Modified Sihler’s Technique for Studying the Distribution ofIntramuscular Nerve Branches in Mammalian Skeletal Muscle

JIE LIU,1 V. PREM KUMAR,1* YAN SHEN,2 HUI-KING LAU,1 BARRY P. PEREIRA,1AND ROBERTW.H. PHO1

1Department of Orthopaedic Surgery, The National University of Singapore2Department of Hand and Reconstructive Microsurgery, National University

Hospital, Singapore

ABSTRACT Background: A largely forgotten technique initially de-signed by Sihler for staining nerve tissue has not been fully explored forstaining intramuscular nerve branches in skeletal muscles.Methods: Fresh, long heads of triceps from locally bred New Zealand

white rabbits were used for this study. Immediately after their removal,the muscles with their motor nerve branches from the radial nerve werefixed in 10 unneutralized formalin, followed bymaceration and depigmen-tation in 3 aqueous potassium hydroxide, decalcification in Sihler’ssolution I, micro-dissection, staining in Sihler’s solution II, destaining inSihler’s solution I, neutralization in 0.05% lithium carbonate, and clear-ance in increasing concentrations of glycerin.Results: A clear three-dimensional orientation of the distribution of the

intramuscular nerve branches within the muscle belly was visualized. Itwas found in all specimens that the long head of triceps in the rabbit wasconstantly innervated by three main intramuscular nerve branches andeach of them supplied different amounts of muscle fibers with somevariation.Conclusion: The Sihler’s neural staining technique can be applied to the

study of the distribution of intramuscular nerve branches in limb skeletalmuscles. Extension of the techniquemay be utilised in the identification ofneuromuscular compartments in skeletal muscles. Such information maybe usefully applied in free muscle transfer of segments of skeletal muscle.Anat. Rec. 247:137–144 r 1997 Wiley-Liss, Inc.

Key words: Sihler’s staining technique, Intramuscular nerve branches,Rabbit skeletal muscle, Neuromuscular compartment

Information on the distribution of intramuscularnerve branches within limb skeletal muscles will en-hance the understanding of their anatomy and functionand will be useful to anatomists, physiologists, andperhaps clinicians. Anatomical dissection has been themost commonly used method, but it has been limited bythe inability to trace the nerve fibers from the extramus-cular branches to the intramuscular terminal branches,as the latter are very fine and invisible even under thedissecting microscope (Markee et al., 1955; Ohtani,1979; Serlin and Schieber, 1993; Yamada, 1986). Fur-thermore, direct dissection would damage muscle fiberand disrupt the normal anatomical relationship be-tween muscle and nerve fibers (Homma and Sakai,1991, 1992). Computer reconstruction of serial histologi-cal sections, another alternative, is not only time-consuming but also inaccurate due to distortion duringtissue cutting, staining, orientation, and reconstruction(Liem and Van Willigen, 1988; Wu and Sanders, 1992;Ying et al., 1992).In the present study, a largely forgotten technique

initially designed by Charles Sihler in the last century(1895) and recently modified by Liem and Van Willigen(1988) and Wu and Sanders (1992) was used to investi-gate the distribution of the intramuscular nerve supplyin the long head of triceps of adult rabbits with the hopeof extending it to the study of limb skeletal muscles inprimates and perhaps humans.

MATERIALS AND METHODS

Eighteen long heads of triceps from locally bred adultNew Zealand white rabbits weighing 2.5–3.5 kg wereused for this study. The animals were deeply anesthe-tized with intramuscular injection of Hypnorm (Jans-sen, Oxford, UK) (0.4ml/kg body weight). The long headof triceps with its primary nerve branch from the radial

ReceivedApril 26, 1996; accepted July 30, 1996.*Correspondence to: Prof V.P. Kumar, Department of Orthopaedic

Surgery, National University of Singapore, 5 Lower Kent Ridge Road,Singapore 119260.Contract grant sponsor National University of Singapore; Contract

grant number RP910468/N.

THE ANATOMICAL RECORD 247:137–144 (1997)

r 1997 WILEY-LISS, INC.

nerve as well as all extramuscular nerve branches werecarefully dissected under a 323 magnification operat-ing microscope (Carl Zeiss, Germany). The muscle andits innervating nerves were harvested and immediatelyimmersed in 10%unneutralized formalin.At the comple-tion of the surgical procedure, the rabbit was overdosedwith sodium pentobarbitone. The surgical procedurewas performed in accordance with International Guid-ing Principles for Biomedical Research Involving Ani-mals (Howard-Jones, 1985).Based on Sihler’s neural staining technique modified

by Liem and Van Willigen (1988) and Wu and Sanders(1992), the following processes were carried out.

Fixation

The muscle was fixed in 10% unneutralized formalinfor at least three to four weeks before maceration. Themedium was changed if it became cloudy.

Maceration and Depigmentation

The fixed muscle was washed under running waterfor half an hour. It was then placed into a solution of 3%aqueous potassium hydroxide with 0.2 ml of 3% hydro-gen peroxide per 100 ml for maceration of the muscleand bleaching of its pigment. The solution was changedonce every one to two days depending on whether thesolution was cloudy. The duration of this step variedwith each muscle. The process was considered completewhen the muscle appeared transparent with the finenerve branches becoming visible under the operatingmicroscope. Usually, two to three weeks were requiredto reach this stage.

Decalcification

The macerated muscle was washed under runningwater for half an hour and decalcified in Sihler’ssolution I (one part glacial acetic acid, two partsglycerin and 12 parts 1% aqueous chloral hydrate) forabout two weeks. The solution was changed once aweek. In Sihler’s solution I, the macerated muscle lostits transparency and shrank. Decalcificationwas consid-ered to be at an end when the muscle became com-pletely transparent again.

Fig. 1. End result of the modified Sihler’s staining. The long head oftriceps muscle in the rabbit demonstrating three main intramuscularnerve branches each supplying different parts of the muscle. Theproximal branch (p) supplies about 40% of the muscle fibers while themiddle (m) and distal (d) branches supply about 30% of muscle fiberseach respectively. In the proximal part of the muscle, there is a ‘‘U’’-

shaped anastomotic nerve (arrow and detailed in Fig. 5) connectingtwo branches arising from the proximal main intramuscular nervebranch. In the middle of the muscle, there is a ‘‘Y’’-shaped anastomoticcommunicating nerve (arrowhead and detailed in Fig. 3) connectingtwo branches arising from the main proximal and middle intramuscu-lar nerve branches, respectively.

Abbreviations

d distal main intramuscular nerve branchm middle main intramuscular nerve branchp proximal main intramuscular nerve branchI insertion of muscleO origin of muscleR radial nerve

138 J. LIU ET AL.

Micro-Dissection

After decalcification, all tissues of the specimen be-came very soft. Under the operating microscope, allconnective tissue surrounding the surface of the musclewas easily removed, especially the fascia-like tissue,which impeded the transparency of the muscle.

Staining

The decalcified muscle was washed under runningwater again and transferred to Sihler’s solution II (onepart Ehrlich’s hematoxylin, two parts glycerin and 12parts 1% aqueous chloral hydrate). The solution waschanged once a week. The muscle was left in Sihler’ssolution II until the very fine intramuscular nervebranches were visibly stained. This step took between10 to 14 days.

Destaining

After staining, the muscle was placed into Sihler’ssolution I again to destain the muscle fibers. Theprocess was stopped when the stained finer nerve fibersbegan to fade. This usually took one to two hours. Thesolution was changed whenever it turned purple.

Neutralization

The destained muscle was washed under runningwater for half an hour and subsequently in 0.05%

lithium carbonate with gentle agitation three times tillthe nerves changed from purple to blue. This step tookone to two hours.

Clearing

This last step made the neutralized muscle fiberstransparent in increasing concentrations of glycerin(40%, 60%, 80%, and 100%). The muscle was kept ineach solution for three days, especially in the lowerconcentrations of glycerin.The whole process took approximately three months

to complete, after which the muscle was stored in adark place in 100% glycerin with a few thymol crystalsadded.During the various steps, a stereomicroscope with

403 magnification (Leica, Germany) was used to decidethe end point of the steps and to study the specimens.

RESULTS

The end result is an almost transparent muscle withits extra- and intra-muscular nerve branches staineddeep blue (Figs. 1, 2). Although the muscle expressedslight shrinkage, it still remained intact to demonstrateclearly the normal anatomical relationship betweenmuscle and nerve fibers.In all specimens, the long head of triceps was consis-

tently innervated by one nerve branch directly from the

Fig. 2. End result of the modified Sihler’s staining. The long head oftriceps muscle in rabbit (separate specimen) demonstrating the proxi-mal branch (p) which supplies about 60% of muscle fibers, the middle

(m) which supplies about 10% of muscle fibers and the distal (d), therest. The distribution pattern of the main intramuscular nervebranches is similar to the specimen in Figure 1.

139DISTRIBUTION OF INTRAMUSCULAR NERVE BRANCHES

radial nerve. Before its entry into the muscle, thebranch divided into three finer nerve branches. Thesebranches entered the muscle belly through the epimy-sium and formed three main intramuscular branches,each of which demonstrated a consistent distributionpattern (Figs. 1, 2). These branches then subdividedinto multiple smaller branches, some of which arosequite proximally from the main branches (Figs. 1, 2, 6).Each main intramuscular branch supplied a differentsegment of the muscle belly with slight variationsobserved in different specimens.The proximal main intramuscular branch was the

thickest (0.5 to 0.6 mm in diameter) and always sup-plied the largest segment of the muscle belly (40 to 70%by volume). This was an approximation and could bedetermined by viewing the entire muscle specimens inthree dimensions.Within the segment, the nerve branchsubdivided into three to five finer branches (0.2 to 0.3mm in diameter) which would branch further up toterminal intramuscular nerve branches (Figs. 1, 2).Some of the intramuscular branches arose very proxi-mally from the main proximal branch after it hadpenetrated the epimysium (Fig. 1).The middle main intramuscular branch was finer

than the proximal one and, sometimes, was the finest(0.2 to 0.3 mm) of the three main intramuscular nervebranches. In addition, it had a wide variation in the

volume of the muscle it supplied. In 13 out of the 18specimens, it innervated 20 to 30% of the entire musclebelly (Fig. 1). In the other five, it only gave branches to10% of the muscle belly in which the main branch (0.2mm in diameter) was finer than the branches (0.3 mmin diameter) arising from the proximal main intramus-cular branch (Fig. 2).The distal main intramuscular nerve branch was also

finer than the proximal one. However, it supplied amore consistent volume of muscle belly (20 to 30%)when compared to the middle main intramuscularbranch (Figs. 1, 2). As in the proximal branch, smallbranches also arose very proximally from the mainbranch after it had penetrated the epimysium (Figs. 1,2, 6).In all of the specimens, there were some communicat-

ing nerve branches within the muscles which could beclearly identified by the naked eye or stereomicroscope(Figs. 1, 3–5). The communications occurred betweenbranches from the same main intramuscular nervebranch (Figs. 1, 5) or between branches from differentmain intramuscular nerve branches (Figs. 1, 3). Commu-nicationswere also seen between two proximal branches(Figs. 1, 5) or between two distal branches (Figs. 1, 3).These communicating nerves were observed to run in a‘‘Y’’-shaped (Fig. 3), ‘‘I’’- shaped (Fig. 4) or ‘‘U’’-shapedpattern (Fig. 5). The pattern of communication was

Fig. 3.Higher magnification (340), showing ‘‘Y’’-shaped anastomotic communicating nerves (arrows).

140 J. LIU ET AL.

never consistent in the different specimens. We wereconfident that these were indeed communicating nervebranches between the intramuscular branches by thenatural way they passed from one to the other.In this study, a few terminal nerve branches (arrow-

heads in Fig. 6) were demonstrated to enter the tendons(arrow in Fig. 6) at the end of the muscle.

DISCUSSION

It has been 100 years since Charles Sihler firstintroduced his method for staining nerve spindles(Sihler, 1895) and more than 50 years since otherworkers reported modification of this technique foridentifying innervation in the kidney, uterus, and ovaryin humans and the foot and leg in animals (Wharton,1937; Williams, 1943). In the last 50 years, only a fewhave reported use of this method in research (Freihofer,1966; Liem and Van Willigen, 1988; Puzdrowski, 1987).A group from the Department of Otolaryngology, MountSinai Medical Center in New York recently used amodified Sihler’s neural staining technique to make adetailed and serial study of the innervation of themuscles surrounding the larynx in dogs and humans(Drake et al., 1993; Mu et al., 1994; Sanders et al., 1993,1994; Wu et al., 1992, 1994). We are not aware of any

report using this technique to study the distribution ofintramuscular nerve branches in skeletal muscles ofthe limb although Guyer (1936) had mentioned the useof the Sihler’s method combinedwith the potashmethodto identify nerve tissue in muscle. In the present study,the distribution of intramuscular nerve branches in thelong head of triceps in the rabbit was satisfactorilydemonstrated by using the modified Sihler’s neuralstaining technique. The actual mechanism of stainingof only the neural tissue by the technique has not beendescribed even in the original paper (Sihler, 1895).However, Liem and Van Willigen (1988) have shownthat myelin in neural tissue takes up hematoxylin.Our study is in agreement with previous reports

(Drake et al., 1993; Freihofer, 1966; Liem and VanWilligen, 1988; Wu and Sanders, 1992), that while theprocess is meticulous and tedious, the results are stillfar from perfect. Much experience and trial and error isstill required with this only available technique (to ourknowledge) for identifying intramuscular nerve branch-ing pattern in the whole muscle and for demonstratingthe innervation pattern in three dimensions. It is mostimportant in achieving a perfect result to accuratelydetermine the end point of each step during the elabo-rate process, especially in themaceration and depigmen-

Fig. 4. Higher magnification (340), showing a ‘‘I’’-shaped anastomotic nerve (arrow) connecting twonerve branches.

141DISTRIBUTION OF INTRAMUSCULAR NERVE BRANCHES

tation and in the decalcification steps. There is noguiding principle on when exactly each step should beterminated. Repeated practice and careful observationof the process in each step would be required to perfectthe technique.Although technically demanding and time-consum-

ing, the Sihler’s process is superior to dissection andcomputer reconstruction of serial histological sections(Liem and Van Willigen, 1988; Wu and Sanders, 1992)in demonstrating intramuscular nerve branching. Al-though as a rule, vessels follow nerves (Taylor et al.,

1994), the pattern is not always consistent and angiog-raphy alone can not substitute for a good nerve stainingtechnique in demonstrating intramuscular nervebranching uniformly in all muscles.The intramuscular branching pattern in the long

head of triceps in the rabbit reveals three separatesegments within the muscle belly and each in theterritory of one of the three main intramuscularbranches. This clarifies the concept of neuromuscularcompartment mooted by English and Letbetter (1982a)and others (Armstrong et al., 1988; Chanaud and

Fig. 5. Higher magnification (330), showing a ‘‘U’’-shaped anastomotic nerve (arrow) connecting twonerve branches arising from proximal main intramuscular nerve branch.

142 J. LIU ET AL.

Macpherson, 1991; DeRuiter et al., 1995; Drake et al.,1993; English and Letbetter, 1982b; English, 1984;English andWeeks, 1987; Hammond et al., 1989; Lau etal., 1995; McIntosh et al., 1985; Richmond et al., 1985;Sanders et al., 1994; Segal et al., 1991; Segal, 1992;Serlin and Schieber, 1993; Thomson et al., 1991; Wind-horst et al., 1989). In previous studies, neuromuscularcompartments were identified by tedious microdissec-tion (Segal et al., 1991; Segal, 1992; Serlin and Schieber,1993) and serial histological sections (DeRuiter et al.,1995; English and Letbetter, 1982a; English andWeeks,1987; Hammond et al., 1989; Thomson et al., 1991).In the rabbit model, the long head of triceps demon-

strated three neuromuscular compartments. Some pre-vious studies by us suggested that each compartment inthis muscle can be stimulated to function indepen-dently from the rest (Lau et al., 1995; Liu et al., 1995).This has raised the possibility that all skeletal musclescan be subdivided anatomically at a submuscular levelinto different anatomical and functional units. Thepractising surgeon may hence transfer functional sub-units of skeletal muscle with its nerve and blood supplyin free muscle transfers (Buncke et al., 1991; Dellonand Mackinnon, 1985; Manktelow et al., 1980; Mank-telow and Zuker, 1984; Taylor et al., 1994) or transferan intact muscle with splitting of its subunits for

separate functions in various reconstructive endeav-ours (Manktelow and Zuker, 1984; Terzis, 1989).The presence of communicating branches between

different main intramuscular nerve branches and be-tween subdivisions of each main intramuscular branchhad also been demonstrated by Drake et al. (1993) indogs and Sanders et al. (1994) in human laryngealmuscles. The role of these communicating branches isunclear. Similar anastomotic patterns have been ob-served in the blood supply to muscles in vascularstudies (Campbell and Pennefather, 1919; Mathes andNahai, 1981; Taylor et al., 1994). We can only postulatea collateral nerve supply function much akin to collat-eral blood supply to these communicating branches. Itmay be relevant to nerve regeneration and reinnerva-tion after nerve injury. A feedback pathway role fromdifferent compartments ormotor unitswithin themuscleis also postulated.As for the nerve branches demonstrated entering the

tendon, they most probably represent conduits forfeedback from postural and stretch receptors.In conclusion, the Sihler’s neural staining technique

has been successfully used to demonstrate intramuscu-lar nerve branches in the skeletal muscle in a smallanimal. Further work on larger skeletal muscles in thehuman or in equivalent primates is indicated so that

Fig. 6. Terminal nerve branches (arrowheads) entering the tendon (arrow) of the long head of triceps.

143DISTRIBUTION OF INTRAMUSCULAR NERVE BRANCHES

the technique can be truly useful to the anatomist, andphysiologist and the clinician dealing with patients.

ACKNOWLEDGMENTS

This work is supported by the grants from TheNational University of Singapore (RP950330), NationalMedical Research Council (RP910468/N), and ShawFoundation (GR05988N). The authors thankDrs. Sand-ers and Wu of the Department of Otolaryngology inMount Sinai Medical Center in New York for introduc-ing the Sihler’s staining methodology to our team. Theauthors also thank Associate Professor E.H. Yap andMrs. Christine Hu of the Animal Holding Unit for theuse of animals and Mr Tan Boon Kiat for the photogra-phy.

LITERATURE CITEDArmstrong, J.B., P.K. Rose, S. Vanner, G.J. Bakker, and F.J.R.

Richmond 1988 Compartmentalization of motor units in the catneck muscle, biventer cervicis. J. Neurophysiol., 60:30-45.

Buncke, H.J., T.M. Whitney, and R.G. Milliken 1991 Facial paralysis.In: Microsurgery Transplantation-Replantation. H.J. Buncke, ed.Lee and Febiger, Philadelphia, pp. 487-506.

Campbell, J. and C.M. Pennefather 1919 The blood-supply of muscles,with special reference to war surgery. Lancet, 1:294-296.

Chanaud, C.M., and J.M. Macpherson 1991 Functionally complexmuscles of the cat hindlimb: III. Differential activation withinbiceps femoris during postural perturbations. Exp. Brain Res.,85:271-280.

Dellon, A.L. and S.E. Mackinnon 1985 Segmentally innervated latissi-mus dorsi muscle: Microsurgical transfer for facial animation. J.Reconstr. Microsurg., 2:7-12.

DeRuiter, C.J., A. DeHaan, and A.J. Sargeant 1995 Physiologicalcharacteristics of two extreme muscle compartment in gastrocne-mius medialis of the anaesthetized rat. Acta Physiol. Scand., 153:313-324.

Drake III, W., Y. Li, M.A. Rothschild, B.L. Wu, H.F. Biller, and I.Sanders 1993 A technique for displaying the entire nerve branch-ing pattern of a whole muscle: Results in 10 canine posteriorcricoarytenoid muscles. Laryngoscope, 103:141-148.

English, A.W. and W.D. Letbetter 1982a Anatomy and innervationpattern of cat lateral gastrocnemius and plantaris muscles.Am. J.Anat., 164:67-77.

English, A.W. and W.D. Letbetter 1982b A histochemical analysis ofidentified compartments of cats lateral gastrocnemius muscle.Anat. Rec., 204:123-130.

English, A.W. 1984An electromyographic analysis of compartments ofcat lateral gastrocnemiusmuscle during unrestrained locomotion.J. Neurophysiol., 52:114-125.

English, A.W. and O.I. Weeks 1987 An anatomical and functionalanalysis of cat biceps femoris and semitendinosus muscles. JMorphol., 191:161-175.

Freihofer, W.C. 1966 The Sihler technique of staining nerves forsystematic study especially of fishes. Copeia, 3:470-475.

Guyer, M.F. 1936 Animal Micrology, 4th ed. University of ChicagoPress, Chicago, Illinois.

Hammond, C.G.M., D.G. Gordon, J.T. Fisher, and F.J.R. Richmond1989 Motor unit territories supplied by primary branches of thephrenic nerve. J. Appl. Physiol., 66:61-71.

Homma, T. and T. Sakai 1991 Ramification pattern of intermetacarpalbranches of the deep branch (ramus profundus) of the ulnar nervein the human hand. ActaAnat., 141:139-144.

Homma, T. and T. Sakai 1992 Thenar and hypothenar muscles andtheir innervation by the ulnar nerve and median nerves in thehuman hand. ActaAnat., 145:44-49.

Howard-Jones, N. 1985ACIOMS ethical code for animal experimenta-tion. WHO Chronicle, 39:51-56.

Lau, H.K., J. Liu, B.P. Pereira, V.P. Kumar, and R.W.H. Pho 1995Fatigue reduction by sequential stimulation of multiple motorpoints in a muscle. Clin. Orthop., 321:251-258.

Liem, R.S.B. and J.D. Van Willigen 1988 In toto staining and

preservation of peripheral nervous tissue. Stain Technol., 63:113-120.

Liu, J., H.K. Lau, W.X. Min, B.P. Pereira, V.P. Kumar, and R.W.H. Pho1995Contractile characteristics on electrical stimulation ofmusclewith multiple motor points: An in vivo study in rabbit. Clin.Orthop., 313:231-238.

Manktelow, R.T., N.H. Mckee, and T.Veltese 1980 Anatomical study ofpectoralis major muscle for functional muscle transplantation.Plast. Reconstr. Surg., 65:610-615.

Manktelow, R.T., and R.M. Zuker: 1984 Muscle transplantation byfascicular territory. Plast. Reconstr. Surg., 73:751-757.

Markee, J.E., J.T. Logue, M. Williams, W.B. Stanton, R.N. Wrenn, andL.B. Walker 1955 Two-joint muscles of the thigh. J. Bone JointSurg., 37A:125-142.

Marthes, S.J. and F. Nahai 1981 Classification of the vascularanatomy of muscle: Experimental and clinical correlation. Plast.Reconstr. Surg., 67:177-187.

McIntosh, J. S., M. Ringqvist, and E. M. Schmidt 1985 Fiber typecomposition of monkey forearmmuscle. Anat. Rec., 211:403-409.

Mu, L., I. Sanders, B.L. Wu, and H.F. Biller 1994 The intramuscularinnervation of the human interarytenoid muscle. Laryngoscope,104:33-39.

Ohtani, O. 1979 Structure of the flexor digitorum superficialis. Okaji-mas FoliaAnat. Jpn., 56:277-288.

Puzdrowski, R.L. 1987 The peripheral distribution and central projec-tions of the sensory rami of the facial nerve in goldfish, carassiusauratus. J. Comp. Neurol., 259:382-392.

Richmond, F.J.R., D.R.R. MacGillis, and D.A. Scott 1985 Muscle-fibercompartmentalization in cat splenius muscles. J. Neurophysiol.,53:868-885.

Sanders, I., B.L.Wu, L. Mu,Y. Li, andH.F. Biller 1993 The innervationof the human larynx. Arch. Otolaryngol. Head Neck Surg.,119:934-939.

Sanders, I., B.L. Wu, L. Mu, and H.F. Biller 1994 The innervation ofthe human posterior cricoarytenoid muscle: Evidence for at leasttwo neuromuscular compartments. Laryngoscope, 104:880-884.

Segal, R.L., S.L. Wolf, M.J. DeCamp, M.T. Chopp, and A.W. English1991 Anatomical partitioning of three multiarticular humanmuscles. ActaAnat., 142:261-266.

Segal, R.L. 1992 Neuromuscular compartments in the human bicepsbrachii muscle. Neurosci. Lett., 140:98-102.

Serlin, D.M. and M.H. Schieber 1993 Morphologic regions of themultitendoned extrinsic finger muscles in the monkey forearm.ActaAnat., 146:255-256.

Sihler, C. 1895 Ueber Muskelspindeln und intramuskulare nervenen-digungen bei schlangen und froschen. Arch. Mikros. Anat. En-twickl., 46:709-723.

Taylor, G.I., M.P. Gianoutsos, and S.F. Morris 1994 The neurovascularterritories of the skin and muscles: Anatomic study and clinicalimplications. Plast. Reconstr. Surg., 94:1-36.

Terzis, J.K. 1989 Pectoralis minor: A unique muscle for correction offacial palsy. Plast. Reconstr. Surg., 83:767-776.

Thomson D.B., S.H. Scott, and F.J.R. Richmond 1991 Neuromuscularorganization of feline anterior sartorius: I. Asymmetric distribu-tion of motor units. J. Morphol., 210:147-162.

Wharton, L.R. 1937 A technique for studying the innervation oforgans. Anat. Rec. 67(suppl. no. 3):469-475.

Windhorst U., T.M. Hamm, and D.G. Stuart 1989 On the function ofmuscle and reflex partitioning. Behav. Brain Sci., 12:629-681.

Williams, T.W. 1943 A technique for the gross differential staining ofperipheral nerves in cleared vertebrate tissue. Anat. Rec., 86:189-194.

Wu, B.L. and I. Sanders 1992A technique for demonstrating the nervesupply of whole larynges. Arch. Otolaryngol. Head Neck Surg.,118:822-827.

Wu, B.L., I. Sanders, L. Mu, and H.F. Biller 1994 The humancommunicating nerve. An extension of the external superiorlaryngeal nerve that innervates the vocal cord. Arch. Otolaryngol.Head Neck Surg., 120:1321-1328.

Yamada, T.K. 1986 Re-evaluation of the flexor digitorum superficialis.ActaAnat. Nippon., 61:283-298.

Ying, R., M. Zhang, Y. Han, G. Wang, K. Chen, and Y. Cai 1992Computer 3-diamensional reconstruction of intraglandular lymphvessels and ductal system of human submandibular gland. ActaAnat. 144:175-177.

144 J. LIU ET AL.