17
MECHANICAL ASPECTS OF THE SINGLE-PIECE
CURVED SHELL FISHHOOK
Carleton L. Crain
University of Arizona
The curved shell fishhooks which appear in California, Polynesia, and
elsewhere have received a great deal of primary and secondary interest in the
literature. Descriptions of the outward appearances of the hooks have led
observers to conclude culture contact purely on form and materials (Ekholm
1950:350; Covarrubias 1954:67), and unusual methods of transportation have
been postulated (Heizer 1949). Neither the form nor the materials have been
investigated in a systematic manner; nor have the mechanical functions of the
hooks received scrutiny. This paper will investigate the structure and func-
tion of this type hook and certain aspects of its distribution in order to
place the problem of diffusion in perspective.
The form of the curved shell fishhook has been variously described as
'IC" shaped, circular (Heizer 1949:93), and "U"I shaped (Buck 1932a:203).  Each
of the above describes the pervading curved nature of the hook. For this
paper, a single-piece curved shell fishhook will be any fishhook made of one
piece of shell where a line may be drawn on the average vector of the point
to intercept either the shank or some other portion of the hook.
-The descriptive analysis of the parts of the single-piece curved
shell fishhook used here is suggested by Buck (1932:202-203) and modified for
general application. (See Figure 1t)
The functions of a fishhook may be described in three general areas:
1. The attraction of the fish to the hook
2.  Setting the hook in the body of the fish
3. Holding the fish from escaping
These functions will be considered separately below.
Hook Attraction
Hooks are designed either to carry bait or to attract fish by means
of a lure. Most fish attraction seems to be based on some property of food,
and the lure may resemble some visual quality of food. From observations
made of feeding habits in aquariums, it would seem that fish estimate the
palatability of materials in their mouths after only one other external cue
such as shape, thus making fishing with hook and line feasible.
The shell fishhook without augmentation has the property of a lure
(Linton 1926:26). Most of the hooks found by Harrington (1928) in a Califor-
nia site were of Haliotis cracherodii (black abalone) and H. rufescens (red
abalone). One side of the shell is iridescent and the other side is colored.
Similar shells such as Pinctada gltsoffi (pearl oyster) in the Pacific and
Aetheria (Nile oyster) of Egypt and the Sudan have the above qualities and
have been used for the manufacture of single-piece shell fishhooks.
Attention is paid to both sides of the hook. The dull side of the
hook may be ground to remove any or most of the colored portion leaving necre
18
(Beaglehole 1938:199). The dull portion may be ground to give a desired
shade of color. Buck (1932b:17h) describes the process of grinding the shell
for the proper color and placing the shell in water to judge the correct
shade. Certain shells from certain localities in Polynesia were sought for
color and shade. Conditions of weather, time of day, and the feeding prac-
tices of fish determine the color of the lure (Buck 1932b:175).
.Shell fishhooks were also used with bait. Buck (1938:293) reports
that they were baited with shellfish in Mangareva. While the bait of aborigi-
nal California, South America, Egypt and Shaheinab is unknown, it may be that
the hooks were baited with similar materials, if bait was used.
The curved shell fishhook is used as a lure, a bait carrier, or both.
It would seem likely that most groups using this type hook would use both
qualities of the hook for the added certainty involved. This combination of
characteristics makes this type of hook more versatile than the hook with
which only bait is used--an important consideration when subsistence i s
geared to the sea.
Hook Setting
The process of securing the fish once the bait or lure has been taken
into the mouth presents an intriguing problem. Some of the curved hooks have
clearance of only 5 mm. from the point to the shank limb (Harrington 1928:23-
168). While it is easier for a fish to take a hook with a wide point clear-
ance, it is also easier for the fish to escape the hook. It takes the fish
longer to force his jaws over the rarrow point clearance, but it is more
difficult to expel it (Buck 1932b:168).
Once the hook is in the mouth of the fish, it must be so designed
that the fish cannot escape. It was found that tithe greatest penetration
power of the hooks occurs when the line of penetration is coincident with the
direction of the force applied" (Gakrulson 1956:120). With the hook resting
at the end of the line this presents no problem, for the point of the hook is
180 to the fish line. The point istien coincident with the direction of
force which would be away from the pole. A hook under pressure from a fish
striking the bait pivots-at the point of attachment reducing the angle of
penetration. (See Figure 2,)
The decline of the angle of penetration is in direct proportion to
the ratio of the shank limb to the point limb. When the two limbs are
parallel and of equal length, and the length of the bendvis equal to the
length of the point limb, the angle of penetration is 90 . At no time can the
ideal angle of 180 be reached when the shank limb is parallel to the point
limb and the point of attachment has a flexible line attached. (See Table 1.)
With modification of the point limb or shank limb or both, this ideal angle
is achieved or closely approximated. The range of degrees included in the
most efficient angles cannot be determined experimentally, but the closer the
angle is.to 180 , the more efficient the hook; conversely, the more the angle
is to 90 , the less efficient the hook.
The curved shell fishhook modifies either the whole hook or the point
limb. Figure 3 represents a functional model of the curved hook without
19
pressure being applied. The point is incurved to the line and gives an angle
greater than the ideal angle of penetration. The same hook under pressure
applied at the point pivots and reduces this angle to near. the ideal. The
effective shank limb is equal in length to the point limb. Thus the approxl-
mation of the ideal angle. of penetration is achieved through the incurve
nature of the point limb. (See Figure 30) A modification of the point limb
would be-unsuitable for certain types o.f fishing. Where the fish strikes on
the-run at a moving object, the portion of the hook most likely taken is the
bend. If the bend of the curved shell fishhook is taken and the point limb
is incurved, the point would be functionally covered by the point limb. The
point limb of the bonita hook (see Figure 4) is either parallel to the shank
limb or slightly outcurved while the shank limb is incurved allowing for the
maximum angle of penetration.  (See Figure 4.)
In summary, the curved shell fishhook can only be used in certain
situations. Ethnographically, they have been used as surf, boat, and rock
fishing. They seem to be designed for a fish that explores the bait and
ferrets portions of food in inedible matrices.
There are three methods of presenting the point of any hook for the
most effective angle of penetration. They are:
1. a long shank
2. an incurved point
3. an incurved shank
The curved hook is designed so-that the hook will set even though the shank
is short.
Hook Holding
Once- the hook is set in some portion of the fish, the-function of the
hook is to hold the fish so that it may be retrieved. This function may be
augmented by either the addition of some auxiliary mechanism such as a barb,
through the action of the fisherman, or by the nature of the hook form.  In
any of the above cases, if the hook is lodged in the gut of the fish, the
problem of holding is solved.
With the bonita hook, the forward motion of the boat keeps the line
taut so that it is difficult to throw the hook.
With a stationary line v-the problem is compounded by the tendency of
the fish to resist the confinement of the line. Some fish run with the line;
others jump and flail and try various motions to dislodge the hook.
Barbs are placed on the hook to inhibit the fish from working free.
Some of the curved shell fishhooks have barbs, most do not, After the hook
-is set, force from the line or force from the fish to free himself generally
results in the fish passing down the point limb and resting on the. bend. The
barb is usually placed just below the point to inhibit the upward movement of
the fish.
*  The shank plays an important part in holding. If the shank is too
long, the point of attachment forms a fulcrum when the fish applies pressure
20
from various angles and acts as a lever to free the fish. A short shank does
not provide leverage to the same degree. The disadvantage of the short shank
in holding is directly related to the angle of penetration. When a short-
shanked hook is under stress, the bite of the hook is reduced often not allow-
ing the fish to pass down the point limb. (See Figure 2.)
The bite of the curved shell fishhook is not reduced under stress for
the hook rotates on the axis of a circle. After the hook is set, the resting
equilibrium is regained bringing the point over the bend. Since there is
little shank to provide leverage for movement of the hook, the fish is firmly
secured.
This rotation of the shell fishhook has the added advantage of not
placing the pressure of the strike totally on the hook, but rather on the
fish line at the point of attachment. Pressure is not on the hook until it
is fully set and the fish has descended to the bend; then only that portion
of the hook from the bend to the point of attachment which receives pressure
is reduced. Because of the brittle nature of the shell material, this reduc-
tion in the hook area receiving stress is important.
All but a small percentage of the single-pieced shell fishhooks
examined were curved. An exception to the curved shape was found in the
Marquesas (Linton 1923:1, LXXI). The point is not incurved and the length of
the shank limb is 50 percent to 100 percent greater than the point limb.
This hook appears to be thicker than the incurved ones found in the same area.
Apparently this thickness is related to the absence of the structural advan-
tage found in the curved hooks.
Materials
. The properties of the materials used in the manufacture of fishhooks
are, ideally, the following:
1. strength
2. resilience.
Shell, bone, wood,, thorns, stone, ivory, turtle shell, and coconut
shell constitute the major materials used in the manufacture of single-piece
curved hooks. Only shell and stone lack both properties of strength and
resilience.
While some of the curved hooks from California were made from the
cross sections of the long bone of deer (Harrington 1928:23 et passim), most
are of shell.
Poverty of materials may be used as an argument for the selection of
shell for fishhooks in some areas of the Pacific, but this argument would not
hold for California, Egypt, South America, or the Sudan. Each of these areas
use shell for the majority of the fishhooks during one period and each has an
abundance of wood and large-boned animals. Seemingly, the lure quality of
the shell was the factor in its selection.
From the measurements of 126 whole and partially curved fishhooks of
Haliotis taken from the Burton Mound in Santa Barbara, California (Harrington
21
1928:90-130), it was found that the average width was 5 mm. and the thickness
was 2.5 mm.  It was noted that the bone hooks in the same site f it in all
measurements into the distributional curves of the shell hooks. (See Table 2.)
A strength-of materials test was made on H. rufescens of 5 sm. thick-
.ness And varying widths. At 2.5 mm. the material. would shear 9.75 kg. applied
5 mm. from the fulcrum. The amount of pressure needed to break the. shell
became less as the thickness decreased and the lever became longer. (See
Table 3.)
From Table 3., it can be seen that shell. is a weak material and it has
no resilience. This limitation of material should be reflected in the meas-
urements of the hooks. .If these relationships exist, there can be reasonable
assurance in making statements to the effect that the form of the hook is .
partially from an appreciation of the material and is imposed by the material.
The assumption has been nade that the incurved nature of the hook is
related to the necessarily shortened shank; and that the shank must be short-
ened to insure hook strength. A partial, yet inconclusive demonstration of
this is shown in the straight shanked shell hooks of the Marquesas which are
characterized by:T thick shafts.  If the above assumption holds,. there should
be a correlation between the total size of the hook and the width of the
shaft. . There should also be a correlation between the width and the thick-,
ness of. the shaft. As the hook increases in size, the shaft should reflect
this .increased width.  As the width increases, the thickness should also.
increase proportionately.
The data from the Burton Mound site is used for testing these assump-
tions.  The number of complete hooks is 10; the number of partial hooks is
116 giving a total of 126 single-piece shell fishhooks.
In testing the correlation between the size of the hook and the width
of the shaft, the outside total width of the hook was used to represent size.
The null hypothesis to be tested is that the probability of correlation be-
tween the width of the hook and the width of the. shaft will be greater than
.05. That-is, any regular relationship between the two could occur by chance
more than 5 percent of the time. Anything less than .05 will be considered
to reject the null hypothesis and accept the assumption that the width of the
hook and the-width of the shaft are functionally related.  (See Table 4.)
The formula-(Edwards 1954:148) for the correlation coefficient is:
(~x1)(~y1)
-x y   _                      )
lr 1.2      ( L x         1.9 y2  EIyl2
22
The ratio between the two is .521. Using a one-tailed test of significance,
the probability of the above occurring by chance is less than 2 percent.
This rejects the null hypothesis. The assumption that hook size and shaft
width are functionally related holds. (See Table 5.) Using the above formula,
we find that the ratio between shaft width and thickness is .486 and the prob-
ability is less than .01. Thus, the assumption of the functional relationship
between the width and thickness of the shaft holds.
Hook size, shaft width, and shaft thickness vary in relation to each
other. A thick shaft on a small hook could occur less than 2 percent of the
time; conversely, a thin shaft on a large hook has the same probability of
occurrence. These relationships seem to imply an appreciation of the
strength of shell and the relative pressure which various hook sizes must
withstand. This further seems to imply that the thickness of the shell is
directly related to the total size of the hook. The gross forms of the hook
are probably directly related to the limitations of the strength of the
materials.
Conclusions and Discussion
The form of the shell fishhook seems limited by materials. While the
shell material offers the advantage of being a natural lure, it is brittle
and lacks strength. In order to have the long point or shank limbs required
for the usual type fishhook functions, the shell would have to be thick and
wide--thus limiting the hook's usefulness. If the shaft is to be made thin,
the lengths of the shank, the angle of penetration can only be maintained by
incurving the point limb or the shank. If the shank limb is incurved and
point limb is straight, the hook is more likely to break than if both limbs
are curved.
Therefore, the curved nature of the shell fishhook seems to be a
product of the limitations of material and of design solutions to these limi-
tations.
The distribution of the single-piece curved shell fishhook is not
restricted to one area, but rather, is world-wide. The hooks have been found
in the Pacific, in North America, in South America, and in Africa- (Buck 1932,
Childe 1956, Harrington 1928, Bird 1943, Brunton 1937, Cole 1954). Arguments
for the diffusion of the hook on the grounds of form and materials are unten-
able. The distribution of the hook form seems to be a product of a type of
ecology and the derived technology used to exploit it.
150% 200% 250% 300%
350% 400%
1450% 500%
Shank Limb/Point Limb
Table 1
The angle Of penetration of a fishhook with the point limb and shank
limb parallel and the length of the bend equal to the length of the point
limb. Calculated in the percentage of the shank length greater than the
point limb length.
23
160
140
Angle of
pene-
tration
a
130
120
110
100
J     m  cn i "    CJ C'J O   r C4 J-4 C4     O   )      J
0
;4
24
4
w~e
tn
(n
U)
IL)
0
1T-CT
C1TzT
ZT11Z
11-01
01-6
6-9
9-L
L-9
9-5
?-Z
Z-l
1-0
5*5-5
5-11
5 *1-fl
V0Z-Z
5 1 -1
1-0
CM
r4
d
0
$-4
~4-i
V)
.
To
$4
41
0
(U)
.14
*0
0
$4
HCU
$4-
0
U)
43
"4-
ow
m
Kg. of force
12           56 7J I   8 9   10 11
rm, thickness
Table 3
Strength of H. rufescens of varied thickness and a width of 5 inn.
nin n 8
7
6
5
Thickness  4
3
2
1
1 2 3 4 .
nmm*
I~~~~~
2    ~~~2
I ~ ~ -I  -\  I  I
- 7 1 -' 1 3 3w 4 3 M2=  -:1
.  2  8 81   7  7  5   4  3  11  -  1 1
6  L 7 7F 3  3I :v1-w=__
R  1  2  _%I"  --  1 .^ - ^-1 1 -
6 7 d 9 10 11 12 13 14 15 16
Width
Table 14
A correlation of shaft width and thickness of 126'whole and partial
fishhooks from the Burton Mound Site.
11
10
9
8
7
6
5
4
3
2
1
'0
Mm. 32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
I - I=     -, =
X          I
-~~~~~~~~I
X.  I I   1 -   I   I  -I=
l%  I  - l
xI
1 2 3 4 5 6 7 d 9 10 11 mm.
Shaft width
Table 5
A correlation table of shaft and bend widths of 10 fishhooks from the
Burton Mound site.
26
Bend
width
*      m      *      0     a      *f   *- m            X      Xf
27
SHANK                                POINT
Figure 1.
Structural aspects of single-piece shell fishhooks.
A
8
F i gure 2.
Rotation of a hook with pressure applied to the point
a. Hook at rest
b. Hook under stress
l
86
Figure 3.
Rotation of a shell fishhook with
pressure applied to the point.
Hook at rest
Hook under stress
Figure 14.
A bonita hook.
28
A
a.
b.
29
REFERENCES
BEAGLEHOLE, E. and P. BEAGLEHCLE
1938   Ethnology of Pukapuka.  B. P. Bishop Museum Bull. No. 150.
Honolulu.
BIRD, J. B.
1943   Excavations in Northern Chile. New York, Anthro. Papers,
American Museum of Natural History Vol. 38, pt. 4.
BRUNTON, G.
1937   Mostagedda. London.
BUCK, P. H.
1932a Ethnology of Tongareva. Honolulu, B. P. Bishop Museum Bull. No.
92.
1932b Ethnology of Manihiki and Rakahanga. Honolulu, B. P. Bishop
Museum Bull. No. 99.
1938   Ethnology of Mangareva. Honolulu, B. P. Bishop Museum Bull. No.
157.
CHILDE, V. G.
1956   New light on the most ancient east. New York.
COLE, S.
1954   The prehistory of East Africa. Baltimore.
COVARRUBIAS, H.
1954   The eagle, the Jaguar, and the serpent. New York.
EKHOLM, G. F.
1950   Is American Indian culture Asiatic?  Natural History Vol. LIX,
No. 8.
GAKRULSON, I.
1956   The fisherman's encyclopedia. Harrisburg.
HARRINGTON, JO P.
1928   Explorations of the Burton Mound at Santa Barbara, California.
Annual Report, Bureau of American Ethnology No. 44:23-168.