cineradiographic study of forelimb movements during quadrupedal walking in the brown lemur (eulemur...

Cineradiographic Study of Forelimb MovementsDuring Quadrupedal Walking in the Brown Lemur(Eulemur fulvus, Primates: Lemuridae)

MANUELA SCHMIDT* AND MARTIN S. FISCHERInstitut fur Spezielle Zoologie und Evolutionsbiologie mit Museum,Friedrich-Schiller-Universitat Jena, D-07743 Jena, Germany

KEY WORDS scapular movement; shoulder evolution; humeralabduction; primate locomotion

ABSTRACT Movements of forelimb joints and segments during walkingin the brown lemur (Eulemur fulvus) were analyzed using cineradiography(150 frames/sec). Metric gait parameters, forelimb kinematics, and intralimbcoordination are described. Calculation of contribution of segment displace-ments to stance propulsion shows that scapular retroversion in a fulcrumnear the vertebral border causes more than 60% of propulsion. The contribu-tion by the shoulder joint is 30%, elbow joint 5%, and wrist joint 1%.Correlation analysis was applied to reveal the interdependency betweenmetric and kinematic parameters. Only the effective angular movement of theelbow joint during stance is speed-dependent. Movements of all other forelimbjoints and segments are independent of speed and influence, mainly, lineargait parameters (stride length, stance length). Perhaps the most importantresult is the hitherto unknown and unexpected degree of scapular mobility.Scapular movements consist of ante-/retroversion, adduction/abduction, andscapular rotation about the longitudinal axis. Inside rotation of the scapula(60°–70°), together with flexion in the shoulder joint, mediates abduction ofthe humerus, which is not achieved in the shoulder joint, and is thereforestrikingly different from humeral abduction in man. Movements of theshoulder joint are restricted to flexion and extension. At touch down, theshoulder joint of the brown lemur is more extended compared to that of othersmall mammals. The relatively long humerus and forearm, characteristic forprimates, are thus effectively converted into stride length. Observed asymme-tries in metric and kinematic behavior of the left and right forelimb arecaused by an unequal lateral bending of the spinal column. Am J PhysAnthropol 111:245–262, 2000. r 2000 Wiley-Liss, Inc.

The aim of this study was to contribute tothe understanding of evolutionary changesof the locomotory apparatus within the The-ria (marsupials and placental mammals).Most profound changes, e.g., the decouplingof the scapula from the trunk, occurred inthe shoulder of the therian stem lineage.Such decoupling resulted in new motionabilities of therians. A characteristic featurein the locomotion of small to medium-sizedtherians is the retroversion of the scapula

about the proximally placed instantaneouscenter of rotation during stance (Miller andvan der Meche, 1975; Jenkins and Weijs,

Grant sponsor: Deutsche Forschungsgemeinschaft; Grantnumbers: Fi 410/1-3, Innovationskolleg Bewegungssysteme INK22/A1-1.

*Correspondence to: Manuela Schmidt, Institut fur SpezielleZoologie und Evolutionsbiologie mit Phyletischem Museum, Fried-rich-Schiller-Universitat Jena, Erbertstraße 1, D-07743 Jena,Germany. E-mail: [email protected]

Received 20 August 1998; accepted 12 September 1999.

AMERICAN JOURNAL OF PHYSICAL ANTHROPOLOGY 111:245–262 (2000)

r 2000 WILEY-LISS, INC.

1979; Fischer, 1994). This movement aloneaccounts for more than two thirds of stancepropulsion in various mammals (Fischer,1998; Fischer and Lehmann, 1998; Schillingand Fischer, 1999). Furthermore, limb geom-etry remains surprisingly rigid throughoutthe stance phase. Only minor angular move-ments occur in the shoulder and elbow joint.The kinematics can best be described aspantograph limb (Fischer and Witte, 1998).

Within primates, humeral retraction re-places scapular retraction with increasingbody size and forelimb length. Whiteheadand Larson (1994) described scapular andhumeral movement during walking in avervet monkey (Cercopithecus aethiops). Inthis species, the retro- and anteversion ofthe scapula differ in amount of rotation andin timing from recordings in other small andmedium-sized mammals (Fischer, 1998;Schilling and Fischer, 1999), and humeralmovements prevail. Based on a single study,it is hard to tell whether these results repre-sent characteristics of locomotion in quadru-pedal primates in general. Small to medium-sized primates could still share a principle ofthe locomotion of other small mammals:crouched posture and scapular movement ina high fulcrum (Fischer, 1994). To test thishypothesis, we analyzed the movements offorelimb joints in the brown lemur (Eulemurfulvus), a typical arboreal quadrupedal pri-mate of medium size (Ashton and Oxnard,1964; Hildebrand, 1967; Napier, 1967; Rose,1973; Walker, 1974). Arboreal quadrupedal-ism is regarded as the ancestral mode oflocomotion in primates (Cartmill, 1972; Jen-kins, 1974a).

Quadrupedal primates are also known todiffer from nonprimate quadrupeds in gaitpatterns (Hildebrand, 1967; Vilensky, 1987,1989) and in stride lengths (Alexander andMaloiy, 1984; Reynolds, 1987).Adult quadru-pedal primates usually use diagonal-se-quence, diagonal-couplets walk. Primateswith such a symmetrical gait pattern canshow asymmetrical behavior of the left andright limbs, particularly in the forelimbs(Hildebrand, 1967). One forelimb can take alonger stride than the other, enabled byasymmetrical bending of the spinal column.Lateral dominance and hand preferenceshave been frequently analyzed for nonloco-

motory movements of forelimbs in primates.Left-hand reaching preferences were provenfor many prosimians (Larson et al., 1989).Surprisingly, other studies (Vilensky andGehlsen, 1984; Demes et al., 1990) analyzedonly one forelimb, probably assuming thatthe contralateral limb would behave in muchthe same way.

In this study, we compare metric andkinematic parameters of forelimb move-ment of the brown lemur with studies ofother primates (Alexander and Maloiy, 1984;Vilensky and Gehlsen, 1984; Reynolds, 1987;Demes et al., 1990; Whitehead and Larson,1994; Jenkins et al., 1978) and of small andmedium-sized nonprimates (Jenkins, 1971,1974a,b; Jenkins and Weijs, 1979; Fischer,1998; Fischer and Lehmann, 1998; Schillingand Fischer, 1999). We are especially inter-ested in whether scapular mobility is re-stricted in favor of an increased range ofhumeral movement in ancestral arboreal-quadrupedal primates, as has been observedin the vervet monkey (Whitehead and Lar-son, 1994) and in the spider monkey (Jen-kins et al., 1978).

As all movements of proximal limb seg-ments are hidden under the skin, fat tissue,and muscles, cineradiography was appliedto study the kinematics in detail.

ANIMALS AND METHODS

One male and one female brown lemurwere obtained through approved sources byour Institute. Experiments were approvedby the Committee for Animal Protection ofthe state of Thuringia. Experiments wereperformed on the male brown lemur, whichwas trained by positive conditioning to walkon a horizontal motor-driven rope-mill, anarboreal analogue of a treadmill. Tread speedwas not fixed, but held manually at a rela-tively constant level during X-ray shots.Only a few cineradiographic records werecollected of the proximal forelimb of thefemale brown lemur, which walked on a flattreadmill during a preliminary X-ray ses-sion. After this session, the animal wasexcluded from further experiments becausewe diagnosed a skeletal disease (spondylo-sis).

246 M. SCHMIDT AND M.S. FISCHER

Cineradiography

The X-ray equipment consists of an auto-matic Philippst unit (type 9807 501800 01)with one X-ray source image intensifierchain. Pulsed X-ray shots were applied (,50kV, 200 mA). Images were recorded using acamera (Arritechno R35-150) on 35-mm filmat 150 frames/sec. The animals were filmedin a lateral projection by a maximum expo-sure time of 10 sec. The image intensifierhas a field of 20.5 3 15 cm and did not allowrecordings of complete forelimbs. Therefore,proximal segments (scapula, humerus) anddistal segments (forearm, hand, and wristjoint) were filmed separately. An orthogonalwire grid, perpendicular to the projectionplane, provided reference points for motionanalysis and for correction of geometricaldistortions in the sagittal plane.

Two copper wires (20 mm) superficiallyattached on the left upper arm and forearmmade it possible to distinguish left and rightforelimb and to estimate abduction and ad-duction angles of the humerus and forearm.As muscles of the upper arm and forearmhave the form of a conical cylinder aroundthe bones, the markers do not lie exactlyparallel to the longitudinal axis of the bones.Thus, the abduction angle of the limb seg-ment inferred from the foreshortening ofmarkers is only an approximate value. Asthe distal end of humerus was frequently offthe screen, the real abduction angle of thehumerus could not be calculated from fore-shortening of the bone in parasagittal projec-tion. Stance and swing lengths were deter-mined from videotapes synchronously takenwith cineradiography, because they ex-ceeded the length of the image intensifier.

Processing X-ray images

X-ray films were copied onto video tapesand analog/digital (A/D)-converted using avideo processing board, and were interac-tively processed by the software ‘‘Unimark3.6’’ (by R. Voss), which was specificallydeveloped for this purpose. It allows theresearcher to interactively digitize previ-ously defined landmarks with a cursor func-tion, to correct distortions automatically,and to calculate angles and distances. Thepositions of digitized landmarks and calcu-

lated angles in the parasagittal plane areillustrated in Figure 1a,b. Calculated anglesare the projections of actual angles onto thesagittal plane, representing their contribu-tion to movements in the plane of forwardmotion.

The error of landmark digitization and itsinfluence on calculated angles were testedby repeating digitization of one sequence(,25 frames) five times. The digitizationerror depends on the size of the animal andthe image contrast of skeletal elements. Forthe brown lemur, digitization error is com-paratively small. It ranges from 0.3°–0.6°for segment angles. It is roughly 1° for jointangles, because the errors of adjacent seg-ment angles may be cumulative in joints.

Analysis of angular movements and theircontribution to propulsion

Maximum amplitudes of joint excursionduring stance and swing phases, and thetiming of segment and limb joint move-ments, were calculated. Effective angularmovements were defined as the differencebetween angles at touch down and lift off.Metric gait parameters (speed, stride dura-tion, and stride length) were also deter-mined. A correlation analysis was applied toreveal the interdependency between metricand kinematic parameters. The statisticalsignificance of left/right differences in met-ric and kinematic parameters was testedusing Student’s t-test.

Fischer and Lehmann (1998) proposed anew approach (‘‘overlay method’’) for calcu-lating the relative contribution of angularmovements to stance propulsion, consider-ing the displacement of fulcra of limb seg-ments. Calculations are based on mean val-ues of typical gait sequences, of which stanceand swing phases are set in the same dura-tion using the method of linear interpola-tion. Based on this method, the data of eachlimb segment are smoothed but their charac-teristics are preserved. Then, a polynomialfit of the sixth order is required to interpo-late the data of the stance phase, and toincrease the number of values (to about 50)in order to reduce the error factor in thefollowing calculations. For calculation, angu-lar values were defined against the verticalplane in such a way that they are positive if

247FORELIMB MOVEMENTS IN THE BROWN LEMUR

the distal end of the segment is in front ofthe proximal end. The horizontal distance(lp) between fingertip and the highest ful-crum of the limb was determined for eachdiscrete limb configuration during stance,using the length of segments and their angu-lar values against the vertical plane (Fig.1c). By turning the proximal segment intothe next configuration without changingangles in the more distal joints, the differ-ence between the horizontal excursion atinstant i (lpi) and at instant i 1 1(lpi11) is thepropulsion caused by rotation of this seg-ment (Fig. 1c). The absolute contribution tostance propulsion of each segment is given

by summation of all data in stance phase.Afterwards, the contribution to forward mo-tion of the remaining segments was calcu-lated in the same way, but under subtractionof the angular movement, which is achievedby the rotation of the more proximal seg-ment or segments.

Terminology

Limb joint angles were defined anatomi-cally and measured at the flexion side ofeach joint. Segment angles were calculatedagainst the horizontal plane (Fig. 1b). Weuse the term anteversion for the cranialdisplacement of the distal end of each seg-

Fig. 1. Kinematic analysis. a: Selected skeletal landmarks. b: Calculated joint and segment angles. c:Calculation of the contribution of segment displacement to stance length.


ment (5 cranial rotation or protraction).Retroversion describes its caudal displace-ment (5 caudal rotation or retraction). Inthe special case of the shoulder joint, we donot use the human-oriented terminology inwhich flexion in this joint refers to cranialmotion of the distal humerus. In quadrupe-dal mammals, flexion in the shoulder joint isdefined as the decrease of the caudal anglebetween scapula and humerus (Boczek-Funke et al., 1996; English, 1978; Goslow etal., 1981; Tokuriki, 1973). As shoulder jointmovements normally consist both of hu-meral and scapular displacements, we can-not follow the suggestion of Whitehead andLarson (1994) to avoid the terms flexion andextension for describing shoulder joint move-ments. The definitions of metric and kine-matic parameters are summarized in theAppendix.

RESULTSMetric gait parameters

Metric gait parameters were availableonly for the male brown lemur. Speed, strideduration, and stride length were measuredfor a total of 90 stride cycles (Table 1, part 1).Observed speeds ranged from 0.5–1.6 m/sec.The brown lemur increases speed by increas-ing stride frequency. Stride duration de-creases significantly (P 5 0.001) with in-

creasing speed (Fig. 2a). Maximum strideduration (1.26 sec) was measured at mini-mum speed (0.46 m/sec). Reduction of strideduration results from both a decreasingstance and swing duration (Fig. 2c,e). How-ever, the relationship between stance dura-tion and speed is not linear. Stance durationdecreases more at lower than at higherspeeds (Fig. 2c).

To test the statistical significance of ob-served left/right differences in metric param-eters, we chose some strides (n 5 12) withidentical walking speeds and compared thevalues of the left and right forelimb usingStudent’s t-test (Table 1, part 2). Left andright forelimbs differ significantly in stanceduration (P 5 0.001). The right forelimbshows a relative longer stance duration (dutyfactor, 62%) than the left forelimb (dutyfactor, 56%).

Only on the left forelimb we did observe asignificant increase of stride length withincreasing speed (P 5 0.1). The swing lengthis positively correlated with speed in bothforelimbs (Fig. 2f ). Stride length of the leftforelimb is roughly 4 cm shorter than that ofthe right forelimb at the same speed (Table1, part 2). We observed these differencesunder treadmill conditions but also in unre-strained locomotion. Videotape recordings ofunrestrained locomotion also exist for the

TABLE 1. Metric gait parameters

Parameter

Left forelimb (n 5 57) Right forelimb (n 5 33)

Meanvalue 6 SD Range

Meanvalue 6 SD Range

Part 1: Complete database for male brown lemur

Speed (m/sec) 0.83 6 0.22 0.46–1.45 0.94 6 0.20 0.58–1.61Stride duration (s) 0.88 6 0.19 0.49–1.26 0.83 6 0.16 0.51–1.17Stance duration (sec) 0.49 6 0.13 0.21–0.85 0.56 6 0.13 0.34–0.76Swing duration (sec) 0.39 6 0.11 0.17–0.63 0.27 6 0.09 0.13–0.52Stride length (m) 0.69 6 0.07 0.49–0.84 0.76 6 0.09 0.55–0.89Stance length (m) 0.33 6 0.04 0.23–0.46 0.36 6 0.05 0.26–0.44Swing length (m) 0.36 6 0.05 0.23–0.46 0.40 6 0.05 0.27–0.46Stride frequency (Hz) 1.20 6 0.28 0.79–2.04 1.25 6 0.26 0.85–1.96

Part 2: Selection of strides (n 5 12) with identical speeds to show the statistical significance of left/right differences

Speed (m/sec) 0.84 6 0.15 0.58–1.07 0.84 6 0.15 0.58–1.07Stride duration (sec) 0.83 6 0.16 0.58–1.19 0.87 6 0.14 0.64–1.17Stance duration (sec) 0.47 6 0.12 0.34–0.76 0.54 6 0.10 0.44–0.71 ⇐ P 5 0.001Swing duration (sec) 0.36 6 0.10 0.17–0.51 0.33 6 0.10 0.17–0.52 ⇐ P 5 0.5Stride length (m) 0.68 6 0.07 0.52–0.77 0.72 6 0.09 0.55–0.85 ⇐ P 5 0.1Stance length (m) 0.32 6 0.04 0.25–0.40 0.34 6 0.05 0.26–0.42 ⇐ P 5 0.1Swing length (m) 0.36 6 0.05 0.23–0.42 0.37 6 0.05 0.27–0.45Stride frequency (Hz) 1.25 6 0.23 0.84–1.72 1.18 6 0.20 0.85–1.56Duty factor (%) 56 6 9 44–77 62 6 8 47–79 ⇐ P 5 0.001


female brown lemur. This animal shows nosignificant differences between left and rightforelimb in metric gait parameters.

Kinematic study

To show that the observed differencesbetween left and right forelimb kinematicsare likely an individual eccentricity of themale brown lemur, we report the data of the

proximal forelimb kinematics of the femalewalking on a flat treadmill. We show somecommon traits, but we do not discuss thedifferences between these animals, whichcould also have been caused by the differentsubstrate.

The description of the kinematics of theforelimb comprises the joint and segmentangles at touch down and lift off, the ampli-

Fig. 2. Changes of metric gait parameters with increasing speed. a: Stride duration. b: Stride length.c: Stance duration. d: Stance length. e: Swing duration. f: Swing length. r, left forelimb; s, rightforelimb.


tude of angular movements, and their effec-tive contribution to linear parameters, to-gether with the intralimb coordination ofjoint movements. We distinguish betweenretro- and anteversion of limb segments,and proper movements of joints. Segmentdisplacement is not necessarily linked withflexion or extension in the adjacent joint butmay merely follow movements initiated moreproximally.

In arboreal quadrupedal locomotion, thediameter of branches is usually smaller thanthe diameter of the animal’s trunk. Thehands are placed in front of the trunk, butduring stance they are positioned under thetrunk. Therefore, the elbow joint is lateral tothe shoulder joint, the humerus moves intoan abducted position, and the forearm isadducted. Abduction and adduction anglesare given separately. Joint angles are pro-jected in the sagittal plane to give theircontribution to propulsion in the direction ofmovement.

Three-dimensional movements of thescapula. Retroversion of the scapula (syn-onymous caudal rotation of Fischer, 1994; orextension in the sense of Miller and van derMeche, 1975; English, 1978; Boczek-Funcke

et al., 1996) begins at touch down at anangle of 45° (65°), which is the minimumangle during the whole stride cycle in themajority of observations (Table 2). It ends inthe last fifth of stance at an angle of about90° (69°). Thus, anteversion begins beforelift off (Fig. 5). Consequently, maximumamplitudes are higher than the effectiveangular movements during the stance phase(Table 2). For the female brown lemur, therange of scapular movement during onestride cycle is comparable with the degree ofscapula excursion in the male (Table 3). Thescapula angles at touch down and lift off areindependent of speed and stride duration.Changes in stance length are only correlatedwith angles at lift off (r 5 0.72, n 5 82,P 5 0.001). No scapular translation alongthe thoracic wall could be observed. Thefulcrum of retro- and anteversion is situatedin the proximal third of the scapular spine,near its vertebral border.

Scapula movements of the brown lemurdo not only consist of retro- and anteversionbut include rotation about the longitudinalaxis of the scapula to a hitherto unknownand unique degree (Fig. 3). The rotationangle is estimated by comparing scapula

TABLE 2. Statistics of kinematic parameters of the male brown lemur (during rope-mill locomotion):angle at touch down and lift off, and maximum amplitude of stance

Angle (°)

Left forelimb Right forelimb Left/rightdifferencesn Mean 6 SD Range n Mean 6 SD Range

ScapulaTouch down 66 49 6 3 43–56 58 47 6 5 31–58Lift off 45 82 6 8 56–98 37 85 6 8 71–100Amplitude 29 42 6 9 24–56 25 56 6 7 31–69

Upper armTouch down 114 126 6 13 86–154 82 127 6 10 96–14Lift off 78 0 6 8 221–20 70 14 6 7 22–33 ⇐ P 5 0.001Amplitude 55 130 6 17 90–158 59 131 6 17 81–170 ⇐ P 5 0.1

ForearmTouch down 23 25 6 3 19–31 11 22 6 5 15–30Lift off 37 118 6 14 76–138 22 126 6 11 105–142 ⇐ P 5 0.001Amplitude 10 109 6 24 56–150

MetacarpusTouch down 10 6 6 3 3–12 8 5 6 3 3–8Lift off 10 69 6 15 48–90 7 71 6 12 52–93

Shoulder jointTouch down 62 168 6 11 141–186 47 167 6 10 145–184Lift off 45 78 6 11 57–103 35 102 6 7 90–124 ⇐ P 5 0.001Amplitude 27 94 6 15 61–130 25 71 6 12 47–89 ⇐ P 5 0.1

Elbow jointTouch down 23 153 6 11 137–171 9 151 6 7 140–157Lift off 26 118 6 18 79–139 15 144 6 8 122–156 ⇐ P 5 0.001Amplitude 12 88 6 19 60–121

Wrist jointTouch down 10 195 6 5 186–200 8 198 6 4 189–202Lift off 10 223 6 10 208–239 7 229 6 9 209–234


shape in the cineradiographic frame with ascapula from a dissected skeleton. The angleis given in the horizontal plane. When theangle is 0°, the scapula is in a parasagittalposition. At 90° it would be parallel to thetransverse plane, with the superior borderfacing medially. The scapula begins to rotatemediad in the middle of swing phase. Attouch down, the angle of median rotationreaches a value of 60°–70°, and the scapulais laid onto the dorsal side of the thorax(Figs. 3a, 4). The scapula stays in this posi-tion until midstance. Only then does it ro-tate back to a parasagittal plane (Figs.3c, 4).

During the stance phase the scapula isalso abducted, with the glenoid fossa posi-tioned lateral to the vertebral border. Theabduction angle of 25°–30° remains nearlyconstant during stance. After lift off andduring the first quarter of the swing phase,the scapula is in a roughly parasagittalposition. The angle of abduction increasesagain to 25°–30° before touch down.

Movements in the shoulder joint. Asexplained previously, the shoulder joint angleis measured on the posterior side of thescapula and humerus (Fig. 1), and flexion ofthe shoulder joint means the diminishing ofthis angle. The shoulder joint displays abiphasic kinematic movement pattern dur-ing the stride cycle (Fig. 5). Flexion beginsshortly before or at touch down and reachesa minimum angle when the wrist joint passes

underneath the shoulder joint. It is thenfollowed by only a slight extension. A secondflexion begins synchronously with the begin-ning of scapula anteversion, even before liftoff. The minimum angle is reached immedi-ately after lift off. Afterwards the shoulderjoint extends continuously until the end ofthe swing phase.

To test the statistical significance of theleft/right differences in the kinematics of themale’s forelimb, we chose some strides ofidentical speed and compared the kinematicparameters using Student’s t-test. Shoulderjoint angle at touch down in both forelimbshas nearly the same value on the left andright side (Table 2, Fig. 5). In contrast tothis, the angles at lift off are significantlydifferent (P 5 0.001). The right shoulder jointis more extended. The female shows nosignificant differences in shoulder joint move-ment between the left and right limb.

Correlation analysis does not show a sig-nificant linkage of shoulder joint movementwith speed or stride duration. Only the leftshoulder joint angle at touch down is closelycorrelated with stride length (r 5 0.79,n 5 27, P 5 0.001). The increase of shoulderjoint angle at touch down leads to an in-crease in stride length. Swing length de-pends on the extension of the shoulder jointbefore touch down in both extremities.

Differences between kinematic param-eters of the left and right shoulder joint inthe male, together with the adjoining scapu-

TABLE 3. Statistics of kinematic parameters of the female brown lemur (during treadmill locomotion):angle at touch down and lift off, and maximum amplitude of stance of the proximal forelimb

Angle (°)

Left forelimb Right forelimb

n Mean 6 SD Range n Mean 6 SD Range

ScapulaTouch down 15 52 6 4 47–58 19 51 6 3 44–57Lift off 16 95 6 6 82–105 18 94 6 6 83–111Amplitude 8 55 6 5 49–65 12 53 6 4 46–60

Upper armTouch down 15 117 6 11 85–132 18 118 6 18 78–136Lift off 17 0 6 5 29–7 19 2 6 7 210–17Amplitude 9 120 6 9 107–135 13 119 6 19 74–138

ForearmLift off 7 125 6 6 115–134 6 125 6 8 117–142

Shoulder jointTouch down 14 169 6 12 132–180 18 169 6 18 132–187Lift off 15 94 6 7 83–105 18 97 6 7 83–109Amplitude 8 76 6 7 64–84 12 77 6 21 34–104

Elbow jointLift off 7 125 6 6 114–132 6 124 6 12 108–143


lar movements, lead to differences in thebehavior of the left and right humerus. Theleft and right humeri are in about the sameorientation at the beginning of the stancephase (Table 2). In the last quarter of thestance phase, retroversion of the right hu-merus ends at an angle of 17° below thehorizontal (Fig. 4). The left humerus ismoved more caudally and reaches a horizon-

tal position at the end of stance phase. Theright humerus is also placed more horizon-tally when stride length increases. The de-scribed scapular anteversion before lift off iscompensated for by a flexion in the shoulderjoint, and the horizontal orientation of thehumerus is kept. Correlation analysis indi-cates that the orientation of the humerus atlift off is intimately linked to stride length(r 5 0.73, n 5 65, P 5 0.1). Only in a horizon-tal orientation may its complete length con-tribute to stride length. Such a horizontalposition of the humerus at the end of thestance phase was also observed for the fe-male brown lemur (Table 3).

The humerus is in an abducted orienta-tion during stance, with a maximum abduc-tion angle of 50°–60° in the first quarter ofthe stance phase. Then a continuous de-crease of humeral abduction occurs, until aminimum angle of 20° is reached at lift off.The abduction angle of the humerus in-creases continuously again during swing.Humeral abduction in the brown lemur isachieved strikingly differently than in man.In the brown lemur, abduction is caused bythe described mediad rotation of the scapula.Whenever the scapula rotates medially andlies on the dorsal side of the thorax, thehumerus follows passively, and is abductedas well as medially rotated around its longi-tudinal axis (Fig. 3). We never observedmediolateral movements within the shoul-der joint. During locomotion the movementin this joint is restricted to flexion or exten-sion. As the flexion/extension plane of thejoint rotates medially together with thescapula, flexion of the shoulder joint in thisscapular orientation results in apparent ab-duction of the humerus.

Elbow joint. Like the shoulder joint, theelbow joint angle has a biphasic kinematicpattern during one stride cycle (Fig. 5). Thefirst flexion of the joint begins at touchdown. Subsequent extension begins in themiddle of the stance phase and reaches itsmaximum at lift off. After lift off, the joint isflexed and remains flexed until the lastquarter of swing phase, when a second exten-sion begins and lasts until touch down.

Fig. 3. Cineradiographic records. a: Touch down. b:Midstance. c: Lift off. Note the change of position andlength of the surface marker (outlined), which indicatesthe median rotation and the abduction of the upper arm.


While elbow joint angles show no signifi-cant differences in both extremities at touchdown (Table 2), their lift off angles aresignificantly different (P 5 0.001): the rightelbow joint angle is more extended. Effectiveangular movement of the right elbow joint isonly about 7°, but it is 36° on the left.Maximum amplitude during stance is morethan twice as high as effective angular move-ment (Table 2). Especially at the elbow joint,the differences between maximum and effec-tive angular movement are striking. Themaximum amplitude is considerably higherthan the effective angular movement, point-

ing to a vertical effect. Flexion of the elbowjoint during stance allows the body to movewhile keeping the center of gravity at aconstant level between forelimbs.

Only the elbow joint shows a significantcorrelation between effective angular move-ment and speed (r 5 0.98, n 5 12, P 5 0.001)or stride duration (r 5 0.89, n 5 12, P 50.001). At increasing speed we observed anincrease of effective angular movement,caused by a decreasing joint angle at lift off.A decrease in this angle also leads to areduction in stance and swing duration, andtherefore again to increasing speed. Joint

Fig. 4. Skeletal orientation of the forelimb at touch down and lift off. a: Dorsal aspect. b: Lateralaspect.


angle at touch down at the end of the stridecycle is closely correlated with stride length(r 5 0.96, n 5 9, P 5 0.001).

As the humerus is abducted during thestance phase, the forearm must be adductedcorrespondingly so that the hand may reachthe branch or rope. At touch down, theadduction angle of the forearm is 30°–33°.This angle is smaller than the humerusabduction angle, because the forearm is

about 20 mm longer than the humerus. Incontrast to humeral abduction, adduction ofthe forearm does not decrease continuallyduring stance. In the first third of the stancephase, the adduction angle decreases toabout 22°. Adduction then ceases until thelast quarter of the stance phase, before itrestarts at about 80% of stance duration.The angle at lift off is approximately 18°. Atthis time the adduction angle of the forearmand abduction angle of the humerus areidentical. While humeral abduction reachesits minimum angle at lift off, adduction ofthe forearm continues further. The mini-mum angle of approximately 13° is reachedat 38% of swing. The following adductionlasts until touch down and brings the handcloser to the rope.

Wrist joint. The wrist joint angle is de-fined as the intersection of the forearm axiswith a line between the metacarpophalan-geal and carpometacarpal joints (Fig. 1b).The kinematics of the more distal handjoints (metacarpophalangeal joint, interpha-langeal joints) were not considered here,because they do not contribute to propulsionin the direction of movement.

At touch down, the hand is placed in asemi-digitigrade position, at an angle of 7° tothe horizontal plane. The wrist joint angle is195° at touch down and approximately 223°at lift off. We did not observe differences inthe angular movement of the left and rightwrist joints.

As stance and swing lengths of one strideare longer than the diameter of the screen,we could not record the course of the wristjoint through one complete stride cycle.Therefore, we determined the effective angu-lar movement of the stance (528°) frommean values of angles at touch down and liftoff. The mean maximum amplitude is 57°.

The wrist joint angle reaches its minimumat touch down. After touch down the jointangle increases continuously up to the lastquarter of the stance phase, and achieves amaximum extension of about 252°, at themoment when the wrist joint passes under-neath the elbow joint. The following flexionlasts until midswing.

Fig. 5. Graphs of joint and segment angles of (a) leftforelimb and (b) right forelimb. Stick figures illustratethe forelimb at several moments of a step cycle. Thewrist joint and hand were omitted, because data aremissing for the swing phase. Lift off is marked by thevertical dotted line.


Contribution of segment displacementto stance length

Scapula retroversion accounts for 63% ofstride length in both forelimbs. Proportionsfor the left shoulder joint are 31%, left elbowjoint 5%, and wrist joint 1% (Fig. 6). Bycontrast, the right shoulder joint contrib-utes only 23% to stride length, and the rightelbow joint 13%.

The elbow joint contributes to stridelength, despite the fact that the resultingjoint movement of stance phase is a flexion.Flexion about a low fulcrum in the first halfof the stance phase and extension about ahigher fulcrum in the second half of stancelead to an overall positive contribution tostance propulsion. Shoulder joint move-ments cause the forward movement of thetrunk in the first third of the stance phase(Fig. 6). Decreasing humeral abduction indi-cates scapula rotation into a parasagittalplane. Then scapula retroversion takes overthe major part of stance propulsion (Fig. 6).

Intralimb coordination

Retroversion of forelimb segments(scapula, humerus, forearm) begins syncho-nously at touch down, in contrast to antever-sion (Fig. 7a,b), which is initiated by cranial

movement of the scapula before the extrem-ity lifts off. The forearm follows mostly at liftoff. The onset of humeral anteversion variesmore than that of other segments. It occursbefore, at, or after lift off with equal fre-quency (Fig. 7b).

Maximum angles at the shoulder andelbow joints coincide with touch down (Fig.7c). During the stance phase, the first flex-ion in both joints ends earlier in the elbowjoint than in the shoulder joint (Fig. 7d).Elbow extension begins in 76% of casesbefore the onset of shoulder extension. Thelatter ceases until late in the stance phase,when only a slight extension can be ob-served. The shoulder joint will not be ex-tended as long as the hand has not passedunderneath the elbow joint. Elbow exten-sion together with scapula retroversion ma-neuvers the humerus into a horizontal posi-tion.

A second flexion occurs in the elbow andshoulder joint at the end of the stance phase.Flexion of the shoulder joint always beginsbefore lift off, and that of the elbow jointbegins at lift off or shortly afterwards (Fig.7e). The shoulder joint is flexed simulta-neously with the onset of scapula antever-sion. Therefore, the horizontal orientation ofthe humerus remains. Flexion of the elbowjoint is synchronized with anteversion of theforearm.

Timing of shoulder and elbow joint move-ments is different during swing (Fig. 7f ).During the first quarter of the swing phase,anteversion of the humerus and forearm ismainly realized by anteversion of thescapula. Additionally, the shoulder jointopens gradually, while the elbow joint keepsa more or less constant angle (Fig. 5). Up tothe last quarter of swing, anteversion of theforearm is achieved only by the displace-ment of scapula and humerus. The elbowjoint only extends at the end of the swingphase, when the humeral angle becomesgreater than 90°. An extension of the elbowjoint would counteract the forward move-ment of the limb as long as the distal end ofthe humerus is directed caudad.

DISCUSSION

On a treadmill, animals perform only apart of their locomotion repertoire. Tread-

Fig. 6. Contribution of scapula anteversion and fore-limb joint movement to stance length. The integral ofeach graph gives the absolute contribution of each jointto stance propulsion. The addition of all integrals givesthe stance length. Each point within the graph indicatesthe instantaneous contribution at a distinct time duringstance.


mill locomotion or restrained locomotion canbe different from that on normal ground(unrestrained locomotion), as described fortree shrews (Schilling and Fischer, 1999),horses (Barrey et al., 1993), and humans(Elliott and Blanksby, 1976). Our observa-tions on unrestrained locomotion are in

agreement with the reports on tree shrewsand horses. The brown lemur shows usuallyhigher stride lengths and lower step frequen-cies during treadmill locomotion as com-pared to unrestrained locomotion. However,only locomotion on the treadmill allows us toanalyze kinematics properly using cineradi-

Fig. 7. Statistics of intralimb coordination of fore-limb segments (a,b) and timing of shoulder and elbowjoint (c–f ) a: Onset of retroversion. b: Onset of ante-version. c: Onset of flexion I. d: Onset of extension I.e: Onset of flexion II. f: Onset of extension II. Note that

the retroversion of the segments is exactly synchronizedand starts at touch down, just like the beginning of thefirst flexion in shoulder and elbow joint. The followingjoint and segment movements are not synchronized.


ography, and especially to record series ofstrides.

Footfall pattern and asymmetricalbehavior of forelimbs

The brown lemur (Eulemur fulvus) per-forms a diagonal-sequence, diagonal-cou-plets walk which is typical for most quadru-pedal prosimians and monkeys (Hildebrand,1967; Vilensky, 1989). Hildebrand (1967)reported that primates with such a gaitpattern may show asymmetrical behavior,particularly of the forelimbs, caused by anunequal lateral bending of the spinal col-umn. We also found a different behavior ofthe left and right forelimbs in the malebrown lemur in metric and kinematic param-eters. Radiographic and orthopedic examina-tion proved that the animal was in goodhealth and showed absolutely no signs ofany pathological affection of the spine orextremities.

When the animal walks on a branch or arope, there is an undulating motion of thespine from side to side, which is achieved byan alternating lateral flexion of the spine.Such a lateral bending may help the indi-vidual to maintain balance (our personalobservation); this occurs not only in lemursbut also in other arboreal-quadrupedal pri-mates, such as the cotton-top tamarin(Saguinus oedipus) and the mouse lemur(Microcebus murinus) when walking onsmaller branches. Demes et al. (1990) alsoreported an extensive lateral flexion of thevertebral column in lorises. The male brownlemur shows asymmetrical bending, withstronger flexion of the left side of the spinalcolumn. This has probably developed in cor-relation with its distinct left-hand prefer-ence in nonlocomotory movements.

Metric gait parameters

In symmetrical gaits, there is an increaseof stride frequency, and often also an in-crease of stride length with higher speeds inprimates (Alexander and Maloiy, 1984; Vilen-sky and Gehlsen, 1984; Reynolds, 1987;Demes et al., 1990) and in nonprimates(Arshavskii et al., 1965; Goslow et al., 1973;Heglund and Taylor, 1988; Fischer, 1998;Fischer and Lehmann, 1998; Schilling andFischer, 1999). Only the left forelimb of the

brown lemur shows a significant speed-related increase in stride length.

The increase of stride frequency is real-ized by decreasing stance and swing dura-tion. This is also the case in slender and slowlorises (Demes et al., 1990). In rhesus mon-keys (Vilensky and Gehlsen, 1984) and innonprimates (Arshavskii et al., 1965; Gos-low et al., 1973; Boczek-Funcke et al., 1996;Fischer, 1998; Schilling and Fischer, 1999),the increase of step frequency is achievedmainly by decreasing stance duration,whereas swing duration is not affected byspeed.

KinematicsMovement of the proximal forelimb.The scapula of the brown lemur has a totalretroversion of about 48°. This is markedlymore than the 28° reported for the vervetmonkey (Whitehead and Larson, 1994). Attouch down the scapular angle is 55° in thevervet monkey, but about 45° in the brownlemur. Maximum retroversion of the scapulaof vervet monkeys is also restricted to 85°,whereas the lemur scapula exceeds thisvalue with maximum angles of approxi-mately 95°. However, Whitehead and Lar-son (1994) reported a distinct scapula trans-lation along the thoracic wall, which we didnot observe in the brown lemur. Usually,such a translation occurs in aclaviculatetherian mammals (English, 1978; Boczek-Funcke et al., 1996; Fischer, 1994), but notin claviculate therians (Jenkins, 1974b;Schilling and Fischer, 1999).

The range of scapula displacement in thebrown lemur is in agreement with the re-sults from small to medium-sized nonpri-mates. Schilling and Fischer (1999) reportedfor the tree shrew (Tupaia glis) a maximumamplitude of 64° for scapula retroversion insymmetrical gaits during continuous locomo-tion on a treadmill. The maximum ampli-tude given by Jenkins (1974a) for the treeshrew in exploratory walk, however, rangesonly between 30°–40°. A maximum ampli-tude of 40°–50° is reported for the opossum(Jenkins, 1971; Jenkins and Weijs, 1979),56° for the cui (Fischer, 1999), 40° for the cat(English, 1978; Boczek-Funcke et al., 1996),45° for the hyrax (Fischer, 1994, 1998), andspeed-dependent 33° or 42° in the pika


(Fischer and Lehmann, 1998). Jenkins(1974b) revealed a maximum scapular retro-version of 30°–40° for the rat during explor-atory walk, but Fischer (1999) observed amaximum amplitude of 55° for rats at nor-mal walking speeds.

In the brown lemur, abduction of thescapula relative to the parasagittal planereaches its maximum (25°–30°) at the begin-ning of stance phase. The scapula is an-terodorsally orientated at touch down androtates into a parasagittal plane at the endof stance phase. The course of the rat’sscapula follows the same scheme, but theangles were not quantified (Jenkins, 1974b).In the cat, abduction was determined at12°–17° using biplanar pulsed X-ray cinema-tography (Boczek-Funcke et al., 1996). Thescapula of the cat remains abducted through-out the entire stride cycle.

Scapula rotation about its longitudinalaxis, which is the third component of scapulamovement in the brown lemur, was reportedpreviously only for the cat (Boczek-Funckeet al., 1996). But the angle of median rota-tion in the cat amounts only to 4°, whereasin the brown lemur this angle normallyexceeds 60°. We pointed out that this me-dian rotation results in humeral abductionin the brown lemur, of more than 50° at thebeginning of stance phase. An abducted hu-merus was also observed in other small andmedium-sized mammals (Jenkins, 1971: treeshrew, oppossum, rat, hamster, and ferret).In these animals the abduction angle of thehumerus ranges between 15°–30°. Humeralabduction in all these mammals is mostprobably caused by the same mechanism asin the lemur. In the cat, humeral flexionaccounts for the slightly higher value forhumeral abduction (10°) as compared toscapular median rotation (4°), but not toactive abduction in the shoulder joint. Flex-ion in the shoulder joint together with amedially turned scapula leads to humeralabduction. In semiterrestrial vervet mon-keys no humeral abduction was observed,and the scapula rotated only in a parasagit-tal plane (Whitehead and Larson, 1994). Inspider monkeys (Jenkins et al., 1978), thescapula lies on the dorsal aspect of thethorax during the whole step cycle andnever assumes a parasagittal position. In

contrast to quadrupedal primates, these bra-chiators show a higher degree of shoulderjoint movements in all dimensions. In hu-mans the scapula is also involved in hu-meral abduction in the transverse plane, buthere most of this abduction movement oc-curs in the shoulder joint (Flecker, 1929;Freedman and Munro, 1966; Bagg and For-rest, 1986, 1988).

Compared to other small and medium-sized mammals, the shoulder joint of pri-mates is more extended at touch down. Thejoint angle amounts to more than 160° in thebrown lemur and vervet monkey (White-head and Larson, 1994) at the end of swing,but reaches only less than 90° in small andmedium-sized nonprimates (Jenkins, 1971,1974a,b; English, 1978; Fischer, 1994; Schill-ing and Fischer, 1999). The relatively longhumerus and forearm in primates are, thus,effectively converted into stride length. Alex-ander and Maloiy (1984) suggest that adap-tation for leaping would predispose pri-mates to use longer strides during walkingon the ground. But it is also plausible thatfor arboreal locomotion longer strides aremore advantageous than higher stride fre-quencies at moderate walking speeds, be-cause a stable foot position can be usedlonger for propulsion. Higher stride frequen-cies would increase the risk of a false stride.Demes et al. (1990) pointed out that highfrequency gaits are disadvantageous in arbo-real locomotion because they produce swing-ing movements of the branches that are notonly dangerous but also energy-consuming.

Vertical orientation of the scapula and thehorizontal orientation of the humerus at theend of stance phase occur in the brownlemur and in all small therian mammals sofar studied (Jenkins, 1971, 1974a,b; En-glish, 1978; Fischer, 1994, 1998; Fischer andLehmann, 1998; Schilling and Fischer, 1999),except the vervet monkey (Whitehead andLarson, 1994), in which the humerus angleat lift off is about 60° below the horizontalplane.

Kinematics of the distal forelimb. Simi-lar to the shoulder joint, the elbow joint ofthe brown lemur also shows an unusuallyhigh angle of about 150° at touch down. Inslow lorises, the elbow joint is completely


extended and the hand touches very farahead (Jouffroy and Stern, 1990). The elbowjoint of small nonprimates is much moreflexed at touch down (80°–90°) (Jenkins,1971; Caliebe et al., 1991; Fischer, 1994;Schilling and Fischer, 1999). In the brownlemur, the elbow joint angle is also moreextended at lift off than in nonprimates withcomparable body size, but the angle at touchdown is still greater, and therefore effectiveangular movement is negative. The maxi-mum amplitude of elbow joint movement ismore than twice as much as effective angu-lar movement during the stance phase, andindicates the function of the elbow joint asan adjusting (‘‘fine-tuning’’) and stabilizingjoint. This is concurrent with observations ofthe hyrax (Fischer, 1994, 1998), pika (Fischerand Lehmann, 1998), tree shrew (Schillingand Fischer, 1999), rat, and cui (Fischer,1999).

Contribution of segment displacementto stance length. The calculation of therespective contribution of segment displace-ments to stance propulsion results in a cleardivision between proximal and distal partsof the forelimb. The scapular (63%) andhumeral displacements (left, 31%; right,23%) are almost solely responsible for propul-sion. The elbow joint and wrist joint contrib-ute only 6% (left forelimb) or 18% (rightforelimb) to propulsion. Data on other mam-mals are only available for the pika (Fischerand Lehmann, 1998), the tree shrew (Schill-ing and Fischer, 1999), the rat, and the cui(Fischer, 1999). In the pika, the values forcontribution to propulsion are: scapula,66.5%; shoulder joint, 23%; elbow joint, 7.5%;and wrist joint, 3%; for the tree shrew insymmetrical gaits: scapula, 42%; shoulderjoint, 17%; elbow joint, 32%; and wrist joint,9%; and for the rat: scapula, 52–57%; shoul-der joint, 14–21%; elbow joint, 18–25%; andwrist joint, 2–9%. The contribution of thewrist joint to stance propulsion can be ne-glected in the brown lemur and is even lessthan the already low values known for mam-mals without prehensile hands. Angularmovements of the elbow and wrist jointsserve mainly to compensate for vertical oscil-lations of the center of gravity caused byextrinsic factors. They are thus mostly non-

propulsive but ‘‘fine-tuning’’ joints (Fischer,1994; Fischer and Witte, 1998). As describedby Lemelin and Schmitt (1998), the long axisof the hand in the brown lemur is notoriented in movement direction. Hence, mostof the length of this segment cannot contrib-ute to stride length.

CONCLUSIONS

In arboreal quadrupedal locomotion, thesupport is usually very small compared tothe size of the animal. More complex three-dimensional limb movements of an arborealprimate are achieved by a high degree ofscapular mobility. It consists of ante-/retro-version, adduction/abduction, and scapularrotation about the longitudinal axis. Medianrotation of the scapula, together with aflexion in the shoulder joint, mediates theabduction of the humerus, which is notachieved by a laterad segment displacementin the shoulder joint. This mode of humeralabduction is strikingly different from hu-meral abduction in humans. A restriction ofscapular ante-/retroversion in favor of anincreasing range of humeral movement, asobserved for the vervet monkey (Whiteheadand Larson, 1994), is not a characteristicfeature of primate locomotion in general. Inthe brown lemur, the length of humerus andforearm is converted into stride length by amore extended shoulder and elbow jointangle at touch down as compared to nonpri-mates.

The calculation of the contribution of seg-ment displacement to stance propulsionshows that scapular retroversion in a ful-crum near the vertebral border causes morethan 60% of the propulsion. Forelimb jointsare flexed during locomotion. An operationaldivision between the proximal and the distalforelimb segments, as has been observed forother small mammals, may also be recog-nized in the brown lemur.

Movements of most forelimb joints andsegments are speed-independent and influ-ence mainly linear gait parameters (stridelength, stance length). Only the effectiveangular movement of the elbow joint in-creases with increasing speed and stridefrequency.

The observed asymmetries of the left andright forelimb kinematics in the male brown


lemur show that hand preferences, whichare distinct in nonlocomotory actions, canalso have an effect on locomotion.

ACKNOWLEDGMENTS

We thank Dr. D. Haarhaus and his staff(Institut fur den Wissenschafflichen Film,Gottingen) for their patience and their com-petence in cineradiography. We are indebtedto Prof. Dr. H.G. Erkert (Zoologisches Insti-tut, Universitat Tubingen) and to Dr. U.Rumpler (Zoo Koln) for kindly providing uswith the brown lemurs. We thank all themembers of our working group in Jena formany discussions, and especially Dr.A. Haas,PD, Dr. J.M. Starck, and PD, Dipl. Ing. Dr.med. H. Witte (all at Jena), as well as thereviewers, for thoroughly revising the manu-script. M. Roser (Jena) skillfully helped withthe illustrations.

APPENDIXDefinitions

Touch down/lift off of footTouch down is defined as the instant inwhich the limb takes up the trunk load(5hard contact). Lift off is defined as thelift off of the metacarpus and the loosen-ing of the grip around the rope of thetreadmill.Metric gait parametersStride cycle: Period from touch down tothe next touch down of the same limb; onlystance and swing phases are distinguished.Stride duration (sec): Time between twosuccessive touch downs of the same limb.Stride length (m): Under treadmill condi-tions, stride length is calculated as thesum of stance and swing lengths becausethe trunk is not moved.Speed (m/sec): Quotient of stride lengthand step duration.Stance duration (sec): Time between touchdown and lift off of a foot.Stance length (m): Horizontal distancewhich is traveled by the fingertips on thetreadmill during stance phase. Stancelength corresponds to stance propulsion ofthe trunk during unrestrained locomo-tion.Swing duration (sec): Time between lift offand touch down.

Swing length (m): Horizontal distancewhich is traveled from the fingertips dur-ing swing phase.KinematicsAngles of segments are given between thelongitudinal axis of segment and the hori-zontal plane. Joint angles were measuredat the flexion side of joints (Fig. 1b).Maximum amplitude: Absolute value ofthe difference between the maximum andthe minimum joint angle during stance orswing.Effective angular movement: Absolutevalue of the difference between joint anglesor segment angles at touch down and liftoff.Abduction/adduction: In human anatomy,the terms ‘‘abduction’’ and ‘‘adduction’’ aredefined as displacement of a limb segmentin the frontal plane. Likewise, the termsdescribe the positions which result fromsuch movements. Here, these terms areused differently, to describe all positions ofa limb segment in which a distal joint islocated laterally or medially to the moreproximal joint, as well as the movementsleading to these orientations.

LITERATURE CITED

Alexander RMN, Maloiy GMO. 1984. Stride lengths andstride frequencies of primates. J Zool (Lond) 202:577–582.

Arshavskii YI, Kots YM, Orlowsky GN, Rodionov IM,Shik ML. 1965. Biophysics of complex systems andmathematical models. Investigation of the biomechan-ics of running by the dog. Biophysics 10:737–746.

Ashton EH, Oxnard CE. 1964. Locomotor patterns inprimates. Proc Zool Soc Lond 142:49–66.

Bagg SD, Forrest WJ. 1986. Electromyographic study ofthe scapular rotators during arm abduction in thescapular plane. Am J Phys Med 65:111–124.

Bagg SD, Forrest WJ. 1988. A biomechanical analysis ofscapular rotation during arm abduction in the scapu-lar plane. Am J Phys Med 67:238–245.

Barrey E, Galloux P, Valette JP, Auvinet B, Wolter R.1993. Stride characteristics of overground versustreadmill locomotion. Acta Anat (Basel) 146:90–94.

Boczek-Funcke A, Kuhtz-Buschbeck JP, Illert M. 1996.Kinematic analysis of the cat shoulder girdle duringtreadmill locomotion: an X-ray study. Eur J Neurosci8:261–272.

Caliebe F, Haussler J, Hoffmann P, Illert M, Schirrma-cher J, Wiedemann E. 1991. Cat distal forelimb jointsand locomotion: an X-ray study. Eur J Neurosci 3:18–31.

Cartmill M. 1972. Arboreal adaptation and the origin ofthe order Primates. In: Tuttle RH, editor. Functionalvertebrate morphology. Cambridge, MA, and London:Belknap Press of Harvard University Press. p 97–122.

Demes B, Jungers WL, Nieschalk U. 1990. Size andspeed related aspects of quadrupedal walking inslender and slow lorises. In: Jouffroy FK, Stack MH,


Niemitz C. editors. Gravity, posture and locomotion inprimates. Florence: II Sedicesimo. p 175–197.

Elliott BC, Blanksby BA. 1976.Acinematographic analy-sis of overground and treadmill running by males andfemales. Med Sci Sports 8:84–87.

English AW. 1978. Functional analysis of the shouldergirdle of cats during locomotion. J Morphol 156:279–292.

Fischer MS. 1994. Crouched posture and high fulcrum.A principle in the locomotion of small mammals: theexample of the rock hyrax (Procavia capensis) (Mam-malia: Hyracoidea). J Hum Evol 26:501–524.

Fischer MS. 1998. Die Lokomotion von Procavia capen-sis (Mammalia: Hyracoidea). Ein Beitrag zur Evolu-tion des Bewegungssystems der Saugetiere. VerhNaturwiss Verein Hamburg 32:1–207.

Fischer MS. 1999. Kinematics, EMG, and inverse dynam-ics of the therian forelimb—a synthetic approach. ZoolAnz 238:41–54.

Fischer MS, Lehmann R. 1998.Application of cineradiog-raphy for the metric and kinematic study of inphasegaits during locomotion of the pika (Ochotona rufe-scens, Mammalia: Lagomorpha). Zoology 101:148–173.

Fischer MS, Witte H. 1998. The functional morphologyof the three-segmented limb of mammals and itsspecialties in small and medium-sized mammals. ProcEur Mech Coll Euromech 375:10–17.

Flecker H. 1929. Roentgenographic study of movementsof abduction at normal shoulder joint. Med J Aust2:122–128.

Freedman L, Munro R. 1966. Abduction of the arm inthe scapular plane: scapular and glenohumeral move-ments. J Bone Joint Surg (Am) 48:1503–1510.

Goslow GE Jr, Reinking RM, Stuart DG. 1973. The catstep cycle: hind limb joint angles and muscle lengthsduring unrestrained locomotion. J Exp Biol 94:15–42.

Goslow GE Jr, Seeherman HJ, Taylor CR, McCutchinMN, Heglund NC. 1981. Electrical activity and rela-tive length changes of dog limb muscles as a functionof speed and gait. J Exp Biol 94:15–42.

Heglund NC, Taylor CR. 1988. Speed, stride frequencyand energy cost per stride: how do they change withbody size and gait. J Exp Biol 138:301–318.

Hildebrand M. 1967. Symmetrical gaits of primates. AmJ Phys Anthropol 26:119–130.

Jenkins FA Jr. 1971. Limb posture and locomotion in theVirginia opossum (Didelphis marsupialis) and in othernon-cursorial mammals. J Zool (Lond) 165:303–315.

Jenkins FA Jr. 1974a. Tree shrew locomotion and theorigin of primate arborealism. In: Jenkins FA Jr,editor. Primate locomotion. New York:Academic Press,p 85–115.

Jenkins FA Jr. 1974b. The movement of the shoulder in

claviculate and aclaviculate mammals. J Morphol144:71–84.

Jenkins FA Jr, Weijs WA. 1979. The functional anatomyof the shoulder in the Virginia opossum (Didelphisvirginiana). J Zool (Lond) 188:379–410.

Jenkins FA Jr, Dombrowski PJ, Gordon EP. 1978. Analy-sis of the shoulder in brachiating spider monkeys. AmJ Phys Anthropol 48:65–76.

Jouffroy FK, Stern JT Jr. 1990. Telemetered EMG studyof the antigravity versus propulsive actions of kneeand elbow muscles in the slow loris. In: Jouffroy FK,Stack MH, Niemitz C, editors. Gravity, posture andlocomotion in primates. Florence: II Sedicesimo. p221–236.

Larson CF, Dodson DL, Ward JP. 1989. Hand prefer-ences and whole (Galago senegalensis). Brain BehavEvol 33:261–267.

Lemelin P, Schmitt D. 1998. The relation between handmorphology and quadrupedalism in primates. Am JPhys Anthropol 105:185–197.

Miller S, van der Meche FGA. 1975. Movements of theforelimbs of the cat during stepping on a treadmill.Brain Res 91:255–270.

Napier JR. 1967. Evolutionary aspects of primate loco-motion. Am J Phys Anthropol 27:333–342.

Reynolds TR. 1987. Stride length and its determinantsin human, early hominids, primates, and mammals.Am J Phys Anthropol 72:101–115.

Rose MD. 1973. Quadrupedalism in primates. Primates14:337–357.

Schilling N, Fischer MS. 1999. Kinematic analysis oftreadmill locomotion of Tupaia glis (Scandentia: Tupai-idae). Z Saugetierkd 64:129–153.

Tokuriki M. 1973. Electromyographic and joint-mechani-cal studies in quadrupedal locomotion. I. Walk. Nip-pon Juigaku Zasshi (Jpn J Vet Sci) 35:433–446.

Vilensky JA. 1987. Locomotor behavior and control inhumans and non-human primates: comparisons withcats and dogs. Neurosci Biobehav Rev 11:263–274.

Vilensky JA. 1989. Primate quadrupedalism: how andwhy does it differ from that of typical quadrupeds?Brain Behav Evol 34:357–364.

Vilensky JA, Gehlsen G. 1984. Temporal gait param-eters in humans and quadrupeds: how do they changewith speed? J Hum Mov Stud 10:175–188.

Walker AC. 1974. Locomotor adaptation in the past andpresent prosimian primates. In: Jenkins FA Jr, editor.Primate locomotion. New York: Academic Press, p349–381.

Whitehead PF, Larson SG. 1994. Shoulder motion dur-ing quadrupedal walking in Cercopithecus aethiops:integration of cineradiographic and electromyographicdata. J Hum Evol 26:525–544.


cineradiographic study of forelimb movements during quadrupedal walking in the brown lemur (eulemur...

Documents