1061

FURTHER LIGHT ON MOUSE HEPATITIS

JANET S. F. NIVENM.D. Glasg.

G. W. A. DICKM.D., B.Sc. Edin., M.R.C.P.E.

A. W. GLEDHILLPh.D. Camb., M.R.C.V.S.

C. H. ANDREWESM.D. Lond., F.R.C.P., F.R.S.

From the National Institute for Medical Research, Mill Hill,London

THIS note is a postscript to a report (Gledhill et al.

1952) of the resolution of the setiological agent of mousehepatitis (Gledhill and Andrewes 1951) into two com-ponents-one (s) relatively stable, the other (L) more.labile. Neither component inoculated alone into weanedmice caused evident illness, but when they were com-bined fatal hepatitis was produced. It now appearsthat the L component is identical with Eperythrozot’ncoccoides-a mouse parasite related to Hcvmobartonellaand variously regarded as a bacterium or a small proto-zoon. Our interest had already been aroused by findingL present in the blood in very high titre, when histo-logical study led, as described below, to the finding ofeperythrozoon in large numbers.

Histological studies of tissues from animals infectedwith the s and L components, together or separately,have yielded interesting results. With the techniques sofar used in this study, no changes have yet been foundwith the L agent. With s small areas of hepatic necrosisoccur, which differ only in degree from the massivelesions caused by the complex infection itself ; bothinfections, and particularly the latter, were accompaniedby varied cytological changes which will be describedin detail later. The most unusual pathological featureof the experimental disease and of that due to the s

component is the appearance of an endothelial andmesothelial reaction throughout practically all the organsand tissues ; this begins within 24 hours of inoculationand has practically disappeared by the 5th day. Theendothelial reaction is focal and is restricted to venouschannels of all sizes ; it is characterised by the formationof multinucleated giant-cells which arise directly fromthe endothelium, project into the vessels, and readilybecome detached. Examination of films of peripheralblood for these giant cells in animals infected 3 dayspreviously with" mouse-hepatitis virus" revealed the

presence of enormous numbers of E. coccoides withcharacteristic ring forms. Many mice infected with s andL separately and combined have now been examined.In the L series E. coccoides appear in the blood within2-3 days of inoculation In the s series no eperythrozoahave been found. In animals infected with " mouse-

hepatitis virus " E. coccoides appear within 24 hours andincrease in numbers until death, the terminal stage ofthe disease being accompanied by many small rod andcoccal forms which may preponderate over the typicalring elements. When s and L are combined to producefatal hepatitis, the same morphological picture is found.So far it has not been possible to dissociate the activityof the L agent from E. coccoides, and, it seems likely thatthey are identical. The L agent and the eperythrozoareact similarly to chemotherapeutic agents, including

neoarsphenamine.Many speculations arise immediately from this associa-

tion ; at present it is sufficient to emphasise the followingobservations :

1.’ E. coccoides occurs in the blood in experimental mousehepatitis without previous splenectomy ; and it can be passedwithout splenectomy in normal mice more readily than isusually believed.

2. Eperythrozoa in plasma can pass through a collodionmembrane of average pore size 0·46 µ and 0·36 µ, and underideal conditions probably also through membranes of 0·3 µła.degree of filterability hitherto unsuspected.

3. Mice of the P strain, which are’relatively resistant tohepatitis, are carriers of E. coccoides, while those of the

v.s. strain, in which the disease -can be readily transmitted,are not carriers of eperythrozoa and do not yield -E7. coccoidesafter splenectomy.We have not yet elucidated the mechanism whereby

the s virus and the eperythrozoon act synergistically toproduce fatal hepatitis. It appears, however, that thes is present to about a hundredfold higher titre in thelivers of mice in which there is concurrent infection witheperythrozoa. It may be that all strains of E. coccoideswill behave as described here ; but this remains tobe proved.

REFERENCES

Gledhill, A. W., Andrewes, C. H. (1951) Brit. J. exp. Path. 32, 559.— Dick, G. W. A., Andrewes, C. H. (1952) Lancet, ii, 509.

Medical Societies

ROYAL SOCIETY OF MEDICINE

Basis of Allergic Reactions

THE pathology section of the Royal Society of Medicinemet on Nov. 18 under the chairmanship of Prof. G.PAYLING WRIGHT, president of the section, to discusscutaneous allergy.

Dr. W. N. GOLDSMITH said that eczema was an epi-dermal hypersensitivity reaction of the late or delayedtype. The process was chiefly cellular, and in -some

instances-e.g., fixed drug eruptions-the reaction wasstrictly localised to circumscribed areas of the skin.

Sensitising agents could be simple substances (evenelements such as nickel or chromium) or more complexorgdnic chemicals, such as paraphenylenediamine. Manyof these substances showed a pronounced affinity for

protein, and the actual allergen might be a complex of thesubstance with protein. Desensitisation had been

attempted with mixtures of serum and the sensitisingsubstance. Products of the skin itself-e.g., blisterfluid-had been shown to act as sensitisers.

In animals intraperitoneal injection of picryl chloridehad produced epidermal sensitisation. Where islands ofskin had been created by means of deep incisions, theapplication of such substances to the islands failed toproduce hypersensitivity, whereas application to the

surrounding skin produced hypersensitivity both locallyand in the islands. It had been suggested that this provedthat the antibodies were elaborated centrally. However, asubject who had been sensitised by the application of9-bromofluorene for 24 hours developed redness and

blistering at the site of application 13 days later ; and theprocess extended steadily towards the periphery. Dr.Goldsmith supposed that this was due to the slow diffusionof the allergen, successive areas becoming sensitised afterthe requisite time interval. This gradual progression waspossible only if the antibody was produced locally inthe skin.

In this type of reaction, passive transfer by serumusually failed, and where it succeeded the reaction wasurticarial, not eczematous. Cross-transplants of skin hadbeen made in uniovular twins, one of whom had beensensitised epidermally. The sensitised transplant lost itshypersensitivity in the normal host, whilst the normaltransplant took on the hypersensitivity of the recipient.Similarly, a- sensitised animal rendered parabiotic witha normal caused the latter to become hypersensitive.The agent was therefore diffusible, but it did not crossthe placenta and its molecules were probably large. Itdid not reside in the plasma, since passive transfer failed ;but possibly it was transported by lymphocytes. Hyper-sensitivity was not induced by injection of such lympho-cytes into the skin itself, but followed intraperitonealinjection in animals. It was not clear how the antibodypassed from the lymphocytes to the epidermis.The injection of histamine did not cause eczema, nor

was histamine released by the eczema reaction. Anti-histamine substances did not have a preventive action.

1062

Cortisone was more effective in epidermal than in dermalreactions; but local application was ineffective, althoughpatch-test reactions might be inhibited.

°

Dr. W. S. FELDBERG, r,.R.s., described a substance-, 48J80’-which on injection caused the ,;release of

histamine without tissue damage. It caused an increase.of capillary permeability, which resulted in local oadema,and permitted the escape of colloidal dyes:from the bloodinto the tissues. Animals so injected showed regional

differences in response to the drug ; the most cedematousareas corresponded with those which normally containedmost histamine. Injection of histamine did not producethis pattern of cedema, which seemed to be due to theaction of histamine at the site of release. Animalsinjected with increasing quantities of 48/80 becamedepleted of histamine so that the induction of anaphylaxiswas no longer accompanied by cedema. Also, animalsrendered light-sensitive with haematoporphyrin did notreact to exposure to light after histamine depletion with48/80. Nevertheless not all the effects of allergic reactionsshould be attributed to histamine alone, since it had beenshown that antigen-antibody reactions released anotherpharmacologically potent substance-5-hydroxy-trypt-amine (serotonin).

Prof. J. R. MARRACK discussed the antibodies involvedin allergic reactions, limiting his remarks to those foundin reactions of the hay-fever type. These were incompleteantibodies, which did not precipitate or agglutinatewhen mixed with antigen. In-vitro work with themwas difficult because of the impurities present in antigen.Even recrystallised ovalbumin might have several anti-

genic components. This could be shown by adding asolution of the antigen to a tube containing a columnof antiserum solidified with agar ; if the antigen wereimpure, a number of zones of precipitation could bemade out. It was therefore often uncertain that theantibodies detected or measured in vitro were those

responsible for the biological reaction, and in-vitro testsshould always be checked against the biological reactions.

Dr. J. F. ACKROYD described the experimental stepsby which the mechanism of ’ Sedormid ’ purpura had beenelucidated. The drug caused in hypersensitive peoplethrombocytopenia with haemorrhages throughout the

body ; this hypersensitivity persisted for 12 years ormore.

If sedormid and the whole blood of a hypersensitive patientwere mixed, there was no clot retraction (this was not so withnormal blood) ; the platelets were seen to be agglutinatedand lysed, and complement was used up in this stage. Normalplatelets were unaffected by the abnormal serum unlesssedormid were added, and platelets from the sensitised casesuspended in normal serum were not lysed by sedormid. Theserum component was found to be present in the gamma-globulin fraction of the proteins.

His hypothesis was that sedormid combined with theplatelets to form an antigen which was, however, of lowantigenicity, so that only a few people became sensitisedby it. In this small minority, taking sedormid renderedthe platelets antigenic and thus capable of union with thespecific serum antibody : their resulting agglutinationand lysis led to the thrombocytopenia. In such peoplepatch-testing with sedormid produced local purpurawithout thrombocytopenia, so that the vascular elementappeared to be independent of the platelet effect. Asimilar mechanism possibly -affected the vascularendothelium. ,

Dr. R. R. H. LOVELL discussed the effects of cortisoneand A.C.T.H. on cutaneous allergy. Patients had beengiven 200 mg. of cortisone or 150 mg. of A.C.T.H. for a weekbefore testing. Reactions to tuberculin (P.P.D.) weresignificantly diminished, as were reactions to local

applications of atropine. There was some reduction ofthe reaction to a direct irritant-manganese butyrate.Triple-responses to histamine, morphine, and grasspollen were not diminished. It seemed clear that the

hormones affected the chain of events comprising thedelayed type of reaction. ’

‘

Supervoltage TherapyThe radiology section of the Royal Society of Medicine

met on Nov. 21, under the chairmanship of Dr. CONSTANCEWOOD, president of the section, to hear two accountsof the therapeutic use of powerful X rays.

, Mr. G. W. BLOMFIELD (Sheffield) has for three yearsused a van de Graaff machine. This is an electrostaticmoving-belt generator that builds up a high potential- offrictional electricity which is applied in increasing stagesalong- an accelerator tube. The electron beam is passedthrough this tube and thereby speeded up to produce2-million-volt X rays from the target-ten times more

- powerful than conventional X rays. Mr. Blomfield saidthat three years was too short a time in which to collectstatistically valuable clinical results with a stabletechnique, particularly since the follow-up of treatedcancer had to cover many years. But survival statisticswere not the only criterion in radiotherapy. The rateat which radiation could be given, its penetration of thetissues, and the shape of the isodose curves in them wereall important. So was the patient’s comfort and theavoidance of high-dose effects, such as skin burns andnecrosis or radiation sickness. They had chosen the 2MeV instrument as the largest which still had somemanoeuvrability and was likely to be economic. Infact the total annual running cost, including salariesas well as replacements (chiefly new belts), was aboutJE3700, which worked out at about ;E8 10s. per patientor 14s. a treatment, which he thought was reasonable.The instrument produced only slight skin reactions,gave a sharply defined beam with little lateral spread,which meant greater precision in irradiating malignanttissue only, and delivered a bigger dose to the deepertissues. At 10 cm. the depth dose was twice that froma conventional 200 kV apparatus. These characteristicsmade supervoltage therapy a useful adjunct to radiumin the treatment of carcinoma of the cervix, since itcould reach secondary growths in the parietal pelviclymph-glands without harming the head and neck ofthe femur. A lead filter was used to reduce the dosein the centre of the beam, and two apphmtionw-of aboutthree minutes from different positions ensured that theglands received an effective dose, while the filter sparedthe uterus to some extent. During 1950 112 cases weretreated with 55% survival so far. including 2 patientswith a stage-4 carcinoma who had received super-voltage therapy alone. Other patients treated in 1950included 26 with infiltrating and fixed bladder carcinoma(11 still alive without recurrence) ; 20 with intra-cranial tumour (7 still alive) ; and 13 with carcinoma oflarynx (5 still alive), who received 6000r in five weeks.In general about 1000r a week was a large enough dose,but in each of 50 patients artificial menopause wassuccessfully induced with single doses of 350-450r.

Prof. J. S. MITCHELL, F.R.S., and his team at Cambridgeare using a synchrotron which is over ten times morepowerful than the van de Graaff instrument, and workson a different principle which requires no moving parts.The electron beam is made to circle faster and fasterinside a hollow ring by means.of a powerful oscillating smagnetic field, and is given further speed from passingat one point through a small coil in which a, radio-frequency alternating current flows. When it-hits thetarget 30-million-volt X rays are emitted.

Professor Mitchell explained that the synchrotronhad presented so many problems that most of the timesince its installation in November, 1949, had been spenton working out techniques and simplifying the controls.Only 5 patients had been treated between May, 1951,and March, 1952. It was difficult to make the machine,reliable, and its low output of 5-lOr per minute at afocal skin distance of 1 metre meant that treatment times

1063

were long and therefore that extensive clinical trialswould be’ hard to achieve. He had chosen 30 MeVas a useful power from -calculations made in 1946 whichshowed that as the power was increased above theconventional level the skin would be more and morespared while a greater proportion of the total radiationwould be delivered centrally where it was wanted.From this point of view the synchrotron was even betterthan the 2 MeV van de Graaff accelerator, and it gavean even sharper beam. Thus it was easy to give a highdose to a tumour while largely sparing surroundingtissues. This also simplified treatment technique, sincea single field of application often sufficed. But sincethe skin was spared it was difficult to gauge how thetumour inside was reacting to irradiation, and this lackof clinical control was disturbing. The fact that thebeam occupied a fixed position was very inconvenient.Noise was less than expected, and patients did not seemto mind it.

They had worked out the isodose curves in phantoms,measuring the dose in various ways and also using a newaudible dose-rate meter which sounded off in proportion

to the dose-rate. It was doubtful whether the r unitcould still be used in this high-power radiotherapy,because of secondary effects of unknown magnitudein the tissues-for instance, the production of positrons,neutrons, and induced radioactivity. Measurements

suggested that some of these effects were negligible insoft, and probably in all, tissues. Absorption of radiationper unit mass by horse-bone was the same as by waterto within 1%, but there was back scatter from bonewhich increased the soft-tissue dose locally by about10%. Another factor which might prove importantwas that the instrument emitted its X rays in brief

rapid pulses and not continuously.For making collimators and applicators they used

lead shot packed in plastic holders. The instrumentwas isolated in a room with walls up to 6 ft. thick,and the patient was observed through a water windowby means of mirrors and could signal to the operatorwith a bell. The cases so far treated included a parotidcarcinoma which had recurred after two separate courseswith 220 kV X rays, a malignant glomus jugularistumour at the skull base which had recurred after opera-tion, and an inoperable glioblastoma where the patientwas still alive fifteen months after irradiation. Inthis case radiation left no mark on the skin; andProfessor Mitchell wondered what had gone on withinthe skull, particularly as in a case of pharyngeal carcinomatreatment left no sign’ externally on the face and neckbut produced a visibly fierce reaction inside the mouthin the tumour area. There was no doubt that the30 MeV synchrotron had a place in therapy for suchcases as these, and also perhaps for carcinoma of thebronchus and the bladder ; it would not be suitable forthe treatment of breast cancer.

CENTRAL SOCIETY FOR CLINICAL RESEARCHAND AMERICAN FEDERATION FOR

CLINICAL RESEARCH ,

. THE Central Society for Clinical Research and themidwestern section of the American Federation forClinical Research met in Chicago on November 6-8.Some 58 papers were given.Thyrotropin in Diagnosis of Obscure HypothyroidismR. D. LEVY and W. JEFFRIES (Cleveland, Ohio) had

measured 1 uptake over the thyroid gland with and withoutinjection of pituitary thyrotropin in normal people, and inpatients with hypothyroidism, with pituitary myxoedema,and with exophthalmos developing after removal of thethyroid gland. They found that a single injection of thyro-tropic hormone (r.s.H.) produced a significant increase in therate of uptake of I131 in people with normal thyroid functionwhether or not they had been taking thyroid. Euthyroid.people, even after having taken as much as 0-26 g. of thyroid

for over eighteen months, uniformly had a significant responseto T.S.H., whereas in those with primary hypothyroidism therewas little if any increase in uptake. Normal response to T.S.H.was also seen in pituitary myxcedema and in postoperativeexophthalmos.

Nature of Circulating Thyroid HormoneW. P. Dniss, E. C. ALBRIGHT, and F. C. LARSON (Madison,

Wisconsin) had given I131 to euthyroid and hyperthyroidpersons, and thereafter separated the plasma-protein fractionsby paper electrophoresis. The pattern of radioactivity in thefractions was measured after cutting the paper into strips.In three days virtually all the radioactivity appeared in anarea just ahead of the a2-globulin peak ; the radio-activity at this site gradually increased. The same sequenceof localisation appeared in the hyperthyroid subjects. Inthem, however, the radioactivity appeared sooner, increasedmore rapidly to much higher levels, and then declined sharply.Radioactive thyroxine added to serum in a test-tube wasbound to a protein in the a-globulin section.

Raben- Westermeyer Growth HormoneJEROME W. CONN and his colleagues (Ann Arbor, Michigan)

said that Raben and associates had reported the isolation ofa hog-pituitary fraction which was growth-promoting in

animals, was not diabetogenic in adult dogs, and causednitrogen retention without effecting carbohydrate metabolism.Conn studied the metabolic effects of this preparation in a17-year-old boy with panhypopituitarism due to a calcifiedcraniopharyngioma who also had diabetes mellitus-a rareassociation. He described a metabolic state resembling thatin the Houssay rat. A detailed metabolic balance study overa twenty-four-day period showed : (a) significant anabolismof body protein with the Raben pituitary-growth preparation ;(b) no intensification of the diabetic state ; (c) no intensifica-tion of the anabolic effect of growth hormone by adminis-tration of insulin ; (d) a definite protein-anabolic effect ofinsulin alone in the diabetic pituitary dwarf, who withoutinsulin maintained nitrogen equilibrium. Conn believed thatthe Raben preparation, when given intramuscularly at a pHbelow 4, could be expected to induce growth in human

pituitary dwarfism without inducing diabetes mellitus.

A.C.T.H., Cortisone, and Aspirin in Rheumatic FeverH. B. HOUSER and other colleagues of C. H. RAMMELKAMP

(Cleveland, Ohio) gave a preliminary report of their long-term study of the comparative effects of A.C.T.H., cortisone,and aspirin on the course of acute rheumatic fever.152 young airmen with acute rheumatic fever were

observed for nine weeks. Treatment with the three agentswas started during the first week of illness and was continuedfor six weeks. The daily dose of A.o.T.A. was 120 rag. initiallyand was reduced at specified intervals so that at the beginningof the sixth week the dose was 20 mg. daily. Intramuscularcortisone was reduced gradually from 300 mg. on the firstday to 50 mg. per day during the sixth week. Aspirin wasreduced during the first week from gr. 1 per lb. body-weightper day to a maintenance dose of gr. 1/2 per lb. per day.Aspirin was most effective in relieving symptoms and signsof joint involvement. Fever subsided most promptly afteraspirin therapy, although there was no great difference amongthe groups in the prevalence of fever after the fifth day oftreatment. A.C.T.H. was most effective in returning the

erythrocyte-sedimentation rate to normal. Abnormal auriculo-ventricular conduction occurred with similar frequencyamong the three, groups, but the duration of abnormalitieswas longer in the aspirin group. There was no distinct differ-ence between the three groups in the appearance or dis-

appearance of murmurs or of other signs of acute carditis.After treatment clinical or laboratory relapses occurred in themajority of patients in all groups ; with few exceptions theserelapses subsided promptly. None of the drugs used had anygreat advantage over the others in the early treatment ofrheumatic fever.

Conversion of Mesobilirubinogen to StercobilinogenPAUL LOWRY, R. ZIEGLER., and CECIL WATSON (Minne-

apolis, Minnesota) said that Baumgartel, in Germany, hadclaimed that cellular dehydrogenases converted bilirubin tomesobilirubinogen in the liver, and that faecal bacteria con-verted bilirubin directly to Stercobilinogen- Since the keystoneof Baumgartel’s hypothesis was the supposed inability offaecal bacteria to convert mesobilirubinogen to stercobilinogen,Watson and his colleagues studied the effect of such organ-isms on N15-labelled crystalline bilirubin. They showed

1064

convincingly that faecal bacteria reduced both bilirubin andmesobilirubinogen to stercobilinogen. Since Watson hadrepeatedly crystallised mesobilirubinogen from human faeces,there was little reason to postulate an extra- intestinal site forits formation, or to doubt its intermediary relationship betweenbilirubin and stercobilinogen.

Refractory Macrocytic AnaemiaD. L. HORRIGAN and ROBERT W. HEINLE described a new

disease entity-a refractory macrocytic ansemia with a defectin vitamin-B12 binding which responded to normal plasma.A patient who had had unexplained anaemia for eighteen yearsdid not respond to crude liver-extract, nor to iron. Free HCIwas present in fasting gastric contents. Heamatological find-ings during relapse included macrocytic anaemia, reticulo-cytosis, and normoblastic bone-marrow hyperplasia. Slightbut significant remissions were obtained with single injectionsof normal human plasma or with single injections of vita-min B12. Such remissions were not maintained either withvitamin Bi2 intramuscularly every one to two weeks, or withoral folic acid 10 mg. daily. Normal blood levels were main-tained, however, when the patient was given vitamin B1°combined with fresh human plasma, at short intervals.

Healthy persons and the patient were studied after intra-muscular administration of 0-5 (Lg. ofCo6O-labelled vitamin B12.In the normals 0-10% of the injected vitamin was excreted inthe urine, and altogether 60% was excreted in twenty-fourhours. These observations suggested that : (a) the anaemiain this patient resulted from deficiency of vitamin B12 causedby failure of proper binding of the vitamin ; and (b) normalplasma contained a factor which corrected this defect.

Haemophilia-like Disease after PregnancyPAUL G. FRICK (Minneapolis, Minnesota) described an

instance of a haemophilia-like disease following pregnancywith transplacental transfer of an acquired circulating anti-coagulant. The disease appeared in a woman three monthsafter her first normal pregnancy and was studied from thetime of its onset until recovery two and a half years later- sixteen months after sterilisation by X rays. The venous

clotting-time was markedly prolonged, and the prothrombinconsumption was minimal. Bleeding-time, prothrombin-time, prothrombin concentration, platelet-count, clot retrac-tion, tourniquet test, and plasma-fibrinogen concentrationwere all within the normal range. There was no evidence of

parenchymal liver dysfunction. The lack of response of the

clotting abnormality to fresh blood or plasma led to the

suspicion of a circulating anticoagulant, which was demon-strated by the delaying effect of the patient’s blood on thecoagulation-time of normal blood. Cohn’s fraction I containingactive antihaemophilic globulin had no effect on the patient’sclotting mechanism. The anticoagulant was therefore believedto inhibit the action of the antihaemophilic globulin (thrombo-plastinogen). The patient delivered a second child fifteenmonths after the onset of the disease, and the anticoagulantwas demonstrated in the baby’s blood during the first twoand a half months of its life. The findings strongly suggestedan immunological mechanism in the development of the

anticoagulant.Endothelial Permeability to Albumin

J. A.’ SCHOENBERGER, H. V. KROLL, and ROBERT M. KARK(Chicago, Illinois) reported on their studies in man on endo-thelial permeability to radioactive-iodine-tagged albumin

injected intravenously or intraperitoneally. They used

1-tagged human albumin to study endothelial permeabilityin health and in various diseases associated with ascites andoedema. The material was injected in tracer doses first intra-venously and a few weeks later intraperitoneally. In healthyvolunteers the intraperitoneal injections were made throughthe posterior vaginal fornix or through the anterior abdominalwall after a small pneumoperitoneum had been established.The studies disclosed a dynamic equilibrium between albuminin intramuscular and extravascular sites. Mathematicalanalysis of the curves yielded quantitative estimates of therate of exchange of albumin across the peritoneal membraneand also out of the vascular endothelium. Quantitativedifferences in permeability between healthy individuals andpatients with ascites were found. In patients with ascitesrepeated studies were made during reablement. Thesedisclosed changing permeability which was correlated withelectrophoretiø studies of the passage of various-sized proteinIllO1e’cules from the plasma to the ascitic fluid. Large proteinmolecules (such as globulins) distributed themselves equally

between plasma and ascitic fluid when permeability waspoor, but did not pass the endothelial barrier freely whenpermeability was improved.

Concentration of. Antibiotics in Brain .

W. E. WELLMAN and his colleagues (Rochester, Minnesota)had sought to find out whether neopenil (the diethylamino-ethyl ester hydriodide salt of penicillin G), a new form ofpenicillin for parenteral use, was concentrated in the brain.The patients they studied were in hospital because of psychi-atric disorders, and prefrontal lobotomy was done for thera-peutic reasons. Before operation, these patients received forprophylaxis one of the following drugs : neopenil, procainepenioillin G in aqueous suspension, aureomycin, terramycin,dihydrostreptomycin, or erythromycin. They found that onlynoopenil and aureomycin were consistently detectable insignificant amounts in brain tissue.L.E. Phenomenon in Penicillin Hypersensitivity and

Serum-sicknessJOHN R. WALSH and HYMAN J. ZIMMERMAN (Omaha,

Nebraska) had studied 8 patients with penicillin reactions and1 with serum-sickness following the administration of tetanusantitoxin. Of the 8 patients with penicillin reactions, 4 hadmild disturbances characterised only by urticaria ; the other4 had fever, adenopathy, joint pain, and skin lesions whichranged from -erythema multiforme to a generalised exfoliativedermatitis. In each of these 4 patients with severe penicillinreactions the plasma L.E. factor was demonstrated, and,furthermore, L.E. cells were found in the concentrated hepar-inised bone-marrow. Study of the 4 patients with mildpenicillin reactions failed to reveal the L.E. plasma factor orL.E. cells in heparinised marrow. Likewise, study of .theheparinised bone-marrow of the patient -with serum-sicknessrevealed L.E. cells ; but an attempt several days later todemonstrate the L.E. plasma factor was unsuccessful.

New Inventions

AN ADAPTER FOR AN INTRAVENOUS NEEDLE

J. MIDDLETON PRICEM.B. Lond.

Formerly ofDepartment of Anæsthetics,

London Hospital

MANY anaesthetists consider that a cannula of theGuest type is the easiest to introduce into a vein, andthat it causes less damage to the vein wall than anintravenous needle.The use of the adapter shown in the accompanying

figure will allow this cannula to be used for intermittentinjections and yet remain patent should an intravenousinfusion be required. The adapter consists of an inner

cannula with a rubber diaphragm at one end, similar toa Gordh needle.Venepuncture is performed with the cannula and

introducing needle connected to a syringe. The syringeand needle are then steadied and the cannula pushed ohup the vein. The needle is then removed and the adapterinserted. Should the adapter become blocked (thoughthere is very little tendency for this to happen) it mayeasily be replaced. Two sizes of adapter have beenused, one to fit a standard case Guest cannula and oneto fit a 19-gauge cannula of similar design, which isused for the back of the hand in adults and for children.My thanks are due to Dr. J. Challis and the anaesthetists

of the London Hospital for their encouragement in thedesign of this needle, and to Messrs, A. L. Hawking whohave made the cannula sets. __

,