ANTHROPOLOGICAL RECORDS

6:4

CULTURE ELEMENT DISTRIBUTIONS: XVIII

UTE-SOUTHERN PAIUTE

BY

OMER C. STEWART

UNIVERSITY OF CALIFORNIA PRESS

BERKELEY AND LOS ANGELES

1942

CALIFORNIAN FISH SPEARS AND HARPOONS

By

J. A. BENNYHOFF

ANTHROPOLOGICAL RECORDS

Vol. 9, No. 4

it

PI

I

ANTHROPOLOGICAL RECORDS

Editors: E. W. Gifford, R. F. Heizer, R. H. Lowie, R. L. Olson

Volume 9, No. 4, pp. 295-338, 7 figures, 3 maps

Submitted by editors, January 6, 1950

Issued August 25, 1950

Price, 75 cents

UNIVERSITY OF CALIFORNIA PRESS

BERKELEY AND LOS ANGELES

CALIFORNIA

CAMBRIDGE UNIVERSITY PRESS

LONDON, ENGLAND

MANUFACTURED IN THE UNITED STATES OF AMERICA

CONTENTS

Page
Introduction  .  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  295

Fish spears  .... .       .  .  .  .  .  .  .  .  .  .  .  .  .  .   ......... .         .  .  .   .  296

Unbarb.ed types  .... .      .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .   .  296
Barbed  types  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  297
Harpoons  (?)  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  298
Simple harpoons    .. .     .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .   .  299

Northwestern California ....   .     .  .  .  .  .  .  .  .  .   ......  .  .  .  .  .  .  .  .  .  .  .  .   .  299
Central California     .        .  .  .  .  .  .  .  .  .  .   . ......  .  .  .  .  .  .  .  .  .  .  .  .  .   .  301

Unilaterally barbed types   .... .      .  .  .  .  .  .  .   ...... .  .  .  .  .  .  .  .  .  .  .   .  302
Bilaterally barbed types .... .       .  .  .  .  .  .  .   . ..... .  .  .  .  .  .  .  .  .  .  .  .   .  302
Southern  California  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  303
Qomposite simple harpoons ....   .     .  .  .  .  .  .  .  .   . ..... .  .  .  .  .  .  .  .  .  .  .  .   .  304
Composite  harpoon  arrows  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  304

Toggle  harpoons  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  305

Analysis.                     . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307

Early horizon fish spears . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .     307
Middle horizon fish spears   . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .    307
Northern maritime harpoons                   ..309
Wider distributions ................ .                                                     310
Central  fluvial  harpoons  .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .   .  312
Southern maritime harpoons                   ..316
Historic fishing harpoons                   ..317

Summary  .... .       .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .   ........  .  321
Concordance of site numbers   .... .      .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .   .  323

ibliography .... .       .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .   .  325
rchaeological specimens, figures 1-7:   explanation  . . . . . . . . . . . . . . . . . . . .    329

MAPS

Settlements and culture sequence in Central California . . . . . . . . . . . . . facing      295
Archaeological sites               ..299
A Ethnographic distribution of toggle harpoQn types         ..319

[iii]

LITTORAL       INTERIOR

ZONE       VALLEY ZONE

Z COASTAL PROVINCE INTERIOR PROVINCE
N    L-AIQ-309              i 9
Q    Q Ala-329        x     ;t

O 0    CCo-295   ~     Scic-43 ~Co

SCI-356             "t

66C

Z                60

~~~J- 0 ~ ~ IJ~.  Scec-107
cc  ~~~~Z~        SJo-56
<cr   (UNKNOWN)       SJo 568

a

Map 1. Settlements and Culture Sequence in Central California: Temporal and Areal Relationsa

a. Early and Middle Horizon Settlements. b. Late Horizon Settlements

CALIFORNIAN FISH SPEARS AND HARPOONS

By

J. A. BENNYHOFF

INTRODUCTION

Fish spears and harpoons are among the most
generally used implements of primitive fishermen.
Too often, however, mere mention is made of

their presence, with the assumption that their
tspecific forms are unimportant details. This

has been largely true of California since its

discove'ry, as well as for most of the territory
.between the Eskimo and the Yahgan.   The wide

idistribution of these implements suggests their

early invention, but insufficient examples have
[been published to reveal any extensive rela-

Ftionships.  The early study made by Mason (1902)

constitutes the only available source for har-
poons of the whole New World, and in it the
,hemisphere is treated in a summary fashion.

'.This work and the ethnological reports from the

state emphasize the presence of the toggle har-
poon in California, while excavation has since
produced an interesting variety of fishing
mplements.

The present paper therefore discusses es-

sentially only the forms found archaeologically
n California and their distribution within the
state. Both temporal and areal differences are
apparent, with the suggestion of local develop-
ment of certain types, while the presence of

still different types is obviously due to dif-
. fusion from centers outside the state.

As a convenience, in the following de-

criptions the typology set up by Gifford (1940)
has been followed wherever possible, the major

ifference being that in this paper separate

ypes have not been established on the basis of

terial, but arbitrarily on that of form only.
uch of the terminology used is that presented
y Drucker in his Northwest Coast Survey.'

pecial appreciation is expressed for the gui-
ance and assistance of Dr. R. F. Heizer, at

soe suggestion this study was begun. I am

ateful also to Professor E. W. Gifford for ac-
es to the University of California Museum of
thropology (UCMA); to Mr. F. Fenenga for his

nterest and for his valuable knowledge of Cen-
ral Californian archaeology; and to Mr. C.

ard for information concerning site CCo-138.

acings and information on Central and Southern
lifornian harpoons in other institutions were

lDrucker, 1943, pp. 34-39; most of the terms are self-

planatory or referred to by illustrations. "Enclosed" re-
rs to barbs which are larger than the intervening spaces;
lsolated" barbs are smaller than these spaces. "High" and
ow" refer to the relative proportion of length of barb to
ickness of shaft.

[29

obtained from Mr. A. Woodward of the Los Angeles
County Museum, Mr. N. C. Nelson of the American
Museum of Natural History (AMNH), Dr. N. M. Judd
of the U. S. National Museum (USNM), and Mr. E.

K. Burnett of the Museum of the American Indian,
Heye Foundation (MAI-HF). Dr. H. H. Stuart, of
Eureka, California, kindly made available for
study his Northwestern California collection.

The site designations using county symbols em-
ployed herein are those assigned by the Univer-

sity of California Archaeological Survey. Sites
not yet recorded in the Survey files or those of
uncertain identification are referred to by name
or previous number. A concordance of new and
old site numbers is given on page 323.

A total of 603 pieces is discussed in this

report, 462 of which are in the UCMA. More than
nine-tenths of the specimens are made of antler,
probably deer or elk. Most of the bone pieces
are small, simple artifacts, particularly of

fish spear type MM2 and simple harpoon type II,
since the thinness and curvature of most bones
would be ill-suited to the manufacture of the
larger specimens. Except for whalebone pieces
(type W) and three of bird bone (Wla) from

Southern California--in part a reflection of the
island habitat of the users--all other bone
specimens are of unidentifiable long bone.

Fish spears and harpoons were occasionally
included as grave furniture, 150 occurring with
37 burials and 24 with 10 cremations. They oc-

curred with both male and female burials, Those
of type 00 perhaps had some other function than
that of mere utilitarian implements, since large
numbers were often buried with individuals.

This is especially characteristic of type 002,

for Schenck and Dawson2 record as many as 51 and
31 included in individual graves. It is prob-
able that fishing and sea-mammal hunting were

the primary, if not the only, uses of these im-
plements, though birds or beaver and other

aquatic mammals may have been taken with them.

However, the only indication that they may have

served as lances for warfare is a dubious artis-
tic arrangement of "Weapons of War"3 containing
a sea-mammal harpoon point permanently attached
to a shaft.

2Schenck and Dawson, p. 369. The 28 specimens from

Beecher Ranch, referred to by Meredith, in Moorehead, 1900,
p. 272, may have been with a burial also.

3Powers, fig. 1, p. 53.

?5]

ANTHROPOLOGICAL RECORDS

The operational principle of the fish spear
is quite different from that of the harpoon, al-
though the terms are often used synonymously. A
fish spear consists usually of a bone, antler,
or wood point permanently attached, often by

means of foreshafts, to a simple shaft retained

in the hand. The harpoon has a detachable point,
connected by a retrieving line to either a sim-

ple or composite shaft. Two basic types of har-

poon, as defined by Jochelson,4 were present in

California: (1) the simple harpoon, "with a heaB
that retains its original position after strik-
ing an animal" (both single-piece and composite
heads were used); (2) the toggle harpoon, "in

which the head assumes a transverse position wheN
an obstruction is encountered."   Except for a

simple bipointed type, only composite varieties
of the toggle harpoan were known in California.

FISH SPEARS

This category includes various types of

artifacts, for many of which several uses have
been suggested.

UNBARBED TYPES

MM3a (fig. 1, a-c). Short length (5.5
to 6 cm.); nearly spherical cross section;
simple sharp tip; bowed, unbarbed shaft;

notch on outside of shaft near square base.
One specimen from site SJo-56 (fig. 1, b)
has a raised girdle one-third of the dis-

tance from the point, and a grooved base op-
posite the notch. Total of 17 specimens, 8

with 4 burials and 4 with 1 cremation. Dis-
tribution: SJo-56 (1); SJo-68 (11); Sac-
107 (5). (Lillard, Heizer, and Fenenga,
pp. 26, 32, 41, p1. 20, fig. i; Heizer,
1949, p. 28).

The notched base and the fact that

nine of these points were found in groups of
three at sites SJo-68 and Sac-107 suggest
their use as tridents. Four tip fragments
were found with a cremation at SJo-68, but

more points may originally have been present.

MM2b (fig. 1, k-u). Medium to short
length (11.5 to 4.0 cm., but average is

closer to 6.5 cm.); thin elliptical cross

section; simple sharp to spatulate tip; un-
barbed; great variability in shaft shape,

which is usually asymmetrical, with a ten-
dency for the side opposite the basal foot

to be straight; oblique distal end with uni-
lateral footlike base. A pair from Sac-66
had seven serrations on upper shaft on side
opposite basal foot (fig. 1, o). Total of
47 specimens, 12 with 8 burials, 7 with 2
cremations. Distribution: Sac-66 (29);

Sac-122 (1); Ala-309 (4); SMa-372 (1); Sac-
43 (4); Sac-60 (4); CCo-139 (2); Sonoma

County (2, Gaffney collection). (Gifford,

1940, pp. 183, 232; Uhle, pl. 9, figs. 11-13;
Lillard and Purves, pl. 10; Heizer and

Fenenga, fig. 1, no. 18; Lillard, Heizer and
Fenenga, pl. 19, a-e, f, g. The three types
set up in Lillard et al, are here considered
as one, since types 2 and 3 have only one or
two examples.)

A variety of uses has been proposed for

these artifacts. Uhle5 suggested their utiliza-

4Jochelson, p. 53.

5Uhle, p. 75.

tion as arrow points, similar to implements of

the Bororo,6 but the pieces seem needlessly ape*
cialized for such a use. Schenck7 interprets

them as composite fishhooks.  At the Early hori-;
zon site of SJo-68 the assemblage shown in

figure 1, j was uncovered in situ.8   The blade"
like object has been considered an MM2b type,9
although both tip and base are atypical.   Its

purpose is difficult to imagine since the hook
is complete in itself and needs no supporting

member.  It has been suggested that it acted as
a wiggling lure to attract fish.   That all im-
plements of this type found in the succeeding

horizon had a similar use seems doubtful.   In no
other instance were these blades found associ-

ated with curved fishhooks or with any artifact
suggesting a composite fishhook. The numerous

rounded tips would make them unfit for fishhook

barbs. Use as a shank would imply wooden barbs'
though the opposite combination, that is, wooden
shanks and bone barbs, is the usual native con-
struction. One general characteristic which

should be noted is the tendency for the side op-
posite the basal projection to be straight,

lacking the angularity needed in a composite
fishhook to separate the barb from the shank.

This straight side would be an advantage if the
implement were lashed against a shaft (possibly
grooved as in fig. 1, u) to be employed as a
fish spear; the attaching cord would then be

held in place by the projecting foot (see fig.
1, t). Many retain evidence of binding at the

basal end. At site Sac-66 twelve of the twenty.

nine found were definitely paired. The four ex-
amples from Sac-60 were found with one burial.
This again suggests a bipointed fish spear.

Dull, roundish points would be less of a hin-

drance when the implement was applied with the
quick thrusting force of a spear.   The frequent
outside angularity and tapered distal ends would
tend to hold the prey, but escapes were a cir-

cumstance which perhaps led to the acceptance of/
barbed types instead.  That barbs were not es-

sential to success is indicated by the unbarbed
leisters used by some historic tribes.

6Lowie, f ig. 53, b, p. 425; or the Bororo harpoon,
Mason, 1902, fig. 14{, p. 217.

7Schenck, p. 226, pi. 43 FH.

8Lillard, Heizer and Fenenga, no. 9, p. 7.
9Beardsley, 1947, p. 152.

.

296

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

Other unbarbed artifacts discussed here as
fish spears are even more problematical. Few

specimens have been found, and it is their gen-
eral similarity which prompts their inclusion
here.

MM2a (fig. 1, d, e). Short length

(4.6 to 5.3 cm.); elliptical cross section;

simple tip; unbarbed; basal shoulder on face,
or slight groove on lower half of face with
transverse ridge midway in length.   Total of
3 pieces. Distribution; Ala-309 (3).

(Gifford, 1940, pp. 183, 232; Uhle, 1907,
pl. 9, fig. 15; Schenck, p. 227, fig. 6,
pl. 43W.)

Uhle and Schenck proposed that speci-
mens of type MM2a were used as arrow points
and fishhooks, respectively, but the objec-
tions noted in the discussion of type MM2b

apply here also. The shoulder or groove and
the straight back of type MM2a would allow
their hafting as fish spears. Their short-
ness is, however, a major deficiency.

MM2d (fig. 1, f-h). Medium to short

length (11 to 7 cm.); thin elliptical cross
section; simple medium-sharp tip; unbarbed;
elongate trianguloid shaft constricted in

lower half, which often lacks polish and re-
tains traces of binding. Total of 7 speci-
mens, 5 with 3 burials. Distribution:

Sac-16 (1); Sac-29 (3); Sac-43 (2); Yol-52
(1).

The few examples of this type prevent
any definite conclusions as to use. Their
angularity and isolated occurrence suggest

they may have been barbs of composite fish-

hooks. The constricted base would also per-
mit their attachment as fish-spear points,
and because of their general similarity to
type MM2b they are included as having the
same use.

MM2e (fig. 1, i). Medium length (9 cm.
restored); rectangulat cross section with
one edge rounded; simple pointed tip; tri-
anguloid shaft, with notches staggered on
each side of upper edge; notched base.

Only a single fragment, from CCo-250, has

been found, and it is tentatively included

because of its notched base and general form.

A variety of other archaeological specimens,
primarily from Southern California, could have
served as fish-spear barbs. Certain of Gif-

.ord's'0 types U2 and U3 quite possibly served as
prongs of the tridents, fisgas, often referred
So by early Spaniards (see p. 303). De Laguna"3

Would also include larger examples of types Tlb,

lh, T2bl, and T2c,12 but the uses of such simple
rtifacts are so manifold that no definite

otion can as yet be suggested without evi-

Xence of-the position of the specimens in situ.

10Gifford, 1940, pp. 177, 223.
"De Laguna, p. 212, fn. 637.

12Gifford, 1940, pp. 176, 177.

BARBED TYPES

That the barbed fish spears developed from
type MM2b is suggested by their distribution,

the similarity of bases, and one example, fig-

ure 1, k, which with some certainty can be con-
sidered a specimen reworked, after loss of its
barb, into an MM2b. It was found at site Sac-

43, one of the two Interior province sites shar-
ing both barbed and unbarbed examples, and prob-
ably represents a transitional form.

These barbed specimens provide greater evi-
dence of use as fish spears. Meredithl3 refers

to a newspaper account of the finding of a group
of these implements in which "2 of the specimens
were bound to a rod or staff, as if it had been
used as a gig or spear." Schenck and Dawsonl4
illustrate such a reconstruction. Rock Creek

mound produced "four or five pair." Large num-

bers of these barbed specimens were often buried
with the dead, and the even numbers frequently
found suggest original pairing. Decomposition
and the possibility that unhafted points were

included would account for the odd numbers which
also occur.

Many pieces could not serve as fishhooks,
as suggested by Schenck in an earlier publica-
tion,'5 for in them the barb occasionally pro-

Jects beyond the shoulder or is of such length
as to touch any straight shank laid against the
flat edge. To explain this one must suppose

that an obtuse-angled or curved shank of wood

was employed. That an arrangement of this sort
is possible is illustrated by an actual speci-

menl6collected during the Vancouver voyage along
the Pacific Coast, Hawaii, and Tahiti, for which
no location is stated. (See fig. 7, i.) It is
quite possible that both the fish spear and

fishhook are represented, since the variation is
so great. Figure 1, d' in particular suggests a
fishhook. But the rare occurrence of these ar-
tifacts in the Late horizon and their complete

absence among historic tribes argue against any
widespread use as fishhooks. The substitution

of the less advanced unbarbed composite hooks of
historic tribes for barbed fishhooks is diffi-
cult to explain. On the other hand, the aban-

donment of the fish spear after the entrance of
the harpoon is easy to understand.

001 (fig. 1, v-_za'-e').  Medium to

short point (12 to 8 cm., with average around
10 cm.); thin rectanguloid to elliptical

cross section, with flat edge often extending
on barbed side from below the barb to the

base; simple sharp tip; one high (occasion-
ally low) unilateral barb; oblique shoul-
dered lower shaft frequently straight on

barb side; unilateral footlike base, on side

131n Moorehead, p. 272.

"Schenck and Dawson, p1. 80, a, pp. 368, 369.
15Schenck, fig. 4, p. 226.

16Dalton, pl. XVI, fig. 22, p. 245.

I

I

297

ANTHROPOLOGICAL RECORDS

opposite from barb. Specimens in figure 1,
d', e' lack this foot but were apparently

notched for attachment. Total of 40 speci-
mens, 19 with 4 burials. Many are only

basal fragments and may be type 003. Dis-

tribution: Ala-309 (1); Ala-329 (1); Col-l
(4); Sac-8 (1); Sac-16 (2); Sac-29 (1);

Sac-43 (13); Sac-66 (1); Sac-117 (2); Sac-49
(8); SCl-356 (2); SJo-59 (1); Sol-2 (1);

Yol-52 (2). (Gifford, 1940, pp. 184, 233;
Schenck, p1. 43 Z, p. 226; Lillard and

Purves, pl. 10; Heizer and Fenenga, fig. 1,
no. 16.)

002 (fig. 2, a-e). Medium to short

point (11.5 to 6.5 cm.); thin rectanguloid

to elliptical cross section, with flat edge
extending on barb side from below barb to
base; simple sharp tip with notch at base;
one high, angular,unilateral barb; concave

arc of shaft below barb extends out to apex
of shoulder in middle of shaft; beveled

shoulder extends to base; unilateral foot-

like basal projection on side opposite barb.
Notching may occur on the arc of the shoul-
der, occasionally on both sides. Four or
five notches is the usual number, but

Meredithl7records twenty-eight on one speci-
men. Set off from type 001 by more careful
wor.kmanship and notched tip. Total of 128
specimens, 85 with 4 burials.18 Distribu-
tion: CCo-141 (2); SJo-87 (31); SJo-140

(7); Beecher Ranch (28); Bethany (52); Rock
Creek (8). (Gifford, 1940, pp. 184, 233;

Moorehead, fig. 412, no. 3, p. 273; Heizer
and Fenenga, fig. 1, no. 15; Lillard,
Heizer, and Fenenga, pl. 29, g.)

003 (fig. 2, f-j). Long to medium
point (14.5 to 9.5 cm., with average at

12 cm.); medium elliptical cross section,

with flat edge occasionally extending from
below barb to base; simple sharp tip; two

medium to high, enclosed, unilateral barbs;
shaft, occasionally shouldered, tapers to
constricted base with unilateral footlike

projection on nonbarbed side. Total of 12
specimens, 7 with 2 burials. Some of the

001 fragments may be of this type. Distri-
bution: Sac-8 (1); Sac-43 (7); Sac-122 (1);
Sac-127 (1); Yol-52 (1); Ala-309 (1). (Gif-
ford, 1940, type NNlb, pp. 183, 233. Com-
plete pieces excavated since indicate that
the example illustrated there is not to be
identified with marine forms used ethno-
graphically in Northwestern California

[ibid., p. 237, fig. 23], and a new type has
been established; Schenck and Dawson, pl.
80, P.)

ITIn Moorehead, p. 272.

18If the 28 specimens reported by Meredith, ibid., from
Beecher Ranch were found with a burial, the burial associ-
ations of type 002 would increase to 113 with 5 burials.

HARPOONS(?)

Wla (fig. 2, k-t). Medium point (16

to 6.5 cm.); cross section varies with bone

used, from curved to thin elliptical; simple
sharp to medium tip; one (occasionally two)
low, usually convex, occasionally concave,

unilateral barb; conical or spatulate base.
One of antler, three bird-bone, two whale-
bone, and six mammal-bone specimens. Total
of 13. Distribution: Hum-67 (1, Stuart
collection); Col-2 (1); CCo-295 (1); San

Nicolas Island (8 [5 in AMNH, Terry collec-
tion]); Channel Island area (1, illustrated
in Amsden); San Francisco shellmound (1, in
Oakland Public Museum). (Gifford, 1940,
pp. 178, 224; the specimen from Hum-67

listed as Wl [p. 178], should be changed to
type II; it is a tip fragment [see Loud,

pl. 21, fig. 13] so the base type cannot be
given. Fig. 2, k, illustrated in Amsden,
p. 26; fig. 2, r, in Stuart collection;

fig. 2, 1, and four other examples in AMNH.)

Type Wla is one of those generalized forms
more easily dealt with in a typological classi-
fication than in a discussion of function, for

no one use can be assigned to it.   The smallness
and fragility of the specimens preclude their
use in hunting large game. The low barb and
small size could be suited to fish harpoon

arrows, and the grooved base of figure 2, n
suggests a line attachment. It is doubtful

whether figure 2, 1 represents a line hole, in

view of the distribution of this form of attach-
ment.  The double-barbed example from the Chan-
nel Islands (fig. 2, k) may have been the point
of a small composite harpoon. Possibly the

Southern Californian pieces served as prongs of
the tridents so often referred to in the early
Spanish accounts of the region.   In one trans-

lation these tridents are described as being of
bone, with a single barb,19 but if the term

fisgas was the original Spanish word, as seems
likely, it could also be translated "harpoon,"
and could more aptly apply to type Wlb, de-

scribed on page 304. Harrington, in describing

the Costanoans, records fish spears with one or
two points, each having a barb.20 Perhaps re-

lated is the Tachi Yokuts single-pointed harpoon,
having a hooked point and made of a pelican wing
bone.2' Although there is no evidence of the

barbed fishhook in the regions where the type

Wla specimens were found, the small size of some,.
especially the San Francisco Bay pieces, would
make them suitable for use as fishhooks. An

ethnographic specimen is known from the Yahi,2
and figure 7, n, o shows prehistoric fishhooks

with wooden shanks and bone barbs obtained from
Humboldt Cave in west central Nevada.

19Fages, p. 51.

20Harrington, nos. 52-54, p. 7.
2"Gayton, p. 15.

22Gifford, 1940, fig. 24, p. 237.

298

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

SIMPLE HARPOONS

Various forms are included in this group,

differentiated chiefly by the location of barbs,
which are either unilateral or bilateral placed
in opposed or staggered position.

NORTHWESTERN CALIFORNIA

The typology proposed by Gifford (1940,

p. 183) is inadequate to deal with the varia-
tion exhibited by harpoons from Northwestern

California acquired since his publication ap-

peared. A new typology is therefore substituted,
using the following symbols: I, large unilater-
ally barbed simple harpoons for hunting sea mam-
mals in Northwestern California; II, small uni-
laterally barbed simple harpoons used for fish-
ing and small game in Northwestern California;

~qua                   I
fiQelg-ua                  i

dding

site 7

I
I

Mrn-27'

n S a Oa rbara

ifSte 10

Santa Cruz Island    Mugu
Santa Rosa Island

Santa Catalina Island,%X
San Nicolas Island'  %San

ClementA.

Island

Map 2. Archaeological Sites

capital letters, method of line attachment (A,

bilateral line shoulder; B, bilateral line guard;
D, unilateral line guard; D, line hole); Arabic

umerals, tip variations (1, simple tip; 2,

lotted tip; 3, grooved tip with inset); lower
ase letters, barb variations (a, simple barb;
k hooked barb). The number of barbs has not

een used as a criterion in classification, for

arpoons were frequently reworked after barbs
ad been broken off. Three forms range in

ength from very long to quite short; since this
bage appears to reflect a difference in the

lame hunted, separate types have been designated
nthe basis of size.

IAla (fig. 4, J-1). Long point (20 to
12 cm.); heavy oblate elliptical cross sec-

tion or rounded anterior face, flat posteri-
or; simple sharp or rounded tip; two (occa-

sionally one or three) high, enclosed (occa-
sionally isolated), unilateral barbs; bi-

lateral opposed line shoulder with maximum

projection on barb side; broad truncated base,
beveled laterally and on the edges, or occa-
sionally rectangular. Total of 5 definite,

1 probable.23 Distribution: Hum-67 (6 [3 in
the Stuart collection]). (Gifford, 1940,
pp. 183, 233; Loud, p1. 21, fig. 3.)

IA2a (fig. 4, m, n, p-r).   Long point

(18 to 15 cm.); heavy oblate elliptical cross
section or rounded anterior face, flat pos-
terior; rounded, slotted tip; two high, en-

closed, unilateral barbs, with torque; 24 bi-
lateral opposed line shoulder with greater
projection on the barb side (occasionally
line guard on side opposite barbs, as in

fig. 4, m); broad truncated base. No com-

plete specimens found, but one base fragment
and one tip fragment found together define

the type. Total of 9 definite specimens, 7
probable. Distribution: Hum-118 (16).

IIAla (fig. 4, s-x). Medium to short

bone point (10.5 to 5.5 cm.); thin to medium
elliptical cross section; simple sharp tip;
two high, enclosed, unilateral barbs; bi-

lateral opposed line shoulder with greater
projection on barb side (occasionally line

guard on side opposite barbs, fig. 4, u, or
unilateral line shoulder, fig. 4, t); broad
truncated base. Total of 9 definite speci-
mens, 3 probable. Distribution: Hum-67 (4,
Stuart collection); Hum-118 (5 definite, 3
probable).

IBla (fig. 5, a). Long point (17 to
14 cm.); heavy elliptical cross section;

simple rounded tip; two medium, high,enclosed,
unilateral barbs, occasionally with torque;

bilateral opposed line guard; truncated base.
Total of 2 definite specimens, 3 probable.

Distribution:   Hum-118 (1 definite, 1 prob-

able; Hum-169 (1 definite, 1 probable); Hum-
170 (1 probable).

IB2a, IB2b (fig. 5, b-g). Long point

(18 to 13.5 cm.); heavy elliptical cross sec-
tion, flat on barb side; rounded, slotted
tlp; two high, enclosed, unilateral barbs,
with torque; bilateral opposed line guard;
truncated base; occasionally incised; one

fragment is made of bone. Total of 19 defi-
nite specimens, 10 probable. Distribution:

23Probable specimeni consist of fragments which retain a
trace of some diagnostic feature or those which came from
significant depths at Hum-169 or Hum-118, where a strati-
graphic succession of harpoon types was found.

2SThe term "torque" is used when one barb is intention-

ally offset from the medial plane of the harpoon. Most fre-
quently it is the lowest barb which is twisted to the right

or left (fig. 6, a) but occasionally both barbs exhibit some
degree of torque (fig. 6, b).

I

299

ANTHROPOLOGICAL RECORDS

subtype a, with simple barbs, Hum-118 (15

definite, 8 probable); Hum-170 (2); subtype
b, with hooked barbs (fig. 5, g), Hum-16925
(2 definite, 2 probable).

IIBla (fig. 5, J-p). Medium to short

bone point (12 to 5 cm.T; thin to medium el-
liptical cross section; simple sharp tip;

two (occasionally one) low to medium high,

enclosed (occasionally isolated), unilateral
barbs; bilateral opposed line guard; trun-
-cated base; often incised. Total of 19

definite specimens, 3 probable. Distribu-
tion: Hum-67 (1, Stuart collection); Hum-
118 (14 definite, 2 probable); Hum-169 (3);
Hum-170 (1 definite, 1 probable).

ICla (fig. 5, h, i).   Long point (16

to 14.5 cm.); heavy elliptical cross section;
simple sharp tip; two high, isolated, uni-

lateral barbs; unilateral line guard on barb
side; truncated or rectanguloid base. Total
of 2, both in Stuart collection. Distribu-
tion: Hum-67 (1); Mad River Beach site (1).

IC2b (fig. 6, a-f). Long point (20 to
17 cm.); heavy elliptical cross section;

flat on barb side; rounded, slotted tip; two
high, enclosed, unilateral barbs with hooked
tips and torque; unilateral line guard on
barb side; truncated base; occasional in-

cising. Total of 20 definite specimens, 1

probable (some of the 8 probable IB2a may be
type IC2). Distribution: Hum-118 (4, one
reworked into punch); Hum-169 (19 definite,
1 probable; two reworked into punches; 4 in
Stuart collection); Hum-124 (2, Stuart col-
lection) .26

IC3b (fig. 7, d'). Long point (27 and
18 cm.); heavy elliptical cross section;

grooved tip, with inset for triangular metal
point (point is beveled on edges and has a

projecting stem for attachment); two (or one)
high, enclosed, unilateral barbs with hooked
tips; unilateral line guard on barb side;
truncated base. Total of 2 ethnographic

specimens, one from the Tolowa (fig. 7, d';

also Gifford, 1940, fig. 23, type NNlb), and
one from the Yurok (Gifford, 1940, fig. 21,
type NNla).

Both specimens also display the same
type of attaching line, made of deerskin

with the hair left on, with a short slit on
the nonbarbed side for a strong double
binding.

IICla (fig. 6, g-i). Medium to short
point (12 to 7.5 cm.); medium elliptical

cross section; simple sharp tip; one, or two
enclosed, unilateral barbs; unilateral line
guard on barb side; truncated base. Total
of 4 definite, 1 probable. Distribution:

Hum-67 (1, Stuart collection); Hum-118 (2);
Hum-169 (1 definite, 1 probable).

ID (fig. 6, J). Single base found,

5.5 cm. long; heavy elliptical cross section,

25Waterman, 1920, site 31, rect. K, map 33.
26Ibid., site 35, rect. I, map 31.

with flat edge on shoulder side; unilateral

line shoulder, with biconically drilled line
hole in upper corner; broad spatulate base,
beveled on both sides of shouldered edge.

Total of 1, from Hum-67. (Loud, p. 384, p1.
20, fig. 8, a, b.)

The use of type I simple harpoon is well
authenticated in reports on historic Indians.
The slotted tip, stone point, and distinctive
harpoon shaft (fig. 7, c'), as well as the ma-

nipulation of the harpoon, are described in de-
tail for the Yurok of site Hum-169 by von Loef-
felholz.27 Kroeber gives an account of the use

of large two-barbed specimens by the Yurok sea-
lion hunters.28 After disguising themselves in

bear or deer skins, they stationed themselves on
convenient rocks and barked to attract attention.-
After an animal had been harpooned "no attempt

was made to hold the bulky prey, but it was fol-
lowed by boat, the shaft regained, and then at
first opportunity the victim was speared again.
Sometimes a canoe was dragged out to sea for
half a day before the animal was dispatched."

Driver gives a somewhat similar account, stating
that a redwood float or drag was attached by
rope to the harpoon.29 The Bear River30 sea-

mammal hunters considered the task so dangerous,
on account of their unstable canoes, that all

possible precautions were observed first. 'The

Coast Yuki3' employed the implement only from the
shore, with swimmers retrieving the quarry.

It can thus be assumed that the large type
I harpoons were employed in hunting large sea-
mammals, particularly sea lion and seal. In

dealing with the use of archaeological specimens
from Gunther Island, Loud suggests that "sharks,

a kind of dog fish"32 were harpooned but he iden--
tifies no such faunal remains from the sites he
refers to. The smaller harpoons (type II) are

not mentioned ethnographically. Although their
shape is the same as that of larger examples,
their small size is unsuitable for large sea-

mammal hunting.   The smallest (e.g., fig. 4, u,
v) were probably used to harpoon salmon and

other fish while the medium-sized examples (e.g.,*
fig. 4, s) were perhaps employed more often for
sea otter.

A characteristic projectile point of stone
having a concave base (fig. 4, n, o) has been
identified as the form used to insert in the

slotted tips of the large harpoons. A stone tip
was found imbedded in a sea-mammal rib from Hum-
118.

A secondary use may be noted for large har-&
poon points. After breakage their tips were fre4

27Von Loeffelholz, p. 120.
28Kroeber, 1925, p. 86.

29Driver, 1939, no. 283, p. 380.
30Nomland, 1938, pp. 111, 112.
31Gifford, 1939, P. 335.
32 Loud, p. 381.

.:

I

I

300

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

quently reworked into pointed implements (figs.
5, b, e, and 6, f, g), probably to be used as

punches, with the base serving as a convenient
handle.  One specimen illustrates the initial
grooving necessary to cut off the broken barb

(fig. 4, r).  A total of six finished specimens

is known, including two type IC pieces from site
Hum-169 (not illustrated). Two slotted tips

were found which had been cut off just below the
slot (fig. 4, i). The tips of two probably IIA
specimens were apparently sharpened for use as
awls (fig. 4,   .

Occasionally, specimens of types IB2, IIB,
(fig. 5, b, J, 1, ?, ) and IC2 (fig. 6, d, f,
M bear incised patterns on their bases and
barbs. These may represent property marks

needed, after a successful hunt, to determine

the owner of the fatal harpoon. Since the in-
cising is not characteristic, however, these

lines more likely reflect the native desire for
decoration which pervaded Northwestern Califor-
nia. At present the known incised harpoons are

limited to site Hum-118. Short horizontal lines
around the base and heterogeneous scratches,

perhaps resulting from manufacturing techniques,
are the most common motifs.

Bilaterally barbed sea-mammal harpoons, un-
,supported by actual museum specimens, are illus-
t, trated in two ethnographic reports. The Coast

Yuki33 obtained a harpoon with single bilateral

;-barbs from the north. A complete description is
[given of the whole implement as well as of its
,use.  The method of attachment is noteworthy

(fig. 7, p), for an archaeological specimen

would probably not be recognized as a harpoon,

but would be considered a fixed projectile point.
This is also true of several British Columbia

ethnographic pieces34 whose simple bases show no
special means of line attachment. The Tolowa35

bilaterally barbed harpoon (fig. 7, b'), of dif-
ferent form and with a line hole, is significant
in view of the rarity of this type of line at-
tachment in California.

CENTRAL CALIFORNIA

The following artifacts from Central Cali-
ornia are also included as harpoons, though

ome specimens may represent fish spears. There
8 no evidence of the use of the simple harpoon
ong the historic Indians of the area. Con-

lusive proof that the principle of the harpoon

known in the Sacramento Valley is afforded

one carbonized specimen from site Sac-16, now
the State Indian Museum at Sacramento (fig.
, a). The harpoon point was inserted in the

plit end of a wooden rod, which still retained

idence of binding. The split ended in a round
le, through which passed a fiber cord., now a
33Gifford, 1939, fig. 6, p. 335.

3PTeit, 1900, fig. 222, g; idem, 1906, fig. 87; Barnett,
34, p. 279.

35Drucker, 1937b, f ig. 2, p. 237.

charred remnant. To the rear of this hole the
wooden rod tapered to a narrow stem. The in-
ference is that this assemblage represents a

simple harpoon, the detachable point originally
having been fastened by a line passing from its
base through the hole in the shaft or foreshaft.
The stemlike projection suggests that the frag-

ment may have been a nondetachable foreshaft in-
serted in a cane or hollowed-out wooden shaft.

Judging from ethnographic use of the toggle har-
poon, it may be safe to assume that the imple-
ment was not thrown, but only thrust, at its
target, probably salmon and other large fish,
and that the short pay-line was therefore no

hindrance. The (pitched?) binding or some im-
pediment inserted in the split end would hold
the line in the foreshaft after the harpoon
point had been pulled out.

Further evidence of the harpoon principle
in the artifacts under discussion is the fre-

quent presence of developed facilities for line
attachment, including various line guards, line
shoulders, and grooves. Some specimens have no
such features (fig. 3, 1, m, but pitch or as-

phalt was probably utilized to make the.connect-
ing line secure and the foreshaft may have been
the detachable element, or some special means of
wrapping and knotting may have been employed, as
in certain ethnographic specimens (see fig. 7,

p, and fn. 34). At the same time the possibil-
ity that these artifacts are fish spears cannot
be ruled out entirely. Unbarbed leisters were

used by some historic tribes, along with toggle
harpoon, and such groups as the Yahgan3 and

SeriM had both the simple harpoon and fish spear,
while the Ona borrowed the Eastern Yahgan har-
poon form but applied it as a fish spear.38

One other aspect of this problem may be

mentioned. It is frequently found that basal

ends of stone projectile points predominate at

habitation sites, the explanation being that the
broken bases were not replaced on retrieved

shafts until the return to camp. This might

also account for the occurrence of fragmentary
fish spear points, while it would be less true
for harpoons. But apparently this is not a

valid assumption. In Northwestern California

basal fragments exceed tip fragments by a ratio
of five to four, indicating perhaps that broken
harpoons were often brought back to camp. A

similar preponderance of bases is found in the
Central Valley for types regarded as harpoons,

whereas tip fragments predominate for types con-
sidered to be fish spears. The greatest number
of fragments are medial sections, probably re-

sulting from decay or accidental breakage at the
site. It is therefore felt that the original

assumption is not applicable to perishable items,
and the ratio of fragments may be regarded as

fortuitous. Since many fragments occurred with

36Mason, 1902, pl. 2.
3Hardy, p. 290.

38Lothrop, fig. 36, p. 82.

301

ANTHROPOLOGICAL RECORDS

burials, it may be correct to assume that decom-
position is largely responsible; the large num-
ber of whole points included as grave furniture
and the rarity of joinable fragments found with
burials do not provide evidence of the "killing"
noted for other utilitarian objects in this area.

Gifford39suggests that the unilaterally

barbed types were possibly correlated with ma-

rine mammals, whereas those with bilateral barbs
were associated with fluviatile uses and fishing.
Fuller information does not support such a broad
generalization, but type I (Gifford, type NN1)

was probably used in the taking of marine forms,
whereas NN2 is characteristically an implement
for river fishing. Unilaterally barbed types

from the fluvial Delta region have been classed
as NN3.

Unilaterally Barbed Types

NN3a (fig. 3, q). Medium point (9 cm.
restored); medium elliptical cross section;
simple sharp tip (?); two convex, enclosed,
unilateral barbs; unilateral line shoulder
on barb side; rectanguloid base. One frag-
ment, from CCo-138.

NN3b (fig. 3, r-t). Long point (26 to
24 cm.); rectangular to curving cross sec-

tion conforming to original shape of antler;
simple medium sharp tip; three convex, iso-
lated, unilateral barbs; unilateral spurred
base on either edge. Figure 3, t has the
suggestion of a bilateral line shoulder.

One medial fragment from site Sac-21 is in-
cluded in this type, although the base may
have been different. Total of 10. Distri-
bution: CCo-138 (9); Sac-21 (1). (Lillard,
Heizer and Fenenga, p1. 29, U.)

Bilaterally Barbed Types

NN2aI (fig. 2, u-z). Medium point (13
to 9 cm.); flattened to heavy elliptical

cross section; simple sharp tip, often long;
four (most numerous) or five enclosed (oc-
casionally isolated, fig. 2, u, y) barbs,
bilaterally opposed; bilaterally opposed
line shoulder, or projecting line guard

(fig. 2, u), occasionally staggered (fig. 2,
v, w); variable base: simple rectanguloid
(fig. 2, z), truncated (fig. 2, u, v), oc-
casionally with lateral bevel (fig. 2, x).
Total of 27 specimens, 2 with 2 burials, 3

with 1 cremation. Distribution: Sac-6 (3);
Sac-21 (7); Sac-64 (2); CCo-138 (15).

NN2aII (fig. 2, a'-d'). Medium point
(10 to 8.5 cm.); medium elliptical cross

section; simple sharp tip of variable length;
three convex, enclosed, bilaterally opposed
barbs, with occasional tendency to angular-
ity (fig. 2, c', d'); bilaterally opposed
line shoulder; rectanguloid or truncated

39Gifford, 1940, p. 183.

base, often with lateral bevel, occasionally `
with faces beveled on sides and end (fig. 2, A
c').  Total of 6 specimens, 2 with 2 crema-   i
tions. Distribution: Sac-6 (1); Sac-21 (1);

Sac-107 (1); Sac-122 (1); CCo-13'8 (1); Nico- X
laus (1).

NN2aIII (fig. 2, e'-i'). Long to medi-..,
um point (15 to 12.5 cm.); cylindrical cross i
section; very long, sharp, simple tip; three ,
angular, enclosed, ridged (fig. 2, g', i')
barbs, bilaterally opposed; bilateral line
shoulders, slightly staggered; rectangular
base, with lateral bevel and sides and end

of faces beveled.   This type is exceedingly
uniform and shows the most careful workman-
ship.  Total of 6 specimens.   Distribution:
SJo-80 (2); SJo-82 (1); SJo-91 (2); CCo-141
(1).  (Gifford, 1940, type NN2a, pp. 184,
233; Heizer and Fenenga, fig. 1, no. 14;
Moorehead, fig. 412, no. 4, p. 273).

NN2aIV (fig. 7, k).   Long point (16
cm.); triangular to flattened elliptical

cross section; long simple sharp tip; single,
low, bilaterally opposed barbs; deep bilat-

eral notches serve for line attachment;       I
stemmed base; probably made from axillary
border of elk scapula. Single specimen,
from Sac216.

NN2bI (fig. 3, a-j).   Long to medium
point (17 to 8 cm.); great variation in

width; flattened to heavy elliptical cross

section; simple medium to sharp tip; enclosed'

(some isolated), convex (occasionally concave 4
or angular), bilaterally staggered barbs, the I
number of which varies from four to six (oneX
specimen has 2, while 5 barbs are most typi-
cal); unilateral or bilateral opposed or

staggered line shoulder (occasionally line'

guard); rectangular, truncated-spatulate, or S
spatulate base.  On occasional examples from
CCo-138 the barbs on one side are longer

than on the other (fig. 3, d). An equal num-
ber of barbs on each side is characteristic.
of both NN2a and NN2b. Figure 3, j is prob-
ably a reworked specimen, having three barbs
on one side and four on the other, the re-
sult, quite likely, of a broken tip. Total

of 19 specimens, 3 with 3 burials. Distribu-
tion:  CCo-138 (7); CCo-139 (1); Sac-6 (1);
Sac-21 (6); Sac-56 (1); Sol-236 (1); Cross

Slough (1); Sacramento Valley, site unknown

(1). (Gifford, 1940, type NN2b, pp. 184, C
233; Lillard, Heizer and Fenenga, pl. 29, h, L
i, v, w; one in Stuart collection).

NN2bII (fig. 4, a-g).   Long to medium
point (15 to 9 cm.; one lacking tip is over
15 cm.); variable width; flattened ellipti-
cal to cylindrical cross section; simple

sharp tip, often long; three (one specimen,

fig. 4, a, has 2 barbs on one side, 3 on the
other, perhaps the result of breakage) iso-

lated (occasionally enclosed or ridged), fre-
quently angular, bilaterally staggered con-
vex barbs; bilaterally opposed or staggered
line shoulder (rarely projecting line guards
and one grooved); rectangular to truncated

302

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

or spatulate base, occasionally laterally
beveled. Total of 10 specimens, 2 with 2.

burials, 1 with 1 cremation. Distribution:
CCo-138 (5); Sac-16 (1); Sac-21 (1); Sol-2
(1); Cross Slough (1); Nicolaus (1).

Out of a total number of 73 type NN2b
specimens, only 28 could be typed. Of the

remaining 45 fragments, 3 were found with 3
burials, 6 with 2 cremations. Distribution
of fragments: CCo-138 (31); CCo-250 (1);
Ker-46 (1); Sac-6 (3): Sac-21 (1); Sac-60
(5); Cross Slough (2); unknown site (1).

NN2c (fig. 3, k-p).   Long heavy point
(20 to 13 cm.); flattened to heavy ellipti-
cal cross section; simple sharp tip, some-

times long; 8 to 5 convex (occasionally con-
cave), enclosed (occasionally isolated), bi-
lateral barbs, with indefinite placement

(both opposed and staggered on the same har-
poon); bilateral line shoulder (1 bilateral
line guard) or no special line attachments;
variable base, with emphasis on simple rec-
tanguloid, but occasionally spatulate (fig.

3, n) or truncated. The bulges on the right
side of the harpoon shown in figure 3, p are
evidently the remnants of broken barbs which
have been smoothed and rounded. Total of 9

specimens, 4 with 2 burials, 1 with 1 crema-
tion.  Distribution:   Sac-21 (1); Sac-107
(2); SJo-139 (1); Bethany (2); Byron (1);
Core (2).

NN2d (fig. 4, h). Long point (12.5
cm.); thin, bowed cross section; medium

sharp tip; two low, enclosed, bilaterally
staggered barbs; wide, pointed, spatulate
base. Single specimen, from CCo-138.

In general, among the bilaterally barbed
harpoons, no conclusive associations seem to

exist between type of barb and type of line at-
tachment or base.  There is a multiplicity of

forms, but examples of most types are not numer-
ous enough to support generalizations about the
linkage of specific features, and too few basal
fragments are known to necessitate the estab-

lishment of anything more than the broadest of
inclusive types of line attachments and bases.

The uniformity and distribution of type

NN2aIII imply that the inhabitants of the south
Delta recognized an ideal form of harpoon; all
specimens known so far, from various sites,

display the bilateral line shoulder and rectan-
'gular base associated with three angular, en-

closed, ridged, bilaterally opposed barbs. How-
ever, only four complete specimens have been
found. The unilateral line attachment is re-

stricted to type NN2bI at present, but this may
be because the sample of basal fragments of
-other types from CCo-138 is inadequate.

Local preferences, expressed by an empha-
sis on certain forms at particular sites, are
more apparent than linkages of construction

features, and may ultimately provide definite

indications of relationships between sites which
can only be suggested now. More samples from

each site are needed, however. The inhabitants

of Sac-66 and Nicolaus made great use of the bi-
laterally opposed line shoulder with truncated

base. This form of bilateral line shoulder also
occurs as Sac-6, Sac-16, and Sac-107 and at the
four NN2aIII sites. The unilateral line shoul-

der and line guard are restricted to CCo-138 and
CCo-139, except for single examples of the line

shoulder at Sac-21 and Sac-60. Also dominant at
CCo-138 are the bilaterally staggered line

shoulder and line guard. Sac-21 yields several
examples of bilateral line shoulders and guards
which show Just the slightest degree of stagger.
Sac-56 and Sol-236 have the staggered line

shoulder also. A long spatulate base with bi-
laterally opposed line shoulders is shared by
sites Sac-6 and Sac-107.

SOUTHERN CALIFORNIA

Records left by early Spanish explorers

leave no doubt of the presence of the harpoon.
Father Ascension, who sailed with Vizcaino, in
discuasing Santa Catalina Island, remarks:   "At
the end of the pole they fasten a harpoon made

of fish bone [whalebone?], and to this they tie

firmly a long strong line like twine.. .When this
is fast in the.fish, they give it the line if it
is a large one, and follow it. "1  Costanso, who

accompanied Portola, wrote of the San Diego area
(Diegueno territory ):41 "Their harpoons (fisgas)
are several yards long, and the point is a very
sharp bone inserted in the wood; they are so

adroit in throwing this weapon that they very
rarely miss their mark." The Santa Barbara
Indians (Chumash) also knew "all the arts of
fishing. tt42

Members of the Martinez Expedition43 in
1792 noted: "They also fish with tridents

(fisgas) and harpoons made of shell and flint."
It thus appears that the word fisgas was used
loosely for both fish spear and harpoon but,
since Costanso mentions that the weapon was

thrown, harpoon is probably the proper transla-
tion. In Priestley's translation, Fages de-

scribes the tridents; if fisgas is the original
term, it may mean harpoon. "The tridents they
use are of bone; the barb is well shaped and
well adapted to its use."44

None of these descriptions is complete

enough to allow identification of archaeological
specimens. However, the following types perhaps
correspond to the bone points referred to in

some of the Spanish chronicles. These artifacts
are considered harpoons because of their long
length and conical bases, which were probably

4?Wagner, p. 236.

4lCostanso, p. 33.
42Ibid., p. 49.

43Martinez, p. 44.
4Fages, p. 51.

I

I

303

ANTHROPOLOGICAL RECORDS

fastened permanently in the hollowed-out end of
a detachable foreshaft.

Wlb45 (fig. 6, l-p). Long point (23
cm. average); heavy elliptical to cylindri-

cal cross section; low, convex (occasionally
concave), unilateral single barb (rare ex-
amples have as many as 6 barbs, fig. 6, 1);

conical, occasionally spatulate base.    Usual-
ly of whalebone, occasionally of ribs.

Total of 23 specimens. Distribution: San

Nicolas Island 16 (9 in AMNH [Nelson, pp.203,
208]; 4 in UCMA [Gifford, 1940, pp. 178,
224, type W3; the last Wl illustrated by

him, ibid, p. 224,is also included, since it
is 12.1 cm. long in its present fragmentary
condition]; 2 in SI [Rau, p. 142, figs. 222,
223]; 1 in Moorehead, fig. 362, no. 7);

SaAta Barbara coast 5 (3 in AMNH [Nelson,
p. 203]; 2 in SI [Abbott and Putnam, p.

224]); Santa Cruz Island 2 (1 in Santa Bar-
bara Museum of Natural History [Orr, p. 117,
type W4]; 1 in MAI-HF [Hodge, p. 223]).

Santa Catalina, an unknown number of single
barbed specimens referred to by Hodge, page
226. San Nicolas Island is probably also
represented in the MAI-HF, for only types
unique to the island were listed on pages
227-229 of Hodge's guide.

W2 (fig. 6, q, r).   Long to medium
point (18 cm. is the only measureable ex-
ample); heavy elliptical cross section;

simple sharp tip; single low, bilaterally

opposed barbs; long narrow shaft ending in
conical base. Asphaltum used for attach-
ment. Usually of whalebone. Total of 2,

both from San Nicolas Island. (Moorehead,
fig. 358, no. 4; fig. 362, no. 2).

Sea otters, highly prized for their fur by
the Chumash and Gabrieleno, 6 were apparently
one of the main targets; other sea mammals and
large fish were harpooned.

COMPOSITE SIMPLE HARPOONS

Several composite simple harpoons are known
from Southern California; in these the foreshaft
becomes the body of the harpoon, and a stone
point and bone (or wood?) barb are lashed on

with string and asphalt. The only ethnographic
specimen known from the Chumash is of this type,
collected by the Vancouver Expedition of 1790-

1795 from Santa Barbara,47 shown in figure 7, z.
Many of GiffordIs types U2 and U348 would be of
the proper size and shape to serve as similar

harpoon barbs. However, these curved artifacts
are also like fishhook barbs illustrated by

Bowers49 and Heye,5 and it will require special
study to separate them according to use. Their
employment as toggle points has been suggested,5
as well as their use as prongs of fish spears,52
perhaps the tridents of the historic groups. It
is thus impossible to identify harpoon barbs

until a detailed study of all such problematical
artifacts from the Channel Islands reveals diag-
nostic differences.

While digging in Big Dog Cave on San Cle-
mente Island,53 Woodward found a composite-

harpoon foreshaft at a depth of ten inches. The
specimen (fig. 7, a), 23.5 cm. long, consists of
a chert point fastened with tarred cordage to a
cylindrical wooden foreshaft. There is a slight
shoulder 3 cm. back from the tip with a shallow
ovoid socket, 1.3 cm. long, 0.5 cm. wide, for
the attachment of a bone barb. An encircling

line shoulder provides means for line attachment.
The conical base probably fitted into a socket
in the harpoon shaft. It is very like the Van-
couver specimen in size, type of point, place-
ment of barb, and basal form, though it came

from an island occupied in historic times by the
Gabrieleno.  Woodward believes that the barb was
of the type shown in figure 7, b, c, and refers
to other barbs from "Wihatset village on Mugu
lagoon and from Simomo village site just north
of Mugu on Calleguas creek."5'

COMPOSITE HARPOON ARROWS

Other composite specimens are smaller and
can probably be regarded as harpoon arrows, to
be used in fishing. A bipointed bone piece

forms both the point and barb, being lashed into
the grooved tip of a wooden foreshaft at an

angle. Woodward found a foreshaft of this sort
in Big Dog Cave on San Clemente Island at a

depth of ten inches, a depth which "can be safe-
ly associated with the historic period." The
specimen is 16.5 cm. long, with a shallow,

slightly diagonal groove in the upper end sur-
rounded by tar and lashing marks (fig. 7, d).

The opposite end is tapered for insertion in the

45Gifford (1940, p. 178) and Orr (p. 117), set up four
types differentiated as to material (bird ormammal bone)
and length. In the present report these four types have
been reduced, on the basis of length, to two.

46Kroeber, 1925, pp. 630, 634.

arrow shaft, probably of cane.             The barb, as re-
constructed by Woodward (fig. 7, e), would seem

47Read, pl. XI, fig. 2, p. 106. Dalton, pl. XV, fig. 8;
on p. 231: "The British Museum possesses several other

harpoon-ends of this type from Santa Barbara, 2 of them be-
longing to the present collection."

48Gifford, 1940, pp. 177, 178, 233.

*SBowers, p. 575.

5oHeye, fig. 10, p. 84.

51Gifford, 1940, pp. 177, 178.
52De Laguna, p. 212, fn. 637.
53Woodward, pp. 284, 285.

54Personal correspondence from A. Woodward, June 21, 194& 8
Fig. 7, a-e, taken from tracings made by him.

I

I
i

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

to correspond with type T2bII of Gifford.55

Robinson56 illustrates a similar example from

Point Mugu.  Possibly those arrow points shown
by Irwin were harpoon arrows.57

T2bII (fig. 7, g-J).   Short length (3

to 7 cm.); circular or elliptical cross sec-
tion; curved or bowed laterally; bipointed;
asphaltum wrapping in middle.   Total of 12

in UCMA. Distribution:58 Santa Cruz Island,
sites 100 (1), 138 (1), 135 (1), 147 (6);
Santa Rosa Island (2); site 10, Ventura
County (1).

Nelson illustrates a "dart point,"59 found
on San Nicolas Island, which also may be a har-
poon arrow (fig. 7, f).

TOGGLE HARPOON

Only the composite variety of toggle head

was known in California, except for a simple bi-
pointed type, and the identification of most

archaeological specimens is virtually impossible
because of the lack of detailed ethnographic de-
scriptions and the frequent simplicity of toggle
points. Kroeber60 states that "the [toggle]
harpoon was probably known to every group in

California whose territory contained sufficient

bodies of water. The Colorado River tribes pro-
vide the only known exception...." He then de-
scribes this rather homogeneous fishing imple-
ment: long, slender, thrusting shaft; single

or, more often, double foreshaft often with un-
equal prongs; simple bone or wood socketed

toggle head wrapped with string and pitch, with
a short line fastening them to the foreshaft

(fig. 7, x, y).  A three-pronged harpoon is re-
corded for the Central Miwok.61 Although prob-
ably this harpoon was most often used with a

thrusting action, Bartlett62 describes how it is
thrown. In one variant, one point was bound

fast to the shaft, while the other detached like
a harpoon.63  The salmon was apparently the most
important game sought with the implement, which
is mentioned for most of the Indians living

within the range of the salmon. However, some
groups restricted the use of the harpoon, e.g.,

"Gifford, 1940, pp. 177, 223.
56Robinson, p. 150.

lIrwin, fig. 7, p. 21.

5Sites given in this paragraph refer to map in Olson, p.2.
"Nelson, fig. 3, d, pp. 204, 205.
60Kroeber, 1925, pp. 815, 816.

61Barrett and Gifford, p. 189; another unique variation
r the Miwok is a tip of obsidian, described by Curtis,
924, p. 131.

62Bartlett, p. 33.

63Kato: Driver, 1939, no. 239, p. 313. Hayfork Wintu:
egelin, 1942, no. 256, p. 174. Some Northern Mono and
rthern Yokuts groups: Driver, 1937, no. 157, p. 112;
insky, no. 156, p. 399; Gayton, p. 146.

the River Patwin employed it for pike but not

for salmon.664 Use of the harpoon was denied by

all Pomo groups living on Clear Lake, presumably
because of the absence of salmon.65

Sargent gives an interesting account of the
use of the toggle harpoon by the Wintu:

"At this instant a long black stick

shoots into the water, very like a pitchfork
with long, slender candle extinguishers on

each point. These caps are about four inches
in length and an inch around at the base,

made of wood, and tipped with a splint of the
deer's shinbone, which is sharp and strong.

The whole is smeared with tar; then each end
of a two-yard rope is fastened into the side
of each cap. The middle of this rope is

firmly attached, with slack-line by a slip-

noose to the handle of the spear, about half-
way up. When thrust into the fish, the spear
goes with such force that it pierces through;
the little caps come off the two prongs and
turn, and the fish swims clear of the spear,
held only by the rope, that tightens on the

handle with the strain. This spear mutilates
the fish, but secures many more than could be
taken with hook and line."66

Chever, probably describing Nisenan Indian

practices, gives an account of toggle fishing by
torchlight on the Feather River.67

Most, if not all, of the ethnographic re-
ports from California emphasize the importance
of varied nets, weirs, dams, and poisons68 in

the capture of large quantities of fish, partic-
ularly for storage. The toggle harpoons, fish
spear, and fishhook were subsidiary implements,
to be used more for supplying daily needs. The
development of two functional families of net
and weir fishermen among the Patwin69   implies
considerable antiquity of specialization in

technique, as does the common insistence in the
first-salmon ceremony that the first fish be

caught with a net or some means other than the
spear or harpoon.

Information available indicates that the
actual toggle varied considerably. The most
easily recognized is the three-piece spurred
type used especially in northern California

(fig. 7, t). Examination of ethnographic speci-
mens suggests consistent differences in spur and
point type, the attaching lines, and the cross
section between Northwest and Central Califor-

64Kroeber, 1932, p. 278.

65Gifford and Kroeber, no. 208, p. 134, groups Ha, Ci,
Ko, El.

66Sargent, p. 441
67Chever, p. 137.

68Distribution of nets and poisons given by Hewes,
1942b, fig. 33, p. 105.

69McKern, p. 248.

305

ANTHROPOLOGICAL RECORDS

nian toggles,70 as well as between those posses-
sed by ethnic groups within each area.

Limited excavation has produced several

varieties of toggle spurs, differentiated by the
form of channel and socket end primarily. The

typology proposed by Drucker7l has been expanded
here.

Ia (fig. 5, x-z). Short length (5.5

to 7.5 cm.); variable cross section; simple

rectanguloid socket end, occasionally simple

sharp end (fig. 5, y); central channel broken
by medial ridge; pointed barb end. Total of
4 definite, 1 probable. Distribution: Hum-
67 (2); Hum-169 (2 definite, 1 probable).

(Gifford, 1940, type MM1, pp. 183, 232; Loud,
p. 381, p1. 21, fig. 12, a, j).

Ib (fig. 5, a'). Short to medium

length (5.5 to 9 cm.); variable cross sec-
tion; forked socket end; central channel
broken by medial ridge; pointed barb end.
Total of 2 definite specimens (a paired

couple are counted as 1), 1 probable. The
probable spur is an unfinished specimen;

since home manufacture is thus demonstrated,
the tip was probably to be forked. Distri-
bution:  Hum-169 (3).

Type II, with scarffed tip, has not
been found in California.

IIIa (fig. 5, b'). Short length (4 to
6 cm.); triangular cross section; rectangu-
loid socket end; central V-shaped channel;

simple sharp barb-end. Total of 1. Distri-
bution: Hum-67 (1, Stuart collection).

IIIb(fig. 5, c'-g'). Short length (4
to 6 cm.); triangular cross section; forked

socket-end; central V-shaped channel extend-
ing from three-quarters of length to full

length; simple blunt barb-end. One specimen
from Hum-169 (fig. 5, e') and one from DNo-2
display flaring sides and better workmanship
and may represent a subtype. Total of 10.
Distribution: Hum-118 (5); Hum-169 (4);
DNo-272 (1).

One unfinished toggle from Hum-169 (11
in. deep) cannot be typed.

Other areas yield d.ifferent varieties of
composite toggle harpoons. The Sinkyone two-
piece spurred type73 probably refers to the

three-piece variety. The Klamath-Modoc74 area
had an unspurred type with a single long, bone,

awl-like point and pitched base (fig. 7, w), not
too different from that of the Yahi.75 Driver76

70Gifford, 1940, pp. 183, 186.

T1Drucker, 1943, fig. 4, p. 39.

72Waterman, 1920, site 10, rect. A, map 5.
T3Nomland, 1935, p. 154.

74Barrett, pl. 22, fig. 4, p. 251.
75Waterman, 1918, p1. 9.

T6Driver, 1937, no. 151, p. 112.

elaborates on a similar straight, unspurred type
found in the Southern Sierra Nevada, usually
made of cannon bone with only the front end
pointed, the base forming the socket. Both

shape and material emphasize the similarity of

this toggle point to awls, and it would be dif-
ficult to identify such archaeological speci-

mens. The Wintu77 also made an unspurred toggle
with a very short bone point, so generalized

that, by itself, it would be problematical (fig.
7, _). Thus very few archaeological specimens
have been suggested as points of composite har-
poons.

Tlf (fig. 5, t-v). Short length (2.2
to 4.0 cm.); rectangular to cylindrical

cross section; bipointed, often with body
broadest near the abruptly beveled base;

bone. Total of 11. Distribution: CCo-141
(1); Col-l (1); Redding site 7 (7); Hum-118
(1); Hum-169 (1). (Gifford, 1940, pp. 177,
222). The form of these artifacts appears

needlessly specialized for use as gorges or
composite hooks, but they might well serve
as points in the Wintu-type toggle; the

largest would fit into a three-piece toggle
spur.

Ul (fig. 5, r, s). Short length (5 to
8 cm.); flat elliptical cross section; sharp
tip; blunt base; bone. Total of 8. Distri-
bution: Hum-117 (1); Hum-169 (7). These

specimens from Northwestern California are
suggested as the tips of the three-piece

spurred toggle harpoon point. The artifacts
fit in the larger spurs satisfactorily, and
there is a similarity with ethnographic

specimens from the Yurok and Wailaki (fig.

7, t; Gifford, 1940, figs.l, 13, 16).  Small
chipped stone points (such as fig. 5, j) may
have been used in some small specimens.

The Central Yokuts claimed a bipointed

toggle point with a string tied around the mid-
dle, sometimes in a medial groove, apparently

similar to pieces often called gorgehooks. The
size varied from 2 inches to 4 inches. The

single or double foreshafts were sometimes made
of deer tibiae. One Chukchansi specimen in the
UCMA is made of metal, possibly a rake tine.78

"One informant claimed the point was drilled and
the cord tied through the hole. This seems un-
likely as no such attachment is reported from

any other Yokuts or neighboring tribe."79 How-
ever, this is the same as the Nisenan harpoon

point, seen before 1870,0 which consisted of a
bipointed toggle with a medial line hole, at-
tached to a quill which served as the socket

77Du Bois, fig. 6, p. 128.

78Driver, 1937, no. 151, p. 112; Gayton, fig. 10, J, lt,
pp. 72, 75.

T9Gayton, fn. 182, p. 75.

"?Chever, p. 137; Beals illustrates the implement, fig.
1, p. 341; fig. 7, q herein.

UM

I

1,

306

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

(fig. 7, q). The Indians of Yosemite Valley

(Southern Miwok) employed the bipointed type of
toggle harpoon without a line hole, similar to
the Central Yokuts form.81

Since the toggle harpoon was used in most
bf Cali!fornia and nails were easily substituted
for the point in the post-contact period, it

may be correct to assume that many other speci-
mens now included in Gifford's "Bipointed Ob-

jects" (type T) and some of the type U examples
were points of the composite harpoon. The

greatest difficulty with this supposition is

that no bone artifacts have been found in the

Central Valley which are distinctive enough to
be classed as composite harpoon spurs. Several

possible explanations may offer a solution: in-
sufficient excavation of Phase II settlements;
unspurred toggles (either awl-like, bipointed,

or Wintu-type); or wooden spurs like those some-

times used by the Pomo,82 River Patwin,83 Maidu,P
and Miwok.85 Some of these groups also had the

three-piece spurred toggle of bone or antler, so
the use of unspurred toggles seems most likely,

with a late diffusion of the spurred type across
the Central Valley.

The presence of single-piece toggle harpoon
heads of Arctic type86 on San Nicolas Island has
occasionally led investigators to assume abo-

riginal contact. T  However, as pointed out by
Heizer,88 there is historical evidence of the

occupation of the Channel Islands by Aleut sea-
mammal hunters imported by Russian fur traders
in the early nineteenth century, and this pro-

vides an adequate explanation for the occurrence
not only-of single-piece toggle heads, but also
of the ground slate point, spear-throwers, and
certain other foreign objects found on the
islands.

ANALYSIS

Recent years have seen the postulation of a
temporal sequence in Central California,89 as

derived from the archaeological record. Three

successive horizons have been proposed and work
is now under way to explore the possibilities of
refining these broad periods in order to recon-
struct the cultural development within each.

There are serious limitations of the extent
to which one element can be used to indicate re-
lationships between settlements, for as many

traits as possible must be studied to gain in-

sight into the position of a culture in relation
to its surroundings, However, suggestions can
be made in order to stimulate further study

which will verify or refute the tentative con-
clusions.  The amount of material from Central

and Northwestern California is sufficient to in-
dicate not only areal affinities, but also

changes through time, with the implication that

these fishing and hunting implements are perhaps
useful "time bearers" and key traits in the un-
raveling of local prehistory.

Perhaps the greatest handicap is the vari-
able sample obtained from each mound. Some

Bites have been thoroughly tested by scientific
excavation, whereas others are represented by
single pieces unaccompanied by data. To this

disparity in the number of artifacts per site is
added the perishable nature of bone and antler
and the fortuitous ratio of fragments. Burial

associations, so essential in present studies of
California archaeology, are rare in comparison

.with the number of artifacts. Nevertheless, the

81Illustrated by Godfrey, p. 62, and described by
Clark, p. 37.

82Gifford and Kroeber, nos. 209, 211, p. 173; bone spurs
seem to have been more common.

83Ibid., no. 211, p. 173; since bone and antler are
denied as materials, wood must have been used.

84Dixon, fig. 49, p. 196; bone or hard wood.

findings seem important enough for considera-
tion.

EARLY HORIZON FISH SPEARS

MM3a. Type MM3a is a characteristic cul-
ture element of the Early horizon, being found
at three of the four settlements recognized as
Early. The greatest number of specimens come

from site SJo-68, which is apparently the latest
of the Windmiller facies settlements.90 Crema-
tion appears earliest in Central California at
site SJo-68, and it should be noted that type

MM3a occurs with both extended burials and cre-

mations at this site. No examples have yet been
found from any Middle horizon site; apparently
the use of this probable trident was abandoned
at the close of the Early period.

MIDDLE HORIZON FISH SPEARS

MM2b. If the specimen shown in figure 1, j
is included in this type, the form first appears
in what may be the close of the Early horizon at
SJo-68. However, its aberrant form and associ-
ations suggest that it should be excluded from
consideration as an MM2b. All other examples

seem limited to the Middle horizon. The local

center appears to have been the Morse facies of
the Interior province, from whose settlements

come 74 per cent of the specimens. Of these, 64
per cent came from Sac-66, 10 per cent from

85Curtis, 1924, p. 131, wooden spurs; Barrett and
Gifford (p. 189) and Godfrey (p. 62), bone spurs.

86Nelson, fig. 3, e; Hodge, p. 229.

87E.g., Leroi-Gourhan, pp. 381, 382.
86Heizer, 1947, pp. 150-152.

89Lillard, Heizer, and Fenenga; Cook and Heizer;
Beardsley, 1948.

9"Heizer, 1949, fn. 15, p. 13.

I

m

307

ANTHROPOLOGICAL RECORDS

Sac-60. The greatest variation in form occurred
at Sac-66, as shown by figure 1, 1-p. Its ab-

sence from other settlements of this facies (Sac-
113, Sac-73, and Sac-107B) as well as the near-
est Middle settlement SJo-142 suggests a time
difference, or may be due to insufficient sam-

pling of these sites. Its presence in the Brazil
facies is also definitely of the Middle culture
period. Little is known of Sac-122 or CCo-139,
but the inference is that these also are Middle
horizon settlements. Since the type is lacking
from Orwood facies, another time or culture dif-
ference is perhaps indicated (see type 002).

In the Littoral zone type MM2b is of rare

occurrence. Of the four specimens from Ala-309,
three were found in Level VIII (probably Middle)
and one in Level V, which would be close to the
transition to Phase I of the Late horizon. Too

little is known of site SMa-372 to establish re-
lationships, but it is probably representative
of the Middle horizon. Such temporal associ-

ations are indicated by associated artifacts for
two specimens from Sonoma County, in the Gaffney
collection. It is thus apparent that these ar-
tifacts, though widely distributed, were not

typical of the Bay culture, for they occur in
the high levels of only one (Ala-309) of ten

sites definitely placed in the Middle horizon.
This suggests that possibly the development of

this type in the Interior province was late, and
that Sac-66 and Sac-60 were among the latest
Morse facies settlements to be occupied.

001. Type 001 adds to this last-mentioned
possibility. Only two Interior sites, Sac-66
and Sac-43, share MM2b and 001. One example
from Sac-43 (fig. 1, k) is evidently an 001

specimen which was reworked into an MM2b, prob-
ably after loss of its barb. This supports the
assumption of a developmental transition from
the unbarbed to the barbed type. Site Sac-43

yielded 37 per cent of type 001, but the rarity
of MM2b at the same site suggests some other

center of development, while Sac-66 is excluded
by the rarity of type 001.

In view of the proximity of Sac-43 to Sac-
60 it seems apparent that Brazil facies was

probably later in time than Morse facies, with
only a slight overlap at Sac-66 and Sac-43.

This theory is strengthened by the additional

occurrence of type 003 at Sac-43 and its absence
in the Morse facies.

Other Middle horizon specimens in the In-
terior province come from Sol-2, Sac-117, SJo-
59, Yol-52 and a northward extension to Col-l.

This wide distribution may Justify the inclusion
of specimens from such unplaced sites as Sac-16,
Sac-29, Sac-8 and Sac-49 in the Middle period,
though some may represent survivals into early
Phase I times of the Late horizon. Since the

type occurs all around Morse facies settlements,
it could be expected that, if these sites had
been occupied, the more advanced fish spear
would have been adopted.

Again, this type is rare in the Littoral
zone, one specimen occurring at each of the

three Middle horizon sites, Ala-309, Ala-329,

and SCl-356. The shallow depth of the Ala-309

specimen within Middle strata and the absence of
the implements from the numerous other settle-

ments of the Ellis Landing facies suggest a be-
lated entrance from the Interior Valley. The
sites yielding these fish spears may be the
latest Middle sites on the Bay. One other

specimen from SCl-356 was found in strata as-

signed to the Emeryville facies of Phase I, in-
dicating probable retention of type 001 at this
site.

The workmanship seems to have little rela-

tion to time. Both crude and well-shaped speci-
mens were found with the same burial at Sac-43,

and frequently both kinds occur at the same site.
There may be a geographical correlation, since
all specimens from Sac-49 (fig. 1, v, y) ex-

hibited the careful workmanship typical of the
area to the south, including Orwood facies (see
type 002).

002. This type, though essentially the
same as 001, is distinguished by the angular

barb and more careful workmanship. The sites at
which the type is found form a distinct unit at
the south of the Delta. Their location at CCo-
141 was in Middle horizon strata. None of the
other sites have been placed temporally. SJo-
140 produced two aberrant examples resembling
type 001, but it would be difficult to say

whether it was the earliest site to receive the
type in the area, or was only peripheral. All

31 pieces from one burial at SJo-87 differ from
the general type in being wider and in having a
thinner, more rectangular cross section. Since
no other examples are known from the site, it

cannot be said whether they represent the work
of one man or a site difference. Meredith re-
fers to a burial "in a mound in San Joaquin

County"91 which had two bilaterally barbed har-
poons in association with one type 002 fish
spear. The distribution of these harpoons
characterizes Phase I of the Late horizon.

Therefore, Bethany mound may represent another
period of cultural transition and suggests a
late development of type 002, an assumption
strengthened by the restricted distribution.
The absence of shared types between CCo-139

(with type MM2b) and Orwood facies (with type
002) also emphasizes this lateness. The fact

that all sites yielding type 002, besides CCo-
141 and Bethany, lack harpoons suggests that

SJo-87, SJo-140, Beecher Ranch, and Rock Creek
sites may also be considered Middle horizon
settlements.

A large number of pieces were often buried
with the dead, but more excavation is needed to
ascertain the significance of this trait. A

specialized fishing occupation may be indicated
(the SJo-87 specimens suggest the work of one
man) or perhaps some ceremonial value was at-
tached to the implements.

91Moorehead, fig. 412, nos. 1, 2, 3, p . 273. Two har-
poons (fig. 3, 1, m herein) in the MAI-HF, from Bethany

mound, are so similar that there can be no doubt of their
identity with those illustrated in Moorehead.

I

I
t

I

I.

w

308

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

The better workmanship of all type 001

specimens at Sac-49 may indicate an influence
from this southern area.

003. The double-barbed fish spear appears
closely associated with the single-barbed type.
It is first noted in the Middle horizon of the
Interior province, with no examples from the

Morse facies. Of the nine specimens known, six

were associated with one burial at Sac-43, along
with eight of type 001. The burial association
of one specimen at Sac-127 also establishes the
Middle culture period, while the lack of any

other type suggests a late adoption at this site.
Probably there were Middle horizon settlements
at Sac-49, Sac-122 and Yol-52. The specimens

from Sac-49 again show the most careful atten-
tion to finish. The only Littoral zone occur-

rence of the type is at Ala-309, where a single
specimen was found at a depth of fifteen feet,
probably late Middle horizon strata.

MM2a, MM2d, and MM2e. Type MM2a is limited
to Level VIII of Ala-309. The depth would sug-
gest a Middle horizon occurrence. Burial asso-
ciations at Sac-43 and Sac-29 allow definite
placement of type MM2d in the Middle horizon,
and the specimens from Sac-16 and Yol-52 are

tentatively assigned to the Middle period also.
The single MM2e specimen was found at site CCo-
250, which according to our present knowledge

exhibits only Phase I traits of the Late horizon.
The unique form of this artifact may indicate a
different use than that of the types being dis-
cussed, but it is also possible that the Middle
form of fish spear was retained at this site.

Despite the inadequacy of the material,

there is considerable agreement between the con-
clusions suggested by fish spears and those in-

dicated by the Middle horizon archaeological re-
mains as a whole. These Middle period specimens
are now being studied by Fenenga, who has found
the same temporal relationships between Sac-43,

Sac-60, Sac-66, and Sac-127 as suggested in this
discussion. There is also similar agreement be-
tween the conclusions presented herein and those
of Beardsley (1947) regarding the Littoral zone.

he retarded entrance of Interior traits is ap-
arent, as well as the unusual relationship of
la-309 with the Interior province during the

iddle horizon and of its close association with
he settlements of the south end of the Bay.

ne factor in the delayed diffusion of the fish
pear toward the west may have been a greater

phasis on net fishing in the Littoral zone, as
uggested by the frequent occurrence of probable
et sinkers in the settlements of the Ellis

ing facies.92

Nothing can be said as yet of the origin of
hese fish spears. Development from the Early
rizon type MM3a is improbable owing to the

issimilarity in form and the seemingly late ap-
arance of the Middle horizon types after the

ly leister had apparently fallen into disuse.
wever, nothing similar has appeared in the

rounding regions. The Middle horizon culture

92Beardsley, 1948, p. 12.

seems to have been well established before the

appearance of type MM2b. The localized distri-
bution in the center of the Interior province

does not suggest any simple diffusion from the
outside. The distribution of types 001, 002,
and 003 again supports the hypothesis of this

central focal point. The similarity of MM2b to
001 strengthens the possibility of local inven-
tion of the latter, with 002 and 003 as later

developments. At the same time, artifacts from
the Eskimo area bear at least a superficial re-
semblance to type 001, though they are classed
as compound fishhook barbs.93  Mathiassen94 il-
lustrates "leister prongs" in situ, but it is
also possible to interpret them as fishhooks.
Both Collins and Mathiassen illustrate arti-

facts,95 apparently similar to MM2b, which are
also considered fishhooks; but these specimens
have sharp tips and an angularity on the side

opposite the foot, to allow separation from the
shank, which are lacking in most California

specimens. Implements having a form similar to
type 003 but a different use include a Hawaiian
netting needle (fig. 7, 1) and an Eskimo bird
spear point.96

The abandonment of the fish spear in the
Interior zone can with some certainty be

ascribed to the entrance of the simple harpoon,
marking the inception of the Late horizon. The
implement was retained into the Late horizon in
the Littoral zone, which never received the har-
poon, but seems to have been replaced by other
fishing implements, especially by nets.

Wla. This type shows considerable vari-

ation; if there were more specimens, it would be
feasible to separate the Central and Northern

California variants (fig. 2, t, r, respectively)
from those of Southern California. Nothing is

known of the temporal associations of the South-
ern California pieces. In Central California,

the specimen from site CCo-295 can be placed in
the Middle horizon. Nothing is known of the as-
sociations of the fish spear from site Hum-67 in
Humboldt Bay; the artifact does not seem to be a
characteristic trait for the area. Figure 2, s
comes from the Colusa province at Col-2, of

Phase II strata; this provenience further empha-
sizes the dubiousness of the inclusion of this
specimen in this group. The simplicity, rare
occurrence, and difference in form (and func-
tion?) of these artifacts confuse any conclu-
sions concerning their significance.

NORTHERN MARITIME HARPOONS

Scientific excavation has been limited in
Northwestern California and there is detailed
information only on sites Hum-118 and Hum-169.

93Colllns, pl. 75, figs. 6, 7.

94Mathiassen, p1. 12, figs. 8, 9.

95Collins, pl. 75, fig. 8; Mathiassen, pl. 22, fig. 1
(no suggested use).

96Nelson, E. W., fig. 42, no. 8, p. 149.

309

ANTHROPOLOGICAL RECORDS

The sequence of harpoons revealed stratigraphi-

cally at these two sites can prouably serve as a
guide for the placement of scattered finds made
elsewhere in the area. Harpoon points were

found throughout site Hum-117 to a depth of 144
inches; the maximum depth of the mound was 156
inches, indicating that fishing and sea-mammal
hunting had been basic economic pursuits along
the coast for centuries.

The limited occurrence of type IAla at site
Hum-67 suggests that the extreme variation in

base form and number of barbs may be the result
of individual preference. Though utilizing the
bilateral line shoulder, the forms do not ex-

actly parallel Hum-118 forms. This type may be
the earliest form found at Hum-67, judging from
the depth association of the bilateral line

shoulder at Hum-118. The IA2a and IIAla speci-
mens from Hum-118 were found only in the lower
half of the shellmound. The examples of IIAla

from Hum-67, though varying in detail, are prob-
ably of the same general period.

At Hum-118 all the IBla and IB2a specimens
were found in the upper half of the site, above
60 inches, indicating the replacement of the

line shoulder by the line guard as a means of

line attachment. This replacement probably also
occurred at Hum-67, where one IIBla was found.
The harpdon with the bilateral line guard was

found in the lower levels only of Hum-169, from
42 to 71 inches. Limited excavation on Hum-170

yielded only harpoons with bilateral line guards,
one type IBla, two type IB2a, and two IIBla.

This site had not been occupied within the mem-
ory of informants who had lived at Hum-169, and
it is not recorded on any map of Trinidad Bay
since its discovery in 1775. Harpoons of type

IIBla also came from the upper levels of Hum-118,
extending as deep as 76 inches, and the lower
levels of Hum-169 from 30 to 67 inches.

More extensive relationships appear with

type IC. Four examples of type IC2b came from
the upper levels of Hum-118, overlapping with
type IB2a. Nineteen IC2b specimens were found

in the upper levels at Hum-169, and two similar
harpoons came from Hum-124, both sites being

historic villages. The unilateral line guard is
also a feature of the ethnographic type IC3b.

Two IC2 specimens from Hum-118 exhibit a slight
indentation on the side without barbs, opposite
the lower edge of the line guard. An ethno-

graphic harpoon from the Tolowa (fig. 7, d') has
a similar indentation. Two type IICla were

found in the upper levels of Hum-118, and the
harpoons of this type from Hum-169 and Hum-67
are probably correlated in time. Single ICla
specimens came from Hum-67 and the Mad River

Beach site; the Stuart collection has two ethno-
graphic specimens from the Requa Reservation,
which are probably of Wiyot origin.

The last change in sea-mammal harpoons

(type IC3b) is limited to single ethnographic
pieces from the Yurok and Tolowa. The antler
tip was modified for the attachment of stemmed
metal points, replacing those of aboriginal
stone.

The frequent occurrence of the same methods
of line attachment at Hum-118 and Hum-67 indi-
cates that this major modification of harpoon

points was not a local phenomenon, but affected
the whole sea-hunting industry in this area.

Since other features vary locally, it is prob-
able that home manufacture was more important
than trade. Conclusions are hampered by the

difference in the two sites: the occupational
deposit at Hum-118 contained abundant harpoons
and sea-mammal bones, whereas Hum-67 was es-

sentially a burial mound with a minimum of har-
poons. The oldest hunters yet known from the

area employed the bilateral line shoulder; this
eventually gave way to the bilateral line guard
and this, in turn, yielded to the unilateral

line guard, whose lateness is emphasized by its
stratigraphic position, wider distribution

(sites evidencing late occupation being the more
numerous), and ethnographic occurrence. Two

barbs seem to have been the preferred number in
the area, with a small number of single-barbed
specimens, especially at Hum-67. Many of these
last represent harpoons which have been reworked
after breakage of the second barb. Only one

three-barbed piece was found.  Barbs with torque
(figs. 4, s, 6, a, S) appear to have been used
throughout the whole period, with an increasing
emphasis in later times. The hooked barb (figs.
5, g, 6, a-c) is limited to ethnographic speci-
mens and type IC, except at Hum-169 where type

IB2b also has this feature; an origin at or near
Hum-169 may be indicated. It is interesting

that the basal forms are duplicated on both the
hunting and fishing harpoons.  The range of IIb
extends as much as 16 inches deeper than IB at
Hum-118, so the bilateral line guard may have
been introduced in fishing first.

The slotted tip is the dominant form

throughout the mound at Hum-118, Hum-169, and

elsewhere in coastal Yurok territory. However,
no examples occur at Hum-67 or at Mad River

-Beach, both within Wiyot territory ethnographi-
cally. Since the region did receive the same

basal forms as farther north, the simple tip ap- I
pears to be characteristic of this southern area.1
Likewise, no hooked barbs have been reported
from sites in Wiyot territory.

Type ID is found only at Hum-67, and may
represent a trade piece from farther north,

since the type was not found elsewhere archaeo-
logically. The only other examples of the line
hole noted in California are in two ethnographic
illustrations, unsupported by actual museum

specimens. One is from the Tolowa,97 the other
is a Nisenan toggle harpoon.98

Wider Distributions

Ethnographic information indicates that the

sea-mammal harpoon was probably used intensively,
by most coastal groups of Northwestern Cali-

97Drucker, 1937b, fig. 2, p. 237. (See f ig . 7, b I herein

98Beals, fig. 1, p. 341.   (See fig. 7, q herein); Chever,
P. 137.

U0

?j

I

2.

I

i

1

7

310

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

fornia as far south as the Mattole,99 while the
Sinkyone "speared" sea-mammals from canoes.100

Gifford'L' records for the Coast Yuki a sea-lion
harpoon head with single bilateral barbs, which
were obtained from the north. Unilaterally

barbed types were probably the aboriginal form;
they are apparently of rather recent adoption
among this Central Californian group, and the
bilateral barbs may well be a postcontact de-

velopment, unless the memory of the informant is
faulty.

Discussion of distributions to the north is
hampered by the paucity of comparative material.
Type IA, with the bilateral line shoulder, the
earliest yet found, seems to be restricted to

California. The closest resemblances seem to be
simple harpoons from the Aleutians, but none are
very similar.

Type IB, with the bilateral line guard, is

more widely distributed. The closest occurrence
is Just across the border in Oregon, at Lone

Ranch Creek.'02 Many elements of former culture
were shared between Hum-118 and the Oregon
shellmound, but, although a representative

sample was obtained from the latter site, only
one simple harpoon was found. This is in sharp
contrast to the 71 specimens from Hum-118, 29

from Hum-169, and even the 5 from two test pits
at Hum-170. It would appear that sea-mammal

hunting was of much greater economic importance
to the prehistoric occupants of Yurok territory
than to those of the Chetco. This is supported
by the greater abundance of elk and deer bones
at Lone Ranch Creek, whereas at Hum-118 sea-

mammal bones were found in excessive quantity,
often in layers, to the virtual exclusion of

land-mammal bones. That the slotted tip never

was used at Lone Ranch Creek is suggested by the
small size and rarity of concave base points.'03
A tip fragment of a small harpoon (II?) was
found in the Willamette Valley.'04

Type IB corresponds in base type with

Drucker's type I,105 which he finds to be charac-
teristic of the Straits of Georgia-Puget Sound
aspect.'06 Future excavation may well reveal a
continuous coastal distribution of this basal

form from British Columbia to California. One

specimen is known from the Interior Tsimshian,'0T

"Driver, 1939, no. 285, p. 314, lists the Tolowa, Wiyot,
Yurok, and Mattole.

l??Nomland, 1935, p. 153.

l?0Gifford, 1939, fig. 6, p. 335. (Reproduced in fig. 7,
p herein.)

102Berreman, pl. V B, no. 28.

3Ibid., type NBb, p. 29, pl. VII, A, nos. 15, 16.

10?Leatherman and Krieger, fig. 24, E, pl. 28, G. Only
the tip was found so the basal type cannot be given.

?05Drucker, 1943, f ig. 3, a, b pp. 36, 38
16Ibid., table 8, p. 120.

1OTIbid.

though other types are characteristic of this
Northern aspect.

One type IC2 specimen, lacking the hooked

barbs, was collected from the Tlingit.108     Type
IC3 occurs ethnographically in identical form

among the Tolowa. The hooked barbs and either
the slotted or grooved-inset tip were used by
the Naltunne Indians of Oregon.109 "Large har-
poon heads of bone, tipped with obsidian" were
found at a site Just south of Port Orford,

Oregon.110 Type ICla was apparently used on the
Upper Columbia."'

Except for the Naltunne and Tlingit speci-
mens, the only slotted tips noted were those of

the Haida harpoons ,12 although these may be more
closely related to certain Aleutian forms which
are made for a different type of point, with a
stemmed rather than a concave base."13

The closest occurrence of the line hole
outside California is in specimens from the

Upper Columbia,114 Vancouver Island,115 and the
Lower Fraser River."16

The following interpretation is derived
from the inadequate information available.

Centuries before white contact the simple har-
poon entered Northwestern California as a dif-

fusion from some northern center. The form for

both long and short harpoons probably included a
simple tip, usually-one or two unilateral barbs,
occasional torque, and bilateral line shoulders.

An apparently uniform culture was found at
Hum-118 from top to bottom, and the fact that
such diagnostic elements as zoomorph clubs

("slave killers")--one from the very lowest
level of the shellmound--tubular pipes,117

steatite dishes, antler wedges, bell-shaped

mauls, and similar items were already in use,
indicates that much of the Northwestern Cali-
fornian cultural pattern had been established
before site Hum-118 was first occupied. The

loSLeroi-Gourhan, fig. 622, p. 342. Specimen in British
Museum.

10Mason, 1902, fig. 18, p. 225.
"loJohnson, p. 50.

'10Collier, Hudson, and Ford, p. 79, pl. 9, i.

112Drucker, 1943, fig. 3, kc, p. 38, type III, table 8,
p. 120.

13Jochelson, fig. 9, p. 55; fig. 54 B, p, 82; pl. 23.
*11Collier, Hudson, and Ford, p. 79, pl. 9, h, _, 1.
15Smith, 1907, fig. 150, p. 368; Curtis, p. 66.
116Smith, 1903, fig. 18, d,_e fig. 50, a.

11TThe tubular smoking pipe is not found in Central Cali-
fornia until Phase I of the Late horizon, when it appears

in all provinces. If a southern origin is assumed for the

pipe, a relative dating may be applied to Northwestern Cali-
fornia, placing the sequence of harpoons being dealt with

here later in time than the inception of Phase I in Central
California. Some of the cremations described by Loud

(pp. 354-356) have most of the characteristics of preinter-
ment grave-pit burning (Schenck and Dawson, pp. 334-336),

which also appears in Central California for the first time
in Phase I.

311

I

ANTHROPOLOGICAL RECORDS

zoomorph clubs from Hum-118 are noteworthy since
most of these come from Hum-67, indicating prob-
able contact at some time between the occupants

of these two sites.118  Local residence had been
of sufficient duration, however, to develop

variant forms of harpoons. Thus the Hum-118

harpoon, with such features as the slotted tip,

occasional combination of line shoulder and line
guard, and the less variable base form differs

from the implement of the not-too-distant Hum-67
with its simple tip, inconsistent number of
barbs, and changeable base form.

Whether the slotted tip for concave-base
points was an invention of Northwestern Cali-

fornia or was derived from unknown forms ances-
tral also to those of the Tlingit, Haida, and
Aleut cannot be said at present. The slotted

tip as a structural feature is extremely char-

acteristic of the Yurok ethnographically and of
their territory archaeologically. In addition

to its use on the sea-mammal harpoons, the Yurok
employed barbed arrows with slotted tips;"19 one

fragmentary archaeological specimen was found at
a depth of 19 inches at Hum-118 (fig. 5, w).
The Yurok toggle spurs have a corresponding
forked tip, with archaeological counterparts

only from sites within Yurok territory, one of
which is an historic Yurok village. Just why

the distribution of the improved tip remained so
restricted is difficult to understand, though

groups which did not depend so much on sea hunt-
ing may have regarded the slotted tip as an un-
necessary complication.

After a considerable lapse of time the line
guard displaced the line shoulder as a means of
line attachment, all along the coast. It is

probably accidental that excavation should re-
veal the greatest concentration of these har-

poons at the peripheries of their most intensive
use, namely in Northwestern California and the

Straits of Georgia-Puget Sound aspect of Drucker.
The theory might be entertained that figure 4,
n, v represents a developing stage of the line
guard.

The hooked barb was added to type IB2 at

Hum-169, but was not accepted by the Hum-118 in-
habitants until the unilateral line guard came
into use. Finally the outside line guard was

eliminated, producing type IC, a form which was
just coming into use at Hum-118 when the site

was abandoned. Since Hum-170, having bilateral
line guard types of harpoons, is not shown on

maps of Trinidad Bay since its discovery, types
IB and IIB were probably in use before 1775.

Although the unilateral line guard was accepted
at Hum-67 and Mad River Beach site, the hooked
barb was not; again, this is perhaps a reflec-

tion of the diminished importance of sea hunting

118Loud, pp. 370, 374, p1. 18. Dr. Stuart, of Eureka,

has a large collection of zo6morphic clubs from Hum-67 also.

119Gifford, 1940, fig. 20, p. 237; von Loeffelholz, fig.
47a, p (nonbarbed); for similar Hupa specimens see Mason,
1889, p. 227, fig. 100, pl. 22.

to the south. Type IC gradually increased in
popularity till it became dominant in the his-
toric period among the Yurok, Tolowa, and some
neighboring Oregon tribes.

Stratigraphic position will have to deter-
mine when the line hole (type ID) made its ap-
pearance in the area. The occurrence of this

form of line attachment has been very sporadic
south of the Northern aspect of the Northwest
coast. In general, there is a suggestion of

late entrance by trade. Bilateral barbs, rep-
resented only by drawings of two ethnographic

specimens,1N apparently made a very late appear-
ance, quite possibly during the historic period.

Just as most articles of the material cul-
ture of Northwestern California, particularly of
the Yurok, exhibit greater technological per-

fection than is found in the implements of sur-
rounding regions,121 so harpoons-from former

Yurok territory display certain specializations
which connote more careful workmanship, such as

the slotted and forked tip, the hooked barb, and
incised patterns.

Sea-mammal hunting was evidently an inten-
sive industry along this coast and may have con-
tributed substantially to the condition of eco-
nomic surplus out of which developed the inte-
rior subclimax on the Klamath River: it might
be speculated that the apparent late spread of
the unilateral line guard north and south, and
the hooked barbs north, is an-expression of the
radiation to be seen in numerous cultural traits
from the Yurok focus.

The only unilaterally barbed marine harpoon.
from Central California is a medial fragment
from Mrn-275 on the Marin coast. The unique

character of the piece in the area has already
been cited as suggesting a transmission from a
northerly region by some wounded sea mammal.'22

There is no evidence that the Indians around San
Francisco Bay ever used the harpoon f6r hunting
sea mammals. Clubs were used by the coastal

Pomo .123

CENTRAL FLUVIAL HARPOONS

The present distiribution of bilaterally

barbed harpoons in Central California indicates
that they are typical only of Phase I of the

Late horizon. The majority occur in Hollister

facies settlements, listed in figure 9, b.  With
two exceptions, discussed below, only Phase I

determinants are associated with harpoons found
with burials. No association of these imple-

ments with clamshell-disk beads is known. One

harpoon fragment from Sac-56, placed in Phase II
by present information on the site, may indicate
retention of the implement at this site, or may

12'Drucker, 1937b, fig. 2, p. 237; Gifford, 1939, fig. 6,
P. 335.

121Kroeber, 1925, p. 1; Waterman and Coffin, pp. 30-33.
122Heizer, 1944, P. 397.
123Loeb, p. 169.

I

I

4

312

BENNYHOFF: CALIFORNIAN
suggest a Phase I occupation of the untested

part of the site. No other occurrence of har-
poons in the Mosher facies is known. The com-
plexity of the bilaterally barbed forms, the

rarity of complete specimens and burial associ-
ations, and the scarcity of bases confuse the

problem of possible relationships within Central
California.

Type NN2c has some indications of early oc-
currence. Included with one burial at Bethany
mound were two type NN2c harpoons (fig. 3, 1, M
and one type 002 fish spear typical of the

Middle horizon. This association suggests a

period of transition, marking the beginning of
Phase I of the Late horizon, at least in the

south Delta. The occurrence of both fish spear
and harpoon together may indicate the adoption
of a new trait by a resident population. The

generalized features of flat, concave, multiple
barbs and simple bases lacking attachments also
imply an ear.ly unspecialized form, when compared
with other types, contrasting with the more com-
mon harpoon of this area, type NN2aIII. The

SJo-139 and Byron site pieces are fragmentary or
too poorly described to be definite, but are not
NN2aIII.

Another burial association, at Core mound,

consisted of two type NN2c harpoons (fig. 3, .),
two large obsidian blades (34 and 32 cm. long,
respectively), and one pipe, with an extended

burial. Although the pipe establishes Phase I,
the obsidian blades seem most characteristic of
the Middle horizon, though they are extremely
rare. Again, however, a period of transition

may be represented. The other occurrences may
represent later modified forms; the Sac-21

specimen (fig. 3, p) has specialized line at-

tachments, while the Sac-107 pieces (fig. 3, n,
o) seem to represent local deviations.

In view of the burial associations, the im-
pression is gained that type NN2c represents the
earliest form of Central California harpoon.

The Bethany specimens might represent the proto-
type, but it is improbable that all the vari-
ation found in the Delta was of local origin.

It can be proposed that various undifferentiated
forms were earliest, with an emphasis on long,

rather heavy points, having numerous concave or
convex barbs with no established placement and
either no specialized means of line attachment
or bilateral line shoulders. It is doubtful

whether all other forms of attachment were de-
veloped in Central California, but the limited
distribution in the area of such types as the

unilateral line shoulder and unilateral and bi-
lateral line guard, as well as the number of
unique harpoon types, convincingly argues in
favor of considerable local inventiveness.

On the basis of quantity and diversity of
rhe remaining types, the center of development
ppears to have been the Hollister facies.

ere are numerous indications that these re-
ated settlements are probably later expres-

ions of Phase I than Bethany or Core sites.

Co-138 and Sac-21 are foremost in quantity and
ariation not only of harpoons, but of other

FISH SPEARS AND HARPOONS

traits as well, and presumably represent fully

developed Phase I settlements. From an analysis
of CCo-138 made by Chard (n.d.)., it would ap-

pear that this site was occupied in late Phase I
and early Phase II times. (No harpoons occurred
with burials having clamshell-disk beads). Par-
tial explanation of the multiplicity of artifact
forms probably lies in the fact that CCo-138,

Sac-6, and Sac-21 have been more thoroughly ex-
cavated than other sites being considered. This
factor, plus the unknown duration of Phase I or
the number of occupants of a settlement, pre-

vents any valid conclusion at this time regard-
ing the importance of fishing. CCo-138, Sac-6,

and Sac-21 are the only sites yielding more than
four harpoons. If the simple harpoon was of as
little importance in fishing in Phase I as the
toggle harpoon was for the historic Indians--as
its infrequent occurrence at most sites, rarity
of burial associations, and lack of standardized
types suggest--then the Valley groups may have
accepted the implement passively, whereas, be-
cause of the environmental stimulus, the Delta

inhabitants welcomed it. Fishing, no doubt, did
assume a larger economic role in this area of

swampy tule marshland, dissected by sloughs, al-
though the number of mortars and charred acorns
indicate that balanophagy was of undiminished

importance in the economy. However, greater in-
tensity of use rather than antiquity may be re-
sponsible for the increased variation in forms
of harpoons from the Hollister facies.

Harpoons with bilaterally staggered barbs
(NN2b) have the widest distribution, being the
only type which spread outside the Delta. The

fragment from Ker-46 in the southern San Joaquin
Valley stresses this dispersal, but the rarity

of the implement indicates its slight importance
in the area. Little can be said of the two sub-
types, owing to the limited number of examples.
NN2bI would be a natural derivative from NN2c

(possibly the SJo-139 and Byron specimens rep-
resent NN2bI), with a spread through the Delta

into the Bay province. The retarded entrance of
harpoons into the Littoral zone is indicated by
the retention of the Middle horizon fish spear

into Phase I times at site SCl-356, which never
received the harpoon, and also at site CCo-250
(assuming type MM2e is a fish spear), which

probably acquired the harpoon late. Sol-236 is
the only Bay settlement which received the im-

plement, and this fact emphasizes both the flu-

vial character of these harpoons and the culture
lag of the Alameda province in Phase I. The

Sol-236 specimen (fig. 3, -) is more like type
NN2c than like the CCo-138 examples of NN2bI.

No occurrence of harpoons is reported from the

Napa area. 124 Type NN2bII extends from CCo-138
to Sac-16, along the Sacramento River. The

peripheral occurrence and rarity of the harpoon
at Sac-16, associated with preinterment pit

12The fish spear from Clear Lake referred to by Hewes,
1942a, p. 100, is not one of the types considered here.

The harpoon illustrated by Beardsley (1948, p1. 1) for the
Late coast culture is an error.

313

ANTHROPOLOGICAL RECORDS

burning, suggest a late arrival at this site.

The limited distribution of the NN2a type
would imply a later local emphasis on the op-

posed placement of barbs. If certain specimens
included in NN2c, in which a close approach to
the opposed placement sometimes occurs, repre-
sent an already established NN2aI type, it is

difficult to understand why there should be such
limited retention, unless staggered barbs ex-

hibit functional superiority, which is not ap-
parent. The production of opposed barbs, then,
would seem to be a simple modification of the

staggered position or possibly a crystallization
of the opposed position from a prototype in the
south Delta having accidental placement, with a
delayed diffusion northward. The form and dis-

tribution of type NN2aIII provide the best argu-
ment for the existence of a recognized ideal

pattern of harpoon construction. This type is

found in a region which seems to have emphasized
artistic elaboration and superior workmanship,

as evidenced, for example, in the fish spears of
the Middle horizon, the angular serrations of

the margins of projectile points and "Stockton
curves," and the balance and symmetry of the

type NN2aIII harpoons. One possible reconstruc-
tion might derive this southern form as the
first type NN2a, itself a refinement of the

earlier unspecialized NN2c in the south Delta.
Influence may then have been exerted to the
north.

The effect of this pattern seems rather
definite in type NN2aII. The CCo-138 example
(fig. 2, d') exhibits slightly ridged barbs,
some angularity, and a somewhat rectanguloid

base. These features are also typical of cer-
tain NN2bII's from the site (fig. 4, d) which

perhaps underwent the same influence. The Sac-
107 specimen (fig. 2, c') retains the southern
base form and some angularity of barbs, while

the Sac-21 and Nicolaus inhabitants may have ap-
plied the three-barb idea to their own local
style of barb and base form (fig. 2, a', b').

The distribution of NN2aIII and NN2aII may re-
flect the stimulated contacts attendant on the

cultural efflorescence in the Delta at this time.
Possibly a further modification occurred at Sac-
21, with the acceptance of the opposed barb po-
sition resulting in a shift from NN2bI to NN2aI;
the limited distribution of the latter type

would support a late origin. Type NN2aI is most
typical of Sac-21. The conditions under which
the CCo-138 pieces were found are not clear,

hence their significance is problematical; the
type is not characteristic of the site numeri-

cally and may represent a brief period of influ-
ence or trade from Sac-21. The Sac-6 specimens
are only tip fragments. However, the above as-
sumptions have not yet been adequately corrobo-
rated. Very little is known of the temporal or
cultural relationship between the south Delta
and the Hollister facies.

A general sequence of change can be pro-

posed as a tentative working hypothesis: the.
earliest harpoons in Central California (type
NN2c) were long, heavy points, with no estab-

lished placement of barbs, and bilateral line
shoulders or no special means of line attach-
ment. It was not long before the staggered

placement of barbs came to be preferred and was
most widely distributed (type NN2b). A later

restricted emphasis on opposed barbs (type NN2a)
seems indicated for the south Delta and the Hol-
lister facies. A reduction in number of barbs
and over all size of harpoons may be a general

trend through Phase I. (The size difference may
be functional to some extent, since the extremes
in proportions of the implements shown in figs.

3, b and 4, c must represent a difference in in-
tended targets.) New variations in the forms of
line attachment were developed, with local pref-
erences becoming established. The significance
of the emphasis on three barbs is problematical
and may be the result of chance variation. It
is associated with traits which appear to be

both early and late developments, being'found

with large size (fig. 4, c) or generalized form
and indefinite barb placement (fig. 4, e).

Since our present information on the relative
contemporaneity of Phase I settlements is in-

adequate, these trends in harpoon construction
will be only suggested, and no attempt will'be

made to arrange the sites in a detailed temporal
sequence until more field data are collected.

Our knowledge of intersite relationships is
hampered by the rarity of specimens from most

sites. CCo-138 and Sac-21 share five of eight

types, with frequent similarity of special fea-

tures, and the occupation of these sites must be
considered at least partly contemporaneous.  The
CCo-138 inhabitants placed greater emphasis on
both unilateral and staggered line attachments
and barbs and developed a greater number of

types. The unilateral attachment may have been
developed at CCo-138, spreading thence to CCo-
139, Sac-21, and Sac-60.

The thorough excavation of Sac-60 revealed
a short Phase I occupation of this site. Its
location between Sac-21 and Nicolaus, the ab-
sence of type NN2aII, and the sharing of the

unilateral line shoulder with CCo-138 suggest
that the Sac-60 inhabitants had late contact
with CCo-138.

The similarity of Sac-21 and Nicolaus

specimens is especially marked, and a resettle-
ment of Sac-21 residents may be represented.

The distribution of the ridged barb and rectan-
guloid base suggests a south Delta influence on
the Hollister facies, being most marked at CCo-
138. Sac-6, Sac-107, and Sac-122 perhaps had
close cultural relationships also.

The archaeological field must be extended
before the conditions surrounding the initial

introduction of the bilaterally barbed harpoons
can be discerned. It is apparent that the im-
plement is typical of Phase I of the Late hori-
zon, a period which is distinguished by the ap-
pearance of certain diagnostic traits in the

Marin, Bay, Delta, and Colusa provinces of Cen-

tral C&lifornia.125  However, these determinants,

125Lillard, Heizer, and Fenenga, pp. 79, 80; Heizer, 1941,
pp. 106-1o8; Beardsley (1948, pp. 17, 18).

314

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

vary in number and kind in the different areas,
a variation which raises the problem of the

nature of the stimulus which produced the Late
horizon. "Late culture, in its general tenor,
is an outgrowth from that of the preceding
Transitional [Middle] period. 11126 Certain

earlier elements disappear, many more merely

change their form, and a few foreign traits are

encountered, among the most distinctive of which
are the harpoons. At present the Delta region
seems to be the center of harpoon development

and dispersal, though this is also the area of
greatest excavation. The bulk of the Phase I
inventory of the Delta province is not regis-

tered in the Colusa province until Phase II, and
it was assumed, in the absence of burial associ-
ations, that the occurrence of harpoons also

followed this pattern,'12  because of their full

development in Hollister facies and their rarity
to the north. However, at present, the simple
harpoon is found only on the periphery of the

Colusa province, Sac-16 being the northernmost
occurrence of the implement. This necessitates

the assumption that harpoons bypassed the Colusa
province in their original entrance of the Delta.

From only two settlements is there suffici-
ent material to allow placement in the Sandhill
facies, and even these are not as well sampled

as those of the Hollister facies. Therefore it
may be premature to claim that the harpoon is

absent in the north. In addition, more speci-
mens are needed from Sac-16, as well as more

adequate information on the cultural affilia-

tions of this site. If existing data are repre-
sentative, diffusion of harpoons within Central
California seems to spread from a Delta center.
The largest number of specimens, the greatest
variation (including the simplest forms), and
the clearest intersite relationships all are

found in the Delta. Sac-16, with two specimens,
one aberrant (type NN2aIV), seems to represent

the maximum diffusion northward. If so, simple
group-to-group diffusion seems unlikely as a

method of entrance, though the harpoon may have
appeared before the occupation of the Sandhill

facies (only two sites at present), or bypassed
them by following the Feather or American River.
The extreme variation in the Delta could then be
ascribed to more intensive use. Equally pos-
sible is the entrance of a migrant group or

there may have been special contacts between the
Delta inhabitants and foreign groups beyond the
Sacramento Valley.

Nothing definite can be said of the center
from which it is assumed that the Central Cali-
fornian types were derived, for archaeological
investigation has not been extensive enough to
reveal forms showing any particular similarity.

A theory of development from unilaterally barbed
marine forms found to the south might be enter-
tained, but this would be an extreme view. En-

trance from the west is negated by an apparently

,26Lillard, Heizer, and Fenenga, p. 79.

2TIbid., p. 80.

complete absence in the Marin, Napa, and Bay

areas (Sol-236, though on the Bay, can be con-

sidered peripheral to the Delta, in view of the
distribution of harpoons). Sierran groups may
have passed it down the rivers flowing into the
Delta from the east. A northern origin, among
groups dependent upon the salmon, seems most

likely, however. Northwest California, Oregon,
Washington, and British Columbia are dominated
prehistorically by the unilaterally barbed

simple harpoon;'128 however, since most of the
excavation in these regions has been done at
coastal sites maritime implements are often

represented. The bases of many lack line at-

tachments, perhaps implying their use as arrow-
or spearheads. The variety of line attachments
found in Central California is nowhere repre-
sented in this southern Northwest Coast area,

though the barb forms are much more heterogene-
ous. A very significant feature is the rarity

of the line hole in this region, a fact which is
in agreement with data from California and may
indicate the late appearance of this character
in the south.129

Bilaterally barbed specimens are rare,130
with little similarity to Californian pieces,
but future archaeology may reveal more types.

The California occurrence is isolated from the
main distribution to the north as outlined by
de Laguna.131 The southernmost occurrence of

either opposed or staggered barbs appears to be
among the northern Salish groups, the closest
resemblance at present being to ethnographic

specimens from British Columbia--the Lillooet,132
Thompson River,133 and Kwakiutl.l34 All seem to
lack line shoulder and guards, or line holes,

but their harpoon principle cannot be questioned.

One other aspect of this problem may be

mentioned. Archaeology reveals an efflorescence
of cultural development in the Delta province in
the Late horizon. White contact may have

shifted the center northward, for Kroeber found
the culture climax to be more in the Valley
proper, though rarity of informants from the

Delta may have influenced this conclusion. At
any rate the Delta province in that region ex-
hibits the most developed forms of the Kuksu

cult, for which some antiquity must be admitted.

128Northwest California: figs, 4, 5, 6, herein; Loud, pl.
21, figs, 3, 13. Oregon: Berreman, pl. V, B, no. 28;
Laughlin, 1943, fig. 24, pl. 28 G. Washington-British

Columbia: Collier, Hudson, and Ford, pl. 9; Hill-Tout, pl.
1, p. 109, pl. 3, p. 111; Rau, fig. 228, p. 145; Smith,

1907, figs. 104, 122, 123, 1141, 150, 159; idem, 1903, figs.
16, 17, 18, 50, 52; idem, 1900, fig. 337; idem, 1899, fig.

20. Most of the greater distribution to the north and east
is outlined by de Laguna, pp. 200, 201.

129Cf. areas farther north, such as Drucker, 1943, fig. 38.
13Smith, 1903, fig. 19.

131De Laguna, pp. 199-200.
132Teit, 1906, fig. 87.

133Idem, 1900, fig. 222, g.
134Barnett, no. 34, p. 279.

315

ANTHROPOLOGICAL RECORDS

Kroeber's135 derivation of the stimuli for these
societies from the Southwest rather than the

Northwest Coast appears valid, but it might be
noted that the appearance of harpoons, presum-

ably with northern affiliations and perhaps in-
dicating the entrance of a migrant group, sug-
gests the possibility that some of the features
of the cult may also have stemmed from the
northern center.

While the simple harpoon may have lingered
into early Phase II times in isolated areas of
the Delta, no association with Phase II deter-
minants has yet been uncovered, and the imple-
ment can be considered characteristic only of

Phase I. It is therefore necessary to seek some
explanation for the abandonment of this type of

implement, which contrasts sharply with the con-
tinuity, often with elaboration and expansion,
of most other Phase I culture elements.into
Phade II.

It is known that early mining operations of
the 1850's so muddied the rivers that fish

spears and harpoons were useless in both North-

ern and Central California, but there is no evi-
dence to support any similar clouding of the

rivers at the end of Phase I. Nor is there yet
any indication of belligerent invasions or war-
like neighbors which might have made fishing
dangerous during this period of change. The

continuance of net sinkers does not support the
theory of taboos against fishing.; other methods
of fishing, such as nets, weirs, and poisons,

may have merely become dominant, but there does

not appear to be any difference in the abundance
of net sinkers between Phases I and II which af-
fected the Interior zone as a whole.

It is therefore felt that the best explan-
ation for the disappearance of the simple har-

poon is that it was replaced by the toggle har-
poon. Evidence supporting such an entrance has
been presented under the distribution of toggle
types. The new form would be superior in many
ways. The operational principle of the toggle,
which virtually ensured the capture of a har-

pooned fish, was a marked improvement over the

simple harpoon whose low barbs probably allowed
more pierced fish to wriggle off the point. As
the toggles were small, they could be used on
two foreshafts, providing greater chances of

hitting the elusive target. The breakage of the
long, slender simple harpoon must have been a
recurrent disadvantage which the shortness of

the toggle point eliminated. The manufacture of
this shorter point required less careful work-
manship than the fragile multibarbed harpoons,
especially if the earliest toggle was of the
awl-like or bipointed variety.

It may thus be suggested that the appear-

ance of the toggle harpoon, as a diffusion from
northern centers, was welcomed in Central Cali-
fornia, and its superiority recognized to the
exclusion of the simple harpoon. The lack of
association between clamshell-disk beads and

135Kroeber, 1932, pp. 4o8-412.

harpoons in the Delta province suggests that the
toggle harpoon entered before clamshell-disk

beads. Present evidence would indicate that the
simple harpoon had diffused northward only as

far as site Sac-16. The Colusa province is in

the direct path of the most probable entrance of
the toggle harpoon, i.e., along the Sacramento
River from the northern center of use of this

implement. The absence of the simple harpoon in
the Sandhill facies supports the hypothesis that
the toggle harpoon had entered this area before
clamshell-disk beads, preventing a further dif-

fusion of the simple harpoon beyond Sac-16.* The
toggle harpoon may have spread southward, re-
placing the earlier harpoon, to be followed

closely by the disk beads. It is significant

that the simple harpoon is absent from all but
one settlement which used clamshell-disk beads
throughout its occupation.

Little can be said of the other types of

Central Californian harpoons, because of their
limited distribution. The single example of

type NN2aIV from Sac-16 may represent an early
form, but is more likely a degenerate specimen.
Types NN2d and NN3a both represent local spe-

cializations of CCo-138. Their rare occurrence
suggests the work of single individuals. Type
NN3b is present at CCo-138 and probably at Sac-
21. All specimens from CCo-138 seem to have
been found together in Phase I strata. The

unique base and limited distribution suggest an
original development in Central California, per-
haps at CCo-138, toward the end of Phase I. In
its lack of interest in unilateral types this

region contrasts with most other fishing areas.

SOUTHERN MARITIME HARPOONS

Little can be said about harpoons from

Southern California since for most specimens de-
tailed data as to their location are lacking.

Only certain composite pieces have any temporal
associations, which Woodward would place in the
historic period.136

"Bone harpoons are rare in the Chumash col-
lections, and when occurring are provided with
several barbs, while the Shoshonean islanders
had numerous harpoons, but all of the single

barb type."137 This observation should be vali-
dated by a complete study of the various scat-
tered collections from the Channel Islands.
Among the published artifacts three single-

barbed specimens were found in Chumash territory,
and the multibarbed specimen shown in figure 6,
J came from San Nicolas Island, though trade or
any number of explanations could account for
this. Until the number and distribution of

multibarbed pieces is available, speculation is
fruitless, but they would appear to be rare and
their concentration in Chumash territory might
be accidental.

136See p. 304 herein.
137Hodge, p. 216.

I

f

I

316

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

The years of "pot-hunting" which ravaged

the mainland and nearest islands may be respon-
sible for the great concentration of specimens
from the islands inhabited by the Gabrieleno,

particularly San Nicolas, the least disturbed of
them, which became the focal point for later

scientific excavation. It is also possible that
the Chumash preferred the composite forms, for

there is no implication by the early Spanish ob-
servers that the Gabrielenio did a greater amount
of sea-mammal hunting or fishing.

The great abundance of harpoon points from
San Nicolas Island is in striking contrast to

their rarity in the other islands, and in view
of the considerable archaeological collections
from San Miguel,138 this preponderance does not

seem due solely to the isolated position of San
Nicolas. Their numbers would seem to confirm
Kroeber's proposal that the restricted island

environment forced attention to the sea with an
emphasis on sea-mammal hunting; sea-otter furs
were probably the major export to the mainland,

in return for which the Nicolenio obtained neces-
sities lacking on their island.'39

Nothing can be said of the antiquity of

these types of simple harpoon, since the loca-
tions within the mound of the great bulk of
specimens are unknown. Present information

stresses the use of harpoons by the late cul-

,tures, but an analysis of possible harpoon barbs
,should reveal greater antiquity. They are not
'included in the cultural inventories of any of

,the successive periods proposed for the regionc40

There is not enough material for proper compari-
son with surrounding areas. The coast north of
the Chumash was probably ill suited to sea-

mammal hunting. The location of these artifacts
iimplies a marine use, substantiated by early

Spanish accounts. These accounts indicate that
the harpoon was used by the Diegueno, though no
specimens from their territory have been noted.
Harpoons were also used in Baja California, at
least in the lower half among the Waicura and

Pericue; a multibarbed specimen of wood has been
illustrated by Shelvocke.141 Actual pieces are
known from the Seri."42

Unfortunately both Seri specimens are made
with iron points, so the aboriginal form is un-

known. The unbarbed turtle harpoon is admirably
suited to the puncture of the carapace, and

McGee suggests that, aboriginally, both point
and foreshaft were made of sea-mammal teeth.

Such a form would be unsuited for securing large
fish and sea mammals, so the barbed harpoon

point of bone may well be the aboriginal fishing
form. Otherwise, the use of this single barb,

138Heye, 1921; Nelson, 1936.
139Kroeber, 1925, p. 634.

1?D. B. Rogers, 1929; M. R. Rogers, 1929; Olson, 1930.

1X1Shelvocke, facing p. 224; see also Costanso, p. 15.

1 'McGee, figs. 20, 21, pp. 187-193; Coolidge and Coolidge,
pp. 121, 122, 258.

with the point fixed in a detachable foreshaft,

would be an odd coincidence in view of the prox-
imity of similar forms in Southern California
and the peninsula.

"What is typical of Seri habits is that the
harpooa shaft is also used for poling, and the
harpoon cord as anchor rope, tow-line along

shore, sheet rope, and for tying a blanket to a
mast and pole to serve as sail...The whole har-
poon...dismounted...seems regularly to be taken
on a transport voyage."'143 This integrated func-
tion does not imply recent introduction of this
implement to the Seri. They once used a bi-

pointed fish spear,"44 but other factors may be
responsible for its apparent abandonment.

The wide use of this implement in this

peripheral region suggests antiquity, which sci-
entific excavation must determine. There is no
satisfactory explanation of a recent adoption,
and a very early entrance may be inferred for
these Southern Californian, Pericue, and Seri
harpoons.

Quite likely the Santa Barbara shafts and
foreshafts represent developments from types

similar to the Seri harpoon of cane or mesquite
root. The Pericue form seems to lack a fore-
shaft, but this was perhaps omitted by the

artist. It may be noted that the stone head,
affixed unilateral barb, shouldered base, and

detachable foreshaft of figure 7, a, b all find
counterparts in the harpoons from the coast of
Peru and Chile.145 A specimen showing super-
ficial resemblance to Southern Californian

archaeological forms is that found at a Develop-
mental Pueblo site in Utah 46 Although its use
was probably different, it may have developed
from the same ancestral type, or may represent
trade.

The harpoon or harpoon arrow with the bi-
pointed bone tip (fig. 7, d, e) also occurs on
the Chilean coast147 and is widely used on the
great rivers of the Amazon and Orinoco basins.

The Natchez,148 and probably other tribes of the
eastern United States, had a similar form.

HISTORIC FISHING HARPOONS

While the toggle harpoon was widely dis-

tributed in California among the living Indians,
the implement has been recognized archaeologi-
cally only in the northern part of the state.

Ethnographic Yurok toggle spurs in the UCMA con-
form to type IIIb, having the open channel and
forked socket-end. The specimens are aomewhat

1*3Kroeber, 1913, p. 21.

Hardy, p. 290.

145Mason, 1902, fig. 13, p. 215; Bird, pl. 123, J, k.
1"Gillin, fig. 22, p. 171.

17Bird, pl. 123, fig. P.

'148Swanton, p. 58; the bipointed arrow point was also com-
mon (see Tyzzer).

I

317

ANTHROPOLOGICAL RECORDS

larger and exhibit flaring sides (fig. 7, r)

similar to single specimens from the historic

sites Hum-169 and DNo-2. Three other examples

of type IIIb from Hum-169 occurred at depths of
26, 36, and 54 inches. Only one of this type

from Hum-118 has a complete socket-end, but at

least three others seem to have had forked ends
also.

One specimen of type IIIa, with a rectan-
gular socket-end, is known from Hum-67. The
type may be present at Hum-118, but complete
artifacts are needed to verify this.

Toggle type Ia is represented by three

specimens from Hum-169 found at a depth between
5 and 48 inches. The two aberrant specimens
from site Hum-67 may represent subtypes, but

more examples are needed. These Humboldt Island
pieces are extreme deviants, being larger, flat-
ter, and cruder than usual type I specimens,

with either an atypical socket-end (fig. 5, y)

or barb-end (fig. 5, x). The flat cross section
is more like Yuki and Coast Yuki ethnographic
toggle spurs than other examples of type I,

though no adequate comparison can be made when
the spurs are assembled in a toggle point. Al-
though Kroeber'49 states that the Yuki obtained
their toggle points from the Wailaki, museum

specimens from the two groups are not similar.'50
An ethnographic spur from the Karok is also

large and flat (fig. 7, s). A general increase
in height and width seems indicated between the
earlier archaeological specimens and the ethno-
graphic toggle spurs.

Two toggle spurs having the medial ridge
and probably the forked tip (type Ib) occurred

in the top 24 inches of Hum-169, while a pair of
specimens were found at about 72 inches. An un-
finished spur at 11 inches depth is larger than
type IIIb and may represent type I.

Type IIIb was apparently the characteristic
form of toggle harpoon in territory occupied

historically by the Yurok. Type I entered rela-
tively late and co-existed with type III at Hum-
169. The forked tip of type Ib was probably a
local modification.

Finished toggle spurs occurred as deep as
67 inches at Hum-118, and only one unfinished

piece was found below this, at 124 inches. This
perhaps indicates an increase in use of the

toggle harpoon in the later period of occupation.
Although two spurs comprise one toggle point,

over twice as many small simple harpoons (type
IIB) came from the same strata as toggle type

IIIb at Hum-118. There were nine toggle spurs
compared to five small simple harpoons at Hum-
169. It is therefore evident that the toggle
harpoon was slow to replace the small simple

harpoon in fishing, but apparently the implement
became increasingly popular among the later oc-

cupants of the area. Toggle spurs will probably
be found more frequently in sites along river-
banks.

149Kroeber, 1925, pp. 152, 174.

15Gifford, 1940, figs. 8, 13, p. 236.

Data on northern distributions are limited.
Type IIIa occurs at Lone Ranch Creek'5' and Bal-
lard152 sites, both in southern Oregon. Type I

is found in the Willamette Valley,153 the Dalles-
Deschutes,154 and is common in the Straits of

Georgia-Puget Sound aspect.155 As with the bi-
lateral line guard of the simple harpoons, the
toggle spurs with medial ridge extend to the

Gulf of Georgia, but no farther. Toggle spurs
with scarffed tips (type II) were used by the
Tlingit.156

It thus appears that the toggle harpoon is
of considerable antiquity in Northwestern Cali-
fornia, occurring to a depth of 124 inches at
Hum-118. Possibly the first settlers at this

site knew of the toggle harpoon but, if so, they
made little use of it until the late occupation.
Although the more intensive use of the implement
is represented in the same levels as those show-
ing the shift to the bilateral line guard in

simple harpoons, more dependence was apparently
placed on the small simple harpoon for fishing
at Hum-118. The occupants of Hum-169 employed
the toggle harpoon more often.

The type I toggle spur occurs from Humboldt
Bay to Vancouver Island; this wide distribution
might indicate that it is the older form. How-
ever, the medial ridge is a superior device

mechanically, providing a firm resistance to the
push of the toggle point as it enters the target.
In addition, the limited occurrence of this type
in Northwestern California, in late sites, indi-
cates a more recent entrance. It is therefore
suggested that type I represents a late diffu-
sion from the Straits of Georgia-Puget Sound

center down the coast, with sporadic acceptance.
The Hum-67 pieces may represent degenerate
copies.

The simple open channel and rectanguloid
socket-end (IIIa) seem to be features of the

earliest form of toggle point in Northwestern

California, being found from southern Oregon to
Humboldt Bay, and possibly in the lower levels
of Hum-118. By the time this implement was in
popular use at Hum-118 the tip had been forked;
increasing size and flaring width seem to have

been developing trends which led to the historic
toggle spurs. To date, the forked tip is re-
stricted to former Yurok territory.

One aspect of the toggle harpoon should be

kept in mind, namely that in the finished toggle
point the distinguishing features of a toggle

spur are concealed by the string-and-pitch bind-
ing. Therefore diffusion of the various forms

151Berreman, pl. V, B, nos. 29, 30.

152Leatherman and Krieger, pl. VIII, fig. 14. It may be

noted that no toggle spurs were found at Schwenn site, which
has indications of greatbr age.

153Laughlin, p1. 28, F, p. 226.

154Strong, Schenck, and Steward, p1. 6, J, n.

155Drucker, 1943, type I, pp. 39, 120, fig. 4, a, p. 4o.
156Leroi-Gourhan, fig. 712, p. 364.

I

e

I

4

318

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

might depend to some extent on observation of

the manufacture of the individual spurs. Thus

the concealment of the forked socket-end may ac-
count for the seemingly limited distribution of

this feature. The sporadic occurrence of toggle
type Ia in Northwestern California may result

from trade, but there is evidence that type lb
was made locally at Hum-169.

Toggle heads having flat points with

scarffed spurs which form a blade slot'57 do not
seem to have entered California, nor is there
any evidence of the leister harpoon.'58

The remaining archaeological specimens are

so generalized that diverse uses can be proposed
for them, and few specimens can be isolated as
toggle spurs or toggle points with any surety.

The type Tlf can best be singled out as possible
toggle points, and the occurrence of the type at
Redding site 7, in territory occupied histori-
cally by the Wintu, is especially significant.
It has been proposed that the specimens were

used in the Wintu-type toggle, although speci-

mens from Hum-118 and Hum-169 may have been used
in the three-piece type. Two other occurrences

have been noted farther south, at Col-l and CCo-
141, but both specimens were unassociated and

their cultural affiliations are unknown, though
Phase II is possible. The other type T examples
have equally valid claims to use as varied fish-
hooks and are not restricted to Phase II levels.
The type Ul specimens described herein are all

from Northwestern California. The artifacts are
so simple that many uses are possible; the other
pieces included by Gifford (1940, p. 177) are

not considered to be harpoon points because of
their location. The awl-like type, if found

archaeologically, probably could not be distin-
guished from basketry awls.

Thus the data obtained by excavation are

inadequate because of their rarity and the lack
of associations of the specimens. The four

types used ethnographically--three-piece spurred,
Awl-like, Wintu-type, and bipointed--with re-

gional variation and wide distribution, do not
support late diffusion in post-Contact times

which is suggested by the apparent absence of
the toggle head archaeologically in Central

California, particularly in view of the subsid-
iary position of the harpoon in comparison with
nets, weirs, and poisons. Unfortunately, ethno-
graphic reports are too frequently not specific

enough to allow any satisfactory distribution of
the four types to be given (see map 3)..

The three-piece spurred type is concen-

trated in Northwestern California and much of
the Central culture province. It is particu-
larly entrenched along the coast, with all

groups from the Tolowa through the Wappo making

15Ibid; Drucker, 1943, type II, p. 39, fig. 4, b, p. 40;
Smith, 1907, pp. 309, 310, fig. 160, P. 388.

158E.g., Ray, fig. 8, p. 60. Hewes, 1942a, p. 95, states
that the trident fish spear was not found south of the
olumbia, except among the Umatilla.

use of it.159 To the east, the distribution is
less uniform, the type being found among the

Shasta, Atsugewi, Valley and Foothill Maidu,160

Plains, Northern, and Central Miwok,161 and per-
haps the Patwin,162 while its possible presence
among the TUbatulabal'63 appears to be an iso-
lated and doubtful occurrence.

I

mM Three-piece spurred #o9e/'-
m   Awl-/ike, Wintu-#ype togg/e
m   Blpoin9ed togg/e

Map 3

Ethnographic Distribution of

Toggle Harpoon Types

159Northwest California: Driver, 1939, no. 231, p. 313;
Gif-ford, 1940, pp. 185, 186, fig. 1, p. 236, figs. 15, 16,
p. 237; Essene, no. 163, p. 6. Yuki: Gifford, 1940, fig.
8, p. 236; Foster, fig. 13, p. 171; Essene, no. 164, p. 6.
Coast Yuki: Gifford, 1929, fig. 7, pp. 335, 336.    Pomo:

Gifford and Kroeber, no. 209,p.211. Wappo: Driver, 1936,

p. 191; the barbed point mentioned was probably similar to
the Pomo three-piece type.

l6OVoegelin, 1942, no. 244, p. 56; Dixon, fig. 49, p. 196.
In an article dealing with the Valley Maidu, Miller, p. 206,
illustrates a pair of Wintu_type toggles, but Mountain
Maidu or Yahi implements may be represented.

l1lCurtis, 1924, p. 131, describes three-piece toggles,

with wooden spurs but obsidian tips, for the Miwok in gen-
eral; Barrett and Gifford, p. 189, refer to an obsidian-
tipped fish spear used in the higher mountains; the same

authors mention the two-pronged harpoon for the Plains and
Northern Miwok, while the Central Miwok are said to have
had a three-pronged harpoon (unique); the toggle for all
groups was the "common Californian type" (spurred?).

162Gifford and Kroeber, no. 208, p. 134. Since antler

and bone are denied as materials by the River Patwin, wood
must have been used. The "toggle head of antler (bone),
pitch and cord" of the Hill Patwin is also applicable to

the Wintu type, which quite possibly extended to the Patwin.

163Information is conflicting. It is assumed that the

barbed toggle head referred to by Voegelin, 1938, p. 14, is
three-piece. However, in Driver, 1937, nos. 147, 151, pp.

62, 112, a Kern River Tubatulabal informant denied any har-
poon, but the Bankalachi claimed the nonspurred type.

319

.11%

.N

1%.

ANTHROPOLOGICAL RECORDS

The awl-like type was used by the Klamath-
(Modoc?)164 and Yahi.165    The Wintu-type is found
among the Wintu,166 Yuki, and Yahi.167      Although
the distinction between the awl-like and Wintu-
type toggles can be made with museum specimens,
it has not been done in the published litera-

ture, and for the remaining groups it is neces-
sary to treat both as variants of a "non-

spurred" type.168 Non-spurred toggles are found

among the Karok, Chimariko (?), Hupa,169 and Hay-
fork Wintu'70 in the Northwest (the first three
groups had both non-spurred and three-piece

toggles); the Achomawi and Mountain Maidu in the
Northern Sierras;'7' and the Western Mono and

many of the adjoining Yokuts, Kawaiisu, and one
Owen's Valley Paiute group.'72

The Nisenan,'73 Southern Miwok,174 and Cen-

tral Foothill Yokutsl75 used bipointed specimens.

The Northern Paiute ani Washo176 possessed
the toggle harpoon, but it is not specified

whether Californian types were employed or the
distinctive form used by the Snake River Sho-
shone177 and probably the Humboldt River Sho-
shone .178

The zouthernmost limit of the toggle har-
poon in California is not clear. Drucker179

found no knowledge of the implement in Southern
California, although only informants from the
interior were available. Opinion was divided
among the Luisenio as to the aboriginal use of

the harpoon.'80 Coastal groups like the Chumash

164sarrett, p. 251, pl. 22, fig. 4. The Modoc denied the
harpoon, Voegelin, 1942, no. 24o, p. 56.

16Waterman, 1918, pi. 9.

s66DuBois, fig. 6, p. 128; Voegelin, 1942, no. 245, p. 56.
167Specimens in the UCMA.

168Driver, 1937, no. 151, p. 112, and 1939, no. 232, p.
379, specifically refers to the Klamath type (Barrett)

rather than the closer Wintu type (Du Bois) for those groups
included in footnotes 169 to 172 herein, but he may not have
intended to distinguish between the awl-like and Wintu

types; the Yokuts specimen sketched by Gayton, p. 72 (fig.
7, y herein) suggests the Wintu type.

169Driver, 1939, no. 232, p. 313.
170Ibid., no. 232, P. 379.

l7lVoegelin, 1942, no. 245, p. 56.
MDriver, 1937, no. 147, p. 63.

173Beals, fig. 1, p. 341 (fig. 7, q herein); Chever,
P. 137.

174Godfrey, p. 62, illustrated.Clark, p. 37: "Their

spears were small poles pointed with a single tine of bone,"
having a "string fastened near its center" to retrieve the
fish.

175Gayton, p. 75; Driver, 1937, no. 151, p. 112.
76Stewart, no. 272, p. 371.

77Steward, no. 272, p. 276, fig. g, h, p. 221.
1T8Ibid., pp. 227, 228.

79Drucker, 1937a, p. 44.
18OSparkman, p. 200.

and Dieguenfo probably had little use for this

fluvial implement, and no indication of its use
has appeared in the area; the single, barbed

toggle noted by Harrington'81 for the Chumash may
refer to the simple harpoon. The toggle harpoon
may have been known to most groups living near
suitable rivers, though in the south, where the
salmon were absent, it most likely was of much

less importance. The muddy Colorado and most of
the Mojave Desert apparently were barriers to

the further southward diffusion of the composite
toggle harpoon.

The distribution and antiquity of the

single and double foreshafted harpoon have been
discussed by Hewes and de Laguna.182 Little at-

tention has been given to the toggle head its,elf,
which exhibits marked variation, though most

composite specimens noted north of California

and the Klamath Indians were of the three-piece
spurred type. One exception is a pair of Wintu-
type toggles (fig. 7, _) which were presented to
the UCMA in 1877 by the Alaska Commercial Com-
pany. They are labeled "spear points used by

Alaska Indians in spearing salmon." These Alas-

kan tips are of bone, while all Californian specl
mens of this type have metal points.

The following description of the Tahltan

salmon harpoon indicates it resembles the Cali-
fornian bipointed type:

"A salmon spear of primitive type was
described [as] a blade.. .about three inches
in length, pointed at both ends, and sharp-
ened along one edge. A line of twisted

sinew or hide passed through a hole about

the center and was secured to a twelve foot

shaft near the head. One end of the toggle-
like blade fitted in a socket in the end of
the shaft.... 1183

The Haida probably used a related form.184

Little can be done with these occurrences
of non-spurred and bipointed toggles along the

Northwest Coast until their presence is corrobo-
rated by additional specimens. However, there
is the suggestion that these two types once had
a wider distribution, surviving only on the

northern and southern peripheries of the area
dominated by the composite spurred toggle har-
poon; perhaps they were stages in the develop-
ment of the spurred form.

The simplicity and wide discontinuous dis-

tribution of the non-spurred types in California
suggest an earlier diffusion than that of the

three-piece spurred type. It is improbable that
three successive forms of the toggle harpoon
could have passed through the Delta region in
Phase II and historic times. The broken and

181Harrington, nos. 46-51, P. 7.

l82Hewes, 1942a, pp. 96-102; de Laguna, pp. 194-198.
l83Emmons, p. 87.

184Niblack, pl. 29, fig. 137, p. 289; fig. 137e, p. 289,
from the Tlingit, is perhaps a variant form.

U

o

p
i

320

:D

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

often peripheral distribution of the bipointed
type suggests that it is the oldest form of

toggle harpoon in California. The toggle prin-

ciple is reduced to its barest essential in this
implement, and it is possible that the type rep-
resents parallel simplification in different

areas.  However, such a theory seems unlikely
for any group which did as much fishing as the

Nisenan or Southern Miwok. Instead a tentative
linkage is proposed between the Californian bi-
pointed toggle and the similar forms in use

among the Tahltan and Haida. This type may well
be related to the simplest single-piece, open-
socket toggle head found in the Bering Sea re-
gion,'85 but more information on these forms is
needed.

It may be impossible to determine when the
bipointed type of toggle harpoon entered Cali-
fornia. No suitable perforated artifacts have
been found archaeologically,and even grooved

specimens would resemble gorge hooks. So far no
.charred remnants of toggle harpoons have been

identified in pre-interment grave pit burnings,
in which we often find perishable materials

which have been preserved. The limited reten-
tion of the bipointed form suggests that it was
less efficient than the more complex toggles.

The type may therefore have spread through the
Valley during Phase I times, to be accepted by
fishermen using the simple harpoon; but it may
have lacked the superiority necessary to dis-
place the simple form. It is also possible
that the toggle is earlier than the latter,

though the fact that the bipointed toggle sur-

vived into the historic period, while the simple
harpoon did not, may indicate the later entrance
of the toggle form.

The awl-like and Wintu-type toggles are

basically the same; the peripheral distribution
and simple form (including the less efficient
foreshafts of equal length) both indicate an

earlier origin and spread than the three-piece
type. Presence of both the non-spurred and

spurred toggles among such groups as the Karok

and Hupa support the conclusion derived from the
distribution that the non-spurred types were in
very general use at one time in California. If

the southern Sierra harpoons do represent diffu-
sions from a northern center, it is improbable
that the Sacramento-San Joaquin River regions
were bypassed in the transit southward. The

double foreshafts and more complex toggle prob-
ably represent improvements over the bipointed

type which were sufficient to cause the abandon-
ment of the simple harpoon in Central California.

The distribution of the three-piece spurred
toggle suggests that the main route of diffusion
was from the Klamath River southward, west of

the Coast Range, with late extensions eastward
to the Atsugewi, Maidu, and Miwok. Excavation
in northern coastal sites indicates that the

spurred form was not in common use until rela-
tively late times even in extreme northwestern
California. One Pomo informant living on the
upper headwaters of the Russian River stated

that the toggle harpoon had been introduced to
his group, replacing the older fish spear.186
The simplicity and perishable nature of non-
spurred toggles, the frequent use of wooden

spurs (Maidu, River Patwin, and Miwok), and the
proposed later acceptance of the spurred toggle

head all contribute to the nonrecognition of the
toggle harpoon in Phase II or historic settle-
ments archaeologically.

SUMMARY

Although fish spears and harpoons are found
in all three culture areas of California, their

orms and temporal associations do not imply any
close historical connection. It is not suffi-

ient to list the harpoon as a culture element
ound throughout California; some distinction

lst be made between marine and fluvial uses and
etween various specific shapes, the only guide
vailable to us i'n the determination of wider

rehistoric relationships. Therefore, an arbi-

rary typology has been defined as an aid in the
iscussion of these implements. The present
per is confined primarily to a descriptive
nlysis, since existing information is too

4mited to support any extensive discussion of

e effect of the fish spear and harpoon on the
boriginal culture.

The sea-mammal harpoon is a characteristic
ement of the Northwest Coast culture area and
well represented on the shores of northern
lifornia, extending back in time to some an-

I

185Leroi-Gourhan, type A, pp. 356-361, fig. 688, p. 358;
irsen and Rainey, p1. 83, nos. 3, 4, 5, p. 165; other
ferences in both.

tiquity. Small harpoons for hunting small mam-
mals and for fishing were used throughout the
same period as the larger harpoons.

Harpoons are now the only element found in
sufficient quantity to demonstrate a sequence of
forms and hence can be used as time-bearers in

this area. Three basic forms of line attachment
serve as fundamental distinctions and occur in
both large and small harpoons from Hum-118 and
Hum-67. The two latest forms are found at Hum-
169. The earliest form recognized is the bi-
lateral line shoulder, which in time was suc-
ceeded by the bilateral line guard, the only

form which has a wide distribution at present,
occurring as far north as British Columbia and
south to Humboldt Bay. This type, in turn, was

gradually displaced by the unilateral line guard,
in use ethnographically among the Yurok and

Tolowa. Emphasis on two unilateral barbs, oc-
casionally with torque, appears to persist

through the entire period. Numerous variants

186Gifford and Kroeber, no. 210, p. 173; one Mono in-

formant claimed the harpoon was modern, Driver, 1937, no.
147, p. 112.

4

321

ANTHROPOLOGICAL RECORDS

and particular structural features will become
of increasing importance as the area known

archaeologically is expanded. At present the
slotted tip, of long duration at Hum-118, and

hooked barbs, appearing in the late occupation
of this site, are concentrated archaeologically
in the area occupied by the Yurok in historic

times and seem to expand a short way into Oregon
ethnographically. Neither feature seems to have
spread southward; no specimens are known from

within former Wiyot boundaries. It is suggested
that such local elaborations provide archaeo-

logical evidence of an intensive sea-hunting in-
dustry among the former inhabitants of Yurok

territory and that this may have been a contrib-
uting factor in the development of the subclimax
centered among the Yurok at the time of white
contact.

The extension of sea-mammal harpoons to the
Coast Yuki would appear to be a late intrusion
of a Northwestern pattern into Central Califor-
nia. Harpoon points were traded in from the

north, hunting was done only from the shore, and
no other use of the implement is recorded from
Central California.

The three-piece toggle point, the most com-
mon salmon harpoon of ethnographic California,

is kndwn archaeologically only along the north-
ern coast. Two basic types have been desig-

nated, one having a V-shaped channel, the other
a central groove interrupted by a medial ridge.
Possibly the first form to enter California had
the V-shaped channel and rectanguloid socket-

end, found from about Coos Bay, Oregon, to Hum-

boldt Bay. At Hum-117 this type was modified by
the addition of a forked tip. Elaboration con-
tinued, until the form used ethnographically by
the Yurok was produced shortly before the his-
toric period.

The medial-ridge type of toggle spur has
the widest distribution, from the Gulf of

Georgia to Humboldt Bay. It is felt that its

sporadic occurrence in California is more likely
due to a late diffusion from the north. Both
forms of toggle spur co-existed at Hum-169.

Central California produces a large variety
of fluvial implements used in fishing. A tri-

dent fish spear is limited to the Early horizon.
The occupants of certain later facies of the
Middle horizon may well have fished with un-

barbed, bident fish spears, type MM2b. They are
most characteristic of the Morse facies. They
seem to have been succeeded by a single-barbed

form, type .001, which quite possibly was a local
development from the unbarbed type. Though cen-
tered in the Brazil facies, type 001 is more

widespread, occurring at numerous sites from San
Francisco Bay to the Colusa province. Type 003,
with two barbs, was apparently a slightly later

invention. Type 002, exhibiting more careful
workmanship, is restricted to the south Delta.

The use of these barbed fish spears was

terminated by the entrance of the simple har-

poon, signifying the inception of Phase I of the
Late horizon. The earliest occurrence of the

harpoon is in the Delta, associated with Middle
culture traits. The prototype may have been a
rather generalized harpoon with no fixed place-
ment of its multiple bilateral barbs nor spe-
cialized line attachments other than rare bi-

lateral line shoulders. A multiplicity of forms
was soon produced, demonstrating considerable

local inventiveness. The greatest use of these
implements occurred in the Delta province, par-
ticularly in the Hollister facies. The south
Delta again developed a form showing more care
in manufacture and artistic elaboration. Site

CCo-138 also seems to have been a center of mod-
ification, having two unique types and a minor
emphasis on unilateral line shoulders, line

guards, and barbs. The different types provide
useful clues to close relationships between cer-
tain settlements. It has been suggested that

the simple harpoon entered the southern margin
of the Colusa province in Phase I from the

south, as well as the periphery of the Littoral
zone. No wider distribution can at present be
given, though the closest resemblances would
seem to be in the Gulf of Georgia-Puget Sound
area.

The absence of the simple harpoon in Phase
II of the Late horizon would seem to result from
the entrance of the non-spurred toggle harpoon
as a diffusion from the north down the Sacra-
mento River system. Types without spurs would
seem to have spread as far as the southern

Sierras.  Three-piece spurred toggle heads ap-
parently were late in entering the Interior

province, and have not been recognized archaeo-
logically.

The data from Southern California are most
unsatisfactory. A variety of forms is known, as
would be expected from an area in which the sea
provided the main livelihood. However, no tem-
poral significance can be stated, and a possible
diagnostic difference between Chumash and

Gabrieleino harpoons cannot as yet be substan-
tiated.

Both composite and simple harpoons are

known archaeologically; the composite specimens
were found with late associations, while no

depth can be given for the one-piece harpoons.
They are perhaps related to the Seri and Lower
California specimens, and might well be the

oldest form in California. A study of artifacts
which may be barbs of composite harpoons might
provide needed answers to the problem of antiq-
uity in Southern California.

i
f

322

CONCORDANCE OF SITE NUMBERS

The site designations employed in this paper are those recently assigned
by the University of California Archaeological Survey (CAS) and use county

symbols. For convenience they are here correlated with the site numbers ap-
pearing in previous reports on California archaeology.

A: new number equated with old site number.

B: old number equated with new site designation.

Sites whose names appear in the text but
been assigned county symbols or have not
cations.

are not listed below have not yet
been discussed in previous publi-

A

New

Ala-309
Ala-329

CCo-138
CCo-139
CCo-141
CCo-250
CCo-295

Old

N.309
N.329

C.138
Simone
C.141

N. 250a
N.295

Co1-1     S.1
Col-2     C.2

Hum-67
Ker-46

*site 67
site 11

Mrn-275   site 275

Yol-52     S.52

B

Old

C.6

C.8

C.15
C.56
C.59
C.66

N.236

N. 250a
N.295
N.309
N.329
N.356
N.372

New
Sac-6
Sac-8

Sac-49
SJo-56
SJo-59
Sac -66

Old

C.68
C .80
C .82
C .87
C .91

C.107

Sol-236
CCo-250
CCo-295
Ala-309
Ala-329
SCl-356
SMa-372

S.1
S.2

S.16
S.29
S.43
S.52
s.6o

New

SJo-68
SJo-80
SJo-82
SJo-87
SJo-91

Sac-107

Old

C.122
C.127
C.138
C.139
C.14o
C.141

S.64
S.66

Col-1
Col-2

Sac-16
Sac-29
Sac-43
Yol-52
Sac-60

Simone
site 11
site 67

site 275

New

Sac-122
Sac-127
CCo-138
SJo-139
SJo-140
CCo-141
Sac-64
Sac-21

CCo-139
Ker-46
Hum-67

Mrn-275

[323]

New

Sac-6
Sac-8

Sac-16
Sac-21
Sac-29
Sac-43
Sac-49
Sac-56
Sac-60
Sac-64
Sac-66

Sac-107
Sac-117
Sac-122
Sac-127

Old

c.6
c.8

S.16
S.66
S.29
S. 43
C.15
S.56
S.60
S .64
C .66
C . 107
C.117
C.122
C.127

New

SC1-356
SJo-56
SJo-59
SJo-68
SJo-80
SJo-82
SJo-87
SJo-91

SJo-139
SJo-14o

Old

N.356
C.56
C .59
C .68
C .80
C .82
C.87
C .91

C.139
C.140

N.372
N.236

SMa-372
Sol-236

I

BIBLIOGRAPHY

ABBREVIATIONS

AA

;- AAA-M

I A Ant

AJPA

AMNH

-A
-B
l.    -M
i BAE

-B
-AR
. MAI-HF

American Anthropologist.

Memoirs, American Anthropological

Association.
American Antiquity.

American Journal of Physical Anthro-

pology, new series.

American Museum of Natural History.
LP        Anthropological Papers.

Bulletin.

[J        Memoirs, Jesup North Pacific Ex-

pedition.

Bureau of American Ethnology.

Bulletin.

Annual Report.

Museum of the American Indian, Heye

Foundation.
-C       Contributions.

MAI-INM        Indian Notes and Monographs.
SI          Smithsonian Institution.

-C           Contributions to Knowledge.
-MC          Miscellaneous Collections.
-R           Report.

UC          University of California.

-AR          Anthropological Records (CED,

Culture Element Distribu-
tions).

-PAAE        Publications in American Archae-

ology and Ethnology.

UCMA        Museum of Anthropology, University

of California, Berkeley.
USNM        U. S. National Museum.

-R         Report.

UW-PA       University of Washington, Publi-

cations in Anthropology.

Abbott, C. C., and F. W. Putnam

1879. Implements Made of Bone. Report upon

U. S. Geographical Surveys West of the
100th Meridian, Vol. 7. Washington.
Aginsky, B. W.

1943. CED:XXIV--Central Sierra. UC-AR

8:393-468.

Amsden, C

1930.
Barnett,

1939.

Homo Californianus. Masterkey, 3:21-
29. Los Angeles.
H. G.

CED:IX--Gulf of Georgia Salish. UC-AR
1:221-296.

Barrett, S. A.

1910.  The Material Culture of the Klamath

Lake and Modoc Indians of Northeastern
California and Southern Oregon. UC-
PAAE 5:239-292.
and E. W. Gifford

1933. Miwok Material Culture.    Public Museum

of the City of Milwaukee, Bull. 2,
pp. 117-376.
Bartlett, J. R.

1854.  Personal Narrative of Explorations and

Incidents in Texas, New Mexico, Cali-
fornia, Sonora and Chihuahua. Vol. 2,
New York.
Beals., R. L.

1933. Ethnology of the Nisenan.    UC-PAAE

31:335-414.
Beardsley, R. K.

1948. Culture Sequences in Central Califor-

nia Archaeology. A Ant 14:1-28.

n.d.   Temporal and Areal Relationships in

Central California Archaeology. Ph.D.
thesis (1947), University of Califor-
nia, Berkeley. MS.

[3

Berreman, J. V.

1944. Chetco Archaeology. General Studies

in Anthropology, 11. Menasha, Wis.
Bird, J. B.

1946. The Culture Sequence of the North

Chilean Coast.   BAE-B 143:587-594.
Bowers, S.

1883. Fishhooks from Southern California.

Science, 1:20. Cambridge, Mass.
Chard, C. A.

n.d.   Hotchkiss Report.   MS.
Chever, E. E.

1870. The Indians of California. American

Naturalist 4:129-148. Salem, Mass.
Clark, G.

1910. Indians of the Yosemite Valley and

Vicinity. Yosemite Valley, Calif.
Collier, D., A. E. Hudson, and A. Ford

1942. Archaeology of the Upper Columbia

Region. UW-PA 9:1.
Collins, H. B., Jr.

1937. Archaeology of St. Lawrence Island,

Alaska. SI-MC 96:1.
Cook, S. F., and R. F. Heizer

1947.  The Quantitative Investigation of

Aboriginal Sites: Analyses of Human
Bone. AJPA 5:201-220.
Coolidge, D.,and N. R.

1939.  Last of the Seris.   New York.
Costanso, M.

1910.  The Narrative of the Portola Expedi-

tion of 1769-1770. Academy of Pacific
Coast History, Publ. 1:4. Berkeley.
Curtis, E. S.

1914. Indian Days of Long Ago. New York.

1924. The North American Indian. Vol. 14.

Norwood, Mass.
325]

I

I
r

I

?i
I

ANTHROPOLOGICAL RECORDS

Dalton, 0. M.

1897.  Notes on an Ethnographical Collection

from the West Coast of North America
(More Especially California, Hawaii

and Tahiti), Formed During the Voyage
of Captain Vancouver 1790-1795. In-
ternationales Archiv fur Ethnologie,
10:224-245.
Dixon, R. B.

1905. The Northern Maidu. AMNH-B 17:119-346.
Driver, H. E.

1936.  Wappo Ethnography.   UC-PAAE 36:179-220.
1937.  CED:VI--Southern Sierra Nevada.   UC-AR

1:53-154.

1939   CED:X--Northwestern California.   UC-AR

1:297-434.
Drucker, P.

1937a. CED:V--Southern California. UC-AR

1:1-52.

1937b. The Tolowa and Their Southwest Oregon

Kin. UC-PAAE 36:221-300.

1943.  Archaeological Survey of the Northern

Northwest Coast. BAE-B 133:17-132.
Du Bois, C.

1935. Wintu Ethnography. UC-PAAE 36:1-148.
Edge-Partington, J., and C. Heape

1898. Ethnological Album of the Pacific

Islands, Vol. 3. London.
Emmons, G. T.

1911. The Tahltan Indians. University of

Pennsylvania Museum, Anthropological
Publ., 4:1. Philadelphia.
Essene, F.

1940. CED:XXI--Round Valley. UC-AR 8:1-98.
Fages, P.

1937.  A Historical, Political and Natural

Description of California. Translated
by H. I. Priestley. Berkeley.
Fenenga, F.

n.d.   The Archaeology of Central California:

The Middle Horizon. MS.
Foster, G. M.

1944. A Summary of Yuki Culture. UC-AR

5:155-244.
Gayton, A. H.

1948. Yokuts and Western Mono Ethnography.

UC-AR 10:1-301.
Gifford, E. W.

1939. The Coast Yuki.    Anthropos, 34:292-375.
1940. Californian Bone Artifacts.    UC-AR

3:153-238.

and A. L. Kroeber

1937. CED:IV--Pomo.    UC-PAAE 37:117-254.
Gillin, J.

1940. A Barbed Bone Projectile Point From

Utah. A Ant 6:170-171.
Godfrey, E. H.

1941. Yosemite Indians.    Yosemite Nature

Notes, 20:49-72. Stockton, Calif.
Hardy, R. W. H.

1829.  Travels in the Interior of Mexico,

1825, 26, 27 and 28. London.
Harrington, J. P.

1942. CED:XIX--Central California Coast.

UC-AR 7:1-46.

Heizer, R. F.

1941. The Direct Historical Approach in

California Archaeology. A Ant 7:98-122.
1944. Artifact Transport by Migratory Ani-

mals and Other Means.   A Ant 9:395-400.
1947. Historic North Pacific Culture Influ-

ences in the Santa Barbara Region.

Masterkey, 21:150-152.   Los Angeles.

1949.  The Archaeology of Central California.'

I: The Early Horizon.   UC-AR 12:1-84.
and F. Fenenga

1939.  Archaeological Horizons in Central

California. AA 41:3.
Hewes, G. W.

1942a. The Ainu Double Foreshaft Toggle Har-

poon and Western North America.   Wash-
ington Academy of Science, Journal,
32:4.

1942b. Economic and Geographical Relation-

ships of Aboriginal Fishing in North-
ern California. California Fish and
Game, 28:103-110.
Heye, G. C.

1921.  Certain Artifacts from San Miguel

Island. MAI-INM 7:17-211.
Hill-Tout, C.

1895.  Later Prehistoric Man in British

Columbia. Royal Society of Canada,

Proceedings and Transactions, ser. 2,
1:103-122.
Hodge, F. W., ed.

1922.  Guide to the Museum, 2nd Floor.

MAI-INM 31.
Irwin, M. C.

1946.  Museum Leaflet, 21:18-21.   Santa Bar-

bara Museum, Santa Barbara.
Jochelson, W.

1925.  Archaeological Investigations in the

Aleutian Islands. Carnegie Institute,
Publ. 432. Washington.
Johnson, F. R.

1944. Fishhooks. Illinois State Archaeo-

logical Society Journal, 2:49-51.
Fairbury, Ill.

Kroeber,

1913.

1925.
1932.

A. L.

The Seri Indians. Southwest Museum
Papers, 6:1-60. Los Angeles.

Handbook of the Indians of California.
BAE-B 78.

The Patwin and their Neighbors. UC-
PAAE 29:253-424.

Laguna, F. de

1947. The Prehistory of Northern North

America as Seen from the Yukon.
A Ant, Memoir 3.
Larsen, H., and F. Rainey

1948.  Ipiutak and the Arctic Whale Hunting

Culture. AMNH-AP 42.
Laughlin, W. S.

1943.  Notes on the Archaeology of the Yam-

hill River, Willamette Valley, Oregon.
A Ant 9:220-229.

Leatherman, K. E., and A. D. Krieger

1940. Contributions to Oregon Coast Pre-

history. A Ant 6:19-28.

I

i
I

7

4

326

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

Leroi-Gourhan, A.

1946. Archeologie du Pacifique-Nord. Travaux

et Memoires de l'Institut d'Ethnologie,
47:325-412. Paris.

Lillard, J. B., R. F. Heizer, and F. Fenenga

1939. An Introduction to the Archeology of

Central California. Sacramento Junior
College, Publ. in Anthro., Bull. 2.
Sacramento.

Lillard, J. B., and W. K. Purves

1936. The Archeology of the Deer Creek-

Cosumnes Area, California. Sacramento
Junior College, Publ. in Anthro.,
Bul. 1. Sacramento.
Loeb, E. M.

1926. Pomo Folkways. UC-PAAE 19:149-406.
Loeffelholz, H. F. and K. F. von

1893. Die Zoreisch-Indianer der Trinidad-

Bai. Mitteilungen der Anthropo-

logischen Gesellschaft in Wien, 23:
101-123.
Lothrop, S. K.

1928.  The Indians of Tierra del Fuego.

MAI-C 10.
toud, L. L.

1918.  Ethnography and Archaeology of the

Wiyot Territory.   UC-PAAE 14:221-426.
Lowie, R. H.

1946.  The Bororo.   BAE-B 143:419-434.
McGee, W. J.

1898.  The Seri Indians.   BAE-AR for 1895-

1896, 17:1.
McKern, W. C.

1922.  Functional Families of the Patwin.

UC-PAAE 13:235-258.
Martinez, J. L.

1938.  California in 1792; the Expedition of

Jose Longinos Martinez. Translated by
L. B. Simpson. San Marino, Calif.
Mason, 0. T.

1889.  The Ray Collection from the Hupa

Reservation. SI-R, 1886, App.

1902.  Aboriginal American Harpoons.   USNM-R

for 1899-1900.
Mathiassen, T.

1927. Archaeology of the Central Eskimos.

Report of the Fifth Thule Expedition,
IV:2.
Miller, M. L.

1896. The So-called "Diggers." Appleton's

Popular Science Monthly, 50:201-214.
Moorehead, W. K.

1900. Prehistoric Implements. Cincinnati.
Nelson, E. W.

1899. The Eskimo about Bering Strait. BAE-

AR 18:1.
Nelson, N. C.

1936. Notes on the Santa Barbara Culture.

Essays in Anthropology Presented to

A. L. Kroeber, pp. 199-209. Berkeley.
Niblack, A. P.

1890. The Coast Indians of Southern Alaska

and Northern British Columbia.
USNM-R 1888, pp. 225-386.

Iomland, G. A.

1935. Sinkyone Notes. UC-PAAE 36:149-178.

1938. Bear River Ethnography. UC-AR 2:91-

126.
Olson, R. L.

1930. Chumash Prehistory. UC-PAAE 28:1-22.
Orr, P. C.

1948. Additional Californian Bone Artifacts.

Appendix, in E. W. Gifford, Califor-

nian Shell Artifacts, UC-AR 9:115-132.
Powers, S.

1877. Tribes of California. Contributions

to North American Ethnology, III.
Rau, C.

1884. Prehistoric Fishing in Europe and

North America. SI-C 25.
Ray, V. F.

1932. The Sanpoil and Nespelem. UW-PA 5:1-

237.
Read, C. H.

1892.  An Account of a Collection of Ethno-

graphic Specimens Formed during Van-
couver's Voyage in the Pacific Ocean.
Journal of the Royal Anthropological
Institute, Vol. 21.  London.
Robinson, E.

1933.  Fishing Arrows from Southern Califor-

nia. Masterkey, 7:149-150. Los
Angeles.
Rogers, D. B.

1929. Prehistoric Man of the Santa Barbara

Coast. Santa Barbara.
Rogers, M. R.

1929. The Stone Art of the San Dieguito

Plateau. AA 31:454-467.
Rostlund, E.

n.d.   A Distributional Study of Primitive

Fishing in California. MS. M.A.

Thesis (1946), University of Califor-
nia, Berkeley.
Sargent, L.

1880. Fishing on the Winniemae. The Cali-

fornian, 2:439-443. San Francisco.
Schenck, W. E.

1926.  The Emeryville Shellmound:   Final Re-

port. UC-PAAE 23:147-282.
and E. J. Dawson

1929.  Archaeology of the Northern San

Joaquin Valley. UC-PAAE 25:289-414.
Shelvocke, G.

1928.  A Voyage Around the World.   Seafarer's

Library Edition. San Francisco.
Smith, H. I.

1899.  Archaeology of Lytton, British Colum-

bia. AMNH-MJ I:III.

1900.  Archaeology of the Thompson River Re-

gion, British Columbia. AMNH-MJ I:VI.
1903.  Shell-heaps of the Lower Fraser River,

British Columbia. AMNH-MJ II:IV.

1907.  Archaeology of the Gulf of Georgia and

Puget Sound. AMNH-MJ II:VI.
Sparkman, P. S.

1908.  The Culture of the Luise"no Indians of

Southern California. UC-PAAE 8:187-
234.
Spinden, H. J.

1908. The Nez Perce.

AAA-M 2:3.

I

k
P,
I

7

327

ANTHROPOLOGICAL RECORDS

Steward, J. H.

1941. CED:XIII--Nevada Shoshoni. UC-AR

4:209-36o.
Stewart, 0. C.

1941. CED:XIV--Northern Paiute. UC-AR

4:361-446.

Strong, W. D., W. E. Schenck, and J. H. Steward

1930. Archaeology of the Dalles-Deschutes

Region. UC-PAAE 29:1-154.
Swanton, J. R.

1911. Indian Tribes of the Lower Missis-

sippi Valley and Adjacent Coast of the
Gulf of Mexico. BAE-B 43.
Teit, J.

1900. The Thompson Indians of British

Columbia. AMNH-MJ I:IV.

1906. The Lillooet Indians. AMNH-MJ II:III.
Tyzzer, E. E.

1936. The "Simple Bone Point" of the Shell

Heaps of the Northeastern Algonkian

Area and Its Probable Significance.
A Ant 1:261-279.
Uhle, M.

1907. The Emeryville Shellmound. UC-PAAE

7:1-106.
Voegelin, E. W.

1938. Tubatulabal Ethnography. UC-AR 2:1-90.
1942. CED:XX--Northeastern California.

UC-AR 7:4-252.
Wagner, H. R.

1929. Spanish Voyages to the Northwest Coast.

Waterman,

1918.
1920.

and
1920.

California Historical Society. San
Francisco.
T. T.

The Yana Indians. UC-PAAE 13:35-102.,
Yurok Geography. UC-PAAE 16:177-314.
G. Coffin

Types of Canoes on Puget Sound.

MAI-INM 6.
Woodward, A.

1941. Notes and News. A Ant 6:284-285.

328

EXPLANATION OF FIGURES

Figures 1-7

Fig. 1. Early and Middle horizon fish
spears. a. Type MM3a, site SJo-68, 1-49054.

b. Type MM3a, site SJo-56, L-19229. c. Recon-

structed form of Early horizon fish spear, site
SJo-68, 1-73521. d. Type MM2a, site Ala-309,
1-8915. e. Type MM2a, site Ala-309, 1-26175.
f. Type MM2d, site Yol-52, 1-13551. g. Type

MM2d, site Sac-43, L-19385. h. Type MM2d, site
Sac-16, L-11457. i. Type MM2e, site CCo-250,

Johnson collection M-19. J. Fishhook assemblage
found in situ, site SJo-68, 1-49056, 1-49055.

k. Type MM2b   (reworked type 001), site Sac-43,
L-19898. 1. Type MM2b, site Sac-66, L-12812.

m. Type MM2b, site Sac-66, 8-34. n. Type MM2b,
site Sac-66, L-18297. o. Type MM2b, site Sac-

66, L-18301. p. Type MM2b, site Sac-66, L-18306.
q. Type MM2b, Sac-122, L-12790. r. Type MM2b,
CCo-139, 1-53962.   s. Type MM2b, site Ala-309,

1-8738. t. Reconstructed form of Middle horizon
fish spear, site Sac-66, L-13169, 1-56344.

u. Alternative method of binding fish spear

points to shaft. v. Reconstructed form of type
001 Middle horizon fish spear, site Sac-49,

1-56246 (2 specimens). w. Type 001, site Col-l,
1-38046. x. Type 001, site Sol-2, 1-72969.

y. Type 001, site Sac-49, 1-56246. z. Type 001,
site Col-1, 1-38107. a'. Type 001, site Yol-52,
L-13590. b'. Type 001, site Sac-43, L-19995.

c'. Type 001, site SCl-356, 1-16910. d'. Type
001, site SCl-356, 1-16909. e'. Type 001, site
Sac-8, 1-56195.

Fig. 2. Fish spears and harpoons from Cen-
tral and Southern California. Fish spears, a-t.
a. Reconstructed form of type 002 Middle horizon
fish spear, site SJo-87, Barr 2264. b. Type

002, site SJo-140, Barr 1092. c. Type 002, site
SJo-140, Barr 1077. d. Type 002, site SJo-140,

Barr 1099. e. Type 002, site SJo-87, Baar 2249.
f. Type 003, site Sac-43, L-19913. g. Type 003,
site Sac-43, L-19946. h. Type 003, site Sac-49,
1-56247. i. Type 003, site Yol-52, L-13467.
J. Type 003, site Sac-127, L-18740. k. Type

Wla, Channel Island area. Same size as average
Channel Island circular shell fishhook, illus-
ttated in Amsden, p. 26. 1. Type Wla, San

Nicolas Island, AMNH, Terry collection, T14611.
m. Type Wla, San Nicolas Island, 1-15953.
n. Type Wla, San Nicolas Island, 1-15950.
o. Type Wla, San Nicolas Island, 1-15951.

p. Type Wla, San Nicolas Island, AMNH, Terry
collection, T14606. q. Type Wla, San Nicolas
Island, 1-15952. r. Type Wla, site Hum-67,

Stuart collection, Eureka. s: Type Wla, site
Col-2, 1-29485.  t. Type Wla, site CCo-295,

1-11259. Harpoons, u-z, a'-i'. u. Type NN2aI,

site Sac-21, L-19091. v. Type NN2aI, site Sac-
21, L-19140. w. Type NN2aI, site Sac-21,

L-19130. x. Type NN2aI, site Sac-21, L-19137.
y. Type NN2aI, site CCo-138, 1-53776. z. Type
NN2aI, site CCo-138, 1-53776. a'. Type NN2aII,
Nicolaus site, L-13783. b'. Type NN2aII, site

Sac-21, L-19133. c'. Type NN2aII, site Sac-107,
L-16386. d'. Type NN2aII, site CCo-138, Johnson
collection, 12-1.   e'. Type NN2aIII, site CCo-

141, Barr 3364a. f'. Type NN2aIII, site SJo-91,
Barr 2104. g'. Side-view of f', showing ridged
barbs. h'. Type NN2aIII, site SJo-91, Barr

2124. i'. Side view of h', showing ridged barbs.

Fig. 3. Late horizon harpoons. a. Type
NN2bI, site Sol-236, 1-22307. b. Type NN2bI,
site CCo-138, 1-49276. c. Type NN2bI, site

CCo-138, 1-53776. d. Type NN2bI, site CCo-138,
Johnson collection, 1-21. e. Type NN2bI, site
CCo-138, 1-49193. f. Type NN2bI, site CCo-138,
Johnson collection, 3-17. g. Type NN2bI, site
Sac-6, 1-82419. h. Type NN2bI, site Sac-21,

L-11495. i. Type NN2bI, site Sac-21, L-11559.

J. Type NN2bI, site CCo-139, Johnson collection,
A-2. k. Type NN2c, Core site, Martine collec-
tion, Sacramento. 1. Type NN2c, Bethany site,
MAI-HF, Dreisher collection, 4/2772. m. Type

NN2c, Bethany site, MAI-HF, Dreisher collection,
4/2772. n. Type NN2c, site Sac-107, L-16385.
o. Type NN2c, site Sac-107, L-16381. p. Type

NN2c, site Sac-21, L-19090. q. Type NN3a, site
CCo-138, Johnson collection, 1-49469. r. Type

NN3b, site CCo-138, L-53747. s. Type NN3b, site
CCo-138, Johnson collection 1-19.   t. Type NN3b,
site CCo-138, 1-53747.

Fig. 4. Harpoons from Central and Northern
California. Central California harpoons, a-h.
a. Type NN2bII, as found in situ, site Sac-16,
California State Indian Museum, Sacramento.

b. Type NN2bII, Nicolaus site, L-13782. c. Type
NN2bII, site Sac-21, L-11496. d. Type NN2bII,

site CCo-138, Johnson collection, 23-2. e. Type
NN2bII, site Sol-2, 1-80411. f. Type NN2bII,
site CCo-138, 1-49972. _. Type NN2bII, site

CCo-138, L-18442.   h. Type NN2d, site CCo-138,
1-53809A. Northern California harpoons, i-x.

i. Cut-off tip of harpoon point, deep levels of
site Hum-118, Patrick's Point State Park (PPSP)
no. 1480. J. Type IAla, site Hum-67, Stuart

collection, Eureka. k. Type IAla, site Hum-67,
Stuart collection, Eureka. 1. Type lAla, site
Hum-67, 1-18428. m. Type Ia2a, site Hum-118,

PPSP 1465. n. Example of concave base point for
slotted tip of harpoons, site Hum-118, PPSP 1219
(harpoon), PPSP 197. o. Concave base point,

[329]

I

ANTHROPOLOGICAL RECORDS

site Hum-169, 1-115787. p. Type IA2a, site Hum-
118, PPSP 409. q. Type IA2a, side and front

views showing torque, site Hum-118, PPSP 1292.

r. Type IA2a, evidencing grooving preparatory to
making a new tool from broken harpoon, site

Hum-118, PPSP 2057. s. Type IIAla, site Hum-67,
Stuart collection, Eureka. t. Type IIAla, site
Hum-67, Stuart collection, Eureka. u. Type

IIAla, site Hum-118, PPSP 1453. v. Type IIAla,
site Hum-118, PPSP 755. w. Type IIAla, site

Hum-118, PPSP 862.   x. Type IIAla, reworked into
awl, site Hum-118, PPSP 1510.

Fig. 5.  Northwestern California harpoons.
Simple harpoons, a-p. a. Type IBla, site Hum-
118, PPSP 1530. b. Type IB, reworked into

punch, site Hum-118, PPSP 44. c. Type IB2a,

site Hum-118, PPSP 402.   d. Type IB2a, site Hum-
118, PPSP 1785. e. Type IB, with reworked tip,

site Hum-118, PPSP 576.   f. Type IB2a, site Hum-
118, PPSP 1066.   g. Type IB2b, site Hum-169,

field 2071. h. Type ICla, site Hum-67, Stuart
collection, Eureka. i. Type ICla, Mad River

Beach site, Stuart collection, Eureka. J. Type
IIBla, site Hum-118, PPSP 270. k. Type IIBla,
site Hum-118, PPSP 2238. 1. Type IIBla, site
Hum-118, PPSP 1417. m. Type IIBla, site Hum-
118, PPSP 1275. n. Type IIBla, site Hum-67,

Stuart collection, Eureka. o. Type IIBla, site
Hum-118, PPSP 283. p. Type IIBla, site Hum-118,
PPSP 1805. Possible toggle harpoon tips, q-v.
j. Possible stone tip, site Hum-169, field 22.

r. Type Ul, site Hum-169, field 2. s. Type Ul,
site Hum-169, field 822. t. Type Tlf, Redding
site 7, 1-37230. u. Type Tlf, Redding site 7,
1-37366. v. Type Tlf, site Hum-118, PPSP 427.

w. Barbed arrow point fragment from Northwestern
California, site Hum-118, PPSP 544.    Toggle har-
poon spurs, x-z, a'-g'. x. Type Ia, site Hum-
67, 1-18429.  y. Type Ia, site Hum-67, 1-8430.

z. Type Ia, site Hum-169, field 4-121.    a'. Type
Ib, site Hum-169, C-5. b'. Type IIIa, site Hum-
67, Stuart collection, Eureka. c'. Type IIIb,
site Hum-169, field 1903. d'. Type IIIb, site
Hum-118, PPSP 1688. e'. Type IIIb, site Hum-
169, field 4-16. f'. Type IIIb, site Hum-118,
PPSP 1777. g'. Type IIIb, site Hum-169, field
2353.

Fig. 6. Harpoons from Northern and Southern
California. Northern California harpoons, a-J.
a. Type IC2b, site Hum-169, Stuart collection,

Eureka. b. Type IC2b, site Hum-169, Stuart col-
lection, Eureka. c. Type IC2b, site Hum-169,

field 4-15. d. Type IC, site Hum-118, PPSP 1995.
e. Type IC2b, site Hum-118, PPSP 751. f. Type
IC, with reworked tip, site Hum-118, PPSP 1678.
g. Type IICla, reworked after loss of first

barb, site Hum-118, PPSP 2174. h. Type IICla,
site Hum-169, field 4-14.   i. Type IICla, site
Hum-67, Stuart collection, Eureka. J. Type ID,
site Hum-67, 1-18423. k. Type I fragment from
Central California, site Mrn-275, 1-54693.

Southern California harpoons, l-r. 1. Type Wlb,
San Nicolas Island, AMNH, Terry collection,
T14462. m. Type Wlb, San Nicolas Island,

1-15954.  n. Type Wlb, Santa Cruz Island, illus-

trated in Orr, p. 117, type W4. Actual size is

9 in. o. Type Wlb, "Santa Barbara," illustrated
in Abbott and Putnam, fig. 103, p. 224. p. Type
Wlb, San Nicolas Island, 1-15955.   q. Type W2,
San Nicolas Island, illustrated in Moorehead,

fig. 358, no. 4. Size not stated. r. Type W2,
San Nicolas Island, illustrated in Moorehead,
fig. 362, no. 2.   Probably about 18 cm. long.

Fig. 7. Various fishing implements. Com-

posite harpoon points and barbs, a-J.   a. Compo-
site harpoon foreshaft from Southern California,
San Clemente Island, Los Angeles County Museum.
b. Restoration of point of a by Woodward. c.
Type of barb associated by Woodward with sea-

mammal harpoon.   d. Composite harpoon arrow from
Southern California, San Clemente Island, Los

Angeles County Museum. e. Restoration of point
of d by Woodward. f. "Dart foreshaft" from San
Nicolas Island, AMNH, Terry collection, illus-
trated by Nelson, fig. 3, d, p. 205, size not

stated. g. Type T2bII composite harpoon arrow
barb from Southern California, site 30, Santa
Rosa Island, 1-6537. h_ Type T2bII composite
harpoon arrow barb from Southern California,

site 147, Santa Cruz Island, 1-31268. i. Type

T2bII composite harpoon arrow barb from Southern
California, site 135, Santa Cruz Island, 1-37159.
J. Type T2bII composite harpoon arrow barb from

Southern California, site 147, Santa Cruz Island,
1-31237. k. Type NN2aIV harpoon from Central

California, site Sac-16, L-11580. 1. Hawaiian
netting needle (see p. 309), 9-1/2 in. long, il-
lustrated in Edge-Partington and Heape, pl. 10,
no. 11. m. Fishhook collected by Vancouver Ex-
pedition somewhere in the Pacific, illustrated
by Dalton, pl. XVI, fig. 22. Length 21 cm.

n. Barbed fishhook from Humboldt Cave,Nevada,

1-43040. o. Barbed fishhook barb from Humboldt
Cave, Nevada, 1-43041. p. Harpoon point of the
Coast Yuki, illustrated in Gifford, 1939, fig.
6, p. 335. q. Nisenan salmon harpoon, illus-
trated in Beals, fig. 1, p. 341.   Size not

stated. r. Yurok toggle spur (same as archaeo-
logical type IIIb), 1-1654.   s. Karok toggle
spur, 1-1825 A. t. Yurok toggle point, with
metal tip, 1-1946. u. Southeast Alaskan

(Tlingit?) toggle point, 2-2757. v. Wintu

toggle point, illustrated in Du Bois, fig. 6,

p. 128.  w. Klamath toggle point, awl-like type,
1-12768a. x. Wailaki toggle harpoon, illus-

trated in Hewes, 1932b, fig. 34, a.   Size not

stated. 1. Yokuts toggle harpoon, illustrated

in Gayton, fig. 10, ?, p. 72. Total length, 15-
20 feet. z. Santa Barbara (Chumash) harpoon,

collected by Vancouver Expedition, illustrated
in Read, pl. XI, fig. 2, Over 37 in. long. a'.

Possible harpoon shaft, probably from Santa Bar-
bara, collected by Vancouver Expedition, illus-
trated in Read, p1. XV, fig. 8, 70 cm. long.
b'. Tolowa harpoon, illustrated in Drucker,

1937b, fig. 2, p. 237. Size.not stated. c'.
Yurok harpoon from Requa, California, illus-
trated in Hewes, 1942b, fig. 34, b. Size not
stated. d2. Tolowa harpoon, type IC3b, with
metal point, 1-2444.

I

m

ii
I

i

330

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

d     e

C

7

m    n   0

I.
I I

/ .

p    q   r  S

Oin.    !     2

O cm. i  2  3  4   5 '

A    A         u

II
II

: 1

w      x    y

z     a'

Fig. 1. Early and Middle horizon fish spears

b

k

j

Ci

v

i

i

[

[
S

b'        c'     d'

e

331

............
..............

I

ANTHROPOLOGICAL RECORDS

m        S-

O.n.  I  2

Ocm.-   3  4

'I

v   X 7C  b' c'   ' f  ' h

Fig. 2. Fish spears and harpoons from Central and Southern California

332

u     v

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS                   333

b

~~~~~~~~~~~~~~~~~~~~C='

LI--g--

C~~ ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ --\

Oin.               2
~~~~~~~~~~~~~~~~Ocm.                                            2i 5
}    z < ~~~~~~~~.0

6<  *\ \_JosV                    r.j                 g      sJsf sn

Fig. 3. Late horizon harpoons

334                                          ANTHROPOLOGICAL RECORDS

: AM

a                             \

EL

n                                                .

Oin.     !        2

Ocm. I  2  34     5-

I                         %

*                  ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~II

(_  _ _  .   :          .    ~  r      S

L      'LS                        g ~~~~~~~~~~~~~r

'4

/                ~~~~~~r

w.-

Fig. 4.   Harpoons from Central and Northern California            A

I

f                                                                                          .     .
, ?- I",                                                                                                          -     :

h

I

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS                           335
U                         --      ~~~~~~a

Oin              2    1

Ocm.i   2   3   4    2

m
p~~~~~

.t
I a                          c S

*C=                                                                                   St

Fig. 5. Northwestern California harpoons

I

I

ANTHROPOLOGICAL RECORDS

I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~P

C-                 ~~~~~~n

Fig. 6- Harpoons from Northern and Southern California

Fig. 6. Harpoons from Northern and Southern California

I1

i
i.

I
r

336

cZZJz

BENNYHOFF: CALIFORNIAN FISH SPEARS AND HARPOONS

-                                 .,--_--~~wm~

D>

-<iff~~~~~~~~~~~~~~~ thill -Avi

/11   n   (\  im.i  2 !3 4  5  A  Wd

V         W    X

Fig. 7. Various ffishing implements

m

n

q

A

U
1

r

S

I

I

t       u

c

- - II I            %.

337

I

. --dmmmmb---

n
IIII

I

i I

b /