THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES 79 T'HE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES. By STANLEY MELVILLE, Dir-ector of thle Radiological Departauint, Broiiptont (Chest) Hospital; Conisutltinig Radliologist to Kinig Edward/ 1I!tZl Sanatoriul1in, lilinih tirst; Consultintg Radiologist to thi' Ro(yal Naitioial Sanatorium, Veuitnor; Honorary Radtliologist to) St. Gcorge's Hospital. IN a short article it will be the initenitioni of the writer to suggest the value of this youngest of the anicillary aids to diagnosis in (a) Pulmonary tuberculosis, (b) pleuritic effusion ; (c) neoplastic disease. PULMONARY TUBERCULOSIS. The value of radiology in the diagnosis of tuberculous affections of the lung par- enchyma would seem to have become an accepted fact. Most clinicians, and certainily all physicianis wvith special knowledge of affections of the chest, appreciate the assist- ance that this branch of medical science gives them, and we have gone beyond the time when it was possible for great clinicians to doubt its diagnostic value in the early stage of pulmonary tuberculosis. The writer is inclined to think that any remaini- ing suspicion of the valtue of radiography in this con'nection, rests largely with the man making the report, and it is up to the radiologist to study the question- from every aspect with such care that he does not fall into the matny pitfalls that await him and induce him to read inito a skiagraph more than is well within the limits (very wvide limits) of the physiologically normal ; not every adventitious shadow in the lung is pathological. The present position of radiography in the diagnosis of pulmonary tuberculosis may be put something as follows: (i) Definite infiltration can be demonistrated upon an X-ray film at quite an early stage of the disease and frequenitly before defitnite physical signs are evidenit. (2) X-ray evidence extencding over a period of some moniths aiid conisisteiitl1 ntegativc may be takeni as conclusive evi- dence of the nioni-existence of pulmoniary tuberculosis. (3) That, in cases in which physical sigins are present, it may be assumed that the initial stage has passed, and in suich cases the X-ray picture will show, as a rule, much more extenisive disease than can be demon- strated by the physical examination. This is especially true in the presence of cleep- seated lesions anid where there is emphysema, both of wvhich conditions are a handicap to the clinician. Thus it comes ab'out -that radiology is of service, not, heaveni forefend, as in the slightest degree taking the place of the clinician, but as anl invaluable aid, possibly the greatest of all anicillary aids in diagnosis. It would seem right to suggest (though writing as a radiologist there is the risk of prejudice creepitng in) that the best method is for the clinical and the radio- logical examiniationi to be unidertakeni by separate individuals. For this thought there are several reasons. Fiirstly, whichever is done first, the mental balance muist be in- fluenced; in the second, place, the patient gets the benefit of two individual examina- tions, and consequently, opinions; in the third place, the clinician who becomes his own radiologist must lose that fineniess of outlook, that power of summing up all the evidence that has been brought together. Those who know the work of the Law Courts will appreciate the analogy of that most trying person, the judge who will insist upon acting the advocate anid thus prejudging his case. To sum up, it would appear that the value of radiology lies in the confirmationi (or otherwise) of clinical suspicion ; in deter- mining the extent of disease prior to the induction of artificial pneumothorax; in noting the effect of the latter procedure. It wouLld be almost a truism nowadays to on April 27, 2021 by guest. Protected by copyright. http://pmj.bmj.com/ Postgrad Med J: first published as 10.1136/pgmj.4.41.79 on 1 February 1929. Downloaded from
Transcript
THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES 79
T'HE VALUE OF RADIOLOGYIN SOME INTRATHORACIC
DISEASES.
By STANLEY MELVILLE,Dir-ector of thle Radiological Departauint, Broiiptont (Chest)Hospital; Conisutltinig Radliologist to Kinig Edward/ 1I!tZlSanatoriul1in, lilinihtirst; Consultintg Radiologist to thi'Ro(yal Naitioial Sanatorium, Veuitnor; Honorary
Radtliologist to) St. Gcorge's Hospital.
IN a short article it will be the initenitioni ofthe writer to suggest the value of thisyoungest of the anicillary aids to diagnosisin (a) Pulmonary tuberculosis, (b) pleuriticeffusion ; (c) neoplastic disease.
PULMONARY TUBERCULOSIS.
The value of radiology in the diagnosisof tuberculous affections of the lung par-enchyma would seem to have become anaccepted fact. Most clinicians, and certainilyall physicianis wvith special knowledge ofaffections of the chest, appreciate the assist-ance that this branch of medical sciencegives them, and we have gone beyond thetime when it was possible for great cliniciansto doubt its diagnostic value in the earlystage of pulmonary tuberculosis. Thewriter is inclined to think that any remaini-ing suspicion of the valtue of radiographyin this con'nection, rests largely with theman making the report, and it is up to theradiologist to study the question- from everyaspect with such care that he does not fallinto the matny pitfalls that await him andinduce him to read inito a skiagraph morethan is well within the limits (very wvidelimits) of the physiologically normal ; notevery adventitious shadow in the lung ispathological.The present position of radiography in
the diagnosis of pulmonary tuberculosismay be put something as follows: (i)Definite infiltration can be demonistratedupon an X-ray film at quite an early stage
of the disease and frequenitly before defitnitephysical signs are evidenit.
(2) X-ray evidence extencding over aperiod of some moniths aiid conisisteiitl1ntegativc may be takeni as conclusive evi-dence of the nioni-existence of pulmoniarytuberculosis.
(3) That, in cases in which physical siginsare present, it may be assumed that theinitial stage has passed, and in suich casesthe X-ray picture will show, as a rule, muchmore extenisive disease than can be demon-strated by the physical examination. Thisis especially true in the presence of cleep-seated lesions anid where there is emphysema,both of wvhich conditions are a handicap tothe clinician. Thus it comes ab'out -thatradiology is of service, not, heaveni forefend,as in the slightest degree taking the placeof the clinician, but as anl invaluable aid,possibly the greatest of all anicillary aids indiagnosis. It would seem right to suggest(though writing as a radiologist there is therisk of prejudice creepitng in) that the bestmethod is for the clinical and the radio-logical examiniationi to be unidertakeni byseparate individuals. For this thought thereare several reasons. Fiirstly, whichever isdone first, the mental balance muist be in-fluenced; in the second, place, the patientgets the benefit of two individual examina-tions, and consequently, opinions; in thethird place, the clinician who becomes hisown radiologist must lose that fineniess ofoutlook, that power of summing up all theevidence that has been brought together.Those who know the work of the LawCourts will appreciate the analogy of thatmost trying person, the judge who willinsist upon acting the advocate anid thusprejudging his case.To sum up, it would appear that the value
of radiology lies in the confirmationi (orotherwise) of clinical suspicion ; in deter-mining the extent of disease prior to theinduction of artificial pneumothorax; innoting the effect of the latter procedure.
80 THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES
assert that where hiemoptysis is due totuberculosis of the lungs, infiltration can bedemonstrated on the X-ray film. There are,however, so many forms of hemoptysisthat are not of tuberctulous prigini, that it isnot correct to take it for granted that thetubercle bacillus is the offending organism.Such things as bronchial varix, cardiacdisease, the indefiniite familial hzemoptysisand vague purpuric tenden"cies, not to men-tion aneurysm, bronchiectasis and neoplasticdisease, are all responsible for definite andstriking lung hemorrhage. The common,acute, onset of. a new glowth of the lungwith a brisk lihemoptysis, is all too commona pictture. Within the last year, the writerhas seen two patients in whom the chief clini-cal evidence was recurrent hamoptysis andboth the patients had been sent to a sana-torium on suspicion of pulmonary tubercu-losis. One patient had the further clinicalevidence of alteration in the breath soundsat the left base, and after careful investiga-tion the cause was found to be due to the.presence of a blastomycetes. The otherpatient had loss of breath sounds to someextent at one base, and after the. intratrachealinjection of lipiodol a small growth wasfound to be blocking the main descendingbronichus; subsequent bronchoscopy alsodemionstrated this, and the growth (a smallfibromna) was successfully r"emoved and .thepatient did well, the h,emoptysis ceasing atonce.
Before concluding this part of the article,it may be fitting to deal shortly with twoschools of thought that would appear tohave arisen since the birth of radiography.One school, and that by far the largest,following the teaching of all the greatclinicians and pathologists, regards the initiallesion of pulmonary tuberculosis as begin-ning in the lung parenchyma, such invasionbeing rapidly followed by reaction in thebronchial lymph glands. The other school,including amongst its teachers men of thehighest reputation and learning, maintainsthat most, if not all tuberculous lesions
begin in the ly'mphatic glands and spreadthence into the lung parenchyma. Fromthis school has arisen the term "hilum-tuberculosis." The more and the longer isone's experience, the less can one subscribeto this teaching. In children, wvith thegreater tenidenicy to necrosis and caseatiotnof the lymph nodes, there may be a directspread, but even this is surely an inhalationspread ? Again, it has very frequently beennoticed that after the production of an arti-ficial pneumothorax, subsequent infiltration,if it takes place in the more sound lung, isoften of the more central type, but even herethe further spread is by inhalation.
Anotlher and more serious heresy, thoughpossibly the same heresy under anothernamne, is that of so-called peribronchialphthisis, of which Sir James Kingston Fowlerwittily remarks, " peribr-onchial phthisis iscommon in America but unknown inFrance." This statement is, however, notquite fair to America, for Wesseler andJaches are most condemnatory of this lineof thought. They draw attention to thelinear striation (uponl which this disease isbased), being caused just as much by blood-vessels and by lymphatic ducts as by thebronchial tree shadows. Once more theradiologist, backed by some clinicians, hasallowed his vision to be obsessed by anexaggeration of the liniear markings anidhas, ascribed to them a pathological signifi-cance. Thus round a feeble structure,grew a new disease; fortunate, possibly, forsome persons who during the Great Warachieved their desire and escaped militaryservice, but nevertheless, short-lived and notheard of to-day.
PLEURITIC EFFUSION.If one disease more than another has of
late been in the minds of us all, it has beenthat affection leading to effusion in thepleural cavity. The ordinary type of pleuraleffusion, whether bacillary in origin, or theresult of chill, giving the classical evidencesto the clinician (Ellis's curve, loss of tactile
THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES 81
and vocal fremitus, displacement of heart,loss of vocal resonance and breath soundswith possibly agophony and Skodaicresonance) are sufficiently definite withoutthe aid of the radiologist. After parencen-tesis, it may be necessary to determine byX-ray examination either the amount ofresidual fluid or the presence of the under-lying cause, e.g., tuberculosis of the lung.Even here there are some possible clinicaldifficulties and possibility of error, especiallyin children, but witlh these the present articlehas niot space to deal. It is in encystedeffusions, and especially those along the lineof the interlobar septum that the aid of theradiologist is sought. Such effusions maybe situated aniywhere in the pleural cavityand may be either serous or purulent. Themost commnon cause of the infection ispneumococcal, more rarely streptococcaland the remainder usually tuberculous.
Wesseler and Jaches divide encapsulatedeffusions into two groups:-
(a) Suiperficial or parietal-apical, axillary,central or basal.
(b) Conlcealed. (i) Between the lung andthe diaphragm. (2) Between the lung andthe mediastinuin. (3) Interlobar.The infrapulmonary or diaphragmatic
empyemata are naturally difficult of recog-nition clinically, especially when they aresituated on the right side. The fluid isencapsulated by adhesion of the lung eitherto the diaphr-agm or to the chest wall.Where a diagnosis has to be made betweeninfrapulmonary empyema and liver abscessor subphrenic abscess, the high-domed andsharply-defined diaphragmn is frequentlydiagnostic of the latter. As a rule, in basalor infrapulmonary empyema the fluid col-lects posteriorly, that is to say, between thedome of the diaphragm and the lower limitof the pleural sac, the main pletural cavitybeing shut off by adhesion.The respiratory movement in the more
mesial part of the chest being limited, it isnore readily cut off by adhesion, thusencapsulated effusions .are least conmmon in
this situation. The diagnosis is all but im.-possible by ordinary clinical methods, butthe sharply-defined opacity near the hiilumof the lung, in company with clinical evi-dence pointing to suspicion of effusion, ismost striking.
Possibly the most puzzling of the encystedeffusions are those situated in the regionof the axilla. The oval or spindle-shapedopacity, with sharply-defined inner wall, isreadily recognized. The clinical evidencehere again may be perplexing, for, oni accounltof the frequent fixation of the lung by ad-hesion to the edge of the empyema cavity,the usual absolute dullness to percussion iswanting; again, and especially where theeffusion is lyinig posteriorly, a loud, almosttympanitic note may be heard anter-iorly.Finally, many of the apparently axillaryeffusioins, seen in the postero-anterior plane,are found to be interlobar effusions whenviewed in the lateral plane. In moderntechnique great importanice is placed uponthe value of the examination in the lateralplane; indeed, in almost all forms of en-capsulated effusion, this view is most in-formatory and, unless absolutely impossible,should never be omitted. Even with a veryill patient, this is readily done withoutmoving the patient. One final word on thesubject of pleural effusions. It is of thegreatest importance to distinguish betweenan effusion in the general pleural cavity(encysted or otherwise) and an interlobarempyema, in regard to the question of treat-ment. An exploring needle may, with im-punity, be inserted in the first condition.In the second condition this is not the case,having regard to the risk of infecting thegeneral cavity and producing a pyothorax.
NEOPLASMS OF THE LUNG.
Prior to the advent of radiology, therewere but few publications and very few dis-cussions on initrathoracic nieoplasms. Suchreferences as were made were chiefly fromthe standpoint of morbid anatomy. Norcan there be surprise at this 4ttitude. In
82 THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES
the first place, intrathoracic new growthswere regarded as being of extremely rareoccurrence, and, in the second place, thehopelessness of the prognosis was not con-ducive to any great enthusiasm.New growths of the lung are usuallv
divided inito two groups, namely, those thatar-e benign and those that are malignanit.It is well, however, to note that only in thepatlhological sense can any such distinctionbe drawn, for the most benign niew growthwill, in time, deprive the patient of life by aprocess of pressure. It is well to stress thispoinlt, if for no other reason than that theclinician and the radiologist may be alive tothe urgenit need for early diagnosis. I nanother place the writer made the suggestionthat on the first evidence of even occasionalbreathlessness, that could not be accountedfor, the aid of the r-adiologist should besought.The benigni neoplasms are for- the most
part extrapulmonary and cause pressuresymptoms by pushing the lung in fronit ofthem and but seldom directly itnvading it,but in time displacing the heart anidmediastin um.Of the commonier benign new growths
fibroma anid the teratomata or congenitalinclusion cysts are the most common. Thefibromata grow usually from the posteriorthoracic wall and are shown to be well-defined opacities pushing the surroundinglung in front of thein but niot dir-ectlyinvading it. Such ttumours are non-irritanitand but rarely produce -any pletiritic irrita-tion. The diagnosis is readily made bycollapsing the lunig, wheni they will be seen*to be extraptulmonary. Operation for theirremoval is now fairly common, and, if thegr-owth is not of large size, i.e., if thediagnosis has beeni madle ear-ly, the result ofsurgical interference is very good.
Teratomitata or congeniital inclusion cystsmay occur anywvhere. In the thorax, themost common are of the type of- dermoidcysts, mainly epiblastic in origin and con-taim any of the constituienits of skin (hair,
sebaceous material and even teetlh or frag-ments of hone)., They grow from theiretnainis of the thymus or from the third orfourth branchial clefts anld are consequenitlyanteriol in position. In sittuation they thusdiffer definitely from-n the fibromata. Thederlmoid cyst is muLch more irritant, andadhesions betweetn the cyst and the sur-rounding pleura are commotn, making it amuch more diffic:ult matter to separate themfrom the lung tissue, which they invade.The shape of the dermoid cyst is generallyirregular- and not so rounded as the fibro-mata. The diagnosis is readily made byexamination in the true lateral plane, theanterior position of the tumouir beingalmost pathognomonic.
Unfortunately, the early clinical diagnosisof initrathoracic niew growth (whether benignor mnalignant) is indefinite atnd at times evenmisleadinig. The only evidence in one caseseeni by the writer was a brisk hemorrhage,wlhich did not assist in diagnosis. Thelotnger a dermoid cyst is left in the thorax,the more difficLult must operation for itsremoval become, and this is to be r-egretted,for if surgery is to have a fair chance,the earlier the diagnosis, the better theoutlook.Of the maligniant neoplasms we are
concernled with carcinloma and sarcoma.Carcinioinata are by far the most common,
indeed it is a matter of doubt if there is anypositive evidence of primary sarcoma of thelung itself.
It is still a vexed question whether or niothere is an actual inicrease in lung cancer.Is the apparetnt increase due rather to thegreater facilities for diagnosis, and the factthat in the pre-radiological days, many caseswere ascribed to pulmonary tuberculosis,particularly of the fibroid type? Even ifthis be the case, surely the pathologist wouldhave noticed changes inconsistent wvith sucha diagnosis. Cur-iously enough the associa-tion between chronic pulmonary tuber-culosis and cancer of the lung has been-nloted by the writer in two cases within the
THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES 83
past mo'nth: in botlh, ani old tubercularlesioni with excavation, and associated withit a definite carcinioma growinlg from themed iasti numn. Friedlander fouLnid squamous-celled carcinoma in the wall of ain oldtuber-culouis cavity. It would seem thenthere is a definiite inc-rease and the writercan onily judge by his owni experience thatthe last few years have revealed shadows inthe lutng which had not been. met withduring nearly twenty years of steady wvork.lt is certainly true that colleagues of thewriter working in other chest hospitals inLondon ar-e finding similar inicrease forwvhich they can account in no other waythani that it is real and not due to otherfactors.Almost without exceptioni, carcinoma of
the lung has ani acute history to start withor acute bronchial irritation appears to beset up. Increasing dyspnzea and unpro-ductive cough are frequently the mainiclitnical signls. Not inift-equentlv, the earlyacute symptoms appear to subside for a timeand the patient seems to be better. A briskhaemoptysis may be anl early clinicalsymptom. At times the discrepancy betweenthe intensity of the physical signs presentand the apparent well-beinig of the patientis of marked significanice to the clinician.
Primary cancet of the lung originateseither (a) from the epithelial lininig of a largebronchus, most comtmonly a mnain bronichusclose to the bifurcationi of the trachea;(b) less commonily fromii- a smaller bronichusdeeply situated, producing the diffuse infil-trating growth in which there is rapiddissemination and destruction of lung tissue.The latter is in all probability identical withaniother form which has beeni described,namely, the " cavernous carcinoma," inwhich necrosis has taken place in the centreof a large mass.
Upper Lobe Catcinomna. - The growthoriginates from the bronchial mucosa ofa large bronchus and causes eventuallystenosis of the tube. What the radiologistsees is not so much the gro-wth itself but the
collapse of that part of the lung supplied bythe large bronchus inivolved, that is to say,the results of the stenosis. Even with thewhole of the upper lobe collapsed andappearinig absolutely opaque the growthitself may still be quite small. The collapsedlobe is sharply demarcated by the lesserinter-lobar fissure. The heart is usuallydrawni over to the affected side partly onaccount of the shrinking of the Collapsedlung anid partly by the other lunig. In theear-ly stages, with the exception of a peculiarspasmodic and uinproductive cotugh, thereis but little clinical evidence.
Hil/a Cadrcinonma.- Some of the -tumoursoriginiating near the hilumii, instead -ofcausing stenosis, appear to traverse thebronchial wall and to invade the lung. Inthese cases, met often in lower lobecarcinioma, there is considerable enlarge-ment of the bronchial glands. The growthis muost irregular in form and outline, andheavy linear processes are seen extendinginto the lung tissue. It is in the diagnosisof these conditions that most of the radio-logical errors have been made. At timesthey may even be mistaken for interlobaremphysemata, and niothing btut the closestco-operation between clinician and radio-logist can avert error.Basal Carciniomiia.-Here the area of lung
involved is seen to be irregularly opaqueand collapsed. Owing to the fact thatdegenerative processes take place morerapidly thanl in upper lobe carcinoma, itwill be appreciated that the differentialdiagnosis between this condition andbronchiectasis and basal abscess is ex'tremely difficult. In a case in the writer'srecollectioni, the finial diagnosis was madeonly at the autopsy, when, after much diffi-culty, a minute fragmenit at the extremeperiphery of al large friable and broken-down mass was found, which, micro-scopically, was carcinoma.The diaphragm in new growth of the
lung is an interesting study. If the growthoriginates in a main bronchus before it
84 THE VALUE OF RADIOLOGY IN SOME INTRATHORACIC DISEASES
enters the lunig, the phrenic nerve may beinvolved, giving the characteristically highdiaphragm with paradoxical movement. If,on the other hand, there is any considerablecollapse of lung due to blocking of thebronchus, the diaphragm on that side willbe raised and almost immobile.There are two aids to diagnosis of some
value:-(i) The production of an artificial pneu-
mothorax for diagnostic purposes is a simpleand harmless expedient, and, although notof invariable value, at times the greaterdenisity of the involved lung compared withthat of mere collapse is striking.
(2) The intratracheal injection of lipiodol.Of the value of this procedure the writercatnnot speak too highly. ln case of com-plete bronchial steniosis the sharply definedlimit of the opaque fluid is strikingly shown.In conditionls in which the bronchus is notcompletely blocked, the fluid is seen to passthr-ough the par-tially constricted tube, anirreguilar outline, even a "rat-tail " appear-ance being shown.
This appearance is so unlike anythingelse, that the writer is convinced that in it,evidence of the greatest value is obtainable.
Sarcomita.-Primary sarcoma of lung is ofextreme rarity. In 1903 Rolleston andTrevor published a case of this nature, anidone or two other authentic causes appear tohave beeni seen. Metastasis from soine un-discovered cause undoubtedly accounts forsome, and a miietastasis from a small medias-tinial primary sarcoma may well have beenmistakeni. In addition, there would appearlo be no doubt but that some tumourslabelled sarcoma are in reality specimens ofsmaller-celled carcinloma.
Endothelioma of Plentra.-Primary endo-thelioma is rare - but the characteristicfeature is chiefly blood-stained effusionwhich shows the typical cells. On produc-tion of an artificial pneumothorax, irregularopacities on the pleura may be seen or thiscondition may be viewed directly with thethoracoscope.
Metastatic Tunmours of Lunttg.-As a rule itis not possible to distinguish with any cer-tainty between the metastases of sarcoma orof carcinoma. Metastatic sarcoma, in themajority of cases, follows a primar-y focus onlthe bony structuires, and grows with alarminigrapidity. Metastatic carcinoma, on the otherhand, is of much slower growth. Therounided " puff-ball" shape anid the fact thatthey are usually multiple are common toboth.
In direct extension to the lung or ratherto the miediastinal glands following breastcarciinoma, the pleura is as a rule rapidlyinvolved through direct pleural spriead. Itis generally recogniized that cancer cells mayremain dormant in the bronchial glands formonths or even years.New Growths of the AMediastinum.-Hare,
in an excellenit analysis of 520 cases ofmediastinal disease, gave the followinigtable:-
The rest were composed of inflammatorydiseases and included inistances of lipoma,enchondroma anid gumma.The differential diagnosis is not (from an
X-ray point of view) without difficulty.Even to the clinician thie physical signis arealike in most forms anid such as there are,chiefly due to pressure effects.
Cyanosis, venious engorgemenit and signsof collateral circulatioin are possibly morecommon in neoplasm. The time factor isof service; if. signis have persisted foreighteen months or thereabouts, the evi-dence is in favour of aneurysm.The differential diagnosis has to be made
between Hodgkin's disease, aneurysm of theaortic arch, sarcoma growing either fr-omthe mediastinal glands or possibly from theremains of the thymus -or from the thyroid.
Lymdphadenomna.-It is seldom that glani-dular enllargement in Hodgkin's disease is
confined to the mediastinum; the associatedsoft and potato-like swellings met with else-where in the body are usually characteristic.A moderate eosinophilia and associatedanaemia with occasional fever and increasein the leucocyte count are,frequently in theclinical picture. The X-ray evidence is thatof a fairly well-defined mass on one or bothsides of the mediastinium.
Aneurysmn of the Aortic Arch.-Pulsationof the tumour was at one time regarded asbeing pathognomonic of this affection. Itis, however, not infrequenitly a matter ofconsiderable difficulty to decide whether alarge, pulsatinig tumour in the mediastinumwvould turn out to be an aneurysm of theaorta or a new growth with transmittedpulsation. The history, a positive Wasser-mann and examination in the obliquediameter are of much greater value; theabsence of " clubbing" being the decidingfactor in a case of aneurysm.
Sarcomtta of the Mediastinuin.-Sarcomasooner or later will be seen to invade bothsides of the mediastinal space and will beseen to cause compression or deviation ofthe trachea with definiite signs of compres-sioIn and interference with the circulation.
Tumtour of the Thyroid.-A substernalthyroid when viewed in the lateral planewill reveal itself as a tongue-like processgrowing downwards and forwards, anddepressing the arch of the, aorta. Suchtumours are of very rare occurrence.The characteristic feature of tumours of
the thyroid, whether benign or malignant,is thie definite and readily observablelateral compression of the trachea.
SURGICALRESURRECTIONS-I.
IT must have -fallen to the lot of everysurgeon of experience to meet with caseswhich proceeded to that pitch of.desperationthat it seemed impossible to believe thatrecovery cotuld take place, in which, never-theless, the forces of nature aided by thesurgical art have triumphed over~ thedepressing influences and led to recovery.In some instances life may temporarilyhave seemed extinct, but has revived againstall expectation. The relation of some ofthese cases should be of general interest,both as to !the particular events whichmay have led up to the crisis and themeans -taken to combat the profound vitaldepression.The first two cases I propose to relate
illustrate recovery after profound circulatoryfailure due to haemorrhage, and may possiblyshow something new in the technique oftreatment.
Case I-SEVERE HAEMORRHAGE- FROM THEBOWEL IN DYSENTERIC TYPHLITIS.
This case was one of the most remarkablerecoveries that I have ever seen and at thetime astonished me.. The patient was aprivate in the Middlesex Regiment, between.30 and 40 years of age. He was servingin an eastern part of the theatre of theGreat War during its last year. Thetime was mid-July and the weather wasexceedingly hot. Admitted to hospital as awalking patient, he was regarded as a trivialcase suffering. from occasional attacks ofdiarrhoea. He was giveni an aperient theday following admission, and two dayslater complained of abdominal pain anddeveloped a temperature of IOJ° F. Wllen1 saw him on that.day a tender lump couldbe felt in. the right lower. abdomeni, and inview of the history -had little hesitation inldiagnosing appendicitis. That same day I
