cutaneous reflexes in small muscles hand · motoneurones innervating the small muscles of the hand....

Journal of Neurology, Neurosurgery, and Psychiatry, 1973, 36, 960-977

Cutaneous reflexes in small muscles of the handM. R. CACCIA1, A. J. McCOMAS2, A. R. M. UPTON, AND T. BLOGG

From the Department of Medicine (Neurology), McMaster University Medical Centre,Hamilton 16, Ontario, Canada

SUMMARY A study has been made of the responses of motoneurones innervating small muscles ofthe hand to electrical and mechanical stimulation of the skin. Both excitatory and inhibitory effectscould be observed in the same muscle after a single stimulus to a given area of skin. The earliestexcitatory and inhibitory responses are probably mediated by group III and the smaller group II

afferent nerve fibres. A later inhibition results from activity in the larger group II fibres which are

connected to cutaneous mechanoreceptors, especially those in the tips of the fingers and thumb. Thislate inhibitory reflex may operate through the fusimotor system. The possible roles of these reflexesare discussed in relation to previous investigations in man and the cat.

In recent years there has been an increasinginterest in the reflex responses of humanmotoneurones to cutaneous stimuli. This interestis appropriate for aberrations in cutaneous reflexactivity underlie some of the most importantphysical signs in clinical neurology. Indeed,possibly as a consequence ofthe great significanceof the Babinski response, nearly all the studieson cutaneous reflex activity in man have beencarried out on the leg. In this last respect thepresent study is different for it has been concernedwith the arm and, in particular, with thosemotoneurones innervating the small muscles ofthe hand. One reason for this choice was theopportunity which it afforded for further analysisof the late waves evoked in contracting musclesby nerve stimulation (Upton et al., 1971).There is, however, a major experimentaladvantage in employing the arm rather thanthe leg for the study of cutaneous reflexes inman. This advantage depends on the fact thatparticularly strong reflex changes can be elicitedby stimulation of the fingers and thumb. Byusing conventional techniques for recordingfrom digital nerve fibres it is then possible todetermine the properties of the afferent impulsevolley responsible for the reflex effects. In thispaper it will be shown that, following a cutaneous1 NATO Reseach Fellow on leave from the C. Besta NeurologicalInstitute, Milan.2 Member of the Canadian MRC Group in DevelopmentalNeurobiology.

960

stimulus, a characteristic sequence of excitatoryand inhibitory responses can be detected in smallmuscles of the hand. The nature of theseresponses, and the factors influencing their sizes,are discussed.

METHODS

SUBJECTS Observations were made in 16 men andwomen aged between 25 and 40 years who were ingood health and had no history of neurologicaldisease. Some of these subjects were examined onmore than one occasion.

STIMULI For electrical stimulation of digital nervesthe electrodes consisted of two strips of silver foil,3 mm in diameter, which were coated with electrodejelly. The cathodal electrode was bent round theproximal phalanx, while the anodal one encircled themiddle phalanx. When only the finger tip was to bestimulated the electrodes were two 6 mm diametertin discs coated with jelly and fixed to the fingerpulp with collodion or adhesive tape. For stimulationat other sites a fold of skin was pinched up andgripped by two spring-loaded clip-on electrodes. Atall sites the stimuli were rectangular voltage pulses50 ,usec in duration and delivered either 1-75 secondsapart or else randomly. In some experimentsmechanical stimuli were used. They were appliedmanually using a probe which contained a gramo-phone cartridge; the output from this piezo-electricdevice triggered a master timing unit (DevicesDigitimer, Model 3290).

Protected by copyright.

on March 1, 2021 by guest.

http://jnnp.bmj.com

/J N

eurol Neurosurg P

sychiatry: first published as 10.1136/jnnp.36.6.960 on 1 Decem

ber 1973. Dow

nloaded from

http://jnnp.bmj.com/

Cutaneous reflexes in small muscles of the hand

RECORDING Recordings from muscles were madewith strips of silver foil, approximately 4 x 0-6 cm,which were coated with electrode jelly on their under-surfaces and fastened to the skin with adhesive tape.One strip served as the stigmatic electrode and wasplaced over the end-plate region of the muscle understudy. The reference electrode was fastened over themuscle tendon or, in the case of a small muscle of thehand, around the terminal phalanx of a digit. A thirdsilver strip was attached to the dorsum of the handand served as an earth. In the text reflex activity hasbeen attributed to certain muscles with the under-standing that, because of the electrode positions andthe nature of the voluntary contractions, thesemuscles were likely to have been the main, but notthe only, sources of the electromyographic (EMG)activity recorded. In some experiments a unipolarconcentric needle electrode (Disa 9013L0501) wasemployed to record the activity of single motor units.

In all subjects sensory nerve action potentialswere recorded from the median and ulnar nerves atthe wrist. The electrodes were chlorided silver discs,10mm in diameter, and mounted in a Perspex holderso that their centres were separated by 3 0 cm. Themuscle and nerve responses were fed through speciallydesigned low-noise amplifiers, using a frequencyresponse which was 3dB down at 2 Hz and 10 kHz.With a superimposition technique alone it waspossible to distinguish nerve action potentials assmall as 1 ,V. The potentials were displayed on astorage oscilloscope (Hewlett-Packard model 141B)and fed through a loudspeaker. In order to measurethe intensity and time-course of a change in moto-neurone excitability, potentials were rectified andthen entered into a signal analyser (Hewlett-Packardmodel 5480B with type 5486B and 5488A plug-inunits). The analyser was triggered by the Digitimerand was used to average the muscle responses to atleast 64 stimuli, unless stated otherwise in the text.During stimulation the subject was instructed eitherto relax the muscle under study completely or tocontract it. The intensity of contraction could becontrolled during the experiment by observing, onthe analyser, the vertical displacement of the tracedisplaying the rectified and averaged muscle activity.At the end of the experiment accurate determina-

tions of the amount of inhibition or excitation weremade by measuring the rectified EMG activity over25 msec epochs. The measurements were performedmanually with a planimeter which was used totrace out the voltage-time integrals on photographsof the averaged responses. The EMG integral wasthen expressed as a percentage of the integratedmaximal M wave of the muscle under study.

In the text all mean values have been given withtheir standard deviations.

RESULTS

EFFECTS OF DIGITAL NERVE STIMULATION In allthe subjects tested a search was made forexcitatory or inhibitory effects after stimulationof a digital nerve. The demonstration of in-hibition requires a special technique. Onemethod which has been used in the triceps suraemuscle for this purpose, that of H-reflex testing(for example, Gassel and Ott, 1970), is notapplicable to the upper limb where H-reflexes

' ;' t' *b~~~~~7Ai.!t,.@|

,;1. 'i"i,.-I . 'd, 1!I

E1~~~2Xjl~~~~~~E'; 9 -l%

r. . . ..

FiG. 1. Upper. Reflex responses evoked in abductorpollicis brevis muscle by electrical stimulation ofindexfinger. Five oscilloscope traces have been superimposed.Middle. Average of 32 rectified responses recordedduring continuation of the above experiment. Lower.Average of 64 responses. The onsets of the variouswaves have been identified (see text); arrow signalsstimulus. Smallest divisions on time scales represent10 msec.

961



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from


962

are not normally demonstrable. The methodwhich has been employed instead requires thatthe subject attempts to make a steady con-

traction; the occurrence of inhibition is signalledby a reduction in muscle activity. There are

various ways in which the muscle activity can

be recorded, of which the simplest is to super-

impose EMG traces on a storage oscilloscope(Fig. 1, upper). While this technique willdemonstrate the presence of relatively stronginhibition, it will not show up minor degrees nor

will it enable determinations to be made of theintensity of inhibition, if this is less than total,or of the onset of inhibitory effect. For thesereasons, an improved method of recording hasbeen employed in which the EMG activity isfirst rectified and then entered into a signalaverager (Gassell and Ott, 1970). In the absenceof a stimulus the EMG activity averaged from anumber of samples will approximate to a smoothhorizontal trace. The vertical displacement ofthis trace from the base-line will be proportionalto the intensity of the contraction. If a stimulusis delivered any inhibition ofalpha-motoneuroneswill cause a temporary reduction in EMGactivity. During this period the averaged recordwill approach the base-line and will reach it ifthe inhibition is complete. Conversely, anyexcitatory effects will cause the averaged trace torise above the initial level, before stimulation.With this technique all subjects could be

shown to have a response to digital nerve

stimulation, using an intensity which was

supramaximal for the fastest-conducting groupof nerve fibres. The most commonly observedresponse was a combination of inhibition andexcitation and a typical example is shown inFig. 1 (middle and lower). In this experimentresponses were recorded from the abductorpollicis brevis (APB) muscle and the indexfinger of the same hand was stimulated. It canbe seen that one inhibitory phase (I1) beganapproximately 58 msec after the stimulus hadbeen delivered and was succeeded by a weakexcitatory phase (E1) at 75 msec. A secondinhibitory phase (12) commenced at 95 msec andwas terminated by a strong excitatory phase (E2)at 133 msec. When a slow sweep of the analyserwas used several further alternating excitatoryand inhibitory phases could be distinguished

M. R. Caccia, A. J. McComas, A. R. M. Upton, and T. Blogg

Isotiv%P

FIG. 2. Observations in a subject who displayedprominent E2 responses in APB following stimulationof her index finger. The upper trace shows averagedEMG activity, while in the two middle traces thepotentials have been rectified beforehand (16 responsesin each instance). In these periods ofobservation therewas minimal background voluntary contraction. Thelowest trace portrays results obtained during a

contraction of intermediate intensity and the ampli-fication has been reduced by a factor of 5. Stimulusshown by arrow. Smallest divisions on time scalerepresent 10 msec.

before a smooth contraction was resumed (forexample, Fig. 8b).The onset of I1 wave-that is, the first

component in this evoked activity-varied from30-76 msec in the population tested and was

sufficiently early to suggest that it was reflexin origin and not consciously directed. Thissuggestion was tested by measuring the voluntaryreaction time to the same stimulus in threesubjects. For example, when the relaxed subjectwas told to move the thumb as soon as he felt thestimulus to his finger, the latencies varied from120 to 160 msec with a mean of 141 + 13 msec. If,however, the subject was performing a weakbackground contraction before the stimulus themean latency was reduced to 123 + 8 msec andthe minimum value to 115 msec. When thesubject was instructed to completely relax themuscle on perceiving the stimulus, the mean



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



voluntary response time was 160 + 24 msec. Forthis latter test it was necessary to use subjectswith incomplete 'reflex' inhibition, in whom a

voluntary component could be searched for. Thevarious reaction time measurements suggestedthat not only the I, wave but also the E1 and I2components were reflex in origin.The reflex responses described, those of well-

marked alternating excitation and inhibition,were found in 13 of the 16 subjects examined. Inthree of these subjects the initial excitatoryphase (E1) was so strong that a response couldbe elicited from the muscle at rest or during very

slight effort (Fig. 2); this latency varied from 50to 75 msec. In a further three subjects, however,neither excitation nor inhibition were prominent.On the basis of serial observations on foursubjects it was our impression that the cutaneousreflexes became more marked with repetition ofthe experiment and that some form of condition-ing might have taken place.

EFFECT OF VOLUNTARY CONTRACTION The degreeof voluntary contraction was an important factorin determining the size of the cutaneously-induced inhibitory reflex. This is illustrated inFig. 3 which shows the results of an experimentin which the strength of voluntary abduction ofthe thumb was varied while the intensity of theelectrical stimulus to the index finger was keptconstant. It can be seen that during relativelyweak effort the inhibition (I1 phase) commencedapproximately 40 msec after the stimulus wasdelivered. The maximal intensity of the inhibitionwas reached 95 msec after the stimulus, duringthe '2 phase; at this time there was almostcomplete suppression of volitional activityat about 118 msec. During stronger contraction

_ _ - T ---_

IFIG. 4. Determination of the fraction of the moto-neurone pool inhibited by cutaneous stimulation (seetext).

A ..

aA.5m'-'rpU,.(

AAk *1M

:6i V

FIG. 3. Effect of increasing voluntary o

APB muscle responses to electrical stinindex finger (stimuli maximal for grc

Results for experiments employing weakand maximum strength are shown in theand lower traces respectively. Smallestime scale represent 10 msec.

1, the inhibition was unchanged in latency butN"- finished sooner, at 111 msec. In order to study

the effect of maximal effort without the com-

lOIJV plication of fatigue the last experiment was

interrupted several times before the full numberof sweeps was achieved for averaging. Figure 3

iV'W . (lower) shows that, although the inhibition was. -:9 unchanged in latency, it now began to decline

earlier, at approximately 102 msec. It is alsoapparent from Fig. 3 that the intensity of the

2OOJV inhibition became less with increasing effort,since a greater proportion of the recruitedmotoneurones were able to continue firing. How-ever, it is perhaps more relevant to consider theintensity of inhibition in terms of the total alpha

contraction on motoneurone pool instead of the active fractioniulation of the only. If it is assumed that the entire motoneurone, intermfediate pool is recruited during a maximal contractionupper middle (Merton, 1954) and if differences in the sizes andd divisions on firing frequencies of motor units are ignored,

then the fraction inhibited will be i/m where i is

I-1

'A1!'

963



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



E~~~~~~~~~~~~~~~~~_ _ _ _ _ _ _ _ _ -

- . -\E~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

- -,_ _ _ _~~~~~~_ AJ ,

*' '- '' - ' ' \-%J <Z.51I/1msec _

FIG. 5. On the left, from above downwards, are shown the effects of increasing stimulus intensity on the reflexresponses. The stimuli were applied to the index finger and recordings were madefrom APB during contractionsofintermediate strength. The corresponding sensory nerve volleys are shown on the right and were recordedfromthe median nerve at the wrist. The smallest divisions on the time scale (left) represent 10 msec.

the maximum drop in the averaged trace during traction i was 0 30 mV; hence the fraction of thethe inhibitory period and m is the vertical dis- motoneurone pool inhibited was 030/0-55 = 055.placement of the averaged trace from the For a weaker voluntary effort the correspondingbaseline during a maximum contraction, before fraction cannot be determined, for the inhibitioninhibition (Fig. 4). could well affect motoneurones other than those

In the experiment under consideration m was actually employed in the contraction. Therefore0-55 mV and during maximal voluntary con- the only certain conclusion to be drawn from this

964



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



40 50 60708090100

Amplitudesnap (MuV)

Stimulus intensity (volts)FIG. 6. Effect of increasing stimulus strength on the sizes of the median sensory nerve action potential(snap) and the E1, E2, and I2 reflex responses. The reflex responses were measured as voltage-time integralsduring 25 msec epochs and expressed as percentages of the maximal integratedM wave. Note logarithmicscale of abscissa.

part of the experiment was that increasing effortcan shorten the period of inhibition but cannotinfluence its time of onset. Because ofthe effect ofvoluntary contraction on reflex activity all theexperiments to be described were performedusing contractions of intermediate intensity.

EFFECT OF STIMULUS STRENGTH ON REFLEX

ACTIVITY In all subjects the strength of theelectrical stimulus to the index finger was

systematically varied so as to alter the size of theimpulse volley reaching the cord. It was foundthat both the excitatory and inhibitory com-

ponents of the reflex were affected by thismanoeuvre. Figure 5 shows the compoundaction potential recorded from the fastest-conducting (group II) fibres of the median nerve

at the wrist after graded stimulation of the indexfinger in a subject with prominent reflex re-

sponses; the Figure also displays the averagedEMG response from the abductor pollicis brevis(APB) muscle.There is a suggestion that 12 inhibition was

already present (Fig. 5 top trace) when thestimulus was at the threshold for consciousappreciation, being felt on approximately halfthe occasions. At this low stimulus strength itwas not possible to record a nerve actionpotential, even with an averaging technique. Asthe stimulus was increased the inhibition becamemore intense. By measuring the EMG voltage-time integral over a 25 msec epoch it could bedemonstrated that the intensity of the inhibitionwas maximal when 65% of the fast conducting

Reflex response(%M) I

965



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



a.

OW"N\

I\~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

. I

0-i

lms 20OuV

FIG. 7. Effect of stimulus strength on reflex muscle responses (left) and sensory nerve potentials (right) in asubject with a low-threshold E1 wave (arrows, top trace). Recordings were made from the APB muscle andelectrical stimuli to index finger were increased progressively (from above downwards in Figure). Smallestdivisions on time trace represent 10 msec. See text.

sensory fibres were activated (Fig. 6). In theseexperiments, in which no special precautionswere taken to control limb temperature, themean maximal impulse conduction velocity fornerve fibres from the index finger was 50+0-7 m/sec. Figure 5 also shows that as thestimulus was raised, the latency of inhibitionshortened from 90 to 42 msec. Among the 16subjects tested the shortest latency observed was30 msec; the mean value for the group was35 + 4 msec. Inspection of Fig. 5 suggested that

this early inhibition was a discrete phaseseparated from the later, and low-threshold,inhibition by an excitatory period (E1). It is ofinterest that in this subject the median nerveaction potential grew no further as the stimulusstrength was increased and the early inhibitiondeveloped. Indeed, in most subjects tested thethreshold for I1 inhibition was substantiallygreater than that for the 12 type, though anexception is shown in Fig. 7.The effect of stimulus strength on excitatory

1001" --- 1-90%

A.k4qel

.,.

'Ai

0IAOONMI064 .pI.vJ.v '06 1,000te II v Ifloom r 114.,

966

a

A6'ov.^4. 16.UIP% VVOW"

& A.1.

El4

. 4i.v



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



(a) (b)FIG. 8. (a) Reflex responses in APB after mechanical stimulation of the tip of the index finger, displayed assuperimposed oscilloscope traces and as an average of the rectified activity. (b) Top trace shows averaged APBresponses to mechanical stimulation recorded during ensuing 500 msec. Alternating phases of activity can beclearly seen. Lower trace shows responses to electrical stimulation of digital nerves in same finger. Smallestdivisions on time trace represent 10 msec.

parts of the reflex was equally complex. The E2excitatory component, starting immediatelyafter the main inhibitory (12) phase, had arelatively low threshold and became moreprominent as the stimulus increased. Figure 6,which relates the size of E2 to stimulus strength,shows a plateau once approximately half of the'fast' sensory fibres had been excited. A furtherincrease in E2 was evoked when considerablylarger stimuli were employed which weresupramaximal for the 'fast' fibres (Fig. 6).The early excitatory component (E1) usually

had a higher threshold than E2 and did notappear until a substantial fraction of the fast-conducting axons had been activated (Figs 5 and6). This was not so in all subjects, however, forin a few the E1 response also had a low threshold(Fig. 7). In this last figure, as in Fig. 5, it can beseen that the I1 and E2 components continued todevelop as the stimulus intensity was increased,even though the sensory nerve action potential,recorded from group II fibres, was maximal.Indeed, augmentation of the stimulus mighteventually cause E1 to become completely

suppressed by the fusion of the I1 and 12 in-hibitory phases, as Fig. 7 shows. Finally, itshould be mentioned again that, in a smallproportion of subjects (three in our study), an E1response could be evoked by a stimulus maximalfor fast conducting fibres with little, if any, back-ground voluntary contraction being needed tobring the motoneurones to the firing level (Fig. 2).

NATURE AND SITE OF RECEPTORS FOR REFLEX Thefibres in the digital nerves arise largely fromreceptors in the skin and, to a lesser extent, fromdeeply situated receptors such as those in jointcapsules and periosteum. The digital nerves donot contain fibres from muscle spindles and it isunlikely that Golgi tendon organs make asignificant contribution, since these structuresare usually found at the junction of a musclebelly with its tendon.To obtain further information concerning the

origin of the fibres involved in the reflex activitydescribed above, two further experiments weredevised. In the first experiment the stimulus wasa mechanical one and consisted of a series of taps

967



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



(1969) for the cat pad. Therefore, in a secondtype ofexperiment, stimulus spread was restrictedby delivering electric shocks through small discelectrodes attached to either side of the tip of theindex finger. From Fig. 9 it can be seen that thefull complement of inhibitory and excitatoryreflex changes ensued. The corresponding mediannerve action potential indicates that the group IIfibres which had been excited by electricalstimulation of the finger tip comprised onlysome 4%0 of those present in the digital nerves atthe finger base.The conclusion from these two experiments is

that a large part of the reflex activity describedin this paper can be evoked by stimulation ofthose mechanoreceptors in the finger tip whichare connected to fast-conducting (group II)axons.

OPTIMAL STIMULATION ZONES In the previoussection the strong reflex responses elicited insmall hand muscles by stimulation of the fingertip were described. In a further set of experimentsthe effects of stimulating widely separated areasof skin were examined. In Fig. 10 the intensitiesof the corresponding '2 inhibitory phases havebeen expressed in terms of the depression ofvoluntary EMG activity during 25 msec epochs(see Methods section). Figure 10 shows that thegreatest inhibition was exerted on APBmotoneurones by electrical stimulation of thethumb and fingers and by stimulating skin overthe elbow (E) and ventral aspect of the forearm(F). In contrast, stimulation of skin in the axilla(A), chest wall (C), ear lobe (L), and the con-tralateral index finger (II, contra) were allrelatively ineffective in producing inhibition,even though the stimuli were large enough tocause pain.The influence of the zone of stimulation was

emphasized more strongly when the latencyrather than the intensity of inhibition wasmeasured. For example, when APB was studiedthe earliest inhibition was observed after stimula-tion of the thumb. Stimulation of the index,middle, ring, and little fingers evoked inhibitoryresponses with progressively longer latencies(Fig. 11). When inhibition of EMG activity inthe abductor digiti minimi (ADM) was studied,differences in latency were also observed follow-ing stimulation of the thumb and fingers. In

1msec

.150Mv

5yV1 msec

f KJV5OiLVFIG. 9. Reflex responses of APB muscle afterelectrical stimulation of tip of index finger or ofdigitalnerve fibres at finger base (upper and lower parts ofFigure respectively). The sensory nerve potentialsrecordedfrom the median nerve at the wrist are shownabove the corresponding averaged responses. Thesmallest divisions on the time scale represent 10 msec.

delivered to different regions of the index fingerby an observer using a blunt probe. In theabductor pollicis brevis, it was found that,although the earlier inhibitory and excitatoryphases, I1 and E1, were both absent the laterphases, 12 and E2, were clearly present (Fig. 8).Following E2, alternating inhibitory and ex-

citatory waves could also be distinguished(Fig. 8b).

This type of experiment not only showed thatmechanoreceptors were capable of elicitingsome features of the reflex, but it also providedsome information as to the site of the activatedreceptors. Thus, of the various regions tested onthe finger, the tip was by far the most powerfulsource of reflex activity. It was still possiblehowever, that a substantial part of the forceimparted to the finger tip during stimulationwould have been transmitted to receptors lying ata distance, in the manner described by Lynn

968

A. Ii; i

I -

I

I8

i%00-^.p

I9

k-k



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



Inhibition(%M wave)

4-

3-1

2-

1-

0I ffm IV F F E A L C

FIG. 10. Amount of I2 inhibition produced in APB muscle by strong electrical stimuli applied to varioussites. Roman numerals indicate digits of same hand; F, volar aspect offorearm; E, elbow; A, axilla;L, lobe of ipsilateral ear; C, chest wall; II contra, contralateral index finger.

Latency(ms)

60- r

FIG. 11. La,following elec(II), middle IOpen andfillekdigiti minimi i

vertical bar aldeviation.

contrast with the findings for APB, however, theshortest latency was produced by stimulation ofthe little finger and progressively later responsescame from stimulation of the ring, middle andindex fingers and thumb (Fig. 11). The results forADM resemble those for APB in that stimulationof skin over the forearm or upper arm alsoyielded inhibition; here again the responseswere very much smaller than those evoked bystimulation of the thumb and fingers.

TI DISTRIBUTION OF REFLEX EFFECTS The effect ofl* * single electrical shocks to various areas of skin

was studied on several different muscles. Thell_]]1 * *- powerful responses elicited in small muscles of[ I I I the hand by stimulation of the fingers and thumb

have already been described. It was found thattency of the eariliest inhibitory phase the same stimuli caused very much smaller

trical stimulation of the thumb (I), index efecs im proxi ma mp le r

(III), ring (IV), and little (V) fingers. effects in more proximal muscles. Examples ared columns show mean valuesfor abductor given in Fig. 12 in which the EMG responses ofand abductor pollicis brevis respectively; the biceps and triceps muscles are shown, to-bove each column indicates one standard gether with those of the extensors and long

flexor muscles of the fingers. The diminished

II (contra)

969



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



If A

.~~~~~~~~~~~~~~~~~~9~~~~~~~~~~~~\~~~~~~'t.1B

~~~M~~~~dWi~~~~~~A\NM4A~~~~~~~l

C

FIG. 12. Distribution of reflex effects in variousmuscles ofthe same subject after electrical stimulationof the ipsilateral index finger. From above downwards,the records were obtained from the abductor pollicisbrevis, abductor digiti minimi, flexor digitorummuscles, biceps, and triceps. The smallest divisions onthe time scale represent JO msec.

responses of proximal muscles were not a con-sequence of the remoteness of the stimulatedzone for weak effects were still encountered ifthe stimuli were applied to skin lying directlyover the muscle belly.

In addition, the reflex effects which have beenstudied could be shown to display a type of'local sign' phenomenon (see Hagbarth, 1952)in that the same stimulus produced differingeffects in muscles situated closely together butsubserving opposing functions. For example, inFig. 13 it can be seen that electrical stimulationof the thumb or index finger evoked stronger

FIG. 13. Ejfects of same stimuli on muscles withantagonistic actions. A. Stimulation of thumb,recordings from adductorpollicis (upper trace of pair)and abductor pollicis brevis (lower trace). B. Stimula-tion of index finger, recordings from adductor pollicis(upper trace) and abductor pollicis brevis (lowertrace). C. Stimulation of little finger, recordings fromadductor (4th palmar interosseous; upper trace) andabductor digiti minimi (lower trace). Smallestdivisions on the time scale represent 10 msec.

reflex effects in the short abductor of the thumbthan in the adductor muscle. Similarly, stimula-tion of the little finger induced more prominentresponses in the abductor of this digit than in thecorresponding adductor muscle.

il"AV#0.

970

p0



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



i~-#Wio !r~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

FIG. 14. Responses from the APBelectrical stimuli to index finger whicheither regularly (top trace) or irregularljSmallest divisions on time scale represe

TIMING OF STIMULI Throughout Iexperiments, it was invariably foiinhibitory and excitatory reflexeslarger in response to the first few stthe succeeding ones. The pherdiminishing responses with regulastimuli is known to characterize m(other reflexes and is termed habituwful examination of this phenomeno

200jpV

V.... . . .. . .. .

FIG. 15. Discharges of a single motoabductor pollicis brevis muscle occurritout a stimulus (upper) or after electricalthe index or little fingers (middle andFigure respectively). For each ofthe thriseveral oscilloscope traces have been supsome potentials have been retoucldivisions on the time scale represent 10

been reported recently by Dimitrijevic et al.(1972).

In the present study an attempt was made toassess the significance of habituation by contrast-ing, in the same subject, the responses to stimulirepeated every 1[75 seconds with those to stimulidelivered irregularly though with a similar mean

!t,a t interval. Figure 14 shows the results of such ans., experiment and it can be seen that both the

inhibitory and excitatory responses were con-. . . . siderably more marked after irregular stimula-

-WWEENDW~ tion.muscle afterwere delivered SINGLE MOTOR UNIT RESPONSES In a smally (lower trace). number of experiments observations were made'nt 10 msec. on single motor units. These experiments

provided further information concerning thepattern of motoneurone discharge after a

the series of cutaneous stimulus. Figure 15 shows theund that the responses of a single unit in the abductor polliciswere much brevis after stimulation of the index finger. It can

;imuli than to be seen that after the 12 inhibitory period is overnomenon of the same unit begins to discharge again at a timeirly repeated when the E2 response normally occurs (90-)st, if not all, 130 msec). So far, we have not seen evidence ofation; a care- units firing during the E2 period which had notbn in man has already been recruited during the background

voluntary contraction. This possibility cannot beexcluded, however, on the basis of the small

,i, number of single units investigated. Of greatinterest was the finding that some units might beinhibited by stimulation of one digit but notanother. Thus in Fig. 15 it can be seen that theunit was strongly inhibited by stimulation of theindex finger but much less affected when the

AAL shocks were delivered to the little finger instead.Although this observation was not pursued, itindicates that the inhibitory receptive fields forindividual motoneurones may be defined quite

s sharply.

EFFECTS OF LOCAL ANAESTHESIA ON REFLEX

. . .. ACTIVITY An attempt was made to study thepossible role of muscle spindles in the cutaneous

or unit in the reflexes by using local anaesthesia to blockng either with- fusimotor axons. In one subject with well markedstimulation of reflexes 20 ml. 0*5/% lignocaine hydrochlorideee experiments solution were injected intravenously after theeerimposed and arm had been rendered ischaemic by an arterialhed. Smallest occlusion cuff (Holmes, 1963; Thorne, 1964).msec. After 20 minutes the finger jerk was reduced to

OM=="

971

i i A II I



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



REFLEX SNAP

,,"AAi%'

,l j#H @s1X~~I -m.

CONTROLv * * *

I~~~~~~~~~

dp . @w@0@@@@@*@***%*.

....... .. ....... .. .. ..I

* * ..........................................

20 3 _8 .

OV-3ERECOVERY

-----~~~~~~~~~~~~~~~~~----------------_ _

0- - Isu ONI IouV

20-22' .

RECOVERY

FIG. 16. Effects oflocal anaesthesia (and ischaemia) on electrically-induced reflex responses in APB (left handcolumn), median nerve sensory action potential (middle column) and finger flexor jerk (right hand column).The fingers jerks were elicited by an observer who used a tendon hammer to deliver blows ofas uniform force as

possible to the semi-flexed fingers. The recordings were made about 20 minutes after the application of the cuffand injection ofanaesthetic, and at 0-3 and 20-22 minutes after release of the cuff. The smallest divisions on thetime scale represent 10 msec (left hand column), I msec (centre and right hand column). For theflexorjerk thevertical bar represents 100 [LVfor the lower record and 50 1iV for the upper three recordings in the column.Notice the slowing of impulse conduction in the two middle recordings of the centre column.

approximately 5000 of its initial value. At this

stage conduction in alpha motor axons appearedto be relatively unimpaired, since strongvoluntary contraction could still be made withmuscles of the forearm and hand. More signifi-cantly, there was little change in the amplitude ofthe median nerve sensory potential which could

be recorded at the wrist after finger stimulation.It was therefore probable that the diminution intendon reflexes resulted from a loss ofmechanicalsensitivity of the spindles and that this, in turn,signified blockage of impulse activity in fusi-motor axons. Of great interest, then, was thefinding that, at the time of tendon reflex de-

FLEXOR JERK

972

oupoev%.tm



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



pression, the '2 and E2 cutaneous reflexes werereduced in size (Fig. 16). A more striking associa-tion between the depression of tendon reflexesand late cutaneous reflexes was noticed as thearm recovered from the effects of ischaemia inthe first few minutes after release of the arterialocclusion cuff (Fig. 16). The delay in the re-appearance of tendon reflexes during this latterperiod, in spite of rapid recovery of voluntarypower, suggested that several minutes wererequired before the anaesthetic blockage passedoff. From Fig. 16 it can be seen that not only didselective anaesthesia appear to abolish '2 andE2 but that I, and E1 might also disappear. Fromconsiderations presented elsewhere (see Dis-cussion section) it is apparent that the latenciesof the I1 and E1 waves are too short for spindlesto have been involved in these responses. Insteadit seems possible that the relatively smallcutaneous sensory axons responsible for thesewaves may also have been blocked by localanaesthesia together with the fusimotor axons(see below).

DISCUSSION

The foundations of our knowledge of cutaneousreflexes in mammals were established bySherrington (1910), who made an extensive studyof the hind limb in the decerebrate cat. Sherring-ton showed that electrical stimulation of aperipheral nerve resulted in flexion of thestimulated limb and extension ofthe contralateralone. In the ipsilateral limb the excitation of alphamotoneurones to flexor muscles was associatedwith inhibition of those to extensors. Sub-sequently, Hagbarth (1952) found, by stimulat-ing areas of skin rather than nerve trunks, thatthe distribution of excitatory and inhibitoryeffects was more subtle and purposeful than hadbeen supposed. Hagbarth observed that extensormuscles in the cat could be excited by ipsilateralstimulation, provided the stimuli were applied tothe skin overlying the muscle belly; stimuli else-where in the limb produced inhibition. Thesituation for a flexor muscle was in contrast, forthe inhibitory field was small and centred overthe antagonist extensor muscle; the excitatoryfield was extensive and covered the remainder ofthe limb. The organization of these responsesmade it possible for the limb to be automatically

withdrawn from a noxious stimulus wherever thelatter was applied. In a later paper Hagbarth(1960) obtained evidence for similar reciprocalrelationships for excitatory and inhibitoryinfluences on flexor and extensor motoneuronesin man. As in the study on the cat, the observa-tions were made in the lower limb and the stimuliconsisted of strong repetitive electrical shocks. Itis of interest that when relatively gentle mechani-cal stimuli are used in animal preparations morediscrete reflex effects can be observed. Forexample, by pressing on the cat pad it is possibleto elicit an extensor thrust reflex (Sherrington,1903), while a tap on a plantar cushion evokesextension of the toes (Engberg, 1964; Egger andWall, 1971). The advantages of studyingcutaneous reflexes in man include not only theability to observe the responses of an intactnervous system but also the possiblity of search-ing for reflex modulations of ongoing voluntaryactivity. The present study differs in several waysfrom most previous ones in man. It has beenconcerned with the upper limb and especiallywith the responses of motoneurones innervatingsmall muscles of the hand. The range of stimuliincluded not only electric shocks of gradedintensity but also mechanical stimuli. In addition,it was possible to ascertain the properties of theactivated sensory nerve fibres from inspection ofthe orthodromically-conducted nerve actionpotentials.The study has shown that electrical stimulation

can induce both excitatory and inhibitory changesin motoneurones. In their time courses, thesechanges are remarkably similar to those describedby Gassel and Ott (1970) in triceps suraemotoneurones after stimulation of the dorsumof the foot. In the present study it was found,by altering the stimulus strength, that two typesof inhibition were involved. One type, designatedI1, had a high threshold and a short latency; theother type I2, had a low threshold and longlatency.

I1 INHIBITION In most subjects the 1, inhibitionwas mediated by fibres having higher thresholdsthan those responsible for the 12 type of in-hibition. While some of the 1, afferent fibrescontributed to the recorded median nervesensory action potential, others were too smallfor their impulse activity to be detected. Accord-

973



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



ing to Buchthal and Rosenfalck (1966), themedian nerve sensory action potential, recordedat the wrist with surface or coarse needleelectrodes, is composed of activity in nerve fibreswith diameters of 6-11 [; these fibres would fallinto the type II category of Lloyd and Chang(1948). It follows that the afferent nerve fibresmainly responsible for I1 inhibition must havehad diameters which were in the lower part ofthis range or even smaller-that is, group IIIfibres. If 6 V is assumed to be the diameter of thelargest sensory fibres contributing to I1 thenthese fibres would have impulse conductionvelocities of 36 m/sec (Hursh, 1939). In a subjectwith a distance between the base of the middlefinger and vertebra prominens of 85 cm, theafferent impulse conduction time would beabout 24 msec. Similarly, the conduction timefrom the cervical cord to the small muscles of thehand would be approximately 15 msec on thebasis of measurements of H-reflex latency madein these muscles (Upton et al., 1971). This valueof 15 msec is the maximum time which would beneeded for an impulse which had just beeninitiated in the axon hillock of the motoneurones,before the arrival of the inhibitory volley in thecord, to travel out to the muscle. It follows that15 msec is also the maximum (efferent) timebefore the I1 inhibitory phase can be detected inthe muscle. The total conduction time for theI1 inhibitory phase would be 39 msec, which isthe sum of the afferent and efferent values,together with any time occupied by transmissionin synaptic pathways within the neuraxis. Thefigure of 39 msec is close to the observed latencyof I1 inhibition in APB which had a mean valueof 35+4 msec after stimulation of the indexfinger in the 16 subjects examined. The onlyestimate ofthe conduction velocities ofcutaneousnociceptor fibres subserving the inhibitory reflexin man is that of Hagbarth (1960) who obtainedmaximal values of 33-40 m/sec. In the catRosenberg (1970) also found that it was thesmaller myelinated fibres which were responsiblefor the inhibitory postsynaptic potentials in ipsi-and contralateral extensor motoneurones. Healso deduced that only one interneurone wasinvolved in the ipsilateral reflex pathway withinthe spinal cord. The present observationsconcerning I1 inhibition are perfectly compatiblewith this being the part of the same process as

that studied by Hagbarth in man and Rosenbergin the cat. The observations of Hagbarth havebeen extended, in that this type of inhibition hasbeen shown to involve motoneurones innervatingsmall muscles of the hand as well as thosesupplying larger and more proximal muscles inthe lower limb.

12 INHIBITION The 12 inhibitory phase was themost marked of the cutaneous reflexes observedin the present study. In some respects it was alsothe most intriguing reflex, since its low thresholdcontrasted with its long latency. We are aware ofonly one previous description of this type ofinhibition after digital stimulation, that recentlygiven by Liberson (1973). Although Libersonrecognized the significance of this inhibition inthe interpretation of silent periods induced byelectrical stimulation of 'mixed' nerve trunk, hisstudy did not contain an analysis of the receptorsand afferent fibres responsible for this reflex or ofthe possible synaptic pathways involved. Thepresent study has provided evidence to show thatthe reflex may be initiated by cutaneousmechanoreceptors and that the tips of the fingersand thumb are particularly sensitive regions.From the morphological study of Quilliam andRidley (1971) several types of receptors wouldemerge as candidates for the reflex-Merkel'sdiscs, Meissner's corpuscles, Pacinian corpuscles,and free nerve endings. The median nerve actionpotential recordings showed that the receptorswere connected to the fastest conducting fibresand that, in a sensitive subject, the reflex couldoccur when less than 400 of these fibres wereexcited in a digit. According to Buchthal andRosenfalck (1966), each of the two digitalbranches in a finger contains about 1,200 fibresat the level of the second phalanx, althoughrather more than one-third of these fibres arebranches from other axons. In any event, itwould appear that the 12 inhibitory reflex can beelicited in some subjects when as few as 50mechanoreceptor fibres have been activated.

Since the stimuli capable of inducing 12inhibition are not injurious to the skin, thereflex cannot be considered nociceptive. Never-theless, it seems possible that the reflex may havea protective function. For example, consider ahand either searching for an object or in-advertently encountering one without visual

974



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



guidance. Once one of the finger tips makescontact with the foreign surface, there areadvantages in pausing, since further movementof the hand would cause tissue damage if otherparts of the object were sharp or hot. In contrast,the checking of the hand would enable sufficienttime for more discrete and appropriate ex-ploratory movements to be programmed by thenervous system once the range of the target hasbeen established.The nature of the neural pathway responsible

for the long latency of the 12 reflex is unknown,but sufficient information is available to justifyspeculation. The high impulse conductionvelocities for the sensory and motor fibresinvolved in the reflex would leave approximately55 msec of the reflex latency at threshold un-accounted for. This central time is substantialbut is still compatible with activity in a poly-synaptic pathway. Thus it is known that thespinobulbospinal reflexes may have a longlatency (Shimamura and Livingstone, 1963) and,in the paralysed limbs of patients with com-pletely transected spinal cords, the secondcomponent of the flexor withdrawal reflex maynot appear for periods ofup to 400 msec (Shahaniand Young, 1971). It is nevertheless of interestthat intracellular recordings from mammalianmotoneurones have not so far revealed inhibitorypostsynaptic potentials from fast-conductingcutaneous afferent nerve fibres (Rosenberg, 1970)or low-threshold mechanoreceptors (Engberg,1964), although it is possible that responses withsuch long latencies might not have been searchedfor. An alternative proposal for the mechanismof the '2 inhibitory reflex is that it involves thefusimotor system. It is now accepted that duringvoluntary movements the alpha motoneuronesreceive an excitatory input from primary end-ings in the muscle spindle. Abolition of thisinput, as in the unloading reflex (for example,Angel et al., 1965), will induce a temporary pausein the discharge of the alpha motoneurones. Areduction of spindle assistance will also occurwhenever the fusimotor neurones are inhibitedand it is known from the work of Hunt andPaintal (1958) in the cat that this may happenafter mechanical stimulation of the skin.The question now arises as to whether 55 msecwould allow sufficient time for the reduction inspindle assistance to be achieved. A satisfactory

answer would require knowledge of conductionvelocities of the fusimotor fibres and of the timecourse of relaxation of intrafusal muscle fibres.Although direct measurements of the latterhave not been made in man, the observations ofThorne (1964) on the spindle reflex (S reflex)have provided useful information. He found thatstrong stimulation of the ulnar nerve at the wristevoked a response in the first dorsal interosseousmuscle with a latency of 36 to 45-5 msec and thatmoving the point of stimulation to the elbowcaused the latency to lengthen. Thorne gavereasons for supposing that these stimuli werestrong enough to have excited the fusimotorfibres; the ensuing intrafusal muscle fibre con-traction had initiated a volley in the primaryspindle afferents which, in turn, had causedmonosynaptic excitation of alpha motoneurones.

It is conceivable that the time required forintrafusal muscle fibre contraction to exciteprimary endings is approximately the same asthat for relaxation to halt the sensory discharge.If so, the extra time involved in the '2 reflex, asopposed to the S reflex, would be the cutaneousafferent conduction time (about 15 msec), theconduction time in proximal segments of thefusimotor fibres (about 16 msec from cord towrist, assuming a distance of 72 cm and aconduction velocity of 45 m/sec; Thorne, 1964),and central time for inhibition of fusimotorneurones (about 2 msec if disynaptic). Thisextra time would amount to 33 msec and, whenadded to Thorne's value of 36-45 5 msec for thelatency of the S reflex after stimulation at thewrist, would yield a theoretical latency of 69-78 5 msec for the 12 reflex. This value is of thesame order as that observed in the present study.More direct analysis of the role of the fusi-

motor system in the 12 reflex was attempted bythe use of local anaesthesia to block conductionin fusimotor fibres. If it were possible to achievea voluntary contraction utilizing only the directdescending neural pathway to alpha mo-toneurones, then, according to the presenthypothesis, cutaneous stimulation should nolonger produce inhibition. In one subject withprominent cutaneous reflexes who was in-vestigated in this way there was a markedreduction in all components of the cutaneousreflex at a time when impulse activity in thesmaller myelinated fibres was thought to be

975



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



blocked (see Results section). If the latencies ofthe reflexes are also taken into account the dis-appearance of 12 phase would be compatiblewith the onset of spindle inactivation, while thediminution in I1 and E1 reflexes might haveresulted from impulse blockage in group IIIcutaneous afferent fibres (see above). Thisexperimental approach is admittedly indirect butthe findings give further inducement for evaluat-ing the contribution of spindles to cutaneousreflexes in man. It is probable that further in-formation on this subject will come from studiesemploying the single fibre recording techniquedescribed by Vallbo and Hagbarth (1968).

E1 EXCITATION One important feature of the E1excitatory reflex was that, like the I1 response, itusually had a high threshold to electricalstimulation and involved cutaneous afferentnerve fibres with conduction velocities of 35-40 m/sec. It is of interest that these velocities aresimilar to those determined by Shahani andYoung (1971) for the withdrawal reflex in the legof a normal subject. This similarity raises thepossibility that the E1 reflex is itself part of awithdrawal reflex and that the full reflex wouldhave been obtained had strong repetitiveelectrical stimulation been employed instead.Further evidence in this direction is that the E1response resembled the flexor withdrawal reflexin the leg in displaying a 'local sign' (seeResults section).

It had been hoped that the present study wouldshed further light on the nature of the V2 wavedescribed by Upton et al. (1971). This wavecould be recorded during voluntary contractionin various muscles, including those of the hand,after stimulation of the appropriate 'mixed'nerve. Upton et al. considered a number ofpossible explanations, including reflexes mediatedby cutaneous or joint afferents and a post-inhibitory rebound in motoneurone excitability(see also Andersen and Sears, 1964). In theexperiments of Upton et al. (1971) the latency ofthe V2 wave could fluctuate by up to 10 msecduring successive trials in the same subject (seetheir Fig. 3). Among the population of normalsubjects studied by these authors the latenciesranged from 48-65 msec. This range of latenciesis very similar to that determined for the E1 wavein the present experiments and could indicate a

common identity for the E1 and V2 waves. Thus,in experiments to elicit a V2 wave, a stimulus toa ' mixed' nerve trunk, such as the median nerveat the level of the wrist, will produce excitationof cutaneous afferents as well as fibres frommuscle, tendon and joint receptors. Against theindentification of the E1 wave with the V2response was the absence, in the present study, ofany correlation between the sizes of the twowaves in the same individual. For example, agiven subject might have a very prominent E1wave and yet not display a V2 response duringvoluntary contraction. The most definitiveanalysis of the V2 wave will undoubtedly comefrom further observations on patients withdivided dorsal roots.

E2 EXCITATION The E2 type of excitation had alatency of about 120 msec and the recordingsfrom the median nerve indicated that it dependedpartly on activity in the larger type II sensoryfibres in the digital nerves. The experimentsusing mechanical stimuli further established thatthese fibres belonged to cutaneous mechano-receptors. In addition, type III fibres must havecontributed to these responses, since the E2wave became larger as the stimulus intensity wasincreased, even though the median nervesensory action potential had already reached itsmaximum size.

In keeping with the interpretation of the E1wave, presented above, one explanation of the E2response would be that it represented the secondcomponent of a withdrawal reflex (Shahaniand Young, 1971). The latency of the E2response also suggests an identity of this wavewith the period of greatest increase in excitabilityof triceps surae motoneurones which Gassel(1970) and Gassel and Ott (1970) were able todemonstrate in normal subjects after stimulationof the skin in the lower leg and foot. In thepresent study it was found that, unlike the E1response, the E2 potential was never recordedunless there had been an earlier inhibitory period.The possibility cannot be denied, therefore, thatsome part of the E2 response, particularly thelowest threshold component, was not reflex inorigin but merely signalled the synchronousresumption of voluntary activity in the moto-neurone pool after inhibition. The increasedexcitability of neurones after inhibitory post-

976



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from



synaptic potentials might also make a con-tribution to this discharge (Andersen and Sears,1964).

Similarly, there must be some doubt concern-ing the reflex nature of the alternating inhibitoryand excitatory waves which follow E2. After thefirst synchronous post-inhibitory discharge, therewill be a tendency for motoneurones to continueto fire in unison, provided that their dischargefrequencies are similar. In fact, these frequencieswill not be completely identical and hence theelectromyographic activity will become pro-gressively smoother as the voluntary contractioncontinues.

In conclusion, it is clear that the present studyhas not permitted a complete analysis to be madeof the neural substrates underlying the variousreflex phenomena which have been described.Nevertheless, the study has shown that inputsfrom the skin exert a powerful influence on thereflex behaviour of motoneurones and that thenature of this influence is possibly more complexthan had previously been supposed. It is hopedthat further studies may lead to a more completeunderstanding of the neurophysiological de-rangements observed in disease.

We are indebted to Drs. F. Petito and R. E. P. Sicafor help with some of the experiments, and to Miss 1.Csatari and Mrs. J. English for secretarial assistance.Financial support from the Canadian MedicalResearch Council to Dr. A. J. McComas is gratefullyacknowledged. Mrs J. Leon, who assisted in thepreparation of the final manuscript, is paid by theMuscular Dystrophy Association of Canada.

REFERENCES

Andersen, P., and Sears, T. A. (1964). The role of inhibitionin the phasing of spontaneous thalamo-cortical discharge.Jouirnal of Physiology, 173, 459-480.

Angel, R. W., Eppler, W., and lannone, A. (1965). Silentperiod produced by unloading of muscle during voluntarycontraction. Jouirnal of Physiology, 180, 864-870.

Buchthal, F., and Rosenfalck, A. (1966). Evoked actionpotentials and conduction velocity in human sensorynerves. Brain Research, 3, 1-122.

Dimitrijevic, M. R., Faganel, J., Gregoric, M., Nathan, P. W.,and Trontelj, J. K. (1972). Habituation: effects of regularand stochastic stimulation. Joiurnal of Neurology, Neiuro-suirgery, and Psychiatry, 35, 234-242.

Egger, M. D., and Wall, P. D. (1971). The plantar cushionreflex circuit: an oligosynaptic cutaneous reflex. Jouirnal ofPhysiology, 216, 483-501.

Engberg, I. (1964). Reflexes to foot muscles in the cat. ActaPhysiologica Scandinavica. 62, Suppl. 235, 1-64.

Gassel, M. M. (1970). The role of skin areas adjacent toextensor muscles in motor neurone excitability: evidencebearing on the physiology of Babinski's response. Jouirnalof Neurology, Neuiroslurgery, and Psychiatry, 33, 121-126.

Gassel, M. M., and Ott, K. H. (1970). Local sign and lateeffects on motoneuron excitability of cutaneous stimula-tion in man. Brain, 93, 95-106.

Hagbarth, K.-E. (1952). Excitatory and inhibitory skinfor flexor and extensor motoneurones. Acta PhysiologicaScandinavica, 26, Suppl. 94.

Hagbarth, K. E. (1960). Spinal withdrawal reflexes in thehuman lower limbs. Jouirnal of Nelurology, Neuirosiurgery,and Psychiatry, 23, 222-227.

Holmes, C. McK. (1963). Intravenous regional analgesia. Auseful method of producing analgesia of the limbs. Lancet,1, 245-247.

Hunt, C. C., and Paintal, A. S. (1958). Spinal reflex regulationof fusimotor neurones. Jolurnial of Physiology, 143, 195-212.

Hursh, J. B. (1939). Conduction velocity and diameter ofnerve fibers. American Journal ofPhysiology, 127, 131-139.

Liberson, W. T. (1973). Silent periods and widespread in-hibition in man. In Proceedinigs of the 4th IltterniatiotialCongress of Electromyography, Brussels, 1971.

Lloyd, D. P. C., and Chang, H. T. (1948). Afferent fibers inmuscle nerves. Journal of Neiurophysiology, 11, 199-207.

Lynn, B. (1969). The nature and location of certain phasicmechanoreceptors in the cat's foot. Jouirnal of Physiology,201, 765-773.

Merton, P. A., (1954). Voluntary strength and fatigue. Jouirnalof Physiology, 123, 553-564.

Quilliam, T. A., and Ridley, A. (1971). The receptor com-munity in the finger tip. Jolurnal of Physiology, 216, 15P-17P.

Rosenberg, M. E. (1970). Synaptic connexions of alphaextensor motoneurones with ipsilateral and contralateralcutaneous nerves. Journal of Physiology, 207, 231-255.

Shahani, B. T., and Young, R. R. (1971). Human flexorreflexes. Jouirnal ofNeuirology, Neutrosuirgery, andPsychiatry,34, 616-627.

Sherrington, C. S. (1903). Qualitative difference of spinalreflex corresponding with qualitative difference ofcutaneousstimulus. Jouirnal of Physiology, 30, 39-46.

Sherrington, C. S. (1910). Flexion-reflex of the limb, crossedextension-reflex, and reflex stepping and standing. Jolurnalof Physiology, 40, 28-121.

Shimamura, M., and Livingston, R. B. (1963). Longitudinalconduction systems serving spinal and brain-stem co-ordination. Journal of Neuirophysiology, 26, 258-272.

Thorne, J. (1964). Reflex responses in man evoked by gammaefferent activation. Natuire, 203, 612-613.

Upton, A. R. M., McComas, A. J., and Sica, R. E. P. (1971).Potentiation of 'late' responses evoked in muscles duringeffort. Journal of Neutrology, Neltrosuirgery, and Psychiatry,34, 699-711.

Vallbo, A, B., and Hagbarth, K.-E. (1968). Activity from skinmechanoreceptors recorded percutaneously in awakehuman subjects. Experimental Neurology, 21, 270-289.

977



http://jnnp.bmj.com

/J N

eurol Neurosurg P


ber 1973. Dow

nloaded from


cutaneous reflexes in small muscles hand · motoneurones innervating the small muscles of the hand....

Documents