643

THE PHANTOM LIMBBy J. DONALDSON CRAIG, M.D., M.R.C.P.

Research Fellow in Medicine, St. Mary's Hospital Medical School

IntroductionThat an amputated limb which has long since

disintegrated should be appreciated in the con-sciousness of its former owner and even give riseto severe symptoms is, at first sight,, a concept sostrange that it is hardly surprising that for manyyears the phantom limb phenomenon was largelyignored in the medical literature and, even now,does not receive the wide recognition which itdeserves. No less to the patient than to theobserver, the idea appears absurd and therefore,in response to casual questioning during routineinspection of the stump, the patient is liable toreply that all is well, preserving silence on thematter of his symptoms for fear of appearingridiculous. Yet in reply to direct and sympatheticquestioning in privacy, rather than in the hurly-burly of an out-patient clinic, almost all of thesepeople will freely admit to symptoms of varyingseverity. They do so with considerable andobvious relief that the experiences which they hadperhaps regarded as imaginary or neurotic, shouldbe accorded the same status and reality assymptoms arising from any other disease process.The pattern of these subjective phenomena is

remarkably constant, and must have been ex-perienced since the days when amputation wasfirst successfully performed. In primitive com-munities phantom pain following amputation hasbeen accepted at face value as a reality. When the'wound ' which occasioned the original loss of thelimb continued to be ' painful,' or if, the ' limb'were felt in an uncomfortable position, the lostmember was disinterred, rearranged in a morecomfortable position, and the wound dressed withsuitable medicaments. Nelson remained con-scious of his fingers after the loss of the arm andmaintained that this proved the existence ofhis soul.During the age of igth century materialism,

when reality was accorded only with reluctance tothat which was demonstrable, it was perhaps in-evitable that these phenomena should receivescant attention. The growth of modern psychiatryhas laid more stress on that which is neither visiblenor appreciable by the ordinary senses and it be-came equally inevitable that the phantom limbshould be explained as a product of disturbed

psychological function, and the distressing symp-toms following amputation are believed by manyto represent an obsessional neurosis. So many ofthe occurrences of everyday life and medicine canbe explained with such facility in terms of presentday psychiatry that it becomes pitifully easy tofall into the error of attributing to disturbances ofthe psyche, any symptoms for which no organicfactor is readily demonstrable. The error is madedoubly easy by the development of secondaryanxiety and other psychological disorders in thosewho have long suffered the tiresome symptomswhich, to the patients, may not appear to have beentreated with the seriousness they merit. The factremains that psychotherapy per se in such casesis of little value.The more or less constant pattern of symptoms,

the disappointing results of psycho-therapy andthe presence of obvious nerve lesions in thesepatients should effectively discourage any attemptto explain the phantom limb phenomenon on thebasis of a neurosis within the strict and limitedsense of the term.

Sherrington showed that the central nervoussystem can no longer be regarded as a series ofindependent neurones resembling a somewhatcomplex telephone exchange and therefore to becomprehended in mechanical terms. Rather is itan infinitely complicated dynamic and organicwhole, flexible and labile, which in the face ofinjury or disease will establish a new integration,arriving at its goal, if need be, by the use of newand less direct pathways. MJtilation of the bodyas by amputation therefore produces less dis-turbance than might be expected on the basis ofmechanical concepts. The arm or leg may be asdead as John Brown's body, yet in the sensoriumit may still live.

Nerve endings, sharing in this attempt to re-sist the mutilation and restore the status quo antebegin to regenerate, but the local anatomicalchanges prevent normal growth and result in theformation of bulb neuromata which, with theproximal ends of the cut nerves intact may re-ceive pain stimuli. Like the less unpleasantsensations which occur more commonly, pain sodeveloping is felt in the phantom.

by copyright. on July 23, 2021 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.24.278.643 on 1 Decem

ber 1948. Dow

nloaded from

644 POST GRADUATE MEDICAL JOURNAL December I948

The Body Image ConceptHead and Holmes postulated the existence in

the sensorium of a' body image 'in terms of whichall sensation is registered, all bodily movementsinitiated. Riddoch, amplifying this concept,suggested that there were three body images,visual, sensory and motor, all being normallyintegrated into a whole, compound body image.Clearly amputation partly destroys the visualimage but the sensory and motor elements surviveand accordingly the lost limb is still ' felt,' and canbe ' moved.' Further, pain arising in its territoryis interpreted as coming from the lost limb. Thefingers, elbow and wrist, having the greatestcortical representation and being the most fullytrained parts of the member, are most deeplyrooted in the memory of the sensorium, andaccordingly it is of these parts that impressions aremost vivid and to these parts that pain, when felt,is most projected. Generally speaking, the lostlimb is ' felt ' in a comfortable position of re-laxation but occasionally it is experienced in thefixed position in which it was last really known tothe patient. For example, Leriche records thecase of a woman who lost her arm in consequenceof a motor accident, her last conscious memorybeing of the blood-covered limb thrust throughthe Windscreen. After amputation the arm was' felt' as.being fixed in the outstretched position.Livingstone records several very similar cases.Here, it seems, a vivid, powerful impression of thelimb had been registered in the sensorium, oustingby its strength the more normal body imagedeveloped from all previously accumulated ex-perience. How far psychotherapy might benefitthis type of case is a matter for conjecture, but anemotional catharsis might act in the same way as,one presumes, the powerful suggestive influenceof magic acts when the savage restores to a morecomfortable position his disinterred limb. Nosuch case has been encountered in the presentseries but the experience of a surgical colleague isdirectly aDalagous. He writes as follows:-

' About three years ago I had to have a cartilageremoved from my right knee, and the operationwas performed under spinal. The surgeon kindlyacquiesced to my desire to see what was going on,and I was allowed to lie on the table propped upin a semi-sitting position, with a mirror arrangedto reflect the operation site for me.

' The proceedings started-after the injectionof anaesthetic-with the application of an Esmarchbandage to my right leg, which was held up ver-tically for the purpose. The anaesthetic had notyet become fully operative while this was goingon, and I had a faint sensation of tightness in theleg, but this rapidly passed off. The leg was thenlowered, the knee flexed over the end of the table,

and the operation begun. I can still clearly re-member, my surprise as I watched in the mirror,the knife cut into my knee-not because it waspainless, for I had had several minor operationsunder local analgesia, and was used to the feeling-but because, in spite of being able to view thewhole proceeding in comfort, I had difficulty inconvincing myself that my leg was not still up inthe air where I had last felt it. The leg beingoperated upon did not seem to bqlong to me at all.So vivid was this sensation of my leg being still upin the air that when, having allowed my attentionand my eyes to wander round the theatre, I lookedin the direction. where my leg had been, I felt ashock at not seeing it there clearly, and indeedalmost convinced myself that I could distinguish aghostly outline.

'This sensation persisted throughout the opera-tion, and continued after my return 'to the warduntil normal sensation began to come back to thelimb.'

Generally, with the passage of time, the limb'shrinks.' Formerly of normal or nearly normallength, it may now be only a few inches long,represented only by the fingers and joints. Thisevolution is not to be explained as a result oflocal alterations but in terms of change in thebody image. The ability to feel and move a non-existent limb generally serves no useful purposeand the loss of visual impulses aids in-the shrinkingof the sensory and motor images. The centralnervous system possessing a great faculty foradaptation, the cortical elements formerly servingthe lost limb now cease to abrogate themselves tofunctions rendered impossible by events outsidethe nervous system.

Phantom Limb CausalgiaDuring this period of adjustment, some degree

of pain is generally felt in the phantom. This isdiffuse and frequently described as cramp-like orburning, particularly when the phantom is ap-preciated as being fixed in an uncomfortableposition. Progressively this pain becomes lesssevere and less frequently experienced until ayear or two after the injury it usually ceases tocause more than slight annoyance.There remain, however, an unfortunate minority

(of which patient five is an example) in whom theredevelops what is essentially a causalgic state. Painis then severe. The stump is often cold, anhy-drotic or hyperhydrotic, cyanosed or red,oedematous or shrunken. Uncontrollable spas-modic movements may occur, accompanied bystabs of pain, but none of these findings is in-variable. Tender neuromata may be present orthe stump may present no obvious abnormality.In such instances permanent benefit may follow

by copyright. on July 23, 2021 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.24.278.643 on 1 Decem

ber 1948. Dow

nloaded from

December 1948 CRAIG: The Phantom Limb 645

sympathectomy or interruption of the sympatheticpathways by procaine infiltration, but excision ofneuromata or neurectomy provide only transitoryrelief. It is' frequently accepted that these painsare caused by pressure on neuromata and for thisreason it has been urged that nerves should not beshortened since neuromata forming after' such'shortening are more liable to pressure fromartificial limbs. Neuroma formation is, however,an invariable sequel of amputation and there is noapparent correlation between the size and dis-tribution of neuromata and the amount of painfelt. Clearly, then, some other mechanismoperates and there appears to be no essentialdifference between these severe phantom painsand those which sometimes follow partial or com-plete section of a peripheral nerve and have beendescribed by Weir Mitchell as causalgia.

Lewis (I936) considered that causalgia wascaused by impulses from the distal cut end of thenerve releasing substances in the skin'and pro-ducing a condition of erythralgia. This ex-planation is not valid in the present instance forthe simple reason that there is no distal nerveending, and so pain impulses must arise from theproximal end of the nerve trunk or ftom a related*nerve system such as the sympathetic. Further,as shown by Nathan (I947), the skin is not ery-thralgic in causalgia. It is therefore tempting tosuppose that either causalgia is caused by abnormalsympathetic activity or that the pain is carried bysympathetic fibres. There is no evidence thatsympathectomy in any way affects sensation in thelimbs, although there is some support for theviewthat deep pain conduction can occur via thesympathetic system. There is no doubt whateverthat in a large proportion of .cases of causalgiasympathetic interruption can relieve pain andthat even when this interruption is temporary, asby procaine injection, relief may be prolonged. Ithas been suggested that the action is the interrup-tion of a vicious circle (Leriche, 1940; Living-stone, 1943), and both vasodilatation and vaso-constriction have been incriminated. Yet incases of causalgia no particular vascular changecan 'be constantly demonstrated, save a generalincrease in the blood flow to the affected limb(Leriche and Fontaine, 1929), nor can the pain beconstantly affected by measures affecting thevascular state, such as reflex vasodilatation.Further, the vascular changes occurring incausalgia are not confined to this condition.' Thereis a very wide fluctuation in the blood flow to thelimb in exercise, climatic extremes and disuse fromany cause, including hysteria, all conditionsgenerally painless. It is therefore much more'probable -that the -variable vascular changes inpainful stumps and causalgic limbs are purely

secondary effects and not causal disturbances.Doupe, Cullen and Chance (i944) therefore con-cluded that causalgia results from the stimulationof peripheral nerve endings by sympathetic fibresat the point of damage. Action current studiesby Granit. Leksell and Skoglund (i944) haveshown the possibility of transmission of impulsesfrom motor to sensory nerve-and vice vers'a at the,' artificial synapse"' where a nerve is cut across,'ligatured or mechanically compressed. Trans-mission occurs much more easily from motor tosensory fibre than in the reverse direction and theeffect is facilitated in a freshly damaged region.The effect is far more pronounced in the proximal..than in the distal portion of a nerve and this.observation may have a bearing on the recognizedfact that phantom pains are more likely 'to besevere after high rather than low amputations.The relief from sympathectomy and the aggrava-tion by excessive heat or cold and by emotional-stress can thus be understood, since all theselatter influences increase sympathetic activity.Likewise pain, in itself a stimulant of the sym-pathetic system (Cannon, I929), may reflexlyaggravate or maintain peripheral pain. In thisrespect at least, causalgia may be regarded as theproduce of a vicious circle, and it is easy to seehow permanent pain may result from excessivehandling of the stump during convalescence,necessarily painful, or from excessive anxietyeither constitutional or induced by poor psycho-logical management.The less severe pains occurring more generally

probably differ from the causalgic state in'degree'rather than nature. All tend slowly to spontaneousimprovement but the explanation of this is notclear.

Histor-iesIt is difficult to elicit an objective and unbiased

history of these events from patients who mayview their symptoms as' signifying neuroticderangement rather than as a more or less physio-logical concomitant of profound anatomicalchange. Accordingly, case histories have beencarefully taken from five men known to thewriter personally, two of them for over ten years.All five are highly intelligent, co-operative, freefrom neurotic personality traits, and they havespoken freely in the knowledge that their symp-toms are accepted as being very real. Their own;descriptions are quoted. These have been editedonly for the sake of brevity.

Patient OneAccountant, aged 30. A very good witness who

received a burst of--machine gun fire in the leftarm between the 'shoulder and elbow on April

by copyright. on July 23, 2021 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.24.278.643 on 1 Decem

ber 1948. Dow

nloaded from

646 POST GRADUATE MEDICAL JOURNAL December I948

i6, I944 No pain was noted immediately, onlymarked coldness of the arm, but by the time hereached the Casualty Reception Station ninehours later, pain was most intense. The arm wasamputated in the mid-humeral region almost atonce. In the post-operative period the stump wasextremely painful, and after I4 days the phantompains appeared. These were irritating and cramp-like, and could be relieved temporarily by' moving' the phantom to a more comfortableposition. Later a feeling developed that the fistwas tightly clenched ' and the nails were digginginto the flesh. This was a very real pain which noaction could alleviate. The only cure was tosubmerge oneself completely in some occupationso that no *other circumstance was felt or ap-preciated. How often I tried this unsuccessfullyand, when successful, as soon as the climax of pre-occupation passed, the same tight clenched fistreturned.' From then till the end of October,1944, was by far the worst period for the phantompains and over two stones in weight were lost.The arm began to shorten but the pains werealways there. ' They were not a thing to complainabout because, after all, I did not expect a doctor totreat a ghost.'

Thereafter, the shortening continued and thepains steadily lessened in severity. By November,I944, it could be divided into three components.

(i) ' Rheumatic' pains in the shoulder-asharp piercing pain in the joint occurring in coldweather aggravated by drafts and relieved bymassage.

(2) A dull pain confined to the bone and broughton by excessive warmth or cold.

(3) Minor aches similar to those experiencedafter muscular overwork,' a stiffness which. causesincreased pain on movement of the stump, butthat very movement will eventually make bothpain and stiffness disappear.'

PRESENT CONDITION, FOUR YEARS AFTER INJURYThe phantom is now so short that 'the hand

appears immediately after the amputation and yetsomehow there seems to be an elbow and a wristbetween it and the end of the stump.' The scaris perfectly healed. The only tenderness is causedby a gpntle stroke across the scars. A firm touchor moderate bang does not cause any paifi. Phantomsensations and minor pains, although no longertroublesome, still persist. In cold weather a sockis worn over the stump, but no other precautionsare taken. An artificial limb is not worn-

Patient TwoWing Commander, fighter pilot, aged 26. Right

great toe shot off in 1940 by cannon fire, the stumpbeing modelled and sutured a few hours later, the

metatarsal head remaining intact. The scarhealed quickly and there was never any pain butwithin a fortnight the sensation arose that themissing toe was still present. A year later thisfeeling persisted though in lesser degree. Duringthis period he frequently stumbled or fell fromattempting to turn on the missing member. Bythe middle of I942 the phantom was veryshrunken, and by i944 it had virtually disappeared.

Patient Three. Medical student, aged 29January, 194I. Compound fracture of right

tibia and fibula and right patella from enemyaction, followed by amputation four days laterat junction of upper and middle third of femur.' Pain in the stump was intense for about threeweeks but its character cannot now be accuratelyremembered. Phantom pains are, however,clearly recollected as starting about two monthsafter operation and they take two forms:

' (i) A fairly constant numb sensation.' (2) A sharp pain, spasmodic and sporadic.' Originally these appeared to come from the

areas where the foot ought to have been but withthe knee slightly flexed. Slowly, almost im-perceptibly, the sensation grew nearer and nearerto the end of the stump, and for the past fouryears the phantom foot has been just below andbehind the end of the stump.

' (i) The numbing sensation like that of wearingshoes which are too tight appears to come mainlyfrom the toes, the plantar surfaces of the distalphalanges and from the dorsal surface of thefoot. This sensation, fairly constant, is notnoticed when my attention is distracted elsewhere,as when talking or concentrating hard on reading.I can always produce the sensation at will, unlikethe pain which I can never produce when I desireto do so.

' Tickling the sole of the left foot produces anunbearable sensation in the phantom foot, almostimpossible to describe. It is as if ice-cold watersuddenly permeated the whole of my phantom leg,starting proximally and pouring down to pool inthe. phantom foot. It is only transient, passingoff within a few seconds. I cannot produce it atwill.

'(2) The spasmodic pain seems to originate at-various sites, most commonly the medial side ofthe big toe, the ball of the foot, the malleoli. Thepain is sometimes severe enough to make me jumpor wake me out of sleep. Again there are twoelements, a deep pain felt in the bone and fairlyconstant for about five to six seconds. The otheris more superficial and severe, coming in shortsharp bursts like machine gun fire, the first twingebeing the worst, the others diminishing in in-tensity and the whole episode lasting only three

by copyright. on July 23, 2021 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.24.278.643 on 1 Decem

ber 1948. Dow

nloaded from

December 1948 CRAIG: The Phantom Limb 647

to five seconds, during which the stump sometimesjerks spontaneously.

' These pains, once they start, continue spas-modically for some time, varying from days tohours. I can find no relation to fatigue, exerciseor weather variations, and they have altered littlein frequency or severity since convalescence,coming on once every four to five weeks.'No treatment has been employed save analgesics

such as codeine, which has only a slight effect.

Patient Four. Civil engineer, aged 47In I943 he was involved in a railway accident in

the Middle East, being pinned under the wreckagefor six hours und.er particularly distressing cir-cumstances, and receiving compound fractures ofboth tibiae and fibulae. It was nearly 24 hoursbefore he reached hospital, seriously shocked butnever having lost consciousness. After con-servative operations on both legs he made a goodrecovery and was able to deal with some business.Knowing that severe psychical symptoms mightfollow such an ordeal if the memory were re-pressed, he deliberately and repeatedly wentthrough all the details of the episode in his mindand has never had any psychological symptomssuch as nightmares or phobias from the accident.A fortnight after the accident amputation of theright leg at the junction of the upper and middlethirds was necessary,'but the left leg continued toheal satisfactorily. Within 24 hours of the opera-tion, the sensation arose that the leg was stillpresent, painless and outstretched in the bedbefore him. To convince himself that the leg wasabsent he repeatedly hit the bedclothes just aheadof the stump. Within two months of the accidenthe was fit enough to be flown to convalescence inSouth Africa, where although the left leg was stillin plaster he got about and led a very active lifementally.To this day he is still constantly awre of the

presence of the ' phantom,' which is of normalsize and shape corresponding roughly with thevariable position of the artificial limb. Practically'from the first pains have been felt in the foot butnowhere else. They are likened to severe pins andneedles, coming in five second bursts. Originallyfairly frequent, these episodes now occur onlyabout two or three times daily. They may bebrought on by strong exertion of. the stump,warmth, jarring or allowing the socket of the,artificial limb to press firmly against the back ofthe stump. They may be accompanied by spas-modic jerking of the stump or this type of un-controllable movement may occur alone.On several occasions the stump has been in-

jured in falls'during the course of his arduousoutdoor occupation. Pain so caused has always

been felt locally and never referred to the phantom.

Patient FiveBusiness executive, born i892, left Cambridge

with a good record to serve in the first worldwar, receiving injuries in I9I6 which necessitateda mid-thigh amputation of the right leg. Hismemory of these earlier events is not to be trusted,but probably phantom sensations began about afortnight after amputation, to be followed shortlyafterwards by definite severe pain in the foot and,to a lesser externt, the knee. From then therefollowed a dismal tale interspersed by variousoperations for remodelling of the stump andsuccessive neurectomies, all producing relief for ashort time only. His intellectual deteriorationwas marked and like so many sufferers from pro-longed pain he developed mild paranoic traits.He saw all too clearly the many faults of societyand overlooked its virtues. He quarrelled withwife and family and recently died, a lonely andembittered man.

DiscussionFrom consideration of the stories told by other

amputees, it is believed that these case historiesare entirely typical, differing only in the peculiarcircumstances which facilitated the taking of aclearer and fuller account of their symptoms thancan be obtained from the ordinary patient.

In case one, a major amputation of the arm,causalgic pains appear to have been fairly severeat one time yet, with the ' shrinkage ' of thephantom they have become only a minor source ofannoyance. Until this man was directly questionedin 1948 he had never before confided his story toany medical attendant-' he did not expect adoctor to treat a ghost.' His morale was ex-cellent, his temperament stable and his outlookbalanced. He was able to accept his symptoms,knowing from conversation with other amputeesin the same ward that t'hese symptoms were afairly common occurrence. He discovered forhimself that concentration on other mentalactivity reduced or obviated sensations of pain,and his recovery has been good. The pain whichcan now be induced by gentle stroking of the scarbears a strong resemblance to the protopathic painof incomplete nerve recovery.

Subjects three and five have noticed ihe closeassociation between uncontrollable stump mnove-ment and spasmodic pain which suggest that botharise from bursts of nerve activity, perhapssympathetic discharge. Whereas subject three hasexperienced the typical shortening phenomenon,subject five continues to feel his leg as of perfectlynormal length and, since he can ' move ' thephantom he is able to identify it with his artificial

by copyright. on July 23, 2021 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.24.278.643 on 1 Decem

ber 1948. Dow

nloaded from

POST GRADUATE MEDICAL JOURNAL December I948

limb. It may be that he has retained normallength because at the time of amputation he wasover 40 and had led a very active athletic life, sothat the image of his limbs was very firmly estab-lished in the sensorium. The stump was painlessfrom the first and he made a particularly deter-mined effort to master the use of the artificial limbwhich has now become incorporated in themodified body image and he may be said to havemade a very good recovery. The reference of un-pleasant sensations to the phantom (patient three)from stimulation of the opposite foot is mostinteresting and argues the presence of abnormalfoci of activity in the cord, probably the inter-nuncial pool of neurones or possibly at a higherlevel, so that abnormal reflex mechanisnis havebeen established. It might be argued thatsymptoms here are sufficiently severe to justifvsome form of active treatment but in view of thetendency to continued spontaneous improvement,this is not contemplated at present.

Patient five and others like him provide thegreatest difficulty. In this, the establishedcausalgic state, one is dealing with a profoundpsychosomatic disturbance, in which both physicaland mental factors operate and adversely affect oneanother in a vicious circle, so that it is hard to saywhich of the two is the more important. Sym-pathetic nerve block earlier in the course of themalady is indicated for such people and should berepeated if necessary, Only when this approachfails should more severe operative measure beundertaken.

ConclusionsPhantom limb phenomena, in one form or

another, are of constant or nearly constant occur-rence after major amputations. This is notsufficiently widely appreciated and patients arenot taught to expect them. It is therefore almostinevitable that when these sensations are noted bythe patient they should act as a source of anxietyand perhaps preoccupation. If he were taught toregard such symptoms as a normal sequel ofamputation, which would diminish with thepassage of time, the amputee might be saved muchdistress.

In a large proportion of instances, causalgicsymptoms occur in greater or lesser severity. Thefactors causing them and the mechanism of theirproduction are not yet clear. But since any in-fluence producing sympathetic hyperactivity mayaggravate these pains, it is reasonable to supposethat unnecessary anxiety or pain during con-valescence may predispose to their development,

particularly when it is remembered that it is duringthis phase of early damage that transmnission ofimpulses from motor to sensory nerves mostreadily occurs. During this healing stage, the un-necessary movement or handling of the stump istherefore to be avoided. Simple sedation as byphenobarbitone may prove valuable. Reassurancemay have to be repeated and every effort should bemade to direct the patient's interest to activitiessuch as light occupational therapy as soon as he hasrecovered from the immediate effects of the opera-tion, always provided that movement of the stumpis avoided. He will then be a much happier manthan the patient who lies abed meditatively con-templating his stump and his phantom, wonderingwhat it is all about.

SummaryI. Phantom limb sensations are believed to

occur after almost every major amputation andcausalgic symptoms of greater or lesser severityare common.

2. These symptoms are not primarily neurotic,though secondary psychological disturbances mayfollow.

3. The mechanism of their production is out-lined, and the probable pathology of causalgia isdiscussed.

4. Five typical histories are reviewed anddiscussed.

5. It is emphasized that much needless anxietyis caused to patients by failure to acquaint themwith the fact that phantom sensations are a realityand not a product of disordered imagination.

6. Suggestions are made for managementduring the early convalescent phase.

BIBLIOGRAPHY

CANNON, W. B. (1929), 'Bodily Changes in Pain, Anger, Fearand Rage.' New York.

COHEN, H. (1947), Lancet, i, 933.CRAIG, J. D. (1948), Lancet, i, 497.DOUPE, J., CULLEN, C. H. and CHANCE, C. G. (1944), J.

Neurology, Neuro-Surgery and Psychiatry, 7, 33.GRANIT, R., KEKSELL, L. and SKOGLUND, C. R. (944),

Brain, 67, 125.HEAD, H. and HOLMES, G. M. (I9Is), Brain, 34, o02.LERICHE, R. (1940), 'The Surgery of Pain,' (Trans. A. Young.)

London.LERICHE, R. and FONTAINE, R. (1929), Lyon Chir., 26, 323.LEWIS, T. (1936), 'Peripheral Vascular Disorders.' London.LIVINGSTONE, W. K. (1938), Arch. Surg., 37, 353.LIVINGSTONE, W. K. (943), 'Pain Mechanisms." New York.MITCHELL, S. W., MOORHOUSE and KEEN (I864), 'Gun-

shot Wounds and other Injuries of Nerves.' Philadelphia.NATHAN, P. W. (I947), Brain, 70, 145.PERKINS, G. (I944), Brit. J7. Surg., 31, 377.RIDDOCH, G. (194I), Brain, 64, 197.SHERRINGTON, C. S. (I906), 'Integrative Action of the Central

Nervous System.' New York.

by copyright. on July 23, 2021 by guest. P

rotectedhttp://pm

j.bmj.com

/P

ostgrad Med J: first published as 10.1136/pgm

j.24.278.643 on 1 Decem

ber 1948. Dow

nloaded from