391

The cut edges of the duodenum and stomach are gastro-jejunostomy for the relief of chronic duodenal sutured in the manner of an end-to-end anastomosis. ulcer with an unobstructed pylorus would soon be I always use thin sterilised silk for this purpose and abandoned. for ligatures. It is tedious and unnecessary to Experience proves that posterior gastro-jeju- describe in detail the methods available for sewing nostomy with an obstructed pylorus is a beneficent the cut end of the duodenum to the hole in the operation, in spite of the risk the patient runs of stomach. It is like stitching a sleeve in a shirt; getting a new ulcer for an old one. The new ulcer the clever sempstress varies her methods according has been evolved in this generation by alterations to the material and the size of the garment. So in the environment of the jejunum brought about with the In a case of extensive resection by surgery. surgeon. ______(Fig. 3) the line of suture in the stomach was 6 inches across. Not wishing to imperil the safety .of such a long line of suturing by joining the TRENCH NEPHRITIS.1 I the cut of the duodenum to it, implanted margins BY W. LANGDON BROWN, M.D. CANTAB., into a fresh in the duodenum opening posterior F.R.C.P. wall of the stomach. In of the careful LOND., spite ligature ASSISTANT PHYSICIAN, ST. BARTHOLOMEW’S HOSPITAL; PHYSICIAN, of vessels there is usually some post-operative METROPOLITAN HOSPITAL ; CAPTAIN, R.A.M.C. (T.F.) oozing from the tissues in the wound area, and it is wise to drain for 48 hours. The abdominal IN May last the Medical Research Committee, at wound is closed with through-and-through sutures, the request of the Director-General of the Army fascia into with inter- the being brought apposition Medical Services, made arrangements for a number mediate sutures of thin silk. of cases of trench nephritis to be investigated in The first I submitted to for patient pylorectomy wards under my care at the St. Bartholomew’s duodenal ulcer was a woman, aged 35, at the Hospital Section of the lst London (City of London) Middlesex in 1909. Hospital Gastro-jejunostomy General Hospital. More than 100 cases have been had been three months but performed previously, under my care, but of these only 58 were specially the pain remained unrelieved. I excised the investigated. pylorus and the segment of the duodenum contain- It is difficult to choose a name for the condi- ing the ulcer. She made a quick and uneventful tion which shall not beg the question of its I have since lost of this " recovery. sight patient. origin. "Epidemic nephritis" assumes what I A few when on a man for a days later, operating am trying to prove. " Trench nephritis " is not duodenal I found the conditions so favourable ulcer, strictly accurate, since I have had cases among for pylorectomy that I excised the pylorus and orderlies from base hospitals, and others who have the bulbus duodeni with the ulcer. Six years not been in the trenches. Still, the condition later I re-examined him with the help of X rays cannot be said to be prevalent outside the trenches, and an opaque meal, and found the gastric efflux so the name will serve. satisfactory. Acute nephritis is not a common disease in civil I have excised the pylorus for chronic duodenal practice; it is not possible to give exact figures, ulcer to the method described in 2Q according since it is not a notifiable disease, but Sir Wilmot patients. In three of them I had previously per- Herringham found that at St. Bartholomew’s formed In one the gastro-jejunostomy. symptoms Hospital, with an average number of 7000 medical returned three months after the gastro-jejunostomy, admissions a year, only 166 cases of acute nephritis, in another and in a third two One six years, years. 120 being males, were admitted during nine years. the sixteenth in the died. He was a patient, series, I found that in a period of five years only 26 cases man from Dr. Essex aged 38, transferred Wynter’s occurred here in men of military age. ward to mine with typical symptoms of chronic Now whereas typhoid and dysentery have in the an ulcer was duodenal ulcer. At the operation past dogged most campaigns, acute nephritis, as found close to the pylorus on the anterior wall of a rule, has not. I have not been able to find that a duodeni. I from well-marked bulbus departed it was prevalent in the Franco-Prussian War, the my usual method and sewed the cut edges of the Chino-Japanese War, the Soudan War, the Hispano- The man made mucous membrane with catgut. American War, or the Russo-Japanese War. It was fifth when his tem- :good progress until the day, certainly rare in the South African War. Thus, the perature rose to 103° F. ; he died on the nineteenth late Dr. Washbourn reported that " acute nephritis The suture line had day after the operation. was singularly infrequent." At Deelfontein there in two I attribute this unfavour- sloughed spots. were only 6 cases of acute nephritis. Dr. Tooth able result to the use of catgut. reported that at the Portland Hospital they only SUMMARY. saw 3 cases, and in these the kidney lesion was Since the treatment of duodenal ulcer passed almost certainly not acquired during the campaign. into the province of surgery it has become the Personally, I did not see a single case in South routine practice to perform gastro-jejunostomy for Africa. its relief in the hope that by diverting the chyme On the other hand, there was a very considerable through the new stoma into the jejunum the ulcer outbreak of the disease in the American Civil War. will heal. If the pylorus is obstructed by the ulcer I have drawn up a curve (see figure) by adding the results are usually good, because the chyme together the cases classified as general dropsy, must pass through the new stoma; but when the Bright’s disease, and inflammation of the kidney. pylorus is patent the chyme flows through it This is necessary, as the diagnosis of Bright’s and in some instances ignores the new route. The disease, which was not made till September, 1862, efflux can be watched with the help of an opaque gradually replaced that of general dropsy, which meal and X rays. I believe it is better whenever was discontinued in December, 1863. The curve practicable to excise the pylorus with the ulcerated will therefore include cases of chronic nephritis, but portion of the duodenum and rejoin the stomach this would not affect the extraordinary rise which and duodenum on the of an end-to-end principle 1 Being an introduction to a discussion held by the Medical and anastomosis. If this method could be made safe Therapeutical Sections of the Royal Society of Medicine on Feb. 15th. 392

I ..... I ...... i Curve showing incidence of nephritis per 1000 troops during first part of American Civil War among the Northern Armies. The armies of the Central Region were attacked earlier and more severely by the disease. Black continuous line = central region. Dotted line = rest of army. took place between March, 1862, and March, 1863. imagine that theory died a natural death as the The case-incidence per 1000 troops rose earlier in summer came on and the number of cases increased. the armies of the Central Region and reached a Still as I am not sure that exposure has been given higher level than that for the rest of the army. up as the cause, I will add the following arguments. In subsequent years there was no similar rise, but a When I was in Bloemfontein during the South small secondary one during the warmer months of African War I saw the mirage between 2 and 4 P.M., 1864. In the Central Region the incidence reached while by 6.30 the ground was covered with hoar- 1’5 per 1000. Throughout the war there were 14,187 frost. There could hardly have been such a violent cases (see table). It is of great interest to recall fluctuation of temperature as this in Flanders; that in the American Civil War the military condi- yet, as I have already said, nephritis was rare in tions resembled those of the present war in some the South African War. In the Russo-Japanese important particulars. There was the sudden dash War during the month of February, 1905, the ther- of the Confederates northwards, which was checkedi mometer showed great differences between the and followed by prolonged . minimum and maximum readings, such as from 14° to 83° F. and from 26° to 75°. Yet I do not read Cases in American Civil War in Northern Armies. of Nephritis that there was an outbreak of Periods. Cases. Deaths. nephritis. Again; Average strength. look at the incidence in the American Civil War. May-June, 1861 ...... 42,500 ... 27... 0 How can a climatic cause account for an outbreak July-June, 1861-62 ...... 279,000 ... 1,790 ... 45 which started in March and lasted for a whole " " 1862-63 ...... 614,000 ... 6,603 ... 148 but was not in the ...... which " z 1863-64 ...... 619,700 2,677 81 year, repeated subsequent

" ., 1864-65...... 574,000 ... 2,744 ... 77 years of the war ?? However, I made careful

... 346 ... 9 " 1865-66...... 101,800 inquiry as to exposure in the 58 cases which were the subject of special investigation ; in 20 cases the - Totals , ...... 14,187 ... 360 men were sure they had been specially wet before the illness whereas in 38 cases no such The present epidemic in the British Expeditionary began, could be obtained. I doubt very much if Force is not assuming such dimensions as these. history mere exposure damages healthy kidneys. On the Up to the end of June, 1915,1062 cases had occurred. other hand, I do think that exposure will exacer- Very few were reported until February, in which bate chronic nephritis, and I have seen a return of month 72 cases occurred, March, 220 in 138jn April, the albuminuria in some of these cases when 211 in and 326 in June. Since then the appa- May, convalescent if the cold. returns have shown a increase. rently patient got Probably monthly steady the theory of nephritis starting from cold obtained It is interesting to note from the discussion on the currency in some such way. The trend of modern subject in Vienna that there has also been an research has been to discredit climate as a cause of outbreak among the German and Austrian troops, disease, past beliefs are enshrined and that it began about the same time as in our epidemic though in such names as malaria and influenza. army. ° Many other causes have been suggested ; of these Etiology. the question of the water-supply demands con. It is a hallowed tradition in medicine that sideration first. There is the possibility of the exposure is a very important cause of nephritis. water having been derived from an unwholesome Yet we notice that while the troops were enduring source, or that chlorination affected the water, or much cold and exposure during the bad winter of some metallic poison might have been derived from 1914-15 there were but few cases of nephritis then. the galvanised water-carts or the water-bottles. It was not until the weather was better that the My cases lend no support to any of these hypotheses; disease assumed epidemic proportions. But so the sources from which the patients had obtained ingrained is the tradition that attempts were made water were extremely variable. Moreover, some had to show that although continued cold was not been careful about boiling the water, others had responsible the violent fluctuations of temperature not ; in some cases the water had been chlorinated, in the early spring played an important part. I in others it had not. , 393

Excessive protein diet has been suggested as a March in each case, and the close similarity ’ possible cause. It seems highly improbable that between the individual cases suggest some common 180 grammes of protein would cause acute nephritis cause and one not operative in all wars. The in a number of otherwise healthy men. If it did, simplest explanation would be a specific infection. it could only be through setting up an intestinal In attempting to elucidate this, our first step was ttoxsmia. Mr. Mackenzie Wallis’s observations are to try by chemical tests whether the nephritis was definitely against the presence of an intestinal of the infective or toxic type. Some authorities roxaemia, since he found neither an increase in the claim that the infective type is glomerulo-tubular ethereal sulphates nor the presence of urobilin in in distribution, while a toxic nephritis, such as that the urine. produced by corrosive sublimate, is mainly tubular. Acidosis has also been suggested as a cause on I may say I am rather sceptical as to the value of account of the frequency with which dyspnoea. this distinction, but so far as it is valid our results occurred in the early stages of the disease. It is would point to the involvement of both glomeruli not clear how acidosis could arise on the diet given and tubules. I do not propose to go into details, to the troops, nor is there a general agreement that but I may say that the tests we employed were the acidosis can cause nephritis. Acidosis was not a chloride test, the iodide test, the lactose test, and constant feature of the cases with dyspnoea in which the diastase test. Mr. J. W. Trevan carried out we investigated the COa content of the alveolar air. the first three and was able to show that both In the discussion held at Vienna on this outbreak glomeruli and tubules were involved, and Mr. various other causes were suggested, even including Mackenzie Wallis found by the diastase test not the deliberate taking of cantharides or chromic only that there was tubular damage, but that the acid by soldiers in order to render themselves unfit degree of that damage could be estimated in this for further service. The combination of alcoholism way, and that the test was thus a valuable aid in and exposure was also suggested. However plausible prognosis. The cases that have a low diastase alcoholism might be as a cause in German troops, it output after the cessation of albuminuria cannot would not explain the outbreak in ours. Dysentery be regarded as completely recovered and are liable and typhoid were suggested as causes, but other to relapse. speakers called attention to a lack of agreement ]}[01’bid Anatomy. between the curves of incidence of the intestinal The morbid anatomy of the kidney in a few fatal and the kidney conditions. Moreover, dysentery cases was found by Professor F. W. Andrewes to and were in typhoid rife South Africa, yet nephritis show signs of a subacute diffuse nephritis. All the was rare. Some speakers regarded the condition as glomeruli showed evidence of inflammation, and an exacerbationof chronic nephritis, but this would there was extensive damage to the convoluted only apply to a few cases. Thus in my series there tubules. The interstitial tissue of the kidney was was a previous history of nephritis in only five. oedematous and infiltrated with lymphocytes and Moreover, there is no reason to suppose that there polymorphs. There was no fibrosis and no micro- are a larger proportion of men with damaged organisms could be found. Pyelitis was not found. kidneys engaged in this war than in other wars. Antityphoid inoculation has also been held respon- Pathology. sible, but there is no evidence that the injection of As there was thus complete evidence of glomerulo- immune bodies can excite nephritis, nor is this con- tubular changes an infective process was indicated. sistent with the rarity of the disease in South Africa Blood cultures were, however, sterile. There and.among the inoculated troops training at home. remained the possibility of a local infection from The French army surgeons report that there which toxic products were disseminated. Attention have been a certain number of instances where the was naturally first turned to the urine. Catheter disease was simulated by putting white of egg into specimens were obtained in 21 consecutive cases the urine and injecting water under the skin. They and cultures made. In 18 of these the urine proved have also suggested that the disease was due to a to be sterile. From the 3 remaining cases B. coli suppressed form of . It is interesting were obtained, but in two of them only with diffi- to note that very few, if any, cases have occurred culty. I mention this specially as Klein and Pulay among the Indian troops, and the natives of India in Vienna and MacWalter in Ireland lay stress on are said to enjoy comparative immunity from B. coli as the causal agent. Our cultural and histo- scarlet fever. It is also interesting to note that logical results are definitely against this conclusion. otitis media, so common in scarlet fever, developed Throat cuUu,rcs.-Of the cases 17 complained of just before the oedema in two of my cases. Sore- a sore-throat as an early or initial symptom. throat was also not uncommon at the outset; 17 of Tonsillitis is known to be associated with nephritis, my 58 cases complained of this. Mr. Mackenzie so cultures were made from all the throats where Wallis has made a comparative study of trench soreness was complained of, and in several where nephritis and scarlatinal nephritis, and has shown no such complaint had been made. In every in- many points of similarity, both in the lesions pro- stance streptococci grew in abundance, sometimes duced and in the characters of the urine. But we practically in pure culture. Pneumococci were do not believe that the two are identical. If the found in 5, and diphtheroids and micrococcus same percentage of nephritis occurred in this pre- catarrhalis were also found in some cases. Before sumed suppressed scarlet fever as in the ordinary it could be concluded that the streptococci were type this would mean that there had been about responsible in any way, it was necessary that con- 10,000 cases of suppressed scarlet fever among the trols should be made, since streptococci are normal troops, which appears to me an impossible assump- inhabitants of the throat. Dr. R. G. Canti there- tion. Moreover, even in "scarlatina sine eruptione"" fore made cultures from the throats of 10 soldiers desquamation follows, which has not been observed with nephritis, 10 wounded soldiers who had not in these cases. got nephritis, and 10 male civilian patients of A point which deserves consideration is the military age suffering from various surgical comparative rarity of the disease among officers. diseases. He found that the soldiers’ throats were, The rapid increase in the number of cases in this as far as bacteriological examination went, in a war and in the American Civil War, starting about healthier condition than the civilians, while the 394

throats of the soldiers with and without nephritis these cases were not so ill that either the kidney were closely similar. He also estimated the or the heart was failing, and none of them showed streptococcal antibodies in the blood of some of signs of cardiac dilatation. The close agreement these cases, but found no evidence of any excess, between the onset of oedema and dyspnoea suggests which is an argument against the existence of a that the outpouring of fluid into the lungs or recent streptococcal infection. pleural cavity may have been responsible for the We next made investigations as to infection by . This is supported by the an ultra-microscopic filter-passer. Mr. Mackenzie frequency with which a cough and also Wallis has produced a definite illness in rabbits and occur. Some of the cases showed a return of monkeys after an incubation period of eight days respiratory distress after their journey from France, by injections of urine, prepared so as to be free from A slight and irregular temperature was common bacteria, but containing any filter-passer present. in the earlier stages, and was sometimes prolonged These results demand consideration, even though where recovery was slow. The blood pressure was he has not succeeded in producing nephritis, for very variable, being raised in some cases but not in the incubation period of eight days excludes a mere others. I did not find that it was much help in chemical cause for the illness and corresponds to prognosis, though probably the best thing is for the the incubation period which has been estimated blood pressure to be raised at the beginning and clinically in some cases. In this connexion the soon to fall to normal. In the discussion at Vienna, results with Wassermann’s reaction are of Schlesinger noted a curious fact-that in a great interest. The reaction was positive in 18 out of number of cases hypertrophy of the heart did not 56 cases, being strongly so in 4 and feebly so in 3- supervene. i.e., it was positive in practically 1 in 3-a higher I noted a great variability in the amount of urine proportion than that of syphilis in the army in secreted. Often it was markedly increased while general. This may mean that men damaged by the disease appeared to be fairly acute-i.e., there syphilis are more liable to nephritis, or that the was hyposthenuria or an inability to secrete a con- positive reaction was not due to the spirochseta centrated urine. It was certainly a more marked pallida, but to some other animal infection, such as feature of these cases than in the acute nephritis a filter-passer is generally assumed to be. met with in civil practice. I did not observe, even in early cases, the extreme reduction in the . ; quantity secreted, such as is usual in ordinary acute I intend only to call attention to those features nephritis. Red blood corpuscles were found in 44 of the 5B of the disease which seem to the I peculiar present cases ; the amount was very variable. The urine outbreak. But there is one on which it is point was sometimes bright red with blood, sometimes that we should evidence from very important get merely smoky, and often chemical and those medical officers who have seen cases in the microscopical tests were needed to detect the blood. As the There are a number of febrile condi- early stages. haematuria cleared there was a tions the and it would be up frequently occurring among troops, flocculent reddish-brown in which alone to know whether acute is precipitate interesting nephritis the blood could be found. I was struck with the to occur among those who have had specially apt fact that in a good many cases blood continued to any such febrile attack. Lieutenant F. G. Chandler, be poured out in some long after all other medical officer to the 2nd Argyll and Sutherland quantity acute symptoms had subsided, as though the told me that in several cases there Highlanders, damaged kidney continued to leak blood rather had been diarrhoea of sudden onset, or a tem- than that there was a inflammation of with in the limbs and haemorrhagic perature 103° F., pain of the kidney. for 24 to 48 hours before he found ------shivering There was a considerable number of albumin in the urine. Other medical officers have generally in the as there was oedema without albuminuria at all. In leucocytes urine, just leucocytic reported infiltration of the kidney in the fatal cases. At the of cases the first thing that the patient majority first I thought this was a feature special to this was the which started in noticed oedema, usually form of nephritis, but Mr. Mackenzie Wallis finds the face and legs and remained curiously localised that it is equally marked in scarlatinal nephritis. sometimes, more it became though frequently Casts were found in all but 3 cases ; usually there ( was almost a constant generalised. was a mixture of epithelial, granular and hyaline feature, in 53 out of the 58 cases; being present casts. Blood casts and fatty casts, on the other and of the 5 in whom it did not occur 4 gave a hand, were uncommon. Crystals were rare in the of In all my cases the history previous nephritis. urine. Calcium oxalate was found twice, and cedema soon cleared up. ... - - of uric acid once. The rarity of organisms One of the was crystals most striking points the frequency in catheter specimens has already been referred to. of as an it was dyspnoea early symptom; definitely As to complications, fits presumably uraemic in present in 49 cases. Of the 9 who did not com- nature occurred in 4 of the 58 cases. They all of it, 2 had no oedema, while the cedema was plain made a good recovery, and I was of opinion that slight in 2 others. As a rule, the shortness of free venesection contributed to this recovery. Mr. breath started at the same time as the oedema, but Foster Moore examined the eyes of patients under did not last so long. The alveolar CO2 was deter- the care of my colleagues and myself, and reported mined by Mr. Trevan in 5 cases during the that he found retinal haemorrhages in some of the In 3 it was normal, while in the remain- dyspnoea. acute cases, but that retinitis occurred in 2 it was reduced from the normal 5 to 3 only ing per those cases which had become chronic.’ cent. Since due to acidosis is definitely dyspnoea always Abercrombie has observed some interest- associated with reduction of alveolar CO2, and Captain ing complications bearing on a possibly infective uraemic dyspnoea is of this type, it would appear origin such as herpes and parotitis. that the shortness of breath was not always asso- Milkiness of the blood serum referred to by ciated with acidosis, ursemic or otherwise. I do Bright in his original account of the disease was not think that dyspncea is usually regarded as a not observed in any of the 56 cases in which the common feature of acute nephritis, apart from uraemia or cardiac failure. But the majority of ( 2 THE LANCET, Dec. 18th, 1915, p. 1348. 395

blood was drawn for Wassermann’s reaction, but agent, as does the unequal degree to which the was present in a fatal case under Major Calvert’s different units were affected. This is supported by care. the extent to which the glomeruli were involved, Course and Prognosis. as shown by the chemical tests and the post-mortem On the whole it be said that the disease has evidence. may 5. So far bacteria have not been found run a favourable course, although I assume that I responsible for the infection, more evidence only got the more protracted cases sent home from though positive France. A patient who developed the disease might be obtained from earlier cases. Results of a have been immediately after he returned from France on suggestive possible filter-passer but are not conclusive. The leave and who was admitted to hospital on the obtained, they point of entrance of the infection is the tonsils second day of the disease was free from albuminuria possibly and I that there is a on the fourth day. Only one of the 58 cases died, upper air-passages. suggest which resembles but and in no other case has life appeared to be in specific organism responsible is not identical with that scarlet fever. Most of the cases have felt quite well as causing danger. 6. The clinical features of the soon as the oedema subsided, though some com- outstanding plained of headache, cough, and digestive disturb- epidemic are : (a) The frequency with which short- ness of breath occurred at the ances for a week or more. Yet many of those who the beginning; (b) in the the felt and looked quite well were still passing great variability duration; (c) tendency albumin, blood, and casts. As they recovered the to remission and relapse; and (d) the ultimately albuminuria often became intermittent, and there favourable prognosis in the majority of the cases. has been a curious tendency to remission and Suggestions.-I would make the following sug- relapse. Some of those who had apparently gestions :- recovered and had gone to convalescent homes 1. That animal experiments with the saliva and soon relapsed. Often the patient has been able to urine from recent cases should be carried out. The tell from his own sensations whether the albumin lack of definite evidence so far may be due either to has returned or increased. It took as much as the infective stages being over before the patients 14 weeks for the albuminuria to cease in some reach England or to the difficulty in inducing cases, and others continued to pass blood and nephritis in the animals employed. albumin even five or six months after the onset. 2. That investigations should be made as to the The diastase reaction in the urine is a valuable exact distribution of the cases. This might be aid to prognosis. Not only are those cases worse in done by means of maps on which the cases could which the diastase output is lowered or absent, but be dotted according to the place of origin. There if the lowered output persists after the albumin- is an impression that certain billets are par- uria has ceased the patient seems liable to relapse. ticularly associated with the disease. A map would A case therefore cannot be regarded as cured until afford valuable evidence as to the kind of place the diastase output has returned to normal, and I where the disease is contracted, and therefore as to would suggest that it is important for this test to the possible source of the infection. Until this is be done in all doubtful cases before return to duty. done it would be useless to try to isolate the cases ; Schlesinger is of opinion that the subjects of probably they are only infectious in the early heemorrhagic nephritis are no longer fit for active stages, and are so no longer when they reach the service, though Strasser thinks that they may still base hospitals. Yet some may still be infectious, be used for less strenuous military duties. On since orderlies at these hospitals have contracted active service such patients are liable to relapse. ; the disease. In the meantime it might be well if the urine and sputa were treated as the excreta in Treatment. the cases of typhoid. I will say little on this head, as I have really 3. That arrangements should be made for the nothing fresh to add to our common knowledge on urinary diastase to be estimated in doubtful cases this. As nitrogen retention is a common feature before return to duty is allowed. of the early cases, I have adopted a diet poor in I would urge that we have a remarkable oppor- nitrogen though varied and fairly liberal in carbo- tunity of studying the pathogeny of nephritis in hydrates and fat for limited periods in the earlier general and of rendering aid to the efficiency of our stages. Sometimes this has produced distinct forces by checking a disease which incapacitates a benefit, but sometimes it seems to make little or no number of otherwise healthy men. The next difference. I am convinced that prolonged nitrogen chapter in the inquiry will have to be carried out in starvation is as bad for the nephritic as for Flanders. I feel sure that anyone investigating the anybody else. disease will find, as I have done, a sympathetic helper in Dr. Walter Fletcher, the secretary of the Conclusions and Suggestions. Medical Research Committee. Conclusions.-1. These cases are true examples In conclusion, I should like to express my of acute nephritis, as shown by (a) the combination indebtedness to the staff of the pathological of oedema with albuminous urine containing department of the hospital under the direction of casts; (b) the nature of the proteins in the urine; Professor Andrewes for their willing cooperation, (c) the tests for renal permeability; and (d) post- and to Mr. Mackenzie Wallis and Mr. Trevan, who mortem evidence. made the chemical investigations. 2. The cases have been too prevalent to be merely Since writing this paper I have seen Sir John Rose accidental, and the close similarity of the symptoms Bradford’s note on the subject in the Quarterly suggests a common cause. Journal of Medicine for January, 1916. It is 3. Exposure, water-supply, metallic poisons, or gratifying to find that from a more extended intestinal toxæmia can be excluded as the cause. number of cases he has come to conclusions The prolonged cases showed a high percentage of practically identical with those I published in the positive Wassermann reactions. July number of the Journal of the Royal Army 4. The curve of incidence both in this war and Medical Corps. in the American Civil War suggests an infective Welbeck-street, W.