such a conclusion:  in three of the four
cases mentioned there is one such pit and no others; at Mer-89 there

are two. Now the big villages along the rivers are repeatedly described
as containing many large houses-up to at least one hundred. The small-
est settlements to which I have seen reference are those spoken of by
Abella (1811) as consisting of two or three houses. Hence it is very
surprising to encounter no archaeological sites showing more than two
large pits. One would suppose that if any of the big dwelling house

pits survived at a given site some of the others also would persist. As
a result we are forced to assume that the pits at the four sites listed
above are the remains of relatively recent special purpose structures,
such as dance houses.

It follows that the quantitative settlement pattern of the pre-
historic Central Valley is all but lost to us, for the population and

house data which we possess cannot be tied to specific sites or villages.
However, in general terms and for working purposes some very broad

estimates can be ventured. These are based upon the evidence presented
previously, which is consolidated for convenience in Table 7.

As a first approximation we have records of 375 sites from Red

Bluff to Bakersfield which show a geometrical mean of 25,400 square feet
in area. Contemporary accounts frequently indicate a population of 750
persons distributed through about 50 houses. From the figures in Table
7 it appears that 1,000 square feet is a fair estimate of average floor

space. This would mean 15 persons per house and 67 square feet of floor
space per person. However, certain adjustments appear necessary. An
average house number of 50 is too high; the mean of six figures cited

for the valley in Table 7 is close to 40. But this (except Dixon' s value
of one to 20) represents the largest villages which came to the attention
of visitors prior to 1850. We must remember Abella's description of the
Tauquimnes who had several villages-one of 60 houses, one of 14, and

several of one or two. The mean here is 16 houses. Hence, in order to
be conservative., we may cut the estimate of 40 in half and allow an

over-all average of 20 houses. It is also probable that the number of

families, or their equivalent, inhabiting each house reached at least to

32

three, that is to say 18 persons. With a floor space of 1,000 square
feet, which seems reasonable, the area per person would be nearly 56
square feet.

Finally, the mean site area of 25,400 square feet seems too small.
If one considers that a house with 1,000 square feet of floor space, plus
peripheral space for all sorts of domestic activities, would have to pre-
empt fully 3,000 to 5,000 square feet of the site area, it is clear that
the total site area would have to exceed 10,000 square feet if more than
two houses were to be accommodated upon it. Following this line of

reasoning, we may delete the 83 sites listed in the Facility records as
being less than 10,000 square feet in area. We then get for the 292

remaining sites a geometrical mean of 41,400 square feet. These values

are admitted to be little more than rough surmises, yet they may well lie
within the realm of probability for the long extinct valley communal
villages.

VI. HOUSE SIZE, SETTLENENT AREA, AND POPULATION THROUGHOUT

NORTHERN AND CENTRAL CALIFORNIA

Part 1

The numerical data which are available to us concerning house

sizes, village or site sizes, and population have been examined for three
areas within the state of California: the Central Valley, the Santa

Barbara Channel, and the lower Klamath River. As a preliminary to exam-
ining demographic and ecologic interrelationships through a world-wide

spectrum of conditions, it is desirable to assemble all possible inform-

ation for a territory of moderate but substantial extent, and to tabulate
this information in such a manner that it can be made the basis for com-

putation. It is also advantageous if the territory involved can be broken
up into smaller units, each representing a region more or less homogeneous
in nature and each containing a sufficient number of settlements and

people to permit an internal analysis as well as a comparison on a broader
basis with other regions.  In other words, we might work on two levels,

one regional or local, as was done with the Yurok, the other territorial,
as could be done with an entire state, country, or civilization.

California provides an admirable laboratory in which to test these
ideas, both because the state can be divided into a series of regional

units on the basis of aboriginal habitation as well as geographic varia-
tion, and because an unusually large amount of information has been

33

accumulated with respect to these units and their ethnography and arch-
aeology.

The first step is to assemble the pertinent data. To this end we
have examined as much of the published material as we have been able to

find and have scrutinized the file of site records at the Archaeological
Research Facility (Berkeley). The results of these inquiries are pre-
sented here in the form of two tables.

The first, Table 8, shows those figures at our disposal for sev-
eral parameters, of which three may be regarded as primary: house size
(floor space); village size (site area); and population. House size is
expressed as area in square feet, and is calculated from written state-
ments or measurements of linear dimensions. In almost all instances

(except the Yurok and Chumash) site area is obtained from the archaeo-

logical site survey records. Population cannot be computed from direct
estimates of tribal or regional totals because we rarely know how many

villages were occupied at a particular time. Hence we rely upon number
of persons per family, number of persons per house, and number of houses
per village to determine population. Occasionally, but infrequently for

California, figures for the number of persons per village are immediately
available.

In addition to the above mentioned categories, Table 8 contains

several derived magnitudes which are of use for various purposes. Among
these are floor space per person, mean total floor space per village,

mean square feet of village space per house and per person, and the ratio
of floor space to village area. The table also shows a great many gaps

and lacunae. This characteristic is referable to the extreme scattering
of information in the sources. A single item or figure must often be

picked from this or that author and inserted where appropriate as an iso-
lated fact in the table. The complete estimate which accompanies the

data for each region must then be formulated, rarely from a single author-
itative statement, occasionally by a solid average, but most frequently by
utilizing sense and judgment. Thus the results are in a certain measure

subjective, and indeed in some instances it has been regarded as preferable
to submit no estimate at all rather than to rely upon an almost wholly
unsupported guess.

The total number of regions for which sufficient house-site-popula-
tion estimates can be obtained is thirty. The division initially followed
ethnographic lines, that is tribal affinity as established by Kroeber and
his students throughout five decades. Thus the Klamath River tribes are

considered separately by tribal group (Yurok, Karok, Shasta), and the Yuki
and the Pomo are kept as individual units. Secondarily, distinctions have

34

been made along geographic lines both intra- and inter-tribally. These
distinctions are based upon two additional factors. The first is the

ready availability of data; for instance there is copious, easily segre-
gated information on four groups of the Wintun stock- (l) the Wintu of

the Trinity River; (2) the Northern Wintun on the upper Sacramento River;
(3) that portion of the Central Wintun who lived on Stony Creek; and (4)

the Southern Wintun inhabiting Cache Creek, Putah Creek, and their afflu-

ents. The second factor is the lack of data imposed by historical accident,
This deficiency in turn necessitates the consolidation of minor regions

which on edaphic grounds might deserve separate treatment. For instance,
the early destruction of the Athabascans necessitates the use of their

territory as a whole. The wiping out of the big river settlements means
that the entire Central Valley has to be regarded as a single region.

Finally, certain areas are omitted simply because there are no data whatever
pertaining to them. Of these the most conspicuous are the periphery of San
Francisco Bay and the habitat of the Costanoans and Salinans in the central
coast ranges. Although numerous archaeological sites have been recorded,
there is complete ignorance regarding details of dwelling construction,
family size, and village population.

Of the 30 regions mentioned, 26 are placed in Table 8. The other

four are discussed separately in Part 2 of this section. The 26 areas are
those in which it is generally acknowledged the predominant, if not the

exclusive, house type was that which held a single family, sometimes perhaps
an extended family, but never more than an average of eight persons. The

calculations for, as well as the characteristics of, these regions are more
or less uniform. They can therefore be handled conveniently as a single
group in the same table.

The remaining four regions all possess the feature which they con-

tain-multifamily or communal houses-to a significant extent, Each region
must consequently be examined carefully and independently in order to arrive
at as much as a first approximation of area- population relationships. Two
have already been discussed in previous sections, the Central Valley and the
Santa Barbara Channel. The others, the interior Pomo and the Wappo, are
considered in Part 2 of this section,

A fifth region constitutes a special case:   this is Santa Rosa Island
for which extensive site records are on file at the Archaeological Research
Facility.  This island was inhabited by the Chumash., and the population was

high at the time of the Spanish occupation. As has been shown by Orr (1951),
there undoubtedly were on the island many houses of the communal type similar
to those seen on the mainland. However it is probable that they were in the
minority. The problem is discussed in Part 2 of this section, but the final
decision has been to include Santa Rosa Island in Table 8 in so far as the

35

single family dwellings are concerned, and then to place the estimates
adjusted for the communal houses in Table 9.

Part 2

The data for single family houses have been set forth exhaustively
in Table 8 and the notes thereto. There still remains the problem of the
communal dwelling and its relation to population and area. The situation
in the Central Valley and along the Santa Barbara Channel have already
been discussed. These areas need not be considered further in detail,
although since the final estimates are included in Table 9 some brief
recapitulation is desirable.

For the Central Valley it was shown that the communal house would
accommodate two to four families, had perhaps 1,000 square feet of floor
space, and was present to the extent of about 20 houses per village.

Actually, if we average four estimates of floor space derived from two

contemporary observers and two modern ethnographers, we get J,130 square
feet for floor space. Similarly, averaging seven estimates of houses

per village we get 42 structures. But it can be shown that some of the

contemporary observers noticed principally the larger villages; hence 42
is probably a serious overestimate. The missionary, Father Ramon

Abella, exploring the delta, found a tribal complex which consisted of
one village with 60 houses, one with 14, and "several" with one or two
houses. Calling "several" equal to three, the average is 16 houses.
Therefore we take 20 as a fair approximation. The maximum number of

families per house is four, three would be more common, and two perhaps
even more frequent. A count of three families, with 18 persons, is

reasonable. For site size it was decided that sites under 10,000 square
feet could scarcely accommodate these big houses in any significant num-
ber; hence those sites in the Archaeological Research Facility records
which were under 10,000 square feet were omitted. The 292 remaining

sites average 41,400 square feet, The figures here underlined have been
included in Table 9.

With regard to the Santa Barbara Channel Chumash, we have calcu-

lated an average of 37.1 houses per village. The direct evidence at hand
does not allow an estimate of more than two families or 12 persons per

house, However it is known that these data include a number of villages
at the far western end of the Channel which clearly contained single

family dwellings. Although these villages had to be considered in an

evaluation of the total population of the region, for the present purpose
of studying the settlements with communal houses they may be disregarded.
Therefore the figures cited above may be arbitrarily increased somewhat,

36

say to 2.5 families and 15 persons per house. The mean area of the sites
was found to be 1011,000 square feet. For the floor space there is no
direct record.

The only statement we have encountered is one by Kroeber (1925:

557) that the Chumash house was large, up to 50 feet in diameter.    If we
regard 50 feet as the extreme, then the range may well have been from 30
to 50 feet, with a possible average of 35 or 40 feet. Using the latter
value, we get an area of 1,257 square feet or, rounding it off, 14,250
square feet.

We next consider the interior Pomo, those inhabiting the Russian
River watershed and the Clear Lake Basin. For the coastal Pomo there is
agreement that the standard house sheltered a single family. Regarding
the others there is room for argument. Our knowledge of this segment of
the tribe comes from several accounts of conspicuously large villages.
Of these the most famous is Senel, or Shanel.

Stephen Powers (1877:168) described this village and drew a careful
map of it. There is no reason to question either his accuracy or his

veracity. He shows 104 houses (plus five assembly houses). The dwelling
houses contained 20 to 30 persons each, which would mean a population of

2,000 to 3,000 persons. However he also says that the five asembly houses
could hold 100 men each, implying a population of 1,500 or 14.42 persons
per dwelling house. The latter number is supported by an informant of
Stewart (1943:45) who said that Senel had 1,500 people.

Stewart also mentions that his informant said the Hopland tribe
had 1,500 people. Kniffen (1939:375) mentions the village of Kacha in

Sherwood Valley with 12 houses and 125 people, or 10.4 per house.    Merriam
(1955:41) speaks of the Yokaia Pomo who built houses 40 to 50 feet in

diameter and whose ceremonial houses would hold several hundred people.

Gifford's (1926) notable study of Cigom, on Clear Lake, showed this village
in 1850 to have had 20 houses, with 2.35 families and 11.75 persons per

house.  The aboriginal value was probably higher, let us say 2.5 families
and 15 persons. Barrett (1908, passim) describes 34 villages in his day,
which had an average of 7.35 houses per village, and this inc luded the

remote hill settlements. Kroeber (1925:241) says that the Russian River
Pomo had houses which "often sheltered several families" and that the

Clear Lake Pomo had elliptical houses with a long axis of 25 to 30 feet,

although certain "poor and old people, and individual families were content
with an humbler abode."

It is clear that there was a great range of house sizes and types.

They held from one to four families, but on the average can be regarded as

37

containing 2.5 families or 15 persons. The floor space must have corres-
ponded. If we think in terms of circular houses, Merriam's figure would
be 1,595 square feet. Kroeber's oval house of 25 by 30 feet would be

about 600 square feet. The Archaeological Research Facility records show
an average of 124 square feet as the mean of the pits at four sites on
Clear Lake. A reasonable, even if not wholly satisfactory, compromise
will be 900 square feet for the area.

The number of houses per village presents a similar problem. The
Archaeological Research Facility records show a mean of 5.0 house pits

for 42 sites. Barrett found 7.35 houses as an average in the decimated
villages of 1908. Gifford saw 20 houses at Cigom. There were 12 at
Kacha and 104 at Senel. The problem of decision here resembles that

encountered in the Central Valley; namely, to allow for numerous villages
with several big communal houses without overemphasizing this element.

Barret's figure of 7.35 houses per village may be taken as a minimum in
spite of the archaeological site records. If we have 34 villages with
7.35 houses each, plus one Senel, one Cigom, and one Kacha, we have an
average of about 11 houses per village. However there were many more
big villages than these three; hence we ought to increase materially

the probable number. An estimate of 15 houses per village is as close
as we can come.

The site area can be computed only from the Facility records

which are reasonably complete for the Pomo territory. The mean of 182
sites is 24,750 square feet.

The valley Wappo, which includes the groups sometimes designated
Western, Central, and Southern Wappo, have been thoroughly studied by
Driver (1936). His analysis of Alexander Valley shows six permanent

towns, of which four had 40 houses each and two had 9 and 11 respectively;
the average is 30. Specifically, in 1870 the village of Unutsawaholma had
11 houses, 21 families, and 92 persons. Thus the average number of fami-
lies per house was 1.91, persons per house 8.35, and persons per family
4.5. Aboriginally, Driver's informants said, there were 17 houses, and
the family number certainly can be taken as 6.0. This gives, at 2 fami-
lies per house, 12 persons per house. After making various correlations,
Cook (1956:122) assigns 23 houses to each village of this group, a figure
probably nearer the truth than the 30 suggested previously.

One factor is still undetermined. What other villages once existed
in-the area besides the six studied by Driver? Driver himself mentions

nine others, most of which he designates as camp sites. However Merriam
and Barrett have both cited names which may have been those of villages

unknown to Driver's informants. Hence the actual number of villages may

38

have reached as high as ten or twelve, many of them small.    If so, we

shall have to reduce the probable average number of houses per village
from 23 to, say, 15.

One other minor bit of evidence is pertinent.    Merriam, in his
village lists (see Cook 1956:124), mentions the village of Mi-yahk-ma,

which occupied the area surrounding Calistoga. Barrett (1908:269) also
mentions this village plus three other smaller ones nearby. It is

likely that here was a tribelet with one large and three small settle-
ments. The total population may well have been close to 500 persons.
Hence, at 12 persons per house, the total number of houses would have

been 41.6 and the average for the four villages 10.4. It seems reason-
able, therefore, to accept 15 as the mean value for houses per village.

With respect to house size, Driver (1936:189) says that the

houses were "typical of the area," and were oval in shape with the long
axis up to 40 feet.   If the short axis be taken at 30 feet, the area is
942 square feet, not far from the area estimated for the Pomo.    Site
size can be evaluated only from the records of the Archaeological

Research Facility. A mean of 48 sites in Napa Valley gives an average
of 19,350 square feet.

The remaining region which contained multifamily or communal

houses and for which we have area figures is Santa Rosa Island. Apart

from the fact that the inhabitants were tribally related to the mainland
Chumash and hence might be expected to have similar customs regarding

dwellings, we have independent evidence from the archaeological work of
Phil C. Orr. In his descriptions of the island sites Orr (1951:221)

distinguishes between six classes of sites. With respect to size, there
are actually only two types: class 1, large house pits; and class 2,

small house pits. The former are said to be 40 to 60 feet in diameter,
the latter 10 to 20 feet.   Using the midpoints of the ranges, the

respective areas would be 1, 960 and 177 square feet.   However, it might
be permissible to use the lower estimate for class 1, and say 40 feet
in diameter and 1, 260 square feet in area.

The difficulty in using Orr's classes lies in tying their occur-
rence to the corresponding site areas. The Facility has records of

sites on Santa Rosa Island which include seven sites with house pits

which are specified as belonging to Orr's class 1, and 48 sites where the
pits are class 2. The respective geometrical means for area are 25,200
and 25,000 square feet, almost an identity. However one site which con-
tains nine class 1 house pits is stated in the record to have an area of
approximately 5, 300 square feet, impossibly small and undoubtedly an

error. Deleting this site., the mean of the remaining six sites is 32,750

39

square feet. Even with this correction-, the difference in area is not
consistent with what one might anticipate with sites containing such
diverse types of dwellings.

Concerning the number of persons per house, we have no information
whatever except that class 2 was clearly a single family structure.    For
the others we might use the Channel Chumash value of 2.5 families and 15
persons., but we have no real warrant for doing so.

The counts from the Archaeological Research Facility records of
number of houses per village give averages of 8.2 for class 1 and 12.7

for class 2. These are both lower than would be expected on the basis of
general description and knowledge of the Channel Islands. On the whole,
we are forced to seek some intermediate solution rather than try to

formulate different sets of numerical values for two utterly diverse

house systems existing side by side and from site to site within a very
small total area.  Consequently we consolidate the data according to the
observed frequency of occurrence of the two types of house, and for sim-
plicity place the results in Table 8 among the regions characterized by

having predominantly single family dwellings. This procedure would have
to be adopted regardless of whether or not Orr' s classification is
accepted as valid.

The records show seven sites of class 1 with 57 house pits, and
23 sites of class 2 with 292 house pits. The floor area of a class 1

house is taken at 1, 260 square feet and that of a class 2 house at 180

square feet. There are in all 30 sites with 349 house pits, an average
of 11.6 per site or village. The average floor space for the 349 houses
would be 356 square feet.  If the big houses held 15 persons and the
small ones six persons, then the average number of persons per house

would be approximately 7.5, and the number of families per house 1.25.
For site area the group mean is close to 25,100.

Part 3

The figures for the 26 regions featured by single family houses
(including Santa Rosa Island) have been repeated in Table 9 and with

these have been placed the final estimates for the four regions in which
communal houses predominated. These are all mean values for the regions
concerned, and when we discuss them we are of course operating on the
interregional or intertribal level, not on the intraregional or local

level. The data will be examined briefly, and detailed analysis will be
reserved for future publication. At this point we wish to determine in
a preliminary manner whether there appear to be any close associations
or systematic differences among the thirty areas.

40

The three key magnitudes shown in Table 9 for each region are:

(1) the total floor space-equal to mean floor space per house multiplied
by mean number of houses per village; (2) the mean village or site area;
and (3) the mean estimated population per village. When we equate the

first with the third parameter we get on a regional basis the equivalent

of Naroll's (1961) function: log total floor space against log population.
The result is shown in Figure 8. The log-log relationship is very clear

in graphic form. The corresponding value for r is +0.945, which is highly
significant. For the slope b we get + 0.622 when we take floor space as
the x-axis, and +1.437 when population is used as the abscissa. There-

can be no further question, therefore, that the logarithmic function holds
with floor space for interregional means as well as for scattered single
village data as reported by Naroll.

The analogous formulation for log mean village size or site area
against log mean population is seen in Figure 9. Here the picture is

very different. Let it first be remembered that high correlations and a
logarithmic relationship were found internally for log village size

against log population in three regions, viz. Tierra del Fuego, Klamath
River, and Santa Barbara Channel.   Figure 9, illustrating 30 regions,

shows rather wide scattering of the points, and r is only +0.374 a value
of distinctly mediocre significance.   New factors have entered to disturb
the logarithmic function.

It is evident that the points on the graph are segregated for

three regional complexes:   the dots designate the hill-coast environment
with single family houses; the crosses denote the five definitely desert
people; the circles indicate the four areas in which the houses were
principally or entirely multifamily.

If in the calculations we now remove the five desert areas., r

becomes +0.747, and if we also delete the four multifamily areas, the

value is-+0.687. These are not impressive, but are definitely signifi-

cant values. Finally, if we calculate r for the four multifamily regions
alone, the value is +0.986, even for three degrees of freedom highly

significant. The provisional conclusion is that in order for the loga-
rithmic relation to be demonstrable for village area against population,
certain edaphic and also probably social factors must be held constant,
and the entire territory must be homogeneous with respect to these.

Another mode of setting forth these distinctions is seen in

Figure 10. Here are plotted the points obtained by taking the ratio of
the total floor space to the mean site or village area for each region.

The numerical values (multiplied by 100) are given in the last column of
Table 9. In Figure 10 both the ratios directly and their logarithms are

41

plotted simply as a continuous distribution, beginning with the smallest
and ending with the largest. It is very striking that the first four

points, which are the smallest, represent four desert areas. Then comes
a sharp discontinuity, broken by only one point by the Achomawi, who

lived under arid but perhaps not full desert conditions. There follows

a row of 16 points, all of them pertaining to single family house regions
of the foothills, coast ranges, and coast. Finally, after a very wide
break there are the points representing the regions with multifamily

houses. Again it is clear that as far as village area or space is con-
cerned, there are three very distinct groups, whereas when floor space

or house area is considered, all the regions form an unbroken exponential
continuum. The interpretation of these findings must await further
analysis.

42

TABLE 1

Values of r, b, and a for Waterman' s (1920) Yurok Data*

No. of    Pop.   j Area      Indep.  I    r   I    b                 a

cases  | method   method   variable           I |         (log) | (sq. ft.)

17        1        1       pop.     +0.895    +1.393   +1.288     1, 940

1        1       area              +0.544    -0.458       --

17        2        1       pop.     +0.824   +1.207    +1.587     3, 870

2        1       area              +0.564    -0.604       --

17        3        1       pop.     +0.751    +1.023   +1.746     5,575

3        1       area              +0.551    -0.560       --

17        1        2       pop.     +0.933    +1.593   +0.813       650

1        2       area              +0.548    -0.319       --

17        1        3       pop.     +0.896   +1.167    +1.508     3, 220

1        3       area              +0.691    -0.849       --

17        1        4       pop.     +0.955   +0.771    +2.346    22,200

1        4       area              +1.163    -2.627       --

17        1        5       pop.     +0.900   +1.117    +1.896     7, 870

1        5       area              +0.626    -0.886       --

17        1        6       pop.     +0.818   +0.850    +2.423    26,500

1        6       area              +0.790    -1.587       --

15        1        2       pop.     +0.894    +1.197   +1.227     1,685

1        2       area              +0.603    -0.515       --

* For methods of calculating population and area see p. 2 ff. of text

43

Values of r, ba and a

TABLE 2
for Lothrop' s

I No. of I
I cases I

-

(1928)

r  I

-I

=

Three Groups of

bh  I     a

(log)

Sites

I

=

-

Group

I. From Fig. 98, p. 182

and table, p. 184

II. From Fig. 100, p. 185

and table, p. 187

III. From Fig. 101, p. 186

and table, p. 188

I

- . .

50
10

5

+0.799
+0.652
+0. 755

+0. 907
+0.697
+0.912

+1.431
+2. 234
+1.600

a

(sq. ft. )

2, 700
17, 150

3, 980

Total                     65     +0.689   +0.792  +1.672      4, 700

44

Notes to Tables 3, 4, and 5

The various sources consulted list in part the same and in part
different villages or sites. Consequently there must be considerable
unavoidable repetition in the tables.

Table 3 lists the 56 recent or "Canaliino" sites mentioned by

D. B. Rogers (1929) in the text of his work. Eight of the names are in
CAPITAL letters. These are the ones which were drawn by Rogers on his

maps and for which house counts were also made by the Portola expedition.
Those sites whose names are underlined are shown on Rogers' maps but no
house counts were made for them. The remainder of the sites are those
for which there are descriptions by Rogers but for which there are no

further data. For the first group of eight sites the house counts are
given as derived from Tables 4 and 5. For the second group of 26 sites
the house counts, as obtained by interpolation in Figure 7, are placed
in parentheses. For the third group, 22 sites, where the average (see
p. 19) is always taken as 37, the house count figures are omitted.

Table 4 shows the data in detail for 16 sites mentioned in the
accounts of ane Portola expedition. The populations and house counts
are as indicated in the references and notes which follow the table.
The areas are from Rogers' maps.

Table 5 summarizes the data from Table 4 and presents the final
estimates offered here for area and population.

45

TABLE 3

"Canaliino" Culture Sites Mentioned by Rogers

Rogers'   I                             I     Area      I   No. of

No.     I   Site Name                  I   (sq. ft.)  I   houses

1

6
7
8
9
13
17
19
20
22
26
27
28
29
30
31
33
35
36
37
39
40
41

42)
43)
44)
45
46
47
51
56
57
58
61
62
63

SHUKU

CARPINTERIA
MISHOPSHNOW
Sandyland

Lyman No. 1
Kolok

SHALWAJ
Miramar
Swetete
Kashwa

.Siuhtun
Amolomol

Burton Mound
El Banos
Mispu
Mispu

Mismatuk

Barger No. 1
Barger No. 2
Ushtahash

Cieneguitas
Hope Cliff
San Marcos

Three sites

east side of

Goleta Slough
Twin Mounds
NESCALITAN

South Side Goleta Slough
Campbell No. 3
Williams No. 3
Williams No. 2
Williams No. 1
Stow

Larson No. 1
Larson No. 2

435, 600
162, 250
682,430
101, 160
103,800
46,000

45
38
110

(30)
(30)
35

31,500
85,400

29,000
46, 140
43,900
33, 210
141, 100
484,700

1,300,000

132,000
158, 900

83, 200
197,100

58,000
53, 200

(12)
(20)

(11)
(16)
(15)
(12)
(39)
(98)

110
(36)
(41)

(25)
(50)
(19)
(17)

46

TABLE 3 [cont'd.]

Rogers'    I                               f     Area          No. of

No.          Site Name                   |   (sq. ft.)   |   houses

64          McCaffery                           80, 000       (24)
71          Winchester No. 3                    65, 800       (21)
72          Tecolote No. 1                     120, 000       (34)
73          TECOLOTE NO. 2                     100, 000        22
74          Tecolote No. 3
75          Tecolote No. 4
76          Eagle Canyon
77          DOS

78              PUEBLOS                        536,000        120
80          Los Gatos

81          Las Llagas No. 1                    79, 600       (24)
82          Las Llagas No. 2                    28, 800       (11)
83          Las Llagas No. 3                    73, 300       (23)
84          Ajuahuilashmu                      143, 500       (39)
86          Refugio No. 1                       16,300        (7)
87          REFUGIO NO. 2 (KASIL)              127,000         80
91          La Quemada                         69, 600        (22)
93          Amolar                             82, 400        (25)
94          Alcatraz
95          Alcatraz

97          Gaviota No. 1

47

TABLE 4

Data Concerning

(For sources

Villages Mentioned by
refer to notes on pp.

Explorers
49-50)

Area        Popu-       No. of     Note
No.*  | Village Name        |  (sq. ft.)     lation     houses       No.

1      Assumpta                              400         30         1

300        30          2
400        30          3
2      Santa Conefundis                                   8         4

3      Shuku                   435,600                              5

60         6
300+       30          7
400+                   8

4      Carpenteria             162, 250                             9

400        38         10
300+       38         11
5      Mishopshnow             682, 430                 100+       12

600+                  13
6      Shalwaj                  46, 000                            14

500                   15
500        40         16
500        30         17
7      Mescalitan            1, 300, 000                           18

100+       19
100+       20
800        80         21

8      Tecolote No. 2          100,000                             22

20+       23
9      Dos Pueblos             536, 000                            24

1, 000+                 25
1,000+                  26
1,600       120         27

Numbers run consecutively for convenience only

48

TABLE 4 [cont'd.]

Area      | Popu-       No. of     Note
No.*   | Village Name       |  (sq. ft.)   | lation      houses      No.

10      Kasil                   127,000                             28

400+        79        29
400        80         30
800        80         31
11      Gaviota                                300        52        32

50        33
300+       50         34
12      San Serefino                           200        24        35

130        25         36
13      Santa Ana                                         20        37
14      Santa Teresa                                      24        38

150        50         39
15      Espada                                 250+       20        40

250        20        41
200        30         42
16      Los Pedernales                          70        10        43

60        10         44
60                   45

*

Numbers run consecutively for convenience only

49

Notes to Table 4

Ref .
No.

No.  Village Name

1   Assumpta

2    Santa Conefundis

Note

1    Bolton, H. E.. 1927:160.  Hereafter referred

to as "Crespi."

2    Smith, D. E. and F. J. Teggart (eds.),

1909:55. Hereafter referred to as
"Portola. 1

3    Priestley, H. I. (trans.), 1937:25.   Here-

after referred to as "Fages."
4    Portola, p. 56.

3    Shuko,

S
6
7
8

Rogers, 1 929:41.
Crespi, p. 161.
Portola, p. 57.

Fages, p. 26: "more populous than Assumpta."

4    Carpinteria

9    Rogers, 1929:46-48, Map 2. This is probably

the southern site, east of Mishopshnow.

10    Crespi, p. 163. Inhabitants "no less" than

Assumpta. Some of the houses "with many
families."

11    Porto1la, p . 57.

5    Mishopshnow

12    Rogers, 1929:48 ff.,

p. 53.

13    Fages, p. 26.

Map 2. House estimate,

6    Shalwaj

7    Mescalitan

14    Rogers, 1929, Map 7. Our estimate of area

according to scale shown.
15    Crespi, pp. 164-165.
16    Portola, p. 57.

17    Rogers, 1929:80.

18    Rogers, 1929, Map 16. Our estimate according

to scale shown.
19    Crespi, p. 167.

20    Bolton, H. E.., 1926. Hereafter referred to

as "Palou."  On p. 156, concerning the

island, "a large town in which were counted
more than one hundred houses."

21    Portola, p. 57.  Says 7 towns, smallest 20,

largest 80 houses.  In latter "we have seen
about 800 natives."

50

Notes to Table 4 [cont'd.]

No.   Village Name

8    Tecolote No. 2

9    Dos Pueblos
10    Kasil

11    Gaviota

12    San Serefino

Ref.
No.

22
23
24
25
26
27

28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45

Note

Rogers, 1929, Map 24.    Our estimate of area

according to scale shown.
Rogers, 1929:201.

Rogers, 1929, Map 25.
Crespi, p. 169.
Palou, p. 157

Portola, p. 57.    Sixty houses and 400 persons

in each village.

Rogers, 1929, Map 28.
Crespi, p. 170.
Portola, p. 59.
Fages, p. 28.

Crespi, p. 171.
Fages, p. 28.

Portola' p. 59.
Crespi, p. 172.
Portola, p. 59.
Crespi, p. 173.
Crespi, p. 174.
Portola, p. 59.
Crespi, p. 175.
Fages, p. 30.

Portola, p. 59.
Crespi, p. 176.
Fages, p. 30.

Portola, p. 61.

13
14

Santa Ana

Santa Teresa

15    Espada

16    Los Pedernales

51

TABLE 5

Summary of

Data for Sixteen Channel Sites

Listed in Table 4

I                  I    Area     I  No.    j No.     I Persons

No. I Village Name     I (sq. ft.)  I Persons  I Houses  I Per House

1
2
3
4
S
6
7
8
9
10
11
12
13
14
15
16

Assumpta

Santa Conefundis
Shuku

Carpinteria
Mishopshnow
Shalwaj

Mescalitan

Tecolote No. 2
Dos Pueblos
Kasi 1

Gaviota

San Serefino
Santa Ana

Santa Teresa
Espada

Los Pedernales

435, 600
162., 250
682, 430
46, 000
1, 300, 000

100, 000
536, 000
127, 000

400

440
400
660
500
800

1, 100

800
300
200

30

8
45
38
110

35
110

22
120

80
50
24
20
24
20
10

13.3

9.8
10.5

6.0
14.3

7.3
9.2
10.0
6.0
8.3

12.5

7.0

250

70

Average

9.5

52

TABLE 6

House Sizes in Central Valley and Adjacent Foothills

(Data are for single house pits unless a range or average is stated)

Dimensions*

Site No. j Probable Tribe I Habitat              I   (feet)    I Notes

Teh- 13
Teh- 22
Teh- 27
Teh- 34
Teh- 53
Teh- 95
Teh-191
Teh- 201
Teh- 235

Wintun
Wintun
Wintun
Wintun
Wintun
Wintun
Wintun
Maidu
Maidu

Wintun
Wintun
Wintun

Western hills

No. of Red Bluff
Western hills

Valley flat land
Valley flat land
No. of Red Bluff
No. of Red Bluff
Eastern hills
Eastern hills

Gle-
Gle-
Gle-

8
10
14

Western
Western
Western

hills
hills
hills

4 to 14
14

4 to 5
7 to 8

8 to 15
30 by 30
12

8 by 11
6 to 8
12
40

8

13 by 15

8 to 12

10 average
33
25
40
35
25

Sweat house
Dance house

Col- 10    Wintun
Yol- 33    Wintun

Valley flat land
Western hills

But-
But-
But-
But-
But-
But-

4
6
65
66
84
94

Maidu
Maidu
Maidu
Maidu
Maidu
Maidu

Eastern
Edge of
Eastern
Eastern
Eastern
Eas tern

hills

valley
hills
hills
hills
hills

Only 1 pit seen
Dance house

Unspecified, 1 pit see
Ditto

Dance house

Eld- 73    Maidu

Sac-ll    Valley tribes
Sac- 117   Valley tribes

Ama- 40    Maidu
Cal-112    Miwok

Eastern hills

Valley flat land
Valley f lat land
Eastern hills
Eastern hills
Western hills
Western hills
Western hills
Western hills

30
25
25
10
12
33
47
35
14

Only 1 pit seen
Only 1 pit seen
Only 1 pit seen

Sta- 129
Sta-130
Sta-130
Sta-130

Yokuts
Yokuts
Yokuts
Yokuts

Only 1 pit seen

1 large dance house
1 sweat house

2 dwelling houses

* Unless otherwise stated all figures represent diameters

53

TABLE 6 [ cont'd. ]

I Probable I                    I Dimensions I

Site No. I  Tribe    H Habitat           I   (feet)   I Notes

Mer- 43    Yokuts
Mer- 58    Yokuts
Mer- 89    Yokuts
Mer- 89    Yokuts

Valley flat land
Valley flat land
Valley flat land
Valley flat land

Eastern lower hills
Eastern lower hills

Mrp- 2
Mrp-  2
Mad-  9
Mad- 29
Mad- 37
Mad- 37
Fre- 57
Fre-126
Fre- 186
Fre-355
Tul- 208
Tul- 225
Tul- 234
Tul- 234
Tul- 241
Tul- 243
Tul- 244
Tul- 245
Tul- 247
Tul- 248
Tul-249
Tul- 260
Tul- 276
Tul- 277
Tul- 284
Tul- 287

Miwok
Miwok
Miwok
Miwok
Miwok
Miwok
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yoku ts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts
Yokuts

Eas tern
Eastern
Eastern
Eastern

hills
hills
hills
hills

18 to 20
43

45
35
52

14 to 22

8
12
18
33

6
20
13

16 by 20

if several"

"dance house pit, If

plus 20 others

1 pit, others doubtful
Ditto

1 dance house pit
15 others

"former sweat house"

"present sweat house"

Eastern hills

Mountains

Eastern hills

Eastern
Eastern
Eas tern
Eastern
Eas tern
Eas tern
Eastern
Eastern
Eastern
Eastern
Eastern
Eas tern
Eas tern
Eas tern
Eastern
Eastern

hills
hills
hil ls
hills
hills
hills
hills
hills
hills
hills
hills
hills
hills
hills
hills
hills

14.7

9.8
13.1

6.5
11.5
13.1

8.2
11.5
11.5
11.5
16.4
13.1
14.7
11.5
11.5
14.7

" large"

[Summary of Table 6 follows]

54

TABLE 6 [cont'd.]

Summary: Taking the Average Size Per Site of Unequivocal Dwelling

House Pits and Averaging Sites by Region

Region and Tribe          No. of Sites           Average Floor Space
Hills Wintun                    7                        83.9
Hills Maidu, Miwok              8                       119.4
Hills Yokuts                   21                       135.7
Valley, all tribes              4                       156.4

55

TABLE 7

Data from Documentary Sources

Source    I
Zalvidea
Zalvidea

Abella
Duran
Duran

Estudillo

Work
Work

Warner

"Trapper''

Dixon
Dixon

McKern
Gayton
Gayton
Latta
Latta

Kroeber
Kroeber

Habitat
Hills
Hills
Delta
Delta
Delta
Hills

Valley
Valley
Valley
Valley
Valley
Hills

Valley
Hills

Valley
Valley
Hills

Valley
Hills

Floor Space I Number of I Houses per I

(sq. ft.)  I Inhabitants | Village    1

41            9
180           28
408         1000           60

40
803-4520                   35

1250          210

Notes

Cuia, 1 family

Sisupistu, 1 fam.

Tauquimnes, com.
Ochejamnes, com.
Chucumnes, com.
Chischa, 1 fam.

608., 000 sq. ft.

Maidu, 1 family
Maidu , 1 family

500-1000

28

40-50

50-100

100-1200

301-1216

53- 211
396-706

78-133

12,000 (?)

180-300

113- 177

1-20

Type 1
Type 2

Communal
1 family
Communal
Communal
1 family
Communal
1 family

56

TAB LE 8

Data from Documentary Sources

Detailed data concerning house sizes., village or site sizes., and population
for 30 California tribal or geographic regions., together with a final esti-
mate for each region. For discussion refer to text and notes which follow.

1  1I         I WC 1  IH
1  1   1    I    11C4 1

W    W  l         I                  1I *   I 1 I ~ ~ I

Cd1W11'-I 0               0        (1$-  1 0 >  0  >1I  >14C  iq4 '.

0           4 4I6- ;At 6nr a)6 uI              (1 tl  ct H *    <N 4Jr4

XtS)r- (IIZ                   ) Wj l  wf  ~   afl V CZ CdC iLtflLZ 4

Q)r          E;I   1             6 81I                    I   I

61               I    I  AI    I    111.7>1  0 >1  >1  14    ;I I-

1lY ro1          147351  1  I    I.5  H1  I.3   6 1  11   1    I   1

81      1          I   I    I    I   11  6.' 1 45!   11   11   I   1I
41            1        I I  I    I      1128,900! 1  1    I   1
2Wiyot17.511                175       18.7.1   1    I1    11   I   1
613 1          11  I    I    I        11 61.5 1  11  11   1    I   1
711e.431           1    19.21    19. 1  I  I   I    1     I1   I   1

111                  11      9.1 .1       111        1 II I
96!            119.1       119.116.o           Il   I8 9001     II
181            1     51   1  1s'      111. 01   I   II 125451  I    1

2011  Estimate_l -.2439_7.51 1 1 7.5215283.8117.61 60 113424.25?.ST0U12631.24LA21AI
12laryok       11  17.51 117.51      11 I.7     1    1     111I11
1311           ll......1  I__  ]I__  I  I__  [_ _  1.1  6.' 1 1 1.........i_ _ _   I ___

2311 Esimt    L...1 9.5J 1     7 .51[ 1    I 4.J1JjJ      11   I    I

24IHup         11    7I.   17.95      1  II   I     1I     11  I    1
2511          11    1 6.411  1 6.41   1192.81  11   1      1  11  I

.261           I I  I    I   I    1   114.411   L  11  1        I

57

TABLE 8 [cont'd.]

11

11

211
Pll

$401
w            W    r-
0  11        $4

11 1 1 1  1  1     1 1 1  1  1  1
1  1  1 11  1  1        K  1CO
10  1  1  1  1I 1  1  K1;.1  2 11  1  1

U) u  I  I1I I  .1  K  K21 S. K1 K1 11X I? 1

$4  V} U)  c 'H) U '   )   U)  (  (  Q U  $)  r-4  'v-I
11 o Ia 10XI  10 U 0  11 4  1  (1   U '  r 1v  U1

1g  6IA K4            cdr41  '$4

C O  (3)  O  -1 Ca ()l UI H  a 4 o1 &4  1  tC -'

1  I  H    I C 1ut1z  ?Z 2 LZ 2 kI  ln 1  X

CO  $4 t4 o4  01  o4  $41  $41H@1   Jl 1 CZ  $411  CZ  .4  1 4  141 t1  -W r-

I11        -  ) Q  ) (3  ) -4Q  ) C

- a

271 p  [cont'd] 1          I l                "1.5     1     1       11   I l   1

2811                 1     1     I    1     1110.61    11   111         1       11
2911            11  400    I     I    1     1  1                     1    1

13011                             _ 11  _ __ I.  . .Ii1  L.3.  -11  _      I  -_-
A3'111 Estimat  il 4003 7.0 1 1  [__ _7.0157.11110.91- 76 il436od    1     I    I

..321IChilula   3] -   ___  3  7. 51  1  [  7. 51  -__   [  7. 0[  -.11  -- 1__   L ______.  1

-3311  E stimate  I I__ 3 7. 5 1  [1__ 7.51  E- 7. 0[ 52 3]I__ 1] ___ L....... I__ 3

34||Shasta      ||  3501   1  1- 21   1     1|  2. 5 |  1    11      l    I  1l

3511            11  2661   1     1    1     113.01           1       1  1       1

jjj,,,3611 ..... --_II...3   176] ___  -1   -.--A   j  _ 1  7. 0j  ---__ 3] __501__.

)2iimat    3i.4       .31     O_.0330       6.0 1 48 11158i?A18, 950113158L123941  8.4
38|lAchomawi       961     1     1    1     N     I    1     11     11     l

3911            1      1211  1   1[        11 5.3&    1L    1114,000      .

-4011  Estimate  3 110]o -6-.03  1  .6.01]18.311 5..31 32 3]_ 5831114, 0001126411 43811  4. 2
411I-, -doc   -   --11  118  1  --     1    11  5.41    I   -.I2 ,1 01   - L  - 1

4211  Estimate  3]_1181 6.03l 1  [6.OILQ9? 11 j2jt3 32 3]_6371[.p27 jhpi9I01150191 8471  2.4

1No.aiute      -I     1     1    [     I    11 3.6L    11   .11_6150011       11

1    Estimate  || 1001 6.01  1   | 6.0116.7ti  3.61 22 ttP .360a61,5001170841279511  0.6
45||Athabascan  11   |  6.01  1  |  6.01    11  5.0j 1|       11    11    1 l
461                  1 4.411     1 4.41     115.01     N

4711            11  125 1        1    1     1     I    1     11         1       1

.AL.......l           I     I     1-   [     Ijj6.31  I- ......6il.L39011Qil__

04911-- Estimate  3]_125j 6. 0J1 1  3_6. 0120. 83] 5.01 30 31_62511.i612.390111278.1121L1- 9.8
501           11    1  6.01  1  1  6.01    1125.01    11   11      11    I     11
5111           11    1  6.01  1  1  6.01    116.31     11   11            I     1
5211          1N   78 1    I     I    I     1       1        1          1

5311           11  314 1   I                N 1              1       1    1     1
5411            I    I     I     I    I      1-  251

55             I  _  78[   1     i      3   13  5.11   11   11] 4J73011 ..  --. -  1 _
5611  Estimate  11- 10031 6.01 1  3_6.03-16.73]_6.0136i&JLfi6001IJ.Z..%2jJ 1 7881 1311- _12.7

58

TABLE 8 [cont'd.]

I  I  I l

11
t11
t11
?R11

0

Q)11

Q) 11
Pfl1

11
11
11
11
11
11
11
11
11
11
11

0

U1
CO

0
0

0

49

I

0I
o

I

41

4-1

49

1n

o

IO
co

0
0

E-H

4-3
.-H

rH-

CTO

0

co

r-4
.r4

0
0
z

cn
~4

I    1   J

I Co  IOIH X
01 X X U Co g

.21t21 O 1t ~

J 0    S4 f l  >

X   o  a)  r  a) o

Z 0 m1 4 GIZ

Co

0

04

L4-4
0

z

11

Co

H

fa4

0

11 0

r- l

41 44

1 CO

r 00

I aZ

11

tIl

i1

*71

04
COi
0
CO
z

l1I-

l1
l1

411s 1,

11   11  D   I

110   1

*n11 1    1

w   bo
* 1- Xr- X4

M 4 v   Q

> I IC  041

11

11

o    Il

0    11

0 b

1 1

0  r4

411

ut 41

C)

CU
Q4

0
0
'4-

co

4-3
0

4-I

0

.r4

4- I

'U

'U

0)
(U
0)

CU

H
*rf

0
JJ

0
0
H-

x

571 Porno,  coast  11 1  6.01  1  1  6.01  116.01   11    1  1        I    11
58111              1241  6. 51  1- 21  6. 51  11  1  1    1           11 I 1
59 111             125 1   1    1    1    1    I     1    11     1    I    1

6 1_1__0                 1    1-L   -   1  11  4.3j  ]]  ]]  2 ,0  20f1_  1.. L

6111  Estimate ]] 1501 6.01 1   1_6.01_25.011 5. 11 31 1176511 4, 27011 8401 _13811 17.9

]]Pomo, interior                See Table 9

ILwappo,-valley      _____       See Table 9        _______-_____

621IWappRo, hills 11 254J  1- .- 11 9...L.31[      1]    1119,000kl   I- 1

6311  Estima.te .i 200] 6.0-1 1 j_ 6.0J33.311 9.31 56 11?i26 111 9, 00011204 21 34011 9.8

-6411Coast Miwok-,1 fi___            j    J[.I 5.4[  1111-. 7,03011    1- --

651   Estimate J]    J 6.0[ 1 j 6.0]     J[ 5.41 32 L ...1a.?7.q2pJlupnL1_22011

66llWintu        I1   1  5.01  1  j  5.01  11  1    11    11          I    1

6711             11   1  6.61  1  I  6.6 1  11  6.0 1  I1111          I    1I
-681[           [[ 111 iii __ L  ..--I...  I___JI-.~.   4.4j  ---I[ -1.......111, 20011 - I  1

-691j  Estimate  I I1111 6.01- 1. 1 6.0j118.5ff_5.41 33 [11 6001111,12001120741[ 33911 5.3

7O1INo LWintun  LL?1291- --J    J    I    11. 7.01 1.I1  1116, 80011  1...JL
7111  Estimate   I 11291_6.01 1-_J 6.0121.5k1_ 7.01 42 J1 903k116, 800112400t -40011 5.4
-721lC. Wintun    - .IL.AZ1  1~.     1]]I_- I16.3j  11    I 8, 03011  jJj

731  Estimate IA 82j 6.Oj 1-- J 6.0113.711 6.31 39 -.1 -518ff 8,0301112741 20611 6.2
-74kLSo. Wintun  ff_-1391[j  1     1 I     1 6. 3]  li ---14,03011-   L   -.I.J...

7511L..Estimate.__f 1391_6.01 1-   6.0[23.2fl 6.31 39 frts?11,001b 1j&2713611 6.2
761IMaidu        11  1921  7.51  1  1  7.51  11  3.41  1I  1     1    1    11
7711            11    I    1         1    1  7.01  11    k1     k    I    1
781                "el 111      I  I    I    I 1 I  1     I11         I    1

7   9 1         11  123 1  I    I    I    I1   I    11    11     11   I 1  1
80111               831 6.01 1  1 6.01    116.51     11  11      11   1

-81ff           J]1064       1           LiL .  --  1 6.31  kili113,431liLi

fi3Jj21  Estimate ]125J- 6.01 1  L 6.0120.811 6.6j -40 ]]_827iliLii30il20351 3361[ --6.1

831Miwok         I  1231j       I    I     I   I          I      I    I    1
8411               1       I  1    1  1              1   1      1

85811          1] 122]     1 __ L   ..1 _11 5.41     11  1114, 15011

8611  Estimate fi 1231 6.01 j1    6.0120.511 5.41 33 il 6641114,1501126201 42911 4.5

59

TABLE 8 [cont'dc.]

.. , . @I

11
1I
I1
II,1

I 1
II1

11    1     1
11    1-     1
11    1      1
11    1     1-

Cod

11   1 6   I

H4 -I     a

cn i 44 C

Q)     O  4

U'    * I
O1 :j O

0 1=      Q)

r-I C)
(1)-40
I cO

Eq 0

C)
0)
.r4

CT
4~4
4-1

0
Z

1    I    1    1

I      11    1
1    I   .11   1

w   IGzW I4

I    O :n   mi

0)1 0  )  C 0 CI  g

QI 6 QI U! 6 W
H41  Hi   HIZ H11

aiZ   __14  1rI  1111Z  _P1

t11t 11

sq- 11A1
coi w 11ow1

Z  H1 CZ HI

.4  r- ll  4-  r-

r- 401   a
44 i- 41  ;

4-i
1-4.

01)

'-40)4
01)  'H

11201:011
11 *1 .11

11 ;U     U ! X1

r-4 H4I
44 > 44

1t   In  11-1

0)
U

Cac

14

0 0)

4-4 r-4

C >

4l-i

0 0

0 0
.r4 C
41J r-4

Ca

87Iono   11      jj]     5. 14  1  [ 5. 1  I  1J 3.6 I   1 1.11       11I_        11
8711  Estimate   11 125] 5.I   1    115. 24.[ s51[ 3.61 _18 11 45411  1L11         11
88IlYokuts,  hillisl  1231   1     1     I    11    I     11    11      11    I    11
8911             11 136   1     11     1    1        1  11     1        1     I

9011             11  1331  6.01  1  1  6.0 1  11    I     II   11      11     I    11
jjj1              11 1311     1     I j   I    11 3.81     11   1111,72011    __I

9211  Estimate   11 1311 6.0  1   1 6.01 21.811 3.81 23 || 498[11720413084[   51011  4.2
931IT'Ubatulabal  1|   1 6.01 1    | 6.01     1 4.01-    11   11      11     I    11
. 94 1 -          .11 -7 5    1 Il        I J . 11 3.3[   11    1133u 1001I   1    11

9511 -Estimate   11751 6.0[ 1 - 1 6.0]1 2.511 3.71 22 11 2781133,100b189491150511 0.8

96IPaiute,          I     I     I     I       I 1   I  I        I I I   1    1     1

I MonoQCo.     I IilOj     I     I    I     11 2.Sj      11  1140,35011        11

9711  Estimate   || 1101 6.01 1_ 1 6.0 118.311 2.51 15 11 2751140%35016140I2690j]      0.7

98l1Paiute, 1a  e      I     I     I    1      1    I     Ii   Il       11   1      I

11 Inyo Co.    11    1     [     I     ]    11 5.51     11    1199[ 2001    I    11

9911  Estimate  || 1001 6.01 1     1 6.0116.7| 5-.5j 33 11 5501199 2%01i180441 3006jJ  0.5

|ISRI - Class 1                      See Table 9               _                   . _
100||SRI  -  Class  211  1771  1    1 T        11    I     11   11      11    I     11
101lI             11    1  5. 01    1    I     11    I     11   11       11   I     11
10211             11       1   1,-L  1   I.    12I11121   %...JL25, 00011     I   11

10311  Estimate   1[J1-J771 6.01 1  1 6.0129.51112.71 76 j1|22501125 000j|1968j 32911   9.0

l|Central Valley                     See Table 9
11channel Chumash                    See Table 9

60

References and Notes to Table 8
No.

1      Kroeber (1925:16) states that the Yurok had a population of

1,052 persons, in 141 houses, in 17 settlements on the Klamath
River from Rekwoi to Kepel. He gives 7.5 as the average family
or household number. This means 8.3 houses per village, and 62
persons per village.

2      Kroeber (1925:17) states 6.0 houses and 45 persons per village.
3      Kroeber (1925:17) cites the count of 1852 as giving 544 persons

in 68 houses, an average of 8.0 persons per house.

4      Kroeber (1925:19) cites the count of 1895 which gives 151 houses

and 384 persons for 17 villages. This means about 2.5 persons

per house, far too low for an aboriginal value.   It is therefore
omitted from the table.

5      Waterman (1920:206). The average of 42 villages is based upon

Waterman's estimate of aboriginal house numbers.

6      Waterman (1920, passim).   This is the average of houses and house

pits shown by Waterman on his maps for 17 villages, and is used
in Section II of this paper.

7      Kroeber (1925:78). This is the average size of 4 houses accord-

ing to measurements made by Kroeber.

8      Cook (1956:83).   This is the estimate of the mean of 78 villages

derived from Waterman, Kroeber, and Merriam.

9      Archaeological Research Facility records.   The mean of 11 sites.,

the records for which are to be found in the files.

10      The average area of 17 sites, maps for which are given by

Waterman. See Section II for discussion dealing with the Yurok.
The figure 25,450 sq. ft. is obtained by Method 2.

11      As for all other regions, this estimate is the best value which

can be assigned to each category in consideration of the available
evidence. The figures are not necessarily averages but may be

weighted according to reliability of source and thoroughness of
coverage. In this case the floor space and household number is

61

accepted directly from Kroeber. The number of persons per
village is derived from the average of houses per village
multiplied by 7.5. The area is that obtained by Method 2
from Waterman' s maps.

12     Kroeber (1925:116). The mean of 8 villages.

13     Loud (1918, passim). The mean of 22 villages.
14     Loud (1918, passim). The mean of 3 villages.

15     Loud (1918:266, 339). A single site, Loud's number 67, in or

about 1850.

16     Loud (1918:266). The same site in 1860., after attrition.
17     Loud (1918:259). Loud's site number 4 in 1850.

18     Loud (1918:269). Loud's site number 112 in 1850.

19     Archaeological Research Facility records. The figure for floor

space is the mean of 3 sites; that for site size the mean of 8
sites.

20     See Note 11. Kroeber's value of 7.5 for household number is

accepted as applying to the Wiyot. The number of houses per
village is based upon the mean values of Kroeber and Loud.

Loud' s sites numbers 4, 67, and 112 were unusually important
and probably exceeded the average size.
21     Kroeber (1925:17).

22     Cook (1956:98). The mean of 61 villages derived from Kroeber' s

(1936) lists. There are no data in the literature or in the
Archaeological Research Facility files on Karok village or
house size.

23      See Note 11.

24      Kroeber (1925:17).

25      Kroeber (1925:131, 138). The mean of 5 villages. Kroeber here

departs from his own rule of 7.5 as household number for the
northwestern tribes.

62

No.

26      Cook (1956:100).   This is according to the 1851 census, for

7 villages.

27      Cook (1956:100).   According to Goddard's map, for 11 villages.
28      Cook (1956:100).   According to Gibbs' map, for 9 villages.
29      Goddard (1903:13).

30      Archaeological Research Facility records.    The mean of 9 sites.
31      See Note 11.

32      Kroeber (1925:138).
33      See Note 11.

34      Dixon (1907:416-418) says the houses were rectangular, 5 x 6-7

meters (350 sq. ft.). They were often occupied by more than

one family, but these seem to have been related, i.e. extended
families. Villages were usually small with no more than 2 or
3 families.

35      Leonhardy (n.d., passim).   This is a careful study of the Iron

Gate No. 2 site on the Klamath River. Leonhardy found 13 house
pits, but thinks only about 5 were inhabited at one time. He

cites measurements of 8 house floors with an average of 266 sq.

ft.  He also states that the houses were conical, not rectangular.
36      Archaeological Research Facility records.    The house areas are

the mean from 4 sites; the site areas the mean from 18 sites;
the house number the mean of pits at 5 sites.

37      See Note 11.   The house floor space is the mean of the three

estimates shown.   Persons per family is taken at 7.0, but persons
per house is increased to 8.0 in order to cover Dixon's claim for
extended families. Number of houses per village is a compromise
between the mean of 7 house pits found by the Facility and

Leonhardy's idea that only 5 out of 13 house pits at Iron Gate
were occupied at the same time.

38      Kroeber (1925:312) states explicitly that the houses were small,

rectangular, 8 by 12 ft. or less. They were necessarily single
family dwellings.

63

39     Archaeological Research Facility records.   For house floor

space the mean of 10 sites; for number of houses the mean of
24 sites; for site area the mean of 95 sites.

40     Floor space is the approximate average. Family and household

number of 6.0 is reasonable in view of the very limited floor
space. However, there is another consideration involved.

Beyond the borders of the northwestern tribes (Yurok, Wiyot,

Hupa, and perhaps the Shasta) and with the exception of those

tribes possessing communal houses (Pomo, Wappo, Central Valley
groups, and the Chumash), all authorities agree in taking the
house as single family, and in putting the persons per family
and per house at 6.0. In view of such unanimity of opinion,
this figure is used consistently hereafter.

41     Archaeological Research Facility records.   The floor space and

houses per village are the means of 8 sites; the area the mean
of 30 sites.

42     See Notes 11 and 40.

43     Archaeological Research Facility records.   Village area is the

mean of 76 sites in Lassen County, and includes a few peripheral
Maidu sites near Susanville. The number of houses (house pits)
per village is the mean of 9 sites.

44     The floor space per house is an approximate average of that

found for the adjacent tribes, the Modoc and Maidu, and that for
the related groups to the south, the Tlubatulabal and the Paiute
of Mono and Inyo counties. The family composition is taken as
the unit conventional for the area (see Note 40).

45      Cook (1956:104). The recent authoritative work by Baumhoff

(1963) on the food ecology of the Athabascans, while treating

exhaustively of population, does not include data covering house
and village areas.

46      Goddard (1924:219). The family number probably relates to post-

contact times. The Athabascans were destroyed by the white man
very early in the 1850's.

47      Kroeber (1925:241).

48      Archaeological Research Facility records. The house number is

the mean of 3 sites, the area the mean of 29 sites. Of these 9

are Mattole (coastal), the remainder Wailaki and Kato. The site

64

No.

48      area mean for the Mattole alone is 16,420 sq. ft., for the

interior groups, 4,180 sq. ft.   However we have no other data
which would justify a segregation of these subtribes, and the
general average must stand.

49      The estimate of 5.0 for houses per village is based upon

Goddard' s ethnographic study, and is accepted in preference to
that of 6.3. The latter figure is derived from only 3 archaeo-
logical sites.

50      Foster (1944:176) and Cook (1956:107). The figure for houses

per village is based upon a single, unusually large village
described by Foster.

51      Gifford (1939, passim).   Referring to the Coast Yuki, Gifford

gives house numbers for 7 villages, based upon the memories of
informants (see Cook 1956:106).
52      Foster (1944:159, 226).

53      Foster (1944:226) states that the "common houses" were 10 feet

in diameter.  The Huchnom, however, had houses like the Pomro,
that were 20 feet in diameter.

54      Foster (1944: 159, 160, 176, 226) says that the Hop Ranch, founded

by fugitives, had 50 houses and was "twice the size of the average
nohot."  But Kicil, a nohot, had 24 houses and was "one of the
largest rancherias."  On the other hand, there were numerous
satellite villages which had no more than one or 2 houses.

55      Archaeological Research Facility records are unusually complete

for this area. Number of houses per village is the mean of 162
sites. Site area is the mean of 402 sites, including both the

coast and the interior. The figure for floor area is based upon
site Men-455 which showed 21 house pits, ranging from 7 to 12

feet in diameter. This agrees with Foster's statement (see Note
54) that the common dwellings were 10 feet in diameter.

56      The value of 78 square feet area conforms with both ethnographic

and archaeologic evidence, yet we have to allow for at least a
few larger structures among the Huchnom for it is not possible
to segregate this group upon any other grounds. The figure 100
square feet is therefore a compromise. The estimate of 6.0

65

houses per village increases the house pit evidence of 5.1 in
order to allow for Gifford' s very clear data on the coast. It
is impossible to compensate for the few large nohots described
by Foster other than to balance them against the very small
peripheral settlements.

57      Kniffen (1939:388).  Mean of 15 villages.   Based upon the

memories of informants.

58      Loeb (1926:159) describes the houses as 10 to 15 feet in diam-

eter and says, "This variety of house...is said to have contained
as many as twelve people."   However two informants stated that
their own houses had contained respectively 6 and 7 persons.

Furthermore 12 people could not find sleeping space in a house 10
to 15 feet in diameter.
59      Kroeber (1925:240).

60      Archaeological Research Facility records.   Floor area is the mean

of 4 sites; houses per village, the mean of 15 sites; site area,
the mean of 71 sites.

61      Floor space is the average of the three estimates reported, as is

the number of houses per village. Mean site area represents an

adjustment. The 71 sites mentioned in Note 60 are almost exclu-
sively in Sonoma County. Thirty-two of them are fishing camp

sites with an area under 1,000 square feet, a size highly improb-

able for a permanent settlement with approximately 30 inhabitants.
Hence it is justifiable to omit the camp sites and average only
those with an area over 1,000 square feet.   The result, with 39
sites, is 4,270 square feet, a figure which is consistent with
that obtained for the coast Yuki just to the north.

62      Archaeological Research Facility records.   Number of houses per

village is the mean of 19 sites; site area is the mean of 99

sites. Floor space datum is based upon one house pit at Nap-221,
near Middletown, which was 18 feet in diameter. Some further
help is provided by two Wintun sites on upper Putah Creek and

two Pomo sites on Highland Creek in Lake County-all four in the

vicinity of Nap-221. The average house pit diameters were respec-
tively 13, 12, 8, and 18 feet.  The over-all average is 12.75 feet
or an area of 128 square feet. As a compromise we have reduced
the Wappo figure from 254 to 200 square feet.

66

No.

63      See Note 62.

64      Archaeological Research Facility records.   Number of houses per

village is based upon 5 sites; mean site area, upon 84 sites.
65      We use the usual family number, 6.0.

66      Du Bois (1935:28) makes some rather ambiguous statements.   She

says, "Each bark house contained a biologic family of three to
seven people, so the population of a settlement might run from
twenty or thirty up to one hundred and fifty or two hundred
people."  This means an average family size of 5.0, with a

single family per house. But with the specified family and

house size, and the population mentioned, there would have been
from 4 to 40 houses. The latter figure is quite improbable.
No such villages are reported by any other observer for the
hills surrounding the Sacramento Valley.

67      Du Bois goes on to quote an informant who had lived in a village

of six houses with 30 to 50 people. This means a family number
of 5 to 8.3 persons, or an average of about 6.6.

68      Archaeological Research Facility records. House size is the

mean of three sites, Teh-13, Teh-22, and Teh-121, all located

in the hills of northwestern Tehama County, close to the Wintu.
There is no evidence for believing that the Wintu houses were

materially different. Number of house pits per village is the
mean of 18 sites, and site area, the mean of 114 sites.

69      We retain the customary 6.0 for family size. For the number

of houses per village, we roughly average the value given by

Du Bois' informant with the site mean. There is no reason to
attempt any adjustment for very big settlements, the existence
of which is implied but not demonstrated by Du Bois.

70      We know of no published ethnographic data for the Wintun proper,

apart from the Wintu and the Patwin. Hence dependence is placed
entirely upon the archaeological site records.

The whole ethnic stock is here divided into three portions: the
Northern Wintun includes the hill and upland villages located
in Tehama and Shasta counties; the Central Wintun segment is

confined to the Stony Creek area in Tehama and Glenn counties,
an area which has been intensively surveyed; and the Southern

67

Wintun includes the Wintun (plus a few Patwin and Lake Miwok)

living in the hills of the Cache Creek and Putah Creek watersheds.
For the Northern Wintun the house size is the mean of 7 sites;

number of houses per village is the mean of 66 sites; site area
is the mean of 143 sites.

71     We use the customary value 6.0 for family size and number of

persons per house.

72     The three figures are respectively the means for 2, 31, and 105

sites. The house area is probably too small, but there is no
evidence to justify a change.
73      See Notes 70 and 71.

74     Archaeological Research Facility records.   The three figures are

respectively the means for 5, 37, and 101 sites.
75      See Notes 70 and 71.

76     Kroeber (1925:397).   Data from American Valley, Plumas County.
77     Kroeber (1925:397).   Data from Indian Valley, Plumas County.
78     Dixon (1905:174).

79     Kroeber (1925:241).

80     Kroeber (1929:260).   The mean of 7 villages.

81     Archaeological Research Facility records.   The three figures are

respectively the means of 2 sites (both in Northern Maidu terri-
tory), 47, and 92 sites.

82     House size is the average of the five figures shown. For the

number of houses per village, the figure for American Valley is
disregarded.

83     These are the Sierra Miwok, from Amador to Madera County.   The

floor space figure is based upon Kroeber's (1925:241) statement
that the Maidu and Miwok both used a circular house 10 to 15
feet in diameter.

68

No.

84      Barrett and Gifford (1933: 200).

85      Archaeological Research Facility records.   The three figures

are respectively the means of 3, 13, and 136 sites.
86      Estimates as previously outlined.

87      The house data for the North Fork Mono are taken directly from

Gifford (1932:57-61).  There were 1,232 inhabitants, living in
239 houses, in 67 villages.  No area data are given by Gifford,

and there are no site records in the files of the Archaeological
Research Facility.  Gifford's figures, of course, apply to times
remembered by informants after 1850, hence the averages may be

too small for consideration as aboriginal. The house floor space
is the approximate average of those found among the Maidu, Miwok,
and hills Yokuts. There is no reason to suppose that the Mono
houses were significantly different.

88      This is for the small, single family house used by the hills

Yokuts according to Kroeber (1925:521).

89      The mean of four sets of measurements given for different tribe-

lets by Latta (1949:87, 91, 92, 98).

90      Data from Gayton (1948).  The family number is stated as 6

(p. 145). the non-communal house is single family (pp. 59, 161,
186, etc.). The floor area is the average of five sets of
measurements given for various tribelets.

91      Archaeological Research Facility records.   The three figures are

respectively the means of 16 sites (in Tulare County), 44 sites
(in Fresno and Tulare counties), and 109 sites.

92      The floor space is the average of the four figures shown. The

number of houses per village and site area are from the Archae-
ological Research Facility record files.

93      Explicit statements by Voegelin (1938:4).

94      Archaeological Research Facility records.   The three figures are

the means respectively of 4, 4, and 18 sites.

95      Estimates as previously discussed.

69

96     Archaeological Research Facility records. Means respectively of

9, 19, and 44 sites from Mono County exclusively.

97     Estimates as previously discussed.   The conventional 6.0 is used

for family number.

98     Archaeological Research Facility records. For Inyo County

exclusively. For houses per village the mean is of 18 sites,

all in Owens Valley. For village area the mean is of 55 sites,
also in Owens Valley.

99     For the estimate of floor space, an approximate average of the

Mono County Paiute and the TUbatulabal was taken.  The dwellings
were all the single family, desert type. For household size the
conventional 6.0 is used.

100     Floor space derived from the statement of Phil C. Orr (1951:221)

that Class 2 houses on Santa Rosa Island were 10 to 20 feet in
diameter.

101     This is the figure used by Meighan and Eberhart (1953:119) in

computing population on San Nicolas Island. We can see no

reason why it should be smaller than that accepted for all other
native groups.

102     Archaeological Research Facility records.  For the number of

houses per village the mean is of 9 sites; for site area the
mean is of 7 sites.

103     Estimates as previously discussed. However it should be noted

carefully that the figures given in the estimate in Table 8 for

Santa Rosa Island apply only to those sites which in the Archae-
ological Research Facility records are specifically stated to

contain houses of Orr's class 2. The distinction between classes
1 and 2 and their occurrence is discussed in full in Part 2 of

Section VI of this paper. The figures which represent the com-
bination of the two types, together with necessary adjustments,
are shown in Table 9. Since the great majority of houses on

Santa Rosa appear to have been class 2, the region as a whole is
treated in the graphs and calculations as an area with single
family dwellings.

70

ON

00   U)

C)

C4-i

-4 Ce

U) Q
.4J

uM  E-4
'U) ~

r4

4-i 'dZ
r4   C)
r >
CeJ -H-

CCa)
U   X4

-74 CO
(a) 4-

S-4co

C)

C) 44

0
7 U)

o4 P
4- cn

U) Cd
Q
4-i

(O
dCl)

H    C0

rLI E-4

-    4-
Ca C

C.)
U

4-I

4-i

.74 4-i
bo

C)  r,r-
N   '74  0

4' Cl)

O    cn    co

C) -o

u)       C

C)  .4  .7

S u

4U)
r4.J

cndC
?     co

:14

'-r4

U )
U) U)
r-4

0C

* ij *bs : aSETTuA

aad saie uealq

PaIB 92eTTTA 01 OOT x
azwds .10T1 1Td101 o0nld

-- - - -- .: -- =- - - 7

a2eTTTA sad
uosaad aad *II *bS

L)  00               4   Nc  4   i.0
en  %Z               00  4    NO

4   00               4   0    N  Lln
4C4                  e   400 N

N

en  00               00   H   ' 4

c.o  en               f  4    H  0

Ot                   Ot  0%  0

en  en               en   N   Ln

0)   0               0    0   0

Lfl  0)              Lfl  0   00

4ON0                0)  H-  U')
Li) 0 0               %   0   %   0
Lfl. 004             00f r4   N  H

N  N             -I   H   N  '.

*i3 'bs UT a2vT1TA sad
aouds 1oo01 W10o u?a3j

a2[ITRA sad
suossad j o *ON

a2uTTIA sad
sasnol jo * ?N

?3l *bs : uosaad
sad aoeds 1ooT1

asnol sad
suosaad 30 *0?N

asnoq sad
saflTmvJ 30 'ON

A1TUW3 slad
suosaad o0 eog?

'*1 9bs ueam
aouds 1003T asnoH

40

4 0
nt  (

Cn  r-

0
'.0

en

.    cn   N    0
00    0   en  '.

0%                       0%
-I                        r.4

0      N            '.0    N      00     N     N     N

n            en     N      Lfl    4     en     Cn    N

00    '.0    '-74  0'     0
N     rN:    4     0      N

7- -  -      -

0  en  4  %0

* L  L*  * en

a2vTITA sad
asnoq sad *tj  bS

U')   00

*')   en

Lfl)  Lfl   Lfl
No NN
0) 0        0

-7    -      -

LI')  Le')  Lfl
en   .s i

4     -N     -
o     o     o

-.I    *

v-   0  en  '.  N
N    en  00  0'  '

Y e n  -  H -

0Ll00 0 0
0 0    00

-7  -  v-  -  H  H

0    -L0 0 0 0

0    4 0 000
0    '0 H H O
4    Ns  -  H  H

10 aTsal

co)

.7

H-  Ce  Ce

:j 4I  ~4-i         5      p
C   C   C   Ce H-  U)  C   C
N   >1  N    4  -74  Ce  4 '

74                X ) 4  4  U

>   3   Z   -   C;)  c rn<  Z  Z

v-  N   en  4   LC '0  N- 00 0'

sqn.

jaqmnN TPuu0l.S

7 1

00 fN ONA 10)  N   00      CO  -It -It N", r-   10      N   00  N-  LO) It~      O

H--  r-  10)  N                                                       -4 10

Cl v-4 00  0   N.-  0  0   0   0   '.o  0   '.0  N          0  ~10 0  .00    1) r-4
H- COe CO v-4  0) C.   Nq  10  0    0)  '0  CO -It          O   ON  0   00   v-A 00
N1  v-A  v  -4 v-l  -I  CO  N   O\f     O   COC1  Cf    1   10)  '.  C   N       v-

O0)0000    0   0   Nl  0   Cf      Cf       1   0       Cf  ON  0D  Cf -.I   0   CO

N00 It   1)  ON  Cft 0)   N   C0  N   N   CO          00  Cft Cf   Cft '0      N
NN   00 %'. NCIA     C 0  C    fIC) '  N  N  0  '.0       ONM      0   v- 0    N

.\  .'  *   .'  *\  *\   "   .'  .'  * IF\        \  .-'  .'\  *\ .'  *

H14v4N            v-A N" N1 r-4I   N   N   N       CO) 00o  .   00  N1 N%   Nq

v-4  v-A

00     0   0   0   0   0   0   0    000  0  0  0  0  0  0  0  0

ON cyl N4  10  10  0   CO  0   0   COCOCO1              N        00

CON    N   N-  CO)  0  0   N   00   0r0Cf        4      N- r-   CO  N\  r     Cf

,' .  ' *  .  ' v  0.~  .   %   -   \       *\  .'\  *'\  .' %   .'  \

'.0   C  -I  N  N       r-4  \.0  00  It  CO  Cf-   r-4  CO'  0  ON  1   v-4  v-I

N   4 r-      v-4I v- v- --   v--A  r4     v-  C   C   ON   N-  Cf  0:

r-4l

Un 0   10) 0    0)  0       0    CO  00  '.0  N-  Cf  Cf0      00  10   0   0   0    0
NO0    '.0  0   0   '.0     0    0   v-A  N- N" %.0   10) ONA  N-   N   10  CO  0

'0 '.0 N-  10) v-A 00       '.0  ON  10) 00  00 'I0   Cf    f      N    10) v-4  0   C

v'  ON   *%

CO)  CfI v-4                                                     Cf     '.

0  '.0 --I 10)  0)  '   N    N       (ON  OIN 0C  N   00  CO   Nq  10  CO)  N    O

l  CY)  CO)  N1  00  10  O CCf)     nCO  C    Cf  CO,  v-I  N11  N   -  CO)  00  '.0  10)

N1 v-4                                                               CO  10)

00     v0 0             -0IO  C  Cft  0  CO  CO  \0  Cft '.0  00  N-  10) 10) C'0  0

U,'0101   0    N  1   10   N   '.   .   '0  1   CO4 CO   CO  Nl  10   ~ v-'A   N

1-4  r--                                                         v-I  N  C

CON-- 0)   0    \.0  CO,    10  10)  N   10  00   10  10  00   10  CO   N   10  00  CO)

0 Z '.  1  0:  00  CO;     00  v   CO~ CO~  0;  0:   Cf       N   00  '.0 N-   N   CO
N1  v-A   N     N,  r  CO   v-       -4  N"  N%   N-  CN  Nq  v--l  r-4  V-4  Cf '0  00

0  0   0   0   0    0   0   0    0   0   0    0   0   v--I  0  0   0    0   10  0

%O'0  .  10  N  '.0  \.0  '.0  '.0  '.   .   .   .   0 '0   .   .   .        0   1

v-   v4                                                rv-4

10)

00     0   100  0  0             0   0        0   0   0    0   0   0    0   N   0   10)

H  H   N   v-   r-4 v-4 v-A v-I  -I  -I  -I  ,   v-   v-4  v-   v-   I CO  N

00     0   0    0   0   0    0   0   0    0   0   0   v4   0   0    0   000  0

'.0 '.   \. '0   .  '. .  \.0  \.0  \.0  '.0 '.  '.0  10~  '.0 '.  '.0  '.0~  ' '0

U-i 0  0   0)   0   0       r-4I ON% NI  ON  10   CO  10  v-   1   0        '0

NOC    10  0)  Cft 0D       r-   N   00  CO  N11 N1   N-  CO   N   r-4  0   10  CO  10

H4  H-  v-A  ON  ON  Nq     v- -4 v-          -   -4i  r-  v-I     r-4  v-4  CO)  r-4  Nl

0   0

s- 4*                                                                >   Cd
0                                                                    Q A  )

v-I  0)                   4) -                  -       0    0   CI)  v-I

*jIi  $-i  v-4  v-I  ,4                    0       v-4  r-4      >-N  H-  v-I
Cl)  0)  -A  rA  0                                  r d   -I  Cd  0      co

O   co  4   Cd  v-                             v-I r   1d  v-0  Z   H    Cd  >   vO
O      0  ~         d   -vA4             4)                    co          cri )

r0 C  v-                    -I4  r_                        -'  . 0  0  v-A4  cd
0                        v-  Cd  Hr      co        C)  j   0)   a)  0Z  CZ  CO

- -  0  Q   4)   j ~  :~          ::jZ4        4-  4-  4)i  4)

0-H4   0   0    a4  04  C)  4)1      4)           5-  0        co   1 j         4)J

Id  14  Iq  lao              g    *        *  'H r  0)     4   vO  -vI  -v-  co  r-  Cld

4)    0   0   Cd  Cd   0   -H  0    )   0   CO  -vI      0   :i$  Cd  co  4)   0)  4)
>4 44     P4           0 ~         0    C3  Z        Z   >)H      P-  a    l)0

0  H   N   CO  Cf10     '.      00   ON  0)   v-I4 N1  CO  Cf  10  '.0      00  ON   0

H4P-  v-I  r-4  v-I  r4  v-I  v-A  v-I  v-I  N C  N4  NA  Nq  N N  N%  NA  N   N  C O

7 2

FIG. 1. The data presented by Naroll (1961) for 17

primitive villages (the city of Cuzco is omitted). Naroll's
values for area, given in square meters, have been converted

to units of 100 square feet and shown as such on the ordinate.
The abscissa gives number of persons. The small rectangle

close to the intersection of the ordinates encloses the space

on the graph which is occupied by the 8 points representing the
smallest villages.

FIGURE I

150                                                     fTr-

S~~~ 10S 1T

5, 000  g <XXX gXX

1, 000

74

F IG. 2.  As Fig. 1 except that floor spc  is
expressed as the logarithms of the values shown in
Fig. 1.

FiGURE 2

t XX

[-I.I''T1- rT-1r-XS

h-      IT    - 0    [<

[S.SE tXX+
tkeSSs-l 44H

J mgkW +s- r      H     +w

4-              :  II-Xt+-- 1- -Wg------

I        t

3.5 i ,t I        J

m 0tttiSHWiWlWl 0t tX>h-AtX   0+i"E

3.0 - 111 1i 11iiii2t4;

:.1 I+

| t t 1 ?  l $ r t t | flI i  -FTTt  I!   -+   t;   4 t$   ; t

2. 5            i 7Fdl  2fi l'    X    X

LttS   WI~~~-t L.;Xl.S ' A

2.=0   1E            t    F id L tF:IhI S ,TI qM

1     .8    t      3 0ijt Xl +t+-0t20

I  .-4F       lT8EILikLTLT  1jl;t0ittL01A4LlX t

1 .4 WX1lX

5,000  10.000   15,000

76

FIG. 3. As Fig. 2 except that number of persons
is also expressed as the logarithms of the values shown
in Fig. 1.

FIGURE 3

I 11VITT-T

2. 5

- - - - - - --- - - -                              If     I  II

2. 0

1.4

1. 8  2. 0 2.02  2.04  2. 6  2.8 3. 0

7 8

FIG. 4. Log plot of area by method 2 against house
count by method 1. Ordinate is log of area in 100' s of

square feet; abscissa is log of number of standing houses

plus pits. The regression line is drawn for all 17 points
with 'b equal to 1.593 (see Table 1). The two points cir-

cled are Omen and Tsurai. The second regression line where
b equals 1.197 refers to the remaining 15 points.

F IGLTRE 4
3 .1 .        If

3.0.

2.9 -------

2.8 .                 - ---------
2.7.
2.6.

2.5 -------
2.4.

2.3 .?                                                        ----------
2. 2
2.1

----------
2.0 .

1. 9 .

------- ---

1.8.
1. 7

1.6

I iI II f

1*5

1.4

I r I I I r FT-T T-

3

-------- ---- --

I L      - - - - - - - - - -

i T

- - - - - - - - - - --                                                                                          f I
1 0

I  I  I  T  I

0. 9

0.8                                                                                         R  FR

0. 6

80

FIG. 5. The average log of the area is determined for
all 17 (or 15) sites according to the six methods described
in the test5 and plotted on the abscissa. On the ordinate

is plotted the value of a (as the logarithm shown in Table 1).
The regression line is also shown. The value of r is +0.967.

FIGURE 5

2. 5
2.4
2. 3

2. 2,
2. 1,
2. 0,
1. 9.
1. 8.
1. 7

1. 6 .
1. 5 .
1.4 .
1. 314
1. 2-0
1 . 1 ?
1.0 .
0. 9 '.
0. 8 .

I   I   I  .   .   .   . I  1   1   1   1   1   1   1  .   .   . ?  . I  .   .   . I  .   .   .   .   .   . . . .   . . .   1 .   I   -   I   . . .   I . .   1 .   I .   I  . . .  I   I   I .   I  I  . . . . . . . . .  1 1 . 1 1 . 1 .

1111111,           llellliiiiii     H H H 1111    H ig                   H H H 11111'14-411+1   HH H H H

0  .   7  IL  I   I               in    1, 11'      .4   1 L I I  I I  H illilil   -  , I I I  I I I I I   I . . .  . . . . .I                             11

1                                                           . . . . . .

w             I I   11                   -  ,                      I                     I i I i i HI  ?IV          v

2.0       2. 1       2:2       2. 3       i. 4        2'* 5   2 0'6   2. 7       2'. 8     2.'9      3 O'O    3-0 1     3.'2     3. 3 ....... 3 '. 4  3. 5

82

FIG. 6. Ordinate is logarithm of site area in 100's
of square feet. Abscissa is logarithm of number of house

pits. The dots denote site group no. 1, the circles group
no. 2., and the crosses group no. 3.

FIGURE 6

4.0,

3. 5

AP"I

3. 0

2. 5

i+                  AL;

2.0

II f

I P i

I III It IIT

5.-

1 1   i f i l  f   I l i  f i l l i l i l i l   1 1 1 1 [ l i t '  ] , l I l i I i i I I 1 I I I I [ l I I !  l 1 f 1 T l I T -   T-11 17 -T-F U T T-

If
l4-

4t-

I    a

3iii f

0. 5

1.0

1. 5

2.0

2. 5      3.0

84

FIG. 7. For the eight sites discussed in the text the
logarithm of the area in 100's of square feet (as drawn on

Rogers' maps) is plotted on the ordinate against the logarithm
of the number of houses as stated by Crespi et al. on the
abscissa.

FIGURE 7

I  I  I  T  I  I  I I  I  I  I  I  I  I I

I  T  I  I  I I  I  I  I  I  I  I I

I  I  I  T  I

I  T  I  T  I  I  I  I  I  I  I  I  I  I  I  I  I  I  1  1  1

I  I  I  I  I  f  I I  I  I I

I  I  I  LA  I  I  I

4. 5.

IA, I II

4.0

II f
I I

r  I  I  I  I       f  I  I  I  I  I  I  I I  I  I I

3. 5

-444?

3. 0                                           -t  14:L

I-L-Ir LL-1

i   j  I  I  I  r

+4

2. 5_

I    T                                       I     T

I   I   f   I  t   I   i

1  7-   t
+                                                                  4

2. 0_

i - Va

4#t

T-t  7

-+t -f f4-    T ix-1-17T

-00[                                                 +4

5..

La  I I  I

1  1  T

+H-f- - - - -

Am  I  I   I

I  I   I  I   I  I  I   I  T  I

1.0                                                                                                                                                              MOW

0.0                       0. 5                          1. 0                         1   5                     2.0                          2. 5                         3. 0

86

FIG. 8. Log total floor space on the ordinate., log
mean number of persons per village on the abscissa. Dots

are single family, circles multifamily regions. The regres-
sion line is shown for b = 1.437.

F IGURE 8

3   7                                 I  .  I  I  I  I  I  I  i . .  I  T  T  I  I  I  I  I  T  I I  I  I  I  I  I  I  I   T  f  I  I  I I I  I  I   I  I  I  I  I  I  I  I  I  I  I  II  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  I  7   I  I  I  I  I   I

I  I   I   I   I   I   I  I  I   I   1  1   1  L  I   I

3  .   6                                                        I  I I I I I  I I  I T f f  I  I  I  I  i I I  I

i      I       f   I   I            I  I  I  I    I  I  I  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1  1

3 5

I  I   I  I  I  I

I  +                                     I      I   I  I      T  I   I

I           T     I   I  I  I  I  I

3  .   4                                                  I    I  I  T  I  I  I  T  I  I  I  I   I  I  II     I          I  II        I     I   I    I  I  I  I  I

I    I  I  I  I  I  I  i  T  I  I  I  I  I  I I      I  II      II I         I     I  I  I  .  I  I  I  I  T
I  I  I  f  I  I  I  I  I  I  T  I  I  T  I  I  I  I  I I  I  I   I  I         1     1  I  I  I  I  I  I  I

3.3.

3. 2..                           ------

T  I    I  I  I  I  T

3.1

3.0,

- - - - - - - - - -

2. 9.-

I  T  I

2.8.1

2. 7.1

I I T
La

I f
I                   I

2. 6

T  I  I  I  I  I  I  I  I

2   5        +                                                                                                                                              I I  f

2 . 4                                                                   If.  I  I  I  I  I   I  I  I  1.  I  I   I.,    I    I  I  I  I  I

I   I   I  f  I   I  I  I   I   I  I   I  I  I   I   I   I   I  I

2. 3

I          'T  I  I I  I  0

2. 2                                                                                                                             I         f 1 1

I    I   f  I  I

I        I  f I  I

2   1

2.0

i I I 1   I r YT_  T  I
4+ +?4

T7--r--I--T--I                                  -   -  -  -  -  -  -  -  -

1   .    9        .   .   .   .   .   .   .   .   .   . .   .   .   .   .   .   . .   .   .   .   .   .   .   .   .   . .   .   .   .   .   .   .   .   .   . .   .   .   .   .   .   .   .   I   I   .   I   .   I   I I   I   I   I   I   I   I   I   I   I I   .   .   .   I   .   I   .   .   I   .

If

4- 4-

I I f

liff      I

I AK]

IIfI

IfI                                                                                                                                                                                                                               t-Tt

: 1 1 1 1  T

I

11 I F

IF

1   .4                                 IIIII-D              70

T I I           I      I       I  I I   I   I  If I   I

I I    f 1 1 1
I I f I I I

1 TI I
1 3

Al                              ....       ..                           If
[ [ f i l l  I                   I   I   I   I   f   I   I I

i    T  I T-

u                                                       '         . - - - - - --.
1. 0  1 , I 1. 2  1. 3  1.4 1 0 5  I's 6 1. 7  1. 8  1. 9  2 0 0 2. 1  2. 2  2. 3  2.4 2. 5  2.6 24

88

FIG. 9. Log village area on the ordinate, log mean
number of persons per village on the abscissa. Dots are

single family, hill and coast; crosses single family desert;
circles multifamily hill and coast.

FIGURE 9

1.  I T T I                            I

IT   I   I I   I  I   I  I   I   I   I
I f   I  I   I   I
f   I   I   I   I   I   I   I   I  I   IT   I  I   I   I   I   I   I   IT   I   I

I   I   I   I  I   I   I   I   IT              I   I   IT   I  I  I           IT   T                                 I   IT   I   I   I   I  I IT   I   I   I   I  I  I   I   I   I
I   I  I   I   I   I   IT                                                            T   I  I   I                    f   I   I   I   I   1  1 1   1   1   1   1   1  1
I   T   I   I   I   I   I   I   I   IT  I   I   I  IT   I

I   I   I   I  I   I   I   I   I   I   IT  I   T   I   I  I  I  I   I   IT   I  IT   I         I  f   I   I  IT  I

I   I   IT   I   I   I   I  I  I   I   I   I   IT                I   IT   I   I   I  IT   I

I  I  IT                                           I IT I  I                             I  IT  I   I  I  I I  I I

I  IT  I   I  T  I I  I I
1 1   IT   I   I  IT   I  I  I   I

I I I F I I I  T I I  I I  I I I  F  I  I I  I  I  I
T I   I   I I   I  IT   T  I I  IT   I   I  I  I  .   I  I  I

IT  I  I I I IT I I I  IT T  I I  I I  I I   I  I I  II  I  I  T  I  I I   I   I IT  I I I  I I
I  I   IT   I   I  I I   I   I I   I  I  II   I  I I  I   I  I   I  I   I  IT  I  I  I  I  T I  I

T  I I    I I      I I I  IT  T  I  I  I                   I  IT I
- - - - - - - - - - - -

I I  IT I

- --- - - - - - - -

I  IT                                                              I  I I IT  I  I I    I

3 .               IT                                                                  T.      I  I IT

I T  I  I  I  T  - T  I  I   T  I

T  I  I  I

I   I   I   I   I   I   I   IT   I          I  I   IT              I  I  I                                  T   I   I

T  I  I   IT             I   IT   I I  I

1      IT       I          I   T   I I

I   f   I   I   I                                I    I   IT   I  I    I  IT

IT  I I  I IT I I  I  IT I I

3.1                                                                         I    I IT

I  I T   I                                    I   IT

I T  I
IT                  IT   I  I I I I  I  I  I  I  I IT  I
I   T   I  I  I  I   I  I   I  IT   I  T   I                I   IT                  IT

1   1  IT  I   I  I   I  T   I   I   r  I   IT I

I   I  I  I   I  I   T  I  I   I   I                                                                       I   I   I   I   IT

I   r  I   r  I   I  I  I   I  I   I   I   I

I  I IT I  T  I  I  I   I  I   I   I  I  I  IT

I IT  I  I I I  I  I  I  I   I I IT I   I  I  I  I  I IT I I  I  I

3  .0                                             TI .-                                                                              [II   I    ITT

I  I IT  I

I. I  I I -  I . .           .  I , I I  .-                        I  I                                             III   I    II I   T

I    I   IT   I  I   .  I   IT   I   I  I IT   I   I   I   I     IT I  I  I  I                                                             I T   I

I    I   I   I  I   I  I   IT   I  I  II   I   I            I    IT   I   I I  I  I                                                        I I   T  IT
I    I IT  I

I     IT  I                                                   I  IT   I   I   I   I                                                        I I   T  IT
I         I    I  I   I   IT   I   I  I   I   IT   I   I   I  T  I  I I I   I   I   I   I  I  I I   I                                        I  I  IT
I                          T  I  I  I I IT I I I                                                                                 T  I        I  I  IT
I  I  I  I  I IT                          I  I I  I  T  I  I  I  I I I I I I I I  I  I . I 1  I                                              I  I  IT
I  I   I   .   I   I   I   IT  I   I   I  I   IT   I               I   IT   I   I   I  I  .   IT   I   I   I  I   I   I

2 .   9       I IT 1.  I 1. I I -                                           I.-    i  T. I 1.    IT   11.

1   I   I   I   I  I   .   I   I   I   I   IT   I  T   I   I  I                                                 I    T

I   T   I   I   I   I   I  I   I   I   I   I   I   I   I

I  IT I
- - - - - - - - - - -           T    I                                                I  IT I

IT

I  I  IT I I I                                                                                                                         I  IT

2.8                                                                                                                                            T I

I IT

ITT   I                    I

IT

2. 7

I T T                          I T T  I

I f  [ I   I       1 .  I

I I F I                                                             I I  IT  I I  I  IT I

I  ir

I f  I               l if   I  I 9   i f  I I  I I     I   r  r  1 .    I   IT        i f I
I T  I  1 1  I                .   I  I -   IT   1 1  I I  T .,  1 1

I  IT  I  I  I  I I I  I                    I   T

I  I    i          I T   I   I

2 . 5  .    ITT            I  I  I          I  I   ITT

TI             T

IT
T I                          I  IT
I   I   I  I   IT    I  I I

IT
I   T                  T  I  I  I  1  1 I  I I I  I I I  I  I

2                                          I  IIIT           I

T

I TI  I                            I TI         I    I   I  i I

ITI

I  I   I   I   T   I  I                                    I  IT   I   I  I   I I  I   I
I  I   I   IT   I  T  I   I   I   I   I  I                                I   I   f   I  I   I I  I

IT        I  I  I  I I  III I  II I  II

2.311                                                                 IJ.

I  TI  I  I

I  I  I II                      I I IT I  I  I  I  II                                                                                  T  I I I I  IT

I   . I  I  I  I  11  I  I  I  I   I   T I    I   I    I   I   I  I  I   T

I ir I  I  I  I  I       I  I  I I     I IT   I        I  F r  .   .   .  I
I  I   I  I  I   I  I   I  I  I   I   IT  I  1  1  I  I     I T  I         . I  I        I   I   I  IT   I   I  I   I

I  I r       I I  I         I I  ITI I  III I

I   IT I  I             . . I  I I
I  I I  I  I I I  r  I  1   1  1  1  1   1  I I  I                   I                      I I F

2   2                 I ITI I I     I I IT  I I  I                                               ITI I

I   I   I   I   I   I                                            I   IT   I   I  . I

IT    I   ITI I  I I

I  I  I      I IT I            I I I       I I I  I   IT    I  I I I                                        IT I I I  I
I   I  f     I                  I   I   ?   ?  I   II   I   T  IT I  T  IT   I   I  I   T  I  I   I  I   I   I  I  I   I   I
I  T  I I  I  I   I  I  I          I I      II I  I  I IT I  I  I I I  I I  I   I  I   I  I

I   I       I T I  I  I             I I     I  I   I   I           I I                  I   I   IT I I  I    I                                I   I  I   IT I

I   I   I  I  I   1111111' I  T  I   I               T   I    I   I  I   I

I IT I   I  II I  I I  I        I I   I     I

I I :  I   . I  I I    I        . I  .  I I  . I  I  I  f I  I I  I I

2 . 1          I  1-   fl ill     I  Il l  I    I    I  I  Ti ff ' .  I  I  I Il l  I  9   I  I   I  I  I   I  I  I  T.  I   I  I  f  I  I  I I  I  I  I

I  II   I I I  I  I  I  I  IT  I  I  I  T  I  I  IT I I  F  I I I IT   T  I  I  I   I  I I
I        I   I   I   I  I  I   IT   I                                   IT   I  I I  I   I

I  I  IT   I  I   I

2. 0

IT

I       I T T

1. 9.-                  ---

I I        I          IT I                  I           I      I  I  T

I  I T  I I

IT  I I

l l f i  I

1.8

-44-
+

-L-L LI L         I I I -L LA -1 j

I   I   I   T  I

1 .6

I  I   I  T  I  I  I  I  I  I                                    I          T    I  I  I  I  I

I

I I T

I   T   I  I  I   II

1    2

I T I
- - - - - - - - - -

I. 0

v  --   -v             I    v    --  v   -  w      'W

1. 0  1. 1 1. 2  1. 3  1.4 1. 5 1. 6   1. 7  1. 8  1. 9  2 00 2. 1  2. 2  2.3 2.4   2. 5  2,.6 21

90

FIG. 10. Ratio of floor space to village area is shown

by dots on the ordinate. The corresponding logarithms are shown
by circles. The abscissa shows simply the numerical sequence of
25 regions spaced at equal intervals on the graph. Lines are
drawn approximately connecting series of points.

FIGURE 10

....        ..... .   .                                                           . .....

. . . . .  I                  . 1  . . . .  1 .  . . ..   . . .   . . .
. . . .   . . . . . .  . . . . . .  . . . . .                . . . . . . . .   . . . .
. .   I   . . . . .  . . . .   . . ..   . . . . .

.. .... ....

.. ... . ... ....
I        I     ...... ..    . ...  1 111  [[] I

iti   t   I  .... ..

. ..... ..  .....                                                ....                 ..... ....  I

. . . . . .    . . . . . .   . . . . .

100

. . . . . . . ..                      . .. . . .  .. . .       . . .                                   .. . . . . . . .

I  I       .  .  .   .  .  IL
I I I t I

90

t       I I t IL
I I I   t    I I

I   I   t  I   I  I  I   I  -   -   -  -   -  -   -   -

.   .   .   .   .   .   .   .   .   .  .   .   .   .  .   .   .   .   .   .   .   .   .   .  .   .   .   .   .   .   .   .   .   .

.   .   .   .   .   .   .   .   .   .                                  I
.  .  .         I   I  I  I  .   I                           .  .   .   .  .  .  I
I.   .   .   .   .   .  .   .   .

80

1.

I t

70

IL
IL I

I I  L t LI t                                    . . . . . .

I I t

60.

.. ........

50

fill [     I I.   I

. . . . . .                                                          . . . . . .

I IL
40.,

...                   . ...        ... I

. . . . . . . .. .   . . . . . . . . .    . . . . . . . . . .. . . . .   . . . . .  I  . . .   . . . ..  I   I  . . .  I  I   .   I  .   I  11 I   -1  I   .   I  .   .   .   .   I   .   I   I   I   I   .   I   I  .   .   .   .   .   .  I

.  .  .  .   .  .  .  .  .  I  I  .  I  I  I  I .   I  .  .   .  .  .  .  .  .  .  .  .  .  .  I  .   I  I  I  .  .  .  .  .   .  .  .  . .   .  .  .  .  .  .   .  .  .  I  .  I  I  .   I -1  t  t   I  I  I  I  I   I  I  I

I   I   I   .   I   I  .   .  .   .

.  .  .   .  .  .  .  .  .  .   .  .   I  I  I  I  .                      .   I  I  .  .  .  .   .  .  I  I  .   I  I  I  I  .  Li  I                     .  .  .  .   .  .

........                ......                    .....   I   tti                                                           L   t

3 0        ... I...               .... ti                    ....  I

. . . . . ..                                      .. . . . .

- - - - - - - - - - -

20

..........
..   ........                                                                          I

1 0

..........

..........

..........
..........
..........
.. .....  I..
..........

H 14111-W
U   Lu i   III1111 II IIIIII                                 ......                            .....                                        j.....

i HHH     .           11111

I

lu

20

92

AA

AAnt
BAE
UC

-AR

-PAAE

Abella, Fr.

1811

Arguello, 3

1821

Barrett, S,

1908

Barrett, S.

1933

Baumhofff, I

1963

Bolton, H.

1926

1927

Bowers, St

1890

BIBLIOGRAPHY

Abbreviations Used
American Anthropologist
American Antiquity

Bureau of American Ethnology
University of California

Anthropological Records

Publications in American Archaeology and Ethnology

. Ramon

Diario de un registro de los Rios Grandes, October 15-31,
1811, San Francisco. Santa Barbara Arch., IV:101-134.
MS in Bancroft Library, Berkeley.

Luis

Expedicion del Sr. Don Luis Arguello, comandante y escrito
por el P. Fr. Blas Ordaz, Capellan de la expedicion. Santa
Barbara Arch., IV:161. MS in Bancroft Library, Berkeley.

. A.

The Ethno-geography of the Pomo and Neighboring Indians.
UCPAAE 6:1-332.

A. and E. W. Gifford

Miwok Material Culture. Bulletin of the Public Museum of
the City of Milwaukee, 2:117-376.

4q. A.

Ecological Determinants of Aboriginal California Populations.
UCPAAE 49:155-236.

E. (trans.)

Historical Memoirs of New California, by Fray Francisco

Palou. 4 vols. University of California Press. Berkeley.
Fray Juan Crespi, Missionary Explorer on the Pacific Coast.
Diaries and letters relating to the expeditions of Portola,

Fages, and Perez, 1769-1774. University of California Press.

ephen

San Nicolas Island. California State Mining Bureau, Ninth
Annual Report of the State Mineralogist, 1889, pp. 57-61.
Sacramento.

93

Cook, S. F.

1955a

The Epidemic of 1830-1833 in California and Oregon.
UCPAAE 43:303-326.

1955b    The Aboriginal Population of the San Joaquin Valley,

California. UCAR 16:2:31

1956     The Aboriginal Population of the North Coast of California.

UCAR 16:3:81-130.

1960     Colonial Expeditions to the Interior of California:

Central Valley, 1800-1820. UCAR 16:238-292.

1962

Expeditions to the Interior of California:
Valley, 1820-1848. UCAR 20:5:151-213.

Central

Cook, S. F.

1950

and A. E. Treganza

The Quantitative Investigation of Indian Mounds.
UCPAAE 40:223-262.

Dixon, R. B.

1905

The Northern Maidu. American Museum of
Bulletin 17:3:119-346.

Natural History,

1907

The Shasta. American Museum of Natural History,
Bulletin 17:5:381-498.

Driver, H. E.

1936

Du Bois, Cora

1935

Wappo Ethnography. UCPAAE 36:179-220.
Wintu Ethnography. UCPAAE 36:1-148.

Duran, Fr. Narciso

1817     Diario de la expedicion de reconocimiento hecha en el mes

de Mayo de 1817. MS in Bancroft Library, Berkeley.
Reprinted in translation by Charles E. Chapman in

Publications of the Academy of Pacific Coast History 2:5.
1911.

Estudillo , J.

1819

M.

Diario que formo yo el tente dn Jose Maria Estudillo en la
campana... emprendo pa el reconocimiento y visita de las
rancherias situadas en los tulares....  Nov. 10, 1819,
Monterey. MS in Bancroft Library, Berkeley

94

Foster, G. M.

1944     A Summary of Yuki Culture. UCAR 5:3:155-244.

Gayton, A. H.

1948a    Yokuts and Western Mono Ethnography:  I, Tulare Lake,

Southern Valley, and Central Foothill Yokuts. UCAR
10:1-142.

1948b    Yokuts and Western Mono Ethnography:  II, Northern

Foothill Yokuts and Western Mono. UCAR 10:143-302.

Gifford, E. W.

1926     Clear Lake Pomo Society. UCPAAE 18:287-390.
1932     The Northfork Mono. UCPAAE 31:15-65.

1939     The Coast Yuki. Anthropos 34:292-375.

Goddard, P. E.

1903     Life and Culture of the Hupa. UCPAAE 1:1-88.

1924     Habitat of the Pitch Indians, a Wailaki Division.

UCPAAE 17:217-225.

Hack, John T.

1942     The Changing Physical Environment of the Hopi Indians of

Arizona. Harvard University Peabody Museum, Report No. 1.

Heizer, R. F. and J. E. Mills

1952     The Four Ages of Tsurai. University of California Press.

Berkeley.

Hester, James J.

1962     Typology of New World Cultures. AA 64:1001-1015.

Kniffen, F. B.

1939     Pomo Geography. UCPAAE 36:353-400.

Kroeber, A. L.

1925     Handbook of the Indians of California. BAE Bulletin 78.
1929     The Valley Nisenan. UCPAAE 24:253-290.

1932     The Patwin and Their Neighbors. UCPAAE 29:253-423.
1936     Karok Towns. UCPAAE 35:29-38.

95

1962     Two Papers on the Aboriginal Ethnography of California.

University of California Archaeological Survey, Report
No. 56:19-58.

Latta, F. F.

1929     Uncle Jeff's Story. Tulare Times Press. Tulare, Calif.
1949     Handbook of Yokuts Indians. Kern County Museum

Bakersfield, Calif.
Leonhardy, F. C.

n.d.     The Cultural Position of the Iron Gate Site.  M.A. thesis,

Department of Anthropology, University of Oregon. 1961.

Loeb, E. M.

1926     Pomo Folkways. UCPAAE 19:149-495.

Lothrop, S. K.

1928     The Indians of Tierra del Fuego. Museum of the American

Indian, Heye Foundation, Contributions 10:1-244.

Loud, L. L.

1918     Ethnogeography and Archaeology of the Wiyot Territory.

UCPAAE 14:221-436.

McKern, W. C.

1923     Patwin Houses. UCPAAE 20:159-171.

Meighan, C. W. and H. Eberhart

1953     Archaeological Resources of San Nicolas Island, California.

AAnt 19:109-125.

Merriam, C. H.

1955     Studies of California Indians. University of California

Press.

Naroll, R.

1961     Floor Area and Settlement Population. Mimeographed by

the War Stress and Culture Project, San Fernando Valley
State College Foundation, Northridge, California.

Naroll, R. and L. von Bertalanffy

1956     The Principle of Allometry in Biology and the Social

Sciences. General Systems: Yearbook of the Society for
the Advancement of General Systems Theory, I:76-89.

96

Nomland, G. A. and A. L. Kroeber

1936     Wiyot Towns. UCPAAE 35:39-48.

Orr, P. C.

1951

Ancient Population Centers of Santa Rosa Island.
AAnt. 16:221-226.

Powers, Stephen

1877     Tribes of California.

Ethnology. Vol. III.

Contributions to North American

Dept. of the Interior, Washington.

Priestley, H.

1937

I. (trans. )

A Historical, Political and Natural Description of

California, by Pedro Fages, Soldier of Spain. University
of California Press. Berkeley.

Rogers, D. B.

1929

Prehistoric Man of the Santa Barbara Coast.
Museum of Natural History. Santa Barbara.

Santa Barbara

Schenck, W. E. and E. J. Dawson

1929     Archaeology of the Northern San Joaquin Valley.

UCPAAE 25:289-413.

Schwartz, D. W.

1956     Demographic Changes in Early Periods of Cohoninna Prehistory.

Viking Fund Publications in Anthropology, No. 23:26-31.

1963     An Archaeological Survey of Nankoweap, Grand Canyon National

Park. AAnt. 28:289-302.

Smith, D. E. and F. J. Teggart (eds.)

1909     Diary of Gaspar de Portola During the California Expedition

of 1769-1770. Publications of the Academy of Pacific Coast
History, Vol. 1:3:31-59.
Stewart, 0. C.

1943     Notes on Pomo Ethnogeography.  UCPAAE 40:29-62.

"Trapper"

n.d.     A newspaper account found in the Hayes Mission Books

[probably written by J. J. Warner]. Bancroft Library,
Berkeley.

97

van Hemert-Engert, A. and F. J. Teggart (trans.)

1910     The Narrative of the Portola Expedition of 1769-1770,

by Miguel Costanso. Publications of the Academy of
Pacific Coast History, Vol. 1:4:91-159.

Voegelin, E. W.

1938     Tiubatulabal Ethnography.  UCAR 2:1-84.

Warner, J. J.

1890     Description of the 1832 Epidemic Among the Indians of

San Joaquin Valley. In An Illustrated History of San
Joaquin County, California. Lewis Publishing Co.,
Chicago. Pp. 28-29.

Waterman, T. T.

1920     Yurok Geography.  UCPAAE 16:177-314.

Work, John

1945     Fur Brigade to the Bonaventura, 1832-1833. Alice Bay

Maloney, ed. California Historical Society. San
Francisco.

Zalvidea, J. M.

1806     Diario de una expedicion tierra adentro, 1806.  Santa

Barbara Arch., IV:49-68. MS in Bancroft Library,
Berkeley.