JOURNAL OF MORPHOLOGY 243:293–299 (2000)
Location of the Vector of Jaw Muscle Force
in Mammals
Walter Stalker Greaves*
Department of Oral Biology (M/C 690), University of Illinois at Chicago, Chicago, Illinois
ABSTRACT Previous work has suggested that the third
molar lies just in front of the point where the resultant
vector of jaw muscle force, estimated from dissections,
intersects the tooth row. This point meets the jaw such
that the vector is 30% of jaw length from the jaw joint.
Thus, the vector divides the jaw in the ratio of 3:7 when
measurements are taken perpendicular to the vector. In
practice, however, distances along mammalian jaws are
typically measured on an easily determined line such as a
line from one end of the tooth row to the other. The
position of the jaw joint is then projected onto this line. As
a rule, such a line is not perpendicular to the vector and so
the distance from the projection of the joint, out to the rear
of the third molar (and the vector’s intersection), is differ-
ent in different mammals. Rarely is this distance 30% of
total jaw length. However, when the location of the vec-
tor’s intersection is measured along the tooth row, this
position varies directly with the inclination of the vector; a
The jaw muscles seem to be attached at the rear of
the jaws in mammals, although some would claim
that this may not be true for all mammals. If this is
true for most, this condition is reasonable, even
though the leverage is relatively poor, if gape is not
to be limited by the amount of stretch in the muscles
or if the avoidance of muscle interference with biting
activities is important.
On the other hand, if strong bite forces are re-
quired, jaw-elevating muscles are expected to be
attached more anteriorly because the leverage is
then better. The out-put forces at some of the poste-
rior teeth also can be larger than the in-put muscle
force. A muscle force located at the rear of the jaw
signals that the typical jaw lever system in mam-
mals is not very efficient and that other consider-
ations are probably more critical.
One such consideration is the location of the re-
sultant force of the jaw muscles. The resultant is
important because the strong jaw-closing muscles
are virtually all active at the point during the mas-
ticatory cycle when large chewing forces are applied
(e.g., de Vree and Gans, 1976; Weijs and Dantuma,
1981). The location of the resultant has an impor-
tant effect on the magnitude of the average bite force
that can be applied along the entire tooth row. This
effect has been studied previously (Greaves, 1988).
This theoretical study considered all the bite forces
© 2000 WILEY-LISS, INC.
vector inclined posteriorly intersects the tooth row far
from the projection of the joint and an anterior vector’s
intersection is relatively close. Only a vector perpendicu-
lar to the line from one end of the tooth row to the other
intersects at 30%. This obvious point suggests a way to
test the above hypotheses when the inclination of the
vector is not known exactly. The predicted relationship
between the distance to the molar, as a percentage of the
total jaw length, and the approximate inclination of the
vector derived from muscle weights (posterior or anterior
depending on whether the temporalis or the masseter/
pterygoid, respectively, is dominant) was observed in a
sample of 46 different mammals. J. Morphol. 243:
293–299, 2000. © 2000 Wiley-Liss, Inc.
KEY WORDS: jaw muscle vector; jaw length; jaw muscle
weights; third molar
along a complete tooth row. Integrating a plot of
these bite forces produced a value for the sum of all
the bite forces along the row. Changing the antero-
posterior location of the muscle resultant force
changed both the length of the tooth row and the
leverage of the jaw mechanism.
Thus, the sum of all the bite forces is different for
each resultant location and is a maximum when the
resultant vector of jaw muscle force is located at a
point 30% of jaw length from the jaw joint. From this
point of view, the tooth row fills the anterior 70% of
the jaw while the posterior 30%, located behind the
line of action of the muscle resultant, is edentulous.
This result is in agreement with what is derived
from dissections and is of considerable interest be-
cause it suggests that posteriorly located muscles
are better after all.
Another important consideration is that if all the
teeth are in front of the resultant force, dislocation of
one of the jaw joints is avoided (Greaves, 1978; Spen-
cer, 1995, 1999). When the force of the jaw muscles
is applied to the jaw anywhere between the most
posterior tooth and the jaw joint, a force at the joint
*Correspondence to: W.S. Greaves, Department of Oral Biology
(M/C 690), University of Illinois at Chicago, 801 South Paulina Street,
Chicago, IL 60612-7213. E-mail: wgreaves@tigger.cc.uic.edu
294 W.S. GREAVES
is directed upward and toward the skull and the
joint virtually always resists a compressive load.
This is true no matter where biting takes place along
the tooth row. This is not true if the muscle force is
applied to the jaw more anteriorly at some point
along the tooth row itself. In this case, a resistant
food item, at any tooth behind the muscle force vec-
tor, could become a temporary fulcrum as the jaw is
closed. Rotation of the jaw around this temporary
fulcrum tends to dislocate one of the jaw joints.
Joints are for the most part compression-resisting
devices that may be damaged by anything more
than modest tensile forces (e.g., Spencer and Demes,
1993; Spencer, 1998, 1999). Thus, joint dislocation
on a regular basis is not expected. A resultant vector
located at some point behind the tooth row com-
pletely eliminates this difficulty.
At the same time, for a given resultant position,
teeth farther back generally have higher bite forces
because the leverage is better. This suggests that,
accepting that joint dislocation is disadvantageous,
the best position for the third molar is just in front of
the resultant vector of jaw muscle force, which is
30% of the way along the jaw. Joint disarticulation is
avoided and bite force along the entire tooth row can
be maximized for a given in-put muscle force. (Note
that because of the three-dimensional geometry of
the jaw, the bite force at the third molar may be
somewhat lower than the force at the first molar
[Greaves, 1978; Spencer, 1995].) In short, reason-
able gape, avoidance of muscle interference with
biting, tooth locations with better leverage, maxi-
mizing the average bite force, and precluding joint
dislocation all require a muscle force in the observed
position near the back of the jaws.
Accepting these points, the correct location of the
third molar (and the resultant muscle force) can be
observed only when measurements are taken per-
pendicular to the resultant vector of jaw muscle
force. This requirement constitutes a major practical
difficulty in real animals because the actual inclina-
tion of the muscle vector can only be estimated from
dissections and is virtually never known exactly (cf.
Weijs and Dantuma, 1981). If the orientation of the
vector is not known exactly, then measurements
that are perpendicular to the vector are only as good
as the estimated inclination of the vector.
Such estimates, as well as theoretical work, are
only partially acceptable and, not surprisingly,
many workers prefer not to rely on them. Rather,
they prefer to use actual morphological landmarks
when measuring skulls and jaws. While this is per-
fectly reasonable, the distances resulting from such
measurements are rarely perpendicular to the mus-
cle vector. Thus, in most studies the apparent loca-
tion of the molar and the muscle resultant will de-
pend on how the measurements are taken.
Accordingly, in many samples the third molar often
appears to divide jaws in a variety of different ratios.
Whatever the true location of the muscle resultant
and the third molar might be relative to the vector,
it cannot be observed with the measurements that
are usually used.
MATERIALS AND METHODS
The Model
Table 1 lists the sample of 46 different mamma- T1
lian species examined in this study. The hypothesis
in this study makes two claims. First, the muscle
vector intersects the tooth row just behind the third
molar. Second, the distance from the joint to the
vector (and the molar) is 30% of jaw length, where
jaw length is the distance from the joint to the inci-
sor, measured perpendicular to the vector. This
model is an idealized geometric construct. Real an-
imals are expected to approach the predictions of the
model, not to mirror them exactly.
In Figure 1, an L-shaped heavy line represents the F1
upper jaw viewed from the side. The horizontal part
of this line runs along the tooth row (Molar–Incisor).
The vertical part of the “L” indicates the distance up
to the small circle (J) that represents the jaw joint.
The projection of this joint onto the line along the
tooth row is j at the corner of the “L.” The arrow (V)
represents the resultant vector of jaw muscle force
and is anteriorly inclined in this example. The point
i is the projection of I onto one of the lines perpen-
dicular to the vector. Thus, Ji is the actual length of
the jaw in this analysis.
Measurements along the two lines Ji and jMI will
be considered below (Fig. 1). Points of interest, (i.e.,
teeth, joint, and vector intersection) are either on or
projected onto these two lines and distances between
these points or projections are measured. Note first
that the vector (V) intersects the jaw just behind the
third molar (M). The (perpendicular) distance from
the vector (V) to the jaw joint (J) and the (perpen-
dicular) distance from the vector to the incisor (I) are
measured along, or parallel to, line Ji because this
line is perpendicular to the vector. The distance to
the joint from the vector is three units and from the
vector to the incisor seven units. Thus, the distance
from the joint to the vector is 30% of total jaw length.
The analogous distances (jM and MI) measured
along, or parallel to, the line that runs along the
tooth row are in the ratio of 1.9:7.1 in this example.
(Measuring jaw length in different ways, such as
from the joint directly to the incisor, gives similar
results, although the effect described here dimin-
ishes with increasing condylar height.) The line jMI
is taken to be the apparent jaw length in this anal-
ysis. The different ratios are due to 1) the fact that
line jMI is not perpendicular to the vector while line
Ji is, and 2) the joint (J) is projected onto line jMI.
Measuring projected distances along different lines
that meet the vector at different angles naturally
gives different results. As obvious as this point is, it
nevertheless cannot be forgotten when studying a
skull or jaw. Measurements along these two lines (Ji
 JAW MUSCLE VECTOR 295
TABLE 1. Species considered in this study

Species Vector jM3/L Species Vector jM3/L

Turnbull (1970) Lagomorpha

Marsupialia Oryctolagus cuniculus ⫺73.9 0.27
Didelphis virginiana 09.4 0.41 Rodentia
Insectivora Sciurus niger ⫺64.2 0.25
Centetes ecaudatus 51.0 0.39 Marmota marmota ⫺60.3 0.18
Echinosorex gymnurus 25.0 0.32 Rattus norvegicus ⫺47.0 0.20
Primates Erethizon dorsatus ⫺50.7 0.28
Perodicticus potto 00.5 0.37 Cavia porcellus ⫺63.8 0.23
Cebus apella 15.8 0.40
Saimiri scirurus 12.2 0.40 Davis (1964)
Macaca mulatta 20.5 0.39 Carnivora
Papio cynocephalus 23.6 0.42 Ailuropoda melanoleuca 21.4 0.29
Mandrillus sphinx 33.1 0.43
Theropithecus sp. 41.0 0.41 Storch (1968)
Erithrocebus patas 19.7 0.42 Chiroptera
Colobus sp. 00.6 0.35 Rhinolophus ferrumequinum 44.5 0.36
Pongo pygmaeus 16.8 0.48 Taphozous nudiventris 55.0 0.41
Homo sapiens 06.1 0.36 Myotis myotis 58.0 0.33
Carnivora Phyllostomus discolor 58.0 0.42
Canus familiaris 43.6 0.32 P. hastatus 61.7 0.36
Ursus americanus 36.6 0.34 Carollia perpicillata 55.5 0.44
Thalarctos maritimus 45.8 0.35 Artibeus lituratus 62.4 0.40
Procyon lotor 40.4 0.34 Noctilio leporinus 68.2 0.34
Perissodactyla Megaderma lyra 67.2 0.41
Equus caballus ⫺73.8 0.23
Artiodactyla
Sus scrofa ⫺37.1 0.29
Camelus dromedarius ⫺08.4 0.27
Cervus sp. ⫺56.0 0.27
Odocoileus virginianus ⫺39.0 0.25
Capreolus capreolus ⫺43.9 0.25
Bos indicus ⫺78.2 0.24
Bison bonasus ⫺78.8 0.30
Kobus sp. ⫺51.7 0.27
Cephalophus sp. ⫺52.6 0.24
Ovis aries ⫺45.9 0.28
O. musimon ⫺48.1 0.27

The numbers in the column labeled ‘vector’ reflect the orientation of the net masticatory muscle vector. The numbers in the column
labeled ‘jM3/L’ is the distance from the projection of the jaw joint (j) to the third molar divided by the length of the jaw measured from
j to I.

and jMI) take on additional relevance when com-
bined with another equally obvious observation.
F2 The diagram in Figure 2 includes three vectors,
each with a different inclination, but is otherwise
similar to the drawing in Figure 1. Each vector is the
same distance (three units) from the jaw joint and so
each is tangent to the circle around the joint that
has a radius of 3. This study accepts the idea that
strong jaw muscle activity at the maximum of the
power stroke of the chewing cycle, in any given an-
imal, will generally produce a vector with a single
inclination. The inclination is expected to change
when muscle activity is reduced or the animal is
using the jaw in a different way.
A dominant temporalis muscle, whose fibers are
posteriorly inclined, produces a resultant vector
(Post) that is also posteriorly inclined (cf. Weijs and
Dantuma, 1981). A dominant masseter/pterygoid
complex is anteriorly inclined and therefore shifts
the resultant vector (Ant) to an anterior inclination.
These two vectors, as well as the vertical one with-
out an arrow (light vertical line), are meant to rep-
resent different animals, even though they are on
the same diagram (e.g., herbivore vs. carnivore).
A line (jI (a, v, or p)) passing along the tooth row is
chosen to represent the horizontal parts of the jaws,
as in Figure 1. The posterior muscle vector (Post)
intersects this horizontal line farther from the pro-
jection of the joint (j) than does the anterior vector
(Ant) [Mp as opposed to Ma]. In this example, the
distances are 35 and 20% of the distance along jMI
(p or a). The vertical vector that is perpendicular to
the tooth row intersects at a point (Mv) in between
these two at 30% of jMI (v and v) [light vertical line].
The inclination of the vector (i.e., the angle it makes
with the line along the tooth row) defines distance
jM (a, v, or p) and vice versa. (Note that total jaw
length when measured along line jMI changes with
vector inclination [jIa, jIv, or jIp], but is always 10
units long when measured perpendicular to the vec-
tors [Jia, Jiv, and Jip ⫽ 10 units].)
The following analysis is based on two simple
ideas: 1) a 30% location of the vector of muscle force
can be observed only when distances are measured
296
Fig. 1. The heavy L-shaped line represents a lateral view of
the upper jaw. The vector of jaw muscle force is represented by
the arrow. See text for further discussion.
perpendicular to the vector, and 2) vector inclination
is related to the point of intersection of the vector
with line jMI. Taken together, these obvious ideas
allow an additional estimate of the muscle result-
ant’s intersection along the jaw. This new approxi-
mation uses a small number of reasonably unambig-
uous measures.
The vector is hypothesized to intersect the tooth
row just behind the third molar. To test this hypoth-
esis approximately, the actual measured distance
from j to the rear edge of the third molar can be
compared with line jM (a, v, or p) that has been
derived from the analysis. This follows because vec-
tor inclinations are correlated with the lengths of
the line segments jM (a, v, or p). That is, when the
inclination of the vector predicts, say, a short dis-
tance from j to the vector’s intersection (jMa), the
measured distance from j to the third molar should
also be short. In part, the hypothesis claims that,
ideally, the distance from j to the third molar equals
line segment jM (a, v, or p).
Measurements
The distance from j to the molar is easily mea-
sured. A measure of vector inclination is also re-
quired, but one that will mirror this variable exactly
is lacking. However, the following analysis only re-
quires an estimate that is reasonable; an exact mea-
sure is not needed. For that reason, the method
ideally requires a relatively large number of in-
stances. Weights of the jaw-elevating muscles can be
used to approximate vector inclination. A posterior
inclination is assumed when temporalis is larger
than masseter and medial pterygoid taken together.
An anterior inclination is indicated by the reverse.
Subtracting the smaller muscle, or muscle group,
W.S. GREAVES
from the larger, dividing by the total weight of all
the elevating muscles, and multiplying this result by
100 gives a number between zero and 100. Numbers
representing larger temporalis muscles, and there-
fore posteriorly oriented resultant vectors, are taken
to be positive. Numbers representing a larger
masseter/pterygoid complex, and thus anteriorly ori-
ented vectors, are treated as negative.
As indicated above, this procedure only gives an
estimate or an approximate indication of whether
the vector is inclined anteriorly or posteriorly (but
see below) and whether it is inclined strongly or
weakly. It clearly does not give the inclination of the
muscle vector exactly, which is what is required in
an ideal case. A correlation is nevertheless expected
between the vector’s inclination and the measured
distance from j out to the rear edge of the third
molar. This follows because the vector’s intersection
with the line along the tooth row is directly related
to the vector’s inclination (Fig. 2).
Jaw muscle weights are unavailable for most
mammals. However, some information is available
(e.g., Storch, 1968; Turnbull, 1970). Unfortunately,
measurements along the tooth row are unavailable
for the specimens that were used to gather the mus-
cle weight data. Therefore, distances along the tooth
row were taken from skulls in the Department of
Oral Biology, from older datasets, and drawings in
the literature (e.g., Davis, 1964; Storch, 1968). This
procedure assumes that all members of a single spe-
cies have similar relative jaw measurements.
RESULTS
Expected Results
Figure 3 is a diagram that indicates the expected F3
results when vector orientation is plotted against
Fig. 2. A diagram similar to that in Figure 1 but including
three representative muscle vectors. Ma, Mv, and Mp ⫽ the
intersection of anterior, vertical, and posterior vectors, respec-
tively. Ia, Iv, and Ip ⫽ the anterior incisors for jaws with anterior,
vertical, and posterior vectors.
 JAW MUSCLE VECTOR
Fig. 3. A diagram to indicate the expected location of data
points when plotting vector inclination (as a proxy for vector
intersection) on the Y axis against the distance from the projec-
tion of the jaw joint out to the third molar (as a percentage of total
jaw length) on the X axis. Distances are measured along a line
drawn along the tooth row. The cartoon (a) shows the jaw mor-
phology of a taxon represented by a point in quadrant A, the
cartoon (b) shows the jaw morphology of a taxon represented by a
point in quadrant B, etc. In reality, taxa are expected only in
quadrants B and C (see Fig. 4).
the location of the third molar. The vertical axis
extends from ⫺100 to ⫹100 and indicates the vector
inclination as estimated from muscle weights. Pos-
teriorly and anteriorly oriented vectors are plotted
above and below zero, where zero indicates a vector
that is perpendicular to the line along the tooth row.
Points near ⫹100 indicate that the vector has a
significant posterior inclination and thus an inter-
section far from j. Points closer to zero represent a
less severe inclination with an intersection that is
closer to j. Anteriorly inclined vectors are plotted
below zero. Strong anterior inclinations are close to
⫺100 and indicate intersections very close to j, while
weak inclinations are found nearer zero and farther
away from j.
These relationships are illustrated with cartoons
positioned at appropriate points along the vertical
axis. In these cartoons, the jaw is represented as an
L-shaped line and the vector as an arrow. As in the
previous diagrams, j is located at the corner of the
“L.” Posterior vectors (large temporalis) point up
and to the left while anterior vectors (large
masseter/pterygoid complex) point up and to the
right. A vertical arrow represents a vector where the
temporalis and masseter/pterygoid complex are ap-
proximately equal. The arrows intersect the jaw at
greater and lesser distances from the corner of the
“L” (j). The tooth row is not shown in the cartoons
along the vertical axis.
The distance out to the third molar as a percent-
age of total jaw length is plotted along the horizontal
axis. The cartoons below the horizontal axis show
only the jaw and a rectangle that represents the
297
posterior teeth. The third molar at the left is close
(0.1) to j (at the corner of the L) and the molar on the
right is relatively far away (0.5).
Four representative cartoons (a, b, c, and d) are
located within the graph space itself. This space is
divided into four quadrants (A, B, C, and D). Each
cartoon shows the jaw morphology of a taxon in its
quadrant. Both the vector and the posterior teeth
are indicated in these cartoons. In the upper left
quadrant (A) the vector intersects far from j (0.4)
because the vector is posteriorly oriented since the
temporalis is large. The third molar on the other
hand is reasonably close to j (0.2). In this quadrant
the vector does not intersect at the rear of the third
molar but rather somewhere in the middle of the
tooth row. The hypothesis predicts that the vector
intersects just behind the third molar and so quad-
rant A is expected to be empty (no data points). The
lower right hand quadrant (D) is also expected to be
empty. In this case an anteriorly oriented vector
intersects close to j (0.2) but the third molar is much
farther away (0.4). Again, the vector does not inter-
sect just behind the molar but much farther back
toward the jaw joint.
The cartoon (b) in the upper right hand quadrant
(B) indicates that both the vector intersection and
the third molar are far from j (0.4). Here the vector
intersects just behind the third molar. This intersec-
tion can only be approximately determined because
of the nature of the data (see above). Data points are
expected in this region. Data points are also ex-
pected in the lower left hand quadrant (C) because
here both the intersection and the third molar are
close to j (0.2). Again, the data are not robust enough
to say that they fall at exactly the same point, only
that this is likely.
Actual Results
The data from Table 1 are plotted in Figure 4. The F4
points on the graph are distributed in the quadrants
more or less as predicted. All of the posterior vectors,
except that representing the giant panda, are paired
with distances from j to the third molar that are
greater than 30% of jaw length and are located in
quadrant B. Anterior vectors are paired with dis-
tances less than this and are located in quadrant C.
Except for the point representing the panda, the
vertical line at 30% of jaw length separates the two
groups of data points.
DISCUSSION
Ideally, muscle weights and jaw measurements
should be taken from the same adult animal. How-
ever, a large amount of information of this kind does
not exist. Multiple examples of each species are also
desirable, but again are simply not available. Fi-
nally, a larger sample of different mammals would
298
Fig. 4. Vector inclination plotted against distance to the third molar in a group of 46 different
mammals. See text for further discussion.
be appropriate, but realistically, data from only a
few hundred different species can be expected.
Another difficulty is that a straight line along
most tooth rows can be drawn only approximately.
Often the row is curved or even bent. Forcing a
straight line along such rows has the effect of in-
creasing the spread of the data points in Figure 4.
More critically, the vertical distance between the
jaw joint and a line along the tooth row varies in
mammals. In Figure 2 a second horizontal line (j⬘I⬘)
represents a line along a tooth row that is closer to
the jaw joint. Distances to the intersection of the
vector measured along jI and j⬘I⬘ are not the same.
Therefore, the apparent distance out to the third
molar varies with the height of the joint above the
tooth row. This relationship also contributes to the
spread of the data points in Figure 4.
In spite of the many difficulties with these data,
very few animals seem out of place in Figure 4. The
point on the graph representing the giant panda is
located just inside the otherwise empty upper left
hand quadrant. The distance from j to the third
molar in the lower jaw of the giant panda is expected
to be longer because a dominant temporalis muscle
W.S. GREAVES
is present. However, this muscle, while large, has a
much more vertical orientation than does the tem-
poralis muscle in a typical carnivoran (cf. Davis,
1964). The resultant muscle vector, in this admit-
tedly odd carnivoran, is thus probably somewhat
anteriorly inclined and not posteriorly inclined. This
suggests that this data point should be located in the
lower left quadrant of the graph with the other
groups with an anterior vector. This case empha-
sizes that one of the assumptions in this analysis,
that the temporalis is a muscle that is strongly pos-
teriorly oriented, is not always true.
Another animal that seems to be out of place is the
bison. With a very large masseter/pterygoid com-
plex, a shorter distance between j and the third
molar is expected. The position of this data point
calls attention to another assumption in the analysis
that was mentioned but not examined. Namely, the
analysis assumed that a vertical vector is perpen-
dicular to a straight line along the tooth row. Exam-
ination of bison skulls shows that the muscle vector
is almost perpendicular to the tooth row even
though it is clearly anteriorly inclined with refer-
 JAW MUSCLE VECTOR
ence to the skull as a whole and according to the
weights of the jaw muscles.
Various other points (e.g., those lying on the hor-
izontal line at 0.0) also seem out of place. Once
again, the orientation of individual muscles, the an-
gle between the vector and the tooth row, or the
nature of the data may explain these poorer fits.
Since muscle weights produce only a general ori-
entation in any case, perhaps skull measurements
that reflect muscle weights, such as areas or dimen-
sions of muscle attachment sites, would be prefera-
ble. Information of this kind, while less robust, is
potentially available from thousands of specimens
from both living and fossil animals. This kind of
information, from a very large sample, is arguably
better than muscle weights from a smaller number
of animals, especially since muscle weights are only
a proxy for vector orientation, which is the unavail-
able information that is actually needed.
Acceptance of the analysis above leads to a con-
sideration of how it can be applied in actual cases. In
a real animal, three points are of interest in lateral
view. These points are the jaw joint (not really a
point), the middle of the rear edge of the third molar
at the alveolar border, and the anterior end of the
anterior incisor. The problem is to determine the
orientation of a vector that has its tail just behind
the third molar. This vector must also be three units
of jaw length from the jaw joint and seven units of
jaw length from the incisor. Imagine such a vector
with an orientation chosen at random. Imagine also
a line perpendicular to the vector that can be slid up
and down the length of the vector while remaining
perpendicular. The (perpendicular) distances to the
jaw joint and to the anterior incisor can be measured
along this sliding line. These distances change as
the orientation of the vector changes. The orienta-
tion of the vector that gives distances to the joint
and to the incisor, that are in the ratio of 3:7, can be
determined by trial and error. According to the anal-
ysis described above, this latter vector orientation is
an estimate of the actual orientation of the muscle
resultant force when the muscles approach maxi-
mum activity levels.
This procedure is obviously easier to carry out on
a drawing than on a skull. Moreover, mammals that
have lost teeth in an evolutionary sense may per-
haps be ignored because problematic estimates will
be required.
Clearly, a three-dimensional rather than a two-
dimensional model is more appropriate for the study
of mammalian jaw mechanics (e.g., Spencer, 1999).
Nevertheless, many functional studies of mamma-
lian jaws are still based on a two-dimensional lever
299
model. Very often these studies find that the lever
arms of jaw muscles and/or teeth seem to vary from
one animal to another. As one common example, a
study will conclude that lever arm differences are
related to different feeding regimes. However, while
individual muscles may have different lever arms,
data have been presented in this article that support
the idea that the lever system of the resultant mus-
cle force is virtually the same in one sample of mam-
mals and that this is probably true for most of these
animals. Moreover, this study implies that if larger
bite forces at the teeth are required, the major al-
ternative is to increase the force supplied by the jaw
muscles because the location of the resultant muscle
force is not expected to change. Muscle force can be
increased by increasing muscle size, changing mus-
cle architecture, increasing the size of the entire
animal, or some combination of these.
ACKNOWLEDGMENTS
I thank B. Patterson and W. Stanley of the Field
Museum of Natural History for permission to study
specimens in their charge. I thank the reviewers for
their insightful comments, M.L. Greaves for helpful
discussion and editorial assistance, and J.S.
Greaves for assistance with the drawings.
LITERATURE CITED
Davis DD. 1964. The giant panda: a morphological study of evo-
lutionary mechanisms. Fieldiana: Zool Memoirs 3:1–339.
de Vree F, Gans C. 1976. Mastication in pygmy goats (Capra
hircus). Ann Soc R Zool Belgique 105:255–306.
Greaves WS. 1978. The jaw lever system in ungulates: a new
model. J Zool (Lond) 184:271–285.
Greaves WS. 1988. The maximum average bite force for a given
jaw length. J Zool (Lond) 214:295–306.
Spencer MA. 1995. Masticatory system configuration and diet in
anthropoid primates. PhD Dissertation, State University of
New York, Stony Brook.
Spencer MA. 1998. Force production in the primate masticatory
system: electromyographic tests of biomechanical hypotheses. J
Hum Evol 34:25–54.
Spencer MA. 1999. Constraints on masticatory system evolution
in anthropoid primates. Am J Phys Anthropol 108:483–506.
Spencer MA, Demes B. 1993. Biomechanical analysis of mastica-
tory system configuration in Neandertals and Inuits. Am J Phys
Anthropol 91:1–20.
Storch G. 1968. Functionsmorphologische Untersuchungen an
der Kaumuskulatur und an Korrelierten Schädelstrukturen
der Chiropteren. Abhandlungen der Senckenbergischen
Naturforschenden Gesellschaft 517:1–92.
Turnbull WD. 1970. Mammalian masticatory apparatus. Fieldi-
ana: Geol 18:147–356.
Weijs WA, Dantuma R. 1981. Functional anatomy of the masti-
catory apparatus in the rabbit (Oryctolagus cuniculus L.). Neth
J Zool 31:99 –147.