SPECIAL C O N TRIBU TIO N
D l
The gliding path of the mandible along the skull
F e r d in a n d G r a f S p e e ( 1 8 5 5 - 1 9 3 7 )
Prosector at the Anatomy Institute of Kiel
Translated and reviewed by
M aria A. Biedenbach, PhD
M argaret Hotz, MD, DMD
H. Perry H itchcock, DMD, MSD
T he follow ing argum ents concern the path of the
m andibular m ovem ents such as the gliding of the
occlu sal surfaces of the dental arches against each
other, either ju st in contact or in full occlusion,
possibly w ith interposition of a thin layer of food. In
other words, these are movem ents w hich could be
selectively perform ed as w ell laterally and sagit-
tally. Furtherm ore, a backward gliding of the m an­
dible often occurs at the end of an incising m ove­
ment. T he course of these specifically m asticatory
m ovem ents does not depend only on the m echanical
array of the m andibular joint in its narrower sense
but also on the configuration of the m asticatory sur­
faces of the teeth. A ccordingly, there is harm onious
interaction. However, up to now, no studies have
disclosed this relationship. Langer’s view ,1 for in­
stance, w hich he derived from a study on the low er­
ing of the m andible is not correct. He stipulates that,
during forward and backward gliding in sagittal
direction, the m andibular condyle m oves around
the lower surface of the tuberculum glenoidale in a
curve supposed to be concentric to the surface cur­
vature of the tuberculum glenoidale, the axis of
w hich is supposed to lie w ithin the tuberculum.
T h is view, however, is incorrect, at least, for m ove­
m ents during w h ich dental arches are in full o cclu ­
sion.
670 ■ JADA, Vol. 100, M ay 1980
M eyer,2 who used his com parative anatom ical
studies successfully in order to establish the sign ifi­
cance of fine structural details of the hum an m an­
dibular joint, does not even m ention an axis for
sagittal displacem ent. N either does H enke.3
However, a definite axis can be established at least
in the m ajority of cases, if we consider the occlusal
configuration of the teeth. The m andible rotates
around this axis w hile perform ing the ordinary
sagittal gliding, w ith maybe ju st a sligh t deviation to
the left or the right.
A profile view of a hum an skull usually shows
that on both sides the m asticatory surfaces of the
molars are aligned in a downward convex curve
along the upper jaw and in an upward concave curve
along the lower jaw. T h is relationship is not consid­
ered in anatom ical textbooks. O ccasionally, it is cor­
rectly reproduced in illu stratio n s, esp ecially in
those made by artists according to nature, but gener­
ally it is not shown. It also received no reasonable
consideration in plastic reconstructions of hum an
dentitions by dental tech nician s, nor in the prepara­
tions o f hum an skulls w ith artificially inserted teeth
w hich were made in Paris. W e rather frequently find
an unjustified tendency to arrange the m asticatory
surfaces of the teeth in a horizontal plane.
A good illustration of the downward convex curve
A ll dentists have heard o f th e “Curve of Spee.” It has m eant d ifferen t things to
d ifferen t dentists. S om e h a v e said that it is an an terop o sterio r cu rv e in the
vertical p la n e o f the low er arch. O thers h a v e said th at it is an anteroposterior
curve in the vertical p lane o f th e u p p er arch (with the opposite arch assum ing a
com p en satin g curve).
In trying to learn w hat F erd in an d G raf S p ee actu ally sa id , it w as fo u n d that a
rea d ily available translation o f G raf S p e e ’s original work did not exist.
Two tran slations were n e e d e d : on e fro m archaic G erm an to m odern G erm an,
an d another from m odern G erm an to English.
M aria B ied en b a ch , PhD, assistan t professor of physiology, U niversity of
Texas H ealth Science Center at San A ntonio, tran slated th e origin al article.
H ow ever, sh e was not en tirely sa tisfied that s h e h a d correctly in terpreted the
den tal terms.
Later, M argaret H otz, MD, DMD, Z ah nartzlich es Institut d er U niverstat
Z urich, agreed to tran slate it using current den tal term in ology.
Now, all E n glish-speakin g d en tists can b e aw are o f w hat G raf S p e e sa id ; an d
on e no longer has to d ep en d on an oth er’s interpretation.
I d id n on e o f th e tran slation but on ly served to p ersu ad e Dr. B ied e n b a c h an d
Dr. H otz to interpret this fa m o u s w ork.
H. PERRY HITCHCOCK, DMD, MSD

in the sagittal plane in w hich the m asticatory sur­
faces of upper and lower teeth occlude can be ob­
tained from the photographic profile of the skull in
Figure 1. Here, as a consequence of attrition, the
molars have lost all the cusps of the crow ns so that
th e m u tu a lly abrad ed m a stic a to ry su rfa ces fit
exactly on top of each other. Thus, the upward con ­
cave curve describes a sm ooth line in the sagittal
plane. T eeth with com pletely intact crown cusps
fundam entally also present the same curvelike ar­
rangem ent of (masticatory) occlu sal surfaces, except
that they show an irregularly em bossed profile line,
w hose in d en tation s are clea rly reco g n izab le as
sm all fluctuations on an assum ed curvilinear base
(Zero line) (Fig 2). As an exact exam ination of the
curve is more com plicated in a dentition w ith pre­
served cusps than in cases in w hich they are reduced
by wear, it is preferable to choose the worn cases in
order to discuss the basic findings.
If, in such a special case, the curve is determ ined
in a profile view of the skull, for exam ple, Figure 1,
or w ith a com pass, on its geom etric projection on the
sagittal plane, the findings are the follow ing:
— That the total visible con tact lin e of the molar
m asticatory surfaces lies on the same arc of a circle;
furthermore,
— That in the sagittal plane the posterior continu­
ation of this arc touches the m ost anterior point of
the m andibular condyle.
A ccordingly, the points of the m andible w hich
glide in contact along the upper part of the skull are
lying on the same cylin drical surface. The location
of the axis of that cylin d er’s curvature could be de­
term ined by reconstruction and m easurem ent w ith
the com pass at the level of the horizontal m idorbital
plane, posteriorly from the crista lacrim alis poste­
rior. In the lateral view it appears as a point (Fig 1, a).
The attritional configuration of the teeth develops
through their respective gliding m ovem ents and
thus conform s to the pathway w hich the teeth of the
m andible follow. As the teeth in the jaw s have to be
considered as fixed, this pathway is not only a m an­
ifestation of tooth m ovem ent but an effect of the
displacem ent of the w hole m andible. C onsidering at
the same tim e Figure 1, it follow s that the m ovem ent
occurs in circular paths ju st as a pendulum m oves
around an axis. T h is axis is given by a transverse
connecting line from the sagittally projected centers
of the masticatory curvature of both sides.
From Figure 2, it further follow s that the m ost
anterior part of the surface of the condyle also m oves
on a circular path w ith the same length of radius as
the occlu sal surfaces of the m olars. In other words,
these parts move on the sam e cylin d ric surface.
Spee : GLIDING PATH OF MANDIBLE ALONG SKULL ■ 671
S P E C IA L C O N T R IB U T IO N
v'enag \e’d & Comp, e
The m ost prom inent anterior part of the m andibu­
lar condyle is also the most lateral. Thus, according
to the explanations of M eyer,2 it belongs to that half
of the condyle w hich during sym m etric forward and
backward m ovem ents of the m andible is moving up
and down very closely along the tuberculum.
T he m asticatory surface of the upper teeth and the
articulating surface of the glenoid tuberculum with
its m eniscus form the analogous parts of a big cy lin ­
drical curve around whose surface area the points of
the low er jaw glide. The radius of the glenoid tuber-
672 ■ JADA, Vol. 100, M ay 1980
Fig 1 a, b ■ Illustrations from
same skull; l,a in exact profile
view; 1 ,b from somewhat anterior
view. Fig l,a shows curve for for­
ward movement of mandible; 1 ,b
shows curve for occlusal surface
without reduction in length. Di­
mensions of illustrations are
comparable as they were repro­
duced at equal magnification (r
indicates radius of curve for mas­
ticatory [occlusal] surfaces and c
indicates center of curve).
Fig 2 ■ Dentition of adult with
well-preserved crown cusps.
Only surface of still-present sec­
ond deciduous molar (m) shows
attrition. This surface prolongs
occlusal line of articulatory sur­
faces of posteriorly located mo­
lars quite naturally.
Fig 3 ■ Mandible of child.
Fig 4 ■ Series of points of hu­
man, worn occlusal surface accu­
rately located in relative position
to each other and to anterior sur­
face of mandibular condyle, k.
Natural size. Two preparations a,
b.
culum is m uch shorter than the radius of the cy lin ­
der. Yet a con cen trically arranged series of basic
points corresponding to its surface can be estab­
lished for the com plete lin e of m otion of the m an­
dibular condyle along the tuberculum , because the
thicker part of the wedge-shaped m eniscus in terca­
lates betw een the two areas and com pensates the
shorter radius of the tubercular curvature. As for­
ward and backward gliding of the m andible takes
place in a path of circular m otion, such d isp lace­
m ents can occu r over longer distances w ithout any

need for the arches to separate from each other.
Thus, m asticatory efficiency is guaranteed. A sep­
aration of the occlusal surfaces is only inevitable in
order to overcom e the contact of strongly protruding
upper and lower canines. But this can also be elimi­
nated by wear.
If the occlusal surfaces of the teeth were arranged
on a plane, and, since the tuberculum articulare is
present, parallel shifting of the closely apposed
arches for purposes of crushing and grinding food
would not be possible. This ought to be considered
in the construction of dentures, not only to enable
better m astication but also in order to avoid lever
effects during chewing. These levering effects are
less manifest with increased length of contact sur­
face in every position of the firmly occluded teeth.
The backward gliding of the m andible, usually fol­
lowing an incising movement, is especially effective
and m ainly responsible for sagittal attrition. Isolated
backward m ovem ents of the mandible are rather
unusual.
In addition, as concerns wear, one has to consider
alternating lateral rotation of the mandible around
nearly vertical axes. During these m ovem ents, the
grinding action on the teeth on the side toward
w hich the movement occurs is in alm ost transverse
directions. During this, the teeth of the lower arch
shift in a laterally descending direction below the
m ost protruding lateral border of the upper crowns
(Fig 2). Evidently the teeth of the other side w hich
simultaneously move forward and m edially cannot
remain in contact with the upper arch. This would
also be a consequence of the unilateral protrusion of
the condyle on top of the m edial half of the tuber­
culum (Meyer). The backward movement, during
w hich the mandibular condyle slips back into the
fossa, can occur by gliding of the arches (backward
and outward), just as they are aligned. This m ove­
ment, of course, also follows the upw ard concave
curve w hich was considered above in the sagittal
projection (however slightly shortened), as the path
of sagittal gliding of the mandible. But this time the
movem ent occurs parallel to the greatest length of
the curve w hich is visible only if view ed somewhat
anteriorly and laterally, since it is flatter. From this
perspective, the transverse distance between con­
dyle and lateral surfaces of the teeth does no longer
appear shortened. Y et one finds that in spite of its
longer radius the continuation of this curve also
touches the lateral anterior surface of the condyle
(Fig 1, b).
Strictly speaking, the worn surfaces of the molars
on each side form together, more or less, a rectangu­
lar area. Its diagonal line, slightly sagittally posi­
tioned, is curved spirally towards the opposite side;
since the surfaces of the third molars have an inward
slant, those of the second m olars are m ore horizontal
and those of the first molars inclined outward. Thus
SPEC IA L CO NTRIBUTIO N
it somewhat digresses from a cylindrical surface.
This effect, aside from the difference in thickness of
the enamel, is due to the tilted position of the teeth.
As is well-known, the crow n of the second molar
and even more so of the third is inclined lingually in
the mandible and labially in the upper jaw in order
to contact its antagonist. The resulting com plication
for the occlusal surfaces is not significant enough to
cause a relevant deviation from the path of m andibu­
lar motion just described. During sagittal sym m etric
displacements, the com plication does not even ap­
pear, because all sym m etric m ovement effects in the
direction of the m edian plane are mutually cancel­
led, the sam e as the medially directed forces of the
external pterygoid muscles.
Regarding the path for sagittal displacem ent,
based on autographic curves about the m ovem ent of
the lower incisor, Langer expressed the opinion that
the mandible is shifting on very flat curves or rather
on straight lines. As a rule, this is not the case.
Furthermore, the photographic curve about jaw dis­
placem ent recently obtained by Luce,4 under Bow-
ditch’s guidance, is not in accord with Langer’s
views, but is in agreement with the results which
derive from the modeling of the occlusal surfaces of
the teeth in the path of sagittal displacement. It
shows that the mandible glides forward in an up­
ward concave curve.
However, the autographic curves are short, be­
cause the greatest possible forward gliding distance
of the mandible amounts to about V-k cm (footnote, p
289). With this displacement, the lower incisors and
canines usually enclose the upper ones from an­
teriorly and laterally. At the same time, the buccal
edges of a lower molar move under the buccal edges
of the next anterior upper tooth. For m astication,
only about two thirds of this gliding distance is
used, and for specific grinding movements, only a
third. For m asticatory function, this rather small
gliding distance is very efficient and m ost important
because of its many repetitions on very long m asti­
catory surfaces. Along the whole length of these
surfaces, the d irection of m ovem ent is clearly
marked through the form of attrition, just as on a
surface of a joint. Therefore, the form of the gliding
surface of the molars offers the best information for
evaluating the course of the movement. It is also the
best criterion for analysis of the autographic curves.
Within certain limits, the ligaments of the m andi­
ble generally adapt themselves to the new position
of the shifted mandible without tension. A s Langer1
already found, only the fibers of the medial a c­
c e s s o r y lig a m e n t, w ith in th e f a s c ia b u c c o -
pharyngea, tighten very early during forward glid­
ing. Since it inserts on the anterior surface of the
condyle, a force develops w hich would m ove the
chin in the direction of the forehead. This is in
accordance with the upward concave curve of the
Spee: GLIDING PATH OF MANDIBLE ALONG SKULL ■ 673
S PEC IA L CO NTRIBUTIO N
m asticatory surfaces of the teeth.
In case of a m aximal mandibular forward dis­
placem ent, the concordant movement of the m enis­
cus is restrained by its posterior attachm ents. Its
anterior half being inserted like a strong wedge be­
tween condyle and tubercle, a further inhibition of
condylar forward movement results in conjunction
with the tension of the lateral ligament. I interpret
the small dip at the end of the autographic glenoidal
curve in L u ce’s Figure 4 as a consequence of a more
pronounced interposition of the m eniscus between
the bony parts of the temporomandibular joint.
Specific findings in different dentitions
Among anim als, as far as my experience goes, the
ruminating and hoofed species as well as some pri­
m ates show the sam e upw ard con cave occlusal
curves in their dentition without exception. The
posterior continuation of the curve touches the an­
terior surface of the condyle just as in humans. The
radius of the cylindric surface on w hich the concave
curve lies, is longer in larger animals and more
prognathic skulls. Different lengths were: horse,
21.9 cm ; Cervus elephas, 18.0 cm ; Cervus dama, 12.9
cm ; small monkey, 8.5 cm; and adult man, 6.5 to 7.0
cm.
A com parison of the measurements found in ani­
mals w ith those found in man shows that, aside from
body size, the relative sm allness of the visceral
cranium coincides w ith a shorter radius and a larger
curvature. This appears to point tow ard an increase
of the m asticatory surface within a lim ited special
area and therefore it seems appropriate. Rodents and
carnivores do not show curves of the m asticatory
surfaces. The tuberculum articulare, present in all
the former anim als, is missing. Therefore, develop­
ment of curved occlusal surfaces as well as the circu ­
lar path during sagittal forward displacem ent seems
to depend on the presence of the tuberculum . Based
on this recognition, several human cases w hich de­
viate from the rule can be explained. These cases
w ill be discussed at the end of this paper. I under­
took the specific exam ination of hum an jaws on as
m any sp ecim ens as possible, in order to avoid
p u re ly in d iv id u a l v a ria tio n s . H o w e v e r, th e ir
number w as not very large because only beautifully
preserved and com pletely developed jaws could be
used. Skeletal jaws w ith indiscrim inately and artifi­
cially inserted teeth were rejected. But, since good
jaws of different races yielded the same findings just
described, they have a general significance.
The observations had to establish the positional
relationship between the occlusal surfaces of the
teeth and the surface of the mandibular condyle.
They w ere made in various ways.
The sim plest way to obtain the necessary m ea­
surements is to evaluate exact geom etric sagittal
674 ■ JADA, Vol. 100, May 1980
projections. Since, in adults, the alignm ent of m o­
lars does not decidedly deviate from the sagittal
plane (on the average less than 20°), projections
through large lenses, for exam ple, photographs, can
be used w ithout relevant errors. By measuring the
crowns w ith calipers, I determined the radius of the
curve approxim ately if strong cusps were present (in
ruminants). In abraded m asticatory surfaces in hu­
mans I did it more accurately: the vertical distances
of a line of arbitrary points on the occlusal surface
and of the m ost anterior point of the condylar sur­
face of the mandible were determined from the hori­
zontal plane and correctly entered in a system of
coordinates. Through accurate reconstruction, the
center of a circle touching as many of the points as
possible was determined; the results could not be
anticipated, but in most cases such a center could be
obtained approxim ately. In some cases with decided
abrasion all points were located exactly on the same
circle (Fig 4). Such exact circular attrition is not
explainable by simple wear. Specific effects or form
relationships, which continuously acted toward this
kind of development, m ust have existed from the
beginning. As mentioned at the beginning, this is
partly reflected from the fact, that even in nonabrad­
ed molars, cusps are arranged along a downward
convex circle, w hich already shows in a ch ild ’s jaw
with deciduous teeth at the same time as the tuber­
culum articulare develops. U ltim ately, develop­
mental relationships peculiar to a species m ust be
considered (Fig 2, 4).
In the beginning, the curve of the m asticatory
surface of the mandible is sometimes not very clear
and is m uch flatter than in the upper jaw where it is
seldom missing. In such cases, the varying pressure
exerted on the teeth appears to gradually induce
com pensation of the curve difference.
In general, it appears that teeth whose antagonists
are missing are more and more elongated by new
bone formation in the alveoli. Roots of teeth with
intact antagonists remain deeply set in the alveolus.
Therefore, we may assume that teeth, as long as they
are subject to little pressure, will continue to grow
more than those exposed to stronger pressure. This
would be a continuously acting m echanism in order
to adjust the curves of the masticatory surfaces of
both arches. In adults, the radius of the curve of
mandibular motion had a length of 6.5 to 7.0 cm,
usually equal on both sides. Variations within these
limits depend on the angular position of the ramus
to the arch and on the width and height of the ramus.
In children with complete sets of deciduous teeth,
the radius is shorter than in adults, because of the
generally smaller dimensions, despite the obtuse
position of the still short ramus to the arch. In a
special case the radius measured approxim ately 4.6
cm.
Among the m any individual variations of the

fixed parts of the m asticatory “ array,” we have to
consider those in w hich the upward concave curve
of the m asticatory surface did not exist. They are not
especially rare.
For all of them , it appears that the effect of the
tuberculum articulare is elim inated (blocked). I
have observed the following:
— An extrem ely low or almost nonexisting tuber­
culum (several cases).
— Shortening of the sagittal and vertical dimen­
sions of the ramus on one side so that its condyle
could never have been in the fossa, while the other
side showed the usual configuration (one case).
— Unusual bulk of the condyle with so short a
distance between tuberculum and anterior wall of
the external auditory canal that the condyle of either
side could never have been in the fossa (one case).
— The m ost protruding part of the condyle does
not lie laterally but opposite the posteriorly tilted
medial part of the tuberculum, with the middle of its
anterior surface on its m edial half, so that it does not
stay as deep in the fossa as usual (three cases).
Extrem e thickness of the m eniscus, w hich I, how­
ever, have not observed, w ould also eliminate the
action of the tuberculum.
S PEC IA L CONTRIBUTION
W ith s u c h c o n f ig u r a tio n of th e te m p o r o ­
m and ib ular joint, the m a stica to ry m ovem en ts
would have to occu r through level forward and
backward gliding within the joints as in rodents.
— K ie l, th e 2 4 th o f M a rc h , 1 8 9 0 .
Originally published in the Archives of Anatomy and Physiology,
16:285-294, 1890.
Translation rights and illustrations use granted by Springer Verlag,
Heidelburg.
Dr. Biedenbach is associate professor, department of physiology, Uni­
versity of Texas Health Science Center at San Antonio. Dr. Hotz is founder
of the cleft palate program, department of orthodontics, Zahnartzliches
Institut der Universtat Zurich. Dr. Hitchcock is professor and chairman,
department of orthodontics, University of Texas Dental School at San
Antonio, 7703 Floyd Curl Dr, San Antonio, 78229. Address requests for
reprints to Dr. Hitchcock.
1. Langer, Das Kiefergelenk des Menschen. Sitzungsberichte der
kaiserl. Akademie der Wissenchaften zu Wien 1880. S. 457.
2. Meyer, H. Das Kiefergelenk VH. Beitrag zur Statik und Mechanik des
menschlichen Knochengerüstes. Dies Archiv 1865. S. 719.
3. Henke, Handbuch der Anatomie und Mechanik der Gelenke, 1863.
4. Luce, C.E. The movements of the lower jaw. Boston Med Surg J, 1889,
Figure 4.
Spee : GLIDING PATH OF MANDIBLE ALONG SKULL ■ 675