I4^ic^^^ Clinical and Pathological PAPERS FROM THE LAKESIDE HOSPITAL, CLEVELAND. RA SERIES 11. 1905. CORNEL L UNIV ERSITY THE l^lompr torrtnarg Sjtbrarg '^'J FOUNDED BY ]^^^ ROSWELL P. FLOWER y , O fo"" the use of the N. Y. State Veterinary College This Volume is the Gift of Dr. V. A. Moore. 356 CORNELL UNIVERSITY UBRARY 3 1924 104 225 549 The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924104225549 CONTENTS. CLINICAL PAPERS. 1. Paracolon Infections, with Report of Thrte Cases. — By Herbert W. Allen, M. D. 2. The Character of the Widal Reaction in the Present Epidemic of Typhoid Fever. — By L. W. Ladd, M. D. 3. Copious Water Drinking and Polyuria in Tjrphoid Fever. — By Edward F. Gushing, M. D., and T. W. Clarke, M. D. 4. Stomatitis in Impetigo Contagiosa. — By Edward F. Gushing, M. D. 5. A Report of Two Cases of "Typhoid Spine." — By George Wilton MOOREHOUSE, M. D. 6. Repeated Copious Haemoptysis from an Aortic Aneurism, extending into the Right Lung and finally rupturing into the Pleura. — By Thomas Wood Clarke, M. D. 7. Hydrophobia with Report of Three Cases. — By Henry S. Upson, M. D. 8. The Hereditary Element in Cataract.— By B. L. Millikin, M. D. 9. Wounds of the Thoracic Duct occurring in the Neck. Report of Two Cases. Resume of Seventeen Cases. — By Dudley P. Allen, M. D., and G. E. Briggs, A. M., M. D. 10. Laparotomy for Perforation in Typhoid Fever. Report of Six Gases.— By C. E. Bmcc-,, A. M., M. D. 11. Intestinal Perforation during the Ci^urse of Typhoid Fever, and its Surgical Aspects.— By C. E. Briggs, A. M., M. D. 12. A Case of Posttyphoidal Infection of a Rib. — By Alfred Irving Ludlow, M. D. 13. Wounds from Blank Cartridges. Observations upon 16 Cases cared for at Lakeside Hospital, together with a Bacteriologic Report. — By Dudley P. Allen, M. D., and A. I. Ludlow, M. D. 14. The Extraperitoneal Relations of the Appendix Vermiformis to the Posterior Surface of the Secum, with the Report of a form hitherto undescribed.— By C. E. Briggs, A. M., M. D. 15. Lumbar Abscess. Report of Six Gases treated by Aspiration and Injection of lodoform-glycerin Emulsion. — By Alfred Irving Ludlow, M. D. 16. A Clinical and Pathological Report of a Case of Splenic Anemia. — By H. L. Sanford, M. D., and David H. Dolley, M. D. 17. Pus in Abdominal Operations. — By Hunter Robb, M. D. 18. The Mortality following Operations for Pus in the Pelvis. — By Hunter Robb, M. D. 19. The Vaginal Incision in Sepsis following Abortion. — By Hunter Robb, M. D. 20. Torsion of a Hydrosalpinx Resulting in Infarction. — By William H. Weir, M. D. 31. Rupture of the Posterior Wall of the Cervix Uteri without involve- ment of the External Os, occurring during Abortion at the fourth month ; escape of the Fetus through the Laceration.— By Wm. H. Weir, M. D. 22. Conservatism in Pelvic Surgery.— By Hunter Robb, M. D. 23. Streptococcus in Gynecological Surgery.— By Hunter Robb, M. D. 24. The Early Diagnosis of Cancer of the Fundus, with Report of Cases. — By Hunter Robb, M. D. PATHOLOGICAL AND EXPERIMENTAL PAPERS. 25. An Experimental and Clinical Research into Certain Problems Relat- ing to Surgical Operations.— By George W. Crile, A. M., M. D., Ph. D. 26. Observations on the Origin and Occurrence of Cells with Eosinophile Granulations in Normal and Pathological Tissues.— By W. T. Howard, Jr., M. D,, and R. G. Perkins, M. D. 27. Eosinophilia in Pelvic Lesions and in the Vermiform Appendix. — By William H. Weir, M. D. 28. Primary Sarcoma of the Esophagus and Stomach. — By William Travis Howard, Jr., M. D. 29. The Origin of Gas and Gas Cysts of the Central Nervous System. — By William Travis Howard, Jr., M. D. 30. Report of Nine Cases of Infection with Bacillus Pyocyaneus. — By Roger G. Perkins, M. D. 31. The Pathology of Labial and Nasal Herpes and of Herpes of the Body occurring in Acute Croupous Pneumonia, and their Relation to the so-called Herpes Zoster. — By William Travis Howard, Jr., M. D. 32. Actinomycosis of the Central Nervous System, with the Report of a Case due to an Unidentified Member of the Actinomyces Group. — By William Travis Howard, Jr., M. D. 33. Streptococcus Mucosus (Nov. Spec?) Pathogenic for Man and Animals. — By William Travis Howard, Jr., M. D., and Roger G. Perkins, M. D. 34. Report of Two Cases of Angio Sarcoma of the Brain. — By Roger G. Perkins, M. D. 35. A Report in Gynecological Pathology. — By William H. Weir, M. D. 36. The Anatomical Findings in the Hypoplastic Genitals from Two True Female Dwarfs. — By William H. Weir, M. D. 37. A Contribution to the Knowledge of Ossification occurring in the Eye, with Clinical Report of Cases occurring in the service at Lakeside Hospital.— By B. L. Millikin, A. M., M. D., and John C. Dabry, a. B., M. D. 38. The Frequency, Site and Course of Tuberculosis in Cleveland.— By John C. Darby, A. B., M. D., and Henry P. Parker, M. D. 39. A Bacteriologic Study of the Blank Cartridge.- By David H. Dolley A. M., M. D. 40. A Case of Tuberculous Salpingitis from which the Tubercle Bacillus was Grown. — By Howard S. Dittrick, M. D. CLINICAL PAPERS. Extracted from The American Journal of the Medical Sciences, January, 1903. PARACOLON INFECTIONS, WITH REPORT OF THREE OASES. By Herbert W. Allen, M.D., RESIDENT PHYSICIAN, LAKESIDE HOSPITAL, CLEVELAND. Within the past six years, and especially within the past few months, increasing attention has been drawn to a class of cases which clinically resemble typhoid fever, but which careful bacteriological examination has shown to be due to infection by organisms intermediate between the typhoid and colon bacilli. For these organisms and the infection which they produce the terms " paratyphoid " and " paracolon " have both been used. Achard and Bensaude,^ who reported the first cases, entitled them " paratyphoid infections." Widal and Nobecourt^ opposed this term, and called their organism the paracolon bacillus, claiming that it more closely resembled the colon than the typhoid bacillus. Libman'' suggests that for the present the term " paracolon " is the preferable one, and would reserve the name " paratyphoid" for those organisms which culturally are identical with the typhoid bacillus, but which are not agglutinated by typhoidal serum. The earlier reported cases of paracolon infection have been so thor- oughly and so recently reviewed that anything more than a brief mention of them seems unnecessary at this time. Achard and Ben- saude, in 1896, reported two cases. The first case resembled typhoid fever, and was complicated by double femoral phlebitis and by cystitis or pyelonephritis. From the urine was isolated a paracolon bacillus. The second case was that of an infant which developed suppuration of a sternoclavicular articulation following a febrile attack of some two weeks' duration. The joint was opened, and from the pus was grown a bacillus identical with that obtained from the first case. In 1897 Widal and Nobecourt isolated a paracolon bacillus from an abscess in the neck, near the oesophagus. Gwyn,* in 1898, reported a case which clinically was typical of typhoid fever complicated by intestinal hemorrhage. From the blood he isolated a paracolon bacillus. Gushing,^ in 1900, isolated a paracolon bacillus from a costochondral abscess. His patient had nine months previously suffered from an illness, with relapse, supposed to have been typhoid fever. 2 ALLEN: PAKACOLON INFECTIONS. In 1900 Sohottmiiller" reported one case, and the following year six additional cases. In all but the last case he isolated paracolon bam ^_ from the blood ; the serum of the seventh case agglutmated the Da from certain of the other patients. , In 1901 Kurth' reported five cases which he considered exampies paracolon infection. From the urine of one patient ^^^^ the teces o another he isolated bacilli which were agglutinated m high dilutions by the serum of four of the cases. The serum of the fifth case-tne one with bacilli in the urine— was not tested. „ • - *• In April of this year Brion and Kayser» reported a case of infection resembling typhoid, complicated by relapse and thrombosis of the left leg. A paracolon bacillus was isolated from the blood, urine, feces, vagina, and rose-spots. In May 1902, Strong' reported an autopsy done forty-two hours after death on a patient who was supposed to have died from typhoid fever. From the spleen was obtained a paracolon bacillus. No lesions were found in the intestinal tract. In June, 1902, Buxton and Coleman'" reported a case resembling typhoid fever from whose blood was isolated a paracolon bacillus. In the same month Berg and Libman" reported an interesting case of typhoid fever with a secondary infection with a paracolon bacillus. The clinical picture was that of cholecystitis, and during life the para- colon bacillus was isolated from the gall-bladder, blood, and urine. ISTo typhoid bacilli were cultivated, but the patient's serum agglu- tinated typhoid bacilli at dilutions of 1 : 250, besides giving a reaction with the paracolon bacillus. At autopsy healing ulcers were found in the ileum. Hume''' has recently reported a case typical of typhoid fever, with relapse, intestinal hemorrhage, and cystitis. A paracolon bacillus was isolated from the urine and feces. The last cases, seven in number, are reported in August of this year. Johnston'' presents four, in two of which paracolon bacilli were isolated from the blood. In the other two the diagnosis was made on the strength of the serum reactions. There were no complications. Hew- lett'* reports one with isolation of the bacillus from the blood. His patient had a relapse complicated with bronchopneumonia. Longcope'* reports two cases, with isolation of the bacilli from the blood in both instances. His second case had a relapse. The first case was very severe, and proved fatal on the twelfth day of the disease. At autopsy the intestinal tract was found practically normal. This is the first com- plete autopsy report on a case of pure paracolon infection. To this list I desire to add the reports of three cases recently observed at Lakeside Hospital. All occurred in the service of Dr. E. F. Gush- ing. The first case is to be reported later from the surgical side. Dr. ALLEN: PARACOLON INFECTIONS. 3 G. W. Crile has kindly allowed me to use the surgical records to com- plete the history. Case I. Paracolon infection; mild course; suppurative cholecystitis during convalescence; incision and drainage of gall-bladder ; paracolon bacilli in pus; right lobar pneumonia; recovery. — M. S., white, male, aged thirty years. Admitted May 13, 1902, complaining of headache, vomiting, loss of appetite, and general weakness. Family and past his- tory unimportant. The present illness began four days before admis- sion, though for two weeks past he had felt weak and drowsy, and had lost his appetite. The initial symptoms were headache, general soreness, slight cough without expectoration, and persistent vomiting. For the past four days he had vomited practically everything taken. Took to bed May 10th. On examination the patient was well nourished ; slightly anaemic. The tongue was dry, coated, and tremulous. Pulse slightly dicrotic. Heart and lungs negative. Liver dulness in the mammillary line extended from the fifth rib to a point two fingers' breadth below the costal margin, where the edge was distinctly felt. The abdomen was Chakt I. itlaj' J II IIP DAY OF uoNTH 1 13 1 14 1 15 1 10 i; 1 16 1 19 1 20 1 21 1 22 1 23 1 24 1 25 1 26 1 27 1 28 1 20 1 30 1 31 1 1 1 2 1 3 1 4 1 5 1 1 T 1 8 i DAY OF DiBEAEE j 6 | | 7 | 8 U | 10 ( 11 | 12 | 13 | 14 | 15 | 10 1 17 | 18 | 19 | 20 | 21 | 22 | 23 j 24 | 25 | 20 | 27 | 28 | 2oJ 30 ] 31 | UJ i ^06 : _ : s tn^" 1"* - * t I 1AQ° 1 ,. 8 \^l 103 S ft . 5 , - J j V- 109° « \^^ K ^ *. i ^*'2.2::2 f5ft^^-,^&. ' : mi" ' - i ^ 2 . ~\ * -.- ^"^2 - Z 2 1^ 4 . <, *^'' - inn" ^ L j\% ^«'** no" " - 'J ~ ^2^ ~ QQ°:::: " " " " "" "i — ^~XZ' PUuaE 100 100 98 U5 100 Oa 88 90 05 00 63 85 65 82 80 75 75 85 85 05 105 120 125 05 80 60 00 flFBPmATiOHB 25 25 28 25 25 25 24 26 25 24 20 20 24 25 25 20 20 25 25 25 25 26 25 23 22 20 22 STODLB 01233321112120 2 111101010101 full, tympanitic, not tender. The spleen was readily palpated. There were no visible rose-spots. Temperature, 100.5° F. ; pulse, 105 j respirations, 25. The urine was high colored, acid, specific gravity 1025; no sugar; trace of albumin; an occasional hyaline cast and a few leucocytes. No diazo reaction. May 14th. Widal reaction negative. No tubercle bacilli in the sputum. From admission until May 30th the course of the disease was mild and uneventful. The highest temperature was 103.8° F. ; it fell slowly and touched normal for the first time on May 29th. On May 18th the patient's serum was tested with a paracolon bacillus, but the result was negative. On May 25th the Widal reaction was again negative. No rose-spots were ever found. The leucocyte count continued normal. In spite of the negative Widal reaction the case was con- sidered to be a mild typhoid. 29day, June 4th, the temperature fell by crisis, and thereafter continued normal throughout the patient's stay in the hospital. The lung signs •cleared up rapidly. The patient convalesced uninterruptedly, and was discharged July ifith, a very small fistula still remaining at the site of the operation. Case II. Paracolon infection ; mild course ; cystitis on seventeenth day ; thrombosis of left femoral vein on twenty-first day ; paracolon bacilli from ■ blood and wine ; recovery. — M. E., male, Australian, aged twenty -six years. Admitted June 6, 1902, complaining of headache, fever, and loss of appetite. Family and past history are unimportant. Present illness began about two weeks ago, with headache, slight cough, and ^oryza. For a few days had mild diarrhoea ; since then has been con- ALLEN: PAKACOLON INFECTIONS. 5 stipated. Appetite gradually failed, and there has been rapid loss of strength. No nausea or vomiting ; no chilliness ; no epistaxis. Took to bed the day before admission. Physical examination showed a well-nourished man looking de- cidedly ill. Expression dull ; face flushed ; tongue moist, coated, and tremulous. Pulse of good volume ; slightly dicrotic. Lungs negative. Systolic murmur at the apex of the heart ; normal area of dulness. Abdomen soft, not tender. Spleen readily palpable. No rose-spots. Liver dulness extended 4 cm. below the costal margin in the mammillary line ; edge palpable. Some reddening, swelling, and tenderness about the left ankle-joint, due to a recent injury. Temperature, 105° F. ; pulse, 112 ; respirations, 28. The urine was clear, acid, specific gravity 1019 ; no sugar ; very faint trace of albumin ; no diazo reaction ; microscopically contained a few coarsely granular casts. After admissioa until June 16th the course of the disease was fairly typical of mild typhoid fever. The temperature was slightly irregular, ranging from lOO'' to 104° F. The pulse was slow, averaging 85 per minute. A few rose-spots were noted, and the spleen was always readily palpable. The leucocytes ranged from 6000 to 9000. The Chakt II. bowels were constipated, requiring enemata. For the first three days it was necessary to catheterize the patient ; subsequently he voided urine without difficulty. The Widal reaction, tested frequently, was never positive at dilutions greater than 1 to 10. A blood culture taken June 7th yielded an actively motile bacillus, which will be described below. On June 16th the patient first noticed slight pain in the left groin, but made no complaint until two days later. Examination at that time disclosed slight swelling of the whole left leg. The foot appeared congested, and there was marked prominence of the superficial veins of the calf and thigh. The surface temperature was not altered. On palpation there was tenderness along the course of the femoral vein and in the popliteal space. The temperature was 103.5° F., which was about 2° higher than it had been running. The leucocyte count was 9000. Two days later, on June 20th, a definite cord-like thickening could be made out in the upper part of the thigh over the situation of the femoral vein. This was decidedly tender on palpation. Nothing was to be felt in the popliteal space. 6 ALLEN: PARACOLON INFECTIONS. On June 13th pus cells were noted in the urine, and on June 21st a culture yielded a large number of colonies of a bacillus identical with that isolated from the blood ; also a few colonies of a streptococcus. The subsequent course of the disease was uneventful. The tempera- ture reached normal on June 29th, and there was no subsequent rise. The tenderness in the left leg persisted for a few days. The leg never regained its natural size, and the superficial veins remained prominent. Under treatment with urotropin the urine cleared rapidly, and at the time of discharge contained very few pus cells, and no paracolon bacilli could be found on culture. The patient was discharged on July 12th. Case III. Paracolon injection; mild course without complications; recovery. — -D. L., male, white, aged thirty-five years. Admitted June 25, 1902, complaining of fever, logs of appetite, and diarrhoea. Family and past history are unimportant. For the past four weeks patient has JmiG July DAr OF MONTH 25 2G|27 |28 2y 30 1 2 U 4 5 C 7 8 9 10 11 12 13 14 15 10 17 18 19 20 21 J2 23 24 25 DAYOFDiSEAse 10 11 1 12 1 13 14 15 10 17 J3 I'J 20 21 22 2a 24 25 2G 27 28 29 30 31 32 ;J3 34 35 Jb 37|38 39 40 (- < t- 1 103 -^ "^'::::_ :: :: : : :: ":: : los ' „ J ' " / ? ft n" A ' A 5 - 1 m° '' 1 i n n' « 101 / t s 1 1 u 3 ' . . r . . ,„„« " i 1 / V - r - - ^-)| \ J } . . " , PULSE 35 85 85 85 75 SO 75 75 80 SO 75 80 75 75 75 70 80 80 73 75 70 72 35 70 80 70 80 70 70 80 70 flESPiHATiONS 25 25 25 24 20 20 20 18 20 20 20 20 20 20 24 20 20 20 20 18 20 20 20 20 20 20 20 20 20 20 18 STOOLB 13323424460154 2 3242344445 2 212] not done any work on account of weakness and loss of appetite. On June 16th had a shaking chill, followed by fever and headache. No subsequent chills ; no epistaxis. The bowels have been loose for a week previous to admission. The patient has remaind in bed since June 16th. Physical examination disclosed a fairly well-nourished man ; slightly anaemic. Tongue moist, coated, and fissured. Pulse regular, and not dicrotic. Heart and lungs negative. Abdomen held rather rigidly ; no tenderness. Border of spleen palpable ; a few scattered rose-spots ; normal hepatic dulness. Temperature, 100.5° F. ; pulse, 90 ; respira- tions, 25. Urine clear, acid, specific gravity 1015 ; no sugar or albumin ; microscopically a few leucocytes and an occasional hyaline cast. Diazo reaction positive. Leucocytes, 4400. The Widal reac- tion, tested on June 29th, July 2d, 11th, and 26th, was persistently negative at dilutions of 1 : 50, though a positive reaction was some- times obtained at 1 : 10. On July 2d and 11th the patient's serum gave a positive reaction with the organisms obtained from the first two cases at a dilution of 1 : 50. The course of the disease was that of a mild typhoid fever without complications. The temperature never rose above 103.5° F. It ALLEN: PAEACOLON INFECTIONS. 7 reached normal July 19tli. The bowels were loose throughout the early stage of the illness. Convalescence was rapid and uneventful, and the patient was discharged July 26th. A blood culture on June 29th, the fourteenth day of disease, remained sterile. A urine culture taken during convalescence was negative for paracolon bacilli. Bacteeiology. The organisms obtained from the first two cases were practically identical, and one description will answer for both. They were very actively motile organisms, having the morphology and staining properties of B. typhosus. The growth on agar and in gelatin was similar to that of B. typhosus. Bouillon was diffusely clouded ; no surface pellicle was formed after ten days' incubation. A slight trace of indol was formed in Dunham's peptone medium after one week. On potato there was a barely perceptible growth appearing as an in- crease of the moisture along the line of stroke. Litmus milk was slightly acidified after twelve hours ; after one week there was distinct reduction of the acidity, and by ten days the color equalled that of the control. After two weeks there was distinct alkali production. The organism from Case I. ("Samuels") produced slightly more alkali that that from Case II. (" Euster "). There was no coagulation of the milk after four weeks' incubation. Both organisms fermented glucose and maltose, with the production of acid and visible gas, but did not ferment lactose or saccharose. No animal experiments were performed. The organisms agree, therefore, very closely with most of the para- colon bacilli heretofore described. Agglutination Tests. In all the tests a time limit of one hour was used, and a reaction was not called positive unless there was good clumping and cessation of all motility. In none of the cases was a positive Widal reaction obtained in dilutions greater than 1 : 10, though tests were made frequently throughout the course of the disease. The serum from Samuels, Case I., agglutinated B. " Samuels" and B. " Euster" in dilutions of 1 : 100 ; likewise the serum from Euster, Case II., agglutinated B. "Samuels" and B. "Euster" in dilutions of 1 : 200. The serum from Lawton, Case III., agglutinated B. "Samuels" and B. "Euster" in dilutions of 1:50. In addition, the sera of Cases I. and II. were tested with Gwyn's paracolon bacillus and one of Johnston's organisms, and in both instances gave positive reactions in dilutions of 1 : 50. B. " Samuels " and B. " Euster " were tested with several difierent typhoid sera of high agglutinative strength, but without any reaction in dilutions of 1 : 10. 8 ALLEN: PAKACOLON INFECTIONS. The following table may assist in making these reactions clear : Senim. Samuels, Case I. , Euster. Case II. . Lawton. Case III. i Typhoid, agglutinative strength f 1-200 \ Organism. Samuels Euster Gwyn Johnston B. typhosus Samuels Enster (jwyn Johnston B. typhosus Samuels Euster B. typhus Samuels Euster 1-10 1-50 + + + + + + + + + — + + + + + + + + + — + + + + + _— - - 1-100 1-200 + + + 0- = no test made. ? = doubtful reaction. Johnston in his recent article has analyzed all the reported instances of paracolon infection. Most of the cases have run a course fairly- typical of comparatively mild typhoid fever, and without the aid of the serum reaction or a bacteriological examination a differential diag- nosis would be impossible. Rose-spots and a palpable spleen Lave been present in a majority of instances. Relapse has been comparatively common — at least five cases. Both of Longcope's cases are unusual in that labial herpes was present. The onset of one of his cases was with a chill. Our third case gave a similar history. The very wide distri- bution of the bacilli in the body is indicated by Brion and Kayser's case. Many of the common and several of the uncommon complications, of typhoid fever have occurred. Of the former there have been re- ported intestinal hemorrhage, cystitis, femoral phlebitis, furunculosis, and bronchopneumonia ; of the latter suppurative arthritis, myositis, osteomyelitis, and suppurative cholecystitis. Our first case is especially interesting as being the first recorded instance of suppurative cholecystitis complicating a paracolon infec- tion. The clinical picture of Libman's case resembled cholecystitis, but at operation the gall-bladder was distended with dark, thick bile, and at autopsy the bladder wall showed no change. Cholecystitis, both, suppurative and non-suppurative, is of course a well-recognized com- plication of typhoid fever, and its occurrence in a paracolon infection serves to again emphasize the marked similarity of the clinical picture of the two diseases. The ease of performance and the success of early operation in this complication is also well shown in this case. Longcope in his paper discusses the pathology of the disease. There Lave been three fatal cases, with autopsies in each instance. Berg ALLEN: PARACOLON INFECTIONS. 9 and Libman's case was a mixed infection, probably primarily typhoid fever, with a secondary infection with the paracolon bacillus. Healing ulcers were found in the ileum. In Strong's case no clinical history is available. The autopsy was done forty-two hours after death, and Strong himself suggests that the bacillus isolated from the spleen may possibly have been a post-mortem invader. The large and small intes- tines were normal throughout. The mesenteric lymph glands were ■enlarged, and a few were hemorrhagic. Fresh smears from the spleen showed a few crescentic aestivo-autumnal malarial parasites and some pigment ; but it seems unlikely that the patient's death was due to malaria. He had been treated with quinine. Longcope's case is complete. Autopsy revealed only the lesions of an acute infection —acute splenic tumor, cloudy swelling of the liver and kidneys, focal necroses in the liver. The mesenteric lymph glands were not swollen, and both large and small intestine were practically normal. While no very definite conclusion can be drawn as yet regarding the pathology of the disease, it is evident that we are dealing with an infection quite distinct from typhoid fever. That all of the cases of apparently pure paracolon infection may not, however, be as simple as Longcope's case might lead us to suppose is suggested by the occurrence in two instances of well-marked intestinal hemorrhages. References. 1. Achard and Bensaade. Bull, et M4m. de la Soo. de M^d. des Hap., November 27, 1896, p. 820. 2. Widal and Nobecourt. Semaine MSdicale, August, 1897, p. 285. 3. Libman. Journal of Medical Research, June, 1902, vol. vili. No. 1. 4. Gwyn. Johns Hopkins Hospital Bulletin, 1898, vol. ix. p. 54. 5. Gushing. Ibid., 1900, vol. xi. p. 156. 6. SchottmUller. Deut. Med. Wochenschr., 1900, vol. xxvi. p. 511; Zeitschr. f. Hygiene, 1901, vol. xxxvi. p. 368. 7. Kurth. Deut. Med. Wochenschr., 1901, vol. xxvii. p. 501. 8. Brion and Kayser. Munch. Med. Wochenschr., April 15, 1902, p. 611. 9. Strong. Johns Hopkins Hospital Bulletin, May, 1902, p. 107. 10. Buxton and Coleman. The Amekican Journal of the Medical Sciences, June, 1902, p. 976. 11. Berg and Libman. Journal of the American Medical Association, June 7, 1902, p. 1493. 12. Hume. Report of Thompson Yates Laboratories, 1902, vol. iv. p. 385. 13. Johnston. The American Journal op the Medical Sciences, August, 1902, vol. cxxiv. p. 187. 14. Hewlett. Ibid., p. 200. 15. Longcope. Ibid., p. 209. (Reprint from The Cleveland Meaical Journal, May, 1904) The Character of the Widal Reaction in the Present Epidemic of Typhoid Fever BY L. W. LADD, M. D., CLEVELAND Lecturer in Clinical Microscopy, Western Reserve University Before discussing the Widal reaction of the present epidemic in Cleveland it will be well, perhaps, to give a brief account of the nature of the Widal reaction and of its importance as a diag- nostic test in typhoid fever. In 1894 R. Pfeiffer first discovered that the blood-serum of animals immunized to typhoid fever, when placed in the abdominal cavity of a previously healthy experimental animal, together with an amount of typhoid culture previously determined fatal to an animal of similar size and kind, net only protected that animal against the action of the bacilli but also induced granular degen- eration of the bacilli and finally their complete solution. In 1896, Pfeiffer and Kolle, and independently Gruber and Durham, found that the blood-serum of human beings recently recovered from typhoid fever when mixed with an actively motile culture of typhoid bacilli soon rendered these immobile and caused their clumping together, that is, their agglutination. This was the first time that the reaction was used as a diagnostic measure. Widal later, in the same year, elaborated the method and showed that the serum of typhoid patients not only exhibited the same agglutinating influence upon the bacillus typhosus after recovery from typhoid fever, but also at the beginning and at the height of the disease. Numerous workers then took up the study of this phenomenon with the result that the following facts were soon established : 1. That the reaction may occur in 1 to 10 and stronger dilution in health and during the onset of numerous other acute infectious diseases besides typhoid fever. 2. That it is exttemely rare when the reaction occurs in dilution of 1 to 20 in other conditions than typhoid fever and that almost never does it occur in dilutions of 1 to 40 or 1 to 50, and above, in other conditions than an existing or past typhoid fever. 3. That the reaction was obtained by Widal and others in dilutions of 1 to 1000 and 1 to 5000, and in one case reported by Widal in which the reaction was positive in dilution of 1 to 20,000. 4. That occasionally, cases clinically typhoid fever, fail to give the reaction during their illness, even though in some of these Read before the Academy of Medicine of Cleveland, March IS, WOl, 2 L.-VDD — Character of the Widal ReactioI'J cases the typhoid bacilli can be obtained from the blood during life and in other cases at postmortem. . 5. That not' for a positive reaction only is agglutination necessary, but also complete cessation of motility m at least 1 to 4U dilution within a time limit of one hour. 6. That competent authorities have obtained the reaction in series of several thousand cases clinically typhoid fever at some time during the course of the disease in from slightly below Mfo to 95% of all cases. . , , 7. That in a series of 2500 cases, not typhoid, the reaction was positive in 2% of all cases. 8. That the reaction in mild cases may disappear even before convalescence is established. 9. That the reaction usually persists after convalescence is established from several weeks to several months with instances in which it has persisted 20 and 30 years after the attack. 10. That in order to determine the exact date upon which the reaction becomes positive, frequent examinations during the course of the disease may be necessary. 11. Absence of agglutination during the first or second week of the disease is no evidence against the existence of typhoid fever inasmuch as the reaction may not be positive until the end of the second month. As regards the day of illness upon which the reaction is positive. Park, in a large series of cases, has furnished the follow- ing data : Positive in first week 20% of cases. Positive in second week 60% of cases. Positive in third week 80% of cases. Positive in fourth week 90% of cases. Positive in two months 75% of cases. At Lakeside Hospital from January 1, 1903, to March 1, 1904, 454 cases of undoubted typhoid fever were admitted, exclusive of some 60 cases admitted to the Private-Ward Service, and exclusive of quite a number of cases probably typhoid fever, though the physical signs and symptoms were not sufficiently con- clusive to warrant my including them in this report in the absence of a positive Widal reaction. The date of onset of the typhoid fever has in all cases been made from the earliest time when some physical sign or symptom was sufficiently prominent to justify the probable onset of the disease. I have compiled a statistic report of the total cases admitted during 14 months, and also for each individual month, in order that to show that there has been a definite change in the character of the Widal reaction during the past six months as compared with the preceding eight months. Ladd — Character of the Widal Reaction 3 In performing the reaction the following regulations were observed : 1. Complete loss of motility and clumping inside of one hour, with a dilution of 1 to 30 and 1 to 50, were necessary before the serum was said to be positive. 2. Those giving a positive result in dilution of 1 to 20 and a partial reaction 1 to 50 were considered suggestive. 3. Those never giving a reaction 1 to 20 or 1 to 50 were considered negative. 4. Cases adfnitted after the fourteenth day with a positive reaction on admitance were disregarded in compiling this report, as there were no means of knowing, when the blood first became positive. 5. Outy those cases which developed positive Widal reaction in the hospital or were admitted with positive Widals before the end of the week were considered. 6. Upon those admitted with negative Widal reaction the test was performed every third or fifth day during the course of the disease until the reaction was positive, provided it became so at any time. The following points were determined for each month of the fourteen : 1. Number of cases. 2. Earliest date upon which the Widal reaction was positive. 3. Latest date upon which the Widal reaction was positive. 4. The average day of illness upon which a positive reaction was obtained. 5. The number of cases ■which gave only suggestive reac- tions. 6. The number of cases which gave no reaction at any time. 7. The number of cases which gave a positive reaction 1 to 20 and 1 to 50. 8. The percentage of cases not positive 1 to 20 and 1 to 50. 9. The number of cases admitted before the end of the second week, 10. The number of cases admitted before the end of the second week which gave positive Widal reactions. 11. The percentage of cases giving a positive Widal reaction before the end of the second week. 12. The number of cases admitted in the first week. 13. The number of cases giving positive Widal reactions in the first week. 14. The percentage of positive Widal reactions in the first week. 15. The percentage of cases during the last six months which did not give a positive reaction. 16. The percentage of cases during the preceding eight months which did not give a positive reaction. 4 Ladd— Character of the Widal Reaction 17. The percentage of cases during the past six months which gave a suggestive reaction. 18. The percentage of cases during the preceding eight months which gave a suggestive reaction. 19. The percentage of cases during the last six months which never gave any reaction. 80. The percentage of cases during the preceding eight months which never gave any reaction. SUMMARY — general, FOR THE 14 MONTHS Of these 454 cases 13% did not give positive Widal reactions in dilution of 1 to 50. 10.8% gave suggestive reactions, that is, positive 1 to 20, but not 1 to 50; 3.4% gave absolutely negative results; 86.7% gave positive reactions 1 to 30 and 1 to 50. The earliest day of illness on which the reaction was obtained was the third, and the latest day of illness on which it was obtained was the fifty-fourth. Of the 454 cases, 405 were admitted before the end of the second week; 68% gave positive reactions 1 to 30 and 1 to 50 before end of second week. Of the 454 cases, 307 were admitted before the end of the iirst week; 41% of these gave positive reac- tions before the end of the first week; 7.5% of the cases admitted during the past six months gave positive reactions before the end of the first week; 46% of the cases admitted in the eight months preceding gave positive reactions in the first week; 58% of the cases admitted in the past six months gave positive reactions before the end of the second week; 71% of the cases admitted during the eight months preceding gave positive reactions before the end of the second week; 15.6% of the cases admitted during the past six months did not give positive reactions 1 to 50 ; 12.5% of cases admitted in the eight months preceding did not give positive reactions 1 to 50 ; 11.4% of the cases during the past six months gave suggestive reactions; 10.6% of the cases admitted during the preceding eight months gave suggestive reactions; 4.1% of the cases admitted during the past six months never gave any reaction; 1.9% of the cases admitted during the eight months preceding gave no reaction. Tliese figures show that the Widal reaction was distinctly delayed in its appearance during the past six months ; that it was less often positive in this period of time than was the case pre- viously and that there were more anomalous cases than was the case previously. I wish to thank Dr T. W. Clarke and Dr John Phillips, of Lakeside Hospital, for the aid which they so kindly gave me in making this report. Extracted from the American Journal of the Medical Sciences, February, 1905 COPIOUS WATER-DRINKING AND POLYURIA IN TYPHOID FEVER. A CONTRIBUTION TO TREATMENT. By Edward F. Gushing, M.D., OF CLEVELAND, OHIO, VISITING PHYSICIAN TO THE LAKESIDE HOSPITAL, AND T. W. Glarke, M.D., RESIDENT PHYSICIAN TO TEE LAKESIDE HOSPITAL. During the summer months of 1903, when, typhoid fever being unusually prevalent in Cleveland, the wards of the Lakeside Hos- pital contained many cases of this disease, an attempt was made, at first in the female ward alone, to give much larger quantities, than usual, of water to drink to the fever patients, to determine its effects on their comfort and condition. The free use of water internally is, of course, accepted as important in the treatment of typhoid fever; and it has been the custom in the hospital to see that a patient with this disease should have what was thought an abundant supply. In the nursing directions, three pints daily has been set as the least amount which such a patient should receive; and besides the water given by the nurse, a quart bowlful of ice- water with a bent-glass tube on a stand by the bed-side has enabled the individual to help himself with a minimum of exertion. The twenty-four-hour amount of urine, always measured and charted, has shown in most cases a daily record of from forty to fifty ounces during the time of the fever, and this has been taken as a fair index of suflficient fluid ingestion. In trying to administer additional water it was soon found, with the efficient help of an admirable head-nurse, that without discomfort or special reluctance on the part of most patients, the unexpected and unusual amounts of from a gallon to a gallon and a half, or even more, could easily be taken. This was accom- plished by giving four ounces of water every fifteen minutes during the waking hours, amounting to from eight to fourteen pints, according to circumstances, in the twenty-four hours. In addition the ordinary patient received every two hours during the day, and once or twice at night, alternately six ounces of milk and six ounces GUSHING, CLARKE: WATER-DRINKING IN TYPHOID FEVER. of albumen -water, representing some three pints more of fluid. These large quantities, so given, were well borne. An occasional patient rebelled at first at the frequent dosage, but most took the water readily and some greedily, and the reluctant were, as a rule, soon persuaded that their comfort was enhanced thereby. The resulting diuresis was marked. The amount of urine passed in the twenty-four hours after admission in the average case was found to be about twenty ounces. After forty-eight hours, or by the end of the third day, with much uniformity in the abrupt response to the copious water-drinking, as the charts show, there resulted a daily elimination of from eight to twelve pints, and even in some cases two gallons or more of urine; and the polyuria was readily kept near this level in an uncomplicated case during the febrile part of Chart I. Fell. Murcli OAVOFMOHTH 28 1'9 1 2 a 4[5 C 7 8 9 10 11 ISJia 14 ID 10 17 IS lii 20 21 22 DAroFDisi=AS£ 5 7 8 Hj|11 L2 13 14 IQ IG 17 isjio 20 £1 22 23 24 2u 20 27 28 a S 0. E 0. JSa " f^tf-\t\/yztt^.i/ji-tTini _, m,,\t ^•^t- ir'^-i,Niti\,i Zz ^ m^^-, ^»r = zr:: = zT-.;S^^ .^^r. :ii::::e::i=5 97 ^ n 1 9"! / . ~! \ ^'^ 1 \ ■ ^ ' ' 19(i -1 C M^;/U -1^ l^^^\MRI/i^-^\- 4-i-f Z vi\iwj ^V'^r ^tl liul lA og /> ' . \; v" .^/U .^y ^/vy ,/ ^ - ,^ A/ ^ w i„(„r DEFECATIONS 1M'L;1'2G|2 85021341210 Y3 3G1 lOO URINE S||Sf|S2 5|S||gg||S|S|g2 2S 270 260 250 220 210 200 ISO ISO 170 160 160 liO 130 120 110 100 90 80 70 v., Med. No. 4231. The broken line shows the urine in ounces. the illness. A daily urine flow of from 120 to 160 ounces was estabUshed and maintained with ease in the average case (Chart I.) ; 220 ounces and more were not unusual (Charts TI. and III.), and in an occasional instance, like that of Annie M., Medical No. 3720 (Chart IV.), ten ounces were passed on the day of admission, to be succeeded on the third day by the enormous flow of 431 ounces, while the average elimination was 270 ounces. The daily number of typhoid fever cases in the hospital at this time was from forty to flfty, and the nursing-staff was taxed to the utmost with the routine care of the patients, and especially with the great number of tub-baths required. In the ward, however, where this experiment in hydrotherapy was undertaken, the head- nurse was soon decided in her judgment that fewer baths were needed and that the total nursing care of the typhoid patients was GUSHING, CLARKE: WATER-DRINKING IN TYPHOID FEVER. 3 Chart narch AprM [I. DAY OF unot-ajoior-tMco-^jaot-waiSg;"'"''*'**'" MONTH -r^rHMrHClWWWNWWMNINOiK. <»«SS?J2333S2SasSSSSgKSSS oI^LI'e =-" = 33233 as ssssaassassss ssgjissssssssg^^sssisi^sgss o°pem"IS sg3ggi|S3|gsgagasgs|ss 1 Ui Q. E 5 rf J iis oort im° 1 i' A A ^ i ^Ki^rHff-j-jjjrlj-iJjj -fM -f-^-i- t J t I^K So H 5 r s g^ij;:-. sJ illftii i ' ' '^ I L 1 I 1 Ql L 1 L it l I Z ai_ 210 99" ^ ' I/' 1 ' r tf, f 1 1 if L It ' L J ^Im^^^--~^-z,-;z^z---zz-W^ «|"-z"S^rpzp_ir^,rzz-zi __1L_|^^^^==„_^^^^_^L.^^ «*__ jgg 77" 1 A P J ^ ^ ffi =, aMXJj^LI^ S m il~juiiiivii\i.twi\i Htsit ^Ji,4j/^ l'^^'^^ ^^ *■ 8» ^'Hli-.i^^ST' ?5 20 V'V! .^ jg ^ -^T^ ^ ^ at ^ c: z = ^ w^Lgfe^ ■£ .^ ,^ i!5S(4w; if iJc: im! Si S^ ^ .o:^ .^ ^ K^ wdrsbii^ i<; V gg DEFECATIONS 2 2 01'20 l" o!l'20 1" oWo 1* 1' 1' 1' 1* 1" 01' 1' 1' 3 01'42 13133344S3 "S URINE S £ S|sp S SJ|g|3 % 3|p|S SSIISSSSSSSS i:::::::::::::::::::::::!! D., Med. No. 4063. The troken line shows the urine in ounces. The record was omitted when the temperature became normal. Chaet III. Jnne Jnly July Aai;. ^ ^ ^ S S 5 g 3 i iF « " ^ - T ■" =• S 3 3 3 3 DAY OF TPujot-cociO'HCiM-Xu'soi-roisSrHOiro^ooi- DISEASE ^ rHr.^^WrHrH-.^rH«o.«S4^J. «« 5|gs3S3SSSsifig$^^5:55^^^g3S DAY AFTER |5!S|3:gS|^SSggSgSggSSS|S OPERATION |c^(^lClc^r-■c^«le^c^Mo^<^^e^<^^lN■^^e^<^^MlNC^^^ IBsSIISisT 1 S s 0. CL S a t:^ j" |5i^:-:::"; ?SJ \ \\ ^ = 3" = -'-i o?S ^5 , .^ j i^-:^ TH ^ 3 ISt" JKjj-,jJ-/it ^1j Xi -1 f^A* t^ itif 1 -,t24 ^ is^i-^J- \\ t- • T^ g ■ 3 ¥ =^ 240 ;2?.t tv ^rti-,^4u^itii - 5 n ■ - 4 SS JSJo i- '^ '^{ttlir'^rtt lit A-I4-I-K-, ^ 210 mi n ii/-fi ^t / t 'zUu^-iUUi^i-^f?^?^^^-^ ::-r— TT---^^^^-7=^5l» 97° icn 1in ^Sn \ 1 y ^, y /K f y , 1 J 1 J ' / /I 3 4^SS SJ iZi^z^'^s: ^"^vsiMyylniifyt. / 1 f / t f ^ / M / nn ^ itlV^^'^ <.^WJ7 tpt i'Jl ?S 20 ^ y..^./v, _ r- /t. '%'-', V-p -T". /■-'/-v^" Tft DEFECATIONS 3Q4442123180 1'0 1'0 1'0 1'0 1' Ol'ijo 130 !■ 121111111 01' 1121122121 uS sgHSSSIiiiiliiSliiiilii issslisis T ii K.. Med. No. 4266. Abrupt rise in the urine from 10 ounces on the first day to 243 ounces on the second day; maximum, 328 ounces on the ninth day. 4 CUSHING, CLARKE: WATEE-DEINKING IN TYPHOIX) FEVKE. less, in spite of the extra attention which the frequent administration of water and the increased use of urinals involved. In the bed- side observation of this group of cases, the general comfort ot the patients so treated seemed apparent. Headaches were not so troublesome, so that the familiar ice-bag was much less in evidence in the ward. Tongues and mouths kept noticeably clean and moist, and the toilet of the mouth cavity was a much simpler nursing task. Apathy, deafness, restlessness, nocturnal delirium and other nervous and toxemic symptoms seemed less in evidence, hypnotics were not so often needed, nausea was unusual, and remissions in temper- ature appeared more frequent. Comphcations, minor and major, were few among the patients, and there were no deaths in this first small series, amounting to twenty cases, though the prevaihng epi- demic was of severe type, and the general mortality in the community and in the other wards of the hospital at the same tinae was large. An additional year's experience, during which this element of treatment has been continued with more or less thoroughness in all the medical wards of the hospital and in the services of Dr. H. H. Powell, Dr. J. H. Lowman, and Dr. H. S. Upson (I am greatly indebted to these gentlemen for permission to utihze their cases), as well as of the writer, has seemed to confirm the early impression of its usefulness, and demonstrated that in all cases of typhoid fever, unless on admission profoundly toxic or with serious complications, or in the ease of children who will not, as a rule, take water so often, this very abundant flow of urine can be secured with certainty in forty-eight hours and kept up during the illness. The method of administration first adopted of having the patient receive, when awake, four ounces of water every quarter of an hour, has been continued. I^arger quantities at a time are less acceptable, and often not well tolerated, while small amounts, frequently given, are usually not unwelcome. The urine in these cases when the daily amount exceeds a gallon, is extremely pale, with a specific gravity of from 1001 to 1005. With an elimination of more than 200 ounces, the specific gravity of the watery fluid is often below 1001. There was found to be a close approximation in the absence of diarrhoea or sweating, between the amount of fluid ingested and the quantity of urine eliminated. Thus in the case of K. (Chart III.), during ten days, 2664 ounces of fluid were given, while the quantity of urine passed was 2626 ounces, an average daily difference of some four ounces. On certain davs the amount of urine passed would slightly exceed the liquid taken. When obvious perspiration was present, or several loose movements occurred, the disproportion was, of course, more marked. In the case of B., Medical No. 4230, where the skin kept moist during most of the illness, 1358 ounces of fluid were given in six days and 844 ounces passed, a daily average of 226 ounces and 140 ounces, respectively. Chart V. shows the effect of repeated hemor- CUSHING, CLARKE: WATER-DRINKING IN TYPHOID FEVER. 5 Chart IV. Si>pt<>iil1ier October ■November DISEASE 3---r^r-?it;i^I?l?iaS=iS?iM«SJe?5'^S!o? ^^55^5-5?^;??S5s:3:ssssteg3s °o^,.VZ 583SgislSiSgslii«SSS3Ss i a. 109^ L \ _n Ji i\ CI m°' 1 h/TJ *]h3jE -Jd -,-.M-l^ io»° f ,1, J i ^ \ir Infii it IV 1 m- Nl11ht V '^"^'^rrW^fll/n/nm nTTT-nTT^nTsmsfTTTT'^v J»'o' 7'^t'^'^i- 1 v-i (-{ fv- ■ llL^Mi Zla 9o;--...J.t M^-^ ip :..^ Ps^^z,.^g..-J./^^"^ 98° ^^ S^--*^ ' ::~„^-:-r? ^^ Z^ ^^}22 96" J i.to - - 2"t int 100 'n7 A/ /.rlti' I'll J''HT , . f m K i, " aS '^^y fVl^y*' V'W;/t,s /rf ,/ /<'»t'^|^|^^a/iWAffl/li.l [AA 1 1 I.IAjI IVIVNJ'r™ ' ■ V H » A /\ , / V •-r\ fl^ '\ • 1"' r ' v •n/ \ / ,__v^ ^ ^ _. - ^ w H-^ J- '^ /" «/V - X 20 , |— — 1 ^" -^ T. . . " OSFEACTIONS 2 12 1 OJl'liOjl'; 208705312220 l":ll 44 43230210 1' 22423444354334 ;;;g '^"-^ S3l^|siillSi^^lSi§K^5§^li ssssssil ii A.M., Hosp. Ko. 3720. Extreme case of polyuria, with increase of urine from 10 ounces on the first day to 431 ounces on the second day. Chart V. Juno July AU IIHt Zl°k s s 8 g ?i s s a s El s Ji s 5 s s s s g B s s » ? ^:!:'5$^S?SSSi33SgKSS§3 3 3 3!3 s I a. E t- D. o ^- i " f M5° 5 u JSt" " A / S J s i- S- J' '^SSal JJJ -, rJ 4 ^-t ' i 2?°;i \ °2'fll^ iT-iPl^tt^t fSJ"-"; t ? "^/^i/i/yv ' ^ *- 1 : .^sA2t\i -diit-ii~. > 'I flO S- 4N0I -*r+-=- ..jE^^.^..l^^^^^^ffl^ 98°I^= = =^ = J= 1"""^ ----1^^ p^--l ^ ^ iL _^ ^ ^ 3 n J^ " _i jEJ ^"i ^^ "^Si] viSijyl ^/ Ta'u^ii\iTfi^i''-n}'y^~^/'^ t i-.^it. iitj-^^,„ J '' tiwy-4-it4/-,„u ::_- , ^ V ■^ c* CO to o CJ '^ eo CO o -■ .H CO to "* (N CO CO CO - o - o CM TJH Tf o «5 o 00 CM o S oo o* l> CO ^: ■sinBq JO jsqran^ *3jni'BJ3dai8i IseqSiH •J9A9J JO SAVQ 3 •B9onno 001 (■Bapas •Bjjniioj) -goei Aioxag 'SMUTio OOI ©Aoq-y 'ss^bo n*V GUSHING, CLARKE: WATER-DRINKING IN TYPHOID FEVER. 9 As the 100 cases of typhoid fever with polyuria reported represent, in some measure, a selected group, the small and diminishing mortality rate is, of course, at the most, suggestive. It is felt, however, that certain conclusions as to the results and usefulness of this mode of treatment, which seems to supply an additional means of combating the toxaemia of the disease, may be submitted with some confidence with the hope that this method of copious water-drinking with its resulting diuresis, may be found by other observers to diminish in some further degree the severity and mortality of typhoid fever in hospital practice. Our experience and conclusions may be summarized as follows: 1. Large quantities of water internally, a gallon or more in twenty-four hours, may easily be taken by typhoid fever patients, if administered in small quantities at frequent and definite intervals. 2. A copious elimination of watery urine at once follows, the degree of polyuria, day by day, closely corresponding to the quantity of fluid ingested, 3. Patients are more comfortable by this mode of treatment and toxic, nervous symptoms are lessened. 4. The mortality, as well as the severity, of typhoid fever, seems to be still further diminished by this method of hydrotherapy employed as an accessory to the cool-bath treatment of the disease. Reprinted from A e chives of Pediatrics, June, 1904, E. B. Treat & Co., Publishers, 241-243 W. 23d St., New York. STOMATITIS IN IMPETIGO CONTAGIOSA.* BY EDWARD F. GUSHING, M.D., Visiting Physician to the Lakeside Hospital, Cleveland. O. The occurrence of a form of stomatitis in impetigo contagiosa has been specially referred to in French pediatric and derma- tological literature. Comby" called attention to this association seventeen years ago, and has several times returned to the topic, describing "stomatite inipetigineuse" as a characteristic and not infrequent complication of acute facial impetigo. Bergeron^ in France, Bohn^ in Germany, and J. F. Payne,* of St. Thomas' Hospital, in England, had previously reported in- stances of ulceration of the lips and mouth in children with im- petigo contagiosa. Payne in his cases noted, also, the occasional inoculation of the disease on other orificial mucous membranes with a resulting purulent coryza, conjunctivitis, keratitis, or vulvo- vaginitis; and from the quite frequent combination of stomatitis and the whitlow-like lesions on the fingers used "hand and mouth disease" as a synonym of his own for the affection. He has again discussed the subject in his recent article" in Allbutt's System. To Bergeron and Comby, however, seem due the detailed clinical picture of this variety of stomatitis. The lesions of impetiginous stomatitis bear a certain re- semblance to those of herpetic or aphthous stomatitis. Their usual seat is on the inner surface of the lower lip, but they may occur on the inner surface of the cheeks, or of the upper lip, on the tip of the tongue, or on the gums, but always in the front of the mouth where the child's fingers might have reached. A pustular or crusted skin patch on the outer border of the lip may be directly connected with an ulceration on the mucous membrane within. The mouth lesions are single or few in number, not grouped like herpetic ulcerations, larger, as a rule, than the latter, irregular in outline, presenting themselves as slightly raised patches of a pearly-gray appearance when recent, and later of a grayish-white * Read before the Ohio State Pedia'tric Society, Cleveland, O., JVIay 16, 1904. Gushing: Stomatitis in Impetigo Contagiosa. or yellowish hue. The patches are quite adherent, and their re- moval exposes a raw and bleeding surface. The surrounding mucous membrane is more or less reddened, and there may be in- crease ot saliva, but no odor to the breath, and little or no local discomfort. The lymph nodes under the chin are slightly swollen and tender. The process remains superficial, and healing occurs, as a rule, within a fortnight. The diagnosis is simplified by the presence of a skin eruption of impetigo, though Comby has seen the lesions in the mouth precede those on the skin. He also, de- scribes involvement of other mucous membranes with purulent rhinitis, conjunctivitis, and vulvovaginitis, and explains all the lesions, as well as the common paronychia, as the result of auto- inoculation from scratching or rubbing and the introduction of the finger into mouth, nose, or elsewhere. Sevestre and Gaston," Dupuy,' Leroux,^ Poulain," Levy'" and others in France, have studied impetiginous stomatitis as well as the other secondary lesions of the nose, eyes and vulva, clinically and bacteriologically. All regard these complications, when looked for, as not infrequent. Bacteriological examination shows in all cases staphylococcus aureus, as in the skin pustules of impetigo. Leroux and Balzer and Griffon,^' however, have isolated a streptococcus from early lesions on both the skin and mucous membranes, which by inoculations has produced char- acteristic lesions on the skin, furnishing anew the same organism, and they regard the staphylococcus as a secondary invader. In recent German literature, also, there is a recognition of the occasional involvement of the mucous membranes in impetigo contagiosa. Lang,'^ of Vienna, in his text book, of 1902, describes these complications; and Mikulicz and KiimmeP^ in their ex- haustive work on diseases of the mouth, refer to the subject. Jadassohn'* in 1896 reported the successful inoculation of im- petigo lesions on the skin from the mouth patches of the disease. Excepting Payne's papers, however, I know of no discussion of the subject in English or American dermatological or pediatric text-books or journal literature. Forchheimer"^ in his admirable monograph on the diseases of the mouth in children, in 1892, alludes to the topic in the single sentence that "the fact that aphthae may be found in the mouths of children with impetigo is of no possible value in establishing any connection between them." H. G. Anthony,'*" of Chicago, in 1898, reporting 50 cases of im- petigo contagiosa notes of one what may serve as a very good Gushing: Stomatitis in Impetigo Contagiosa. description of impetiginous stomatitis, as follows: — "A child of three had the ordinary form on the chin, together with lesions on the mucous membranes of the mouth, which were not aphthous, but greatly resembled syphilitic mucous patches ; they had ap- peared shortly after the chin lesions. There were no syphilitic lesions on other parts of the child's body and no history of syph- ilis in the child or its parents. Scrapings from these lesions con- tained staphylococci, but no oidium albicans. They healed prompt- ly under treatment by peroxid of hydrogen." It is obvious that the possible involvement of the mouth, the mucous membrane of the nose, the conjunctiva, or the vulva in impetigo contagiosa present problems of differential diagnosis of interest to the pediatric physician. In so common a disease, where a glance usually suffices for diagnosis, and in the uncleanly type of children, commonly affected and seen in dispensary clinics, it is easy to appreciate that a running nose, inflamed eyes or sore mouth may be overlooked, or not suspected to represent an in- tegral part of the disease. These complications of impetigo contagiosa were brought to the writer's attention by the following cases : — Case I. — Impetiginous Stomatitis. Charles P., six years old, was admitted to the Children's Ward of the Lakeside Hospital, under suspicion of typhoid fever. For a few days he had been listless, complained of headache and seemed feverish at night ; the bowels had been constipated and he had vomited once. The boy was well-nourished and of good color, but was apathetic. Tem- perature 99.5° F., pulse 100. Examination was negative, except for lesions of impetigo on the face. In the middle of the forehead, on the upper lip at the orifice of the left nostril, and on the chin, were typical crusts of impetigo contagiosa, those on lip and chin brownish-red from blood admixture. There was slight muco- purulent discharge from the left nostril. The tongue was coated, the mouth and throat otherwise negative. The submaxillary and submental lymph nodes were slightly enlarged. There was no Widal reaction, and the leukocyte count was 11,000. The urine was normal. The child continued apathetic and drowsy for three or four days, the temperature ranging from 99" to 100° F. On the fifth day the temperature was normal, and the boy bright and hungry. The facial lesions were tending to heal. On the next day there was found on the inner surface of the lower lip near the right angle of the mouth a pearly-white, irregularly rounded Gushing: Stomatitis in Impetigo Contagiosa. lesion, 14 inch in diameter, slightly raised above the mucous membrane, closely adherent, without red areola, but presentmg a superficial resemblance to a lesion of herpetic stomatitis. The next day, a slightly larger patch of the same appearance was noted on the inner surface of the right cheek opposite the first molar. There was general redness of the mouth and some increase of saliva. The lymph nodes under the chin on the right side showed increased enlargement and tenderness. The thin, smooth, gray- ish-white patches, dull and opaque in look, did not resemble a diphtheritic membrane and were not pultaceous. The surface layer was closely adherent and left a raw, red surface on removal. There were no further lesions, and in the course of eight days, the patches gradually cleared. Meanwhile, the facial lesions had healed. Cultures from the under surface of one of the mouth lesions gave a pure cuhure of staphylococcus aureus. Case II. — Impetiginous Rhinitis. Mabel B., six years old, was brought to me because of a profuse purulent discharge from the left nostril of some days' duration. The character of the dis- charge suggested an inflammation of one of the sinuses, or the presence of a foreign body in the nostril. Examination, except for redness and swelling of the tissues, was negative. An older sister with whom the child slept, had been seen a few days before, with facial impetigo. No suspicion of the bearings of this fact on the purulent rhinitis was excited, until a typical pustule of im- petigo appeared on the upper lip of the patient. Bacteriological examination of the nasal discharge showed staphylococcus aureus in pure culture. With simple treatment, the purulent secretion from the nostril subsided as the facial lesion healed. BIBLIOGRAPHY. 1. Gomby. La France med. 24 Dec, 1887. De quelques stomatites de FEnfance. Rev. mens. d. Mai. de i'Enfance. Sept., 1888, p. 916. Soc. med. des hop. d. Paris, 1891, p. 330. (Discussion of Sevestre's communication.) Traite des mal. de I'Enfance. (Grancher-Gomby-Marfan) Stomatite impetigineuse. 1897. Vol. VII., p. 362. 2. Bergeron. Stomatites. Diet, encycloped. d. sci. med. (Dechambre.) 1880. 3. Bohn. Die Mundkrankheiten. Gerhardt's Handbuch d. Kinderkrank. Vol. IV., p. 2, 1880. Gushing: Stomatitis in Impetigo Contagiosa. 4. Payne, J. F. St. Thomas' Hosp. Rep., Vol. XIII., p. 301, 1883. 5. Payne. Impetigo Contagiosa. Allbutt's System. Vol. IX., p. 526. 6. Sevestre et Gaston. Sur une variete de stomatite diph- teroide a staphylocoques. Soc. med. d. hop. d. Paris, 26 June, 1871, p. 316. 7. Dupuy. Sur I'impetigo et certaines de ses localisations chez I'enfant. These de Paris, 1891. 8. Leroux. De I'impetigo des enfants, affection contagieuse, inocuable, et microbienne. J. d. clin. et. d. ther. enfant. 8 March, 1894, p. 210. 9. Poulain. These de Paris. 2 March, 1892. 10. Levy. These de Nancy, 1897. 11. Balzer et Griffon. Le streptocoque, agent pathogene constant de I'impetigo. C. R. Soc. d. Biologie, 29 Oct., 1897, p. 916. Stomatite diphteroide impetig. a streptocoque. Rev. mens. d. mal. de I'Enfance. Jan., 1898, p. 23. 12. Lang. Lehrbuch d. Hautkrankheiten, 1902, p. 224. 13. Mikulicz, u. Klimmel. Die Krankheiten des Mundes. Jena. 1898, p. 233. 14. Jadassohn. 73 Jahresber d. schles. Ges. f. vaterl. Kultur. Med. Sektion, 1896, p. 89. 15. Forchheimer. Diseases of the Mouth in Children, 1892, p. 38. 16. Anthony, H. G. J. Cutan. and Genito-urin. Dis., 1898, Vol. XVI., p. 227. [Reprinted from the Boston Medical and Surgical Journal, Vol. cxlvii, No. 3, p. B9-63, July 17, 1902.] A REPORT OF TWO CASES OF "TYPHOID SPINE."! BY OBOROB WIIjTON.MOOREHOUSE, M.D., CLEVELAND, OHIO, Visiting Phyelcian to the Ditpensary of the Lakeside Hospital and Western Reserve University. Since 1889 several examples of a rare sequel or coinplication of typhoid fever have been reported under different names. That of " typhoid spine " is the one first used and possesses the advantage of not binding one to a definite theory of the underlying pathological cause. I have succeeded in finding references in the literature to twenty- one cases having some resemblance or relation to the condition first reported under this name, and to these I take pleasure in adding the histories of two others which came under my care at' the Lakeside Hospital during my service as resident physician. Ca.se I. J. F. C, male, age twenty-eight years, agent structural steel work ; entered the Lakeside Hospital on the eighth day of his illness with ty- phoid fever. The attack, while not marked by extremely high temperature, was a severe one complicated by intestinal hemorrhage, relapse and furunculosis. The Widal and diazo reactions were obtained. Although there were no symp- toms referable to the back during his stay in bed, it was noticed later that it was exceptionally weak and ached after the patient was allowed to sit up or walk about. There was no actual pain. The back was examined several times with en- tirely negative results. • Read before the Academy of Medicine erf Cleveland, June 20. 1902. The patient left on the 6th of June, 1899, after a stay of seventy days in the hospital. On the 23d of August he re-entered the hospital, com- plaining of pain and weakness in the lower lum- bar region, and gave us the following account of his illness : For ten days after his discharge his back was weak and ached a little ; then without assignable cause he began to have pain which was slight at first but increased rapidly in severity. At the beginning it was throbbing in character but later became dull. This dull pain began to be punctuated by sudden short attacks of severe pain ; the sharp pains were associated 'rt'ith a great tendency to twitch, and any movement in- tensified and prolonged the attacks. The pains were somewhat more severe when the patient drew up his legs, and were a little relieved if the legs were in a horizontal position or nearly so. The duration of these attacks was usually an hour or two, with intervals of from three to four hours of comparative relief. Movement caused pain, and was very apt to bring on an attack of this kind, which, however, often began without assign- able cause. The attacks were more frequent at night than in the day time, and disturbed his sleep. The localization of the pain was in the lower lumbar and sacral region or buttock ; most commonly on the right side, sometimes on the left, never on both. Previous to his admission to the hospital the patient states that opiates were discontinued, because on trial they were found to have so deleterious an effect on the action of "the kidneys and bowels. During the severest part of the illness the patient had periods of a few days .each with normal temperature, alternating with febrile periods of similar duration in which the temperature sometimes rose to 103° F. At this time he became much emaciated and nervous, "going all to pieces " if he lost control of himself. His bowels were constipated. About four weeks before his read mission to the hospital he began to improve. The attacks of pain became less fre- quent and- severe, the appetite improved and he gained in weight. At entrance the appetite was good, the bowels were regular, and the pain was confined to the lower lumbar region, was dull in character, with an occasional sharp twinge. The physical exam- ination was largely negative. The patient was somewhat emaciated ; the pulse 80, regular and rhythmic ; the patellar reflexes were very weak, the plantar reflex was not obtained. There was no evidence of destructive disease of the spinal column ; no deformity ; no tenderness over the spine or elsewhere ; no indication of involvement of the sciatic or other great nerve trunk ; no evi- dence of psoas abscess. The urine is reported not to have contained albumin ; there was no diazo reaction; the sediment showed a few hya- line and fine granular casts and leucocytes. Examination of the blood revealed the presence of 5,280 leucocytes to the cubic millimetre. The Widal reaction was reported negative. During the first month of his second stay in the hospital the patient was not kept continuously in bed, and was at times almost free from discom- fort, and again suffered considerable pain in his back or in his buttock. At the end of his first month in the hospital the temperature, which had varied between 98° F. and 99° or 99.5° F., began to go up gradually, reaching its highest point, 103.6° F., on the 25th of September. The febrile period extended over two weeks. Shortly after this rise in temperature began, an examination of the blood showed 8,800 leucocytes to the cubic millimetre. On the 2d of October the spleen, which is said to have been normal in size at en- trance, was felt below the costal margin, and a Widal reaction was obtained. There was no rose- rash, the abdomen was neither distended nor tym- panitic. In spite of the enlargement of the spleen and the apparent recurrence of the Widal reac- tion, it is hard to see how this could be either a reinfection with typhoid or a relapse of that dis- ease. At first it was not thought necessary to confine the patient strictly to his bed. The Paquelin cautery was used on several occasions, the back was strapped, and tonic remedies were employed. When the more severe pain began, antirheumatic, antineuralgic, sedative and hyp- notic remedies were employed, with opiates when necessary. The case progressed without the aid of any mechanical contrivances aside from the very inadequate one of adhesive straps. When the patient was in bed with severe pain, with a view to determine whether relief would be af- forded by the use of extension by weight and pulley, manual extension was tried without any apparent benefit. Until perhaps the last of Octo- ber there was very little change in the condition of the patient as to pain, but from that time on there seemed to be a steady, but very slow im- provement in this respect. On the 2d of Novem- ber he began to have a bed-rest very cautiously, and this was gradually increased, as the procedure seemed to be tolerated, and the patient's strength bore it. He was discharged from the hospital on the 13th of December, 1899, the note stating that he had still some trouble with his back. The patient was seen in July, 1901, and said that, except for two to three days about three weeks after leaving the hospital, he had no pain or ache in the back after his discharge, but that he had so much weakness in the back that he greatly feared a return of the trouble on several occasions. He did not attempt to go to work in the oflUce until May, 1900, thirteen months from the onset of the typhoid, and, for some time after returning to the oflBce, did very little. During all the next summer had a good deal of soreness in the back for a day or two after any misstep or overexertion, and experienced considerable dif- ficulty in getting about on this account. At the present time he does about the same as he did be- fore his sickness, but still has a little soreness in the back after a misstep or overexertion. He is quite certain, however, that this last is decreasing and will entirely disappear. Case II. J. M. H., male, age forty-two years, a salesman in a large hardware store, entered the Lakeside Hospital July 17, 1899, in the service of Dr. J.^E. Cook, by whose kind permission I am permitted to report the case. Unfortunately the records of the case were not preserved, but the following history has recently been secured from the patient, a very intelligent man, and I think that the genei'al features of the attack have not been in any way misrepresented by the loss of the original records. To Dr. Cook I am indebted for certain facts about the case not remembered by the patient or myself. He was taken ill with typhoid on the 7th of April, 1899. About seven weeks later the fever left, the course being one of only moderate severity and entirely without com- plications in the preliminary attack. At about the usual time after the disappearance of the fever, thp patient was propped up in bed, and felt a little stitch or catch in his back such as one has at times on twisting it. This trouble increased continuously, and within twenty-four hours he was unable to turn in bed, and suffered from deep, cramp-like pain in the lower lumbar region, a little more to the right than to the left. This pain was not absolutely continuous, but came in paroxysms, which seemed to be relieved, or at least to be made more bearable, if the patient grasped the bed or some other object tightly while it persisted. After this first bed-rest and its very unpleasant sequel, he was kept flat upon his back fully ten days, and at the end of that time was so much relieved that he was again propped up in bed. The time of the bed-rest was very gradually increased, and he was allowed to put on his clothes at the end of another ten days. He was not,- however, free from pain or soreness when sitting up, but when lying down he had no dis- comfort. There was no elevation of temperature. At the end of about two weeks the pains again began to increase, and one day when on his feet he had so severe a " cramp " that he would have fallen if it had not been for support. He there- upon took to his bed and for three or four days had some elevation of temperature. On the 17th of July he entered the hospital. The patient was anemic, emaciated and nervous, being almost hysterical at times. There was no evidence on physical examination of a neuri- tis •, no deformity of the spine was detected ; no ten derness to deep pressure was found in either iliac region or over the painful area in the back. Tenderness over the spine developed later and was for a time very marked. There were no rectal or vesical symptoms. Fortunately I have been able to find the original record of his tem- perature for the first thirty-eight days of his stay in the hospital. He entered with a normal tem- perature, but during the first month he had febrile periods of from twelve'hours to four days in dura- tion, in which the temperature rose to 102.5° F., 103.5° F., and once to 104.7° F. They alternated with rather longer periods of normal temperature. After the first month in the hospital his tempera- ture is supposed to have remained normal. The patient was never free from discomfort at en- trance, and at relatively short intervals would have a succession of spasms of pain lasting at intervals for a day and a night, or even two days, after which he would be relatively free from them for a short time. Absolutely no cause could be as- signed for the onset of these attacks, although any movement except one executed with the greatest circumspection increased his discomfort, whether he was in pain or relatively free from it at the time. For about one week from the time of entrance heat was applied to his back. During the second week cold was applied by means of an ice bag. . Morphia was used throughout when it was con- sidered necessary. Dr. J. H. Lowman saw the case in consultation during one of the attacks of pain, and suggested the use of extension by means of a weight and pulley, on finding that manual extension afforded relief. Extension was applied, and was continued intermittently for about four weeks. In spite of the fact that it afforded an appreciable! amount of relief, it was not continued more than a few hours each day, since its appli- cation disturbed the patient, and it was thought desirable to allow him to turn on bis side with the hope of thus avoiding bed sores. In addition to tonic remedies, potassium iodode was admin- istered in moderate doses, and opiates were used as needed for the relief of pain. For a consider- able time after entrance the prognosis was con- sidered very grave. He was thought to have a tuberculous spondylitis. From the time of the first bed-rest till the patient was discharged from the hospital, "improved," advances were made very gradually. He had his clothes on for the first time about three weeks before he went home. The patient was seen in July, 1901, and stated that he returned to the store in January, 1900, but did very little work for eight or nine months, and had an aching, weak back all the time. Any jar or misstep or stubbing the toe would hurt his back. It was well on in the spring of 1901 be- fore he could lift anything at all heavy, and when seen was unable to scuffle, and felt a hurt in the back at times when "doubling up," especially before a storm. He would hesitate to lift weights greater than one hundred pounds, which was the most he was called upon to attempt, but within these limits he did anything that came his way in the hardware business. He had considered him- self well for the few months preceding the time he was seen in 1901. The case reported by Eskridge ' is quite in- accessible, and the history is of such interest that it seems desirable to give an extended abstract in this place. E. S., male, Germany, hostler. Mother died of brain fever in the thirty-fifth year. Maternal grandfather died of hemiplegia. Family history otherwise unimportant. The pa- tient had been a hostler for the preceding six years. For the preceding eight or nine years he had joint pains in damp weather. Five years ago he contracted syphilis. Four years ago he was lame for two months from pain in the sacral region, and the pain extended to the left side of the pelvis. Two years before the present attack he moved from his former home in Illinois to Denver. Since his stay in Colorado his joint pains seem largely to have disappeared. In July, 1892, he vi^as admitted to the Arapahoe County Hospital suffering from an attack of typhoid fever. During his convalescence he began to complain of pain in the sacral region. At first it was simply stiffness after sitting, with some pain when he attempted to get up. This passed off after he walked about for a short time. About the middle of September he left the hospital and returned to work, but the pain in the back and the parts around the left hip became so great that he was compelled to give up work and return to the hospital. After his return to the hospital he was confined to bed on account of pain. When, the patient stood both legs were straight and the gluteal folds were normal. The legs could both be abducted and adducted without pain. Ex- tremes of flexion and extension of the left thigh caused great pain. The back was painful on pressure over the first sacral spine, and the tenderness was limited to one spinous process. Pressure here caused pain to shoot down the posterior portion of the left thigh, and in the region of the small sciatic nerve on the same side. When the left leg was straightened and brought forward, it gave rise to pain in the sacral region of the spine, in the left hip, and in the posterior portion of the left thigh. When the leg was brought backward, the pain complained of was chiefly in the sacral region of the spine. There was no paralysis or paresis of any of the muscles, the limitation of the movements of the left leg was due simply to pain. Reflexes : Knee jerks, both greatly increased; ankle clonus, ab- sent; plantar reflexes, right fair, left more marked than right; cremaster reflex, right normal, left absent; lower abdominal reflex, absent; epigastric reflex, right present, left absent. Temperature, localization and muscular senses were all normal. A condition of slight hyperesthesia was found over the left leg, and over the space one-half the size of a quarter on the front of the right thigh there was found the condition of • anesthesia. There were no other disturbances of sensation found. Pressure over -both ilia at the same time, so as to press the ilia upon the sacrum, at the sacro-iliac synchrondrosis, caused considerable pain in the left sacro-iliac joint, and the pain extended from one side of the pelvis to the other. The hip-joint seemed entirely free from pain. There was no tenderness over any of the nerves of the legs. In the differential diagnosis myelitis, sciatica, hip-joint disease, tumors of the bones of the pel- vis, malingering, localized pachymeningitis, and some form of bone disease were discussed,. After consideration, all of the above were ruled out ex- cept the localized meningitis or the bone lesion. The meningitis if present was considered to be an affection of the external surfaces of the dura, in- volving the sheaths of the nerves leaving the cord, on only one side, but not involving the cord itself. Now as an external pachymeningitis so rarely oc- curs in the absence of bone disease, it is fair to presume that we have bone disease and pachymen- ingitis associated, although it is possible to explain 10 all the symptoms without the presence of a men- ingitis. Meningitis, however, would not account for all the symptoms in the absence of bone dis- ease. The pain in the left sacro-iliac synchron- drosis and the tenderness over the first sacral vertebra are due to an affectioh of the bone or its periosteum. There is no statement as to the final outcome of the case, although it seems to have been well on the way to recovery when reported. An accurate pathological diagnosis of the cases here reported is very difficult, and the same dif- ficulty has been met by other reporters of similar conditions. This difficulty is reflected In the titles under which they have reported their cases. Al- though Gibney had formed a theory that the under- lying pathological condition was a perispondylitis, he reported' his cases, the first, so far as I know in the literature, under the title, "The Typhoid Spine." Osier, writing later, reports three cases as a neurosis. In the last three years histories of cases presenting certain points of similarity to those reported by American observers have been published in Germany as examples of typhoid spondylitis. While an exact determination of the pathological condition in a disease in which the outcome has been uniformly favorable is prac- tically impossible, the writer wishes to review some of the facts upon which his conclusion is based, that in a certain number of the cases re- ported the symptoms have been due to an inflam- matory process involving one or more of the vertebrae, or their periosteum or cartilages. In 80 far as I have been able to collect it the litera- ture of the subject consists at present of fifteen articles reporting twenty-one cases, the majority of which I have been able to consult at first hand. I have had access to seventeen or eighteen of the histories of these cases in the report of the ob- server or in a few instances in a satisfactory ab- stract. From the standpoint of diagnosis I wish 11 to call attention particularly to the occurrence of deformity, and to the frequent association of fever with the affection. As to the occurrence of deformity in eighteen cases, the records are silent on this point in six, deformity is definitely stated not to have occurred in six, but in six others there is fair evidence that a deformity was present at some time in the course. In three cases, 12, 13 and 14, a kyphosis is definitely stated to have been observed by the re- porter. Konitzer," in the introduction to a case he reports, remarks that a new symptom complex has been made by Quincke " which appears as a disease of the lumbar and sacral region after ty- phoid fever, causing very severe pain and swelling in these parts. This statement leads me to believe that Quincke's cases, two in number, had some de- formity. I have not been able to consult his orig- inal article, and the abstract of the histories of his cases given by another writer makes no mention of deformity. In another case " the patient claims to have noticed a prominence in the lower lumbar re- gion which had disappeared, however, before she was seen by the reporter. It is well known that inflammatory processes secondary to typhoid fever are usually, or at least often destructive, and it is a rather remarkable fact about these cases, if they are due to an inflammatory process, that they have in no case gone on to suppuration, nor does the deformity usually persist. In five of the cases in which a deformity was noted it disappeared in a short time, in one only being present at the time of the report. Gibney' has reported a case of tor- ticollis in which there was a well-marked deform- ity of' a number of the cervical vertebras with an enlargement of their lateral masses in a man of forty-five who at the age of twenty-two had a febrile disease of several weeks' duration, pro- nounced by his physician typhoid fever, and fol- lowed by a painful affection of the cervical spine. Gibney reports, also,, that a painful affection of 12 the hip which he observed following an attack of "' typhb-malarial " fever left some limitation of motion. He mentions these cases as items which strengthen his belief in the correctness of his opinion that the condition is due to a low grade of inflammation. The presence of a febrile re- action associated with the condition under consid- eration, and not to be explained by any coexisting condition, might, doubtless, be considered, next to the appearance of a kyphosis, the most cogent reason for thinking that the underlying patho- logical condition was inflammatory. In a total of seventeen cases no statement is made by the reporter in this jjoint in five, in two cases it is stated that no febrile reaction was observed. In the remaining cases, ten in number, the tempera- ture was elevated during some portion of the af- fection. In two cases slight febrile reaction is reported. In the remaining eight it seems to have reached at least 103° F. at some time in its course. For these and other reasons the writer feels that the true type of the " typhoid spine " is a symp- tom complex due to an inflammatory process, and that pain in the back following typhoid fever but due to other conditions should not be confused with it. At least one case has been reported ^^ in which there was a painful condition of the mus- cles of the spine, associated with degeneration, leaving behind a contraction of the muscle af- fected. There have also been observed cases of neurosis apparently not unlike that known as the railroad spine, although in some of these cases it is open to question whether this neurosis was the original condition or a later complication. The prognosis of painful affections of the spine subsequent to typhoid fever, in so far as recovery is concerned, is excellent. No deaths are recorded in the literature. Even allowing the two cases reported by Gibney, one a painful affection of the cervical spine with permanent deformity, the other a similar painful affection of the hip with 13 limitation of motion, to influence our prognosis as to deformity this is also excellent, for in no other cases in which the course has been ade- quately followed to determine the final outcome has any permanent deformity or limitation of motion been observed. From another standpoint the prognosis is very different. In very few cases has the trouble been of short duration. It has caused a very marked disability for weeks and more often for months, and in many cases has been characterized by a series of exacerbation arising spontaneously or brought on apparently by the most trivial causes. In view of the probable duration of a well-de- veloped case of typhoid spine it seems desirable to emphasize the necessity for the greatest cir- cumspection in the management of any case which during the febrile period or during conva- lescence, suggests the development of the typhoid spine by the existence of^ a weak, aching, painful back. For a well-developed attack, relief from pain is the first requirement. This end may be attained very largely in some cases by rest in a recumbent position. Additional relief may be given by various mechanical devices affording ad- ditional support to the spinal column, as by a jacket or brace. In a similar fashion, extension may be serviceable. Later a jacket may be of further assistance in supporting the back, in reas- suring the patient and in permitting him to get about at an earlier date than would otherwise be possible. For the relief of pain also antirheu- matic, sedative and hypnotic remedies may be employed, but unless the suffering is mild, opiates will be found necessary, from time to time, to make the patient's condition bearable. The use of nourishing food and of tonics will be indicated at appropriate times. Other medication seems not to be of any great value, although potassium iodine has been used in a number of oases with apparent benefit. 14 BIBLIOGKAPHY. 1. Cribney. The Typhoid Spine. New York Medical Journal, .1889, i, 596-598. Reports three cases; also one of an acute, extremely painful affection of hip, hearing the same rela- tion to the fever as did the affections of the back, and a fourth case in all probability of different nature as pointed out by Osier. Discusses probable pathology of the affec- tion. The literature of bone lesions in typhoid, etc. 2. Fussell. The Typhoid Spine. ^New York Medical Journal, 1898, i, 635. Reports one case. 3. Gibney. A Further Contribution to Typhoid Spine. Univer- sity Medical Magazine, Philadelphia, 1891, vol. iv, No. 2, p. 32. Reports case of torticollis, said to have originated after an attack of typhoid fever occurring twenty-three years previously. Reported for its bearing upon the pathology of the affection. Mentions the fact that a slight impair- ment of motion has persisted in the hip alluded to in first communication. 4. Eskridge. Periostitis of the pelvic bones following typhoid fe\er. Kansas City Medical Index, January, 1893. Reports one case. 5. Osier. On the Neurosis following Enteric Fever known as the Typhoid Spine. American Journal of the Medical Sci- ences, Philadelphia, 1894, N. S., cvii, 23-30. Reviews cases of Gibney's original article. Refers to possible case and reports two others. 6. Study. The Typhoid Spine. Medical Record, New York, 1894, xlvi, 109. Reports one case. 7. Romme. Le "Typhoid Spiife" etles neurosis post-typhiques Gaz. hebd. de med., Paris, 1894, xii, 662-664. Reviews Osier's article. Summary of Gibney's and of Osier's cases. 8. Newcomet. A Case of Typhoid Spine. International Medical Magazine, Philadelphia, 1898, vii, 697-599. Discusses the subject with references to literature. Reports one case 9. Quincke. Ueber Spondylitis Typhosa. Mittheil aus deu Grenzgeb., 1899, Bd. iv, H. 2. Reports two cases. 10. Konitzer. Ein Fall von spondylitis typhosa. Miinch. Med. Wooh., 1899, p. 1145. Reports one case. 11. Schanz. Ueber Spondylitis Typhosa. Arch. f. klin. chir., Bd. Ixi, H. 1. Gives history of two cases reported by Quincke and of one by Konitzer. Reports one case. 12. Neisser. Exact title not known. Deutsch. Aerzte Zelt., 1900, H. 23. Vergl. ferner. herz. Zeitsohr. f . Orthop. Chir., Bd. viii. No. 1. Reports two cases. 13. Kuhn. Ueber Spondylitis Typhosa. Miinch. Med. Woch., 1901. Gives histories of cases reported by Quincke, Konitzer Schanz and Neisser. Reports one case. 14. Taylor. The Typhoid Spine. Philadelphia Medical Journal. „ „ ^°'' ■>'"'' ^O' '^^' P- 113*- Reports one case. M. Crittenden. Salient Points in an Epidemic of Typhoid Fever Based upon Fifty-five Cases. New York Medical News, 1901. vol. Ixxix, No. 1, p. 12. [From The Johns Hopkins Hospital Bulletin^ Vol. XVI, No. 168, March, 1905.] EEPEATED COPIOUS HEMOPTYSIS FEOM AN AOETIC ANEUEISM, EXTENDING INTO THE EIGHT LUNG AND FINALLY EUPTUEING INTO THE PLEUEA/ By Thomas Wood Clarke, M. D., Late Medical House-Officer, Johns Hopkins Hospital, Baltimore. Resident Physician, Lakeside Hospital, Cleveland, Ohio. The case I wish to report is that of J. McD., an Irishman, [98] 39 years of age, who was admitted to the Lakeside Hospital, Medical Number 4303, in the service of Dr. Edward P. Gush- ing, on July 13, 1904, complaining of pain in the right upper chest, cough, and expectoration of blood. His father had died of consumption. Beyond this the family history was negative. The patient had always been a healthy, strong man, by occu- pation a laborer, and had done hard work all his life. He had rheumatism twelve years ago, and slight attacks of the same disease ever since then. As to specific history, the pa- tient admitted having gonorrhoea three times, the last attack eight years before, and a soft chancre eighteen years ago. No history of a primary sore or secondary syphilitic lesions could be obtained. Except for the minor diseases of childhood, the patient had had no further illnesses. He used alcohol to excess. The patient dated his present trouble from a year before admission to the Hospital, when he began to have discomfort in his chest, especially the right front, shortness of breath. ^Read before the Clinical and Pathological section of the Cleveland Academy of Medicine, on Nov. 4, 1904. (1) [98] dizziness, ringing in the cars, and attacks of weakness. His general condition grew worse until two weeks before admission to the Hospital, when he began to cough and expectorate small amounts of blood. At this time, the pain in the chest grew much more severe, and on admission it extended to the right shoulder and down the right arm to the elbow. On the day- he came into the hospital, he had had an especially profuse haemorrhage, but was quite indefinite as to the exact amount. During the past year, he had lost twenty pounds in weight. On examination, the patient was found to be a well built, muscular man, coughing frequently and expectorating con- siderable quantities of blood-stained mucus. His voice was husky but not brazen, and the cough not the typical "goose cough." His respirations were normal, the pulse regular and of good volume, 88 per minute. The left pulse seemed slightly more full than the right. They were synchronous. The face was of good color; pupils equal and normal; slight arterio- sclerosis. There was a very slight tracheal tug. On inspec- tion of the chest, it was found to be symmetrical and the ex- pansion equal. Over the first and second right inter-spaces, for a distance of 11 cm. from the mid-line and over the third inter-space for 4 cm., there was a marked visible pulsation, [99] but no bulging of the chest. On palpation over the pulsating area a slight systolic impulse could be felt, followed by a marked diastolic impact. There was no thrill. On percus- sion, dullness was obtained in this region over an area of 11.5 cm. laterally, and 9 cm. vertically downward from the ri^ht clavicle. The dullness did not extend to the left of the sternum. In the right back, the dullness extended from a point 3 cm. above the angle of the scapula upward for 8 cm., and included the entire space between the spinal column and the scapula. No pulsation or tumor was noted in the back. On auscultation over the right upper chest could be hoard a markedly accentuated diastolic shock, and a loud friction rub, synchronous with the respiration. No bruit was audible. The lungs were elsewhere normal. The apex of the heart was slightly more to the left than normal, in the fifth interspace 10.5 cm. from the mid-line, and the cardiac dullness 13.5 cm. out, or 4 cm. outside of the (2) nipple. Except for a marked increase in the intensity of the [99] aortic second sound, the heart sounds were normal. The phy- sical examination was otherwise negative. The unquestioned diagnosis of aneurism of the thoracic aorta was made, and the attempt made to put the patient on a modified Tufnell treatment. He, however, objected so ser- iously to the dry diet, and so absolutely refused to remain quiet, that this had to be given up. For ten days everything went well. After the second day, the bleeding stopped, and the cough and pain were greatly relieved. On July 23, how- ever, after a sudden movement, the patient commenced cough- ing, and spat up four ounces of almost pure red blood. This was the beginning of a remarkable series of sixteen hgemor- rhages. The next was on July 39, 8 ounces, and from then they occurred as follows: August 2, 12 ounces; August 6, 32 ounces; August 9, 16 ounces; August 11, 36 ounces; Au- gust 17, 16 ounces; August 20, 6 ounces; August 21, 20 ounces; August 24, 28 ounces; August 28, 10 ounces; Sep- tember 2, 4 ounces; September 3, 2 ounces; September 5, 16 ounces; September 13, 8 ounces; and September 14, 7 ounces; in a period of seven and a half weeks a total of 225 ounces or slightly over 14 pints. This was almost pure blood, mixed with very little mucus. The hsemorrhages in each case, coming on after some exertion, as sitting up, or reaching for an article on the floor, lasted only a very few minutes, and accompanied paroxysms of coughing. At no time did he raise any fibrin or blood-clot. The haemorrhages were in each case stopped by morphia, at times as much as a grain being required. The patient became steadily more pale. The red blood count dropped from 4,864,000 to 2,304,000 and the haemoglobin from 65 per cent to 38 per cent in the first month after which, owing to an accident to the hsmoglobinometer, no record could be taken, but from the patient's appearance, it must have been much below this. Two days after the last hemorrhage on September 15, 1904, the patient during a fit of anger, suddenly sat up in bed, became very weak, broke out into a cold sweat, and said he was dying. The pulse rose from 80 to 125 but remained fairly strong. Th*e respirations be- came labored and finally stopped, the heart continuing to beat (3) 199] for some time after this. The patient died in about twenty- minutes. During the last few daj'S of the patient's life, signs of fluid were evident in his right chest, causing flatness at the base. The probable size of the aneurism and the danger of puncturing it made it seem inadvisable to use the exploring needle. Several attempts were made to examine the patient's larynx, but no good view could at any time be obtained. During the illness, it was a matter of much speculation among the' attending physicians why an aneurism, with a per- foration large enough to cause such repeated large hsemor- rhages, should stop bleeding at all, and not cause an imme- diate lethal haemorrhage. The explanation of this is perhaps found in autopsy findings for the records of which, and for aid in summarizing the same, I wish to thank Dr. William T. Howard, Visiting Pathologist, and Drs. D. H. DoUey and J. H. Bacon, Eesident Pathologists of the Lakeside Hospital. AUTOPSY REPORT. The following abstract is taken from the autopsy protocol : Autopsy by Dr. David H. Dolley, September 16, 1904, aut. Ko. 536. The body is that of an adult male, 170 cm. long. Eigor mortis marked. Pupils equal and dilated. The chest is flat, the supra- and infra-clavicular fossse well marked. There is no bulging or other abnormality of the thorax. The abdomen is scaphoid. The extremities are free from wounds, scars, and deformities. The superficial glands are not pal- pable. Thorax. — The sternum, ribs, and costal cartilages are normal. The left lung over-laps the heart to more than the xisual degree. The upper lobe of the right lung is the seat of a rather firm mass. The middle and lower lobes are not visible. The right pleural cavity contains 3000 cc. of dark blood-tinged fluid containing flakes of fibrin. The right parietal pleura is smooth and free from adhesions. The left pleural cavity contains a small amount of clear fluid. The pleura is smooth and free from adhesions. The thoracic con- tents were removed en masse. Heart. — The pericardial cavity contains 200 ec. of clear fluid. The parietal layer of pericardium is somewhat thick- (41 ened, as is the visceral layer over the right ventricle. At the [99) latter point, the epicardium is red and rough. The peri- cardium is adherent to the upper lobe of the right lung, over the surface of which it extends for a considerable distance. The heart muscle on section is pale and flabby. The segments of the mitral valve are moderately thickened. The segments of the aortic valve are normal in appearance and show no thickening. The tricuspid and pulmonary orifices are of normal size and appearance. The right auricle is markedly compressed by the aneurism to be mentioned later. All the valves are apparently competent. The Aorta. — The aorta just above the valve measures 7.5 cm. in circumference, at the junction of the ascending and transverse portions of the aorta, 8 cm. Below this, the vessel ib not dilated. The aorta throughout is the seat of a number of large and small irregular thickened areas, which project slightly into the lumen. Some are calcified and rough. The large branches of the aorta are normal. On the right side [lOO] of the aorta, 3 cm. above the valve, there is an oval open- ing, 3.5 by 1 cm. in diameter, leading into a spherical sac, 9 cm. in diameter, which is filled with laminated fibrin, the superficial layers of which, are soft and of a grayish-red color, while the deeper layers are white and firm. The opening of the sac into the aorta is partly closed by a rather firm mass of grayish-white thrombus. This aneurismal sac, which is round in outline projects upward and somewhat backward, directly into the upper lobe of the right lung, which incases it through- out two-thirds of its extent. Anteriorly the sac is adherent to, and is covered, to a considerable extent, by somewhat thickened pericardium. The walls of the aneurism are com- paratively thin, being thickest at the base, where they average 8 mm. and thinnest at the point just opposite the opening into the aorta, at which point the walls are lost in the lung tissue. Over a large part of its extent, the walls of the sac and the visceral pleura are continuous. Microscopically, the wall con- sists of firm fibrous tissue, organising fibrin and inflamed pleura, with occasional scattered traces of markedly com- pressed alveoli. At the point directly opposite the commun- ication between the aorta and the aneurismal sac, there is (5) [1001 an opening between the latter and the pleural cavity. At this point, the pleura is ruptured, and a rough mass of grayish- red fibrin projects into the pleural cavity. This opening is a linear tear, 5 cm. long. Adherent to the upper lobe of the lung, near this point, there is a fresh blood clot, weighing 700 grams. On transverse section through the sac and the right lobe, the whole interior of the sac is filled with, at some places firm, and at other places friable, clot. The aneurism occupies nearly the whole of the upper lobe of the right lung, and two-thirds of the extent of the aneurism is situated in this lobe. The bronchi leading to the upper lobe contain no blood, but on section, small bronchi can be traced into the aneurism sac. The lower and middle lobe of this lung are collapsed, airless and rather firm. The pulmonary arteries are normal throughout. The Left Lung. — The left lung is voluminous and markedly emphysematous. On section, it is oedematous throughout. The bronchi of both lungs, the trachea, and larynx are normal in appearance. Abdomen. — The abdominal muscle and subcutaneous tissue are normal. The diaphragm on the right side projects 4 em. below the costal margin. The liver is displaced downward and to the left, and reaches a point 5 cm. below the umbilicus, the right lobe lying directly in the middle line. The stomach is- displaced downward into the left iliac fossa. The ileum and jejunum and the transverse colon are displaced into the pelvis. The other abdominal organs are in their usual posi- tions. The abdominal cavity contains a moderate amount of clear fluid. The peritoneum is smooth and free from ad- hesions. The liver is of ordinary size and shape. The biliary system is normal. The spleen, kidneys, pancreas, stomach, intestines, and other organs show no pathological changes. Anatomical Diagnosis. Arterio-selerosis of the aorta with sacular aneurism of the ascending arch of the aorta, projecting into and occupying a greater portion of the upper lobe of the right lung. Compression atelectasis of the right lung. Eup- tured aneurism and extensive heemorrhage into the right pleural cavity. General enteroptosis. Pressure upon the right auricle. Oedema and chronic passive congestion of (6> both lungs. Hydro-thorax, hydro-pericardium, and hydro- [lOO] peritoneum. Communications between aneurism and small bronchi. Right Lung Showing Aneurism. U. L. = Upper Lobe and aneurism. M. L. = = Middle Lobe. L. L. = Lower Lobe. A. C. = Aneurism Cavity opened. R. = Rupture. P. = Pericardium. LITEEATUEE. The point of especial interest raised by this case is that of uon-lethal hjemorrhages in cases of thoracic aneurism. After looking up with some care, all the literature available in Cleveland, I have been compelled to conclude that such cases are very rare. Though cases have occurred, in which one or two severe hemorrhages have been followed by months or even years of exemption, I have not found reference to any such series of large hemorrhages as are here reported. With the (7) [100] facilities at my disposal, I have been unable to make a com- plete review of the literature, but have collected six cases showing one or more points of similarity to the present one. In 1847, the case of the English surgeon, Mr. Liston was reported. This gentleman, after one profuse haemorrhage, was absolutely well for three months. Then, after two months in which he steadily failed, and expectorated considerable rusty sputum, he died, without further haemorrhage. At autopsy were found three old perforations into the trachea, blocked by blood clot. (1) Gairdner, in 1859, reported a case in which the patient was ill ten years. Four years before he died, he had two profuse rioi] hemorrhages from his lungs, followed by some staining of the sputum. After this, he had no more haemorrhages, though for the last six months, some staining, until a hsemorrhage of eight ounces caused suffocation and death. The autopsy re- vealed an aneurism of the descending aorta which had perfor- ated the left bronchus and trachea, the perforation being filled with old clot. The left lung was collapsed but not involved. (2) Gairdner also states that up to that time, 1859, nine cases of thoracic aneurism had been reported which had had hemorrhage two months or more before death. He gives no references except to the case of Mr. Liston, and I have been unable to find these cases. Two cases are reported by Dr. Osier somewhat similar. The first, in which the patient had a haemorrhage of two quarts and three weeks later slight haemoptysis. Dr. Osier re- ports eight months later as a cured aneurism. (3) The sec- ond is a man, who, after one profuse haemorrhage, lived four weeks, and then dropped dead without further bleeding. At autopsy, this case was very similar to the one here reported. It was an aneurism of the ascending arch, a large part of the wall of which was made up of pulmonary tissue, the trachea not being involved. In this case, as in the present one, death followed perforation of the aneurism into the right pleural sac, with internal haemorrhage. (4) Peacock reported in London still another case where the descending arch was involved, the aneurism wall being com- posed of the upper lobe of the left lung. This case spat up (8) small amoiants of blood for some time and finally died from [ loi] rupture into the left pleura. (5) Involvement of the lung in aneurism is a comparatively rare occurrence, but is one thing to be thought of in the considera- tion of pulmonary hasmorrhage, especially where there is a sus- picion of thoracic aneurism. That an aneurism may cause extensive destruction of lung tissue without haemorrhage is ghovm by Johnson's case. In this, the patient had symptoms of chronic pleurisy, and had no haemorrhage until the fatal one. Aneurism was not suspected, but at autopsy one was found, of the descending aorta involving the lower left lobe of the lung, the aneurism walls being only one quarter of an inch thick, composed of fibrin, indurated lung and pleura. (6) This of course, must have existed for some time, but the bronchi were presumably compressed and obliterated in the wall of the sac. Prom the above cases it would appear that, while haemop- tysis is a common fatal termination in thoracic aneurism, a very large haemorrhage may occur in this disease, without causing the death of the patient, and that occasionally a pa- tient may have one or more large haemoptyses, and finally die from some other complication of the aneurism. The point of bleeding may be from rupture into the trachea or bronchi and be stopped by the opening being plugged with fibrin, or it may occur from involvement of the lung tissue itself. In the latter case, it is probable that the small bronchioles open into the wall of the sac, but are so compressed by it, that they normally remain closed, or are covered by fibrin. On exertion, or mov- ing, either the fibrin shifts its position, or the straining of coughing opens the ends of the small bronchi and the blood leaks out. On the patient again becoming quiet, either nat- urally or by means of narcotics, the original condition is re- sumed and the bleeding is temporarily arrested. In closing, I wish to express my thanks to Dr. William Osier of Baltimore, for some useful suggestions and for information concerning his cases, and to Dr. Edward F. Gushing of Cleve- lajid for permission to report this case from his service at the Lakeside Hospital. (9) [101] BIBLIOGEAPHY. (1) Lancet, London, December, 1847, p. 633. (2) Trans. Eoy. Med. and Chir. Soc, London, 1859, p. 189. (3) Phila. Med. Times, 1888, XIX, p. 149. (4) Ibid., 1889, XIX, p. 223. (5) Trans. Path. Soc, London, 1863, XIII, p. 39. (6) Lancet, London, January 12, 1867, p. 44. (10) [Reprinted from The Cleveland Journal of Medicine, November, 1901] Rydropbobia witb Report of Cbrce Cases. BY HENRY S. UPSON, M. D., Professor of Diseases of tlie Nervous System in the Western Reserve Medical School, Cleveland. OF the three cases which I wish to report tonight, two occurred under my own observation in 1899, and the third is of recent occurrence in the practice of Dr Thomas Hubbard of Toledo, who kindly allows me to report it with my own. The first patient, Mr L., 71 years of age, was seen at the request of Drs J. F. and W. S. Hobson on May 30, 1899. Until the development of this difficulty he was in good health. During the very cold weather early in February he went to the barn to milk, and was bitten on the hand by a large tomcat, which sprang upon him without obvious reason. He had been in the habit of petting the cat, which seemed to be fond of him. He allowed about one-half hour to elapse, and the wound was then cauterized and treated with an antiseptic dressing. The cat had been observed to be sick for two or three days previous to this time, was very irritable, wild, and apparently delirious. About three or four days after the bite was inflicted, thecat died. The body was frozen stifif. A rabbit was inoculated from this cat's spinal cord by Dr Perkins, and at the time when the patient was first seen, some four months after the bite, the rabbit, and a cat which had been bitten by the presumably rabid one, were both under observation in the laboratory of the Lakeside Hospital. They were living and well. The patient was quite well until May 25 when pains began to shoot up from the hand along the arm. This continued for two days. On the 28th the patient was restless, but had a fairly good night's sleep after taking two teaspoonfuls of bromidia and 1/50 grain of hyoscyamin sulphate. When seen by me on the morning of May 30 he had developed considerable diffi- culty in swallowing. This had been noticed for the first time the night before. It was especially hard for him to swallow water or other fluids. He could not eat any food, although the night before he had eaten a few straw- berries and that morning he had eaten two strawberries. He found, how- ever, not much difficulty in taking pills or capsules. Tn attempting to swal- low water he manifested great distress, breathed very fast, sighed deeply, and usually gave it up, although he did manage at times to swallow some fluid. His mind was perfectly clear, he talked rationally, not only on the subject of his illness, but on other subjects, and did not seem very mi:scope all tlie char- acteristics of chyle." The fluid at this time amounted to about IJ ounces. During the next 2 days the cavity again tilled up with about 2 ounces of fluid, and was then packed with iodo- form gauze, after which the discharge of chyle increased greatly, amounting to IJ to 2 pints in 24 hours. This soon resulted in constitutional effect, and a few days latei- the patient was removed to a hospital, the lower end of the wound opened, and a small vessel readily found from which chj-le was escap- ing; this vessel liad been torn tiirough, the opeuing being about the size of a knitting needlp. The vessel was clamped, the clamp remaining on .'5 days. Tlie condition of the patient began to improve immediately and recovery was complete. Ca.se 7. — Operator, C. A. Porter. Reported by H. W. Gushing." Operation, December, 1897. Diagnosis, tuberculous adenitis. Glands in left side of neck. Duct reached 4 cm. above the subclavian vein. Wound of duct J inch above entrance into vein caused the escape of a clear yellowish fluid at the rate of about 1 to 2 drams in 10 minutes. Wound repaired with 3 fine Lembert sutures. There was still a slight discharge of lymph which was thought to come from a small radicle. Wound opened on the second day, discharged 15 days. No constitu- tional effect. Recovery complete. Case S.— Operator, W. S. Halsted. Reported by H. W. Gushing. Operation, November 5, 1895. Female, age 66 years. Diagnosis, candnoma. Tumor of left breast with enlarged axlflary glands. Complete excision of breast, pectoral muscles, axillary and supraclavicular contents. • Wound healed by first intention. Ten days after operation wound above clavicle opened, and 4 or 5 ounces of milky fluid evacuated. Gauze drainage introdncied, following which there was profuse dis- charge for 8 days. During this time the patient lost 10 pounds. Wound explored, discharge observed coming from level of omo- hyoid muscle where it crossed the jugular vein. Gauze packing introduced. No subsequent leakage. During the next 15 days patient gained 23 pounds. Subsequent recovery uneventful. Died in 1896 of internal metastases. Case 9.— Operator, H. W. Gushing. Reported by H. W. Gushing. Operation February 25, 1898. Female, age 62 years. Diagnosis, carcinoma. Recurrent supraclavicular glands on left side, following excision of left breast 2 months previous. Dur- ing dissection the thoracic duct was wounded about 1 cm. above the subclavian vein. The vessel extended 4 cm. above its entrance into the vein. " A large branch, about the size of the usual silver probe, entered into this vessel just below the point of injury, and disappeared behind the subclavian about 2 em. to the left of the main duct." Wound 3 mm. long, allowing intermittent flow of considerable clear serovts fliud. Wound in duet closed with flne silk Lembert suture; no subsequent leak- age. Wound closed ; recovery uneventful. Writer claims first case of suture without leakage. Case 10.— Operator, ^Y. H. Lyne. Reported by W. H. Lyne.' Operation May 5, 1896. Male, aged 24 years. Diagnosis, stab wound in neck, left side. Rleeding had stopped when patient Avas seen. There persisted, hoAvever, an abundant milky fluid discharge. The wound was 1 inch long behind the left clavicle, parallel with the outer border of the sternomastoid, near its attachment, " necessitating a longitudinal wound of the thoracic duct." Wound cleaned, and packed with iodoform gauze; repacked 7 hours later, when discharge was found to have ceased. No subsequent leakage. Recovery uneventful. No examination of fluid reported. 5 Operative wounds of the thoracic duct. Report of a case with suture of the duet. Annals of Surgery, 1898, vol. 27, p. 719. 7 Stab wound of the thoracic duct, recoverv. Virginia Medical Semi-JTonthly, 1SH8, vol. 3, p. 278. Case 11. — Operators, W. E. Schroeder and S. C. Plumber. Reported by W. E. Schroeder and S. C. Plumber.^ Operation, date not given. Male, aged 24 years. Diagnosis, not given. Cystic tumor at base of neck on both sides. During dissection on left side the internal jugular vein was injured and ligated. Escape of lymph " was seen to come from small orifice, evi- dently an opening in a large lymphatic vessel, either in the thoracic duct itself or its large subclavian branch." Attempts to ligate the vessel were unsuccessful ; no attempt at suture was made. " From the horizontal position of the vessel it was taken to be, most probably, the large subclavian branch of the thor- acic duct." Discharge at operation was transparent, yellowish, and not excessive. Wound packed with iodoform gauze ; slight serous discharge lor 3 weeks. Recovery otherwise uneventful. Case 12. — Operators, W. E. Schroeder and S. C. Plumber. Reported by W. E. Schroeder and S. C. Plumber. Operation, date not given. Female, aged 23 years. Diagnosis, tuberculous adenitis. Glands in left side of neck. During the operation a milky fluid was seen to escape from "an injury to the thoracic duct, or one of its large branches." The vessel was found and lifted with forceps. Air was inspired through the wound. " The vessel was ligated," but chyle still discharged through a small opening, "which was also ligated." No subsequent leakage. Recovery uneventful, except for slight rise of temper- ature for a lew days following operation. Case 13.— Operator, F. B. Lund. Reported by F. B. Lund." Operation, December 18, 1898. Female, aged 34 years. Diag- nosis, carcinoma. Tumor in left breast. Excision of left breast with axillary and supraclavicular contents. Internal jugular and subclavian veins dissected out. Thoracic duet sought, but not found. No lymphatic injury suspected at operation. Pour days after, cervical wound opened and a " partially coagulated milky fluid" evacuated. The leakage was found to be quite brisk, requiring a change of large dressings every 3 hours. This continued for 4 days. The wound was packed lightly with gauze without effect, but subsequent firm gauze packing stopped the discharge. Convalescence uneventful. No consti- tutional effect mentioned. Case 14.— Operator, J. C. Warren. Personal commuuiea- tion.i" Operation, 1896. Female, age not given. Diagnosis, carcinoma. Recurrent supraclavicular glands 3 years after first operation. No injury of tlie thoracic duct was observed at the time ol operation, but 5 days afterward, a fluctuating tumor about the size of a Messina orange was evacuated. Discharge con- tinued from 1 to 2 weeks, but was finally stopped by a graduated compress held in place over the wound by a strip ol adhesive plaster. The character of the fluid was not mentioned, other than that it was " chyle." A similar operation was performed one year later, and the same occurrence took place. Patient died of internal metastases in 1898. The case was published in Dr. Warren's cases of cancer of the breast, but no allusion was made to the injury of the thoracic duct. Case 15.— Operator, A. Miinter. Reported by A. Miiuter." Operation, June 16, 1898. Male adult. Diagnosis, incised wound at base of neck, on left side, caused by a 1)1onv from an axe. The patient had bled a considerable amount, and was pale and weak. The wound had been dressed temporarily witli 8 Report of two cases of injury of the thoracic duct in operations on the neck. Annals of Surgery, ISilS, vol. 2S, p. 229. 8 A case of operative injury of the thoracic duct, Boston AUdical and SurgicalJournal, IS89, vol. 140, p. 354. 10 Also mentioned by V. B. Lund, and J. C. Warren, at a nieetuig of the Surgical Section of the Suffolk District Medical Society, January 4, 1899. Boston Medical and Kurgiiral Journal, 1899, vol. 140, p. 353. n Schnittvurletzung des Ductus Ihoracicus. Deutsche medlcini.sche Wochenschrift-, lS!)y, Mo. 48, p. 79Si. horse dung, which was a favorite means of stopping hemor- rhage in that community. The wound was 12 cm. long in the left supraclavicular fossa, reaching the clavicle, and dividing the clavicular attachment of the sternoma.stoid muscle. The external jugular vein was divided. From the depth of the wound a milk-white stream about the size of a straw was (low- ing from a small vessel which was thought to be the thoracic duct. No attempt at ligating the vessel was made, as conveni- ences and assistants were not at hand. The wound was cleaned and packed with iodoform gauze, securely bound down, and the patient sent to a neighboring hospital. The next day the wound was opened and the chylous discharge continued. The vessel from which it was e.scaping was readily ligated, the divided sternomastoid muscles sutured, and the wound closed. There was no further discharge, and recovery was uneventful. Cask 16.— Operator, D. P. Allen. Reported by D. P. Allen and ( '. E. Briggs. Case A, reported in full above. Case 17.— Operator, 1). P. Allen. Reported by D. P. Allen and C. E. Briggs. Case B, reported in full above. A numljer of these cast's have been collected pre- viously. Cheever, Case No. 1, 1875, is the first report found. Boegehold, in reporting his case, No. 2, 1883, made no mention of Cheever's case, although he appears to have searched for previous instances. Keen's report, 1894, 'included the cases of Cheever and Boegehold, to which he added the case of Phelps, No. -t, and his own, No. 5. dishing, 1898, again reported the four cases mentioned by Keen, collected in addition the cases of Vagedes, No. 3, Schwimm, No. 6, and added the pre- viously unreported cases of Porter, No. 7, Halsted, No. 8, and his own. No. 9. To tliese are now added the collected cases of Lyne, No. 10, Schroeder and Plumber, Nos. 11. 12, Lund, No. 13, Warren, No. 14, Miinter, No. 15, and the report of two new cases, Nos. 16 and 17. This last series includes the only two reported cases of accidental (nonoperative) wounds of the thoracic duct in the neck, Cases Nos. 10, 15, both of which, however, were handled in tlie same manner as operative wounds. Information contained in the standard textbooks of surgery on the injuries uf tlic thoracic duct is very meager and with reference to injuries of the duct in the neck there is practically nothing. One looking for information, therefore, is obliged to go to a good many sources. Boegehold, in l.S8;i, was the first to give any detailed consideration of the subject. He was followed by an admirable article by Kirchner in ISS,").'^ Both of these writers, however, approach the subject from tlie standpoint of wounds of the th(jracic duct in general and say practically nothing with reference to cervical wounds of the duct. Keen, in 1894, was the first to give any detailed consideration of cervical wounds of the duct, (lushing, in 1898, supplemented Keen's considerations by a slightly more extended report. As the number of '2 Ein Fall von rechtsseltigen Chylotliorax in Folge von Ruptur (ics Ductus thoracicus. nebst Htatistik and Kritikderbisherbekannten cinschlagig-cii Fiille, Archlv fiir klinisclie t'hlrurgie, 18S5, B. 32, S. ISfi. reported cases, Iiowever, has been nearly doubled since 1898, and as a number of points have arisen for consider- ation which had been previously disregarded, it seems permissible to add somewhat extensively to the work already done, even at the risk of a certain amount of repetition. Aiiafoiii!/. — The anatomy of the thoracic duct is of importance in this connection, with reference to two considerations : First, the relations of the duct in the neck with reference to preserving its integrity in low cervical dissections ; second, the anatomic distributions of the duct commonly spoken of as its anomalies, through which collateral circulation may take i)lace in case the duet is irreparably injured. What is considered the normal course of the duct in the neck is ilescribed in the standard works on anatomy as the course taken between the esophagus and the left subclavian artery, reaching as high as the lower portion of the seventh cervical ver- tebra, curving over the apex of the jjleura and the left subclavian artery, and terminating in the outer side of the angle formed by the internal jugular and subclavian veins. In taking this couise its highest point is about 3 to .'i..') cm. above the upper margin of the sternum, but lying as it does to the inner side of the arch of the left subclavian ai'tery, it is so de(;ply j)laced in the neck that there is seldom danger of injury, altliough its height above the sternum is considerable. Nevertheless, a num- ber of instances have been observed in which the arch of the duct has reached considerably above the point men- tioned, and an instance is mentioned '^ in which the arch of the duct was .5.5 cm. above the sternum, reaching the thyroid gland. In the case of Gushing, Ko. 9, and Por- ter, No. 7, the duct was 4 cm. above the innominate vein, and in Case No. 16 reached about 5 cm. above the sternum. Keen, Case No. ">, does not mention the height of duct above the vein, but it was probably abnormal, as the vessel itself was injured in open dissection. It seems probable that operative wounds of the main duct have occurred only in instances of abnormally high arch, and that the other wounds mentioned have been injuries of some of the radicles only. With reference to the anomalous distribution and ending of the thoracic duct, marked variations have been observed. They are of practical interest only so far as they tend to strengthen the belief in adequate col- lateral circulation following injury of the duct itself, or one of its branches. Without going into extensive details, the following variations may be mentioned : ^V double duct entering the innominate vein on both sides of the neck ; a duct entering the vena azygous major, the lumbar or vertebral veins, or the A'ena cava ; a duct '■1 Honle'.s Anatomy, 1S76, vol. 3, p. 458. entering the right internal jugular or right innominate only, which condition, however, is usually associated with an anomalous right subclavian artery or a right aortic arch. The number of cases of multiple termina- tions of the duct in the left internal jugular, the subcla- vian and innominate veins are too numerous to record, and it seems probable that subsequent careful observa- tions will confirm the opinion that in a considerable proportion of cases such multiple terminations exist. From an anatomic standpoint, therefore, tiic probability of adequate collateral circulation after injury to one of the large terminal brandies of the duct seems extremely strong, and as injuries of the main duct itself occur so near its entrance into the vein, it seems quite likely that satisfactory c(jllateral circulation would folloA\' in such instances. Histologic features of tlie duct have a practical value relative to the contraction of the duct following injuries. In general the structure resembles that of the artery, with an intima, a muscular media, and connective tis- sue adventitia. The striking difference, however, is the abundant distribution of elastic fibers through all three of these layers, being quite marked even in the muscular coat ; the adventitia, also, is considerably thicker than is seen in arterial structures. The presence of these elastic fibers doubtless explains the variation observed in the size of the duct. When distended, it may be as large as 4 mm . in diameter, while in its collapsed condition it is very small and almost unrecognizable. For the use of those who are interested in looking up cases of anomalous distribution of the duct, a section of anatomic references is given at the end of this article. Physiology. — Consideration of the physiology of the lymphatic system is of very marked practical importance from several standpoints. That the thoracic duct pos- sesses a high degree of contractility would be strongly suggested by its histologic structure, and has been proved by theoretic and practical experimentation. Boegehold mentions the observations of Dittrich, Ger- lach, and Herz" on hanged criminals, where the elec- trical contractility of the duct was observed as long as 36 minutes after death ; and the observations of Henle and Pfeuffer, in which contraction was observed I4 liom-s post mortem. Experimentally with dogs, the thoracic duct has been tied above, and after consider- able pressure has accumulated by the filling of the recep- taculum chyli, the duct has been incised longitudinally between the receptaculum and the ligature ; in several instances the contractility was sufficient to close the opening and preserve the integrity of the duct. It is of » Prager Vierteljahrsschrift, Feb. 3, 1851. JO interest here to mention tiiat the pressure in the thoracic duct during the interval of digestion is equal to from 9 to 15 mm. of mercury, '» which is considerably more than the blood pressure in the jugular vein, 3 to 4 mm. of mercury, and less than that of the carotid artery, about 70 mm. of mercury (in rabbits).'* The physiologic aspect of complete occlusion of the thoracic duct, as observed in disease and animal experi- ments, are of importance with reference to the establish- ment of collateral circulation and rupture of the recep- taculum chyli on the one hand, and the state of nutrition sustained by the organism on the other. The complete occlusion of the thoracic duct by disease without serious interference with the flow of chyle, has been observed. Such an instance, due to tuberculous thrombosis of the duct, is mentioned by Welch." Other cases have been recorded. Occlusion of the duct, due to a pathologic process is, however, gradual to a greater or less extent, and the possibility of collateral circulation under such circumstances seems much greater than where the duct is suddenly occluded, as by ligature. A considerable amount of experimental work has been done on animals, mainly dogs and horses. This work was carefully reviewed by Boegehold, who collected the detailed results of these observations. He found that ligation of the duct resulted in complete obstruction of lymphatic circulation with rupture of the duct in but few instances, and that collateral circulation as a rule was rather readily established. With reference to the degree of nutrition sustained after ligature of the duct, the experimental observations of Schmidt-Mulheim have been of marked importance. It was found that the absorption of albuminous material was practically unafftected, and that the nutrition of an animal could be maintained for a considerable time on such a diet. These experiments were performed on dogs on the supposition that the thoracic duct is always single in these animals, and that ligation of the duct in the neck prevented absolutely the entrance of chyle into the circulation. It has been since observed that the thoracic duct in the dog is subject to some variation, as in man, which deprives the experiment of a certain amount of value. It is a well-established physiologic fact, however, that the absorption of fat occurs almost entirely through the medium of the thoracic duct ; that albuminous material is absorbed directly without enter- ing the lymph flow of the thoracic duct, and that carbo- hydrates find their way through the duct only to a very limited degree, about 1^ of the total absorption. It 16 Weiss, M., Vlrchow's Arohlv, B. 22. S. 526. 10 Foster, M., Text Book of Physiology, 189S, pt. 1, p. 205. " Transactions of the Association of American Physiolans, 1889, vol. 4, p. 76. n can, therefore, be stated with considerable confidence that independently of the lymphatic circulation, a high degree of nutrition can be sustained by the ingestion of proteids and carbohydrates. Several other physiologic facts may be mentioned here, the pertinence of which will be indicated later. With reference to the rapidity of the absorption of fat, the observations of Erben in a case of chyluria are of interest. He found that fat appeared in the urine within 2 or 3 hours after ingestion. Schafer states that absorp- tion of fat in a dog, after a full meal, continues for 30 hours. He also quotes Zawilski, who claims that the amount of fat in the intestine during digestion is practi- cally constant, and the rapidity with which it passes from the stomach depends largely upon the rapidity with which it is absorbed by the intestine. This view appears to be generally accepted by physiologists. It seems rea- sonable to conclude, therefore, that it would pass more rapidly from the stomach following a period of starvation, after fat absorption from the intestine had largely ceased. But as it is known that fasting diminishes, to a certain degree, the power to form normal digestive secretion, it is probable that fat digestion would be slightly less rapid after starvation. Harley, in speaking of normal absorp- tion of milk fat, concludes from a number of experiments performed on dogs that the maximum rate of absorption is reached about 7 hours after taking food, after which time it continues at about the same rate until completed, regardess of the amount of fat introduced into the stomach. With reference to the length of time liquid or semi-liquid food remains in the stomach, Beaumont's observations on Alexis St. Martin are extremely importr ant, although it is not at all unlikely that the artificial conditions existing in St. Martin altered somewhat the peristaltic action of the stomach. He found that bread and milk introduced through the gastric fistula was completely passed frorn the stomach into the intestine by the end of 3 hours. Absorption of fat would, of course, not occur until after the material had entered the intestine. The nature of lymph and chyle are of importance in diagnosis. Foster, speaking of lymph, says, " Broadly speaking, we may say that all the substances present in blood plasma are present also in lymph, but are accom- panied by a larger quantity of water." The presence of fat in lymph forms the distinction between lymph and chyle. The amount of fat in chyle in the thoracic duct varies, about 5% being the common amount. In dogs it has been found to vary from 2/c to 15^ . This increase is due almost entirely to the presence of neutral fats. A small part of the fat is present in fat globules of con- siderable size, but the large proportion of it exist- in a very minute stage of subdivision, rcrieiiibling under the microsfdpe anioriihdirs urates, and possessiug " Brown- ian " rnovenients. This minute subdivision of the fat constitutes what is commonly spoken of as the "molecu- lar basis" of chyle. Lymph is almost colorless, resem- bling serum which one sees coming from wounds. Chyle reseml)l('s milk in appearance, and ma.\- or may not coagulate spontaneously. Spontaneous coagulation is due to the presence of fibrin, and when observed in wounds of the duct is probably due to the small amount of blood collected with the fluid, as chyle removed from an internal cavity, as in cases of chylous ascites, does not coagulate spontaneously. The most careful and exhaust- ive examinations of chj-lethat have l)een found recorded are mentioned by Wliitlain an article on chylous ascites. This contains a careful investigation by himself of the fluid drawn from the abdomen in a case of this charac- ter, and a very extensive chemical examination by Hay, of Aberdeen ; to these we w(juld refer those wishing more extensive infiDrmation. For clinical ])urpose8, however, the presence of a milky fluid containing fat in minute subdivision is sufficient for making a diagnosis of chyle. The detailed references mentioned above may be found under Phymology at the end of the paper. Having considered some theoretic and practical ques- tions from an anatomic and physiologic standpoint, a more detailed study of the cases will be made. In 6 of the cases, Nos. .3, 6, 8, 13, 14, 17, injury of the duct or its bran(;hes was not suspected at the time of opera- tion, while the remaining 11 cases, Nos. 1, 2, 4, 5, 7, 9, 10, 11, 12, 15, If), it was known beyond reasonable doubt to have occurred. The subclavian and internal jugular veins were exposed, but uninjured in 7 cases, Nos. 2, T), 7, 9, 18, 16, 17. Exposure of the veins was not mentioned in 4 cases, Nos. 1, 8, 12, 14, although in case No. 8 they were undoubtedly exposed. The subclavian was injured in 2 cases, Nos. 1, 3 ; the internal jugular vein in 2 cases, Nos. 4, 11. In tlie 2 cases of accidental injury, Nos. 10, 15, the question of injury to the large vessels could not be determined, though such an injury seems probable in case No. 15, where there was consider- able hemorrhage. In 4 cases only, Nos. 5, 7, 9, 16, was the injury to the duct itself absolutely determined, and the anatomy of the parts satisfactorily developed. In case No. 6 there was a longitudinal tear } inch in a ves- sel \ inch in diameter ; the distance above the innomi- nate vein was not mentioned. In case No. 7 the duct was wounded \ inch above the innominate vein, size of opening not being given. In case No. 9 there was a longitudinal wound of 3 mm., 1 cm. above the subclavian vein. In case No. 16 the wound was 3 mm. long and 1 to 5 cm. above the clavicle. The height of the duct in the neck was determined in 3 cases : In t-ases Nos. 7, 9, it 1.-; was 4 cm. above its eutraiice into the vein, and in ca.se No. 16 it was -t cm. above the sternum ; in the remain- der the anatomic relations of the duct were not men- tioned. The discharge of lymph or chyle when it occurred at the time of operation was slight in two cases, Nos. 10, 12, only moderate in five, Nos. 1, 2, 7, 9, 11, and was quite profuse for a greater or less time in four cases, Nos. 4, 5, 15, 16. In two cases, Nos. 1, 16, the pressure seems to have been considerable, in the lattei- the fluid spurting up into tlie wound about 5 to 7 cm. In these cases the fluid appears to have been lymph only in six instances, Nos. 1, 4, 5, 7, 9, 11, while chyle was dis- charged at operation in the remaining Ave cases, Nos. 2, 10, 12, 15, 16 ; Nos. 10. 15, however, were accident cases and not prepared for operation. In case No. 1, although the fluid appeared to have been lymph and the patient had fasted for 12 hours, there was enough lymphatic pressure to cause an appreciable degree of spurting, as already mentioned ; in case No. 16, where the patient had had breakfast 1.} hours before operation, chyle spurted up 5 cm. to 7 cm. In those instances where the discharge did not appear until after operation, it became evident the first dav in case No. 3, the fourth day in Nos. 18, 17, the fifth day in No. 14, and the tenth day in Nos. 6, 8. The nature of the fluid in these cases varied of course with the period of digestion, and was undoubt- edly lymph or chyle in each instance. In the entire series, however, the results of microscopic examination of the fluid was mentioned in but three cases, Nos. 1, 5 and 17, and in the two latter fat was demonstrated in very minute subdivision. In No. 4 the fluid was said to be "lymph," and in cases Nos. 6, 14 it was men- tioned as " chyle," but the character was not given. In the remaining cases the diagnosis seems to have been made on the gross appearance of the fluid, as no micro- scopic or chemic tests were recorded. The treatment of the 11 cases where the injury was recognized at the time ■ of operation was by packing, suture and ligation. Packing was employed in five cases, Nos. 1, 2, 10, 11, 15, and another case. No. 4, was undoubtedly also treated in this way, although the line of treatment was not mentioned. Lembert sutures were employed in four cases, Nos. 5, 7, 9, 16, in the first three fine silk being used, in the last fine catgut. In case No. 9, the head and neck were encased in a plaster-of-paris bandage to insure immobility. Ligature was employed in one case. No. 12, and was unsuccessfully attempted in case No. 11, but the latter was finally packed. The cases that were packed, with one exception. No. 2, con- tinued to discharge from the first dressing, but in No. 1 this discharge was very slight, scarcely staining the 14 dressings, and was of special interest, as tlie innominate and subclavian veins were both ligated. The patient lived for 36 hours, taking nourishment and stimulants. The ilow of lymph, therefore, must have been through a collateral circulation or backward along the internal jugular vein, since it did not come out through the wound. The presence of valves, though imperfect, in the internal jugular vein about 1 inch above the innom- inate, renders the supposition somewhat stronger that the flow of lymph was accommodated by collateral cir- culation. In case No. 10 the leakage continued but a short time, while in No. 11, though slight, it persisted for three weeks. In Nos. 4, 15, the discharge was of large amount and was checked by secondary operation, as will be mentioned later. In the cases in which suture of the duct was employed, subsequent leakage occurred in two, Nos. 5, 7, but in each instance was thought to be due to wounded radicles, rather than to escape from the main duct. In No. 5 this discharge was profuse for five hours and considerable for the next 24 hours, after which it ceased entirely ; in No. 7 the dis- charge continued for 15 days, though slight ; in No. 9 the closure appears to have been perfect, the wound was closed without drainage, and no subsequent leakage occurred ; this was the first case of suture without sub- sequent leakage, and the only case of suture in wliich the wound was closed successfully without drainage. In No. 16 there was no subsequent leakage, but a gauze drain which was introduced into the lower end of the wound was removed on the third day, healing taking place by first intention. The case in which ligation was employed, No. 12, showed no subsequent leakage ; the method of closing the wound was not mentioned. Packing was employed as a method of treatment in 5 cases, Nos. 3, 6, 8, 13, 17, in which the injury was not recognized until after operation. In No. 14 of this same series a graduated pressure pad was applied over the wound, and no mention was made of packing. In No. 3 packing appears to have been immediately successful, as there was no subsequent discharge. In cases Nos. 6, S, 13 the wounds were packed but lightly, and profuse dis- charge continued. In No. 6 packing was abandoned, and subsequent operation performed, as will be men- tioned later. Cases Nos. 8 and 13 were subsequently securely tamponed, and the discharge ceased immedi- ately. In Nos. 14, 17 the discharge continued 14 and 18 days respectively ; in the former case the employment of packing was not mentioned, in the latter secure packing was employed after a few days, but without very appa- rent eifect. An interesting and instructive feature in all the cases in which packing was used is the fact that in those cases where light packing or drainage was eiu- 15 ployed there was subsequent discharge, whereas flrm packing controled the discharge completely with the possible exception of one case, No. 1 7. Subsequent or secondary operation was resorted to in 8 instances, Nos. 4, 6, lo. In case No. 4 the discharge was profuse and the constitutional efiect was marked. The discliarging point was successfully clamped, the clamp remaining on 3 days. In case No. 6, following light packing, there was abundant discharge, which soon caused marked general physical depression. A few days after the original operation, the wound was explored, the discharging vessel readily clamped, the clamp remain- ing on for 3 days. There was immediate improve- ment. In case No. 15, in which the injury was due to an accident, the wound was packed when iirst seen. The following day the wound was opened, the discharging point readily ligated and the wound completely closed. There was no further discharge. Constitutional effect due to lymphatic discharge was evident in 3 cases, Nos. 4, 6, 8. In No. 4 the discharge was estimated at 3 pints daily, and the patient lost flesh rapidly ; the vessel was subsequently clamped, and the patient gained flesh at the rate of a pound a day. In case No. (i the discharge was estimated at IJ to 2 pints daily, and the constitutional effect was marked; the vessel was clamped, and immediate recovery occurred. In No. 8 there was profuse discharge for 8 days, during which time the patient lost 10 pounds ; the wound was securely and successfully packed, and during the next 15 days the patient gained 23 pounds. In case No. 1 the patient died from "shock and exhaustion," and as there was no appreciable lympliatic discharge subsequent to operation, it seems unreasonable to attribute any consti- tutional effect to the injury of the duct. In No. 15 the patient was in a condition of exhaustion due to loss of blood. In No. 17, while there was considerable depres- sion following the operation, the lymphatic discharge was so slight that it was not thought to bear any causal rela- tion to the condition. In the remaining 11 cases no con- stitutional efftect was observed, although in several there was sliglit discharge for a considerable time. When constitutional eff'ect was noted, the rapidity with which it occurred, and the quickness and completeness of the recovery after the discharge had ceased, permitted no question as to the cause. The prognosis of the series was very favorable, only one death occurring. No. 1 , and that bearing no apparent relation to the injury of the duct. Gay, however, in reporting this case, ventures the opinion that the wound of the duct would have precluded the patient's recovery. One author, Schwimm, Case No. 6, speaks of injury of the thoracic duct as of "frightful importance." 10 Gerrish" remarkH that "the prosi)Cc-ts of success are not brilliant," and one writer''' remarks that a wound of the thoracic duct is beyond surgical treatment. All others, liowever, who have ventured to express an opinion, are agreed that wounds of the thoracic duct are serious only to a very moderate degree, and it might be added that the most serious feature lies in the possibility of entirely occluding tlie duct by ligature, suture, or clamps. Recommendations ottered l)y writers for operative treatment are somewhat varied. The method of suture, among the more recent writers, is recommended first where possible, and is referred to by Koenig and Keen and later by Gushing. Clamping the vessel and leaving tlie forceps on for several days, although practised in 2 cases, Xos. 4, 6. was recommended only in one instance; Scliroeder and Plumber remark that clamping is Fen- ger's method of dealing with such injuries. Ligation has been recommended l)y Souchon-", he having great faith in iminediate collateral circulation. Cushing recom- mends, where suture is impossible, the placing about the duct of a provisional ligature which can be secured subsequently if packing is ineffectual and the failing condition of the patient justifies such a radical measure. He also recommends tying all visible lymphatics when working near the duct. Schwimm recommends ligating the lymph-vessels in the pedicle of the lower glands of the neck before thej' are entirely removed. An objection to all ligatures in the neighborhood of the duct is that e recommended, as liga- tion of the duct should not be performed until all other means, including a secondary operation, have been tried without avail. The ligating of lymphatic branches of any considerable size, or blindly ligating pedicles deep in a wound in the neighborhood of the duct does not appear justifiable, unless the duct has been developed by dissec- tion, or unless the operator is sure he is not occluding the main channel. The same objection exists with refer- ence to the use of clamps. The value of packing is un(iuestioned, and mentioned by all. Study of the t'ases would certainly justify the " Deunis' System of Surgcrv. 1S95. vol. 2, p. 4BS. " American Text Book of SiUKt-ry, 1893, p. -US. »' Park'sSurgery, 18'Jli, vol. 2, p. -Tij. 17 belief that accurate flrni packing will control the dis- charge in all cases. In those instances in which the discharge was not controled by packing, it was said to have been lightly applied in several, and no mention was made of firm packing in the others. The objection that by packing, the lymphatic circulation may be entirely shut off, seems unreasonable, as the pressure of the lymphatic circulation would seem sufficient to pre- serve to a certain degree the integrity of the duct, especially as any packing is sure to become slightly loosened in the course of a few hours. One great value of packing, as mentioned by Gushing, is that it increases largely the chance for collateral circulation. The tampon is a safety valve allowing the escape of a certain amount of lymphatic fluid in case the lymph-pressure becomes too high before the new channels of circulation have been fully opened. It is not meant to be said that packing will control the entire discharge at once in every case, but that it will control it very largely at first, will do so completely after a short time, and will sometimes be effectual from the very beginning if properly used. The use by Warren of a graduated pad applied for pressure over the wound, is merely a less accurate method of reaching the same results. The employment by Gushing of a plaster-of-paris bandage, employed so extensively at the Johns Hopkins Hospital, seems an unnecessarily severe procedure. Absolute immobility cannot possibly be obtained, and it seems possible to secure a sufiicient degree of support to the head by large cervical dressings carefully applied, and the use of sandbags in bed. The theoretic explanations offered in those cases in which lymphatic discharge has not appeared until sev- eral days after operation, are not altogether satistiactory. It seems highly probable, in these instances, that only small radicles of the duct were injured, as discharge from a vessel of considerable size would probably have been recognized at the time of operation ; whereas the escape of a small amount of clear fluid could readily be mistaken for serum. In two eases, however, the wound was subsequently investigated, and the exit of escape found to be of considerable size, in one instance being clamped, No. 6, and in another successfully packed, No. 8. It is noticeable also that tlie discharge in these 2 cases was delayed longer than in any others of the series, not appearing until 10 days after operation. The explanation offered by Schwimm and later by Gushing, that the discharge is due to increased lymph-pressure after feeding has been resumed, is not entirely satisfac- tory. The lymph-pressure in all the cases was probably at its height within a few days after operation, while in several instances the leakage did not appear to have commenced until sevei-al days after this period . In the 18 meanwhile the processes of repair must certainly have been going on. In the two instances in which the dis- charge was delayed for 10 days the above view seems justifiable, in spite of the possibility of there having been a very slight leakage during the entire period. The question of subsequent enteric treatment, although of the utmost importance, appears to have received no careful or thoughtful consideration, and this is one point to which we would draw special attention. Gerrish i^ says " the patient is to be kept perfectly quiet, and be given such food as will barely save liim from actual starvation, in the hope that by keeping the duct as nearly as possible empty the accidental opening will close." This is the most important suggestion found,, but it lacks any specific recommendations. Agnew's suggestion, 1^ tliat no food should be given in order to keep the walls of the duct collapsed, does not recommend itself ; a further suggestion in the same article, that milk may be introduced directly into the venous circulation, seems very properly to have never been attempted. Gushing recommends that the patient should receive a "meager diet," but says notliing furtlier. From a physiologic standpoint, tlie indications for feeding in such cases seems very direct, namely, that nourishment sliould consist solely of proteids, with a possible addition of a small amount of carljohydrates, and that all fat should be avoided. That a sufiicient degree of nutrition can be maintained on these materials has been demon- strated. The use of properly prepared beef juice and egg albumin from the white of an egg, meet the require- ments in its most rigid aspect. Centrifugalized milk contains no fat, although it has considerable sugar. The slight absorption of carboliydrates through the lymph, however, does not make it a serious objection. Subse- quently the use of oysters, fish, and raw vegetables may be satisfactorily employed. By such feeding the mate- rial passing through the thoracic duct can be reduced to a minimum, and at the same time the nutrition of the patient satisfactorily maintained. That the amount of lomph passing through the duct is somewhat increased by the ingestion of even albuminous fluids, is undoubt- edly true. The amount of this increase, however, it has not been possible to ascertain from previous observations, and it is our future purpose to investigate this question by animal experiments. One other very practical suggestion arises from a consideration of the physiology of digestion, and some of the experiments on which this suggestion is based were briefly enumerated at the end of the section on physiology. The appearance of lymph during the interval of digestion so closely resembles serum that its presence in a wound, unless in considerable amount. 19 would entirely escape detection. This undoubtedly has been the case in a number of instances where injuries of the duct or its branches were not suspected at the time of operation. The appearance of chyle, however, is so distinctive that its presence in a wound, even in small amount, would be almost sure to attract attention at once. Not only this, but the increased lymphatic pres- sure would surely Open up and bring to notice slight injuries of the duct itself, or injuries to radicles which might very readily contract sufficiently to prevent the escape of lymph during operation, but would be opened up and discharge subsequently on the resumption of feeding. Low lymph-pressure and the absence of chyle are two of the most favorable conditions for overlooking a wound of the lymphatic system. No mention has been made of this point previously, so far as can be ascertained. The physiologic suggestion then, to induce the absorption of fat during operations in the neighbor- hood of the thoracic duct, seems eminently practical. The objection will be at once raised that distended lym- phatics will be injured much more readily than those in a partially collapsed condition. The gross difference between partially filled and distended lymphatics would probably not be very noticeable owing to the small size of the vessels, and it seems probable that most, if not all of the wounds of the lymphatics, while distended, would occur if the vessels were only partially filled with lymph, the difference being that in one instance the wound would probably be apparent at the time, while in the other it might not be realized until after the operation, when high lymph-pressure was increased. The question of lymph-pressure must be met either at operation or soon after, and it seems preferable that this pressure should be exerted at the time when its effect can be observed and corrected immediately, if necessary, rather than later when its effect must be met blindly by inaccurate packing or secondary operation. Another objection which may be raised is that of introducing material into the stomach before giving an anesthetic. It must be remembered, however, that absorption of fat does not occur until the material readies the intestine, after \yhich its objection, from the standpoint of the anes- thetist, does not exist. It would be desirable, however, to produce the effect by a material small in bulk and which will be passed out of the stomach readily and quickly. These requirements would seem to be most satisfactorily met by the use of a small quantity of cream, perhaps from 4 to 6 ounces. From experiments already men- tioned, it is safe to assume that all, or nearly all, of this amount would be passed out of the stomach by tlie end of 3 hours if the patient had received previously an adequate enteric preparation for operation. The solid 20 material of trearn being almost entirely fat, and in such a ijhysical condition as to be most readily absorbed, so far as we linow from physiologic experiments, the absorption of fat \\-ould probably begin within a short time, perhaps an hour, after the cream had entered the stomach. It has been shown that in animals the height of fat digestion is reached after about 7 hours, and that after a full meal it continues in dogs for nearly 30 hours. The previous enteric preparation for operation is men- tioned in only 3 cases in the series. In cases Nos. 1, 5, the patients had fasted for 12 and 18 hours respectively and the discharge at operation was lymph and not chyle. As is customary, hevever, these patients had probably received alimited diet, largely albuminous, for sometime previous. In case Xo. 16 the patient had been fed 11 hours before operation, the pressure in the duct was consider- able, and the fluid was chyle. It is to be regretted that the facts have not been recorded in more of the cases. The period of fat absorption being so prolonged, it seems certain that one would find chyle in the thoracic duct for at least a period of from 2 to 7 hours after the inges- tion of cream on an empty stomach, and that if the material were given 3 hours before operation little if any cream would remain in the stomach as an objection- able feature from an anesthetic standpoint. In conclusion, the following thoughts suggest them- selves : First. — The increasing frequency of extensive dissec- tions in the neck makes it desirable to consider means of avoiding injury to the thoracic duct. Second. — It is desirable that if wounds of the thoracic duct or its branches occur, they should be recognized at the time of operation. The result in Case A makas evi- dent the advantage of such early recognition. As this case was operated upon after a meal, it suggests the ingestion of 4 to 6 ounces of cream 3 hours before opera- tion. This might be especially desirable in secondary operations undertaken for the purpose of locating the point of injury. The desirability of such a procedure may be further studied by observations made upon animals. Third. — That suture of the duct with fine silk or cat- gut be accomplished where possible ; that all small dis- charging lymph radicles be ligated ; that the ligatingand clamping of lymphatic vessels of considerable size be avoided, unless the integrity of the thoracic duct itself has been demonstrated ; that where suture of the duct or large radicles is impossil)k', gauze packing, firmly and accurately applied, be used ; that the head and neck be kept at rest, the use of morphin to a considerable degree being recommended if necessary-. Fourth. — That until repair of the duct is thought to 21 be complete, nutrition should be sustaiiiwl on albumin- ous material, with possibly a small amount of carbo- hydrates, but with an absolute exclusion of fats. REFERENCES. Cases. Allen, LI. P., and Briggs, C. E. FiTM-nt article. Operator, D. P. Allen. Oa.ses Nos. l(i, 17. Boegeliold, E. Arehlv fur kliuische Chirurgie, 1.SS3, B. 2'.), .S. W3. Ueber die Verletzungen des Ductus thoraciciis. Operator, ^^'ilms. Case No. 2. Cushing, H. W. Annal.s of Surgery, 1898, \'ol. 27, p. 719, Operative Wounds of the Thoracic Duct. Report of a Case with Suture of the Duct. Operators, C. A. Porter, \\ . S. Halsted, H. W. Cu.sliing. Cases Nos. 7, 8, 9. Gay, G. W. Boston Medical and .Surgical .Journal, 187o, Vol. 92, p. 422. Surgical Operations at the Boston City HospitiU. Operatoi", D. W. Cheever. Case No. 1. Keen, W. W. The New York Medl(.-a,l .lournai, 1891, Vol. .59, p. 569. Operation Wounds of the Thoracic Duct in the Neck, with a Resumfi of Two Prior Recorded Cases and Two Additional Cases. Operators, A. M. Phelps and W . \\ . Keen. Cases Nos. 4, b. Lund, P. B. Boston Medical and Surgical Journal, 1899, \'ol. 140, p. 354. A Case of ' ipcrative Injury of the Thoracic Duct. Opcratoi's, F. B. liund and .T. C. Warren. Cases Nos. i:^, H. Lyne, W. H. Virginia Medical Semi-Monthly, 1898, Vol. S, p. 278. Stab Wound of the Thoracic Dud, Operator, W. H. J^yne. Case No. 10. Jliinler. X. Deutsche niedlcinische Woclicnsclirift, 1899, No. 48, p. 799. Schnittverlezung des Ductus thoracicus. Operator, A. Miinter. Schroeder, W. E., and Plumber, S. C. Annals of Surgery, 1898, Vol. 28, p. 229. Report of Two Cases of Injury of the Thoracic Duet in Oper- ations on the Neck. Operators, W. E. Schroeder and S. O. Plumber. Cases Nos. II, 12. Schwirara, J. Annals of surger.y. 1896, Vol. 23, p. 582. A Case of Operative Injury of the Thoracic Duct at the Root of the Neck. Operator, -J. Schwinim. Case No. 6. Vagedes, T. Inaugural Dissertation, Wurzburg, 18S5. Ueber Ver- letzungen des Ductus Thoracicus. Operator, Maas. (.'aso No. 3. Anatuiny. Albin. Gesch. u. Beschr. des Sangadersystems. .Mascagnls. B. ;3, S.42. Brinton, .1. H. New Vork .Medical .Journal. 1891, Vol. 59, p. 568. The Surgical Relations of the Thoracic Duct in the Neck, with E.vhibitions of two Dissections made by Mr. Ward Brinton. Gray, H. Anatomy, Descriptive and Surgical, 1.SH3, p. 682. The Thoracic Duct. V. Haller, A. Elementa Physiologic, T. 7, p 222. Henlc. Anatomy, l.S7(i, Vol. 3, p. 45). Hyrtl. Lehrbuch der Anatomic. 12 Aufl., S. 9.3(i. Klein, E. Elements of Histology, 1892, p. 90. The LymiJhatic ^'essels. Morris, H. Human Anatomy, 1893, p. 683. The Thoracic Duct. V. Patruban. Seltene Anatomische Beubachtungen. Oesterreich Jahr. 6, Juli, 1844. Picrsol, G. A. Normal Histology, 1898, p. 117. The Lymphatic System. Schilfcr, E. A. t^uain's .\.nat4.>my, Vol. 1, pt. 2, p. 378. Lymphatic System, Stbhr, P. Lehrbuch der Histologic, 1S92, p. S2. Lympligefussjs- tem. Svitzer. Mullei's Archiv, 1845. Thane, (i. D. Quain's Anatomy, 1892, Vol. 2, pt. 2. p. .')17. The Thoracic Duct. Watson, .lournal of Anatomy, 1872, \'ol. 10, p. 427. Wutzcr. Mullcr's Archiv, Heft 4, 1834. jPhy.notoffy. Beaumont, W. The Physiology of Digestion, 1847, p. I'j'i. Colin. Traitfi de Physiologic com parge des Anlmau.\. Paris, 1823. T. II Cooper, A. Medical Records and Hesearches. London, 1798. 22 Three Instances of Obstruction of the Thoracic Duct, with Some Experiments Showing the Effects of Tymg that Vessel. Dittrich, Gerlach, Herz. Prager Vierteljahrsschrift, Februaiy 3, 3851. Flandrin. Handworterbuch der Physiologie, Nasee, B. 1, S. 246. Poster, M. Textbool!; of Physiology, 1891, pt 2, p. 496. The Nature and Movement of Lymph (including Chyle). Harley, V. Journal of Physiology, 1895, Vol. 18, p. 1. The Nor- mal Absorption of Fat, and the Effect of Extirpation of the Pancreas on it. Henle and Pfeuffer. Zeitscbrift, 18.52, B. 2. Leuret and Lassaigne. Rechorches Physiologiques et Chimiques, pour servir a I'Histoire de la Digestion. I'aris, 1825, 8, p. 178. Magendie and Dupuytren. Journal de Physiologie, T. 1, p. 21. Bchafer, B. A. Textboolr of Physiology, Vol. 1, p. 463. Bchmidt>Mulheim. Archiv filr Physiologie, von du Bois-Reymond, 1877, S. 553. Weiss, M. Virchow's Arehiv, B. xxii, S. 526. Welch, W. H. Transactions of the Association of American Physi- cians. 1889, Vol. 4, p. 56. Whitla, W. British Medical Journal, 188S, Vol. i, p. 1089, " Chylous Ascites." Miacellaneous. American Textbook of Surgery, 1893, p. 448. Rupture of the Thoracic Duct. Dennis, P. S. System of Surgery, Dennis, 1895, Vol. 3, p. 220. Injuries of the Thoracic Duct. Dennis, F. S. Surgery by American Authors, Paris, 1896, Vol. 2, p. 270. Injuries of the Thoracic Duct. Gerrish, F. H. System of Surgery, Dennis, 1895, Vol. 2, p. 468. Wounds of Lymphatic Vessels Hamilton, J . B International Textbook of Surgery, Warren-Gould, 1900, Vol. 1, p. 918. Rupture of the Thoracic Duct. Klrschner. Archiv fiir klinische Chirurgie, 1885, B. 32, S. 156. Ein Fall von recht.seitigen Chylothorax in B^olge von Ruptur des Ductus thoracicus nebst Statistik und Kritik der bisher bekannten einschiag- Igen Falle. Koenig, F. Lehrbueh der Speciellen Chirurgie, 1898, B. 1, S. 553 Verletzung des Ductus thoracicus. Murray, J. International Textbook of Surgery. Warren-Gould, 1900, Vol. 2, p. 204. Injury of the Thoracic Duct. Souchon, E. Surgery by American Authors, Park, 1896, Vol. 2, p. 265. Wounds of the Lymphatic Duct. LAPAROTOMY FOE PERFORATION IN TYPHOID FEVER. ' REPORT OF SIX CASES. By C. E. Briggs, A.M., M.D., OF CLEVELAND, OHIO. {From the Surgical Clinic of Dr. Dudley P. Alhn, Lakeside So.spitaL} The report is based upon the six cases of laparotomy for intestinal perforation during the course of typhoid fever, occurring in Lakeside Hospital since its opening in January, 1898. These were performed in 1899 and 1900 during the -writer's service as resident surgeon, and, with the exception of the last case, came under his immediate observa- tion and care. The inclusion of this last case in the series is perhaps of doubtful propriety, since the patient entered the hospital as a case of general peritonitis and was treated as such, the case containing very little of interest with reference to the surgical treatment of typhoid per- foration. For the purpose of completeness, however, it has been added. It is due to the courtesy of the visiting staffs of Lakeside Hospital that the present report has been made possible, and to these gentlemen the writer wishes to acknowledge his obligation. Case I. — Male, aged forty-seven years. Moderately severe case. At entrance, leucocytes 5700 per cubic millimetre. Widal reaction negative, diazo positive. Bowels rather loose, no hemorrhage. Symp- toms of perforation on the twenty-first day of the disease ; abdominal pain, sensitiveness, rigidity, and moderate distention. Else in pulse and temperature ; slight vomiting ; leucocytes 4460 per cubic milli- metre. Condition unfavorable. Operation. Five hours after appearance of symptoms. Chloroform. Time, thirty minutes. Right lateral incision. Dark-brown abdominal fluid. Perforation of ileum 8 centimetres from caecum ; closed with continuous Lembert catgut sutures. No threatening perforation found. Irrigation with sterile water ; intestines not removed. Gauze drainage. Cultures from abdominal fluid, bacillus coli communis. Death seventy- five hours after operation, apparently from exhaustion. Autopsy, restricted. Local peritonitis about lesion and gauze pack- ing, catgut sutures absorbed, and perforation patent. Four other neighboring perforations opening into the same area of local peritonitis. Confluent ulcers of caecum, not perforating. No general peritonitis. Cultures from local peritonitis, bacillus coli communis ; from general peritoneal cavity, negative. C. K., medical No. 769 ; surgical. No. 909. Service of Dr. H. H. Powell. Male, aged forty-seven years, white, German, married, car- penter. Admitted August 5, 1899 ; discharged August 15th. Dead. 1 Reprinted from The American Journal of the Medical Sciences, January, 1902. i BKIGGS: PERFORATION IN TYPHOID FJEVEK. Family History. Negative. Previous History. Had always been well, strong, and active. Twelve years ago he had malarial fever ; six years ago he had what was taken to be typhoid fever, being ill for nine weeks ; no other illnesse.^ vpcalled ; habits good ; no venereal history. Pre.ient Illne-ss. Patient eQtered hospital conij)lainiDg of headache, lo?s of appetite, diarrhoea, and fever. Headache was the first symptom, beginning two weeks previous to entrance, and persisting. There was also anorexia and diarrhoea, the latter being quite troublesome. There was some abdominal pain in the region of the umbilicus. Physical Examination. Patient well-nourished and developed ; mind clear ; face flushed ; tongue coated. Pulse 80, good quality. Tem- perature 101.8.° Lungs negative. Heart not enlarged; soft systolic murmur at apex. Spleen enlarged on percussion, not palpable. Abdo- men soft, slightly sensitive about umbilicus ; numerous rose-spots. Leucocytes 5700 per cubic millimetre. Urine, no albumin ; faint diazo reaction. Widal reaction negative. During the next six days the symptoms were those of a moderately severe case of typhoid, the temperature running 101° to 103°, during which time he had six tub baths, four of which were given in the last two days. The bowels remained rather loose. The urine continued to show a diazo reaction, and there was a slight trace of albumin. No subsequent Widal tests were made. On August 12th, the twenty-first day of the disease, the patient began to have abdominal pain, coming on rather gradually and being first noticed at 6.30 a.m. At first it was confined to the lower half of the abdomen, gradually spreading upward and becoming general. Pain was also noticed at end of penis. He vomited a slight amount of curdled milk at 8 a.m. The temperature at 6 a.m. was 101°, and at 9 A.M. had risen to 102.2°, and the pulse had risen from 72 to 104 during the same period. The leucocytes at 9 a.m. were 4460 per cubic milli- metre, and this unfortunately was the only count made. At 9.30 a.m. the patient was seen with Dr. W H. Nevison. The abdomen was moderately distended and tympanitic. There was marked sensitiveness over the entire lower half of the abdomen, and the abdominal wall was generally quite rigid. Respirations 28, largely costal. Pulse was 124, dicrotic, readily compressible. The lungs were clear, the heart as previously described. The general condition of the patient was poor, although not altogether unfavorable for operation. Operation, August 12th, 10.30 a.m., by Dr. W. H. Nevison, five hours after first appearance of symptoms. Anaesthetic, chloroform, preceded by morph. sulph., gr. one-quarter, hypodermicaljy. Time thirty minutes. Light anaesthesia. Incision eight centimetres long in right linea semi- lunaris opposite anterior superior spine. Considerable dark-brown fluid escaped, fecal odor, containing small particles of extraneous matter thought to be feces. Caecum readily found and ileum carefully inspected for about one hundred and twenty centimetres. A perforation was found about eight centimetres from the cascum. It was round, about three millimetres in diameter, on the free side of the bowel, and from it fecal matter was escaping. The area of induration surrounding this opening was about two centimetres in diameter. No other perforations were found in the explored ileum or ciecum, nor were any perceptibly BKIGGS: PERFORATION IN TYPHOID FEVER. 6 thinned areas appreciable, although careful search was made. The perforation was turned in without excision by a double row of con- tiuuous Lembert catgut sutures. The ileum was dark and congested, the ctecum somewhat so. Other coils of intestine seen were slightly congested. Appendix normal. Parietal peritoneum seemed but slightly congested. Presence of enlarged glands not observed. No fibrinous flakes seen. There were no adhesions. Abdominal cavity thoroughly irrigated with a large quantity of sterile water, special attention being paid to the pelvis and the lumbar fossse. The intestines were not handled, but the irrigation was thoroughly accomplished by means of a hard-rubber flushing tube. Site of suture left near the surface, and to it iodoform gauze drainage was placed, brought out through the lower end of the wound. Above this the wound was closed with inter- rupted silkworm-gut sutures through all layers. During the operation the patient received strych. sulph., gr. one-fifteenth, hypodermically, and an infusion of seven hundred cubic centimetres of normal salt solu- tion. The radial pulse was very rapid and weak, imperceptible at times. The operation was associated with considerable shock, the patient being in a poor condition on leaving the table. The subsequent treatment of the case was by strychnine, nutrient stimulating enemata, liquids without milk by mouth, subcutaneous infusions of salt solution, sponge baths, and oxygen. Patient took nourishment well during entire time. He was slightly delirious and semi-conscious at times, but for the most part his mind was clear and active. There was slight abdominal distention, but practically no pain except on the day following the operation, when it was quite marked in the lower half of the abdomen for a short time. The discharge from the abdomen was very profuse during the last twenty-four hours, and was dark colored, which, together with the pain mentioned above, suggested strongly a second perforation in the locality of the gauze drainage. The temperature remained about 102.5° to 104° rectal, and the pulse 120 to 130. On August 15th, three days after the operation, the patient began to show evident signs of dissolution, the pulse-rate increased steadily, and the quality became progressively worse. There was some general abdominal distention, most marked in the epigastrium. There was an accumulation of coarse, moist r&les throughout the chest, and the respirations were labored. Oxygen was given during the last seven hours. He was unconscious for six hours before death. He died August 15th, at 3.05 p.m., seventy-five hours after operation. The bacteriological examination of the abdominal fluid at the time of the operation showed on cover-slip preparations a medium-sized bacillus and a small coccus, singly and in pairs. Plate cultures showed a pure growth of bacillus coli communis. No growth of the coccus mentioned could be found. Autopsy. This unfortunately was restricted. About the gauze pack- ing and the site of the lesion in the ileum there was a localized peri- tonitis, as was to be expected. Opening into this localized area was the original ulcer. The catgut suture had been almost entirely absorbed. There were four small neighboring perforations of the ileum, which also opened into this same localized area. There were several confluent ulcers of the csecum which had not perforated. There was no general peritonitis. Cultures taken from the area of local peritonitis showed 4 BRIGGS: PERFORATION IN TYPHOID FEVER. bacillus coli communis. Cultures and cover-slip preparations from general peritoneal cavity were negative. Case II. — Male, aged seventeen years ; severe case. At entrance, leucocytes 5980 per cubic millimetre ; Widal reaction negative ; diazo reaction negative. Marked constipation ; slight distention. Hemor- rhage on seventeenth and eighteenth days of disease, preceded by con- siderable abdominal pain. Symptoms of perforation on eighteenth day of disease ; abdominal pain, sensitiveness, rigidity, moderate distention, rise in pulse, temperature unaffected, slight vomiting. Condition very unfavorable. Operation. Eight hours after appearance of symptoms. Chloroform . Time, twenty-five minutes. Right lateral incision. Dark-brown abdominal fluid with curds. Perforation of ileum twenty-five centi- metres from cajcum ; one threatening perforation close by ; both secured with continuous Lembert silk sutures. Irrigation with salt solu- tion and sterile water ; intestines not removed. Gauze drainage. Cul- tures from abdominal fluid, bacillus coli communis. Death sixteen hours after operation ; general peritonitis. Autopsy, restricted. Intestinal suture intact. No other perforations found. General peritonitis. Cultures from general peritoneal cavity, bacillus coli communis. J. B., medical, No. 779 ; surgical. No. 925. Service of Dr. J. H. Lowman. Male, aged seventeen years, white, American, single, student. Admitted August 13, 1899 ; discharged August 22d. Dead. Family History. One brother died of typhoid fever ; one sister ill with typhoid fever at present time. Previous History. Children's diseases. Always strong and active ; habits good ; no venereal history. Present Illness. Patient entered hospital complaining of headache, constipation, and general malaise. Headache was the first symptom, and was first noted ten days previously. Bowels constipated, no move- ment for forty-eight hours before entrance. No hemorrhages. Chill}' sensation during last two nights ; no distinct rigor. Epistaxis on day of entrance. Physical Examination. Patient well nourished and developed ; athletic. Face flushed ; tongue slightly coated, trembles when protruded. Pulse 116, low tension, dicrotic. Temperature 102.8°. Lungs nega- tive. Heart slightly enlarged to the left, loud systolic murmur at apex. Spleen slightly enlarged on percussion, not palpable. Abdomen rather tense, not tympanitic, no especial sensitiveness, no rose-spots recorded. Leucocytes 5980 per cubic millimetre. Urine, no albumin, no diazo reaction. Widal reaction negative. During the next four days the symptoms were those of a severe typhoid. The nervous sj^mptoms were marked. There was a little abdominal pain on two occasions following tub baths. Baths were given every three hours with but few intermissions. Reactions following baths were not veiy good. The pulse ranged from 90 to 120, and was of rather low tension. On one day patient complained of considerable pain on inside of right leg, extending down to the knee. On August 18th, at 8 p.m., following tub bath, there was sudden sharp pain in lower half of abdomen, which continued for some little time, but with rather less severity. The temperature was unaffected, but the pulse-rate was increased somewhat. Sponge baths were sub- BRIQGS: PERFORATION IN TYPHOID FEVER. 5 stituted. No blood examination was made. During the next day there was slight abdominal pain with an occasional increase in severity. The abdomen was somewhat distended, but this passed off after several free bowel movements. Following this the pain practically disap- peared, and the patient was comfortable. There was no vomiting. Pain was present along inner side of right thigh. Leucocytes were 8000 per cubic millimetre in the morning, and 10,000 per cubic milli- metre in the afternoon. The next day, August 20th, the abdominal pain had practically disappeared, but there was slight sensitiveness above the pubes and in the left inguinal region. Tub baths were resumed, and were borne rather better than previously. A small amount of old clotted blood was passed in one of the stools. August 21st, the eighteenth day of the disease, the patient had been rather more comfortable than usual. At 2 a.m., while in tub bath, there was sudden severe pain in the lower abdomen. The abdomen was tense, not distended, and quite generally sensitive. The tempera- ture was unaffected. The pulse rose from 80 to 110, and was not well sustained. At 2.45 a.m. the patient passed about two drachms of blood in a bowel movement. The patient vomited curdled milk at 3 a.m. Abdominal pain continued ; sensitiveness was quite marked, more so in the right lower quadrant. Sponge baths were substituted. No blood examination was recorded. At 9 a.m. the patient was seen with Dr. W. H. Nevison. His general appearance was that of a very sick and much reduced man. The finger nails and lips were slightly cyanotic. Pulse was 120, regular, small volume, low tension. Kespira- tions were 24, thoracic. The abdomen showed moderate general dis- tention and tympanites. There was general sensitiveness, slightly more marked in the lower half, and marked muscular resistance. The con- dition of the patient seemed quite unfavorable for operation. Operation, August 21st, 9.45 a.m., by Dr. W. H. Nevison, eight hours after perforation was thought to have occurred. Ansesthetic, chloroform, preceded by morph. sulph., gr. one-quarter, hypodermically. Time, twenty-five minutes. Light anaesthesia. Incision in right linea semilunaris nine centi- metres long, at level of anterior superior spine. A moderate amount of dark fluid containing particles of curdled milk escaped. Caecum readily found, and ileum searched for about one metre from csecum. About twenty-five centimetres from caecum a perforation was found on the free margin of the gut, three millimetres in diameter; base gen- erally thickened. About six centimetres from this thickened area a very thin centre was made out. Nothing more approaching perforation was found. Perforation was turned in without excision by two rows of running fine silk Lembert sutures for a distance of about two and five-tenths centimetres. The thin ulcer was also turned in with a single running silk suture. About the perforation was a little gray fibrinous deposit. The ileum had lost its gloss in places, but in general looked well. No fibrinous exudate outside the region of perforation. The caecum and appendix normal. Remaining coils of intestines and parietal peritoneum appeared to be in good condition. No adhesions were found. Abdomen flushed with hot salt solution and sterile water in large quantities, special attention being paid to the dependent por- tions. This was done by means of a hard-rubber flushing tube. The intestines were not removed from the abdominal cavitv. The site of 6 BEIGQS: PERFORATION IN TYPHOID FEVER. suture left near wound, and to it was placed iodoform gauze drainage, brought out through lower end of wound, above which the incision was closed with interrupted silkworm-gut sutures through all layers. Dur- ing operation patient's pulse became very feeble, respiration shallow, color poor. He was given strych. sulph., gr. one-fifteenth, hypoder- mioally, divided, and a saline infusion of eight hundred cubic centi- metres, under which the condition improved somewhat. The subsequent treatment was by strychnine, nutrient stimulating enemata, subcutaneous infusions, and sponge baths. _ The reaction to the sponging was very marked, and the liaths were given with consid- erable care. The patient's condition improved slightly for a few hours following the operation, after which time he failed steadily, and died August 22d, at 2 a.m., sixteen hours after the operation. Bacteriological examination of the free abdominal fluid at operation showed a pure growth of bacillus coli communis on plate cultures. No cover-slip preparations were made. Autopsy. This was very restricted. The intestinal suture was intact. No further perforation was found. There was a general peritonitis with thin turbid fluid. Cultures from the general peritoneal cavity showed bacillus coli communis ; no cover-slip preparations were made. Case III. — Male, aged thirty-one years ; mild case with a moderately severe relapse. At entrance leucocyfes, 5700 per cubic millimetre ; Widal reaction positive ; diazo reaction negative. Bowels rather con- stipated ; no hemorrhages. Symptoms of perforation on forty-third day of disease, twelfth day of relapse ; severe abdominal pain, general sensitiveness, moderate distention, marked muscular rigidity on right side ; rise in pulse, temperature unaffected ; slight vomiting ; leucocytes, 31,600 per cubic millimetre. Condition of shock, but fairly favorable for operation. Operation. One hour after symptoms. Cocaine solution, 1 per cent. Time, twenty-five minutes. Right lateral incision. Little brownish abdominal fluid. Perforation of ileum forty-five centimetres from caecum, closed with double continuous Lembert silk suture ; no threat- ening perforations found. Irrigation with sterile water, intestines not removed. Gauze and glass-tube drainage. Cultures from abdominal fluid showed staphylococcus pyogenes aureus. Death fifty-eight hours after operation ; typhoidal toxaemia. ^Miops!/, restricted. Sutures intact ; no other perforation ; no general peritonitis ; no peritoneal adhesions except around drainage. Cultures from general peritoneal cavity, negative. T. D., medical No. 992 ; surgical No. 1304]. Service of Dr. E. F. Cushing. Male, aged thirty-one years, white, Welsh, married, coach- man. Admitted January 25, 1900 ; discharged February 28th. Dead. Family Hldory. Negative. Previoia History. Always strong and active, no definite diseases recalled ; habits good ; no venereal history. Present Illness. Patient entered hospital complaining of headache, weakness, and lack of energy. Symptoms began ten days before entrance, with malaise. Following this there was severe headache, anorexia, slight cough, and marked weakness. He also had several attacks of abdominal pain described as " colic." The bowels were loose, but there had been no hemorrhages. Patient had vomited several times. There was no epistaxis. BKIGGS : PERFORATION IN TYPHOID FEVER. / Physical Examination. Well nourished and developed. Mind clear ; tongue coated. Pulse 100, full, not dicrotic. Temperature 102°. Lungs and heart negative. Spleen enlarged on percussion, not pal- pable. Abdomen slightly distended and slightly sensitive on pressure ; covered with scattering rose-spots. Leucocytes 6700 per cubic milli- metre ; haemoglobin, 90 per cent. Urine, no albumin, no diazo reac- tion. Widal reaction positive. For the next three weeks the patient ran a course of mild typhoid fever without complications ; the temperature rarely rose above 102.5°, and for a large portion of the time was below 102^. The patient suffered no abdominal pain, and there were no intestinal hemorrhages. On February 14th, the thirty-first day of the disease, the tempera- ture began to rise, and the patient suffered a well-marked relapse. During the next twelve days the symptoms were somewhat more aggravated than during the primary attack ; the temperature ran somewhat higher, though at no time was extreme ; on several occasions there was nausea and a limited degree of vomiting. There was no abdominal pain, the abdomen was soft, not distended. Urine con- tained a faint trace of albumin ; diazo reaction positive. Patient was somewhat constipated ; there were no intestinal hemorrhages. On February 26th, the forty-third day of the disease and the twelfth day of the relapse, the patient was given an enema at 9 a.m., which was but slightly effectual. Following this he complained of feeling a little distended, and there was slight abdominal pain. This pain, how- ever, soon passed off, and at 10 a.m. the patient was found sleeping quietly ; pulse 100. When seen fifteen minutes later he was in a con- dition of evident shock ; he was cyanotic and trembling ; the pulse had reached 120, and was very compressible ; temperature unaffected. There was quite severe general abdominal pain, the abdomen was some- what distended, but not very sensitive. The liver appeared to be pushed upward. The patient vomited once, about an ounce of greenish fluid. Leucocytes 31,600 per cubic millimetre. The patient was seen with Dr. D. P. Allen at 10.30 a.m. Cyanosis was still present. Pulse was 130, low tension. Respirations were 30, thoracic. The abdomen was moderately distended ; there was general, not localized, sensitive- ness, and there was marked muscular rigidity on the left side ; peris- talsis was present. The heart sounds were weak, the lungs clear. Patient's condition was fairly favorable for operation. Operation, February 26th, 11 a.m., by Dr. D. P. Allen, one hour after appearance of symptoms. Local anaesthesia, cocaine solution, 1 per cent., preceded by morph. sulph., gr. one-quarter, hypodermically. Time, twenty-five minutes. Incision six centimetres long in right linea semilunaris opposite anterior superior spine, carried into peritoneal cavity without causing the patient pain. Abdominal cavity contained a little brownish fluid. Caecum readily found, normal. Appendix normal. Perforation of ileum about forty-five centimetres from caecum, on free border, about two millimetres in diameter. Perforation closed without excision by two rows of continuous silk Lembert sutures. No other perforations or threatening perforations found, but the ileum was not examined beyond the lesion. Abdomen irrigated with large quantities of hot sterile water with flush tube, intestines not being removed ; special attention was given to the lumbar fossae and pelvic cavity. Ileum con- 8 BRIGGS: PERFORATION IN TYPHOID FEVER. siderably congested. Peritoneal surfaces otherwise apparently unaffected. No fibrinous flakes. No adhesions seen. Glass-tube and gauze drain- age placed to seat of suture, on either side of which the wound_ was closed with interrupted silkworm-gut sutures. The patient's condition did not appear to be affected by the operation : the pulse was as good as before operation. Very little pain was experienced except during irrigation, when it was moderate. Subsequent treatment of the case was by strychnine, nutrient and stimulating enemata, subcutaneous saline infusions, sponge baths, and liquids by mouth. The patient appeared to be unaffected by the opera- tion. During the next two days the course was that of an increasingly severe typhoidal toxic condition with the accompanying evidence of exhaustion during the second day. There was failing pulse, marked trembling, moderate delirium at times. The abdomen was only slightly distended, was generally tympanitic, and was but moderately sensitive. At times there was sharp abdominal pain. There was but little dis- charge from the wound. Urine showed a large trace of albumin, and contained an abundance of hyaline and granular casts. The patient died at 9.10 p.m., February 28th, fifty-eight hours after operation. Bacteriological examination of the free abdominal fluid at the time of operation showed on cover-slips a few cocci and numerous bacilli. Tube cultures showed a pure growth of staphylococcus pyogenes aureus. Autopsy, restricted. The intestinal sutures were intact, the perfora- tion being firmly closed. No additional perforations found. There was no general peritonitis. There were no peritoneal adhesions except those about the gauze packing. Cultures and cover-slip preparations from general peritoneal cavity were negative. Case IV. — Female, aged sixteen years ; very severe case. At entrance leucocytes 9932 per cubic millimetre ; Widal reaction posi- tive ; diazo reaction positive. Bowels moved readily with enema, no hemorrhages. Marked delirium. Symptoms of perforation on the fifth day of the disease ; abdominal pain, rigidity, distention ; rise in pulse and temperature ; no vomiting ; leucocytes 15,080 per cubic millimetre ; general constitutional change marked. Exploration advised. Operation. About four hours after first appearance of symptoms. Cocaine solution \ per cent. Time, twenty minutes. Short right lateral incision. No perforation ; peritoneal cavity apparently normal. Wound sealed with collodion dressing. Cover-slips and cultures from abdomen negative. JMedical treatment uninterrupted by operation. Subsequent course of fever severe ; convalescence slow. H. B., medical No. 1035. Service of Dr. H. S. Upson. Female, aged sixteen years, colored, American, single, cleaner. Admitted January 4, 1900 ; discharged March 24th. Cured. Family History. Negative. Previous History. Children's disease ; no other illness recalled ; habits good. Present Illness. Patient was transferred from gynecological service complaining of general weakness. Two days before patient complained of genera] malaise and ached all over. There had been a little cough, occasional epistaxis, anorexia, and vomiting with moderate frequency. Bowels regular. Physical Examination. Patient well nourished and developed ; mind clear ; tongue coated. Pulse 100, fair strength and volume, slightly BEIGGS: PERFORATION IN TYPHOID FEVER. 9 dicrotic. Temperature, 104.2°. Lungs and heart negative. No en- largemeat of spleen perceptible. Abdomen soft, no rose-spots. Leuco- cytes, 9932 per cubic millimetre ; haemoglobin, 85 per cent. Urine, albumin faint trace ; diazo reaction present. Widal reaction positive. From the start the symptoms were those of a very severe typhoidal infection. The temperature was high, averaging 10-4° to 105°. The reaction from the tub baths was not satisfactory, and the drops in tem- perature very moderate. The pulse for the first three days averaged 110 to 120, but was of fair quality. There were marked nervous symp- toms, the patient being delirious most of the time and rarely giving an intelligent answer to questions. The amount of stimulation required was considerable. On January 6th the patient complained somewhat of abdominal pain, and there was slight distention but no rigidity. On January 7th, the fifth day of the disease, the patient's condition began to get markedly worse in the early part of the evening. The patient was delirious, so that her subjective sensations were of little avail in diag- nosis. In spite of the delirium, however, she complained of abdominal pain, which it seemed must be at least moderately severe to arrest her attention at all. The abdomen was moderately distended, and was everywhere tympanitic. There was a considerable degree of muscular rigidity. No definite sensitiveness could be localized. The pulse had risen from 120 to 140, was weak and dicrotic. The temperature had risen from 102.8° at 5 p.m. to 105.5° at 8 p.m. The leucocytes at 8 P.M. were 15,080. The patient was seen about 8 p.m. with Dr. G. W. Moorehouse, and essentially the same condition found as noted above. A definite diagnosis of perforation could not be made. The constitu- tional effect, however, from some cause or other had been very marked, and owing to the delirious condition of the patient and apparent insen- sibility to ordinary degrees of pain it was felt that a perforation under the circumstances would be very easily overlooked, and that the physi- cal condition was such as one might readily expect in a patient so severely ill if a perforation had actually occurred. On account, there- fore, of the possibility of overlooking a perforation, together with the degree of abdominal symptoms afforded by the patient in even such a delirious condition, an exploration under local ansesthesia was advised. The risk of such an operation was very slight, whereas the danger from an unrecognized perforation was of the gravest possible character. Operation, January 7th, 9.45 p.m., by Dr. C. E. Briggs, about four hours after the appearance of symptoms. Local ansesthesia, cocaine solution, i per cent., preceded by morph. sulph., gr. one-quarter, hypodermically. Time, twenty minutes. lucision three centimetres long in right linea semilunaris opposite anterior superior spine. On opening the peritoneum no free fluid was found. There was no injection of the csecum or adjoining portions of the ileum. No adhesions were found, and the peritoneal cavity appeared to be perfectly normal. Cover-slip preparations from right iliac fossa and pelvis showed no organisms. Wound closed with buried silk sutures in layers. A collodion dressing was applied, consisting of several alternate thin layers of sheet wadding and collodion, making a firm and absolutely impervious dressing. Condition of patient apparently unaffected by operation. Permission given to continue baths within four hours if desirable. Patient remained on the medical service. 10 BRIGGS: PERFORATION IN TYPHOID FEVBK. The subsequent history of the case is of little interest from the present stand-point, except for one attack of abdominal pain. She ran the course of a very severe typhoidal infection, the toxic symptoms being^ so severe on several occasions that her recovery seemed hopeless. ^ The temperature remained persistently high for nearly ten days, and was but slightly affected by tub baths. On only one occasion, January 20th, the eighteenth day of the disease, was there any subsequent sug- gestion of an abdominal complication. On this day she was awakened at 2 A.M. by severe abdominal pain. The pulse was but slightly affected. The temperature was unchanged. The abdomen was soft and not sen- sitive, and there was very slight sensitiveness on deep pressure in right lower quadrant. The pain could not be definitely localized, although the patient was perfectly rational, and it passed off within a short time. The leucocytes, however, at 2.30 a.m. had risen to 60,660 per cubic millimetre ; two days before the leucocytes were 8060, and five days later 9600 per cubic millimetre. The convalescence was slow and uneventful. Following the operation, baths were given as required without reference to the incision. The incision healed by first inten- tion. The patient was discharged March 24th, cured. Bacteriologieal Examination. Cultures taken from the right iliac fossa and pelvis showed no organisms in cover-slip preparations, and there was no growth on cultures. Case V. — Male, aged seventeen years ; mild case. At entrance leucocytes 4900 per cubic millimetre ; Widal reaction positive ; diazo reaction positive. Bowels rather constipated, no hemorrhages. Symp- toms of perforation on twentieth day of disease ; severe abdominal pain, general sensitiveness, slight distention, marked muscular rigidity in lower half of right side ; rise in pulse ; fall in temperature of 1° ; no vomiting; leucocytes 9200, 13,000, 13,200 per cubic millimetre. Con- dition of patient favorable for operation. Operation. Three hours after first appearance of symptoms. Cocaine solution, 1 per cent. ; chloroform. Time, fifty-five minutes — forty minutes under chloroform. Right lateral incision. Yellowish-gray abdominal fluid. Perforation eleven centimetres from cfecum, closed with continuous Lambert silk sutures ; no threatening perforations found. Irrigation with sterile water. Intestines not removed. Glass- tube drainage from pelvis. Cultures showed bacillus mucosus capsulatus. Recovery. C. W., medical No. 1208 ; surgical No. 1530. Service of Dr. E. F. Cushing. Male, aged seventeen years, white, American, single, laborer. Admitted June 16, 1900; discharged August 19th. Cured. Family Hvdory. Negative. Previous History. Has always been well and strong. A few years ago had erysipelas, and later a fever the nature of which he did not know. Habits good ; no venereal history. Present Illness. Patient entered hospital complaining of headache, backache, and pain in stomach. Two weeks previously symptoms began with general aching sensation and tired feeling. Shortly after this he began to have nausea, and there was anorexia. Epistaxis was frequent. Bowels rather costive. Had been in bed one week. Physical Examination. Fairly well nourished and developed ; mind clear ; face flushed ; lips dry ; tongue coated. Pulse 80, full, but readily compressible, dicrotic. Temperature 101.8°. Lungs and heart BRIGGS: PERFORATION IN TYPHOID FEVER. 11 negative. Spleen apparently not enlarged. Abdomen not distended, soft, and showing numerous rose-spots. Leucocytes 4900 per cubic millimetre ; haemoglobin, 85 per cent. Urine, albumin faint trace ; diazo reaction present. Widal reaction positive. Daring the next five days the symptoms were those of a rather mild case of typhoid fever, the temperature coming, down gradually, with good drops following tub baths. There was no diarrhoea, no intestinal hemorrhages. The abdomen was soft and flat. The progress of the case was very satisfactory. On Jane 22d, the twentieth day of the disease, the patient had been as usual all through the day. No tub baths had been given. At 6.30 P.M. he was seized with a sharp abdominal pain which was not definitely localized. When seen a few minutes after this there was no abdominal distention, bat marked muscular rigidity and general sensitiveness. The palse was 76 and of good quality ; the temperature 102.8", which was 1° higher than at 4 p.m. The leucocytes at 6.45 p.m. were 9200 per cubic millimetre. The pain was sufficient to make the patient groan audibly. A small enema with turpentine, two drachms, was given, which was followed by a light-yellow stool and a small amount of gas, but afforded no relief from the pain. Morph. sulph., gr. one- quarter, *as given hypodermically. At 7 p.m. the pulse was 88, and the temperature had fallen 1°. At 7.30 p.m. the patient was seen witih Dr. J. L. Martin. The pulse was 110, low tension, readily compres- sible. The patient had a dusky appearance, and there was marked cyanosis of the lips and finger-nails. Respirations were shallow, mostly thoracic. The abdomen was not distended, but there was marked rigidity over the lower half, and was a little more evident on the right side ; there was considerable general sensitiveness. The patient still complained of moderately severe pain in lower half of abdomen. At 8 P.M. the pulse had risen to 120. The cyanosis was slightly less marked, but otherwise the condition was about the same. The leucocytes were 13,000 per cubic millimetre. At 8.30 p.m. the leucocytes were 13,200 per cubic millimetre. Operation, June 22d, 9.30 p.m., by Dr. C. E. Briggs, three hours after first appearance of symptoms. Local anaesthesia, cocaine solution, 1 per cent., preceded by morph. sulph., gr. one-quarter, hypoder- mically ; chloroform. Time, fifty-five minutes, forty minutes of which were under chloroform. On reaching the operating-table the abdominal pain was still present, and there was slight distention. The cyanosis was considerably less. Pulse 120, Under local anaesthesia an incision was made six centi- metres long in right linea semilunaris. Although the patient suffered no pain, he became very nervous and quite beyond control, and light chloroform anaesthesia was substituted. Owing to the boy's nervous condition the opening of the abdomen under cocaine required great patience, consuming fifteen minutes. On opening the peritoneum a moderate amount of thin, yellowish-gray fluid escaped. Caecum readily found, ileum reached, and perforation found on its free border about eleven centimetres from the ileocaecal valve. It was about two milli- metres in diameter and exuded bowel contents when pressed. About it the bowel was red and thickened for about two centimetres in diameter. The perforation was wiped off with bichloride and turned in with two rows of running fine silk Lembert sutures for a length of three centi- 12 BKIGGS: PERFOEATION IN TYPHOID FEVEK. metres. Over these were placed three interrupted Lembert sutures to relieve the strain, as the bowel was very friable. The cseeum and ileum for about one metre were searched with eye and finger for perforations or thin ulcers, but nothing suggesting a threatening perforation was found. The ileum for about half this distance was injected and seemed somewhat thickened. . Appendix normal. No fibrinous flakes were observed. No adhesions were found. Incision enlarged to nine centi- metres. Abdomen was thoroughly irrigated with a large amount oi sterile water, with special reference to the pelvis and flanks ; the intes- tines were not' removed. The pelvis contained a large amount or trnn, purulent-looking material, rather different in appeara,nce from that seen on opening the abdomen. Examination of cover-slip preparations from this fluid showed numerous large and small bacilli and a few small cocci. Intestinal coils near pelvis were injected, but remainder of abdomen appeared nearly normal. Abdomen flushed till perfectly clear. Glass drainage-tube placed in pelvis, brought out through lower end of wound, above which the wound was closed with interrupted silk- worm-gut sutures. The site of perforation was left immediately below the incision. During the operation the patient's pulse remained at 120, good quality, rising a little during recovery from anaesthesia. Patient was but very slightly under chloroform, and was moving most of the time except during closure of the abdominal wall. The forty minutes under chloroform were largely spent in careful examination of the bowel, and in very thorough irrigation of the peritoneal cavity. The subsequent treatment of the case was by nutrient and stimulating enemata, strychnine, sponge baths, and liquids by mouth. During the first night there was very profuse sweating, which was relieved by atropine. The day after the operation, June 23d, the patient began to vomit a dark-brown fluid, and this continued for several hours, the patient's general condition becoming considerably affected. The stomach was washed out, about three pints of dark-greenish fluid being evacuated, after which the vomiting ceased, and there was immediate and steady improvement. From this time on the patient made a progressive and uninterrupted recovery, the pulse coming down to between 80 and 90 on the second day after the operation, and the temperature falling gradually to 99° by the end of the first week. The leucocytes on the second, sixth, and eighth days after operation were 15,400, 5400, 5600 per cubic milli- metre, respectively. The discharge from the wound was moderate. The glass tube was gradually withdrawn, being replaced by gauze pack- ing on the eighth day, which was entirely omitted four days later. The wound was very indolent, and there was considerable sloughing, but no active suppuration. The wound was entirely healed four weeks after operation, but owing to the persistence of a slight degree of fever — 99^ to 99.5° — the patient was kept in bed nearly two weeks longer. The cause of this temperature was never discovered. The patient was discharged August 19th, fifty-eight days after operation, in excellent condition. Bacteriological Examination. Cover-slips and cultures taken from the abdominal fluid when the peritoneum was first opened were neo-a- tive.. Cover-slips taken from the seat of the ulcer and from the turbid pelvic fluid mentioned showed medium-sized bacilli and a few small cocci. Plate cultures from these two localities showed bacillus mucosus BEIGGS: PERFORATION IN TYPHOID FEVER. 13 capsulatus. Cultures taken from the bottom of the drainage-tube four days after operation gave a pure growth of the same organism. Case VI. — Male, aged twenty-eight years. Patient entered hospital with general peritonitis ; no history obtainable. At entrance leucocytes 10,600 per cubic millimetre ; Widal reaction positive ; diazo reaction positive. Diarrhoea followed by constipation ; no intestinal hemor- rhages. Marked abdominal distention, rigidity, pain, and sensitiveness. Date of perforative symptoms not determined. Operation. Period after perforation not known. Ether. Time, thirty minutes. Right lateral incision. Turbid abdominal fluid. Per- foration of ileum thirty centimetres from caecum ; clo.sed by continuous Lembert catgut sutures. Irrigation with sterile water ; intestines not removed. Gauze aod glass-tube drainage. Cultures from abdominal fluid, unidentified bacillus, probably bacillus coli communis, or bacillus mucosus capsulatus. Death four hours after operation, general peri- tonitis. Autopsy. Intestinal sutures intact, no other perforations. General peritonitis. Cultures from heart, lungs, kidneys, and brain, bacillus mucosus capsulatus ; from general peritoneal cavity, negative. J. K., medical No. 1302 ; surgical No. 1660. Service of Dr. E. F. Cashing. Male, aged twenty-eight years, white, German, single, laborer. Admitted August 20, 1900 ; discharged August 21st. Dead. History was obtained subsequently for purposes of record, but no information was at hand at time of entrance for aid in diagnosis. Family History. Negative. Previous History. No former illnesses known. Habits good. Present Illness. Patient entered hospital complaining of abdominal pain. The symptoms began with headache eight days previously. For the last six days he had abdominal pain with some distention. There was diarrhoea at first, but during the last few days there was but one movement. There was no vomiting or epistaxis. No fever was noted. Was in bed only pa,rt of this time. Abdominal pain had been constant. Physical Examination. Patient entered in the evening, and no his- tory was available to aid in diagnosis. Moderately well developed, poorly nourished. He had the appearance of a very sick man. Tongue coated and dry. Pulse 120, irregular, but moderately strong. Tem- perature 100.8°. Lungs : on right side below clavicle there was slightly increased vocal fremitus ; otherwise negative. Heart appeared not enlarged, sounds weak. Liver dulness from fourth to sixth rib in mammary line. Spleen : dulness anterior to about mid-axillary line ; palpation impossible on account of distention. Abdomen distended, tympanitic, moderately rigid, sensitive to pressure, more marked in lower half and possibly slightly increased on right side ; no rose-spots ; doubtful dulness in both flanks. Leucocytes 10,600 per cubic milli- metre at 9 P.M. Urine : albumin i per cent. ; diazo reaction faint. Widal reaction positive ; both of these examinations made the follow- ing morning. The case was thought to be one of general peritonitis from perfora- tion of a typhoid ulcer, although the existence of typhoid fever was not absolutely determined until the next morning. Circumstances, how- ever, made an immediate operation impossible. During the night he was very actively stimulated. The next morning the Widal reaction determined the diagnosis. The case was seen by Dr. D. P. Allen about 14 BRIGGS: PERFORATION IN TYPHOID FEVEE. 9 A.M. The abdominal distention was extreme. There was genei-al tympany, marked muscular rigidity, and general sensitivenes;^. -Liie patient was in a condition of profound shock. Pulse 140 and very poor quality ; temperature 105.6° ; respirations 40. The case was con- sidered practically hopeless. Operation, August 21st, 10.30 a.m., by Dr. D. P. Allen. ^ Time, after onset of symptoms not known. Anresthetic, ether. Time, thirty minutes. Incision seven centimetres long in right linea semilunaris opposite anterior superior spine. On opening the peritoneum a moderate quan- tity of straw-colored fluid and thin pus escaped, containing fecal matter. Perforation of the ileum was found on the free border about two milli- metres in diameter, about thirty centimetres from the caecum. This was closed with two rows of running catgut sutures, strengthened by four interrupted silk sutures. There was considerable agglutination of the intestinal coils in the neighborhood of the perforation, together with congestion of the remaining coils of intestines and the parietal peritoneum. No other perforations found. A considerable amount of thin pus was found high up under the liver and a considerable amount in the pelvis. The abdomen was thoroughly irrigated with sterile water, the intestines not being removed. Gauze and glass-tube drain- age from pelvic cavity and subhepatic region. Pulse during the opera- tion 180, very weak, often imperceptible. The patient was highly stimulated with strychnine and subcutaneous saline infusions. There was no reaction, however, and he died at 3 p.m., four hours after the operation. Bacteriological Examination. Cover-slips from the abdominal fluid showed cocci and small bacilli. Tube cultures showed a pure culture of bacillus coli communis or bacillus mucosus capsulatus, the exact identity apparently not having been established. Autopsy, complete. No. 194, August 26th. Anatomical diagnosis, typhoid fever, ulceration, and perforation with general peritonitis ; apical tuberculosis and hypostatic pneumonia ; acute splenic tumor. Special lesion ; thirty centimetres above caecum there was a perforating ulcer of the ileum, closed with catgut sutures, which were still intact ; there were several other ulcers with very thin bases in the neighborhood, but no perforations. Cultures from heart, lungs, kidneys, and brain showed bacillus mucosus capsulatus ; from general peritoneal cavity, negative, owing probably to the recent irrigation. GENERAL Considerations. That all six cases were typhoid fever is readily established. The Widal reaction was positive in four cases — III., IV., v., and VI. In the first two cases of the series the Widal reaction was negative at entrance, and for some reason was not recorded later. The course of the disease was perfectly typical in all instances. A perforation was found in all cases except Case IV., and this case had a positive Widal reaction. Autopsy in the fatal cases — I., II., III. and VI. — showed typical bowel lesions. The cases with one exception — Case V. — were all severe infections and this severity, not the complication, was held responsible for the death in two cases — I. and III. One case — III. — occurred dnriua; a moderately severe relapse, following a rather mild primary course. BRIGGS: PERFORATION IN TYPHOID FEVER. 15 The bowels were constipated in three cases — II., III., and V. — in the first of which the condition was very marlced. In one case — Case I. — there was moderate diarrhoea. In Case IV. the bowels moved easily with enema, neither marked constipation nor diarrhtea being observed. In Case VI. the bowels were loose at first, followed by rather aggravated constipation. Intestinal hemorrhages occurred in only one instance — Case II. In this case it was present on the seventeenth and eighteenth days of the disease, was slight in amount, but was accompanied with considerable abdominal pain. The last hemorrhage was apparently coincident with the perforation. The ages were sixteen, seventeen, seventeen, twenty-eight, thirty-one, and forty-seven. There was but one female in the series — Case IV. She was colored, and no perforation was found. The remaining cases were males, and white. In Finney's series but one case occurred in a negro, and that was fatal. SiGss AND Symptoms. As compared with the large proportion of the more extended collections, the cases presented show a remarkably distinct set of symptoms. In Case VI. the patient entered with unmis- takable signs of general peritonitis, and the primary symptoms of per- foration in this case were not observed. The discussion of the signs and symptoms is restricted, therefore, to the first five cases. The date of appearance of symptoms was the fifth day in Case IV. ; the eighteenth day in Case ' II. ; the twentieth day in Case V. ; the twenty-first day in Case I. ; the forty-third day, the twelfth day of a relapse, in Case III. In Case VI., judging from the probably inaccurate history, the perforation occurred early, probably on the fifth or sixth day. The abdominal symptoms were quite well marked in all cases. Pain appeared suddenly in three cases — Cases II., III., and V., and came on gradually in two cases — Cases I. and IV. It was confined to the lower half of the abdomen in three cases — Cases I., II., and V., but in Case I. it eventually radiated to the upper half and became general. In Case III. it was general, and in Case IV. it appeared to be general, but the delirious condition of the patient made accurate observation impos- sible. The pain was severe in three cases — Cases II., III., and V. — and moderate in two cases — Cases I. and IV. In Case I. pain also radiated to the penis, which was the only instance. Sensitiveness to pressure was slight in two cases — Cases II. and IV. ; moderate in Case III. ; marked in Cases I. and V. It was general in three cases — Cases III., IV., and V., and was confined to the lower half of the abdomen in two cases — Cases I. and II. Increased sensitiveness in the right lower quadrant so frequently spoken of was not observed. Muscular resistance was a very marked sign in all but one of the 16 BRIGGS: PEEFOEATION IN TYPHOID FEVEE. cases — Case IV. In this instance the resistance was but moder; le, and no perforation was found. The resistance was general in three cases Cases I., II., and IV., and was confined to the right half of the abdomen in one case— Case III. In Case V. it was confined to the lower half of the abdomen, and was considerably more marked on the right side. Distention was observed as moderate in Cases I., II., HI-. and IV., but was present to a slight degree only in Case V. Tympany to a greater or less extent was observed in all of the cases. Slight vomiting was observed in Cases I., II., and III., and was absent in Cases IV. and V. The pulse showed a marked rise in all instances— 20 to 50 beats. The quality of the pulse also became progressively worse with the increase in the number of beats, and was a very valuable sign. In Case IV., where no perforation was found, the pulse rose very gradually from 120 to 140. The marked severity of the case, however, made all of the signs and symptoms in this instance particularly difficult to estimate with much accuracy. The temperature was unaffected in two cases — Cases II. and III. — and in two cases — Cases I. and IV. — there was a distinct rise, 101° to 102.2°, and 102.8° to 105.5°, respectively. In but one case — Case V. — was there a fall in temperature, and in this instance of but 1° only, 102.8° to 101.8°. The respiratory rate was markedly increased. The nature of the respirations, however, was noted as largely thoracic in all the cases. This was especially marked, together with a very shallow quality, in Case V. The condition of general systemic shock, as observed shortly before operation, was very marked in Case II., and was present to a considerable degree in Case IV. It was noted as moderate in Cases I., III., and V. Cyanosis, especially of the lips and finger-nails, was marked in Case V., moderate in Case III., slight in Cases I. and II., and absent in Case IV. The white blood count as recorded showed a distinct rise in all but one instance — Case I. The records are as follows : Case I., entrance, 5700 per cubic millimetre ; perforation, six days later, 4460 per cubic millimetre. Case II., entrance, 5980 ; perforation, eight days later, no count recorded. Case III., entrance, 5700 ; perforation, thirty-two days later, 31,600. Case IV., entrance, 9932; symptoms of perfora- tion, two days later, 15,080. Case V., entrance, 4900 ; perforation, five days later, 9200, 13,000, 13,200. Case VI., entrance, 10,600. It is to be noted in Case IV. that subsequent to the operation the leuco- cytes on one occasion rose from 8060 to 50,660, associated with consid- erable abdominal pain, but no other constitutional effect. The Condition of Patient at Operation'. The cases with but one exception — Case V. — were rather unfavorable to operation, two being quite so, and one practically hopeless. As this is regarded as of considerable importance, the conditions will be given in more detail ; BRIG(J8: PER?-(IRATION IN TYPHOID FEVEE. 17 Case I. The patient's condition was poor, although not absolutely unfavorable to operation. The perforation occurred on the twenty-first day of the disease, in a moderately severe case, before the patient had begun to recover at all from the effects of his illness. He was weak and emaciated, and had comparatively little recuperative power. Case II. This case was quite unfavorable for operation. The patient was a strong, athletic boy, but the case had run a severe course, and the per- foration occurred on the eighteenth day during the very height of the disease. The patient was weak and exhausted, and the rapidity with which he succumbed to the peritoneal infection was not surprising. Case III. This case was somewhat more favorable, although not very promising. The perforation occurred during the height of a rather severe relapse, being the forty-third day of the disease, the twelfth day of the relapse. He was quite weak, but appeared at the time to have a fair degree of strength. Case IV. The case was desperately sick at the time perforation was feared, but the symptoms appeared so early, the fifth day, that the patient's constitution had not been badly reduced. On this account it is believed she would have stood fairly well a more extensive operation. Case V. The condition of this patient was very favorable for operation, in marked contrast to the other cases. The case was mild, and when the perforation occurred, the twentieth day, the temperature had begun to come down ; he was not badly reduced. Case VI. The case, as stated, was in extremis at entrance, and was regarded as practically hopeless. DuEATiON OF Symptoms. The lapse of time between the first appearance of symptoms and the operation was short in all but one instance — Case II. In Case III. the operation was performed one hour after the appearance of symptoms. Cases V. , IV., and I. were operated upon within three, four, and five hours, respectively, although the exact time of suspected perforation in Case IV. was somewhat indefinite, owing to the delirium of the patient. Case II. was opened eight hours after definite symptoms first appeared ; three days previously the patient had had abdominal pain following a tub bath, which grew less the next day, disappeared entirely the day before active symptoms, and was entirely unassociated with constitutional signs suggesting perforation. The duration of symptoms in Case VI. could be only roughly estimated, as perforation occurred before entrance, but it was probably twenty-four to forty-eight hours. Operation. The technique of the operations presented very little that was distinctive. The anaesthetic was general in four cases, local in two cases. Chloroform, very lightly given and preceded by morph. sulph., gr. one-quarter, was administered in Cases I., II., and V. Ether was similarly given in Case VI. Local anaesthesia, cocaine solution, ^ and 1 per cent., preceded by morph. sulph., gr. one-quarter, was employed 18 BKIGGS: PERFOEATION IN TYPHOID FEVEE. in Cases III. and IV. In Case V. the abdomen was entered under cocaine, but the nervous condition of the patient required the adminis- tration of chloroform. The incision in each instance was in the right linea semilunaris opposite the anterior superior spine, three to nine centimetres long. In each instance, except Case IV., the cKCum was readily found, the ileum, csecum, and appendix examined. In Case IV. the peritoneal cavity was so visibly normal that the ileum was not withdrawn. This, however, was considered a fault in the operation. An examination of the ileum was made for about one metre in Cases I., II., and V., but in Cases III. and VI. the inspection was not carried beyond the perfora- tion. In all cases careful search was made for thin areas, threatening perforations, as well as for complete perforations. The perforation was excised in no instance. The closure of the perforation was by two rows of continuous Lembert sutures in each case, in two of which — Cases V. and VI. — these were strengthened by another line of interrupted silk sutures. The suture material was catgut in Cases I. and VI., and silk in the remainder. The irrigation was with very large quantities of sterile water, by means of a large flushing tube. With this tube all parts of the peritoneal cavity were readily reached, special attention being given to the lumbar fossse and pelvic cavity. Only the portion of the ileum examined was removed from the abdomen, and an effort was made to handle the intestines as little as possible. Drainage was used in all cases except the exploration — Case IV. Iodoform gauze alone was used in Cases I. and II. ; glass-tube and iodoform gauze in Cases III. and VI., and a single glass drainage tube from the pelvis in Case V. The site of the suture was left close to the abdominal wall in all instances. The wounds were all partially closed with interrupted silkworm-gut sutures through all layers, except Case IV., which was completely closed with interrupted buried silk sutures. The time of operation was twenty minutes in one case, twenty-five minutes in two cases, and thirty minutes in two cases. In Case V. fifty-five minutes were consumed, fifteen minutes in going through the abdominal wall under cocaine, and forty minutes, or the remainder of the operation, under chloroform. The condition of the patient was apparently unaffected by the operation in Cases III. and IV.; it was excellent in Oase V. ; was poor in Cases I. and II., and very bad in Case VI. Pathology. Pathological observations of the cases are of special interest, and it is to be greatly regretted that more complete autopsies could not be obtained. There was a moderate degree of peritonitis at the time of operation in all the cases of perforation, judging from the amount of free fluid in the abdominal cavity. In Case I. there was considerable dark-brown fluid with fecal odor ; in Case II. a moderate amount of dark fluid : in BRIGGS: PEEFOEATION IN TYPHOID PEVEE. 19 Case III. a small amount of brownish fluid ; in Case V. a moderate amount of thin, yellowish-gray fluid, and in the pelvis a large amount of thin, purulent-looking fluid ; in Case VI. a moderate amount of straw-colored fluid with a large accumulation of thin, purulent fluid under the liver and in the pelvis. The perforation was single in each instance, and occurred on the free border of the ileum. In only one case — Case II. — was a threatening perforation found. The perforations were all small, two to three milli- metres in diameter, and were surrounded by an indurated friable base about two centimetres in diameter in several instances. In one case — Case II. — there was a small amount of fibrinous exudate in the neigh- borhood of the perforation. The perforations were eight, eleven, twenty-five, thirty, and forty-five centimetres from the caecum in Cases I., v., II., VI., and III., respectively. The ileum, aside from the perforation, was altered in each instance, ranging from slight congestion to a thickened dark appearance. The appendix was normal in all cases. The csecum was appreciably con- gested in Case I. only. The remaining coils of intestines were slightly congested in Case I. In Case V. those lying in the pelvis were con- siderably congested ; in Case VI. there was a marked congestion, and agglutination of the coils in some places. The parietal peritoneum was appreciably congested in Cases I. and VI. only. In none of the cases were there fibrinous flakes in the peritoneal cavity, except for a small amount of deposit about the perforation in Case II. No adhesions were observed in any of the cases except in Case VI. In Case IV., in which the symptoms appeared on the fifth day and no perforation was found, the abdominal cavity was perfectly normal so far as could be observed. Bacteriology. Bacteriological examination of the free abdominal fluid was made in all cases, but in only one instance was a culture taken from the perforation. Cover-slip preparations from Case I. showed a medium-sized bacillus and a small coccus ; plate cultures grew bacillus coli communis only. Case II. showed a pure growth of bacillus coli communis in plate cultures. Case III. showed in cover-slips a few cocci and numerous bacilli ; in plate cultures, a pure growth of staphylo- coccus pyogenes aureus. Case IV., in which no perforation was found, cover-slips and cultures were negative. Case V., cover- slips and cul- tures from fluid on first opening the abdomen were negative ; cover-slip preparations from the perforation and the turbid pelvic fluid showed numerous medium-sized bacilli and a few cocci. A plate culture from these two localities gave a pure growth of bacillus mucosus capsulatus ; a tube culture taken from the bottom of the drainage-tube in the pelvis four days after operation showed a pure culture of the same organism. In Case VI. cover-slip preparations showed cocci and small bacilli, the 20 BRIGGS: PEEFOEATION IN TYPBOID FEVEK. nature of which was not definitely determined, but was apparently bacillus coli communis or bacillus mucosus capsulatus. Autopsies, complete, Case VI., or restricted. Cases I,, II., and III-, were made in the four cases of death. General peritonitis was present in two cases only — Cases II. and VI. — in the other two cases the gen- eral peritoneal cavity being free from fiuid and adhesions, and the cul- tures from the same being negative. Case I., death seventy-five hours after operation, showed a local peritonitis about the gauze packing, into which the original perforation and four small neighboring perfora- tions opened ; the catgut sutures were almost entirely absorbed ; no general peritonitis ; the ileum showed several large confluent ulcers, not perforating. Cultures from the local peritonitis showed bacillus coli communis ; cover-slips and cultures from general peritoneal cavity were negative. Case II., death sixteen hours after operation, sutures intact, no other perforations ; general peritonitis, cultures growing bacillus coli communis. Case III., death fifty-eight hours after opera- tion ; intestinal sutures intact, no additional perforation ; no general peritonitis, cover-slips and cultures from general peritoneal cavity nega- tive. Case VI., death four hours after operation ; suture intact, no other perforations : general peritonitis ; cultures from heart, lungs, kid- neys, and brain showed bacillus mucosus capsulatus ; cultures from gene- ral peritoneal cavity were negative, probably owing to recent irrigation. Results. In considering the results of these cases from the stand- point of operations for the relief of typhoid perforation, it is manifestly improper to include Case VI., which was suffering from general peri- tonitis before the case was seen at all, and was operated upon solely for the relief of the abdominal infection. Case IV., in which no perfora- tion was found, adds one other to the increasing list of explorations without untoward results in cases of suspected perforation, but cannot, of course, be included in the list of cases where perforation has actually occurred. In the four remaining cases with perforation, one alone recovered. Case V. In all four cases there was a sufficient degree of peritonitis to cause an accumulation to a greater or less degree of free abdominal fluid, in which organisms were found in both cover- slip preparations and cultures. Only one case — Case II. — as shown at autopsy, died of general peritonitis. In this instance the operation was deferred eight hours, the longest delay in the series, and an evident mistake. The other two fatal cases — Cases I. and III. — showed local peritonitis about the packing, as occurs in any abdominal wound that is drained, but the general peritoneal cavity was free from fluid and adhesions, and no organisms were obtained from it in cover-slip preparations or cultures. It can be safely said, then, that whatever else one may wish to consider the cause of death in these two cases, it was not general peritonitis. A very large percentage of the fatal BEIGGS: PERFORATION IN TYPHOID FEVER. 21 cases following operation so far reported, at least 80 per cent., died from general peritonitis. The percentage cannot be given with any great degree of accuracy, however, owing to the small number of post- mortem examinations. The fact that it was possible to prevent the occurrence of general peritonitis in three out of four cases of perfora- tion is worthy of note despite the fact that only one case recovered. The two fatal cases without general peritonitis — Cases I. and III.— lived seventy-six and fifty-eight hours, respectively, after operation. In Case III. the perforation occurred at the height of a moderately severe relapse, in which the clinical evidence of increasing typhoidal toxaemia was apparent before operation. The operation was performed under local anaesthesia without shock to the patient that was apparent to careful observation. It is confidently believed that this death was due to typhoidal toxaemia, and was independent of the operation. As a complete autopsy was not obtainable, however, this statement cannot be maintained with absolute certainty, although it is very strongly supported by clinical symptoms and the absence of general peritonitis. In Case I. perforation occurred on the twenty-first day during the height of the fever in a moderately severe case, before the patient had begun to recover from the effects of the disease, and it was evident that he had very little recuperative power. The clinical appearance of the patient on the last day of life was one of exhaustion and gradual heart failure, and was not unexpected, owing to the condition of low vitality which was apparent before operation. This of course was added to very materially by the shock of operation, which was, unfor- tunately it is believed in this instance, performed under general rather than local anaesthesia. It was also undoubtedly added to in a consid- erable degree by four subsequent perforations near the original lesion. The closure of the perforation was not successful, owing to the subse- quent absorption of the catgut sutures. By rare good fortune the four subsequent perforations had opened along the line of drainage, and the extravasation was being cared for completely, as far as could be seen at autopsy, by the gauze drainage going down to the original perfora- tion. The general peritoneal cavity was entirely free from evidences of peritonitis, there being no fluid or adhesions, and the cover-slip preparations and cultures from the same being negative. It is believed that the patient died of exhaustion. It is certain he did not die of general peritonitis. Here, again, however, the absence of complete autopsy leaves the absolute cause of death somewhat in doubt. Extracted from The American Journal of the Medical Sciences, May, 1903. INTESTINAL PERFORATION DURING THE COURSE OF TYPHOID FEVER, AND ITS SURGICAL ASPECTS. By C. E. Briggs, M.D., OF CLEVELAND, OHIO. Frequency and Mortality op Perforation. While the per- centage of deaths resulting from typhoid fever, and more particu- larly the proportion of these deaths due to perforation of the bowel is a -well-established fact and quite generally known, we believe that the specific appreciation of the appalling number of these individual cases of perforation is by no meaus appreciated by the medical profession in general. While we all appreciate that the death rate from typhoid fever is from 7 to 14 per cent., averaging, probably, for the entire United States about 10 per cent., and that of these fully one-third are due to perforation, we think that few could have read without marked surprise the statistics recently collected by Dr. H. M. Taylor in an article on " Typhoid Perforation: Its Frequency, Prog- nosis, Diagnosis, and Treatment," New York Medical Journal, Feb- ruary 1, 1902, vol. Ixxv. p. 193. From various statistical sources gathered in 1896 the author has determined the number of cases of typhoid fever in the United States as about 500,000 per annum, and the death rate about 50,000, which is doubtless a very fair average mortality. The frequency of perforation has been so definitely estab- lished from so many sources as being about 33J per cent, of the deaths, that this number can hardly be questioned. This would give a death rate of something over 16,660 cases from perforation in typhoid fever in one year in the United States. Of this number it is now definitely known that 5000 to 8000 cases annually can be saved by operative interference. These figures represent, we believe, an adequate appreci-7 ation of the usefulness of surgical interference in the United States at the present time, and are certainly striking and instructive. Owing to the very rapid medical and surgical progress along the lines of diagnosis and treatment in typhoid perforation during the last five years, the usefulness of operation in such cases has become estab- lished so far beyond the possibility of any doubt that it is no longer found necessary to urge and justify the procedure in the large majority of the cases. The question of recovery in such cases without operation 2 BRIGGS: INTESTINAL PERFORATION. is a matter of very grave doubt, although it has been absolutely known to occur in a few isolated instances where the damage was well local- ized near the caecum, the cases assuming more the characteristics of appendicitis than those of perforation. It is an unwarrantable assump- tion, however, that unoperated cases recovering from mild or even marked symptoms of perforation are actual cases of recovery from per- foration itself. At a time when medical treatment was practically the only means employed, the possibility of recovery was a very serious question even among the best authorities. In an article by Dr. R. H. Fitz (Transactions of the A-^yoeiation of American Physicians, 1891, vol. vi. p. 200) on "Intestinal Perforation in Typhoid Fever: Its Prog- nosis and Treatment," the author states: "Since perforation of the intestine in typhoid fever may take place without any subjective symptoms, and since suggestive — even so-called characteristic — symp- toms may occur without any perforation having taken place, it must be admitted that recovery from such symptoms is no satisfactory evi- dence of recovery from perforation." Any practitioner who has been intimately associated with typhoid fever for any length of time has seen cases of typhoid suggesting perforation that have gone on to re- covery, but too often there has been an unwarranted assumption that perforation actually occurred, and the instance has been mentioned as a remarkable cure. Since the usefulness of operation has been estab- lished in these cases it has become a well-known fact that the gravest symptoms of perforation may be present without the actual lesion, and, on the other hand, a rather premature assumption from vague symptoms has been absolutely justified by the discovery of perforation at operatioQ. It is needless to say that these remarks apply to the diagnosis of perforation and not to the diagnosis of general peritonitis following perforation. The same position, however, the lack of signs and symptoms corresponding to the actual pathological condition, has been long since established with reference to general peritonitis that has recently been demonstrated with reference to typhoid perforation. The man who assumes an unassailable position in either one of the above-mentioned conditions in instances which cannot be subjected to actual demonstration, can no longer be considered to hold a tenable and justifiable position. It must be admitted at the present time that it is a far safer and much more justifiable assumption that the mor- tality in unoperated cases of intestinal perforation in typhoid fever is practically 100 per cent. It was eighteen years ago that the first operation for perforation was performed. Dr. W. W. Keen, in 1898, in his book on Surgical Complications and Seqii.elce of Typhoid Fever, published a table of 83 cases collected by Dr. T. S. Westcott, and in the Journal of the Ameri- can Medical Association for January 20, 1900, in an article on " The BRIGG8: INTESTINAL PEEFOEATION. 6 Surgical Treatment of Perforation of the Bowel in Typhoid Fever," the same author gives a table of 75 additional cases collected by Dr. M. B. Tinker, 158 cases in all, which were considered to include all the cases published to January 1, 1900. Since that time the number of published cases has rapidly increased, but owing to greater familiarity with the subject and the increased frequency of operation it is believed that the number of published cases is far below the number of opera- tions performed, especially in unsuccessful cases. The majority of the reports have been glaringly incomplete, indicating either a deplor- able lack of scientific observation, or failure on the part of the writers to appreciate the increasing necessity for careful and detailed reports in such cases. The grosser and more evident facts with reference to- diagnosis have already been well established, but the refinements in diagnosis which are going to make it possible to detect the occurrence of this comj)lication at a much earlier stage and in a much larger pro- portion of the cases are still lacking. The question of general peri- tonitis ought not to enter at all into the consideration of perforation. Many of the pathognomonic signs, so-called, of perforation are now known to be merely an indication of general peritonitis. It is only by keeping these two conditions distinctly separated in the general medi- cal mind that the diagnosis of perforation is to be more frequently and more accurately made, and it is only by strict and careful attention from now on to the refinements of diagnosis that we shall be saved the humiliation of operating for perforation and so frequently finding extensive peritoneal infection. Much improvement in operative tech- nique at the present time cannot be expected, as this has already reached as advanced a stage of perfection as our present surgical knowledge will permit. Improvement is to be looked for almost entirely along the line of early and accurate diagnosis, which can only be the result of careful clinical observation, and what is of equal or even greater importance, of accurate pathological study. The path- ology of these cases, especially the information afforded by extended post-mortem examinations in unsuccessful cases, has been even more neglected than the clinical study, whereas the whole question of prog- nosis and of the causes of death depend absolutely upon post-mortem findings. With but relatively few exceptions it has been assumed that the mortality following operation has been due to general peritonitis. It cannot be denied at the present time that this is not for the most part a correct view, but the comparatively few post-mortem examinations recorded certainly indicate that, a good many of the unsuccessful cases die from causes entirely independent of the operation, and which at the present time are beyond means of control. It is in the hope of stimu- lating interest in the more uncertain points of diagnosis, and especially in the pathological knowledge of these cases, as well as affording some 4 BEIGGS: INTESTINAL PEEFOBATIOK. definite suggestions for accomplishing these objects, that the present article is presented. Many of the observations contained in this paper have been suggested by cases of typhoid perforation studied at the I/akeside Hospital in Cleveland during the last four years, several of ■which were reported by the writer from the surgical clinic of the hos- pital in The American Journal of the Medical Sciences, Jan- uary, 1902, vol. cxxii). p. 38, " Laparotomy for Perforation in Typhoid i^'ever." It may seem at first that it requires considerable temerity to speak •of the diagnosis of typhoid perforation from a purely surgical stand- point. The view of the medical profession, however, has changed so radically with reference to the treatment of perforation, that it may ■almost be said to come under the head of a surgical rather than a medical complication, in the same manner in which appendicitis is now ibrought to the observation of the surgeon, not solely at a time when •operation is immediately demanded, but for the sake of diagnosis in 'doubtful or somewhat obscure conditions. Surgical aid in the diag- mosis of appendicitis has become of the utmost value, and we feel equally certain that much the same consideration will prevail with refer- ence to typhoid perforation. It is largely on this account that the surgical interest in typhoid fever in general has recently become so manifest. The physical conditions of the abdomen are often so altered during the course of typhoid fever that the surgeon who is unfamiliar with the typhoidal abdomen is utterly incapable of giving a well- considered opinion in cases of suspected perforation. It was in view of this fact that Dr. Osier, a few years ago, so strongly urged the desirability of surgical observation of the typhoid cases at the Johns Hopkins Hospital, and it is largely on account of the lack of oppor- tunity for such observation among surgeons in general that so little has been done on the subject outside of large hospitals. The desirability of surgical familiarity with the abdominal conditions in all stages of typhoid fever cannot be too strongly urged, and its importance cannot be over-estimated. Etiological Considerations. In considering the subject of in- testinal perforation in typhoid fever, there are a number of points in the etiology of the condition which have already been determined with all the accuracy that is desirable, and which can be dispensed with in a few words. The time at which the perforation oeeurs, so far as diag- nosis is concerned, is of no consequence. It most frequently occurs during the third week, but in one of the Lakeside series there were very marked suggestions of perforation in one case on the fifth day of the disease, although no perforation was found, and in another case general peritonitis had followed a perforation occurring on the sixth day, as nearly as could be made out from the history. It may occur BRIGOS: INTESTINAL PBEPOEATION. 5 late in a prolonged convalescence, or in a relapse, so that from a stand- point of diagnosis the stage at which the patient has arrived in the illness must be entirely overlooked. The time of perforation is of much consequence, however, with reference to the general condition of the patient and to the way in which operation is borne, as will be mentioned later. The severity of the fever can be overlooked entirely in diagnosis, except for the fact that diagnosis is more difficult in severely sick cases, where one scarcely ever finds the indications so marked. It is the opinion of Dr. Osier that it occurs more frequently in severe cases ; but medical opinion in general is quite divided, and from the study of the reported cases one is certainly justified in a con- servative opinion. It certainly appears to occur about as frequently in mild as in severe cases, and in any event it occurs so frequently in both classes that the mildness or severity of the disease can be entirely laid aside with reference to diagnosis. The matter of constipation and diarrhoea is of little consequence in diagnosis, the cases being so equally divided that these conditions can also be overlooked. The question of age and sex is of slightly more importance, but this also can be over- looked almost entirely. It occurs much more frequently in young adults, but three or four case^ have been reported in children under fifteen years. It also occurs much more frequently in males than in females. Neither of these conditions can influence the diagnosis to any appreciable degree, however, except that one might be justified in delaying a little longer in cases of young children. It is of somewhat curious interest that only two instances of operation for perforation in a negro have been reported, and that in only one of these was perfora- tion actually found. The question of intestinal hemorrhage is of some- what more importance, but mainly from the fact that its occurrence at the time is very apt to mask or simulate perforation and render diag- nosis much more difficult. As an etiological feature, however, ante- cedent to perforation, we believe it can be entirely overlooked. The occurrence of hemorrhage is purely fortuitous, and depends solely upon the chance of the ulcerating surface encroaching upon a vessel of con- siderable size. It cannot be assumed that the ulcer is approaching the peritoneal surface more rapidly on that account, or that the process will even eventuate in perforation merely because the hemorrhage has occurred. A comparatively small number of the reported cases of perforation have been preceded by hemorrhage, and it is only a small percentage of the cases of hemorrhage that have eventuated in per- foration. In instances where the relation is suggested it is almost invariably a mere assumption that hemorrhage has come from the particular ulceration that has gone on to perforation. So that in con- sidering the diagnosis of perforation we think one can entirely overlook as etiological factors the period of the disease in which the perforation 6 BRIGGS: INTESTINAL PERFOEATIOK. occurs, the severity of the disease, the occurrence of constipation or diarrhoea, the question of age, sex, and color, and the occurrence of hemorrhage from its purely etiological standpoint. Symptoms and Signs of Perforation. In considering the symp- toms and signs of perforation we believe it is of the utmost importance to preserve a sharp distinction between the two. Some of the signs of perforation are doubtless always present, although possibly unobserved or misinterpreted. The amount of dependence one can place on symp- toms, however, depends entirely upon the mental condition of the patient and his appreciation of sensations. The patient whose facul- ties are moderately alert can afford very material aid through his per- ception of pain and his own conception of being very much worse or very much better. In proportion to the amount of apathy so often present to a greater or less degree, these sensations are much less acute, and the patient's ideas regarding himself are of much less con- sequence. In deeply apathetic or delirious cases the question of symp- toms must almost invariably be largely or entirely set aside. Conse- quently in the latter class of cases, which are almost always the severely sick cases, one must rely almost entirely upon physical signs and phenomena. As in so many other conditions, in the cases in which one is most desirous of making an early and careful diagnosis, the means at hand for making this diagnosis are much more strictly limited than in less severe cases where one can risk a little larger margin of error. It is thus evident that in most severe cases one is forced to be more radical in his diagnosis and establish the presumption of perforation on somewhat less sufHcient grounds. These two distinctions, then, between signs and symptoms on the one hand, and the clear-minded as against the apathetic and delirious patients on the other, at once assume pro- portions of the very highest importance, and it does not seem to the writer that these considerations can be too strongly urged. These are distinctions to which we feel justified in making frequent reference. Symptoms. In considering first the symptoms of perforation one finds for consideration pain, sensitiveness, the sensations associated with systemic shock, and the sensations of altered respiration. These have been named in the order of their relative importance in the vast majority of cases in which the sensibility of the patient is sufficiently acute to make them of any use whatsoever. Abdominal pain is by far the most important symptom, and if present demands the utmost consideration in forming a diagnosis. This is especially true where the patient has been comparatively free from previous abdominal complications and the symptom appears as a sud- den, sharp, often agonizing sensation, quite circumscribed, located in the lower part of the abdomen near the median line or toward the right side ; and especially if the pain remains severe and circumscribed for BKIGGS: INTESTINAL PERFOKATION. 7 an hour or more. Such a picture one can scarcely mistake, and with a very slight suggestion along the line of physical signs one must feel justified in making the diagnosis. The pain, however, may come on gradually, commencing almost anywhere in the abdomen, but more frequently starting near the median line, near or below the umbilicus. If the pain is general at the start it usually becomes more closely con- fined to the lower half of the abdomen within a short time. Wlien pain becomes more general after having been moderately well localized at the start, it suggests rather strongly that peritoneal infection is pro- gressing from a point of perforation. The pain, whether sudden or general in its first manifestations, localized or diffuse, may at the same time be either moderate or severe ; in general it can be said that the better localized the pain the more keenly it is appreciated, but the distinction is of vastly less consequence than the fact itself that pain is present. So that the occurrence of pain is a feature of the utmost importance when present. The location of the pain, its limits, its severity, the manner of appearance, and persistence are matters of secondary importance, except possibly the last-mentioned feature. In proportion to the degree to which these various features of the pain are accentuated these secondary considerations of pain may be said to be of importance, but the absence of such accentuation must be almost entirely overlooked in arriving at a diagnosis. Sensitiveness is, of course, very closely allied to pain, and may almost be considered as a leas strongly marked manifestation of the same sen- sation, so that many of the things mentioned with reference to pain are true to a less degree with reference to sensitiveness. All patients with abdominal pain are sensitive, but one is often obliged to rely con- siderably upon sensitiveness as a symptom in the absence of pain. The limits of sensitiveness are usually more narrow than those of pain, and on this account it is sometimes of aid in assisting the patient to more definitely locate his pain. Localized pain is accompanied by localized sensitiveness, but diffuse pain is not infrequently associated with cir- cumscribed sensitiveness, and this fact is often an aid of considerable importance. Spreading sensitiveness is open to the same interpretation as spreading pain, and is if anything a little more to be relied upon. The occurrence of sensitiveness is more common to the lower half of the abdomen or to the entire right half. Its frequent occurrence in the right lower quadrant, to which attention has been strongly called by several writers, does not appear to exist in many cases. The truth of this statement has been strongly suggested by the Lakeside series and seems to be justified by the large proportion of reported cases. The degree of accentuation increases the usefulness of sensitiveness, also, as a symptom, possibly a trifle more so than in cases of pain ; but here again the characterizing features of sensitiveness are of only 8 BRIGGS: INTESTINAL PERPOEATION. secondary importance as compared with the presence of sensitiveness itself. The sensations associated with systemic shock vary so greatly in different patients, and are so closely allied to certain signs to be mentioned later, that one is obliged to speak of them in a very general way. These sensations, general and varying as they are, however, in a responsive and clear-minded patient are worthy of careful consideration when present. Perhaps their meaning is best described by the expression so often used by the patient himself, literally or implied, of an " all-gone feeling," as if the "bottom had dropped out of him;" a feeling vaguely described but strikingly realistic, that some serious catastrophe has occurred. It is doubtless this that sometimes causes in these patients the intensely anxious and apprehensive countenance not infrequently seen, the exact meaning of which they are unable to put into words. It is scarcely possible to speak of these sensations in more detail. Their presence is unmistakable and of considerable importance ; their absence, again, may be entirely disregarded in making a diagnosis. The sensations of altered respiration are of little consequence com- pared with the other symptoms mentioned, or compared with respira- tion as a sign. The patient is apt to feel short of breath, and finds that his respirations for some reason or other must be largely costal. As a sensation it is present in almost direct proportion to the intensi- fied characteristics of pain already mentioned, and in the absence of pain, at least in the absence of pain and sensitiveness, it does not exist. The symptoms of altered respiration are of so little consequence that they may be entirely disregarded. Signs. But it is upon the signs suggesting perforation that one is forced to rely largely, and in apathetic and delirious cases one must rest his diagnosis upon these signs almost absolutely. Here, however, it is necessary to draw the distinction even closer between perforation and general peritonitis, and keep in mind that the diagnosis of general peritonitis is not the diagnosis of perforation. And against maintain- ing this distinction one cannot with propriety set the fact that both these conditions at times are very bizarre and misleading in their mani- festations, and that occasionally the postmortem examination reveals a condition of which the manifestations before death gave absolutely no recognizable indication. These signs suggesting perforation include, first, indications gathered from the abdominal examination alone, such as muscular resistance, the presence of gas within or outside of the in- testine ; and, second, certain general systemic indications, such as vomiting, altered respiration, temperature, pulse, altered blood condi- tions, and general systemic shock. Muscular resistance is, we believe, the most important of all these signs. It is a sign of the utmost importance in almost any acute intra- BEIGGS: INTESTINAL PEEFOEATION. \) abdomiaal condition, and its reliability can rarely be questioned witli impunity. It indicates very clearly that the abdomen is trying to pro- tect itself against injury from the outside, and is a condition largely independent of the volition of the patient. The presence of pain and sensitiveness is, to be sure, usually associated with muscular resistance to a greater or less degree, but the same patient, if he is deeply apathetic or delirious, and abdominal protection is still needed, would show muscular resistance. This muscular resistance is not infrequently general, but in cases of perforation in typhoid fever, since the lesion is almost invariably located in the lower half of the abdomen or on the right side, the resistance is accentuated in these localities. Its useful- ness is impaired to an inexperienced observer, however, in that many cases of uncomplicated typhoid show a varying amount of muscular resistance which must not be mistaken by one seeing the case for the first time. Here, again, is a very striking example of the necessity for surgical familiarity with the abdominal conditions in uncomplicated typhoid, and also of the great desirability of having typhoid cases under frequent surgical observation. Muscular resistance is probably never absent in perforation. If resistance was present before perfora- tion it is probably always increased, and in such cases an accurate esti- mation of the degree of increased resistance is of great value.' Distention and tympany from gas within the intestines are signs of no consequence with reference to perforation per se. Their absence is of equal inconsequence. As signs of spreading peritoneal infection, however, they are both of well-known and merited consequence and need not be discussed here. They are confirmatory evidences only in so far as one has been willing to run the risk of spreading infection secondary to the perforation. If there was gas in the intestines pre- vious to perforation, as is frequently the case, it would not be increased by the accident ; if the abdomen was soft and there was no distention one would not find a sudden accumulation of gas when perforation occurred. So that aside from the fact that distention and tympany may tend to obscure the diagnosis and render it more difficult they can both be set aside entirely with reference to the diagnosis of perforation itself, and are confirmatory only in proportion to the degree of spread- ing infection. The presence of abdominal gas outside the intestine, however, if its presence can be demonstrated with certainty, is practically a proof of 1 The presence and significance of muscular spasm nave been receiving considerable atten- tion of late. Muscular spasm is mentioned in comparatively few reports, and has doubtless been overlooked in many instances. The data is insuflicient to judge with any degree of accu- racy as to the frequency of its occurrence, but it is probably very often absent. We feel it to be a sign not infrequently confused with voluntary muscular contraction, and believe with Munro ("The Clinical Diagnosis of Typhoid Perforation," Boston Medical and Surgical Journal, February 5, 1903, vol. cxlvlii., No. 6) that at the present time it is often difaoult to properly estimate its significance. * 10 BRIGGS: INTESTINAL PEEFOEATION. perforation. Such evidence is usually sought through obliteratiou of the liver dulness. This manifestation, however, is a matter of such uncertainty that it renders the interpretation of the sign of practically no value. The obliteration of liver dulness has been demonstrated again and again to be due to a distended colon or small intestine as well as to free abdominal gas, and in several cases in which diagnosis was considered practicaly absolute on account of this sign, no perfora- tion was found. The absence of altered liver dulness is, of course, a matter of absolutely no consequence in diagnosis, and in cases where the liver dulness is altered to an appreciable extent, it is quite impos- sible to say whether it has been effected by free abdominal gas or dis- tended intestine. If the liver dulness is altered one may be justified in being suspicious of the circumstance, but we feel he is unjustified in any further assumption. Nausea and vomiting occur so infrequently that they are of little aid in the diagnosis, and their absence may be entirely set aside. They are more frequently associated with cases in which there is severe pain, and in these cases are probably open to the same interpretation as nausea and vomiting under any condition of severe abdominal pain. They are also seen, however, as manifestations of severe systemic shock, but in such cases the presence of shock is so manifest that one need not look for the confirmatory evidence of vomiting. They occur in less than 25 per cent, of the cases. Their presence is of slightly confirma- tory value only. Altered respiration is a sign of consequence, and is probably to be interpreted as muscular resistance, namely, a means of protecting the abdomen. The prevailing type of respiration where perforation has occurred is almost invariably distinctly thoracic. This type of respira- tion is also seen in severe cases of typhoid with distention and rigidity, but the phenomenon is considerably accentuated when perforation occurs. It was a very striking sign in all the Lakeside cases. ^ It is rather more evident, to be sure, in clear-minded patients with consider- able pain, assuming somewhat the quality of a symptom ; but it is sufficiently recognizable as a sign alone where the mental condition of the patient is to be entirely disregarded. The rate of respiration, also, is nearly always more or less increased, but is of largely secondary con- sequence. If the patient is in pain, respiration may be unduly increased on this account alone ; if he is insensible to pain, the restricted respira- ' We were considerably surprised to see in a recent report of eight cases of perforation, by McCrea and Mitchell (" Surgical Features of Typhoid Fever," JohnsHopkins Hospital Reports, 1902, vol. X., Nos. 6, 7, 8, and 9), that the obEervation of altered respiration had been very in- constant, and was apparently of indiflfeient significance. Our observations have been so con- stant and well borne out, however, that we feel in no way inclined to belittle the frequency or significance of the sign. BRIGGS: IJfTESTINAL PERFORATION. 11 tory excursion will necessitate somewhat more rapid breathing^. Altered respiration, then, of the costal type warrants much the same assumption as muscular abdominal resistance, and we believe this altera- tion can be found in practically all cases if sufficiently well observed. As this alteration is largely relative it is probably less reliable than muscular resistance, and as it is open to the same interpretation its practical value, aside from being merely confirmatory, is not great. Its absence, however, as in the case of muscular resistance, must be con- sidered of considerable importance. Alteration in temperature may be of some positive value, but the absence of such alteration must be set aside entirely in making a diagnosis. An accompanying chill is of rather infrequent occurrence. Two of the Lakeside cases showed no alteration in temperature, and many other reported cases are equally negative in this respect. Many cases will show a rise or fall generally not exceeding 2° F., which, when considered with the usual irregularity of the typhoid chart, must be admitted to be of little value. A very decided drop in temperature, as is sometimes seen in hemorrhage, is rather less frequently met with in perforation, we believe, than has been generally supposed. As an indication of spreading infection the value of altered temperature has long since been well established, as was mentioned also with reference to distention and tympany ; but as an indication of perforation itself the matter of altered temperature may be very largely overlooked. The puke is an indication of greatest importance. It is, of course, an evidence of systemic shock accompanying perforation, but is a matter of such consequence that it deserves special mention. One can almost invariably expect a rise in pulse that is quite marked, perhaps twenty beats or more, the alteration occurring rather suddenly. This rise in pulse is affected in a measure by the increased respiration, but is almost always too marked to be accounted for entirely on such grounds. The quality of the pulse also is markedly altered, the ten- sion being reduced, and the pulse rendered readily compressible. These strikingly distinctive changes in the rate and quality of the pulse are, of course, less strongly marked in more severely ill cases, in which the pulse is already rapid and the quality poor, but even in such cases, to one familiar with the daily condition of the patient in question, the relative alteration is evident. It may also be of some aid in differen- tiating between shock and hemorrhage, as alteration in rate and quality is likely to be more gradual in hemorrhage ; it is a distinction, however, that is by no means always recognizable, but may serve as a some- what confirmatory observation. Sudden marked alteration in rate and quality, then, is an indication of great importance. Definite knowledge regarding altered blood conditions occurring in typhoid fever is so meagre that we feel that little reliance can be 12 BEIGGS: INTESTINAL PERFORATION. placed upon it in making a diagnosis of perforation. This may appear at first sight to be a rather striking and radical statement. In the large proportion of cases it is very doubtful if the white blood count is affected by the perforation itself. It is true that in many instances a leucocy- tosis is observed to follow perforation at an interval of anywhere from one to six hours ; but, on the other hand, in a very considerable number ■of cases no such leucocytosis is observed. It has also been observed that moderately sudden and quite extreme leucocytosis occurs in patients in whom perforation was supposed to exist, when operation showed -definitely that none had occurred. One of the Lakeside cases showed -such a rise from approximately 8000 to 50,000. It has been assumed by some and considered demonstrated by others that following perfora- tion there is a " wave of leucocytosis" extending over a period of a few hours, and that in cases where no leucocytosis is observed the ■observation has been made at the close of this wave. Some cases show ithis condition, others do not. Whatever one's belief regarding it may Ibe, it is known to be a certainty that the leucocytic equilibrium in man is very easily disturbed to a greater or less extent. Some of the causes that at times effect this disturbance are known with a moderate degree of certainty, but the number of causes with which we are somewhat acquainted is probably very small as compared with the number of causes for which we are utterly unable to afford any adequate explana- tion at the present time. Slight temporary differences in the white blood count are usually beyond explanation, and while it may be assured that such temporary flights occur in perforation it is an assumption that must be based upon far wider and much more accurate observation than has been made thus far. These waves of leucocytosis have been demonstrated to exist in many conditions, and for that matter in sup- posedly normal individuals, which had they occurred in conjunction with symptoms suggesting perforation might be assumed to be due to that cause. One such set of observations has been made by Cabot, Blake, and Hubbard in an article on " A Study of the Blood and Its Relation to Surgical Diagnosis," Annals of Surgery, vol. xxxiv. p. S61. These views are based upon the observation of only ten cases, to be sure, four of which were typhoid fever, but they are as extended and certainly more accurate on the whole than any observations in favor of the leucocytic wave theory in perforation, and serve to put the subject where one is unable to attach to it much diagnostic impor- tance at the present time. What may, however, prove to be of some consequence is a steadily increasing leucocytosis reaching a well-marked maximum, which is attained after a considerably longer period than the height of the so-called wave. The interpretation of this increase, however, is largely an assumption ; it may be due to spreading infection which has not yet become sufficiently marked to cause the fall in the BKI6GS: INTESTINAL PERFORATION. 13 number of leucocytes not infrequently seen in cases of extensive peri- toneal infection. Be this as it may, it must be admitted to be of relatively small practical importance, for if these cases are to be saved one must not wait for a well-developed leucocytosis from any such cause. There is a great though apparently decreasing tendency to over- estimate the value of leucocytosis in surgical abdominal conditions at the present time, and we feel that this is an example. The irregularity noted in the blood examinations in cases of demonstrated typhoid per- foration is so great that the subject needs far wider and much more accurate study than has thus far been given. But while from the standpoint of diagnosis the question of leucocytosis must be largely set aside, from the standpoint of careful and complete observation of cases the importance of such examination cannot be overestimated. The subject may have possibilities ; before leaving it the absence of such possibilities must be thoroughly demonstrated. Alteration in the amount of hccmoglohin and the number of red blood corpuscles in these cases is a matter that has been almost entirely over- looked, but from a theoretical standpoint it would appear to be of somewhat more consequence than the white count. The most difficult cases for diagnosis are those in which the decision lies between hemor- rhage and perforation. If the diagnosis of hemorrhage is incorrectly made, and instead of this perforation has occurred, the desirable opportunity for operation is lost while waiting for the demonstration of hemorrhage through the bowel discharges. It may be possible to make use of the red count and the estimation of hsemoglobin as affected by hemorrhage to aid in the differential diagnosis in these cases. The white count can be of little aid in this differentiation. It is felt that some careful observations along this line will prove valuable. Systemic shock is probably always present, although not always recog- nized, and is an indication of the very greatest importance. The con- dition is usually so evident and its indications so readily recognizable that it is scarcely necessary to consider it in detail other than to insist on its importance. Altered respiration to a slight degree possibly, and especially alterations in pulse, temperature, and blood are indications of general systemic change. The pallid countenance, the drawn facial expression, slight cyanosis of lips and finger-nails, the sudden occur- rence of perspiration, and the slight tremor that is sometimes present, together with the altered pulse and temperature, make a very striking picture when well marked. Here, again, however, in proportion as the organism is deeply affected by the disease itself, the margin of vitality susceptible to even the severest degree of shock is narrow, and the system may be so deeply affected that it fails entirely to respond in any appreciable degree to the shock received, and the catastrophy passes unrecognized. It must be kept in mind, however, that this is 14 BEIGGS: INTESTINAL PEKFOEATION. said solely with reference to the severity of the infection, and is entirely independent of the mental condition of the patient. Shock is distinctly recognized by physical, not mental demonstrations, and if any reac- tionary power remains in the organism the condition can be recognized . in spite of delirium or apathy. This condition of systemic shock was observed in all of the Lakeside cases. In one case with severe infec- tion and moderately well-marked shock no perforation was found, and the cause of the suddenly altered condition was never demonstrated. In summary, then, the signs and symptoms upon which one is forced at the present time to rely mainly in making a diagnosis of per- foration are pain, sensitiveness, muscular resistance, altered respiration, alteration in rate and quality of the pulse, and evidences of systemic shock. Differential Diagnosis. The question of differential diagnosis is a matter upon which there is little to be said with satisfaction, owing largely to our fragmentary and quite indefinite knowledge of the causes of pain during the course of typhoid fever. Pain must be con- sidered a most important symptom, and at times the most important feature in making the diagnosis of perforation, but at the same time it is the indication which most often leads us astray. A most instructive and commendable paper by Dr. Thomas McCrae, " Abdominal Pain in Typhoid Fever," Neio York Medical Journal, May 4, 1901, vol. Ixxiii. p. 749, is perhaps the most valuable article of its kind that has occurred in the more recent literature. His report is based upon the analysis of 500 cases at the Johns Hopkins Hospital, in which it was found that only two-fifths of the patients were free from pain or ten- derness ; one-fifth showed tenderness only, while two-fifths showed pain sometime during the attack. Of this number there were only 13 cases of perforation, while in 70 cases there was no discoverable cause. It must be admitted that in a large proportion of cases in which the existence of perforation has not been demonstrated at operation, the causes of the indications leading to the diagnosis have not been discov- ered, and in instances in which the causes have been assigned it was largely an assumption. Among these causes, known or supposed, are diaphragmatic pleurisy, pneumonia, iliac thrombosis, appendicitis, peri- tonitis from undiscoverable causes, intestinal obstruction, suppurating mesenteric or retroperitoneal glands, cholecystitis, and intestinal hemorrhage. Regarding pleurisy and pneumonia, they have both been known to be associated with abdominal pain, the entire aspect of the case resem- bling very closely that of perforation. If the pneumonic process is not sufiiciently near the surface to be recognizable, one can have little hope of avoiding the mistake. A careful pulmonary examination in all cases is the only possible precaution. Such cases are, however, fortunately, very rare. Iliac thrombosis is usually associated with sen- BRIGGS: INTESTINAL PBRFOKATION. 15 sitiveness in the groin and for a varying distance along the course of the femoral vein, and a careful examination may reveal this cause of &. misleading symptom and should always be carefully kept in mind. Appendicitis occurring during the course of typhoid fever can very rarely be differentiated from perforation occurring in or near the caecum, and there is fortunately no need for so doing, as the indications for operative interference are so nearly identical. Peritonitis without discoverable cause has been mistaken for perforation, but can scarcely give the early appearances of perforation ; in either event, however, immediate interference would be indicated. Intestinal obstnietion has simulated perforation, but here, too, the instances in which it -would simulate the early symptoms of perforation must be extremely rare, while the indications for exploration would be equally certain. The considerations of consequence in connection with appendicitis, periton- itis, and obstruction with reference to perforation are not so much in relation to the difference in diagnosis, but that the conditions must be kept in mind as possible causes of the symptoms and signs in cases in which no perforation has been found. The same may be said of cases of suppurating mesenterie or retroperitoneal glands; this condition can scarcely be mistaken for perforation pure and simple, but has been mistaken for resultant indications of perforation due to spreading infection. Cholecystitis has simulated perforation very closely, as in the case reported by Dr. H. B. Allyn, " Typhoid Fever, with Perfora- tion of the Colon and Gall-bladder ; Operation ; Death ; Autopsy," Philadelphia Medical Journal, August 3, 1901, vol. viii. p. 193. In this case no perforation was found, the existence of cholecystitis was not discovered at the operation, and the cause of the misleading indi- cations was found only at autopsy. In such cases, however, the pain is usually high in the abdomen, and is very likely to be associated with slight jaundice or a trace of bile in the urine. Here, again, however, the important consideration is not the diagnosis, but the possibility of overlooking such a condition where no perforation is found. From hemorrhage, however, differential diagnosis is of the utmost importance. Less than half of the cases of hemorrhage are associated with pain, and it is probably the proportion is small in which perforation might be reasonably suspected. The diagnosis is of importance not so much because hemorrhage sometimes suggests perforation, but because the indications of perforation itself are sometimes attributed to hemorrhage, which is a mistake of the utmost gravity. Where the two conditions do suggest each other it is a diagnosis of the greatest difficulty, and in the present state of our knowledge there is relatively little to aid one. The pain and sensitiveness may be the same, the signs themselves may be almost identical. Alteration in the rate and quality of the pulse, however, may be more gradual in hemorrhage, but the difference is 16 BRIGGS: INTESTINAL PERFOBATION. purely relative and not always of assistance. The evidences of sys- temic shock are usually more severe in perforation, but this is also a relative difference and merely a suggestion. As the case progresses the diagnosis of hemorrhage is usually rendered clear by the subsequent appearance of blood in the stools, but if perforation is strongly sus- pected one cannot be justified in delaying for this confirmation of the diagnosis of hemorrhage. The reduction of haemoglobin and the number of red corpuscles, it is strongly believed, may be of material aid in differentiating these two conditions, as has been suggested by McCrae, and was of demonstrated value in one case at the Johns Hopkins Hospital. We feel that this is a most important subject, and would urge that it be widely investigated and reported for the sake of accumulated statistics. But when all is said it must be admitted that differential diagnosis between perforation and hemorrhage is sometimes extremely difficult and misleading, and in such instances one must feel justified in diagnosing perforation rather than hemorrhage, since the gravity of delay in perforation far exceeds that of exploration. It may be well to mention here several things not already notfd which may obscure or aid in the diagnosis. To one who is familiar with the not infrequent effects of tub baths, especially in severely ill cases, it would be suggested at once that the abdominal pain, muscular rigidity, and varying degrees of systemic shock sustained might lead one astray in a case where perforation was suspected or even where na such thought had been entertained previously, if the effects of tubbing were unusually marked. It is, therefore, a wise precaution in sus- pected cases to substitute sponge baths, in order that the diagnosis may not be obscured by ever so slight an alteration in the patient's condi- tion from other than the suspected cause. In one of the Lakeside cases the sudden change in the patient's condition was thought to be due to the effects of the tubbing, and caused a disastrous eight-hour delay before operation. Again, in administering morphine in cases of hemorrhage one is courting danger on one hand while attempting to forestall it on the other. It is at once evident that in cases suggesting both perforation and hemorrhage the diagnosis should be clearly defined if possible before administering morphine. There are occasionally cases, however, in which one feels obliged to give morphine regardless of the consequences. One other suggestion which we have not seen mentioned elsewhere seemed at the time to be of material aid in the diagnosis in two of the Lakeside cases, in one of which perforation had actually occurred, while in the other operation was delayed and the case recov- ered. In both of these instances the question was raised whether the rather moderate abdominal pain might not be due to flatus. A small glycerin enema administered in each instance was expelled by both patients with the evacuation of considerable gas. In the case of actual BBIGGS: INTESTINAL PERFOEATION. 17 perforation no relief from the pain was experienced, while in the case recovering without operation the pain was considerably though not entirely relieved. It is a procedure that ia scarcely applicable in cases of suspected hemorrhage, but may occasionally afford information of more or less value where perforation is suspected independently of hemorrhage. It may be permissible to consider in this connectioQ the so-called " pre-perforative stage," upon the existence of which as an absolutely recognizable condition Dr. Gushing, of Baltimore, has been so per- sistently insistent. In the Johns Hopkins Hospital Bulletin, November, 1898, No. 92, p. 267, he defines his position as follows: " Under the ' pre-perforative ' stage let it be understood that the whole period is in- cluded between the first involvement of the serosa with the customary formation of adhesions at this point, until these adhesions, which may for the time constitute the floor of the ulcer after the serosa has given way, have themselves become broken down and general extravasation has taken place." What he refers to is a pre-extravasation and not a pre-perforative stage, and this distinction the writer himself mentions in a subsequent paper. That adhesions form in all cases of perforation is probable in spite of the fact that they cannot always be demonstrated at operation. That every case of perforation has what may be called a pre-perforative or pre-extravasation stage is too evident to need mention, just as anything that occurs has a pre-existent condition ;. but when the same writer mentions it as a " definite recognizable con- dition " he is insisting on a refinement of abdominal diagnosis that is quite beyond the possibility of recognition with any degree of cer- tainty. The impossibility at the present time of recognizing a distinc- tion depending upon the thickness of the peritoneum is self-evident. An early diagnosis is the ne plus ultra in perforation, and doubtless in some of these cases one may feel satisfied he has found the condition mentioned by Dr. Gushing, but it is seldom to be hoped that one will strike the exact time of perforation with such accuracy. It is firmly believed that the best that can possibly be hoped for is the possibility of operating upon these cases within a considerably wider limit than Dr. Gushing urges, a little before or a little after the catastrophe has actually occurred, and the cases in which one is able to place his operation between the involvement of the peritoneum and actual extravasation are, we believe, purely fortuitous and not a matter of judgment. Operative Treatment. It is hardly necessary any longer to urge that the treatment of typhoid perforation is solely operative. It may not be always possible to operate, but when the diagnosis of perforation is once made one can no longer consider the patient as being under treatment if operation is not employed. So far as the patient is con- 18 BEIGG8 : INTESTINAL PERFORATION. cerned, however, it can be safely said that operation should be resorted to in all instances. This is the only reasonable ground in even the most desperate cases. The list of recoveries contains a number of the most hopeless instances of surgery as a last resort, notably the remark- able case of Dr. Abbe, New York Medical Record, January 5, 1895, in which operation was performed sixty hours after perforation, and in which two pints of extravasated fecal matter were removed from the abdominal cavity; also the case of Dr. Champlin, No. 155 of Dr. Tinker's series, published by Dr. Keen in his article already mentioned, in which the interval between perforation and operation was estimated at three days; and the case of Dr. Pearson, British. Medical Journal, 1899, vol. i. p. 1097, in which recovery followed operation nine days after perforation was thought to have occurred, although in this instance there was a localized abscess about the csecum, the case resembling more a slowly progressing appendicitis. The time lohen the operation should he performed, that is, the number of hours after perforation, is a matter upon which statistics seem strangely at variance with common sense and experience. Cases of perforation are always associated with a greater or less degree of shock. To operate during the severity of this condition can scarcely be con- sidered wise unless the period of shock is unduly prolonged. The diagnosis having been made, however, the matter of an hour or two of heat and stimulation in the severest cases will serve to get the patient in as good condition for operation as one can hope. If the systemic shock is quite inconsiderable we need not delay even this length of time. If the period of shock is prolonged in spite of treatment, we believe that the danger of spreading infection will far exceed any possible benefit derived from delaying in the hope that the patient's condition will improve. It is owing to the deplorable lack of post- mortem examinations that we are obliged to assume this position instead of accepting it from demonstrated facts. The cause of death in the vast majority of reported cases has not been accurately ascer- tained, and where stated has been merely a clinical assumption. The number of autopsies obtainable from the reports is far too small to justify any conclusions. It is probably true, however, that general peritonitis is the cause of death in the majority of fatal cases operated upon at an early date, and in a very large proportion of cases in which operation has been delayed. It must certainly be admitted that a case dying of general peritonitis following early operation could not possibly have been saved from this catastrophe by delaying a number of hours. The statistics of Dr. Westcott and Dr. Tinker, published by Dr. Keen, show that the largest percentage of recoveries follows operations per- formed during the second twelve hours after perforation. We cannot but feel, however, that these statistics, convincing as they first ap- BEIQGS: INTESTINAL PEHFOKA TION. 19 pear, are quite misleading. They are largely made up of the older cases, in most of which operation was long delayed and among which the recoveries were largely instances of fortuitous and unlooked-for good fortune. The fact, also, already mentioned, that the cause of death in a.11 but a few of these cases was never definitely known affords a possi- bility of error that renders the statistics practically useless. In the more recent cases recovery has been much more frequent among those in which operation was early performed. But even here, again, unless sufficient interest is exercised to demonstrate positively the cause of death in unsuccessful cases, the decision of this question must be largely a matter of general experience and not scientific demonstration. The nature of the ancBsthetie will often prove to be a matter of con- sequence, especially whether one will employ local or general anaesthesia. There are frequently serious objections to both, but we cannot at all subscribe to the feeling of certain writers who claim that local anaes- thesia should be employed in all cases, and that the use of general ansesthesia is a deplorable mistake. By the employment of local anses- thesia it has been thought to obviate entirely the additional shock of the administration of a general anaesthetic. In some instances this is so, in others it is purely an assumption. It is too often the case that what may be spoken of as purely operative shock is largely attributed to the use of a general anaesthetic, and no one will maintain that lapar- otomy under cocaine, in which the abdominal contents are disturbed, is entirely relieved of shock by the omission of a general anaesthetic. The mental anxiety, worry, and dread incident to a laparotomy under local anaesthesia in a hardy, strong individual, or even in a patient whose sensibilities are dulled to no inconsiderable degree by the severity of the illness, is a consideration never to be overlooked and will not infrequently render the decision against local anaesthesia. What is to be accomplished is the greatest possible reduction of sys- temic shock to the patient incident to the operation, and in some instances this is best accomplished by local anaesthesia, in others by general anaesthesia. Stolid and more or less apathetic patients are more favorable for local anaesthesia. The amount of self-control an individual is able to maintain is a very uncertain indication, as the greatest self-control may sometimes be associated with extreme and very depressing mental strain. Even in those cases in which a local anaesthetic has been properly and successfully employed, it is doubtful if the extensive and careful abdominal irrigation usually required, can be as well accomplished in most instances without resorting to a general anaesthetic at this stage of the operation. As so much depends on the care and thoroughness of this part of the procedure, it does not seem always wise to complete it under local anaesthesia unless the indications are very direct. 20 BRIGGS: INTESTINAL PEEFOKATION. One of the maia objections to general aafesthesia is the fact tliat so- much time is usually lost in getting the patient sufficiently under. This may be obviated by using chloroform after the administration of about one-quarter grain of morphine hypoderniically ; but what we feel will eventually prove a still more rapid and considerably safer method is the employment of nitrous oxide and oxygen followed by ether. By using the gas and ether alternately in getting the patient under, com- plete anaesthesia can be produced readily in one to two minutes, and the method is devoid of the unforeseen but ever-present dangers incident to chloroform. The prolonged use of nitrous oxide and oxygen alone may eventually displace the use of chloroform and ether altogether in these cases, but at the present time its usefulness and safety for pro- longed ansesthesia has not been sufficiently demonstrated. When gen- eral anaesthesia is employed it is unnecessary to add that it should be given into the hands of one of very considerable experience, and the administration should be as light and as short as possible. On& advantage of the general antesthetic thus administered over local anses- thesia, which we feel is not always kept in mind, is the greater rapidity with which the operation can be done under general ansesthesia, and the considerable reduction of the exposure and handling of the patient, to say nothing of the added thoroughness with which the operation may be performed. The use of local ansesthesia in these cases is an extremely simple matter to one who has even a limited abdominal experience, but its usefulness and employment is something which should be carefully considered and not entirely assumed. The requirements of operation are so direct and clear that the technique is simple, and the steps of the operation well defined and logical. As an operative procedure it has been so well developed that we caa look for comparatively little improvement in this direction, and there is consequently much less of interest in this portion of the discus- sion. The incision in the right lower quadrant of the abdomen is the only logical opening in these cases, and the linea semilunaris opposite or a little above the anterior superior spine we believe to be preferable. The lower part of the ileum, the csecum, and the appendix include the vast majority of perforations, and are, on the whole, most readily accessible through the incision mentioned. An additional advantage of this incision is the well-recognized fact that the search for perfora- tion is best begun from the csecum. The incision should be sufficiently generous, perhaps 8 to 10 cm. long, to allow a careful and somewhat visual exploration. It occasionally happens that the perforation is completely protected by well-formed adhesions and fecal extravasations prevented, as was the case in an operation performed by Dr. J. C. Warren, reported by Dr. K. B. Greenough, Boston Medical and Surgical Journal, May 8, 1902, vol. cxlvi. p. 491. If these adhesions BRIGGS: INTESTINAL PEEFOBATION. 21 are blindly broken up through a small incision unnecessary fecal extravasation may occur, the danger of which, owing to the certainty of the infection, cannot be estimated. After entering the abdomen the first search for the perforation should be made in the ileum immediately above the csecum, as the vast majority of perforations occur in this portion of the gut. By far the best way of accomplishing this is by starting at the csecum, a fixed point very quickly found. We feel that this is a very wise routine pro- cedure to employ, since one knows accurately the exact location of the small intestine he is handling and avoids the discouraging loss of time almost invariably incident to a haphazard and undiscriminating inspec- tion of coils of small intestine appearing in the wound. Starting with the csecum, one should inspect the ileum for a distance of about one metre. There is no object in going above this unless one intends to search the entire intestinal tract. This search for the perforation can be hurriedly but carefully conducted if the incision is sufficiently large to afford even a limited view of the general field. It is usually con- sidered wise to replace the coils of intestine within the abdomen as the search is made, although some feel, and not without reason, that keep- ing the explored portion of the intestine well covered with hot towels and finally replacing all the coils at once can be accomplished with less handling of the intestines and less incident shock to the patient. We feel it is wise to make this first hurried search of the ileum for the perforation only, which is immediately recognized if present, a more careful examination of the gut for very thin ulcers being made after the really serious features are corrected, providing the condition of the patient will permit. It also affords a comprehensive conception of the portion of the bowel almost invariably affected, and prevents the possibility of spending valuable time over relatively unimportant conditions. It is, therefore, well to make this entire rapid inspection of the bowel even before closing any perforation that may be found. It is a very simple matter for an assistant to retain the location of these perforations outside the abdomen with sponges. The method of closure of the perforation is relatively unimportant, but should always be done with silk ; catgut alone has been demonstrated on several occasions to be insufficient. A short double row of continuous Lem- bert sutures is all that is required, and can be very rapidly placed without elaboration. Excising the ulcer before suturing is an utterly useless and time-consuming procedure. It is a rare exception to find an ulcer of such size that the suture encroaches to an important degree upon the lumen of the gut ; in such instances the suture must be placed with greater care. If there are several neighboring or con- fluent perforations, or the integrity of a considerable portion of the gut is severely impaired by deep ulcerations not yet perforated, it is 22 BEIGGS: INTESTINAL PEEFOEATION. wise to secure this portion of the intestine to the edge of the wound rather than consume time by resection or an elaborate method of suture, and to close the fistula subsequently.' Having secured the perforation, if the condition of the patient permits, it is always wise to make a retrograde search of this last metre of the ileum for deep ulcerations threatening perforation. These can be recognized with con- siderable accuracy by the trained touch. Running the finger down over the bowel the dangerous areas are indicated by small, soft, circular places in the centre of a moderately large, thickened area. The sen- sation is readily acquired by post-mortem examination of an ulcerated ileum, iu which the observation can be verified by subsequently opening the bowel. These areas when found should be reinforced by a run- ning silk Lembert suture. The caecum and appendix being immedi- ately at hand, it is well to give them a passing inspection. One is scarcely justified, however, in removing the appendix unless it is seriously affected, and not in instances where it is " at all abnormal," as was recommended by Dr. Finney, in his article in the Johns Hop- kins Hospital Reports. If no perforation has been found in the search so far instituted, it is inadvisable to make further examination of the intestinal tract, excepting, possibly, the sigmoid flexure, which is occa- sionally, though quite rarely, affected. Typhoid perforations in other locations are so extremely rare that one would lose many more patients in a prolonged intestinal search than in running the slight risk of having overlooked such a condition. Irrigation of the abdominal cavity we feel should be somewhat pro- longed and very thorough, even if there are no visible evidences of peritonitis or extravasation. The stimulating usefulness of hot irriga- tion has been even urged in explorations where no perforation has been found, but at the present time one hardly feels inclined to irrigate in such cases. In many instances this abdominal irrigation should be deferred until the inspection of the gut has been made, unless the inspection is seriously hindered by the amount or nature of the fluid present, in which case a rapid preliminary irrigation may be useful, the main irrigation, however, to be reserved until after the » Since the writing of this paper we have had occasion to feel that certain cases of rather widely distributed severe lesions may be suitable for resection. A recent case of the Lakeside series, not yet published, presented at operation two perforations and nine deep ulcers dis- tributed over 40 cm. of the ileum. Two perforations and three ulcers threatening perforation were sutured, and a chance taken on the remainder, although the gravity of so doing was recognized. The patient died five days later, primarily from pneumonia. No further perfora- tions were found, but two of the unsutured ulcers were saved from actual extravasation only by light adhesions. The presence of a slight beginning peritonitis not over a few houra old was apparently due to infection through the thin adhesions covering these two ulcers, and doubtless contributed to and would certainly later have caused the death of the patient. We feel that the patient would have stood resection of the affected area and anastomosis with a Murphy button, and on reviewing the case it is a source of deep regret that the procedure was not employed. We know, however, of no attempt at resection up to the present time. BEIGGS: INTESTINAL PERFORATION. '23 treatment of tlie bowel. Salt solution is possibly preferable to sterile water, but either one will do. It should be about 115° F., copious in amount. This, we feel, is best and most wisely accomplished through one or several irrigating tubes without removal of the intestines, and should be especially thorough in the region of the pelvis and lumbar fossae. It is often well to irrigate the pelvis first, introducing a large rubber tube which will take up a large part of the return flow ; in this way one will not infrequently avoid general dissemination of a large proportion of the abdominal fluid and fecal extravasation which very frequently settles to the pelvis. If extensive peritonitis is already present, one must treat this complicating incident as such, and that will vary greatly according to individual belief. We feel, however, that the extensive intestinal, manipulation sometimes advised in such cases is of doubtful wisdom unless the condition of the patient is unusually favorable. We would feel inclined to rely on thorough and prolonged irrigation of the abdomen without removal of the intestines, as this is about all such patients can endure. In many instances, to properly accomplish this irrigation to the best advantage, we feel that a general anaesthetic is advisable, although it is occasionally satisfac- torily accomplished under local anaesthesia. It may seem occasionally wise to close these cases without drainage, but such has certainly been felt to be the rare exception. It is possible that having the temerity to close these cases after thorough irriga- tion may add considerably to the recovery rate ; such cases, however, should certainly be restricted to those in which examination of the abdominal fluid during the operation shows the absence of organisms, or, possibly, those in which only bacilli are found to be present ; but the advisability of drainage is scarcely questioned at the present time. The way in which one drains will vary according to the operator's personal belief. Whatever the method, however, it seems advisable to drain the pelvis. In the Lakeside cases what proved to be a very satis- factory method was by means of glass drainage tubes leading to the pelvis, and along the side of the tubes in some instances gauze drainage was also placed. It seems wise, also, to leave the seat of perforation or doubtful portions of the bowel near the incision, introducing gauze drainage to this portion of the bowel in order, if possible, to forestall the disastrous consequences of insecure suture or subsequent perforation. Around the drainage the wound should be closed as completely as pos- sible, as the granulating capacity of such patients is sometimes consid- erably reduced, although the healing of wounds by first intention seems to be rapid and secure. While it is needless to say that the operation should be done as rapidly as possible, it must still be remembered that these patients bear operation much better than one would ordinarily he led to suppose who 24 BUIGGS: INTESTINAL PERFORATION. had not had the opportunity for personal observation. The incision, the search for perforation, the repair of the bowel, can all be very rapidly and at the same time thoroughly performed, so that this por- tion of the operation may well be considered the minor part. The careful and thorough irrigation, however, is a matter that cannot be unduly hurried, although here the time can be somewhat diminished by having several irrigating tubes in use at the same time. This por tion of the operation, while possibly consuming the most time, is the least depressing to the patient, as it is associated with a considerable degree of stimulation. We fear not a few cases are sacrificed through lack of proper irrigation, owing to the overwhelming desire of operators to complete the operation in as short a time as possible. AVe would urge the desirability of the utmost speed during the really operative part of the procedure, which is easily possible for any operator of moderate dexterity and good judgment, but would deprecate hurried and inadequate cleansing of the abdominal cavity. The superior ease and rapidity with which one can accomplish the operation under a genera] ansBsthetic is often in marked contrast to the slow, painstaking procedure not infrequently required under local ausesthesia where the sensibilities of the patient are so keenly alive to what is going on, and this consideration, as previously mentioned, will not infrequently incline the operator to the use of general anaesthesia as of marked advantage. The subsequent treatment of these cases presents nothing distinctive or striking, it only being necessary to keep in mind the requirements of the patient from the standpoint of a laparotomy on the one hand and that of typhoid infection on the other. While we do not mean to sug- gest that the treatment of such cases does not require the greatest care and judgment, it presents no distinctive feature which need be dwelt upon. One thing, however, is extremely desirable, namely, to keep constantly in mind the possibility of a subsequent perforation, and avoid as far as one is able the possibility of obscuring the manifesta- tions of this occurrence in any way by the line of treatment. The question of exploratory incision in rather doubtful cases of sus- pected perforation we feel is a matter of great importance and increas- ing interest. This has been most advantageously urged and considered by Dr. H. W. Gushing, and is, we feel, his most useful contribution to the surgery of these eases. It has been exceptionally well expressed in his article, " Sur la Laparotomie Exploratrice Pr^coce dans la Per- foration Intestinale au cours de la Fidvre Typhoide," Archives Gen- erales de Medecine, January, 1901. In instances where perforation is suspected, but the diagnosis cannot be considered as definitely made, exploration affords the only hope of avoiding the catastrophe of allow- ing a perforation to go unrecognized until more or less general infection BRIGG8: INTESTINAL PERFOKATION. 25 has occurred. Even at the present time it has been proven beyond a reasonable question that these operations when properly performed are practically devoid of danger and affect the condition and course of the patient to a scarcely appreciable degree, whether performed under local or light and rapid general anaesthesia. If no perforation is found the abdomen can be immediately closed, a collodion dressing applied and the baths resumed within a few hours, so that treatment of the patient need scarcely be interrupted. Such was the case in No. 4 of the Lakeside series. The impervious dressing necessary for continued tub- bing is conveniently applied by using six or eight alternate layers of collodion and sheet wadding, each layer being a little larger than the one previously applied. It is well to use only half of the thickness of the ordinary sheet wadding. Exploration was first suggested by Dr. Finney, who advised making a very minute incision for the purpose of taking cultures from the abdominal cavity. This method was employed in the Lakeside case mentioned, but the restricted nature of this exploration was considered a mistake, and can certainly be no longer recommended. One must now be able to detect more than the presence of free abdominal fluid or organisms within the abdominal cavity. The perforation may be protected by adhesions or the ulceration may be scarcely through the peritoneum, either of which conditions may occur without being dis- covered by such restricted exploration. If one is to look at all he should examine the ileum and caecum with care ; it will take scarcely more time and will afford information of consequence. It must be readily admitted that the very slight danger incident to a rapid, skilful exploration is not to be compared with the serious consequences of an unrecognized perforation. Considering this as an aid to diagnosis,- it is largely along this line that we are to look for a considerably larger percentage of recoveries from perforation. It will necessarily result occasionally in a premature operation, but will add very materially to the number of lives saved. In instances where no perforation is found, either with or without evidences of peritoneal infection, at once arises the question of how far one may feel justified in exploring the abdomen for such other intra- abdominal conditions as have been known to suggest perforation. This is a matter which will be decided entirely by the individual judgment of the operator, and must be affected to a greater or less degree by the condition of the patient. Where, however, one finds an abnormal intra-abdominal condition, as peritonitis, for instance, it seems hardly justifiable to limit the search for the seat of the trouble to a rapid glance at the ileum and caecum. The appendix is so close at hand that its inspection should never be omitted. The detection of suppurat- ing mesenteric glands is important for the sake of adequate disposition 26 BRIGGS: INTESTINAL PERFORATION. of drainage. Exploration of the region of the gall-bladder may be accomplished without serious objection. The presence of intestinal obstruction may possibly be detected by a hurried examination of the intestine from several parts of the abdomen. Unless the indications are pretty direct, however, one would scarcely feel inclined to risk much time on this very remote possibility. The extent to which these explorations are to be pushed is a matter for much thoughtful consid- eration. The advisability of secondary operation for subsequent perforation or other operative indications is no longer a matter of general question, but of individual judgment in each case. The remarkable vitality displayed in a number of reported cases in which two or more subse- quent operations were performed, has established this point beyond reasonable question. Operative indications are to be followed, and can no longer be set aside, for the simple reason that the patient has already undergone one operation. Prognosis. At the present time it is extremely difficult to form even a reasonable idea of the prognosis in these cases. We know that it is affected by a number of conditions. In general, the recovery rate is higher among young people. The time during the disease at which perforation occurs has an influence, since the recuperative powers of the individual are altered at different times. Recovery is less fre- quently to be expected when perforation occurs, during the height and severity of the disease, whether of a primary attack or in the midst of a severe recurrence. When perforation occurs early in the disease, before the vitality of the patient has been drawn upon to any consider- able degree, the prognosis in general is better, and also during conval- escence after the recuperative powers of the individual have begun to assert an advantage. The prognosis is always severely affected, of course, by the length of time the case has been allowed to go after perforation, the extent to which a spreading abdominal infection has occurred, and the bacterial nature of this infection. It is also largely affected by the judgment, skill, and dexterity of the operator. But when it comes to an estimation of the percentage of recoveries the figures obtainable from the present statistics are, we believe, absolutely unreliable. In Dr. Westcott's table, comprising the first 83 cases reported, the recovery rate was 19.3 per cent., while in Dr. Tinker's table, comprising, presumably, the next 75 cases reported, the recovery rate was 28 per cent., making an average recovery rate for the entire 158 cases of 23.41 per cent. This recovery rate is far below that obtainable under the present conditions of diagnosis and operative technique ; but the prognostic suggestions afforded by even the most recent cases which the writer has carefully collected are, we feel, too inaccurate and undiscriminating to deserve mention. This analysis BKIGGS: INTESTINAL PEEFOEATION. 27 affords a certain percentage of deaths associated with operation, to be sure, but many of the cases dying after operation bear no relation whatsoever to the true mortality in direct connection with either the perforation itself or the operation. This last statement has been most satisfactorily demonstrated by several post-mortem examinations in the Lakeside series. With but rare exceptions it has been impossible to draw this distinction in our study of recent cases owing to the incom- pleteness of the reports, but we are sure such a discrimination would alter materially our statistical mortality. Dr. Osier has divided these cases into three classes: those which recover from operation, those which die from causes immediately related to perforation and operation, and those which die from causes entirely unrelated to either. It is of the utmost importance that this distinction be kept closely in mind in forming an estimate of the mortality in these operations, but it is a distinction which, curiously enough, seems to have almost entirely escaped the thought of those attempting to form a statistical opinion of recovery and mortality. It is absolutely impossible to acquire the knowledge necessary to form this more accurate opinion without wide and accurate pathological observation and careful post-mortem examina- tions. An attempt was made by the writer to form such an estimate irom the number of autopsies reported, but the number at present is so small and the reports are so fragmentary and inaccurate that the analysis has proved utterly worthless. Unless the operators under whose observation these cases are occurring can contribute generously to our pathological knowledge, any further reports are practically useless and might much better be omitted, since superficial reports only burden the literature without contributing to our present knowl- edge. Dr. Keen estimates that we can look forward to 30 per cent, of recoveries. The recovery rate at the Johns Hopkins Hospital of cases occurring while under treatment in the wards is 45.4 per cent. Dr. Osier estimated a possible recovery rate of 50 per cent., and Dr. Gush- ing of 50 to 60 per cent. We cannot but believe that a discriminating analysis in well-observed cases with complete pathological examina- tions would give a recovery rate considerably in excess of this, possi- bly 70 per cent. , Pathological Considerations. The nature of the abdominal infection in these cases is purely fortuitous, being drawn from the wealth of bacterial flora of the ileum, the possibilities of which have probably never yet been exhausted. The proportion of cases in which bacterial examination of the abdominal fluid has been made is very small, but such examinations have been more frequent in recent cases as their usefulness and prognostic value have become more evident. The infections as a rule, however, appear to be mild, largely bacillary, bacillus coli communis and bacillus mucosus capsulatus being the most 28 BEIGGS: INTESTINAL PEEFOKATIOJST. frequent. One strange incident of a pure infection of staphylococcus pyogenes aureus has been reported. Some half-dozen cases of infection with bacillus typhosus have been collected, from which it appears that these cases are attended with a high mortality. It will be necessary, however, to have a large number of cases in which a complete and careful bacterial examination has been made to give these results any- thing more than a passing interest. It is very desirable that our knowledge of the nature of these infections be widely increased, as it may eventually be found that certain features in the treatment can be- altered to advantage according to the nature of the infection. It is possible that certain infections which are recognizable from cover-slip examinations will require very much less thorough irrigation, for instance, and that in some cases we are at present subjecting the patient to a considerably more severe strain than the circumstances may require. Cover-slips and cultures should be freely taken, and as these procedures require but a moment they need scarcely interfere with the- rapidity of the operation. These examinations should be made not only from the fluid first encountered on opening the abdomen, but also from the lower and more dependent levels, as it sometimes happens that the upper part of the abdominal fluid appears to be free from organisms, and one is led to the erroneous conclusion that no bacteria] infection has occurred. It is very desirable, also, to make an examina- tion from the site of the perforation itself, and this, of course, is- especially necessary in early cases where no appreciable amount of abdominal fluid is present. It has been thoroughly demonstrated pathologically, and in several instances has apparently been clinically confirmed, that infection of the peritoneal cavity may occur through the base of an ulcer without perforation, or even through the inflamed portion of the intestine itself without deep ulceration. Such a case was reported by Dr. loison,. " Du Traitement du Chirurgie de la P^ritonite SuppurSe Diffuse," Revue de Chirurgie, February 10, 1901, No. 2, where no solution in continuity of the bowel was found either at operation or autopsy. It is extremely probable, also, that such was the condition in the case of Dr. Dandridge, No. 28 of the Westcott series, in which no perforation was found, but the abdomen contained gas and free pus. The patient was irrigated and drained, and recovery followed. Such instances,, however, are quite unusual, and it is hardly safe to assume that the case is of this variety, simply because no perforation is found in the- rapid search of the ileum. It is, we feel, rather safer to assume that infection has occurred from some-other source, which should be found and corrected as circumstances may permit. It is probable that adhesions more or less extensive occur in all of these cases and that the infrequency with which they are reported is BEIGG8: INTESTINAL PEKFOEATION. 29 ■due to inaccurate observation necessitated by a very restricted incision through which the bowel has been drawn before the examination was made. As mentioned above, this procedure does not seem wise, since adhesions forming a perfect protection for the peritoneal cavity against infection maybe blindly broken up and the opportunity offered' for general infection. It is not meant in recommending a rather more generous incision that there is a necessity for actual visual observation of the entire field, but that by this means sufficient care can be exer- cised in the withdrawal of intestine to prevent fecal extravasation and infection where such has not yet occurred. In cases of what appear to be sudden perforation it is doubtful if these adhesions are of sufficient •consequence to afford much protection. In the Lakeside cases, while no adhesions were found in those operated upon for perforation itself, it was probably true that they were overlooked through insufficient oare in the search for the lesion. With reference to the great practical value of complete postmortem ■examinations in these cases, too much cannot be said. The greatest interest thus far seems to have centred in reporting certain clinical features and the circumstances of recovery, and from the small ■number and paucity of the postmortem examinations it is evident that very little interest has been taken in this part of the subject. It is perfectly useless to continue reporting simple recoveries for the sake -of apparently increasing the recovery rate. What is most essential at the present time is a large collection of well-observed and well-reported autopsies. A complete examination, however, does not necessarily mean a complete autopsy. Through even a restricted incision it is pos- sible to gain a fairly accurate conception of the extent and nature of the peritoneal infection if present ; the appearance of the bowel for the sake of comparison with that observed at operation ; the presence of additional perforations, overlooked or subsequent, and the existence ■of threatening perforations ; the degree of success with which the sutur- ing has been accomplished, and other respects in which the operative technique has been unsuccessful. Cultures may also be widely taken -through such an incision, and the spleen, heart, and lungs examined. Blood cultures should also be taken. It is only in this way that we are going to be able to arrive at some definite conclusions as to the •cause of death in these eases and the respects in which the treatment has been unsuccessful, and it is only by these means that the mistakes that are constantly being made in every case of perforation, either in ■diagnosis or treatment, can be rectified. For the sake of the present patients the clinical features and the operation are of sole consequence ; the interest of future patients, for the present at least, lies almost ■entirely in pathological examinations. In the apparent hope of giving some general direction to the clinical 30 BRIGGS: INTESTINAL PEEFOEATION. features of these cases, Dr. Osier has published the specific instructions which are followed in the medical wards at the Johns Hopkins Hos- pital, " Perforation and Perforative Peritonitis in Typhoid Fever," Philadelphia Medical Journal, January 19, 1901, vol. vii. p. 116. This is as admirable a schedule as can well be imagined, and could most profitably be adopted for the sake of complete and uniform observation. To this schedule we would feel inclined, from an operative and pathological standpoint, to add certain suggestions which may be an aid in determining with more accuracy some of the unrecognized features mentioned in connection with these cases. Doubtful clinical features of so much consequence in establishing the all-important early diagnosis are covered by the Hopkins schedule. For the sake of a general and more accurate idea of the amount of systemic depression associated with the ansesthetic, the operation, or lioth, it is well to note with as minute accuracy as possible the condition of the patient imme- diately before and immediately following operation, as well as the changes in his condition for at least twenty-four hours subsequent to the operation. In taking cover-slips and cultures these should be obtained from various parts of the abdomen and also from the upper and lower portions of the abdominal fluid if present, and from the site of perforation. It is desirable to note with much care the general con- dition of the bowel, the nature and extent of adhesions, and the extent and severity of the general infection if present, for the sake of com- parison with post-mortem examination in case the patient does not recover. In the post-mortem examinations, whether complete or re- stricted, the condition of the bowel and the extent of infection should be carefully noted ; cover-slips and cultures should be widely taken from the abdomen, organs, and blood ; the condition of the operative field in the bowel should be carefully observed, also the presence of additional perforations, either those overlooked during the operation or those apparently subsequent to the operation, and the presence of threatening perforations. These are all points of special consequence, which, it is believed, will lead to an accumulation of facts of much practical value if carefully and generally observed. It is necessary to make so many of these post-mortem examinations without the direction of a trained pathologist that the circumstance seems to justify the men- tion of these details. It is from well-equipped hospitals that we must first look for such a class of reports, as elsewhere it is scarcely possible to have both medical and surgical observation of all typhoid cases, and especially of all suspicious cases. A means which will add very materially to the unity and usefulness of these observations and reports, and which will tend to develop a degree of judgment in diagnosis and operation that cannot possibly be obtained in any other way, may be afforded by having all these cases brought under the immediate control BEIGGS: INTESTINAL PERFORATION. 31 of one member of the hospital staff. The number of perforations occurring during a year in any one institution is not large, and if the opportunities for observation and treatment are divided among a number of men the experience gained by each is very limited, the observations are very apt to be restricted and unmethodical, and the combined judgment rather biased and uncertain according to the nature of the particular case or two which has come to the hand of any one man. We feel that if this suggestion could be carried out our knowledge of this subject would be widely, rapidly, and very bene- ficially increased. While, however, it is true that a large portion of these operations will still be done in hospitals, mainly on account of the difficulties in diagnosis, it is still interesting to note that there has been reported by a general practitioner one successful case in a country house under the conditions which one finds there ordinarily existent. The case was reported by Dr. R. T. Davis, "Perforation in Typhoid Fever; Operation ; Recovery," American Medicine, January 18, 1902, vol. iii. p. 116. The operation was performed only six hours after perfora- tion was thought to have occurred, which betokens a degree of watch- fulness that one can scarcely hope to obtain with much frequency under such conditions. As this is the first case of its kind reported it is of special interest in showing the changing attitude of the medical profession in general with reference to perforation during the course of typhoid fever. [Reprinted from American Medicine, Vol. VI, No. 19, pages 744-745, November?, 1903.1 CASE OF POSTTYPHOIDAL INFECTION OF A RIB. BT ALFRED IRVING LUDLOW, M.D., of Cleveland, Ohio. Second Resident Surgeon, Lakeside Hospital. Many cases of posttyphoidal infection of the ribs have been reported. Keen ^ published a list of 40 dis- eased ribs following typhoid. Horsley^ has tabulated a list of 48 cases, and adds, that "Many reports are in such a fragmentary state that they cannot be used for tabulation." The following case, occurring in the service of Dr. Dudley P. Allen at Lakeside Hospital, is presented with the hope that it may be of value in future tabulations. Dr. Henry S. Upson has kindly allowed me to use the medical history of the case. N. P., a Hebrew, male, aged 37, married, was admitted August 23, 1902, complaining of headache, backache, and fever. His family history is good. His personal history shows he had pertussis and measles in childhood, and gonorrhea 13 years ago. Five years ago he was operated upon for fistula in auo. He had malarial fever 4 years ago. His present illness began one week before admission. Pre- vious to this time he was well. The initial symptoms were chills, sweats, headache, and general weakness, followed in 2 or 3 days by loss of appetite and inability to be about his usual work. On examination the patient was seen to be well-nour- ished. Mucous membrane was of good color. The tongue was dry and heavily coated. Pulse was regular, of fair volume, not dicrotic. Heart and lungs were negative. Abdomen soft, not tender. Spleen was not palpable. Several well-marked rose- spots were seen on the abdomen. Temperature was 104.6° F., pulse 90, respirations 25. The urine was amber colored, acid, specific gravity 1,022, no sugar, a trace of albumin, a few hyaline casts and leukocytes. No diazo reaction. Examination of the blood was made at various times as follows : August 25 : White blood count, 8,250. Widal negative. August 28 : White blood count, 9,400. Widal negative. August 31 : Test of patient's serum with the paracolon 1 Surgical Complications and Sequels of Typhoid Fever. 2 Annals of Surgery, February, 1903. bacillus gave a negative result. Widal positive (1 to 10 and 1 to 50 dilution). For a week irom the time of admission the patient's tem- perature showed a daily variation between 102° to 104° P. Dur- ing the second and third weelis there was a gradual decline of the temperature until the fourth week, when it became normal and continued so during the period of convalescence. The patient passed through an apparently uncomplicated course of typhoid fever and having gained rapidly in weight and strength was discharged October 11, 1902. The subsequent history is as follows : The patient said he experienced a slight pain in the region of the left lower ribs a few days after he left the hospital (Octo- ber 11, 1902), and he noticed a slight swelling over the sternal end of the seventh rib. About November 1, 1902, the swelling opened spontaneously and he came to the dispensary for treatment. The opening was enlarged and a small amount of pus evacuated. The patient continued to come to the dispensary for dressings. The wound, however, did not entirely heal, continuing to discharge consid- erably and causing severe pain at times. On January 16, 1903, he was admitted to the surgical ward of the hospital. On examination a sinus, extending down to bone, was found on the sternal end of the seventh rib. Its opening was surrounded by a red, swollen area 3 cm. in diameter. On January 20, under ether anesthesia, an incision 8 cm. in length was made over the sinus along the line of the seventh rib. About J dram of pus was obtained. The tract of the sinus was excised and the necrotic tissue about the rib, includ- ing some of the structure of the rib, was thoroughly cureted. The wound was packed with iodoform gauze. In addition to this the patient also had an ischiorectal abscess, which was also incised and drained. Cultures from this abscess showed Bacillus coli communis. The wound over the rib was repacked dally and continued to granulate. On February 3, 1903, 2 weeks after operation, the patient was referred to the dispensary for further dressings, as there was still a slight discharge from the wound. BAOTEKIOIiOGIC KEPOBT. A coverslip made from the smear showed a considerable amount of debris containing many pus cells, among which were a few small bacilli. A stain for tubercle bacilli was negative. Cultures were made upon the following media : Qlycerin-agar. — Whitish semitransparent growth upon the surface and a faint growth along the line of stab. Glucose-agar. — A growth similar to that on glycerin-agar. No gas formation. Potato. — The surface has a moist appearance but there is no visible growth. Milk.— Slightly acid. No coagulation. Oelatin and Blood-serum,. — No liquefaction. Bouillon. — A uniform cloudiness. Indol reaction negative. The bouillon culture was used in the agglutination test (2). Coverslips from these cultures show a very actively motile bacillus varying in form, according to the medium on which it was developed, staining with the anilin dyes but not with Gram's stain. N s * 1 . T ""i" ^^" "7" ^MH "^"' ■k < _ -t- 1 1 , 5 4- ■-)-• --:— - -^- ..[.. •-;-- -!-- ""("■ "> -'.- e*> ..;— --■- -i- -J -■:- -:- ..;.. --;— '^^ -;- :^ X s ■4- ._;.. ..;-. --:- -!.. ~^- «- — :. ;i -;-■■ •-:— ..:.. -+- -■'. — ■-f" --;-- ■-'. — < ... 1 — r ■ 5 •< • 1 " t 1 J j 1 ■ j i" ■ '*^ ■ » ^ — :— 1 --I-- -I-. ..'_. . .'. _ i -:.- 4^- - —I— <^ --:-- •-i- .-!-. ._!-. ,..;.. "!- 1 -.;_. -4e =+? :f; - -'■ ^ N ^ ; : „.t ..'-. --*— -^ ^., -•■ "tl , , ; 1 - 1 i s _L, -I-- --'— 1 -i-- "1" ■f- -;- -(-- -'< - -;-| 1 •^ ^ _.!_. T-;— -•:-- * 'i" .... — !-■ -■!— «s; ^ - "'ii ^ •> ■ -!■- „•.. - -; — ■i" ■ -!- --!- -!-- -:.. S^ --:— ... -U r In v. ..J.. --i— --:— -i- -1- -!-- -!- -J!< -:-■ ..■_. . ... -H N !i -4-- -i- -:— i --U -i" -i-- -i- *^ .J.J . .. 1 > • -!-- -:-- -•!— 1 i "T* -■'(- --i— • -i< ^ -■i" 1 ... —^ 5s ^ "S -->■ ■r ••;-- "!-- -f- -■!- •■!< ^ •-;-- ..... •-; — ■•■ 1 ft ^ , • -V- .J.. -■f— ~i"' ,.;— --:-« ^ ^ -4- 1 ... -i- t ^ 5 --i— — ;— ■-!- --:-- "h S ' -v~ ..;.. _.j. .. - ..;.. s :i _.'_. -;- 1 --■-1 -.•.. -.!.« •+• -■!- 1 ..... ... ..;.. > -r- -i-- -;-■ ..;. . .L. --;-' --i-« ..'-_ ..;.. 1 ... 1 ^ ^ 5S < ! 1 "tl '^ -■-- _.;. . 1 ... --;-■ ▼h V, s» . 1 1 J ; <; -r- J_. -f- .f.. ... -;- \ ! ^ (i. -i" — r- -;- — ^ - -;-- --;- -K 1 — ,— -;-- -!-- -I--- • ■ -i- i N -;-- 'f" ...-- -:-- -J— ^ ^ _.;_. 1 ..:.. . . .. .J.. 1 el 1 S 2 1 S 2 2'2 2 I *g"S 1 Agglutination Tests.— In all the tests a time limit of 1 hour was used and the test was not called positive unless there was good clumping and cessation of all motility. Serum was obtained from the patient February 16, 1903, and gave the fol- lowing reactions : L. With a known typhoid bacillus : 1-10 Positive 10 minutes. 1-.50 ■' •■■ *-^ 1-100 Negative— some motility after 12 hours. 2. With the bacillus above mentioned (bouillon culture 18 hours) : 1-10 Positive ■ 5 minutes. 1-50 " 10 1-100 ■' 15 1-200 " 20 1-300 " 30 1-500 Small clumps, but slight motility. 1-750 ' 3. Serum from a typhoid patient agglutinated the same bacillus (2) as follows: l-'iO Positive 15 minutes. 1-100 " 20 l-iOO " 25 1-300 " 45 " 1-500 Small clumps, but still motile after 2 hours. It will be seen by the foregoing that the organism is Sacil- lu.i typhi abdominalis or Eberth's bacillus. PATHOLOGtIC KEPORT. The curetings were hardened in Orth's fluid and prepared by the celloidin method for microscopic study. The section shows a few bone trabeculas, the matrix of which takes a deep pink staia and contains numerous bone cells. One trabecula shows a light pink homogeneously staining matrix, within which are a few cartilage cells. The structure about the trabec- ulas is made up of a framework of connective tissue which. In some places, particularly between the marrow cells, is of a delicate type. The larger portion, however, appears as large bands of fibroblasts infiltrating and taking the place of the marrow tissue. Here and there are areas of marrow tissue showing cells of many varieties, including lymphocytes, plasma cells, mononuclear and polymorphonuclear eosinophiles, together with an occasional giant cell. Both the connective tissue bands and the marrow show a scattered, small, round- celled infiltration. A few congested capillaries are seen in the marrow struc- ture, also several large nemorrhagic areas which are probably due to traumatism. The remainder of the section is made up of granulation tissue, rich in newly formed and congested capil- laries and showing large numbers of plasma cells, small round cells, a few eosinophiles, polymorphonuclear leukocytes, and red blood cells. The process is evidently a chronic osteitis with the formation of chronic granulation tissue. stain for the bacillus in tissue : A section of the curetings was stained by the following method: 1. Stirling's gentian violet for 10 minutes. 2. Acetic acid, 1-1,000 for a few minutes. 3. Dehydrate in 95% alcohol. 4. Oil of cloves to clear. 5. Xylol. 6. Xylol balsam. Microscopic examination of the section revealed a small, light-blue staining area near a small arteriole in one of the con- nective tissue bands. Within this area there were numerous deep-blue staining bacilli. The foregoing data show conclusively that the case is one of a posttyphoidal infection of the -rib producing a chronic osteitis with the formation of an abscess from which the typhoid bacillus was recovered in pure culture. At jihe present writing, June 1, 1903, the patient still comes to the dispensary for dressings. There are 2 small granulating areas over the junction of the seventh costal cartilage and sternum. There is still some discharge, and a small amount of induration in this region. I am indebted to Dr. W. T. Howard, Jr., and Dr. L. W. Ladd, for reviews of this work. [Reprint from The Cleveland Medical Journal, December, 1903] Wounds From Blank Cartridges OBSERVATIONS UPON 16 cASES CARED FOR AT LAKESIDE HOSPITAL, TOGETHER WITH A BACTERIOLOGIC REPORT BY DUDLEY P. ALLEN, M. D., CLEVELAND Professor of Surgery, Medical Department, Western Reserve University, Cleveland, and A. I. LUDLOW, M. D., CLEVELAND Second Resident Surseon, Lakeside Hospital About the fourth of July, of this year, a very unusual num- ber of casualties were reported resulting from the firing of toy pistols. There was nothing unusual in the weapon itself, but the report made by the explosion of the cartridge seemed much louder than that which had formerly resulted from similar explosions. The shell of the cartridge itself is of brass. The nature of the explosive has not been determined. It seems, however, to possess unusual power. The number of injured patients who were brought to Lakeside Hospital suiifering from wounds made by these blank cartridges was 15. There was one additional case of a wound of the hand by a sky-rocket. In the 15 mentioned injured by blank cartridges, the wounds were distributed as follows : There was one of the abdomen, one of the scrotum, one of the finger, and 13 of the hand. All of the wounds save one presented the same general characteristics. The- wound of the abdomen was somewhat peculiar, and may, therefore, be described by itself. The patient, a boy of 10 years, was playing with a toy pistol. He had on trousers and a shirt. The explosion of the cartridge pro- duced a wound on the left side of the abdomen. The external wound was small and seemingly superficial, and the boy did not seem to be seriously injured. I saw the boy on the morning of the fourth day following the injury. At that time there was a small wound the size of a slate pencil about three inches to the Read before the Academy of Medicine of Cleveland, October 16, 190S 2 Allen-Ludlo\\' — Wounds from Blank Cartridges left and a little above the level of the umbilicus. The abdomen wsls decidedly tympanitic, and there were evidences of peritonitis. A very grave prognosis was given, since it was evident that the peritonitis was already far advanced and the boy's condition was extremely serious. An anesthetic was administered and a small incision was made at the point of injury. It was found that the wound had penetrated the abdominal cavity, and that immediately behind it lay a coil of gut which had been so damaged that the integrity of its wall was destroyed. Immediately against the wall of the intestine was found the black wad of the cartridge. This wad had preserved its original form. The damage to the gut was so great that repair was impossible. The gut, therefore, had to be drawn into the wound, and an artificial anus was made. Notwith- standing what was done, the boy continued to fail, and died from general peritonitis the same day. This case has been reported first because it is unique, and also because it shows the explosive power and danger of these cartridges. The wounds of the hand were almost all located in the palm. In all of them, so far as could be judged, the muzzle of the pistol must have been held close to the hand. The external injury was slight in most cases, consisting of a small irregular opening through the skin. On enlarging the wounds it was found that the channel broadened from the point of entrance, so that the wound in the interior of the hand was much larger than that of the integument. The wound extended, in some cases, almost the entire length of the hand, and in others penetrated the hand so that the wound passed from the palm of the hand upward between the metacarpal bones, and it was manifest that the back of the hand was involved. in the injury. The method of treatment pursued in the cases treated in the hospital was to anesthetize the patient, to lay the wound open thoroughly, to excise all lacerated and discolored tissue, to disin- fect it with the utmost thoroughness, and to pack it open with iodo- form gauze. In the majority of cases thus operated upon the wad of the cartridge was found intact in the deepest portion of the wound. The cases treated in the hospital have been gathered under two heads, first, those which recovered without serious symptoms, and, second, those which died from tetanus. Omitting the case of wound to the abdomen, there were nine injured hands, which virtually received their primary treatment in the hospital. The wound of the injured finger alluded to above had been carefully washed, the opening packed with gauze, and had been repeatedly Ali.en-Ludlow — Wounds from Blank Cartridges r soaked with carbolic acid before admission to the hospital, 13 days after the injury. The treatment which the hands had received previous to admission to the hospital had in all cases been super- ficial dressings. In none of these nine cases had the wound been opened or excised before admission to the hospital. The length of time which the wound had existed before admission to the hospital was as follows: Three had been injured less than 12 hours, six had been injured two days, and one had been injured nine days. In addition to treatment by excision and packing with iodoform gauze, as stated above, two cases, viz., one of my own and one of Dr H. A. Becker, received prophylactic treatment by the injection of serum. Every one of the 10 cases treated in the hospital by the method described recovered. The second class of cases were those which developed tetanus. There were five of these. Four had injuries of the hand and one had an injury of the scrotum. The previous treatment of these cases had been as follows : In four of them an incision and counter-opening had been made, how long after the receipt of the wound it is impossible to say. In no case, however, had the patient been anesthetized or the wound excised and cleared out. The development of tetanus occurred in one case six days, in two cases seven days, and in two cases eight days after the receipt of the wound. All .cases developed trismus and the deaths occurred in from 12 hours to four days after admission to the hospital. The treatment given in these cases was as follows : In all cases the wounds were opened under ether and excised. All thi'se cases received treatment by intermuscular injection of antitetanic serum. In one case antitetanic serum was injected, in addition, into the lateral ventricles, through an opening made above them in the skull. In addition to this the patients received antispas- modics and anodynes to control the violence of the spasms, and to some of them chloroform was administered to relieve the sufifer- ing. The case in which the injection of antitetanic serum was made into the lateral ventricles of the brain lived for four days. No other patient lived longer than 48 hours. There is a two- fold reason for presenting this series of cases to the Academy, the iirst is their professional interest, and the sec- ond their public importance. The professional interest of the cases depends upon various factors. The first of these factors is that although the external wound caused by the wad of a toy pistol is small, the injury to the deeper structures is much more extensive than one would be led to anticipate from what appears upon the surface. The wounds are relatively greater than those. 4 Allen-Ludlow — Wounds from Blank Cartridges which have resulted from the toy pistols heretofore, and are also far greater than those which are produced by the ball of an ordi- nary pistol. Another factor of importance is that the serious symptoms in the cases did not develop at once. The wound may remain quiescent for some days, presenting no conditions which would give rise to serious apprehension. Later it becomes evident that there is deep inflammation, either upon the side of the hand which is wounded or upon the opposite side. The appearance of serious symptoms on the opposite side of the hand in several cases of tetanus treated in the hospital had resulted in counter-openings being made previous to their admission. In none of the cases in which tetanus developed had anything further been done than to make counter-openings or to incise the original wound. In no case had the original wound been cleared out and excised under an anesthetic. The cases primarily operated upon in the hospital were, in all excepting a single instance, brought in the hospital within 48 hours after the injury. The one which did not enter the hospital until the ninth day after injury was a case of a wound of the finger. The wound had been cleaned out and packed with gauze although it had not been excised. Although the number of cases is too few to permit of any positive conclusions as to the exact time after which it is impos- sible to avoid tetanus by excision of the injured tissues, the facts pointed out in these cases as to the length of time which ihe wound existed before treatment had been undertaken in those which recovered and in those which later developed tetanus, are most instructive. The patients treated primarily in the hospital were generally cared for under the supervision of the resident surgeon, Dr San- ford, and to him largely is due the credit for the successful results obtained. In all the cases treated primarily in the hospital the [jatients were anesthetized, the wounds were laid open, excised, thoroughly cleansed and packed open with iodoform gauze.. It seems, therefore, that this treatment must have been a large factor in the results obtained. To what part of the treatment the success is due is, of course, a matter of uncertainty. It may have been due to the thorough cleansing of the wound, but inasmuch as the bacillus of tetanus is an anerobic growth it seems that the suc- cess of the treatment may have been due to the fact that by pack- ing the wounds thoroughly open there was a free access of air to every portion of the wound and, on this account, the pathogenic germ failed to develop. With this idea in mind an attempt was made to gain some Allen-Ludlow — Wounds from Blank Cartridges 5 added information by means of experiments conducted upon ani- mals. These experiments have been carried on by the second resident surgeon, Dr Ludlow, and will be detailed in a paper which he will read this evening. Without anticipating what Dr Ludlow has to say in his paper, it is perhaps proper to state that the experiments have not been so successful in demonstrating the effect of treatment as had been anticipated. It is our intention to conduct these experiments further if it is found feasible to do so. Our idea in conducting the experiments was that if animals could be regularly inoculated with tetanus by shooting into them the wads from cartridges of toy pistols, it might be possible to dem- onstrate in a considerable series of cases that this development of tetanus could be avoided by certain treatment, viz., first by cleans- ing of the wound, second, thoroughly incising and excising it, and third, by packing it wide open with gauze and giving the air access to it. We have not, however, been successful in producing tetanus in animals by shooting them with the toy pistol. There remains of course the open question in this connection as to whether the wad of the cartridge, the explosive of the cart- ridge, or the dirt which gained access to the wound at the time of or subsequent to the injury, is the source of infection. This also is a question which we have not thus far been able to solve. The conclusions which it seems proper to record as the result of our observations are, first, that the cartridges of the present toy pistol are far more dangerous than those which were formerly used, and that the wound which is produced is far more extensive and more prone to produce tetanus. It seems also strongly prob- able from the fact that all cases treated primarily in the hospital by excision and open drainage recovered, that this treatment is one to be recommended highly, although it cannot yet be said that it is an absolute guarantee against the development of tetanus. The fact that five cases not so treated developed tetanus and that ten treated in this way did not develop tetanus is a strong recom- mendation for the treatment. The vast amount of injury done by the toy pistol in Cleve- land and in other cities of the United States, in July last, would seem to demand that scientific bodies, like our own, should point out in an authoritative way to the governing boards of our various cities that the manufacture and use of pistols of this kind should be absolutely prohibited. Allen -Ludlow — Wounds from Blank Cartridges Q a Pi o J w w Q a < « < P a; < a Eh < to O K o S u H Iz; Ce) < lays. 1 1 i s 1 " w 'f % 0) 1 ?. excis move -Intr in (1- .Anti ynes rum rbolii Antis ound adrei turn, rum- Anod lil Wound Wad re Serum- CarboU. modici ound ad re rum- irmal sior« Anod 1 ^& ^ ^ eg ^cg P^cofc 8 ^ 13 ^ ^, o c 1 S C s J3 B i g s E w cij rt % d f- a a a c. • o a . M . 02 . COTi 1 II 11 is il ^% 2K 3 3 ^ af •3 S 3 £3 E e 3 13 tJ J .gs .£g .^E H ^ (- H E- tS-" s B 1 dressing. ening dor- cleaned U 1 1 15 .sll 4) ? ^ o. +? c o-SS Sac K Jl C .9 •§2 S «3 ^■■S S C 4J C+J ■i3.2 C^i 0.2*^ j3.a V 5:^ S 3 3 S 3 3 .— 1 M Tj J S'o o o 3 E u 9-5 « o o o o 3 c ^ 3 CU cgcS f^ ^ (££& ^Q <3£ 1? s >. >. >. ^ 1 3 "3 ■3 •-1 A •—1 o CO 3 >. 3 r^ N , ~- II c 5, t "3 M *-t >-. i-v •-n « ■* « -* ■<*' 'd' , •ci •d •d £ g § tS CO K s a •a ■3 "d ■d •d c s a c p 3 3 3 >v o S fe ^ ^ ^ ^ c tf t; ti l4 U c3 O . 6 ^ 6 ^1 ^ 3 JS S^ A iS S 1 03 00 S n m 'sM '•« '.M 'fV^ •liS ts !■•§ &•§ •&•§ ;?s !?= ;^e :fl's ^'s ^b |8 ^o § ^■^lOtOtQlQi-ICPT-l 2. 3 S if -^ A ■-i t^ >-i 2 S 5 •B f C 13 ■£ ■^ d c3 •g C 1 I I I ^ m J ^ ^ ^ to 1^ o CO O < O CO pi m •< 8 Allen-Ludlow — Wounds from Blank Cartridges BACTERIOLOGIC REPORT A. I. LUDLOW, M. D. In accordance with our usual routine an effort was made to determine, as far as possible, the bacteriologic findings of. the cases of blank cartridge-wounds. Inasmuch as the tetanus bacillus developes only in the absence of oxygen it was necessary to make anerobic cultures. When the wound was thoroughly opened, such part of the wad as was obtainable, together with the tissue in its immediate vicinity, was placed in .a sterile test-tube and as soon as possible was transferred to the culture medium. The cultures were made according to Wright's method. Two kinds of media were used, a 1% glucose bouillon and agar agar. The medium fills the test-tube to a considerable height so that oxygen can less easily penetrate to the deeper portions. The contents of the tube are boiled for a few minutes to expel the excess of oxygen from the medium. The tube is then immersed in cold water to cool its contents rapidly, and then before the medium becomes solid, the tube is placed in a water bath at 38° C. for a few minutes. When the medium may be assumed to have reached this tem- perarture, it is inoculated with material from which the growth is to be obtained. After the culture medium has been inoculated, the cotton stopper is thrust down almost to the medium and a second absorbent cotton stopper is inserted sufficiently far down into the test-tube so that the upper end is about one centimeter below the mouth of the tube. Next there is run into the absorbent cotton stopper a small quantity of a watery solution of pyrogallic acid and about 1 c.c. of a 50% sodium hydrate solution. The tube is immediately closed air tight by firmly inserting a rubber stopper sealed with wax or paraffin. Cultures made in this way gave the following results : 1. Cases which did not develop tetanus : Five cases are included under this heading. The cultures from three cases vvere sterile. In the fourth case, streptococcus pyogenes, staphylococcus pyogenes albus and bacillus mucosus capsulatus were found. The fifth case showed cultures of strepo- coccus pyogenes and staphylococcus pyogenes albus. 2. Cultures from cases which developed tetanus : In all of these cases symptoms of tetanus were apparent at the time the cultures were taken. In the first case, the one on whom the intracranial injection \\as made, the culture was sterile. In this case the wound had Ali-en-Ludlow — Wounds from Blank Caktridges 9 been cleansed and soaked with bichlorid before the patient came to the hospital. The culture from the second case showed a growth within 24 hours. After a week had elapsed, in order to give time for spore formation, the culture was examined. The medium pre- sented a considerable amount of gas formation and a diffuse growth spreading out from the line of inoculation, particularly in the deeper portions. Coverslips made from this culture showed many bacilli which were nonmotile. Some of these bacilli were straight or slightly curved with somewhat rounded ends and no spores. These bacilli were enclosed in a transparent capsule and stained by Gram's method, the bacillus urogenes capsulatus. Many other bacilli were found appearing as slender, straight bacilli, with rounded ends. Some showed at the extremity of the bacillus a spore, spherical in form and considerably greater in diameter than the rods themselves, giving the bacilli the shape of a pin. This bacillus stained by Gram's method, thereby present- ing both morphologically and in its staining properties the appear- ance of the bacillus of tetanus. In addition to the above, arobic subcultures showed strepto- cocci and the bacillus mucosus capsulatus. From the third case the organisms were identical with those found in the second case with one important exception. In this case the bacillus which appeared almost identical with the tetanus bacillus, was more javelin shaped, decolorized by Gram's method and developed in^^the presence of oxygen. This was evidently the pseudotetanus organism. More will be said later with reference to these cultures in ' animal inoculation. The fourth and fifth cases showed the bacillus mucosus cap- sulatus and bacillus arogenes capsulatus but no organism like the tetanus organism could be found. The startling feature in all these last four cultures was the large amount of gas formation, in one case it being so great as to force out the rubber cork. ANIMAL INOCULATIONS Bouillon cultures from the second and third cases were heated to 80° C. for 20 minutes in an eflfort to destroy all but the suspected tetanus bacillus. One cubic centimeter of each solution was then injected subcutaneously into the hind leg of two guinea- pigs. A second anerobic culture made from the culture thus heated still showed both the tetanus bacillus and bacillus arogenes capsulatus, so both organisms were introduced at the inoculation. 10 Allen-Ludlow — Wounds from Blank Cartridges Twenty-eight hours after inoculation the guinea-pig which had been inoculated from the culture containing the tetanus bacil- lus showed rigidity of the leg inoculated. Within six hours this leg showed tetanic contractions. The next morning the guinea- pig was found dead. At the seat of inoculation there was no suppuration. The internal organs appeared normal. Cultures were made from the tissues excised about the point of inoculation and showed bacillus arogenes capsulatus in pure culture. The tetanus bacillus could not be recovered. The bacillus arogenes capsulatus might there- fore have had some part in the death of the guinea-pig. The culture of bacillus arogenes capsulatus should have been injected into a guinea-pig but unforli.nately this was not done. The guinea-pig inoculated from the culture containing the pseudotetanus and bacillus arogenes capsulatus was unaffected. SUMMARY In the cases which did not develop tetanus the bacillus of tetanus was not found. In only one case of the five which developed tetanus was the tetanus bacillus found, and this organ- ism was not recovered from the guinea-pig inoculated. The finding of the bacillus arogenes capsulatus in four cases which developed tetanus is interesting. Has this organism any- thing to do with the phenomena assigned to tetanus ? The action of light, especially sunlight, is very destructive to the tetanus bacillus. This suggests an experiment in regard to the action of the X-ray upon the growth of the tetanus. If it should prove destructive to the organism it might be of some value to submit the opened wounds to the X-ray treatment. The data in hand is too fragmentary to be of any value except as a suggestion for more work along this line of research. [I(eprinted from The Medical News, July 16, 1904.] THE EXTRAPERITONEAL RELATIONS OF THE APPENDIX VERMIFORMIS TO THE POSTE- RIOR SURFACE OF THE SECUM, WITH THE REPORT OF A FORM HITH- ERTO UNDESCRIBED.* BY C. E. BRIGGS, A.M., M.D., OF CLEVELAND, OHIO; INSTRUCTOR IN SURGERY, WESTERN RESERVE UNIVERSITY; SURGEON TO THE DISPENSARY, LAKESIDE HOSPITAL. The case to which attention is invited in this article presents a very unusual and, so far as we can find, a unique anatomical placement of the vermiform appendix. A girl of twenty years, a patient of Dr. I. N. Oakes, of North Ridgeville, Ohio, was referred for intercurrent appendectomy. The patient had suffered from three attacks of what appeared to be appendicular colic. The first attack was in June, 1903, and was manifest by nausea followed by pain in the right iliac fossa; the pain was not severe, and disappeared in two days ; there was no fever. About a month later she had a similar attack, but even less severe. On August 20, she suffered a third attack, manifested by naus,ea, to- gether with pain in the right iliac fossa, the pain being somewhat more severe than on *the former occasions, and persisting for five days. On this occasion, as on the others, there was no fever, no chill and no vomiting, except once immediately after taking a dose of whisky. Turpentine stupes were applied in each instance. In none of the attacks was she confined to bed, although in the last attack she felt pretty miserable. These facts in the clinical history are mentioned, as they seemed to indicate the absence of a definite in- flammatory condition. Since the last attack, how- ever, there persisted a slight degree of sensitive- ness on deep pressure in the region of the ap- pendix. The patient lived in the country a con- siderable distance from ready communication, which fact, together with the manifest tendency toward increasingly severe recurrence, was the occasion for operative interference. * Read at the Clinical and Pathological Section, Cleveland Academy of Medicine, March 4, 1904. Received for publication March 15, 1904. / The operation was performed at Charity Hos- pital, Cleveland, October 21, 1903, and presented nothing of interest aside from the anatomical con- dition. The cecum and appendix were readily located, there being no adhesions. The appendix emerged from the cecum at the point noted in the large majority of cases, a little internal and slightly posterior to the projection of the caput, as in the classical third type of the cecum de- scribed by Treves.^ Curving gently upon itself in a downward and backward direction, what ap- peared to be the tip of the appendix seemed at- tached to the cecum by an adhesion about 1.5 cm. below the appendiceal base. This at first appeared to be the entire appendix, and measured 2.5 cm. There was a mesoappendix arising from the under or left layer of the mesentery of the ex- treme distal portion of the ileum. The mesoap- pendix was about 2 cm. long at the base of the appendix, and was about 4 cm. along its free border, reaching what appeared to be the adher- ent tip of the appendix. This free border was' no- where attached to the parietal peritoneum, so that the apparently adherent tip of the appendix could be completely surounded by introducing the finger from the outer side underneath the appendix, be- neath the mesoappendix, and out again below the free border of the mesoappendix. The lack of any adhesions along this free bor- der of the mesoappendix, in the presence of what appeared to be an adhesion of the tip of the ap- pendix, was the first consideration that led to an understanding of the actual condition present. The end of the cecum was drawn upward, mak- ing prominent the line of reflection between the cecal and parietal peritoneum. On careful in- spection it was now seen that what appeared to be the adherent tip of the appendix was merely a constriction narrowing the appendix to about one-fourth its proximal diameter. From this constriction the remainder of the appendix was seen in dim outline to lie extraperitoneally along the cecum as far as the line of peritoneal reflec- tion mentioned, from which point it was lost to view in the retrocecal connective tissue. The extraperitoneal portion of the appendix was i cm. long. A short incision was made along the line of peritoneal reflection and the retrocecal connective tissue separated along the appendix to its tip. This portion of the organ was i cm. long, making the entire appendix 4.5 cm. in length ; the free mesenteric portion 2.5 cm., the extraperitoneal portion i cm., the retrocecal por- tion I cm. The organ wa's readily removed. The mesen- teric portion was 6 mm. in diameter, narrowing down at the constriction to a trifle over 2 mm. ; beyond this point the diameter of the extraperi- toneal and retrocecal portions widened out to 8 mm. The mucous membrane of the portion distal to the constriction was slightly congested, but this, together with the slight general thickening of the wall of the same portion, was the only gross evidence of inflammation. There were no adhesions. There were no foreign bodies. It is possible the distal thickening was due to muscular hypertrophy through efforts to expel mucus from this portion through the constriction. At all events the inflammation was slight and had not extended through the appendiceal wall. That we have here a primary anatomical con- dition and not a condition resulting from in- flammatory adhesions, there can be no reasonable doubt. An acute inflammatory condition gluing the appendix to the cecum would also have left the appendix adherent to the parietal peritoneum against which it rested posteriorly, which did not occur. An inflammatory process would undoubt- edly have bound down the distal portion of the mesoappendix to the parietal surface against which it rested, but the distal border was not ad- herent and the finger could be freely run under it. If the extraperitoneal appearance of the mid- dle portion of the appendix had been produced by recent adhesions, one would expect to find the remains of peritoneum on the cecal side of the appendix, but this portion of the organ was sepa- rated from the cecum only by loose connective tissue. Finally, no explanation along the line of inflammatory adhesions can possibly account for the location of the distal centimeter of the appen- dix lying in the retrocecal connective tissue. The most frequent peritoneal relation of the appendix is where the entire organ lies freely in the peritoneal cavity supported by a mesoappen- dix extending only along the proximal half of the organ, the distal portion being without a mes- entery (Quain,- Sappey'') ; exception is taken to this statement by more recent investigators, notably Monks and Blake,* the opinion prevailing that the mesentery extends further along the ap- pendix in a much larger proportion of cases than was formerly believed. An extraperitoneal rela- tion of all or a part of the appendix is quite rare, except as the result of an inflammatory process. Of 577 cases mentioned by Stroud,^ only nine were partially or wholly extraperitoneal. Bryant*' mentions three out of 144 cases examined by H. M. Biggs. Turner, quoted by Bryant," found in 105 examinations four partially, and two wholly extraperitoneal. An examination of the records of 480 autopsies, performed in the Pathological Department of Lakeside Hospital, affords a series of 440 cases in which the vermiform appendix was free from gross inflammatory conditions or operative interference. Among these 440 cases only four instances appear in which any portion of the appendix was extraperitoneal, three of which occurred in the last few autopsies. Only one of these cases sustained a relation to the pos- terior surface of the cecum, as is mentioned later. There have previously been recorded several types of extraperitoneal relations of the appendix to the posterior surface of the cecum and ascend- ing colon, all of which are so well recognized that they merit no extensive reference list. 1. As the end of the cecum is drawn forward and upward, the appendix may be seen to be ex- traperitoneal, lying against the posterior surface of the cecum. There is, of course, no free mesen- tery. It is made possible only in the case of a short or much curved appendix, or where a con- siderable portion of the posterior cecal surface is covered by peritoneum, and the cecoparietal peritoneal fold is high up under the upper portion of the cecum or lower end of the ascending colon. (Deaver'^). 2. A second form occurs when only the prox- imal portion of the appendix is in extraperitoneal relation to the posterior surface of the cecum, while the distal portion to a varying degree pro- jects freely into the peritoneal cavity, and is com- pletely invested with the peritoneum. There may or ma}- not be a short mesentery extending from the cecum or ileum to this free distal part of the appendix. In such cases the distal portion may be thought at first to be the entire appendix, but it is noted that the longitudinal bands of the cecum converge on the real, not the apparent base of the appendix, and on careful inspection the remain- der of the organ can be seen beneath the posterior peritoneal covering of the cecum. Such an in- stance is mentioned by Holmes,'*, and a figure given by Huntington. ° 3. A third variety is seen where the proximal portion of the appendix is situated extraperi- toneally against the posterior surface of the cecum, while the distal portion, instead of being free as in the second variety, passes upward along the posterior aspect of the cecum and ascending colon beyond the line of the cecoparietal perito- neal reflection, and lies in the connective tissue be- hind the large gut, entirely out of relation to the peritoneum ( Huntington "^^ ) . 4. Another relation exists when the ceco- parietal peritoneal reflection is ver}- low clown, close to the tip of the cecum, so that the entire posterior surface of the cecum is extraperitoneal, and lies in direct contact with connective tissue. An appendix in this situation lies entirely in the retrocecal connective tissue, bearing no direct re- lation to the peritoneum (Huntington^^). 5. A fifth variety, a modification of the third or fourth, is observed when the tip of the appen- dix, for a varying length, projects beyond the cecum on either side from the retrocecal connec- tive tissue in which is buried all the remainder of the appendix, or at least its distal portion exclu- sive of the tip. The projecting tip is, of course, completely invested by peritoneum. Such an in- stance has been mentioned by Deaver,^^ and has also been recently observed by the writer. 6. Still another form is seen when the tip of the appendix lies behind the peritoneum against the posterior wall of the cecum, the proximal re- mainder being invested with peritoneum and sup- ported by a mesentery that may be entirely free, or more or less adherent to the cecum. The ex- traperitoneal aspect may vary from what appears to be a mere adhesion, to a degre involving the entire distal half of the organ. In the first in- stance it resembles the case described by Treves,'-^ mentioned later, in which the tip was adherent to the under surface of the mesentery of the ileum instead of the posterior cecal surface. The other extreme has been recently observed in the Path- ological Department of Lakeside Hospital by Dr. D. H. Dolly, resident pathologist, the entire distal half of the appendix being extraperitoneal. Bv reference to the description of the case under consideration, it will at once be seen that the situation of the appendix bears a certain rela- tion to the third variety mentioned above. This relationship, however, is confined entirely to the distal half of the appendix, the first centimeter of which lies extraperitoneally against the under surface of the cecum, while the extreme centi- meter or tip of the organ lies in the retrocecal connective tissue above the cecoparietal peri- toneal fold. The proximal half of the appendix lies within the peritoneal cavity, and has an un- adherent mesoappendix with a free margin, thus resembling in a degree the sixth variety de- scribed. It has been impossible for us to find, after a careful independent search, any distinct reference to a similar condition as an anatomical entity, not the result of an inflammatory process. The nearest suggestion we can find to the peri- toneal relations existing in this case is a very gen- eral statement given by Huntington,^* in speak- ing of the peritoneal relations of an appendix which lies in the connective tissue behind the cecum, he says, '" Even in these cases, however, the dorsal surface of the cecum and the root of the appendix retain their free serous investment." By referenc« to Fig. 517, described in this con- nection, it seems apparent that the remark applies to an appendix in which the root lies against the cecum, and " the root of the appendix " has a " free serous investment " as does the dorsal sur- face of the cecum " ; in other words, that the outer or free side only of the appendix is covered by peritoneum, the same laj'er covering the dorsal surface of the cecum. This brings the description under the third va- riety mentioned above, from which reference was made to Huntington. At all events it must be observed that the text is not altogether clear, that the reference is not specific, and that no allusion is made to a free mesoappendix with an unadherent distal margin. Treves^' mentions a case in which the appendiceal tip was adherent to the under or left layer of the mesentery of the ileum forming a loop, but no reference is made to an extraperitoneal relation of any portion of the appendix. It is instructive in this connection to consider certain embryological facts which may shed some light on this very unusual location of the appen- dix. Excellent figures illustrating these points may be found in the text of Huntington^'' or Kollman.^" About the sixth week of the embryo the cecum develops as a bud on the posterior sur- face of the ascending limb of the umbilical loop. The twist toward the right of the posterior limb of this loop around the duodenocolic isthmus places the cecum in the right hypochondrium be- low the liver. This occurs about the fourth month, and the appendix at this time is already differentiated from the cecum by the overshadow- ing growth o fthe body of the latter. During the next three months, with but rare exceptions, the cecum gradually descends to the right iliac fossa, reaching this location about the seventh or eighth month. During the descent of the cecum with the formation of the ascending colon, the identity of the appendix is rapidly developed; a meso- appendix is doubtless present at a very early stage of this development In most instances as the cecum descends, the mesentery of this portion of the gut becomes ad- herent to the dorsal parietal peritoneum of the abdominal cavity, and identical with it, the coal- escence progressing from the median line out- ward toward the ascending colon and also from the above downward the peritoneal covering of the posterior part of the colon also unites with the dorsal parietal peritoneum, with which it comes in immediate contact, the two adherent lay- ers eventually giving place to the retrocolic con- nective tissue. The failure of this peritoneal coal- escence results in the unusual occurrence of a cecum and ascending colon, completely invested with peritoneum to a varying degree from below upward, with the presence of a longer or shorter mesentery from without inward toward the med- ian line, according to the degree to which the coalescence has been deficient. The retrocecal relations of the appendix ap- pear to bear a relation to these embryological changes. While the explanations offered for the development of these relations are of necessity largely hypothecal, still it can be said of them that they appear to be logical and sufficient. As the cecum descends, it seems that the tip of the appendix occasionally becomes adherent to the posterior abdominal wall in the line of this de- scent. The peritoneal relations of the appendix produced in this way would vary according as the appendiceal tip became adherent .high up or low down in the cecal course. Occurring high up, the cecal descent would carry the base down, even- tually rendering the appendix taut. Still further descent would cause the cecum to overhang the base of the appendix to a varying degree, so that the appendix would appear to spring from the posterior cecal wall ; this condition is also ac- centuated by the subsequent right lateral develop- ment of the cecum inferiorly. The descent pro- gressing still further would eventually place the base of the appendix so far up on the posterior cecal wall that the coalescence of the cecal and parietal peritoneum with the formation of the retrocecal connective tissue would naturally reach below the base of the appendix, forming the ceco- parietal peritoneal fold at this point, and thus leaving the appendix lying completely in the retrocecal connective tissue and entirely out of relation with the peritoneum. This is the condi- tion met with in the fourth variety mentioned above. Having these changes in mind, it is readily conceivable how some of the other relations of the appendix previously described may occur. The appendix becoming adherent lower down, or the descent of the cecum being less marked, only the distal portion of the appendix may come to lie in the retrocecal connective tissue, while the re- maining proximal portion, merely being drawn snugly against the lower part of the cecum, still lies in relation to the peritoneum. The peritoneal surface of this proximal portion now lying in contact with the peritoneum of the posterior cecal wall coalesces with the latter and changes to connective tissue precisely as in the case of the cecum and parietal peritoneum, so that eventually the proximal portion is covered by peritoneum on only one side, while the distal portion lies above the cecoparietal fold in the retrocecal connec- tive tissue. The mesoappendix also becomes ad- herent to the cecum ; and it is sometimes thought to be traced as a thickened peritoneal surface on the cecum to the inner side of the adherent prox- imal portion of the appendix. In this way is ex- plained the third form previously mentioned. For an explanation of the case under special consideration we must naturally turn to changes produced in this same line of development, since it bears such a close and suggestive relation to the third form, the development of which we have just considered. It seems most probable that in the case reported the tip of the appendix became adherent rather late in the descent of the cecum so that only a smair portion, i cm., became lodged in the retrocecal connective tissue. It would seem that this condition eventuated when the cecum had nearly reached its lowest position and the appendix, instead of being drawn taut against the cecum, still had a considerable proximal portion free in the peritoneal cavity with a mesoappendix extending 2.5 cm. outward from its base. The portion of the appendix between the retrocecal tip and the mesenteric portion, a distance of i cm., became adherent in some way to the cecum, the adherent peritoneal surface changing to con- nective tissue. That this process of adhesion. of the appendix to the cecum should have stopped exactly at the point where the mesoappendix ceased, leaving the distal edge of the mesoappendix entirely free and the proximal 2.5 cm. of the appendix in the usual relation to its mesentery, is strange and very unusual, but entirely conceivable in the order of development. Had the distal edge of the meso- appendix become adherent, a fossa would have been formed extending under the free portion of the appendix and the remainder of the meso- appendix. This fossa might later have been en- tirely obliterated by adhesion of the mesoappen- dix to the cecum, the free portion of the appendix also becoming adherent, and thus produced a con- dition quite similar to that described under the development of the third form, though brought about in a different wav. 10 Aside from its anatomical interest, this very unusual relation of the appendix possesses two elements of clinical interest. The unavoidable presence of a bend or kink in the appendix at the termination of its mesenteric portion has, of course, the same significance as a similar con- dition existing under other circumstances, the importance of which has been so admirably in- sisted upon by Allen. ^^ The other clinical fea- ture is the possibility of an intestinal obstruction owing to the insinuation of a nuckle of small gut under the free mesenteric portion of the ap- pendix, beneath the mesoappendix, and out below its unadherent distal margin ; such a possibility was noted by Treves^ ^ in the case showing a free appendiceal loop, to which reference has been made. Although adding nothing fo the present dis- cussion, a contribution by Peronidi^^ gives an ex- tensive bibliography on the anatomy of the cecum and appendix, containing many articles to which reference would otherwise have been made in the paper. REFERENCES. 1 Treves, F. The Anatomy of the Intestinal Canal and Peri- toneum in ^lan, p. 34 (1885). 2 Quain's Anatomy, Vol. Ill, Pt. IV, p. 109 (1898). 3 Sappey, P. C. Traite d' Anatomic Descriptive, T. IV, p. 224 (1889). 4 Monks, G. H., and Blake, J. B. The Normal Appendix; its Length, its Mesentery, and its Position or Direction, as Observed in 656 Autopsies. ~ Boston Medical and Surgical Jour- nal, Vol. CXLVII, p. 581, Nov. 2y, 1902. 5 Stroud, B. B. Notes on the Appendix (Abstract), Pro- ceedings of the Tent^ Annual Session of the Association of American Anatomists, Dec. 28, 1897, p. 127 (1898). 6 Bryant, J. D. The Relations of the Gross Anatomy of the Vermiform Appendix to Some Features of Clinical History of Appendicitis, Annals of Surgery, Vol. 17, p. 164 (1893). 7 Deaver, J. B. A Treatise on Appendicitis, p. 34, PI. V (1900). 8 Holmes, E. W. The Appendix Vermiformis, The Pennsyl- vania Medical Journal, Vol. 31, p. 205, Jan., 1902. 9 Huntington, G. S. 'ine Anatomy of the Human Peritoneum and Abdominal Cavity, p. 276, Fig. 574 (1903). 10 Huntington, G. S. Loc. cit. p. 241, Fig. 517. 1 1 Huntington, G. S. Loc. cit. p. 258, Fig. 558. 12 Deaver, J. B. Surgical Anatomy, Vol. Ill, p. 174, PI. 342 (1903). 13 Treves, F. Loc. cit. p. 44. 14 Huntington, G. S. Loc. cit., p. 251. 15 Fluntington, G. S. Loc. cit., p. 237. 16 Kollman, J. Lehrbuch der Entwickelungsgeschichte des Menschen, p. 375 (1898). 17 Allen, D. P. The Origin of Appendicitis. Transactions of the American Surgical Association, Vol. XV, p. 561 (1897). 18 Peronidi, G. Recherches Anatomiques sur le Caecum et son Appendice. Rez'uc de Chirurgie, T. XaII, p. 221, August, 1900. LUMBAR ABSCESS; REPORT OF SIX CASES TREATED BY ASPIRATION AND IN- JECTION OF lODOFORM-GLYCERIN EMULSION. ALFRED IRVING LUDLOW, M.D. Second House Surgeon Lakeside Hospital. CLEVBLAND, OHIO. No attempt is made, in this report, to establish the absolute value of aspiration and injection of lumbar abscess with tJie iodoform and glycerin emulsion. The investigations already made in regard to this method show that it is of great benefit to many patients. Although the cases cited below are too few in number to be of positive value in themselves, yet when added to other reported cases they may be of some confirmatory value. It is with this idea that we present the following cases, occur- ring in the service of Dr. Dudley P. Allen at Lakeside Hospital. Case 1. — A male, aged 25 years, white, single, admitted to the hospital Feb. 7, 1898. Four months previous to the time of entrance the patient noticed a swelling about the size of a walnut in the right inguinal region. The tumor gradually increased in size, but at no time did it cause pain. Family History. — Father, mother, four brothers and one sister living and well. One sister died of pulmonary tuber- culosis at the age of twenty- three ( six years ago ) . Personal History. — Measles during childhood. For the last three years, particularly in winter time, he has been troubled with a cough. His physician told him that he had pulmonary tuberculosis. In April, 1898, the patient complained of severe pain in the right lumbar region. This lasted about three weeks, during which time he experienced some fulness in the right groin whenever he flexed his right thigh on his abdomen. After this period, however, no symptoms were noticed until the distinct swelling appeared. Physical Examination. — The tumor is somewhat pyriform in shape, about one-half larger than the ordinary incandescent lamp, and extends from the right external abdominal ring up- ward, backward and outward, just above Poupart's ligament along the crest of the ilium to the miaaxillary line. The tumor is fluctuant. The skin is normal in color and there is no in- crease in surface temperature. The remainder of the physical examination is negative except for some dulness in the right supraclavicular and infraclavicular spaces. Treatment. — The patient received no treatment before coming to the hospital. In this, as in the following cases, the opera- tive treatment consisted of a, thorough sterilization of the skin over and about the tumor, followed by aspiration of its contents with a trocar. As a rule, about three ounces of an emulsion of iodoform and glycerin, fifteen grains to the ounce, were injected into the abscess cavity. In the majority of cases the operation can be done under local anesthesia. In this case ten aspirations and injections were made. The -first two at intervals of a week and the remaining eight at intervals of two weeks. During the periods between the last eight aspira- tions the patient was permitted to return to his home. At each aspiration, except the last, from six to ten ounces of fluid were withdrawn, the first being of a whitish color, con- taining a cheesy material, while the remainder were of a dark brown color. At the last aspiration only one ounce of a red- dish brown fluid was obtained. Results. — The bacteriologic findings were negative, all cul- tures being sterile. The patient has been seen within the last few days and shows no evidence of return of the abscess. It might be interesting to add that in February, 1901, while he was troubled with a severe cough, tubercle bacilli were found in his sputum. At present, October 21, 1903, however, he is gaining in weight, lias had no cough for many months, and seems to be in excellent condition. Case 2. — This patient, a female, aged 19, white, single, was admitted to the hospital April 15, 1899. Family History. — Negative. Personal History. — General health good except for the present trouble. She gives a history of three injuries to her back, but does not remember the time or location of the in- juries. A year previous to admission the patient began to have pain in the lumbar region, especially when arising from a chair or after jarring the body by a misstep. The pain and discomfort gradually increased. In October, 1S98, she was advised by her physician to try absolute rest. Accordingly she remained i bed for three months, which resulted in marked improvement In February, 1899, she noticed a swelling just above Pouna f ligament on the right side. A smaller swelling, too-ether w^flf a general fulness more externally, appeared just "to the llff of the vertebral column in the lumbar region. Physical Examination.— This was negative except for thn swelling above noted. Treatment. — -April 19, 1899, twenty-five ounces of light green cloudy pus were removed from the left lumbar region, the trocar being inserted near the apex of Petit's triangle. Although only two ounces of the emulsion were injected, the urine, on the next day, gave a strong iodin reaction. On May 2 the right inguinal tumor was aspirated, and five ounces of greenish pus obtained. The same amount of emul- sion was used, but the urine gave no iodin reaction. Ten days later the trocar was inserted in the left inguinal region where a swelling had appeared and four ounces of grayish pus were withdrawn. One week later this same region was again aspirated and three ounces of grayish pus removed. A month intervened before the next aspiration. Ihiring this period the patient was out of doors in a wheel chair every pleasant day. At the last aspiration, on the left side, about five ounces of greenish pus were evacuated. Cultures made at the time of each aspiration were sterile. Results. — Three months after this last aspiration the patient was examined by Dr. Allen and no indication of any reaccumu- lation could be detected. At the present time her physician says that there has been no reappearance of the abscess and that her general health is excellent. Case 3. — Female, aged ten y^ars, white. Admitted to the hospital Feb. 7, 1901. Family History. — Father died of tuberculosis at the age of 35. Mother died of the same disease at the age of 30. Personal History. — Patient fell down stairs when one year old. No trouble was apparent until a year later, when the spine showed some kyphosis in the lumbar region. Since this time she has sufl'ered repeated attacks of pain in the back at various times. About a year ago (November, 1899) she began to complain of pain in the left knee. Frequently the pain has been so severe as to cause her to remain in bed while in the intervals between these attacks she could play and walk about as usual. One month ago the pain became especially severe, and since that time it has been almost constant. In addition to this trouble she gives a history of pertussis, measles, scarlet fever, varicella and diphtheria. She has coughed considerably from time to time for the last four years. Physical Examination. — ^The heart was found normal. The lungs show a few rales at the right apex, posteriorly, and rales over the larger portion of the left lower lobe. The lumbar vertebrEe show a marked kyphosis and the dorsal vertebrse a mild scoliosis. Examination of the extremities negative, except that percussion of the left lower extremity in its axis causes mild pain in the hip. Treatment. — ^The first treatment employed was rest in bed and constitutional treatment. Extension was applied for two weeks, giving considerable relief. On April 1, 1901, a swelling was noticed in the left inguinal region. It began to increase in size and became tender. The next day a trocar was inserted 3 centimeters above the left anterior superior spinous process of the ilium and several ounces of thick yellow pus aspirated. The •usual amount of emulsion was injected into the cavity. Two weeks later, the abscess having reaccumulated, the same proc- ess was employed. For a few days the patient was somewhat depressed, and iodin was demonstrated in the urine. Results. — Within a, week, however, she commenced to im- prove greatly, and on June 1 she was placed on a Bradford frame. As often as the weather permitted she was taken out of doors. Under this treatment she gained rapidly in weight and strength. There was no evidence of reaccumulation of the abscess when the patient was discharged. At the present time (October, 1903) there is still no indication of return of the abscess. Her general health is good and she is able to attend school regularly. Cask 4. — Female, aged 39 years, white, single, was admitted to the hospital Jan. 31, 1902. Family History. — Negative. Personal History. — She had typhoid fever three years ago, but otherwise her general health had been good until July, 1900, when she began to have "neuralgic" pains in her right hip and thigh. Later the corresponding parts on the other side of the body were affected. From the time of the first attack the pain has been fairly constant, being aggravated after exertion. She experienced difficulty in rising up after stooping down to pick up objects from the floor and often was obliged to assist herself by taking hold of a chair or table. In December, 1901, she noticed a mass just above the left iliac crest, anteriorly, which has gradually increased in size and has caused her to favor that side in walking. She has never had any symptoms localized in her back at the pite of the kyphosis. Physical Examination. — The heart and lungs are normal. At the second and third lumbar spines there is a kyphosis slight in degree, but distinct. A mass can be made out in the left iliac region extending about half way from the anterior superior spinous process of the ilium toward the median line, downward toward the pubes and upward to a little above the crest of the ilium. It is distinctly fluctuant, and the amount of tissue covering it anteriorly is evidently not great. Treatment. — Previous to entering the hospital she had re- ceived no treatment. After rest in bed for a week, the left iliac region was aspirated in the usual manner, the trocar being inserted just above and to the inner side of the left anterior superior spine of the ilium. Sixteen ounces of green- ish pus were drawn off and three ounces of the iodoform emul- sion injected. Within a week the abscess cavity began to refill, and February 24 a second aspiration was made at which time the same amount of pus was removed. During the in- terval between this and the next aspiration the patient was up in a wheel chair almost every day. March 19 it was evident that another aspiration was necessary, for distinct swelling and fluctuation could be made out in the left iliac region. Ac- cordingly, this was done, and again sixteen ounces of greenish pus were evacuated. A week after this aspiration the patient was allowed to return home after being instructed to live out of doors as much as possible. On April 21 she returned to the hospital. The abscess had reappeared and distinct fluctuation was present. The nest day the abscess was again aspirated and this time about eight ounces of yellowish pus were ob- tained. Bacteriologic Report. — The coverslip and cultures from the pus obtained at the first aspiration were negative. Pus from the second aspiration showed the presence of the Bacillus proteus vulgaris. Cultures from the third and fourth aspira- tions were sterile. No tubercular bacilli could be detected. Results. — The patient was allowed to go home after the fourth aspiration with instructions to report from time to time. She was seen by Dr. Allen June 27, 1902. Her general health was very much improved and she felt much stronger, although she had not gained much in weight. A little thick- ening was made out in the iliac fossa, but no fluctuation could be perceived. The patient was again examined in December, 1902. There was no evidence of return of the abscess. Her general health was excellent. At the present time (October, 1903) there is no return of the abscess. Case 5. — Female, aged 7 years, was admitted to the hospital Jan. 2, 1903. Famil-j/ History. — Father, mother, two sisters and three brothers are living and in good health. Her grandmother and uncle died of pulmonary tuberculosis. Personal History.— The patient's general health has been fairly good. About two years ago her parents noticed a bulg- ing of the spine. Until six months ago the patient complained of no pain, but at that time she commenced to limp and suffer pain in the left knee. The kyphosis has been gradually in- creasing to the present time. Physical Examination. — ^There is a marked kyphosis with a slight right scoliosis beginning at the eleventh dorsal ver- tebra and extending to the fourth lumbar. There is no ten- derness on pressure. Some resistance can be made out in the left iliac fossa. The remainder of the physical examination is normal. Treatment. — Three days after admission to the hospital a plaster jacket was applied, and three days later the patient was allowed to go home. June 18, 1903, the patient returned to the hospital with a large fluctuating mass on the upper and outer side of the left thigh. This mass has been gradually increasing in size for the past few weeks. It was unattended with pain or redness. The day after admission, under ether anesthesia, the abscess was evacuated and 325 e.c. of thick dirty yellow pus removed. The usual amount of iodoform emulsion was injected. The next day lodin was found in the urine. June 29, the abscess having reappeared, it was again aspirated, and this time 300 c.e. of pus were removed. A third aspiration was done on July 7, at which time about 300 c.c. of yellowish fluid were obtained. July 24 a fourth aspiration was made and eight ounces of brownish fluid were withdrawn. This time four ounces of the emulsion were injected. Two days after the aspiration there was the only marked rise of temperature during the course of the aspirations. On this day the tem- perature was 102 F. Bacteriologic Report. — Cultures taken at the time of each aspiration were sterile. Results. — July 29 the thigh appeared practically normal, so the patient was sent home with instructions to her physician to send her back to the hospital if the abscess reappeared. Up to the present time (October, 1903) there has been no return of the abscess, and the patient is in excellent health. Case 6. — A female, aged 33, white, married, was admitted to the hospital Aug. 6, 1903. Family History. — One sister died of tuberculosis. Personal History. — Patient had diphtheria two years ago. She has always been of a nervous disposition. Five years ago she had an attack of pain in the lumbar region. At that time she thought the trouble must be rheumatism. These at- tacks of pain continued to come on at varying intervals. Be- tween the attacks she was fairly well, although she had some dilHculty in arising from a reclining position or on attempting to pick up any object from the floor. At night this pain was so severe that she would cry out. About ten months ago a. swelling appeared in the right inguinal region. Physical Examination. — In the region of the lumbar ver- tebrae there is a well-marked scoliosis to the left and a well- defined kyphosis in the same region. In the right groin there is a swelling about three centimeters in diameter, fluctuant and becoming smaller when the patient reclines. The re- mainder of the physical examination is negative. Treatment. — Aug. 8, 1903, under local anesthesia, an as- pirating needle was inserted into the tumor mass and 300 c.c. of greenish yellow pus removed. Two ounces of the iodoform emulsion were injected. The day after the operation the urine gave a reaction for iodin. A second aspiration was made August 13 and eight ounces of yellowish serous fluid withdrawn. One week later very little swelling could be detected and by August 30 the mass had entirely disappeared. On August 31 the iodin reaction disappeared from the urine. The patient was sent home on a cot with instructions to re- main in bed several weeks. Bacteriqlogio Examination. — The pus ' contained many leu- cocytes, but both cultures were sterile. Results. — -At present the physician who is attending the patient reports that he has detected no return of the abscess. SUMMARY. 1. Four cases gave a family history of tuberculosis. 2. Five cases occurred in females whose ages ranged from 7 to 39 years and one case in a male 25 years old. 3. Two patients gave a history of injury to the back. 4. In three cases two aspirations were made, in one case three, in another four, and in another ten. 5. The urine from four cases out of the six gave a reaction for iodin the next day after the aspiration. This reaction persisted only for two or three days, except in one case in which it persisted for two weeks. 6. Slight mental depression was noticed in two cases. 7. As a general rule, there was an elevation of temperature from two to four degrees following each aspiration. 8. The cultures were sterile in every case except one, in which the Badllus proteus vulgaris was obtained. 9. In all the six cases there has been no indication of return of the abscess after a period ol five years in one case, three years in another and two years in a_ third, while in the remain- ing three one year or less has elapsed since the last aspiration. There was a marked improvement in the general health of Reprinted from The Journal of the American Medical Association, July 2, 190J,. ^Extracted from the American Journal of the Medical Sciences, May, 1905 A CLINICAL AND PATHOLOGICAL REPORT OF A CASE OF SPLENIC ANAEMIA. By H. L. Sanfoed, M.D., AND David H. Dollet, M.D., OF CI.EVEIAND, OHIO. (From the Surgical Service and the Pathological Laboratory of Lakeside Hospital, Cleveland, Ohio.) 1. Clinical Report (Dr. Sanford). Case of anaemia of long -duration with gastrointestinal disturbances and pain in splenic region; recent severe and repeated hemorrhages by stool and vomitv^ following traumata; blood characteristic of secondary anwmia of chlorotic type without leukocytosis, with no general lymphatic enlargement; idio- pathic splenomegaly; exploratory laparotomy; subsequent splenectomy; death. Mr. B. R. T., aged twenty-eight years, married, a farmer, of Greenwich, Ohio, was sent to Lakeside Hospital on the evening of October 14, 1903, with symptoms of vomiting of blood, tarry stools, profound anaemia and a tumor in the lower mid-abdomen. The patient stated that three weeks before he had been thrown over the head of a spirited horse and had struck flat on his back, but had not been prevented from working about as usual. Two weeks later, which was a week before entrance, he was jerked off his feet and thrown violently some distance by this same horse which he was leading. The following morning he had a copious black liquid stool, after which he vomited much blood, he thinks at least a pint, of both dark blood clots and some fresh blood. This vomiting continued at intervals that day, and then stopped, but he had large tarry stools nearly every day up to the time of his -entrance. He was in bed during most of the time, felt greatly prostrated, and had no appetite. His physician on the day before he was brought to the hospital first noticed a tumor in the lower mid-abdomen, which he took to be a distended bladder, but on catheterization its size was not altered. The urine appeared normal and contained no blood. Dr. Maynard, of Elyria, who was called in consultation, then sent him to the hospital. On entrance the patient seemed greatly -fatigued from his journey and could answer questions with diflB- ^culty. 2 SANFORD, dolley: splenic anemia. Physical Examination. Patient is a young American, slight in frame, very sallow, somewhat emaciated, in considerable state of shock. Temperature, 100.6°; pulse, 120, small volume, regular, compressible; respiration, 28. Sclera subicteric; teeth poor; tongue furred and moist. Chest, slender and shallow, expansion equal and good; clavicles and ribs prominent; supraclavicular and infraclavicular fossae deep; no glands palpable. Interspaces sunken; epigastric angle narrow. Percussion note dull over clavicles and supraclavicular spaces, elsewhere hyperresonant. Breath sounds clear front and back; musical rales on expiration in left foreaxillary line. The area of cardiac dulness is hard to define; it apparently extends somewhat to the right of the sternum; a blowing systolic murmur is heard all over the base, and not at the apex or in the axilla. The abdomen is sunken, scaphoid in general outline, somewhat sensitive to palpation in the left upper quadrant where there is slight muscular rigidity. Liver dulness begins at the fifth space in nipple line and extends to costal margin. Splenic dulness not made out. The pole of right kidney is palpable on deep inspiration. Left kidney not felt, and it seems with one hand in the left lumbar region and the other under the left costal margin that the two hands can be brought nearer together with less intervening structures on the left than on the right side. On pressure here the patient is evidently sensitive, although apathetic to examination elsewhere. Just below the umbilicus is a mass, oval in general outUne, lying diagonally, its higher pole to the right, smooth, hard, not tender, quite freely movable, edges rounded, no notches felt. The abdomen is everywhere tympanitic except over the mass described where the note is flat. The inguinal glands are slightly enlarged. Rectum empty except for a few small clots; no hemorrhoids; extremities thin; no tibial scars or oedema. The patient was given artificial heat and stimulation by strychnine and saline subcutaneous infusions every six hours. Milk and whites of eggs; morphine for restlessness. A glycerin enema resulted in a black liquid stool, which on examination was found to be largely decomposed blood. The immediate history of severe repeated traumatisms combined with the patient's statement that he had always been well previously, quite naturally led one to regard the present symptoms as the direct result of the injury rather than to seek a more remote cause. Various possibilities were considered, among which hemorrhage from a gastric or duodenal ulcer, and a dislocated viscus were foremost. Neither diagnosis seemed to account for all the symptoms, and in view of the later developments in the case, a further discussion of these points is unnecessary. SANFORD, DOLLEY: SPLENIC ANEMIA. O The patient reacted fairly well to stimulation, and was kept under observation the next day, during which he had seven more black liquid stools. Rectal feeding was added to the treatment which was otherwise the same. A blood examination on the day after admission showed: Reds, 2,800,000; leukocytes, 8000; haemoglobin, 25 per cent.; coagulation time 45 seconds; no plasmodia seen. The next day the patient's unchanged condition and the uncertain diagnosis seemed to warrant an exploratory operation which was decided upon. Operation, October 16th (Dr. Dudley P. Allen. Gas and ether). Laparotomy ; exploration. An incision 9 cm. in length was made in the median line below the umbilicus and over the tumor. A small quantity of clear peritoneal fluid was met with from which cultures were taken. The intestines were normal. The mass was identified as the spleen, greatly enlarged and displaced, freely movable, otherwise of normal appearance. Accurate measurements were not taken, but the organ seemed at least two or three times its normal size. The surgeon considered any operative procedure on the spleen unwise in the patient's condition and closed the incision. Another incision in the median line of the epigastrium was then made and the stomach inspected, and found normal. Nothing abnormal felt about the liver, gall-bladder, appendix, or pancreas. Palpation of the kidneys showed them apparently normal in position and size. The incisions were then closed. The patient received very little shock from the operation, which took but fifteen minutes. The conditions found at operation raised the question whether the traumata might not be coincidences or, perhaps, exciting causes in producing the appearance of symptoms primarily due to some as yet unrecognized organic lesion. The arrival of the patient's friends gave the first opportunity of getting the family and previous history in the case, which was as follows: Family History. Father died of "broken back" at the age of seventy years. Mother and one brother living and well. An uncle died of heart disease. No history of malaria, syphilis, cancer or tuberculosis in the family. Previous History. Patient had children's diseases. He has always been pale since childhood. Between seven and eight years ago he had an attack of very violent vomiting which continued all night and was not connected with any indiscretion in diet; vomitus contained no blood. He has always had poor digestion, often dis- tressed after meals, sometimes vomiting, with relief to the discomfort. In the summer of 1902 he had "bowel trouble," in which there was diarrhoea, griping pains, severe vomiting, with some blood in stools and vomitus. From this time up to the injury the patient was not well, and was under a doctor's care for general "ill feeling'* 4 SANFOED, dolley: splenic anemia. and frequent urination. In the winter of 1902, he also complained of a good deal of dragging pain in the back and left side. No history of venereal disease; tobacco in moderation; no alcohol. Patient has not lived in a malarial district and never had periodic chills. With this suggestive history and the enlarged and dislocated spleen found at operation, a detailed blood examination was then made and showed the following conditions: Reds, 3,480,000; leukocytes, S500; haemoglobin, 25 per cent. Differential count of 600 cells: polymorphonuclears, 74 per cent.; small mononuclears, 18 per cent.; large mononuclears, 4 per cent.; transitional forms, 1.5 per cent.; eosinophiles, 2.5 per cent. Red cells took stain poorly, and were variable in shape and size, centre quite thin, no rouleau formation, some poikilocytosis, six nucleated reds seen; no plasmodia. The urine was amber, clear, specific gravity, 1020, acid, no sugar, faint trace of albumin, sediment, few leukocytes, epithelial cells. The case then presented the following data: A patient with (1) a previous history of an anaemic condition with gastrointestinal disturbances and pain in the splenic region; (2) an immediate history of severe and repeated hemorrhages by stool and vomitus following traumatism; (3) a present condition of profound anaemia, with the blood characteristic of a secondary anaemia of the chlorotic "type without leukocytosis, with no general lymphatic enlargement; and (4) a splenomegaly apparently not due to malaria, syphilis, tuberculosis, leukaemia, or amyloid disease. From this symptom- -complex the possibility of the existence of splenic anaemia was -Strongly suggested and a course of supporting treatment was instituted while further study could be carried on. During the week after operation the patient's general condition improved slightly. The stools no longer contained fresh blood, but were dark, copious, watery, varied between three and seven ■daily and were at times involuntary. There was no further vomiting. Liquid nourishment was well taken. The temperature ranged between 99.5° and 101.5°, and the pulse from 100 to 140. The outline of the tumor was marked out on the abdomen in silver nitrate solution to watch for changes in its size or position. On the tenth day after operation thfe tumor was found to have rincreased in size by almost one-half, and the gain in size seemed mostly toward the right and below. The new outlines of the enlarged organ were marked on the patient's abdomen. Two days later a still further very rapid increase in size was apparent, so that the anterior rsurface of the mass presented an area twice as large as when the patient entered the hospital. The incisions of the original explor- ation were firmly healed by first intention. To-day for the first time petechial hemorrhages were seen on the sides of the chest in areas the size of the palm of the hand, and smaller similar purpuric spots appeared on the right forearm and hand. SANFORD, DOLLET: SPLENIC ANiEMIA. 6 No change in subjective symptoms accompanied these new phenomena. The patient was still apathetic and weak, though he declared he felt better, and his only complaint was the exhaustion attendant on frequent bowel movements which still were of a tarry color and averaged five to six daily. The blood condition was examined from day to day, and showed practically the same con- dition as given above. The question of splenectomy in the case had been constantly before the surgeon's mind, but he desired first to improve the patient's general condition in view of the recent enormous loss of blood he had suffered. The rapid and sudden increase in the size of the spleen, however, convinced Dr. Allen that ground was being lost instead of gained, and an immediate operation was proposed and accepted by the patient and his friends. Operation October 29th (Dr. Dudley P. Allen. Gas and ether). Laparotomy; splenectomy; closure. A vertical incision 15 cm. in length was made separating the fibres of the left rectus muscle in order to avoid the former incisions. On opening the peritoneum some clear fluid was met with from which cultures were taken. The spleen was found to be lying free in the abdomen without adhesions to the intestines or other structures, its only attachment being a large long pedicle containing four twists, composed of large congested vessels, some of which contained thrombi. A more detailed, microscopic description of the organ and its pedicle will be given later. The pedicle was securely tied high up with silk, clamped off distally and removed with the organ, which was with difficulty squeezed through the skin incision. No hemorrhage accom- panied the removal; the pedicle stump was reinforced by over- and-over catgut stitches, and showed the openings of several large vessels. The abdomen was closed in layers. The patient seemed to receive no shock from the operation which lasted twenty minutes. For two days after operation the patient's general condition was encouraging. The temperature approached the normal, and the pulse rate decreased to an average of 90, with improvement in its quality. Stimulation with strychnine and saline subcutaneous infusions was continued as before. On the third day a marked oedema of the whole right leg appeared which increased rapidly to alarming proportions. This was followed some days later by oedema of the right side of the scrotum and left side of the neck. Vomiting reappeared on the sixth day, and it became difficult to nourish the patient, as rectal nutrition was not retained except at intervals. There was no further appearance of blood, however, either in the vomitus or stools. Purpuric areas remained unchanged. Later the vomiting ceased and the patient seemed much brighter again. An attempt was made to give Fowler's 6 SANFORD, dolley: splenic anemia. solution, but this had to be abandoned, as on the eleventh day the patient again rejected all nourishment. The abdomen, which up to this time had been flat, now began to show distention, with appearance of free fluid in the flanks. This collected rapidly, and on the thirteenth day the abdominal wound, which had healed and from which the stitches had been removed, burst open and a large amount of clear fluid escaped, from which cultures were sterile. The patient now was irritable and very hard to rouse, sleeping nearly all the time with some delirium. Nourishment was refused, and rejected when given. Stools and urination were involuntary. This condition of progressive weakness, accompanied by emaciation, dyspnoea, and cyanosis, continued until the patient's death on November 15th, seventeen days after operation. The changes in the blood following the removal of the spleen were very interesting, and in the main coincided with observations reported under like conditions in other cases, though, unfortunately, the patient's death prevented their being followed out over a long period. Immediately after operation there was a great diminution in the red corpuscles from 3,480,000, four days before operation, to 1,836,000 two days after operation. This could not be accounted for by hemorrhage, either before or during the operation, as there was no great loss of blood immediately before the operation and no blood was lost in tying off the pedicle of the spleen. It must be attributed to the effect of the removal of the organ. After this initial drop the red count began steadily rising and had reached 2,800,000 just before death. The white count which had been 8500, jumped to 34,000 two days after operation and gradually decreeised to 22,000 at death. The percentage of haemoglobin which before operation was 25 per cent., dropped to 22 per cent, after operation, and then gradually rose to 28 per cent. All these changes, the marked postoperative diminu- tion in red cells, and to a less degree in haemoglobin, and the sudden rise in leukocytes, corroborate previous experiences following splenectomy. In the differential counts the polymorphonuclear form showed a marked increase from 74 per cent, before operation to 93 per cent, before death. The eosinophiles varied in amount. Nucleated red forms which averaged 6 in a count of 600 cells before operation increased to 36 before death. To facilitate comparison the results of the various counts are tabulated : SANFORD, DOLLEY: SPLENIC ANEMIA. Before operation After splenectomy . Date. Red cells. Leukocytes. Hsemoglofcin. 1 October 16 2,800,000 8,000 26 per cent. • -1 25 3,480,000 8,500 25 " November 2 1,836,000 34,800 22 3 1,914,000 30,000 23 7 2,212,000 22,000 22 10 2,848,000 23,720 28 11 2,464,000 26,000 26 12 2,568,000 23,000 27 13 2,632,000 22,000 28 14 2,682,000 22,000 28 Differential Counts of 600 Cells. Before operation After splenectomy Date. Oct. 25 Nov. 2 Polymorpho- nuclear. 74.0 per ct. 83.5 " 89.0 " 93.0 " 92.3 " Small mononu- clear. 18.0 per ct. 12.75 " 4.66 " 3S " 3 6 " Large mononu- clear. 4.0 per ct. 1.75 " 2.33 " 1.4 " 1.3 " Trans- itional. 1.5 per ct. 1.50 " 2.0 '• 0.6 " 1.0 " Eosino- philes. 2.6 per ct. 0.6 " 2.0 " 1.2 " ].8 " 2. Pathology. (Dr. Dolley.) Autopsy Protocol. (Autopsy performed seventeen hours after death.) The body is that of a "well-formed but emaciated white man, 180 cm. long. There is moderate rigor mortis, and slight posterior hypostasis. The con- junctivae are somewhat yellow. The mucous membranes are pale. The skin is sallow but not pigmented. A few small petechial areas are scattered just above the costal border. There is no glandular ■enlargement. The chest is barrel-shaped. The supraclavicular fossae and sternal notch are deep. The abdomen is scaphoid. Two recently healed wounds of operation appear in the midline; one, 6 cm. long, beginning at the xiphoid cartilage, the other, 7 cm. long, beginning at the imibilicus. A third laparotomy wound, healed with the exception of one stitch, 12 cm. long and 3 cm. to the left of the umbilicus, is present in mid-abdomen. The right leg is larger than the left throughout, and moderately cedematous. Brain and spinal cord not examined. Abdominal Cavity. The spleen is absent. The splenic veins become very tortuous from the end of the pancreas, one and one-half coils being left after the splenectomy. They measure from the end of the pancreas to their ligated extremity 2.5 cm. coiled and 5 cm. 8 SANFORD, DOLLEY: SPLENIC ANEMIA. uncoiled. The apposing surfaces of the coils are united by well- organized adhesions. Beginning at the junction with the portal vein, the splenic vein and its branches become more and more distended with palpably soft material until at their ligated extremity the larger of the two branches is 2 cm. in diameter. The veins are attached to the jejunum at one spot, 2 cm. in diameter, by organizing fibrinous exudate. On incision they contain soft mixed thrombi, which in places are loosely attached to the wall. There is sUght thickening of the walls but no evidence of calcification. The splenic artery is rather tortuous and dilated. It measures 2.5 cm. from the end of the pancreas to its ligated extremity. Its elasticity is poor, the intima is somewhat roughened and extremely reddened toward the distal end. It contains a similar grayish-red thrombus. Mediastinal fat is scanty. The thymus is absent. Pleural Cavities. Each contains about 10 c.c. of clear serous fluid. There are no adhesions. The pericardial cavities contain 30 c.c. of similar fluid. The sac wall is not thickened. Thyroid Gland. Both lobes are somewhat enlarged. The tissue appears normal. The tracheal glands are markedly anthracosed. One at the bifurcation is studded with small fibrocaseous tubercles. The lungs are voluminous, externally smooth and slightly anthra- cosed. They are hypercrepitant and cushiony except in the lower posterior portions. On section there is moderate hypostasis of ihe lower lobes. Elsewhere the tissue is more moist than normal. The left weighs 550 grams, the right 549 grams. The bronchi contain some frothy watery mucus. The bronchial glands are like the tracheal. One is composed of a calcified envelope with a caseous centre. The heart is of normal size. The right ventricle averages 4 mm., the left 15 mm. in thickness. The auricles and right ventricle contain post-mortem jelly clots. Fluid blood is scant. The muscle is pale and flabby. The valves are normal except for a few ather- omatous patches on the ventricular surface of the right mitral segment. The circumferences of the valvular orifices are: aortic, 6.2 cm.; pulmonic, 7.3 cm.; mitral, 9.5 cm., and tricuspid, 13 cm. The coronary arteries have a roughened intima with considerable atheroma. They are not tortuous. The endocardium elsewhere is smooth. Aorta. The elasticity is normal. The wall is not thickened. The intima in places is roughened by yellowish or whitish plaques. The pulmonary artery and venw cavce appear normal. The liver weighs 1356 grams. It measures 27 x 14 x 8^ cm. The left lobe measures 13J x 14 x 5 cm. It is contracted, but its shape is fairly preserved. The capsule is slightly thickened, and the SANFOKD, DOLLEY: SPLENIC ANiEMIA. _ 9" capsular veins are moderately injected. The veins of the round ligament are not dilated. The surface is moderately roughened and nodular, especially over the inferior and lateral portions of the right lobe. Over the inferior surface are several nodules softer in consistence than the rest of the tissue. On section, the organ is moderately hyperaemic. The cut surface is uneven and finely,, though not uniformly, granular. The tissue is firm and cuts vfith increased resistance, especially in the lower and lateral portions of the right lobe, just under the capsule. The lobular outlines are not distinct; the tissue is markedly bile-stained, of a dirty brownish- yellow color. Owing to the bile staining the character of the apparently increased connective tissue is not evident. Glisson's capsule is most affected. The portal vein is not thickened and its intima is smooth; it is free from thrombi. The gall-bladder is constricted at the junction of its proximal and middle thirds and bent upon itself, so that the two portions are in partial apposition and united by old adhesions. Elsewhere it is non-adherent and apparently normal, containing dark bile. The ducts are patent. (Esophagus. Congested sub-mucosal veins are distinct about the cardiac orifice. The (Esophageal veins are considerably dilated and tortuous. The coronary veins of the stomach are not so dilated, but the cardiac branches communicating with the oesophageal veins show marked varicosity. The stomach is somewhat dilated, but the wall is not appreciably thinned. It contains 500 c.c. of brown sour-smelling fluid. The mucosa shows considerable post-mortem change and is not indur- ated. The sub-mucosal veins about the cardia and along the greater curvature are dilated. The folds of the mucosa only appear along the greater curvature, where their prominence is due to dilated veins. No eroded veins are apparent. There are no ulcers nor scars. The pancreas is moderately hyperaemic. The head and body are of normal size, but are distinctly indurated. The tail is very con- tracted and firmly attached to the splenic vessels. On section there is great increase of dense fibrous tissue. The small intestine is not distended. It shows only an occasional congested but not swollen area in the mucosa. The veins of the serosa are not congested. The appendix is normal. Large Intestine. The mucosa of the caecum and the ascending colon is markedly congested. There are no hemorrhoids. The suprarenals show post-mortem softening of the medulla. Kidneys. Right weighs 215 grams, the left 285 grams. They measure respectively 12 x 7.5 x 3.8 and 13 x 8.6 x 7 cm. Perirenal 10 . SANFORD, DOLLEY: SPLENIC ANEMIA. fat is scanty. The cortical veins are not dilated. The capsule is normal. The cortex of the right is 6 mm., of the left 8 mm. in thickness. The cut surface is smooth, but mottled with grayish areas corres- ponding to the labyrinths. The organs are moderately pale and the glomeruli are barely visible. The arteries are not sclerotic. The ureters, bladder, seminal vesicles, prostate, and testicles appear normal. The mesenteric glands are small and barely palpable. The retro-peritoneal glands are enlarged, firm, and of a glistening grayish-white appearance on section. Hcemolymph Glands. Lying on both sides of the sternum, along the carotid arteries, the renal vessels, the ihac vessels, and about the splenic vein and the oesophagus appear glands varying in size from a pinhead up to that of a large bean. These are soft, dark- red, spleen-like and fairly abundant. Between thirty and forty of convenient size were removed. The right iliac vein just above Poupart's ligament is filled with a red, non-adherent thrombus. The right femoral vein contains a similar thrombus. This changes in character 5 cm. below Poupart's hgament, having a white friable core which at one point is attached to the wall. The inferior hemorrhoidal veins are moderately dilated and tortuous. Bones. Several ribs and the right femur were opened and found to contain bright-red and succulent marrow. No lymphoid area^ appear. Spleen (received from the surgical department. Gross description by Dr. Howard). The specimen consists of a spleen, irregularly oval in shape, presenting a rounded anterior margin from which the notch has been obliterated. It weighs 1650 grams and measures 22 X 16 X 7 cm. The vessels, which form a pedicle, are covered with fine pinpoint fibrin masses, but present no adhesions. This pedicle is twisted four times, and its uncoiled state measures 17 cm. in length and 4 cm. in its greatest diameter. It is composed of several large tortuous veins surrounded by a considerable amount of fat, and several branches of the splenic artery, which is divided a considerable distance from the hilus. The pedicle also includes a soft body measuring 1.5 x 1 cm., covered with peritoneum, and resembling an enlarged lymphatic gland. At the distal end mixed gray and red thrombi protrude from the veins. The largest vein is 2.5 cm. in its greatest diameter. On incision they are filled with thrombi, but contain a variable amount of thin fluid blood. The thrombi are partly red and partly mixed, some being yellowish-gray in color. As a rule they are non-adherent, or can be easily separated. In general the mixed thrombi are soft or friable. The thrombi can be readily traced into the large and the small veins of the organ. SANFORD, DOLLET: SPLENIC ANEMIA. 11 The walls of the veins are not much thickened, and show no signs of recent or old inflammation except the thickening of one near the hilus. The splenic tissue is light brown in color and cedematous. The trabecule are well marked, the lymph nodes obscure. The tissue is not friable. There are no infarcts. Microscopic. Material from all organs was hardened in Zenker's and Orth's fluid, in formalin and in alcohol, and stained by ordinary and special methods. Lungs. There is moderate emphysema and anthracosis. The sections from the posterior portions show moderate congestive oedema. The bronchi and pulmonary arteries are normal. Careful search for the bone-marrow giant cells described by Warthin in his cases was negative. Heart. There is marked segmentation and some fragmentation of the muscle. Liver. The connective tissue is moderately increased, but not at all uniformly even in the same section. The increase is greatest in the right lobe under the capsule, as noted macroscopically, Glisson's capsule about the larger vessels being particularly thickened. The lobules are usually sharply defined, but occasionally the connective tissue extends somewhat into them. The connective tissue is dense fibrous in character, with few nuclei, and slight round cell infiltra- tion. In places of more marked fibrosis, fibroblastic cells are more abundant, but the tissue is nowhere actively proliferating. More rarely an increase of fibrous tissue about the central veins is noted. Mallory's connective-tissue stain shows here and there a slight increase of the reticulum of the lobules. This is more prom- inent in areas of atrophy of the lobular centres. This reticulum takes a light red with Van Gieson's stain. A very few fibroblastic cells appear in these areas and several capillaries are filled with proliferating endothelial cells with a few karyokinetic figures. Excepting these few instances endothelial proliferation does not occur. In the more cirrhotic portions many lobules show dilatation of the central veins and capillaries with some atrophy of the liver cells, but this is appreciable in but few lobules in the less fibrosed areas. The atrophic liver cells present more or less fatty degen- eration and necrosis, with a little brownish-yellow pigment. Numer- ous attempts to demonstrate the presence of iron were unsuccessful. Proliferation of bile canaliculi is found in only two areas. No bone- marrow giant cells are found. The stomach and intestines show chronic passive congestion, associated in the small intestine with slight chronic catarrhal inflammation. Pancreas. Sections from the head and body show a moderate increase of connective tissue, both perilobular and interacinous. Some islands of Langerhans also have an increased amount of 12 SANFORD. DOLLEY: splenic AN.EMIA. stroma. The sections from the tail show marked fibrosis, dense fibrous and hyahne tissue replacing many lobules, while the remain- ing ones are small and distorted. The ducts have a thickened wall and a few are tortuous. The islands of Langerhans are diminished in number. A few are unaffected. In some there is fibrosis, but in others fatty and granular degeneration of the cells is more pro- nounced. The veins are moderately congested and have thickened walls, particularly in the tail. The kidneys show a very slight chronic interstitial nephritis. There is no deposition of pigment. The portal vein is normal. ' Splenic Veins. The intima is shghtly and irregularly thickened and is composed of dense fibrous tissue with few nuclei. Thickening of the adventitia is more pronounced. In the latter, particularly about the vasa vasorum, there is moderate fibroblastic proliferation. But few fibroblastic cells occur in the intima and media. Organ- ization of the thrombi is well advanced in places. The mesenteric glands are fibrosed. There is some proliferation of the endothelium. Hcemolymph Glands. A few are of the marrow lymph type described by Warthin, but show partial transformation into or- dinary lymph glands. The lymph spaces are full of endothelial cells, some of which are phagocytic for red blood and lymphoid cells. Only a moderate number of bone-marrow giant cells appear. The majority of the haemolymph glands resemble splenic tissue in structure. The follicles contain no germinal centres and usually show small hyaline areas. The blood sinuses, which are much dilated, are only partly filled with blood. The most striking feature is the presence within them of numerous large cells of endothelial type packed with red blood cells. Mononuclear eosinophiles appear in great numbers, and normoblasts are moderately numerous. A few mastzellen are present. Numerous cells resembling myelocytes appear in the blood sinuses, but these are more abundant in the lymph spaces of the glands of marrow lymph type (Wright stain). Pigment is in moderate amount. It gives the iron reaction with ferrocyanide and HCl. Some phagocytic cells give the same reaction in a diffuse way. The bone-marrow (rib) is of a poor lymphoid type. Giant cells are not apparently reduced in number. Plasma cells and poly- morphonuclear leukocytes are in small proportion, but mono- nuclear eosinophiles with nuclei varying greatly in size are extremely abundant. Polymorphonuclear eosinophiles, while absolutely increased, are less numerous relatively. Normoblasts appear in large numbers with active mitosis. Myeloblasts are only occasion- ally seen. Neutrophilic myelocytes are almost entirely absent. A pecuhar feature with the Wright stain is the presence of cells SANFORD, DOLLEY: SPLENIC AN.EM1A. 13 Tesembling lymphocytes, but with a red protoplasmic border. Ex- •cluding these, ordinary lymphoid cells are not increased in number. There is a scanty amount of iron containing pigment, both in ■cells and free. Spleen. The capsule is moderately thickened and composed of a dense hyaline connective tissue. The trabeculse are also thickened and are farther apart than normal. The splenic tissue is con- ;structed of considerably dilated blood spaces, more or less filled with blood, and separated by a moderate amount of usually dense fibrous tissue with a varying number of fibroblasts. Under the •capsule this newly formed connective tissue is much more abundant, and it is not uniform in the other sections, thick fibrous bands appearing here and there. Hyaline transformation of the con- nective tissue appears nowhere outside of the trabeculse proper. 'The venous spaces are lined either by an almost flat or a somewhat swollen endothelium with an occasional free cell, but no proliferating ■endothelium appears. Scattered through the pulp are relatively few lymphoid cells and an almost equal number of polymorpho- nuclear leukocytes. The lymph nodes are farther apart than normal, but it is not possible to say they are absolutely diminished in number. They are usually not atrophic but show an increase of ■coarse reticulum with a few fibroblasts. Often the central arteriole is thickened and hyaline. The veins are greatly dilated, their walls Are moderately thickened and many contain fibrinous and mixed thrombi. Not infrequently small fibrinous clumps appear in the blood spaces. A very moderate amount of pigment is deposited in phagocytic ■cells in the blood spaces or in the stroma or lies free in the trabeculae. It gives the iron reaction. A few normoblasts, plasma cells, and mononuclear eosinophiles are to be seen, but neither myelocytes nor bone-marrow giant cells are found. With the Van Gieson stain there is considerable variation, but a large part of the newly formed stroma takes a deep-red color. A section throught one of the splenic veins removed at operation shows the same degree of fibrous thickening as in the stump. The adventitia is the seat of a much more marked fibroblastic prolif- eration. The thrombus is just beginning to organize. Pathological Summary. There is nothing in the histological structure of the spleen which cannot be explained simply by chronic passive congestion. In the causation of this, two factors have to be ■considered — i. e., the abnormality of the splenic vessels and the <;irrhosis of the liver. The dislocation of the spleen must have been of long standing, for the splenic vessels uncoiled measured 19.5 cm. from the end of the pancreas. Granting even that the multiple twists all occurred as a result of the accidents, which is hardly probable, the elongation which existed prior to this must have ■caused considerable interference with the splenic circulation. 14 SANFORD, DOLLEY: splenic AN.EMIA. In two cases of splenic anaemia, reported by Dock and Warthin,^ there were stenosis and calcification of the portal vein, a condition identical with this one as far as the effect on the spleen is concerned. The question they raise is whether the portal lesion is primary and the splenic fibrosis secondary to it, or, on the other hand, whether the splenic condition is primary, while the condition of the portal vein is coincidental or results from a toxic condition of the portal blood dependent upon the disturbed splenic function or dependent upon a portal or general intoxication. The dislocation of the spleen could not have been of congenital origin, for the alteration does not indicate such a long-standing condition, and an enlargement of the organ at once suggests itself as the main factor in its production. The abnormality of the splenic vessels is essentially a mechanical, not an organic one. Granted that some enlargement of the spleen, from whatever cause, started the elongation of the splenic vessels, once produced, with probably some torsion, the reaction on the circulation of the dislocated organ would be so great that the subsequently increased congestion and fibrosis would result largely from the condition of the vessels. The condition of the spleen must be considered then as largely secondary to that of the vessels, which gives them a very probable etiological significance. If splenic anaemia existed before the dislocation of the spleen, and the initial enlargement and the dislocation of that organ were manifestations of the pathological processes of the diseaise, the splenic vessels lose their casual relationship. But the histological structure offers no evidence of this. There is not even the prolifer- ation of endothelium so commonly found in the spleen in this disease to complicate the picture. The relationship of the cirrhosis of the liver to the condition in the spleen presents the greatest difficulty in interpretation. In part the process in the liver is regarded as a mild and not uniform chronic passive congestion, with a just beginning central cirrhosis, shown by the increase of fine reticulum in the atrophic portions of the lobules, with some endothelial proliferation. The moderate emphysema of the lungs offers a probable explanation for this condition. More important there is also a moderate though irregular increase of periportal tissue, but the histological pictures of the fibrosis in the liver and in the spleen are too nearly identical to draw conclusions regarding time relationship. What part the liver played in the production of the splenic condition it is impossible to say. Was the hepatic process primary, and was the initial enlargement of the spleen leading to its dislocation secondary to it, or were the two processes coincident? The fibrosis in the liver and in part that in the spleen may have resulted from a general or portal intoxication. The changes in the heemolymph nodes are interpreted as compen- satory for the failing splenic function. It is probable that the bone- SANFOKD, DOLLEY: SPLENIC ANEMIA. 15 marrow shared to some degree in the assumption of this function. But the presence of pigment and phagocytes in the spleen would, however, indicate that this function was not entirely lost, while the moderate evidence of old haemolysis in the haemal glands would further bear this out. But these glands show an excessive number of phagocytes full of intact red blood cells which is readily explained by the results of Warthin's work. He concludes from his experi- ments with sheep that haemolysis and leukocyte formation are the functions taken up by the haemal glands after splenectomy. But the haemolytic action of these glands exceeds that of the normal spleen, and the resulting anaemia is compensated for by increased activity on the part of the bone-marrow. The marked fall in the blood count after splenectomy was due then to this Excess of haemo- lymphatic haemolysis over that of splenic. No iron pigment could be demonstrated in the liver and kidneys. This fact, together with the paucity of megaloblasts in the bone- marrow, indicates that the anaemia was not primary in type. While the hyperactivity of the hsemolymph glands would explain the anaemia which undoubtedly existed before the accidents, the intestinal hemorrhages following them reduced the patient to his state on entering the hospital. Pathological Diagnosis. Dislocation of the spleen with marked elongation and torsion of its vessels, and thrombosis. Chronic passive congestion and fibrosis of the spleen. Interlobular atrophic cirrhosis of the liver. Chronic passive congestion with beginning central cirrhosis of the liver. Chronic passive congestion of rest of portal system. Compensatory hyperplasia of haemolymph nodes. Hyper- plasia of the bone-marrow. Secondary anaemia. Moderate diffuse emphysema of the lungs. Tuberculosis of bronchial and tracheal glands. My thanks are due to Dr. W. T. Howard, Jr., for the use of material and for his interest in the work. 3. General Summary. The patient's previous history of pallor, digestive disturbances, and pain in the splenic region, symptoms of which there is a history for a number of years, strongly suggest the early stage of a splenic anaemia. The frequent urination and the pain in the splenic region also would imply a long-standing dis- location of the spleen. This is corroborated by the enormous elongation of the splenic vessels. Whether or not all the twists in the splenic veins occurred at the time of the two accidents, it is impossible to state, but it is at least more reasonable to suppose that from their multiple character they did not all occur at one time, but that some had existed previously. The thrombosis of the splenic vessels must have been the cause of the sudden enlargement of the spleen while the patient was in the hospital. The lack of organization in the thrombi of the vessels removed at the splenectomy proves this. 16 sANFOED, dolley: splenic anemia. If Banti's stage of splenic ansemia is limited only to those cases in which the cirrhosis is a terminal manifestation we cannot include our case in that category, for the cirrhosis of the liver is certainly of as long standing as the fibrosis in the spleen. We prefer to regard the case as one of splenic ansemia associated with a cirrhosis of the liver, not as a result of the process in the spleen, but rather accompanying it. The enormous elongation and torsion of the vessels associated Tvith the dislocation of the spleen possess a very probable etiological rsignificance in this case. ilReprinted from The American Joubnal of Obstetrics and Diseases of Women and Children, Vol. XLIV., No. 2, 1901.] PUS IN ABDOMINAL OPERATIONS.' HUNTER ROBB, M.D., Professor of Gynecology, Western Beserve University; Gynecologist-in-Chief to the Lakeside Hospital, Cleveland, O. The mortality following operations for suppurative disease of the tubes and ovaries is variously estimated as from 8 to 20 per cent. In the face of these figures it is certainly incumbent upon us not to rest satisfied with present conditions, but ever to be looking for more perfect methods and to adopt more strin- gent precautions in our operative procedures. The most perfect degree of asepsis obtainable before, during, and after these operations, must always be insisted upon. Whenever pus is met with at the time of the operation, if a fatal result occurs, we are too apt to content ourselves with the explanation that infective material was already present and that its spread was only a natural result. As a matter of fact, however, it must not be forgotten that the pus met with under such circumstances, as a rule, is free from virulent bacteria; and whenever an infection follows a "pus operation" and no organisms can be demon- strated in the secretions encountered, it is not only possible, but even highly probable, that in the majority of these instances the fatal results have been due to the introduction of some septic material during the operation. Again, as a result of my observations during the past seven years, I have become convinced that operators not infrequently err in carrying out radical abdominal procedures, when the pa- tient's resistance is in such a lowered condition that she is very apt to succumb to the shock of the operation per se. Such a con- dition must always be given careful consideration when deciding for or against operative interference during an acute attack of localized or more or less generalized pelvic peritonitis. Believing that this factor has a very important significance in influencing our results, I have made it a rule, during the acute stage of a pelvic abscess, to defer an operation while the patient's condi- ^Read before the American Gynecological Society at Chicago, May 31, 1901. Copyright, William Wood & Companr. 2 ROBB: PUS IN ABDOMINAL OPERATIONS. tion is improving. In the meanwhile the patient is kept perfectly quiet on her back in bed, and heat, in the form of flaxseed poul- tices or turpentine stupes, is applied to the abdomen. In addi- tion, a vaginal douche of a gallon of a warm one per cent solution of carbolic acid or a saturated boric acid solution is given twice daily. For nourishing the patient we depend upon nutritive injections entirely for several hours. Anodynes and heart stimu- lants are given if necessary. As a result of these measures we have found that, in the great majority of cases, the acute symp- toms will subside within a few days and we can then operate upon the patient, who is now in a much better condition, either by the abdominal or the vaginal route or by the combined pro- cedures. If, however, no improvement takes place under the above treatment within a reasonable time (eight to ten hours), and if we can make out a pelvic mass, we puncture through the vagina, and, after irrigating the sac or the pelvic cavity, pack with sterilized gauze. We have noticed that a number of well- known general surgeons in this country are treating eases of acute appendicitis in a similar manner. The micro-organisms that are most frequently met with in cases of suppurative disease of the tubes and ovaries are Staphylococcus pyogenes aureus, Streptococcus pyogenes, Gono- coccus, and Bacillus coli. Other forms are occasionally met with, such as the tubercle bacillus and Proteus Zenkeri.^ When we are able to demonstrate the presence of Staphylococcus pyo- genes aureus at the time of the operation, we expect an unin- terrupted convalescence in the majority of these instances. The organism most to be feared is Streptococcus pyogenes, and its presence always makes the prognosis very grave. We recall in particular an instance in which at the time of the operation macroscopical evidences of a small amount of pus were demon- strable, but no organisms could be shown from the pus at the examination made at this time. The culture tubes that were in- oculated at the time of the operation, however, subsequently showed a profuse streptococcus growth. We drained in this case, but, despite our efforts, a streptococcus peritonitis developed and the patient died. The same organism was found at the autopsy. We unfortunately infected another patient with this organism although before the second operation we carefully sterilized our hands, instruments, etc., in the usual manner. We re- moved without much dififlculty the adherent tubes and ova- ries on both sides. Cover-slips made at the time of the op- 'Johns Hopkins Hospital Bulletin, No. 70, January, 1897. EOBB: PUS IN ABDOMINAL OPERATIONS. 3 eration were negative. In two days, however, this patient be- gan to show evidences of a peritonitis, and a fatal result took place a day and a half after the death of the first patient. Streptococcus pyogenes was found, at the time of the autopsy, in the exudate present in the peritoneal cavity. In the first case there was a history of an induced abortion at the third month of pregnancy with subsequent fever six years before admission to the hospital. This suggested a streptococcus infection at that time. In the second case the patient had had two normal labors without any complications. Such a stern lesson taught us once for all that even in cases in which there are only adherent organs, even when no pus is apparently present, we should always be on our guard against carrying infective material from one case to another. From the fact that we found no organisms present by the cover-slip ex- amination at the time of the operation, however, we felt that we were justified in going on with the second operation. But for several years now, after having had such an experience, we do not perform more than one abdominal section in the same day, preferring to operate much more frequently rather than run risks. I certainly believe that the adoption of this precaution has saved a certain number of cases from being infected. I should say that, even prior to this regrettable accident, when- ever we encountered a pus case we always postponed the carry- ing out of any further abdominal work, except in emergencies, until forty-eight hours had elapsed. The question of using drainage in pus cases is always of in- terest, and all operators are not in agreement as to its advan- tages and disadvantages. In my clinic we have not used a drain- age tube during the past seven years, and we now, as a matter of fact, seldom drain, even in pus cases, through the abdominal incision. Occasionally, when there has been a septic area left behind, we drain per vaginam, but even in these cases I do not feel that this procedure is by any means always necessary. It has always seemed to me that we do not sufficiently protect from infection during the operation the portions of the pelvic contents which are apparently intact. Inasmuch as pelvic in- flammatory disease frequently involves the intestines to a wide extent, it is sometimes impossible to prevent the infectious ma- terial from being spread more or less during the manipulations of the diseased structures. We can, however, to a great degree limit the free distribution of this purulent material throughout the abdominal cavity. In order to effect this, as soon as the ab- 4 ROBB: PUS IN ABDOMINAL OPERATIONS. domen is opened we are in the habit of placing large gauze sponges high up in the flanks on either side, the patient being kept in the horizontal position while the masses are being enu- cleated. If we break into a pus sac during our manipulations, we at once make cover-slips of the escaped pus, and while these are being examined we attempt, as far as possible, to remove the material that has escaped by mopping it up with gauze sponges. After this we wash out the abdominal cavity with large quantities of sterile salt solution, in order to get rid of as much as possible of the pus and to dilute any that may remain. The enucleated mass is now surrounded vrith gauze moistened in a 1 : 1000 bi- chloride solution, which is not removed until the structures have been cut away. The pedicle is thoroughly cauterized with the Paquelin cautery, and the abdomen is again washed out with salt solution, sponged dry, after which 300 to 500 cubic centimetres of warm sterilized salt solution are introduced and left there. The incision is closed without drainage. The following is the clinical and bacteriological analysis of 72 consecutive, unselected abdominal sections for suppurative diseases of the tubes and ovaries, two deaths occurring : ANALYTICAL REPORT OF SEVENTY-TWO CONSECUTIVE PUS CASES.^ Age.— The oldest patient was 41, the youngest 17, the aver- age age being 26.37. Fifty were married, 16 were single, 6 were widowed. OccMpa^iow.— Housework, 43; prostitutes, 11; dressmaking, 3; cooks, 2; canvassing agents, 2; seamstresses, 2; boxmaker, 1 ; dining-room girl, 1 ; clerk, 1 ; domestic, 1 ■, laundress, 1 ; cigar packer, 1 ; match factory girl, 1 ; pianist, 1 ; telephone clerk, 1. In 23 pregnancy had not occurred. In 13 miscarriages alone had taken place. In 8 full-term births alone had occurred. In 28 there was a history of births at term and miscarriages. The highest number of miscarriages in any one case was 5. The highest number of births at term in any one case was 7. The average number of miscarriages in the 41 cases having mis- carriages was 1.94 per cent. The average number of births at full term in the 36 cases having births at term was 2. Thus it will be seen that in 49 of the 72 cases the number of births at term was 2. These figures go to show that the accidents inei- ' The first case of this series was operated upon October 11, 1898, and the last one February 28, 1901. Since the analysis of these cases we have had 15 additional instances, making the number 87, but a sufficient inter- val of time has not elapsed from which to draw conclusions. EOBB: PUS IN ABDOMINAL OPERATIONS. 5 dental to labor and abortion have to be carefully considered as factors in the production of pelvic disease. Infection.— '^Q were able to establish a gonorrheal history in 14 cases (probable in 4 others). Infection after labor at term had occurred in 7 cases (2 of these cases also gave a history of a previous gonorrhea) ; infection after miscarriage in 17 cases (4 of these patients giving also an old gonorrheal history). In the remaining cases (31) no definite history of a specific infection or any relation between the disease and labor or abor- tion could be made out. In many cases there was a gradual onset with exacerbations of the symptoms at the men- strual period, but with no definite acute attack before ad- mission into the hospital. In many the symptoms had per- sisted for long periods of time, in 1 case for 20 years. It was frequently noted that an attack of peritonitis had occurred be- fore the patient was admitted into the hospital. In 33 cases these attacks had lasted under three weeks, and for the most part they were of about two weeks ' duration. The shortest attack was one of two days immediately preceding admission. In eases of more than three weeks' duration various periods of time were represented — six weeks, two months, six months, a year, etc., with no special uniformity. Definite and similar previous attacks had occurred in 7 gonorrheal eases, the largest number of previous attacks being 3, except in 1 case in which there was a history of a great many. From this it will be seen that the infec- tions following gonorrhea on the one hand, and labor or abortion on the other, are about equal in number. Thus, in 14 cases (19.44 per cent) we were able to get a positive his- tory of a previous gonorrheal infection, and in 20 eases (27.77 per cent) there had undoubtedly been an infection following labor or miscarriage. Allowing, however, for those cases in which there had been a previous gonorrhea as well as a history of an infection following labor or abortion, and placing these doubtful cases in the column of the gon- orrheal infections, we have the following figures: Gonorrheal cases 23 (?), or 33 J/^ per cent of the cases; infections after labor and miscarriage, 20, or 27.77 per cent of the cases. Here, as has been said, we have added to the gonorrheal list by taking six eases which would seem to rather come under infections fol- lowing labor or miscarriage. It is, of course, often extremely difficult to feel sure that a patient has had a specific vaginitis, unless cover-slip examinations have been made of the secretions at the time of the supposed infection. In many instances it is EOBB: PUS IN ABDOMINAL OPERATIONS. not difficult to obtain a history of a vaginitis, but to prove that it has been specific in origin is not always easy. Many of the mis- carriages had been criminally produced. The infections in these cases could generally be traced to the production of the abortion. The temperature, pulse, and respiration before operation were as follows : Highest temperature, 105.2° F.; pulse, 142; respiration, 42. Lowest maximum temperature, 98.18° F. ; pulse, 82; respiration, 22.18. Average maximum temperature, 102.09° F. ; pulse 106.8; respiration, 27.2. The chief clinical symptoms were as follows, many patients- showing several: Pain in lower part of the abdomen was present in 67 cases. Backaclie " Leucorrhea " Dysmenorrhea " Headache " Dysuria " Bearing-down pains were Menorrhagia was Constipation " Gastralgia " Chilly feelings were Metrorrhagia was Cough " Painful defecation " Nausea and vomiting were The most prominent symptoms were pain in the lower abdo- men, backache, a leucorrheal discharge, dysmenorrhea, head- ache, and painful micturition. Pus was found as follows : In the ovary, unilaterally, 24 times ; bilaterally, 5 times, including tuberculous cases. In the tube, unilaterally, 25 times; bilaterally, 39 times, with tuberculous cases. In the appendix, in 1 ease. In the walls of the uterus, in 1 case. In almost every case in which the pus was unilateral, the other tube and ovary were found to be adherent. Infection of the tube and ovary together (tubo-ovarian abscess) was noted unilaterally only and in 17 eases. The tubes are more liable to a bilateral infection ; the ovaries to a unilateral infection. The following operations were performed, all being carried out at the Lakeside Hospital : Appendicectomy 26 Dilatation and curetting 24 Evacuation and drainage of pus sac, when removal was Impossible 3 • 43 ' 34 ' 30 ' 21 ' 13 ' 8 ' 5 ' 6 ' 2 ' 2 ' 3 ' 1 case. ' 1 " ' 2 cases EOBB: PUS IN ABDOMINAL OPEEATIONS. 7 Myomectomy 3 Partial resection of ovary 2 Perineorrhapliy 2 Salpingectomy, unilateral' , 17 Salplngo-oophorectomy, unilateral 20 Salpingo-oophorectomy, bilateral 46 Supravaginal hysterectomy 1 Suspension of the uterus 9 "Vaginal puncture 10 Total number of operative procedures carried out upon the 72 patients 163 • Wlieneyer tlie tube or ovary was simply adherent, but not disorganized, the adhesions were separated and the structures allowed to remain. Despite this conservative surgery, in not a single case up to the present time have there been complaints from these patients, nor has there been any necessity for fur- ther operative procedures. As a supplementary procedure ap- pendicectomy was carried out 26 times, or in 36.1 per cent of the total number of cases. We believe that, in cases of suppurative disease of the tubes and ovaries, in about one-third the appendix will be found to be adherent. In those cases in which vaginal puncture was carried out, we found it impossible to thoroughly remove the abscess wall by the abdominal route, and in such instances we employed drainage per vaginam for a few days. In 3 cases, after opening the abdomen, it was found impossible to remove the pus sacs owing to the patients' weakened condition, and we were therefore obliged to drain through the vagina. Besults.—2 patients died, 70 recovered. Mortality for all cases, 2.77 per cent. Morbidity: So far as we have been able to ascer- tain, there are no patients in this series that make any com- plaints referable to disorders in the pelvis. Drainage.— In 9 cases drainage was carried out through the cul-de-sac alone; in 2 cases through the abdominal wall alone; in 1 case through the cul-de-sac and abdominal wall combined. In 2 cases in which infection occurred subsequently to the original operation, the abdomen was opened and drained. One of these patients died, the other recovered. Thus abdominal drainage was carried out in 2 instances only immediately fol- lowing the operation, or in 2.77 per cent of the cases. In 10 cases (13.6 per cent) drainage was carried out per vaginam. 'In these cases the opposite tube and ovary were removed in all but one case, in which this procedure was impossible, drainage being used after evacuation of the pus. . 8 ROBB: PUS IN ABDOMINAL OPERATIONS. Suppuration of the abdominal wound occurred in 12 cases, or 18.7 per cent. In 2 of these abdominal drainage was employed at the time of the operation. In 2 abdominal drainage was insti- tuted several days'after operation, 1 of these patients recovering. In the remaining 8 eases primary union occurred on the surface, but pus developed deeper down, usually about ten to fourteen days after the operation. Wherever there occurred a secretion from the wound in which it was possible to demonstrate micro-or- ganisms, we classified the ease as one of infection. In the secre- tion from the incision in 6 of the 12 cases we were able to demon- strate Staphylococcus pyogenes aureus and in the other cases Staphylococcus pyogenes albus. In this series of cases we con- sider that the infections above referred to were to some extent due to the fact that we had previously operated upon a case of post-puerperal infection in which Staphylococcus pyogenes aureus was found at the time of the operation. In six of the abdominal sections performed during the two weeks foUomng (although we left an interval of four days) the incision became infected to a slight extent, and in the sero-sanguinolent secretion we were able to demonstrate Staphylococcus pyogenes aureus. In the tubes and ovaries the following organisms were found: Staphylococcus pyogenes aureus, 2 cases; Staphylococcus pyo- genes albus, 1 case ; Bacillus coli communis, 1 case ; Streptococcus pyogenes, 2 cases ; Friedlander 's bacillus, 2 cases ; Bacillus tuber- culosis, 1 case; Gonococcus, 7 times positive (3 others probable). Thus in 12 cases (16 per cent) organisms were found. We were not able to cultivate the gonococcus, and the diagnosis was made entirely from the cover-slip examinations from sections of the tissues. The diagnoses at the time of the operations were as follows (these differ somewhat from the pathological diagnoses, since a considerable number of cases of apparent salpingo-oophoritis proved on further examination to be pus tubes) : Abortion, retained secundines. 1 Endometritis, Septic 1 Appendix, abscess of 1 " and stenosis 1 " adherent 11 Follicular hypertrophy of ovary 6 " concretions in 2 Hydrosalpinx 1 cystic 1 Lacerated cervix 1 " flexure of 2 Myoma J (periappendicitis) . 6 Ovarian abscess, unilateral'... 1 Endometritis 24 " " bilateral 1 'Seventeen cases of tubo-ovarian abscesses have been classified under pyosalpinx and ovarian abscess; in each instance the abscess was uni- lateral, with a pyosalpinx upon the other side in 12 cases. EOBB: PUS IN ABDOMINAL OPERATIONS. 9 Pelvic abscess, probably pyosal- Relaxed vaginal outlet 2 pinx 1 Retroversion 9 Perforation of uterus 1 Salpingo-oophoritis, unilateral" 21 Phthisis 2 " bilateral . 19 Pyosalpinx, unilateral 20 Vulvo-vaginitis 1 -Pyosalpinx, bilateral' 32 Perhaps a few words about the two fatal cases may be of in- terest. In the first case, in which there was a uterus septus par- tialis, a hystero-salpingo-oophorectomy for a tubo-ovarian ab- scess involving the uterine cornu, with appendicectomy, was performed. A considerable amount of pus escaped into the pelvis at the time of the operation, although the adjacent ■structures were protected by gauze pads. The previous history suggested the possibility of a streptococcus infection, since her illness had followed a miscarriage which had occurred ten weeks before the patient was admitted to the hospital. The temperature at the time of her entrance ranged between 100° -and 102° F., but for the most part was under 102° F. At the time of the operation we were not able to demonstrate any or- ganisms by cover-slip examination in the pus that escaped. "We accordingly followed our usual plan of washing out the abdomen and closing without drainage. In eight hours the patient's tem- perature rose to 104° F. and the pulse to 148. She was, how- ever, feeling comfortable. The next day her temperature fell to 101.5° and her pulse to 128. For the succeeding twenty-four hours her symptoms were on the whole favorable, and, with the exception of some difficulty in respiration, she seemed to be pro- gressing satisfactorily. Her temperature varied during the next three days between 101° and 103.8° F. ; at the same time there was marked dyspnea and the pulse gradually increased in rapidity. She finally died on the fourth day following the operation, with- out having shown any marked evidences of any peritoneal in- volvement, as there was no nausea, vomiting, or tympany and the bowels had been thoroughly well opened. Autopsy showed the ease to be one of pelvic abscess with gen- eral sepsis. Cultures made from the lung showed a pure strep- tococcus pyogenes infection, the same organism being also found in the liver, pleura, and peritoneum. In this case, then, we had to do with a streptococcus infection which had as its origin either 'This includes 3 tuberculous cases. There were 3 tuberculous cases -of bilateral pyosalpinx; in 1 of these there was also a bilateral tubercu- lous abscess of the ovary. "Many showed pus on microscopic examination. 10 EOBB: PUS IN ABDOMINAIj OPERATIONS. the introduction of this organism at the time of the operation or the miscarriage which occurred some months previous to the time of the operation. From my previous experience I am of the opinion that in this ease a fatal infection would have fol- lowed even if drainage had been used. In the second ease we were not able to establish a history of an infection following an abortion or labor. Here we had to deal with an abscess involving the left tube and ovary, and an adherent right tube. We removed the diseased structures and made a cover-slip examination of the pus, but no organisms were demonstrated. We also inoculated culture tubes. The ab- domen was closed. The patient's temperature the day following the operation rose to 104.4° F., and for the next two days varied between 102.5° and 105.7° F. On the tubes that were inoculated from the pus streptococci were found. On the second day after the operation the abdomen was reopened under cocaine anesthesia and washed out, drainage being afterward instituted. The ab- domen was irrigated a second time and salt infusions were car- ried out at stated intervals. She died on the third day. Cultures made from the lung at autopsy showed the Diplo- coccus pneumoniae and Bacillus coli communis; in the liver, Bacillus mucosus capsulatus was found, and in the peritoneum Diplococcus pneumoniffi. In this case, then, streptococci, though not demonstrable by cover-slip examination at the time of the operation, appeared in the culture tubes two days later. At the time of the autopsy only Diplococcus pneumoniffi was found in the peritoneal cavity. In conclusion I wish to thank my former associate. Dr. Wil- liam H. Weir, and my present associate. Dr. C. D. Williams, for kind assistance in making up this analj^tical report. 702 Rose Building. THE MOETALITY FOLLOWING OPBEATIONS FOE PUS IN THE PELVIS. HUNTER ROBB, M.D. Professor o( Gynecology, Western Reserve University, and Gyne- coIoglst-in-Chief to the Lalseslde Hospital. CLEVELAND, OHIO. In the course of his everyflay reading, as well as in his own operative work, the busy surgeon is always meet- ing with new data, so that at the end of a certain time he finds his mind loaded up with a mass of undigested knowledge, much of which is of no practical utility, or even apparently contradictory. Under these circum- stances he finds it necessary to pass in review from time to time all these accumulated ideas, and to sift out the chaff from the wheat so that he can once more start with a clear field and find himself in a position to answer the questions: 1. What have I learned lately which I can utilize in improving my surgical Judgment ? 2. What omissions from or additions to my operative technic will give me better results in the future ? Prob- ably the most satisfactory way by which he can attain this end is to make a careful classification of all his cases and as soon as he finds himself in possession of a sufficiently comprehensive series exemplifying a given pathologic condition to make a careful analysis of all the cases, and from the various data thus obtained to draw any conclusions that seem to be warranted, discard- ing what has proved to be worthless, and reserving fqr further consideration any points which may still be doubtful. It goes without saying that deductions based on a few isolated cases would be worse than useless and might be altogether misleading, and that the series should present instances of approximately all the vari- ous complications which may be met with in this par- ticular class of operative cases. During the past three years we have had in our clinic two series of unselected abdominal sections without a death. In the first there were 114 and in the second 108 cases. In the two together we encountered prac- tically all the usual forms of pelvic lesions that occur in the female generative organs. Prom these cases, to- gether with some prior and other subsequent instances, we have been able to make up a series of 100 consecutive abdominal operations for pus in the pelvis with two deaths. An analytical report of 73 of these cases has already appeared in a paper read before the American Gynecological Society in Chicago, May 30, 1901. Since the date of this publication we have had 28 additional instances. To-day I will briefly consider some of the bacteriologic questions connected with such pus eases, together with some points in the operative technie. So far as I can gather from my reading, many sur- geons would appear to still hold to the belief that the finding of pus in the Fallopian tubes and ovaries in a given ease affords sufficient ground for assuming that the patient's chances of recovery following radical operative procedures are necessarily bad. So gloomy a nrognosis, I feel sure, when based on the presence of pus per se, is by no means always warranted, and the general con- dition of the patient, as well as her capacity or inability to react after the shock of a severe surgical operation, are often of paramount importance. This is liable to be the case more particularly when dense adhesions have to be separated. Again, when pus is already present, not a few operators appear to assume ithat the carrying out of an aseptic technie, so far as' they themselves are con- cerned, becomes no longer necessary, and when a fatal result follows an operation under these conditions or when the abdominal wound subsequently becomes in- fected, they are too apt to content themselves with the explanation that infective material was already present and its spread was only a natural result. Against such conclusions, however, we have the evidence of repeated bacteriologic examinations to prove that the pus found under such circumstances seldom contains living organ- isms. This observation has been demonstrated in my clinic in a large number of instances, in which no growths were obtained on the culture media, and the fact that the micro-organisms are dead undoubtedly plays a large part in contributing to the favorable re- sults which often follow operative procedures. There is another point of interest connected with the treatment advisable in these cases, namely, the influence of the employment of palliative measures to the acute stages of purulent invasions of the Fallopian tubes and ovaries. I have long been convinced that it is an unwise routine to proceed at once to radical steps during an acute attack of pelvic abscess. Even if the patient does not improve somewhat in a short time under general hygienic meas- ures and external applications together with the exhibi- tion of suitable drugs, by simply opening up the cul-de- ssLC, in the great majority of instances the pus can be given an outlet through this channel, and the absorption of the toxic substances present in the pelvis can be diminished so that as a result the patient's condition will generally become better. She will then be able to with- stand the shock if it should later become necessary to institute radical abdominal measures for the removal of the pus-sac contents. After having tried this plan of treatment in a considerable number of cases, and com- paring the results with those following immediate radi- cal operative procedures during the acute stage of pelvic abscess, I have become convinced of the great ad- vantages of the former method in the majority of instances. The micro-organisms that are most frequently met with in cases of suppurative disease of the Fallopian tubes and ovaries are Staphylococcus pyogenes aureus. Streptococcus pyogenes, Gonococcus and B. coli. Other forms are occasionally met with, such as B. tuber- culosis,^ Proteus zenkeri. Of these organisms the one which is the most to be feared is undoubtedly Strepto- coccus pyogenes. But although its presence always makes the prognosis grave, it must be borne in mind that there are many instances on record in which this organism has been found during abdominal operations, and yet the patient has recovered. This favorable out- come is to be explained by the wide range of virulence possessed by Streptococcus pyogenes. In my experience a fatal peritonitis has sometimes followed operative pro- cedures in which there were no macroscopic evidences of pus at the time of the operation, and I am inclined to be- lieve that in such cases, with adherent lateral structures, the prognosis is often worse than when a considerable quantity of pus is met with. For this reason, and also because one can never be absolutely sure that the organ- ism may not be very virulent, I have for a long time followed the rule of not performing more than one abdominal section on the same day, preferring to have more frequent sessions rather than to run risks. When- ever I find pus t always postpone carrying out any fur- 1. Recently my assistant, Dr. Howard H. Dlttrlck, has been able to grow the tubercle bacillus from the pus of a tuberculous sal- pingitis. ther abdominal work, except in emergencies, nntil 48 hours have elapsed. Drainage. — The question of drainage in pus cases is always of great interest, and all operators are by no means in agreement as to its advantages and disadvan- tages. In my clinic we have not used a drainage tube dur- ing the past seven years, and we now, even in pus cases, seldom drain through the abdominal incision. When should we and when should we not carry out drainage in pus operations ? As has already been suggested, the pus that is found in the Fallopian tube and ovary is in the great majority of instances sterile, and if this is the case, what advantage is to be obtained by the use of drainage? Again, if the pus encountered in a given case contains virulent organisms, will the use of drain- age prevent a spread of the infection after a complete operation? After a thorough trial of both procedures, I have been forced to the conclusion that no form of drainage will prevent the occurrence of a certain amount of peritonitis if the organism present is virulent. There is, however, a class of cases in which drainage may pos- sibly be indicated, namely, those in which it has been impossible to remove the adherent pus sac or structures completely.. In these instances we employ for the pur- pose gauze sponges, which are introduced through the abdominal incision and carried out through the vagina. If the operation has lasted for a considerable length of time and the patient has been in a more or less marked condition of chronic sepsis, the drainage is kept up for from 24 to 48 hours. I must say, however, that I do not feel altogether sure that even under these cir- cumstances the procedure is always absolutely called for. In passing, I might say that the only other cases which in my judgment call for drainage are those in which it has been impossible to control the bleeding that occurs after the separation of dense adhesions, or in which, during the separation of adhesions, a rupture of the bowel has taken place, and it has been an exceed- ingly difficult case to close the rent. In the former case the gauze is used more for the purpose of compression, while in the latter it serves to protect the bowel, and should a leakage of the contents take place they can find an avenue of escape by means of the opening made in the cul-de-sac through which the distal end of the gauze sponge is carried into the vagina. On the other hand, if we have not drained and symptoms should arise after an abdominal section, as the result of the ab- sorption- of septic material that has been introduced at the time of the operation, or from remnants of the re- tained diseased foci, we have plenty of time, as a rule, to reopen the abdomen and wash it out again, and then, if necessary, institute drainage in order to prevent the absorption of more septic material. Such cases are, however, fortunately very rare, and if we employ drain- age as a routine in order to anticipate such a condition we may run many risks of infecting the patient. THE TECHNIO TO BE EMPLOYED IN PUS OPERATIONS. The teehnic to be carried out in operations for pus in the pelvis differs somewhat from that carried out in ordinary abdominal operations, and although the pus met with under these circumstances iff, as a rule, sterile, practical experience has shown it to be a wise precaution to protect as far as possible from infection the portions of the pelvic contents which are apparently intact. In order, therefore, to limit the free distribution of noxious material throughout the abdominal cavity, it is our prac- tice, as soon as the abdomen has been opened, to place large gauze sponges high up in the flanks on either side. If during our manipulations a pus sac is ruptured, we at once lower the patient and then make coverslips of the escaped pus, and, while these are being examined by an assistant, we try to remove as much as possible by mop- ping it up with gauze sponges. The pelvic cavity is then irrigated with large quantities of sterile salt solution in order to get rid of the greater part of the septic material and to dilute thoroughly any that may still be left. The enucleated mass is now surrounded with gauze which has been moistened in a solution of 1-1000 warm bi- chlorid of mercury. This gauze is not removed, if it can possibly be kept in place, until the structures have been cut away. If we unexpectedly encounter a pus sac before we have had an opportunity of applying our gauze protection pads, and if the contents escape into the pelvic cavity, the? patient is at once lowered to the horizontal posture, and the pelvic and abdominal cavities are irri- gated with hot sterile salt solution. The excess having been sponged out and the gauze protection pads placed in position, we then proceed with the enucleation of the mass. The pedicle is thoroughly cauterized with the Paquelin cautery, the abdomen is again washed out with salt solution and sponged dry, after which 300 to 500 c.c. of warm sterile salt solution are introduced and left in the cavity. The incision is closed without drainage. For 24 hours following the operation the lower end of the bed is elevated. . The situation of the pus in this series of 100 cases, 6 ' the number of eases 'in which we were able to demon- strate organisms, and the classification of these organ- isms may be of some interest. The pus was found as follows: In the ovary, uni- laterallyy 30 times ; bilaterally, 6 times, including tuber- culous cases. In the tube, unilaterally, 36 times; bi- laterally, 52 times, with the tuberculous cases. In the appendix once. In the walls of the uterus, once. In almost every case in which the pus was unilateral the tube and ovary on the other side were found to be ad- herent. Infection of the tube and ovary together (tubo- ovarian abscess) was noted, unilaterally, 21 times, _and bilaterally twice. The tubes were more liable to a bilateral infection;- the ovaries, to a unilateral in- fection. In the tubes and ovaries the following organisms were found: Staphylococcus pyogenes aureus, 3 cases; Sta- phylococcus pyogenes alhus, 3 cases; B. coli com- munis, 4 cases ; Streptococcus pyogenes, 4 cases ; Fried- landers hacillus, 2 cases; B. tuberculosis, 2 cases; Gonococcus, 8 times positive (six others probable). Thus in 24 cases (24 per cent.) organisms were found. We were not able to cultivate the gonococcus, except in one instance, and the diagnosis was made in the other cases entirely from the coverslip examinations of sec- tions made from the tissues. In the two cases in this series which terminated fa- tally, we had to deal with a streptococcus infection, as was demonstrated by the growth on culture media. In the first case there was a previous history of infection following a miscarriage which had occurred ten weeks before the patient was admitted to the hospital. At the time of the operation we were not able to demonstrate by coverslip examination any organisms in the pus that escaped. No drainage was employed.. The infectious material may possibly have been introduced at the time of the operation, although it is quite likely that the process dated from the time of the miscarriage. In the second case the coverslip examination of the pus re- vealed no organisms, but on tubes inoculated from the pus streptococci were found. The abdomen was re- opened on the second day after the operation under co- cain anesthesia and'^ washed out, drainage being after- wards instituted. The patient died on the third day. Cultures made from the lung at autopsy showed Diplo- coccus pneumonice and B. coli communis; in the liver, B. mucosus capsulatus was found, and in the peritoneum Diplococcus pneumonice. In this case, then, streptococci, though not demonstrable by coverslip examinations at the time of the operation, appeared in the culture tubes two days later. At the time of the autopsy only Diplo- coccus pneumonicE was found in the peritoneal cavity. CONCLUSIONS. A careful analysis of this series, therefore, seems to justify the following conclusions; 1. The mortality following operations for suppurative diseases of the tubes and ovaries can be kept under 5 per cent. 2. The death rate is largely influenced by a, tha viru- lence of the specific organisms present ; b, the individual resistance of the patient; c, the time and manner of carrying out the operative technic. 3. The micro-organism most to be feared is Strepto- coccus pyogenes^ but it must bo borne in mind that this organism varies considerably in virulence. 4. Abdominal drainage following operations for pus in the tubes and ovaries is seldom called for, as the or- ganisms are generally dead. Drainage becomes necessary only when it has been found impossible to remove the suppurative structures, or where perforation of the bowol from the separation of dense adhesions is to be feared. Under these circumstances the best route is by the va- gina. • 5. The employment of sterile salt solution for irrigat- ing the pelvic cavity will satisfactorily remove the pus or its products and the filling of the abdomen with salt solution will dilute and promote the rapid absorption of any inflammatory products that may be left behind. 6. The elevated position for 34 hours following the operation, with the abdomen filled with salt solution, tends to prevent the intestines and omentum from com- ing in contact with the immediate field of operation, and as a Result adhesions are not so likely to form between the viscera and the incised surfaces. 7. Should symptoms of infection follow the closing of the wound in pus cases, we have, as a rule, sufficient time to reopen the abdomen and wash out the infective- ma- terial that may have been left behind, or that may have been introduced at the time of the operation. 8. Operations for pus in the tubes and ovaries from the standpoint of the pus per se are ijot surrounded by more danger, as a rule, than those in which a purulent focus is not present. Reprinted tkom The Jouknai- oi' the American Medical Association January 17, 1903 [Reprinted from American Gynecology, June, 1903.] THE VAGINAL INCISION IN SEPSIS FOLLOWING ABORTION.* By Hunter RobBj M.D., Professor of Gynecology, Western Reserve University, and GynecoIogist-in-Chief to the Lakeside Hospital, Cleveland, Ohio. I believe that statistics justify the two following statements: (i) Of all married women who die between the ages of 20 and 40 nearly 10 per cent succumb to a puerperal infection; (2) At the present day, in private practice at least, the mortality from child-bed fever is nearly as high as it was 40 years ago. Such a confession is at once astounding and humiliating, and we must of necessity come to the con- clusion that the advantages of aseptic and antiseptic procedures are either chimerical or that the belief in them is still to a large extent more theoretical than practical. But if the mortality is still as high, what shall we say of the mor- bidity? No reliable statistics dealing with this subject are at hand, and how many cases of septic infection are still classed as instances of milk fever, rrialaria, nervous chills and the like, remains uncertain. Again it would appear that criminal abortion is very widely practiced at the present day, and certainly it is a matter of surprise that the gross ignorance and recklessness of those who carry out such unlaw- ful procedures do not lead to consequences even- more disastrous than those of which we have cognizance. Hence we may conclude that even although every reputable physician should observe the strictest asepsis and that as a consequence child-bed fever should occur only as a rare accident in his practice, there will always remain a number of cases of sepsis after abortion for which the medical profession is in no way responsible but which, nevertheless, it will certainly be called upon to treat. But even if the unfortunate patient survives an infection following labor or an abortion, she often remains a wretched invalid, sometimes for the rest of her life, or at other times until her sufferings are re- moved or palliated by a mutilating operation. It can hardly be doubted that fatal or bad later results are often due to an imperfect or an improper treatment of the condition. It is now a generally accepted fact that shortly after the idea of antisepsis had been promulgated, many patients died from sublimate or carbolic-acid poisoning following intra-uterine injections of solutions of these drugs. As a result, such *Read before the Cleveland Academy of Medicine, Apr. 17, 1903. Hunter Robb, M.D. strenuous procedures fell out of favor, especially after experience had shown that quite a considerable proportion of puerperal, even of streptococcic, infections will recover, if the patients receive only symptomatic treatment. Nevertheless, since, as has been pointed out, such recoveries are often very incomplete, and the patient rarely re- gains her former health, or at least only after a major operation, it is our bounden duty always to be on the lookout for some procedure which may not only reduce the mortality, but may also render re- covery, when it occurs, more perfect. It is on this subject that I wish to speak, my remarks dealing mainly with cases of abortion, although it is not impossible that they may be applicable to a large proportion of septic cases following labor at term. So far as my own experience goes, the remote results of a septic process following labor or an abortion, implicate more particularly the Fallopian tubes and ovaries in from 40 to 60 per cent of the cases. When an infection occurs after an abortion, there is usually a chill fol- lowed by an elevation of temperature, an accelerated pulse rate, and more or less pain in the lower part of the abdomen. Together with these symptoms we often have a bloody discharge from the vagina, or one mixed with pus or mucus, which not infrequently has a very offensive odor. As a rule the infection begins in or near the external genitals, and providing that it is of a virulent nature it implicates successively the structures of the vagina, the cervix, the uterus and the Fallopian tubes, and thus extends to the peritoneal cavity. At other times it reaches the peritoneum through the uterine walls by way of the lymphatics or blood vessels. Accompanying the inflammation there is often an exudate, which may accumulate in the peritoneal cavity or may remain sealed up in the Fallopian tubes. This exudate may be mucous, serous, purulent or bloody in character ; or more often it is of a mixed form. It may contain virulent organisms, although not infrequently it is sterile. If the fluid in the cul-de-sac is purulent the Fallopian tubes and ovaries are very liable to become mfected, and the ends of the former may become sealed up. If absorption of the exudate does not occur, organization takes place, with the fonnation of adhesions, which bind down the tubes and ovaries more or less firmly and thus interfere with their functions. In this way various symptoms are produced which cause the patient a greater or less amount of inconvenience until the adhesions are separated. Under such circumstances we have two problems to solve, (i) How can we ensure to the patient the Vaginal Incision in Sepsis Following Abortion. 3 best chance for her life? and (2) How can we at the same time pre- vent the occurrence of unfortunate sequelae? Our present method of dealing with an infection after an abortion is as follows : The cavity of the uterus is first cleaned out in order to prevent the continued absorption of toxines from any infectious ma- terial that may be present, and thus at least limit the intensity of the process. For this we employ the finger alone, or combined with the curette. The cavity is then thoroughly washed out with a hot saline solution and two ounces of peroxide of hydrogen; it is then cleansed again with the saline solution and sponged dry. Two or three drachms of iodoform powder are carried into the uterine cavity, after which the latter is packed with strips of sterile gauze. We next make a free opening into the cul-de-sac, and after evacuating any fluid that may be present, we irrigate it freely with hot saline solution, followed by one or two ounces of hydrogen peroxide. After this the cavity is again washed with saline solution and sponged dry. Lastly two or three drachms of sterilized iodoform powder are dusted into the cul-de-sac, which is then packed tightly with strips of sterile gauze. By this procedure we believe that we obtain two distinct advantages : ( i ) By evacuating the fluid we prevent further absorption of toxines from the cul-de-sac. (2) We save the Fallopian tubes and ovaries by preventing adhesions which would almost inevitably form as a result of the 'or- ganization of the exudate. Pryor, of New York, who has written at some length on this sub- ject, does not carry out the vaginal incision unless he can demonstrate the presence of organisms in the uterine cavity. Moreover, even in the latter case, he opens into the cul-de-sac, not for the purpose of drainage, but in order to fill the pelvic cavity with iodoform gauze, whereby he aims at isolating the uterus completely, and bringing all parts of the pelvis into contact with the gauze, which eliminates free iodine when it meets with a serous membrane. Our experience, how- ever, has shown that even when no infective organisms can be demon- strated in the secretions from the uterine cavity, there are not infre- quently present in the cul-de-sac purulent or muco-purulent collections containing bacteria. In some of our instances, indeed, a considerable amount of free pus, serum or blood has been found, even when a bi- manual examination under full anesthesia had failed to suggest its existence. I shall briefly present the clinical history with the bacteriologic analysis of 10 cases of sepsis following abortion treated in this way during the past year. The operations were performed by myself and Hunter Robb. M.D. my assistants at the Lakeside Hospital, and the bacteriologic work was carried out by Dr. Charles D. Williams, my former Resident Gyne- cologist, in Prof. Howard's Laboratory. In these cases the results both immediate and later, following the procedure have been so uniformly satisfactory, that we have great hopes that a more extended experience will demonstrate it to be not only a life saving method, but a reliable prophylactic measure against the subsequent loss of function of the lateral structures, and a pre- ventive against much of the wretched health that so many of our patients date from an abortion or a labor. Clinical and bacteriologic analysis of ten cases of septic infection following abortion, in zvhich dilatation of the cervix and curetting of the uterine cavity was supplemented by incision into the vaginal cul-de-sac. Age. The oldest patient was 39, the youngest 19 years of age. The average age was 26 years. Eight were married and 2 were widows. Occupation. Housework, 7; cashier, 1; prostitute, i. In 4, full-term births had occurred; there had previously been 11 miscarriages among the 10 patients. Abortion had been induced in 7 cases, a rubber catheter having been inserted into the uterine cavity. Symptoms. The most frequent symptoms were a sharp or dull pain in the lower abdomen, chills and fever, and a bloody discharge from the vagina. Temperature and pulse on admission to hospital. The highest temperature was 105.7° F- J highest pulse, 160. The lowest temperature was 100.3° F- ; lowest pulse, 96. Average maximum temperature 102.7° F- Average maximum pulse 129. Lowest maximum temperature T00.3" F. Lowest maximum pulse 96. The vaginal discharge. In 7, it was bloody in character; in 3, purulent. Operation. In each case the uterine cavity was curetted and the pouch of Douglas was opened. In 2 cases. a slight amount, in 2 cases a large amount and in 6 cases a moderate quantity of debris, was removed from the uterine cavity. In 3 cases the debris was foul-smelling in character. It consisted of portions of the placenta, fetal tissues and endometrium. Vaginal Incision in Sepsis Follozving Abortion. Material in cul-de-sac. In 2 cases there was no perceptible amount of fluid in the cul-de-sac and in these cases there was placental tissue in the uterine cavity. In the second case there were cocci on coverslip examination from the cul-de-sac. In 8 cases in which fluid was found in the cul-de-sac the following amounts and characters were noted : ( 1 ) 800 cc. of sanio-purulent fluid with a fecal odor. (2) 125 cc. of pus. (3) A small amount of blood-stained fluid. (4) About 40 cc. of a clear fluid. (5) A small amount of clear fluid. (6) 200 cc. of a purulent fluid. (7) 70 cc. of pus. (8) 180 cc. of a sanious fluid. Thus in 80 per cent of the cases fluid was found in the cul-de-sac varying in amount from a small collection of a clear fluid to 800 cc. of a sanio-purulent fluid. In 4 cases (40 per cent) it was purulent in character. Leucocytosis. In 5 cases in which the blood examination was made, the following counts were recorded : 18,000, 16,000, 12,000, 22,000, 28,000. In the remaining 5 cases the patients were taken to the operating room immediately after admission to the hospital and the blood was not examined. In 4 of the 5 cases there was a marked increase in the number of the leucocytes. Results — 9 recovered, i died. Coverslip and cultural examination of contents of the uterine cavity and the cul-de-sac. Case I. Coverslip: — Numerous cocci and bacilli from the uterine cavity and from the cul-de-sac. Cultures : — B. coli and streptococcus. Case II. Coverslip : — Negative. Cultures : — Negative. Case III. Coverslip : — Cocci and bacilli from uterine cavity and from cul-de-sac. Cultures : — B. coli communis. Case IV. Coverslip from uterine cavity and cul-de-sac: — Nega- tive. Cultures : — Negative. Case V. Coverslip from uterine cavity and cul-de-sac : — Negative. Cultures : — Negative. Case VI. Coverslip: — Numerous bacilli from uterine cavity and cul-de-sac. Cultures : — B. coli communis. At autopsy, coverslips from uterus. Fallopian tubes and peritoneal cavity showed strepto- Hunter Rohb, M.D. coccus and B. coli communis. Cultures made at autopsy from Fallopian tubes gave .B. coli communis and B. mucosus capsulatus. Case VII. Coverslips : — No organisms from uterine cavity. Bacilli from cul-de-sac. Cultures : — B. coli communis. Case VIII. Coverslips :— Uterine cavity negative. Cul-de-sac, Staph, pyogenes albus. Cultures :— Staph, pyogenes albus from cul-de-sac. Case IX. Coverslips : — Negative from cavity and cul-de-sac. Cul- tures : — Negative. Case X. Coverslips : — Cocci from cul-de-sac. Negative from uterine cavity. Cultures : — Negative. Thus in 5, or 50 per cent, of the cases organisms were found on coverslip and culture-tube examinations. The following organisms were demonstrated in coverslips and cultures : B. coli coipmunis, 4 times. Streptococcus, twice (together with the B. coli). In one case in which the streptococcus was present it was found in the peritoneal cavity and Fallopian tubes at the autopsy. It was demon- strated only by the coverslip examination. In 3 cases no organisms were demonstrated either in coverslips or in cultures from the uterine cavity, but in each case organisms were present in the cul-de-sac. In one, B. coli communis and in the other, Staph, pyogenes albus was found, while in the third case no growths could be obtained from the cul-de-sac, although cocci were demon- strable on coverslips. In 4 cases no organisms could be detected on coverslip or in cul- tures from the uterine cavity or from the cul-de-sac. In 3 cases the results of the coverslip and culture-tube examinations were similar. These findings are not altogether in accord with those of certain writers who state that organisms can always be detected in the uterine cavity when the peritoneum is infected. Moreover, our observations show that organisms may be present in the cul-de-sac when none can be detected in the uterine cavity. The fact that the B. coli communis was the organism present in about 40 per cent, of the cases afiforded a more favorable prognosis than 'would be justifiable in the presence of Streptococcus or Staph, pyogenes aureus. The B. coli communis met with fortunately was not of a high order of virulence. It may be of interest briefly to detail the history of the fatal case. History. L. Mc. Age 25. Admitted to the hospital, July 17, 1902. Married 3 years. Nullipara. One miscarriage at 4 months. Vaginal Incision in Sepsis Following Abortion. 7 Occupation, housework. On admission she stated that, after missing four menstrual periods, she had had an abortion induced on July 6. Six days later she had chills and fever. The fetus was removed 8 days later by manual manipulation. The fetal sac, however, did not come away at this time. Two days later the patient had a severe chill, fol- lowed by a fever and a great deal of pain in the lower abdomen. She was in a marked septic condition ; the temperature was 103° F., the pulse 150; she was slightly delirious. On examination the uterus was found enlarged to about the size of a 5 months' pregnancy. There was a foul-smelling, purulent dis- charge from the vagina. She was put under the influence of ether immediately and a large amount of necrotic, foul-smelling material was removed from the uterine cavity by means of the fingers and the curette. The cul-de-sac was incised, allowing from 150 to 180 cc. of a sanio-purulent fluid to escape. The uterine cavity and the cul-de-sac were irrigated with hot salt solution and peroxide of hydrogen, and then sponged dry, after which sterilized iodoform powder, and sterilized gauze were placed in the cul-de-sac. The operation was performed in twenty minutes. The patient soon after went into collapse and died within seven hours. The following is an abstract of the autopsy protocol (Dr. H. T. Parker) : "General peritonitis, considerable amount of greenish yellow pus in abdominal cavity. Uterine cavity necrotic and of a blackish green color. This process extends from 1.5 to 2 mm. into the uterine wall. The necrotic surface of the uterine cavity is covered with cocci. "Coverslips made from the Fallopian tubes, uterine and peritoneal cavities showed streptococcus pyogenes with a bacillus. Culture tubes gave B. coli communis and B. mucosus capsulatus." In this case, then, we had, as the result of an induced abortion, a general peritonitis, produced by a mixed infection. yo2 Rose Bldg. [Reprinted from The American Jouenal of Obstetrics and Diseases op Women and Childeen, Vol. XLIV., No. 4, 1901.] TORSION OF A HYDROSALPINX RESULTING IN INFARCTION. BZ WILLIAM H. WEIE, M.D., Demonstrator In Gynecology, Western Beserve UniTerslty ; Resident Oynecologlit to the Lakeside Hospital, Cleveland, Ohio. Torsion of the Fallopian tube of a grade sufficient to produce pathological changes and marked clinical symptoms is a very un- common occurrence, especially when the tube alone is affected. The following case, however, presents a typical picture. N. H., set. 46, white, married, was admitted to the gynecologi- cal service of Lakeside Hospital November 18, 1899, suffering from severe pain in the right side and lower abdomen, which be- gan suddenly the day before admission; nausea, difficulty in mictujition, and extreme tenderness in the lower abdomen. The family history was negative. With the exception of the usual diseases of childhood, patient has always been perfectly healthy and has never had any attacks similar to the present one. Menstruation began at 16, always regular, and painful before marriage, but not since ; duration one week ; the last period oc- curred one week previous to admission. Patient has had two miscarriages, the last twelve years ago, convalescence in both instances being uninterrupted. No viable child has been born. No history of previous pelvic inflammation could be elicited. Upon examination of the pelvis the uterus was found to be in retroposition, somewhat enlarged, movable and sensitive; the os uteri externum was quite patulous. Apparently in front of the uterus was a rounded, fluctuant, sensitive mass about the size of a baseball ; behind and to the right was an ill-defined, smaller, sensitive mass. A provisional diagnosis of right ovarian cyst was made. The examination was unsatisfactory on account of the extreme sensitiveness of the parts. Copjriglit, WUliim W«od St Company. 2 WEIE: TORSION OF A HYDEOSALPINX. The temperature on admission was 100° F., pulse 100, respirar tion 24. The urine showed a trace of albumin with a few hyaline and granular casts. The patient presented symptoms of a local- ized peritonitis, which subsided under palliative treatment. On November 23, five days after admission, abdominal section was performed by Dr. Hunter Eobb. The examination under anes- thesia just before operation showed that the uterus was of mod- erate size and movable; the fundus was directed forward and toward the right side ; to the right of the uterus, and movable independently of it, was a fluctuant mass, the size of an orange, suggesting an ovarian cyst. The left ovary seemed small and movable. An incision was made in the median line and an adherent hematosalpinx, the size of the closed fist and blackish red in color, was found on the right side of the pelvis. The pedicle formed by the mesosalpinx showed two complete twists. The ovary was adherent, but was otherwise imaffected. The adhesions having been separated, the tube and ovary were Mgated with sUk; they were then removed and the pedicle was cauterized ; the left tube and ovary, which were quite adherent, were removed in a similar manner. A small myoma the size of a hazel-nut, attached by a short pedicle to the posterior surface of the fundus upon the left side, was removed by myomectomy, and the incision in the uterus was closed with catgut. The appendix was found bound down by adhesions to the cecum, so that no meso-appendix was apparent. It was therefore removed and the cauterized pedicle covered with peritoneum. The abdomen was flushed out with salt solution and sponged dry, no oozing occurring. Five hundred cubic centi- metres of sterile salt solution having been left in the cavity, the abdomen was closed, catgut being used for the peritoneum and the skin and silver wire for the fascia. Convalescence was un- interrupted, the highest temperature being 101.2° F., highest pulse 126. The bowels were opened thoroughly on the day after operation; the dressing was removed on the tenth day, when perfect union was found to have occurred. Since the patient left the hospital attempts to reach her by letter have been unsuccessful. Pathological Examination.^— The right tube is occluded at the fimbriated extremity and forms a thin-walled, distended sac con- taining a thin, bloody fluid. The tube measures 24 centimetres in length around its convexity and 5 centimetres in its greatest 'From the Pathological Laboratory of the Lakeside Hospital. WEIE: TORSION OF A HYDKOSALPINX. 6 diameter; the weight of the tube and ovary together is 160 grammes ; in color it is dark red, almost black in places, and upon the surface there are a number of separated adhesions. The twists formed at the cornual end of the tube were straightened out at the operation. After being hardened in formalin for twenty- four hours, the tube, when cut open longitudinally, shows some half-dozen compartments, increasing in size from the cornual end outward and separated by incomplete septa. Along the sides of the compartments can be recognized the atrophied folds of the mucosa running in a longitudinal direction. The fluid contained in the cavity has not become hardened, but es- capes when the tumor is incised. Microscopically the whole structure is seen to be the seat of a hemorrhagic infarction, being necrotic and infiltrated with red blood cells. The folds are few in number and scarcely recognizable. A small amount of the epithelial covering is found, the cells being irregular, flattened, and degenerated, being arranged in a single layer. The connective tissue of the folds is crowded with red blood cells, only a few nuclei of the original structure being seen. The muscularis is also infiltrated with red blood cells, the greater part of the muscle either staining very faintly or not at all; around the periphery, however, the nuclei show up more dis- tinctly. The vessels are distended with blood, and the walls are often degenerated. No peritoneal cells are found. That a distended hydrosalpinx had existed previously to the twisting of the pedicle and the resulting local peritonitis seems highly probable from the complete and evidently long-existing closure of the fimbriated extremity, the atrophic condition of the folds of the mucosa, and the chronic appearance of the septa dividing the lumen into compartments. A further proof of this is to be found in the fact that the contained fluid did not coagu- late when subjected to formalin, as would probably have hap- pened if it had consisted of a large proportion of blood or serum, whereas the fluid in most cases of chronic hydrosalpinx fails to coagulate in this hardening agent. Furthermore, it is extremely improbable that a small occluded tube would become twisted, such an accident occurring almost always in cases of tumors of considerable size with a small pedicle such as would accompany a distended hydrosalpinx. The right ovary, measuring 3.5 x 2 x 1.5 centimetres in its vari- ous diameters, shows a few slight adhesions, but otherwise ap- i WEIR: TORSION OF A HYDROSAIiPINX. pears normal. Microscopically a few peri-ooplioritie adhesions are seen. The left ovary and tube show a number of separated adhesions. Microscopically the tube presents a "healed salpingitis." The appendix, 9 centimetres in length, microscopically appears nor- mal. The small myoma, 2 x 15 x 1.5 centimetres, presents the usual gross and microscopic appearances. In the literature a considerable number of cases are found; about thirty have been reported altogether. Forselles^ describes a case of his own and gives an abstract of fourteen other cases reported by as many writers in the literature. Hartman and Raymond- report two cases, and, in addition to giving some ref- erences found in Forselles' article, quote from five other writers who have met with this condition. In a subsequent article Hart- man reports three more cases. Legue^ describes three cases, one of which had been previously reported ; to this reference had also been made by Hartman and Raymond. Ries* met with a case of spontaneous amputation of both tubes; later the hydrosalpinx which had formed upon the right side became twisted and in- farcted, producing the characteristic symptoms; he refers to an article by Prager'* in which a number of cases are described. Montgomery also reported a case before the Philadelphia Obstet- rical Society. Prom the data supplied by the literature the following con- clusions may be drawn : This accident, in almost every instance, is found in organs the seat of previous pathological change, the most common condition being a hydrosalpinx. Its occurrence may be explained by the fact that in these cases the tube walls are usually thin, the comual extremity is narrow, while the fim- briated end is enlarged and distended with fluid, heavy, and is apt to be free from dense adhesions, a considerable range of mobility usually existing. On the other hand, a pyosalpinx is almost always densely adherent, the walls are more thickened and rigid than in a hydrosalpinx, and, the mobility being quite limited, the torsion less frequently occurs. Two or three in- stances, however, are on record. Torsion of the tube has also 'Forselles: Ueber Axendrehung der Tube. Deutsche Zeltschrift fur Chirurgle, 1898. ' Hartman et Raymond : Le torsion de pedlcule des salpingo-ovarltes. Annales de Gyn. et d'ObstSt., 1898. "Legue: La torsion de salpingitis. Presse m6d., Jan., 1900. 'Ries: American Gyn. and Obstet. Journal, April, 1900. 'Prager: Arch, fur Gyn., vol. Ivlii. WEIE: TORSION OP A HTDEOSAXPINX. 5 occurred in connection with tubal pregnancy, hematosalpinx, uterine fibroid, parovarian cyst, and hydatid cyst. The tube may also be included in the twisted pedicle of an ovarian cyst, as had happened in the ease of a patient upon whom I operated in July, 1899. In two instances the tubes were the seat of malig- nant disease, and in one case (Hartman and Raymond), occur- ring during pregnancy, the tubes had evidently been previously healthy. This last is the only instance in which evidences of previous lesions were not found. The tube alone may be involved, or the ovary may also be in- eluded in the torsion. The direction of the twist is inconstant, but usually seems to follow the hands of a watch; the number of twists varies from one-half to four and a half complete turns. In size the tumor may be as large as the fetal head at six months, while in form it is variable, being usually somewhat globular and readily mistaken for an ovarian cyst. The wall is usually thin and the tension rather high. The pathological changes depend upon the degree of constric- tion. In one of Legue's cases the pedicle was twisted one and a half times and yet there was no apparent compression or stran- gulation, although the sjnnptoms were well marked. The first change produced by the torsion will be a venous stasis associated with edema and later with interstitial hemorrhage. In appear- ance a twisted hydrosalpinx in this condition closely resembles that of an incarcerated intestine, being tense, lustreless, and plum-colored or black. In fact, on several occasions it has been mistaken for the strangulated gut. The later effects would be thrombosis of the vessels with the degenerative changes described by Sanger as hemorrhagic necrosis and by Rundl as a hemor- rhagic infarct. Actual necrosis is usually, however, prevented by the formation of adhesions in the less severe cases, while operative measures are, as a rule, carried out in the more grave cases owing to the severity of the symptoms. The symptoms are practically identical with those arising from a twisted ovarian cyst, and it is often impossible to differen- tiate between the two until the abdomen is opened. At the site of the lesion there is usually sudden sharp pain, which may be so severe as to cause fainting or collapse. Vomiting may occur, and, in connection with the abdominal pain, may suggest intestinal obstruction or appendicitis, but it does not persist or become fecal. There is rarely obstipation and the temperature is nor- mal or only sUghtly elevated. A history of one or more previous 6 WEIE: TORSION OP A HTDEOSALPINX. attacks may often be obtained, as in torsion of an ovarian cyst. Upon examination the tense, fluctuating pelvic tumor can usu- ally be detected ; the mobility is often considerable, and in most of the cases a diagnosis of ovarian cyst with torsion of the pedicle has been made. The prognosis is usually good if proper operative measures can be carried out ; the occurrence of a lesion in a pyosalpinx of course adds to the gravity of the ease. The treatment is operative, and, as many of these patients have had previous attacks, if the condition seems to be improving it would seem advisable to wait until the acute attack subsides. 702 Rose Bthlding. [Reprinted from American Gynecology, August, igo2.) RUp'fU^E .OF THE POSTERIOR WALL OF THE CERVIX UTERI WITHOUT INVOLVEMENT OF THE EXTERNAL OS, OCCURRING DURING ABORTION AT THE FOURTH MONTH ; ESCAPE OF THE FETUS THROUGH THE LAC- ERATION. By Wm. H. Weir, M.D., Demonstrator of Gynecology, Western Reserve University; Resident Gynecologist to Lakeside Hospital, Cleveland, Ohio. Lacerations of the cervix, resulting from labor or abortion, usually extend upward from the os externum ; they oceur^ on one or both sides of the cervix or they may be multiple ; they vary, greatly in extent, from a slight fissure to a deep tear, extending up into the vault of the vagina, opening up the parametric tissue or even communicating with the peri- toneal cavity. Tears of the upper part of the cervix not involving the external os are generally due to a rupture of the uterus below the "contraction ring" of Bandl. They may communicate with the abdominal cavity or merely with the parametric tissue. Again, as in the case reported here, the laceration may open into the vagina without implicating either the vaginal vault or the os externum. The only other instance of this sort that I have been able to find is one seen by Gebhard {Path. Anatomie der Weihliche Sexualorgane, 1899), in a woman who had died from heart disease just after an abortion at the fifth month. In his case, there was found in the posterior wall of the cervix a large rupture com- municating with the cervical canal but not implicating the external os, the latter being quite small and admitting only the finger tip. The abortion had occurred through this tear in the cervix. Complete separation of the lower border of the cervix, in the form of a circular strip of tissue around the external os, has been mentioned by Winckel. Complete separation of the cervix from the body of the uterus has resulted also from high circular rupture of the uterus, the peritoneal cavity being opened. Fistulous openings through the anterior cervical wall may result from necrosis induced by long continued pressure of the fetal head against the pubic bone but this accident is of a different nature from Wm. H.-Weir. M.D. the ordinary lacerations, which are caused by the forcing down of the fetus against a rigid or insufficiently dilated cervix. The patient, aet. 21, was admitted to the gynecologic service of Lakeside Hospital, May 14, 1900. She presented symptoms of an impending abortion — severe bearing down pains and uterine hemor- rhage. Examination showed that the cervix was small and hard and Fig. I. Exposure of posterior surface of Cervix: shows ragged tear in posterior lip of cervix, through which abortion occurred. the external os contracted and rigid ; the uterus was the size of that ■of a four months' pregnancy ; the breasts were somewhat enlarged and colostrum could be squeezed from the nipples. Three and one-half months previously the menstrual period had not begun when expected and the patient, fearing that she might be pregnant, had inserted a rubber catheter into the uterus. Nothing resulted from this procedure until three days later when she had appar- ently a normal period. The next menstrual flow also appeared and Rupture of the Posterior Wall of the Cervix Uteri. 3 vaginal examination after it had ceased did not suggest pregnancy, the cervix being quite hard and small and the uterus but slightly en- larged. A third menstrual period also occurred and a week later the present symptoms began spontaneously. Abortion being considered inevitable it was decided to hasten mat- ters by dilating the rigid cervix. This was attempted with the Goodell- EUinger dilator, under cocain anesthesia, but the resistance of the tissues was so great that the canal could not be stretched to more than I cm. in diameter. The procedure was executed with the utmost care, apparently without causing the slightest laceration. The patient's con- dition being excellent, it was decided to await further dilatation by the natural process and a small strip of gauze was left in the cervical canal. During the afternoon the pains were frequent and severe, bromide of potassium and chloral being given without much effect. In the evening the gauze was removed in order to relieve the pain and the cervix was found still rigid and admitting only the tip of the index finger. The pains continued during the night and early r^n the next morning a four months' fetus within the unruptured amniotic sac and the pla- centa were expelled. On vaginal examination, made shortly afterward, the external os was found unruptured and surprisingly small; the uterus was contracting well and there was no hemorrhage. No un- favorable symptoms having developed, further examination was post- poned until three days later, when a large oval laceration was discov- ered in the posterior wall of th' cervix about i cm. above the margin ■of the external os, involving neither the os nor the vaginal vault. This tear communicated with the cervical canal and had served for the pas- sage of the fetus. It had been entirely overlooked at the first exam- ination. The opening was sutured with silk-worm gut and excellent union ■occurred. The convalescence was uninterrupted and the patient's sub- sequent health has been perfect. yo2 Rose Building. [Reprinted from The American Journal of Obstetrics and Diseases of Women and Children, Vol. L., No. 6, 1904.] ^ CONSERVATISM IN PELVIC SURGERY.' HUNTER ROBE, M.D., Professor of Gynecology Western Reserve University, and Visiting Gynecologist to the Lakeside Hospital, Cleveland, Ohio,^ A THOUGHTFUL coiiservatism in operative procedures, when dealing with pathological conditions of the female organs of generation, has fortunately become the rule in the majority of surgical clinics. And more particularly is this likely to be the practice where the surgeon in charge has kept in touch with the progress that has been made from this standpoint, and from his own experience and that of others has learned to appreciate the advantages of conservative procedures in the treatment of pelvic inflammatory disease. There still remain, however, operators who believe it to be their duty to resort to extreme measures when dealing with pathological conditions of the tubes and ovaries, although the same men often favor more conservative procedures when they encounter general surgical abnormalities. Now, with few exceptions, the so-called cystic or cirrhotic ovary is still capable of performing its functions, and when the organ is bound down by adhesions, the symptoms are apt to be due mainly to their presence. Hence, in the restricted sense of the term, the condition of such organs is not pathological, and when they have been removed under these circumstances they seldom, if ever, show any evidences of inflammation even upon microscopical examination. But even when the ovaries or tubes have undergone actual in- flammatory changes, or where they are occupied by tumor for- mations, or are bound down by adhesions so that their func- tions are interfered with and the necessity for operative pro- cedures becomes imperative, we still have to decide how far we ought to go and how we can get the best results for the patient not only immediately but later on. After an experience of more than five years in the application of conservative measures in vari- 'Read before the Ohio State Medical Society at Cleveland, May i8th, 1904.. Copyright, William Wood & Company. 2 ROBE : CONSERVATISM IN PELVIC SURGERY. ous forms of pelvic abnormalities, we have been able to thorough- ly convince ourselves of the great advantages that may be ob- tained by preserving as far as possible, the integrity of the pelvic organs. It is true that in a small percentage of cases, after such a line of treatment has been followed, the patient will still have to undergo a second operation before she can be completely re- lieved of her discomfort, and it is also possible that in a few in- stances, by the introduction of infection under these circumstances, her condition may be rendered even worse. Such cases are ex- ceptional, however, in our experience. Before choosing the more conservative operative procedure, we always make it the rule to carefully explain to the patient, or to her friends, that such measures will be carried out if in our judg- ment at the time of the operation they seem to be advisable. But we further state, that even though we remove what seems to be the inflammatory area, it may later become necessary to institute a second operation, before relief is obtained. In the great ma- jority of instances, the patients are perfectly willing to take a good many chances if there is a reasonable prospect that the con- servation of the pelvic organs will be compatible with future health and comfort, and from actual observation we have found that it does not become necessary to perform a second operation in more than from three to five per cent, of all such cases. If an ovary or a portion of an ovary can be saved, before the normal menopause has begun, or even during the time in which the pa- tient is experiencing these changes, we have found that not only the immediate convalescence, but also the subsequent condition of the patient is in every way more satisfactory. Anyone that has cared for patients that have had their ovaries removed (even though diseased) can testify to the fact that many of them suflfer more or less acutely for varying periods of time, following the operation, in some instances as long as five years. And unfortunately duriiig this time the morphin or some other drug habit may be formed in endeavors to relieve their distress. The prevention of the artificial menopause is the most important reason for leaving, the patient her ovaries whenever this is pos- sible, the question of possible pregnancy following conservative measures being in our opinion only of secondary importance, as in the majority of these cases, the patients are in an unhealthy con- dition for the bringing of a child into the world. This criticism, of course, does not usually apply to those cases in which a tumor is present, implicating only one ovary. Where the question of ROBB : CONSERVATISM IN PELVIC SURGERY. 3 pregnancy is to be considered, one has to deal with the condition of the Fallopian tubes, as well as that of the ovaries. ! The vairious methods to be employed in these conservative operative measures when dealing with the tubes and ovaries are well known to everyone and a description of them would prove to be merely a tiresome repetition. I would say, however, that in the light of our experience it is a wiser procedure to remove the Fallopian tube, whenever a pyosalpinx exists, i.e., where there are macroscopical evidences of pus. When, however, the ovary is involved in an abscess formation, the same radical treatment is not always indicated, as the abscess in most instances does not involve all of the ovarian stroma. Moreover, microscopical ex- amination of many of these ovaries will show that the abscess is walled off, and the ovarian stroma beneath is frequently in- vaded only to a slight extent. In such instances the abscess may be excised and the line of incision be brought together with a fine silk or catgut Suture. I shall nqw give a brief analysis of the work in this line carried out in the : Gynecological Department of the Lakeside Hospital during the past five years. I wish to acknowledge the valuable assistance ) pf Dr. Howard Dittrick, the Resident Gynecologist, and Dr. Wm. J. Abbott, the First Assistant Gynecological Interne to the Lakeside Hospital, in the preparation of the. analytical tables. In all the cases considered the lateral structures showed macroscopically marked evidences of inflammatory disease, and -there were adhesions which bound down the structures. We have only included in this analysis those cases in which we were able tQ^carry out conservative measures, and not those in which we were obliged, ; on account of the technical difficulties of the opera- tion, to leave the lateral structures in on one or both sides, although removal was indicated. ANALYSIS OF CASES. Total number, 237. Age : The oldest patient was 52 ; the youngest 17 years of age. Average age, 28.12 years. Five were 17; six were 18; nine were 19; eight were 20; twelve were 21 ; thirteen were 22; twenty. were 23; nineteen were 24; fourteen were 25 ; nine were 26 ; eight were 27 ; eighteen were ,28 ; eight were 29; fourteen were 30; four were 31 ; six were 32; ten were 33 ; four were 34 ; seven were 35 ;. three were' 36 : three were 37 ; six were 38 ;. four were 39; five were 40; four were 41 ; three were .42 ; two were 44 ;, one was 45 ; three were 46 ; two were 47 ; two 4 ROBB : CONSERVATISM IN PELVIC SURGERY. were 48; one was 49; one was 50; two were 52. The majority of the patients were between 17 and 30 years of age. Menstrual History. — The menstrual history was abnormal in 169, and normal in 68 cases. The menopause had taken place in one. The symptoms presented in the abnormal cases in most in- stances were those of dysmenorrhea, menorrhagia, prolonged and irregular flow. Leucorrhea.— One hundred and seventy-six patients gave a his- tory of a leucorrheal discharge. Of these patients, 124 were mar- ried, 38 were single, and 14 widowed. Married Life. — 165 were married; 52 were single; 20 were widows. Longest time married, 28 years; shortest time married, three months. Of the 165 married, 115 had borne children. The total number of children borne by the 115 patients was 271, the average being 1.14 per cent. The greatest number of children borne by one patient was 11. The next greatest number was 9. Still-born children, 2. Twins, r. Women having borne one child, 51; 2 children, 21; 3 children, 12; 4 children, 13; 5. children, 4; 6 children, 5; 7 children, 2; 9 chUdren, 2; 11 children, i. A History of infection following labor was recorded in 40 cases. Instrumental delivery was followed by infection in 13 cases. The total number in which there was a history of infection following labor, 53. . Miscarriages. — Number of patients having had miscarriages, 104; number of miscarriages, 170; married, 83; single, 10; wid- owed, 10. Abortions or miscarriages had been induced in 52 cases; in the married patients, 34; in single patients, 10; in wid- ows, 8 ; patients having had one miscarriage, 65 ; two miscar- riages, 22; three miscarriages, 13; four miscarriages, i ; five mis- carriages, I ; six miscarriages, i ; seven miscarriages, i. Infection follo^wing a miscarriage had occurred in 42. Gonorrheal Infection. — In 51 patients there was a history of gonorrheal infection; positive in 32; probable in 19. Of these 28 were married; 19 single, and four widowed. Among the married the history was positive in 16, probable in 12, and in one was complicated with syphilis. Among the single the history was positive in 15, and probable in 4 cases. Among the widows, positive in one case, and prob- able in three. A positive history of a specific infection is generally difficult to obtain, and unless the infection can be surely proven, we are not justified in making positive deductions from this standpoint. " ROBB : CONSERVATISM IN PELVIC SURGERY. 5 It will be seen , from a study of the cases of labors and miscar- riages, that infection in these instances plays a very important part in the causation of inflammatory diseases of the tubes and ovaries. Thus, there were 53 cases of infection following labor and 42 following miscarriage — in all 95 cases. Bowels. — In 73 cases there was a history of constipation. Forty- seven of the patients were married; 14 were single; 12 widowed. Micturition. — There was some complaint with this function in 124 cases. Ninety of the patients being married, 23 single; 11 widowed. The general condition was good in 138 cases; in 82 fair; and poor in 17. The uterus was adherent in 121 cases. Eighty-six of the pa- tients were married ; 29 single ; 6 widowed. The bowels were adherent in 165 cases. One hundred and ten of the patients were married; 42 were single; 13 were widowed. The vermiform appendix was removed in 113 cases. Married, 80 ; single, 23 ; widows, 10. In 56 married patients it was ad- herent; in 14 flexed; in one occluded; in 9 hypertrophied. Sin- gle: adherent, 18; flexed, 3; occluded, i ; hypertrophied, i. Wid- ows : adherent, 8 ; occluded, i ; hypertrophied, i . It was adherent and not removed in two cases on account of an extreme condition of shock after removal of the pelvic struc- tures. One of the patients was married, the other single. The appendix was adherent in 82 cases " flexed in 17 " " occluded in 3 " " " " hypertrophied in 11 " Ovaries: 113 STRUCTURES SAVED. Right 93 Left 76 Both (47 times or) 94 Right (partial) . .■ 17 Left (partial) 17 Both (partial) 3 Both ovaries with tubes 17 times or . . . 34 334 o robb: conservatism in pelvic surgery. This number was saved in 237 cases, a little over one ovary, and a third to each patient. Tubes: Right 17 Left 25 Both 25 times 50 Right (partial) 15 Left (partial) 11 Both (partial) 5 times or 10 Total 128 in 237 cases, or a little more than half a tube to each patient. STRUCTURES SAVED IN THE PUS CASES. Pus was met with in 64 cases out of 237, or in 27.004 per cent. They were divided as follows : Married 36 Single 22 Widow 6 64 The pus was found as follows : Double pyosalpinx: Married 22 Single 5 Widow 3 30 Single pyosalpinx: Married 7 Single 5 Widow I Tubo-ovarian abscess with pyosalpinx (single) : Married 6 Single I Widow o 7 robe: conservatism in pelvic surgery. 7 'Tubo-ovarian abscess (double): Married '. o Single 8 Widow o ~8 Tubo-ovarian abscess (single): Married i Single 2 Widow I 4 ■Ovarian abscess (double): Married o Single I Widow I 2 STRUCTURES SAVED IN PUS CASES. Ovaries. Tubes. ^gM 27 Right 5 Left 15 Left 2 Both 12 times, or 24 Rjght (partial) 2 Right (partial) 4 Left (partial) 2 Both (partial) 2 72 II Thus 72 ovaries were saved in 64 pus cases, or about one and a fifth ovary to each patient. In 12 cases both ovaries were saved. Eleven tubes were saved, or i to about every five and a half patients. In these cases the following organisms were found : Times. ■Gonococcus 6 Streptococcus pyogenes 4 (i doubtful) Staphy. pyog. aureus 3 Staphy. pyog. albus , 2 B. coli communis 2 B. mucosus capsulatus i Cocci (no growths) 3 21 in all or J^ of the cases. 8 ROBB : CONSERVATISM IN PELVIC SURGERY. In the pus cases the abdominal wound became infected 4 times^ or 6.25 per cent. From the infected abdominal wounds in the pus- cases the following micro-organisms were isolated : Times Streptococci i B. coli communis r Staph, pyog. aureus i Staph, pyog. albus i B. mucosus capsulatus ■ i 5- Micro-organisms found in abdominal wounds other than pus cases : Times. B. coli communis 2- Staphy. pyog. albus z Cocci and bacilli on coverslip, no growth i 5 The leucocyte count in the pus cases : Highest 36,000 Lowest 10.000 Average count 21,615 Drainage was employed in the pus cases as follows : Abdominal alone none Vaginal 13 times Abdomino-vaginal 2 " Fifteen times, or in 23.43 per cent, of all pus cases. Drainage was carried out 21 times in 237 cases, or in 8.86 per cent of the total number of cases. Once by the abdomen alone; twice by the abdomen and vagina combined ; eighteen times by the vagina alone. The convalescence was interrupted by the following conditions : Bronchitis in four cases ; pneumonia in one case ; pleurisy in three cases ; phlebitis in one case ; abdominal fecal fistula in one ; post- operative mania in one; suppuration of abdominal wound twenty times. Total number interrupted, thirty-one. ROBB : CONSERVATISM IN PELVIC SURGERY. 9 In those cases interrupted by infection of the abdominal wound (eight of which were in the pus cases), nineteen of the infections were sHght and one marked. Total number of abdominal wounds infected, 8.43 per cent. Number of cases requiring a secondary abdominal operation for the relief of symptoms, seven. Number of cases under observation on account of pelvic discomfort, eight. In seven of these cases the symptoms of which the patients com- plained disappeared after a year's time; one is still complaining. Deaths in the pus cases two, or 3.1 per cent. Deaths in the whole number of cases (237) four, or 1.68 per cent. Case I. — Among the pus cases the diplococcus pneumoniae was found in the secretions in the peritoneal cavity, together with B. mucosus capsulatus and B. coli communis. Case II. — The patient also died from the .effects of a pelvic peri- tonitis ; autopsy not allowed. The two remaining cases were in the non-suppurating class. In one there was an ectopic gestation which had involved the right tube. On the eleventh day following the operation she de- veloped an acute obstruction of the bowels. The temperature and pulse were practically normal. The abdomen was reopened and the obstruction relieved, but she succumbed from shock one hour following the operation. In the secon.d case the disease of the lateral structures was complicated by an adherent, much thick- ened and contracted gall-bladder, which contained three good- sized gall-stones. The opening made into the gall-bladder was difficult to close. The patient developed a localized peritonitis, which resulted fatally five days after the operation. 702 Rose Building. [Reprinted from- The American Journal of Obstetrics and Diseases of Women and Children, Vol. L., No. 6, 1904. ] THE STREPTOCOCCUS IN GYNECOLOGICAL SURGERY.' HUNTER ROBBj M.D., Professor of Gynecolosy. Western Reserve University; Visiting Gyiiecologist to the Lakeside Hospital ; WINFORD H. SMITH, A.B., M.D., Assistant Gynecological Interne to the Lakeside Hospital, Cleveland, Ohio, Bacteriologists have still much to learn and teach us about the streptococcus, especially with reference to the differentiation of this organism into distinct species or varieties. Clinically, Tiowever, we know a good deal about this organism, and experi- ence has taught us to regard its capacities for evil with respect not altogether unmingled with a wholesome, dread. We know that the streptococcus in some form or other is' the pathogenic agent in erysipelas, acute septicemia, the puerperal fevers and post-operative peritonitides, and thus far researches have taught us to wonder not so much at the frequency, but rather at the rela- tive infrequency of acute processes due to the agency of the strep- tococcus pyogenes. Doderlein has shown that of the vaginal secretions taken f^om about 200 cases, nearly 100 were found to be abnormal, and that 10 per cent, of th6se pathological secretions contained the streptococcus pyogenes, which in 50 per cent, of the cases at least was pathogenic for animals. Under these cir- cumstances when we remember the bruised and lacerated con- dition of the genitalia after labor, which would naturally offer easy access to invading organisms, it would seem surprising that puerperal infections are not much more common. Czerniewski found streptococci in the lochia of 33 out of 81 women suffering from puerperal fever, whereas in those from 57 healthy women he was able to find this organism only once. In ten fatal cases he demonstrated its presence in the various organs of the body after death. On the other hand, Pryor reports 36 recoveries in 37 cases of 'Read before the American Gynecological Society, May, 1904. Copyright, William Wood & Company. 2 ROBB AND SMITH : STREPTOCOCCUS IN GYNECOLOGY. puerperal sepsis from which the streptococcus pyogenes was iso- lated, but attributes his good results to the effects of the iodine which was set free from the iodoform gauze employed by him. Unfortunately our own experience is far less encouraging, and after carefully explaining errors in technique, we have been forced to the conclusion that the organism with which Pryor had to deal was either less virulent than the form encountered by us, or that his patients had more resistance — or that his good results were possibly due to a combination of these two factors. In order to arrive at some definite conclusions with referenct- to the streptococcus pyogenes as a cause of death in our own work, we have made an analysis of all our cases in which this organism has been found during the past six years. It will be shown from our observations that quite a large number of our patients died, and one was unimproved. It will also be noticed that in the great majority of cases in which this organism was met with, there was a previous history of infection following labor, or an induced criminal abortion. In the analysis of all cases the points noted were as follows : Name; age; condition; occupation of the patient; number of children and ages; miscarriages; most frequent symptoms; tem- perature and pulse on admission; highest temperature; highest pulse; lowest temperature; lowest pulse; average maximum tem- perature ; average maximum pulse ; lowest maximum temperature ; lowest maximum pulse ; leucocytes ; operation ; results ; urinalysis ; drainage; condition of the appendix; suppuration; class accord- ing to the operation. Abortion Cases (including a few cases of labor after which treatment was necessary) : Total number of cases 137 Average maximum temperature 100.8° F. Average maximum pulse 1 1^ Operations. Abdominal and vaginal . 3 D. & C. with irrigation 71 D. & C. with irrigation and vaginal puncture 29 D. & C. (following labor with irrigations) 2 D. & C, vaginal puncture and irrigation (following labor) ... 3 Pelvic abscess ; vaginal puncture i Miscellaneous operations , 6 No operative procedure 12 ROBB AND SMITH : STREPTOCOCCUS IN' GYNECOLOGY. 3 ( (None stfeptococcus) ) Abortions completed in hospital ■] /Que died) V . . lo Recoveries 104 or 75.9 per cent. Improved 17 " 12.4 " " Unimproved , i " .8 Died 15 " 10.9 " '' )■/■■ Operations in the case of the patients who died : Abdominal and vaginal 2 D. & C. and irrigation 5 D. & C. and irrigation and vaginal puncture 5 D. & C. and vaginal puncture (following labor) 1 Miscellaneous i Abortion completed in the hospital i IS Streptococcus pyogenes in cases of abortion or following labor : Total number of cases 16 Average maximum temperature 101.6° F. Average maximum pulse 124 Operations. Abdominal and vaginal 3 D. & C, irrigation 2 D. & C, irrigation, and vaginal puncture 6 D. & C. and irrigation (following labor) i D. & C, irrigation, and vaginal puncture (following labor) . . 4 16 Recoveries 4 or 25 per cent. Improved 3 or 18.75 P^^ cent. Deaths 9 or 56.25 per cent. In 16 then, of the 137 cases, the streptococcus was found. The total number of all our cases (from all sources) in which the streptococcus was found is 40. Consequently, those in which this organism was found following an abortion or labor formed 40 per cent, of the total number of streptococcus cases from everv source. Operations in the Cases which Died. Abdominal and vaginal 2 D, & C, vaginal puncture and irrigeition , , . , 4 ROBB AND SMITH : STREPTOCOCCUS IN GYNECOLOGY. D. & C. and irrigation (following labor) i D. & C, irrigation and vaginal puncture (following labor) ^ . . 2 Furthermore, we find 5 cases and i doubtful case which were operated upon here (abdominal operation) not for abortion, but in which there was a history of a previous abortion or labor fol- lowed "by chills and fever and other symptoms pointing to in- fection, mild or otherwise. Cases in which the streptococcus pyogenes was found (from all sources) : Total number of cases 40 Average maximum temperature 101.2° F. Average maximum pulse 116 Operations. Abdominal operation only 5 Abdominal and vaginal 14 D. & C. and irrigation (following abor- tion) 2 D. & C, irrigation and vaginal puncture (following abortion) 6 D. & C, irrigation and vaginal puncture (following labor) 4 D. & C. and irrigation following labor. . i Pelvic abscess evacuated by vaginal puncture i Miscellaneous 7 Recoveries 20 or 50 per cent. Improved 6 " 15 " " Deaths i4 " 35 " " Operations in the Fatal Cases. Abdominal operation only 4 Abdominal and vaginal 3 D. & C, irrigation and vaginal puncture following abortion 4 D. & C, irrigation and vaginal puncture following labor 2 D. & C. and irrigation following labor . , l 19; of which 7, or 36.8 per cent, died. 16 cases in which the vagina alone was opened; 6, or 37.5 per cent., of the patients died. ROBB AND SMITH : STREPTOCOCCUS IN GYNECOLOGY. 5 Of these 14 deaths in the streptococcus cases, 9, or 64.3 per cent., were in cases following labor or abortion. We find that in all cases the streptococcus was found alone or in combination in the following order of frequency : ( i ) Strepto- coccus alone; (2) Streptococcus and Staphylococcus aureus; (3) Streptococcus, Staphylococcus aureus and B. coli communis. As to the regions from which these organisms were obtained we might say that in all cases (except 3 in which they were found in the vagina), the organisms were obtained from the uterus, the adnexa, the cul-de-sac or from, several of these situations. In other words, the organisms were present in places which were admittedly not their normal habitat. In the past six years we have had 724 abdominal sections with a total of 25 deaths — a mortality of 3.45 per cent. In 7, or 28 per cent, of the deaths, the streptococcus pyogenes was demon- strated. Ntimber of streptococcus cases in the abdominal sections that recovered 12 Number of cases operated upon by abdominal section 19 Number recovered: 12, or 63.2 per cent., leaving a mortality of 36.8 per cent, for the streptococcus cases in which an abdominal section was carried out. 702 Rose Building. I Reprinted from The American Journal of Obstetrics and Diseases of Women and Children, Vol. LI., No. i, 1905.] THE EARLY DIAGNOSIS OF CANCER OF THE FUNDUS, WITH REPORT OF CASES.' BY- HUNTER ROBB, M.D., Professor of Gynecology in the Western Reserve University and Gynecologist-in- Chief to the Lakeside Hospital, Cleveland, Ohio. (With three illustrations.) The routine examination of uterine curettings will undoubtedly enable us in many instances to make a diagnosis of cancer of the fundus in its very earliest stages, while, at the same time, it will probably demonstrate that cancer of the body of the uterus is of more frequent occurrence than has been generally supposed. The exact ratio between the incidence of cancer of the body of the uterus and that of the cervix has not yet been clearly shown. Tesson, in some recent statistics, found it to be from i to 4 in 10. Some authorities give the proportion as i in 50. Pozzi, in 214 cases of cancer of the uterus, found only 6 in which the proc- ess was confined to the body. Within the past six years we have met with 6 cases of cancer of the fundus in 42 cases of cancer of the uterus, i.e., once in every 7 cases of cancer of the uterus. Unfortunately, as happens also in cancer of the cervix, we do not often meet with the cases in their incipiency. Clinically there may be few, if any, symptoms of cancer of the body of the uterus, even although the condition may have been present for some time. Hemorrhage is less frequent than in cancer of the cervix, and does not as a rule appear until late in the disease, first, because the increase in the number of blood-vessels is not so great as in cervical growths; and secondly, because the diseased tissue is much better protected from any external injury, being enclosed on all sides by the uterine wall. In fact there are no pathognomonic symptoms and in many cases the patient appears to be in the best of health. As the disease generally occurs in women between 50 and 60 years of age, we should always be on our guard whenever we meet with a patient who gives a history of a delayed menopause, or with a watery or 'Read before the Qinical and Pathological Section of the Cleveland Academy of Medicine, October, 1903. Copyright, William Wood & Company. 2 ROBB : THE EARLY DIAGNOSIS bloody discharge between the periods, or if after the menopause has been passed at irregular or stated intervals a slight watery or bloody discharge is noticed on her linen. Provided that cancer of the body is recognized before it be- comes macroscopically evident and the uterus can be entirely re- moved, the prognosis for a permanent cure is good, since extension does not usually take place until late in the disease, and generally is not found in the operable cases. Nevertheless, several cases have been reported in which there was secondary involvement in the tubes and ovaries at a relatively early stage (Ries, Lohlein, Wehmer, Reichel). In an analysis of 30 cases made by Cullen at the Johns Hopkins Hospital, 20, or 60 per cent., of the patients had remained free from any recurrence after periods of timt; ranging from 1 1 months to 6 years. Steinbach reports 23 cases, 13 of the patients remain- ing well after 3 or 4 years. Winter reports 30 cases, 16 of the pa- tients showed no recurrence after 5 years. Two out of our three patients are still living and well; one two years and one three years after the radical operation. The association of myomatous disease with adeno-carcinoma of the body of the uterus is not uncommon, inasmuch as myomata are so prevalent. As a rule, however, they are small and of no im- portance. In some instances they are large enough to produce symptoms that mask the cancerous condition. The history of our cases, in brief, is as follows : Case I.^— M. E. D. (patient of Dr. C. M. Hoover, of Alliance, Ohio). Single; age 58; occupation, housework; menstruation be- 'This patient returned June 18, 1904 — three years and eight months after the radical operation — complaining of having had some colicky pains in the lower abdomen off and on during the previous three months. In addition to these pains she had had shooting pains in the vagina and rectum. In January, 1904, she noticed the presence of some blood and mucus when the bowels were moved. In other respects she had been feeling very well. On examination the vaginal walls seemed perfectly free everywhere. In the pelvis and just beneath the promontory of the sacrum a rounded, somewhat irregular and rather hard tumor, of the size of a lemon, could be palpated. The tumor mass was firmly attached to the anterior rectal wall at this point. Per rectum the mass could be felt just beneath the mucosa of the bowel. On opening into the cul-de-sac on June 29th the tumor mass was found to have infiltrated the rectal wall to such an extent that a radical operation was contraindicated. This case demonstrates that although the disease focus was removed while it was apparently confined to the fundus of the uterus, metastases occurred in less than four years. OF CANCER OF THE FUNDUS. 3 gan at 12, irregular until about 20; menopause at 48. Leucorrhea slight in amount. "On admission, Oct. 23, 1900, the .following record was made: Has had 'female complaint' for 20 years. One year ago last July (1898) had aching and swelling in lower abdomen. In April had colicky pains and bleeding from the vagina as if she was going to menstruate ; this lasted for several weeks ; in August, 1900, she had a similar attack and a 'regular hemorrhage.' " "Her previous and family history is negative. She looks worn ; weight 148 pounds ; bowels constipated ; urination frequent." Physical examination of pelvis and abdomen. — Outlet relaxed : vaginal walls prolapsing. Cervix in axis. Uterus in right lateral flexion, somewhat enlarged, movable. Left ovary cystic, not ad- herent; right ovary not made out. The urinary analysis showed a trace of albumin and a few leucocytes before and after operation, but otherwise nothing ab- normal. On October 24, curetting was performed; the material removed was more abundant than usual and there was consider- able hemorrhage. Malignant disease having been diagnosed by examination of the curettings, on October 30 a vaginal hystero- salpingo-oophorectomy was performed. The convalescence was uninterrupted, and the patient was discharged from the hospital on November 27, in good health. Abstract of pathological report. — Curettings. — These are abund- and and present a varying appearance on microscopic examination ; some fragments show a very marked glandular hyperplasia, while in others an interstitial endometritis is most marked. Where the former condition prevails the surface epithelium consists of a single layer of tall columnar cells showing cilia in places ; the cells are crowded together and form a fairly regular surface, but oc- casionally small projecting tufts are seen. The glands are enor- mously increased in number, the interstitial tissue being reduced to a minimum; the size is variable; most of them are small, al- though a number of large glands with secondary gland formations within them are encountered ; the lumina are usually regular in outline and empty, but some contain blood, a clear mucoid secre- tion or a small amount of debris. The glandular epithelium forms a single layer of tall columnar cells closely crowded and showing frequent karyokinesis. The stroma is small in amount and seems completely crowded out from between many of. the glands ; it shows considerable cellular infiltration and the vessels are nu- merous. The appearance of the fragments present a typical pjc- 4 ROBE : THE EARLY DIAGNOSIS ture of the malignant adenoma as described by Gebhard. Other fragments show only a few atrophic glands lined by normal epi- thelium, while the stroma is- very dense and presents considerable cellular infiltration. Still other fragments show a condition inter- mediate between these two extremes. Diagnosis. — Adenoma malignum corporis uteri. The uterus measures 7.5 cm. in length, 4 cm. in width at the cornua and 2.75 cm. in thickness ; it weighs 38 gm. ; the external appearance is normal. On section the cervical canal shows submucous ecchy- moses due to the former curetting; the uterine cavity measures 3.5 X 1.75 cm. It is slightly yellowish in color with hemorrhagic areas. Near the fundus on the posterior surface is a small area, 7 cm. in diameter, slightly elevated, granular, showing a few minute, dilated and cystic glands. Lower down on the posterior uterine wall is a second similar area, about 3 cm. in length. The remainder of the uterine cavity is smooth. In the left lateral wall is a small, interstitial myoma, less than i cm. in diameter (Fig. i). The right tube and ovary are normal, except for a few light adhe- hesions. The left tube and ovary are bound together by adhe- sions; the tube is thickened and occluded; the ovary is distended by a Graafian follicle cyst the size of a walnut. Microscopically the uterine mucosa is scanty, most of it having been curetted away. Vvhen present, it shows a condition similar to that seen in the curettings. Remnants are found on the rough- ened areas noted macroscopically. The muscularis is normal and is not invaded by the neoplasm. The right tube shows a slight OF CANCER OF THE FUNDUS. 5. chronic endosalpingitis with a few surface adhesions. The right ovary presents surface adhesions and the usual senile changes^ The left tube is thickened by a diffuse chronic salpingitis. The left ovary shows surface adhesions and a follicular hypertrophy. Case II.— J. S. (patient of Dr. Dorman, of Ashtabula, Ohio). Age 40; married for 17 years ; Il-para, the last labor 8 years ago, the first instrumental; i miscarriage 12 years ago. The menses began at 11, regular, occasionally painful, duration one week usu- ally, but of late more profuse and of longer duration; flowing when admitted. Leucorrhea, moderate in amount. The present trouble began about one year ago with profuse menstruation which is becoming more pronounced every month. She has had several severe hemorrhages and on admission had been flowing continuously for 17 days. During the past week she has noticed low down in the right iliac region a tumor as large as an apple ; there has been occasional pain in the lower abdomen ; the bowels are loose ; the patient is anemic, the hemoglobin being 55 per cent. The inguinal glands on the left side are slightly enlarged. The urine showed a faint trace of albumin before, together with a few hyaline casts after, operation. Pelvic examination. — Outlet relaxed, cervix far back and crosses axis of vagina, moderately deep bilateral laceration. Filling the cervical canal is felt a friable polypus. The uterus is enlarged and irregular in outline, suggesting a myoma. Both ovaries are small and freely movable. On March 14, patient was curetted, the material being abundant. On March 20, a combined vaginal and abdominal pan-hysterec- tomy Was performed. Convalescence was uninterrupted except for a brief rise of temperature with slight aphasia for a few days after the operation. The hemoglobin was 75 per cent, when the patient was discharged, April 22, 1901. The microscopical examination of the curettings shows the sur- face of the cavity to be very irregular, numerous finger-like growths, varying in length and shape, being present. The central stroma of these is somewhat loose and densely infiltrated with leucocytes. The epithelial covering consists of a single layer, or more often of two or more layers. Many of the processes con- sist of epithelial cells without any stroma. When occurring in a single layer the epithelial cells show elongate oval nuclei, stain- ing fairly uniformly. Hyperchromatic forms are also found. The nuclei of the stratified epithelium vary in size, shape and intensity 6 KOBB : THE EARLY DIAGNOSIS of Staining ; a few very large forms with excess of chromatin oc- cur ; karyokinetic figures are often seen. There is marked leuco- cytic infiltration of the epithelial covering and degenerated areas occur in it. The glands are enormously increased in number and vary from the normal to typically malignant forms. They are very irregular in size and shape, are often lined by stratified epi- thelium and present villous-like epithelial tufts, with or without ■stroma, extending into the gland lumen and cfften reaching the ■stage of secondary gland formation. Some of the gland spaces are completely filled with epithelium ; others contain desquamated and degenerated epithelium, leucocytes or a mucinous material. The epithelium for the most part is pale, swollen and somewhat ■degenerated. It is infiltrated with leucocytes and shows a few karyokinetic figures ; in some places the nuclei are very irregular in size and shape and show an excess of chromatin. The stroma is fairly abundant, although crowded out from between many of the glands, it is densely infiltrated with leucocytes, except in the deeper part. Blood-vessels are very numerous. In places the tissue is quite necrotic. Diagnosis. — Adeno-carcinoma corporis uteri. The tissues re- •moved by hysterectomy consist of the myomatous uterus with TDOth ovaries and Fallopian tubes. The uterus measures '14 cm. in length, 8 cm. in its antero-posterior di- rameter, 13 cm. in width at the cornua, 5.5 cm. in width at the internal os. The peritoneal surface is smooth except near ithe left cornu, where a small myomatous nodule, 1.5 cm. in di- ameter, can be felt in the posterior wall ; near it are several small- er ones the size of peas. The uterine walls are thick and pale, averaging 3 cm. in thickness; the uterine cavity is 4.6 cm. in width. In the anterior wall of the uterus and filling the greater part of its cavity is a myoma 5.5 x 4.4 cm., partly submucous and partly interstitial ; its surface is more or less rough, especially at the upper margin. The tumor is quite firm, except at its lower part where it is necrotic and yellowish-white in color. The left «ide of the uterine cavity extending down to within 1.5 cm. of "the internal os, and part of the anterior wall present numerous little outgrowths, 2.5 mm. high, separated by furrows or sulci. They are most in evidence at the left cornu. The growth has in- vaded the underlying muscularis to a depth of i cm. in one place. The remaining mucosa of the uterine cavity is smooth but some- what injected near the lower margin of the large myoma ; the ■mucosa of the cervix is apparently hypertrophied (Fig. 2). OF CANCER OF THE FUNDUS. 7 Microscopically, the cervix shows a marked interstitial cervi- citis. The mucosa of the body, except in the papillary area, is thin; the glands are few in number and lined by normal pale epithelium. .The surface epithelium, usually in a single layer of irregular cells, is occasionally stratified, but shows no apparent signs of malignant disease. The stroma shows a marked inter- stitial endometritis; the papillary area shows typical adeno-car- cinoma, the appearance being similar to those described in the curettings. The margin is rather abrupt both in the mucosa and also in the muscularis ; the latter shows a slight cellular infiltration before the advancing growth, but at a short distance from it, it is normal. The characteristic feature of the growth is the large number of glands of moderate size and lined by a single layer of •epithelium suggesting the malignant adenoma of Gebhard. The large myoma shows more or less complete hyaline degeneration. At its lower border it is quite necrotic with dense leucocytic infiltration. The right ovary shows a slight oophoritis and chronic peri- oophoritis. The right tube shows an acute catarrhal salpingitis. There is dense cellular infiltration and the lumen contains a small amount of purulent exudate. The left ovary and tube are normal. Case III. — J. McG. ; age 52 ; married 26 years ; Ill-para, oldest •child 25, youngest 20 ; 2 miscarriages, last 22 years ago. Patient O ROBB : THE EARLY DIAGNOSIS well nourished, weighs 165 pounds. The menses began at 13; last regular period 2 years ago. Present trouble. — For the past ten months the patient has had a slight leucorrheal discharge with some odor, and more or less bloody, the color having been brighter of late. The general physical examination was negative. The urine was normal before operation; afterwards it showed a faint trace of albumin and fed and white blood cells. Pelvic examination. — Outlet relaxed; cervix in axis; posterior lip somewhat irregular ; uterus sagging in pelvis, somewhat small- er than normal, movable ; fundus forward ; lateral structures not clearly made out but no adhesions detected. On June 18 the uterus was curetted and a portion of the cervix was excised for microscopical examination. The curettings were abundant, 1 5-20 ' times the normal arnount. Anedo-carcinoma was diagnosed from the microscopical examination. Vaginal hys- terectomy was performed June 27, 1902. The convalescence was uninterrupted. The pathological appearance of the curettings varies in the dif- ferent fragments. As a rule the surface epithelium forms little tufts with or without a stroma basis ; the epithelial cells are mostly -columnar but some are cuboidal. They vary from two to many layers in thickness ; the cells differ in size ; the nuclei lack uni- formity in size, shape and position within the cell, and the in- tensity of staining. A formation of new glands from the surface can be seen. Large numbers of leucocytes occupy spaces between the epithelial cells. The glands are very numerous and in some areas completely displace the stroma. They vary in size and shape, some being much hypertrophied and showing secondary glandular formation within their lumina. The glandular epitheli- um is stratified and often completely fills the cavity ; otherwise the lumina are of various sizes and shapes and contain desquamated epithelium, leucocytes and necrotic debris. The epithelium is in- filtrated with many leucocytes. The cells vary in size and the nuclei also differ markedly in size and staining qualities; many karyokinetic figures are found. The stroma in the most affected areas is almost completely crowded out by the large masses of epithelium. Elsewhere it is less reduced and shows large numbers of vessels and more or less cellular infiltration. Diagnosis. — Adeno-carcinoma corporis uteri. The uterus re- moved by vaginal hysterectomy measures 11.5 cm. in length and 8 cm. transversely. At the right cornu a small part of the OF CANCER OF THE FUNDUS. 9 tube is found, 3.5 cm. long and .5 cm. in diameter. The cer- vix appears about normal except for the loss of a wedge of tissue removed for examination at the present operation. The uterine mucosa, about 2 mm. thick, is somewhat injected, but otherwise normal, except for an area about 2.5 cm. in width near the right cornu where a finely lobulated, soft, friable, yellowish- white papillary or cauliflower-like growth is found. This projects above the surrounding level and also hangs downward somewhat like a polyp over the normal mucosa at its lower limits. The growth also invades the muscularis below it, the boundary line be- tween the two being indistinct. The uterine wall, as a rule, is somewhat thickened, but beneath the neoplasm it is but half as thick as on the opposite side in a similar position. The portion of Fallopian tube seems normal (Fig. 3)- Microscopically the cervix uteri shows a slight cervicitis; the endometrium of the. corpus, as far as the limits of the polypoid growth, is thin with few glands of normal appearance and with a dense stroma, the surface epithelium being nearly normal. The border of the neoplasm is rather sharply marked. It is a typical udeno-carcinoma, consisting of large glands showing the forma- tion of secondary glands within the main lumina and usually com- pletely filled by the proliferating epithelium. The stroma is scant\ and infiltrated with leucocytes, the appearances being similar to lo robb: early diagnosis of cancer of the fundus. those described in the curettings. The growth extends rather deeply into the muscularis, the latter presenting a marked leuco- cytic infiltration before the advancing margin of the neoplasm, while a short distance away the muscularis is practically normal. For the report on the pathological findings and diagnoses, I wish to thank my colleague, Prof. Wm. T. Howard, Jr., and Drs. Wm. H. Weir and Charles D. Williams, of. the Gynecological Department of the Lakeside Hospital. 702 Rose Building. Pathological and Experimental Papers. An Experimental and Clinical Research into Certain Problems Relating to Surgical Operations BY GEORGE W. CRILE, A. M., M. D., Ph. D. Professor of Clinical Surgery, Medical Department Western Reserve University; Surgeon to St. Alexis Hospital; Associate Surgeon, Lakeside Hospital, Cleveland {Abstract of an Essay awarded the Alvarenga Prize for 1901 by the College of Physicians oj Philadelphia) INTRODUCTION THE several subjects comprising this research were suggested by practical experience in the operating room. As most of our exact knowledge of human functions has been gained by animal experimentation, it would seem to be safe to apply the laws thus obtained to surgical practice. This is more particularly true of the vascular and nervous systems. Upon the ready and the exact application of such laws depend the immediate results of many operations. This research extended over three years, during which time the results were clinically compared and applied in an active operative practice. It was exceedingly gratifying to note the close correspondence between the phenomena observed in the operating room and in the laboratory. While surgical anatomy occupies a conspicuous place in literature, surgical physiology, although almost as important, is all but unknown. Unknown, because the physiologic laboratory is a comparatively recent creation, whose important surgical relations are scarcely appreciated. The experimental data, tracings, specimens, etc., have been preserved, and there is no statement made that may not be verified. A sufficient number of illustrations have been added to elucidate the text. Unless otherwise specified the experiments were. made on dogs, taken unselected as they were supplied by the laboratory servant. Every precaution was taken to inflict as httle pain or distress as possible. 2 Crile — Surgical Operations on the el'fect of severing and of mechanically irr't THE VAGI . ^ (The Protocols of the nine experiments ci dogs, of ■"^f^?' *? summary, are published lis the original et,»».j'-/ ^ In the dog the sympathetic fibres run ^^^ --T3^^:S:dt; the vagus proper. Therefore this nerve has been a ^ vago-sympaihetic, so that allowance must be made lor me , ■ r n- i • ,,1-, - open th. ,.o.. d, I ihrust my hand mto the upper angle from nhich tiie hemorrhage came. Although this was quickly done there was a great loss of blood. With my hand grasping his throat he was replaced upon the ward ambulance and taken to the operating room, the house surgeon maintaining, as well as he could, artificial respiration on the way. On returning to the operating room the large vessels were clamped and the forceps allowed to remain. The patient was removed from the ward to the operating room on account of want of light. Ice and heat were alternately applied, which helped to restore the respiratory action. He was given subcutaneous injections of saline solution and 1/200 grain of strychnin every half hour. At the end of twenty-four hours the forceps were removed, and at no time after recovery from immediate effects of the operation were there observed any unusual respiratory or circulatory symp- toms. Patient made a good recovery. The microscopic exami- nation of the growth showed a large number of inclusion cells, indicating a marked malignancy. There was considerable hoarse- ness for two weeks, after which it gradually disappeared. The respiratory failure in this case was in full accord with the experi- mental evidence. Case 3 (Abstract): In removing glandular metastases, fol- lowing total laryngectomy for carcinoma, the left carotid artery and vagus were involved. The pulse and respirations were observed during the excision and after it. A hypodermic injec- tion of 1/100 grain of atropin was given twenty minutes before the operation was begun. The nerve was severed by a quick snip of the scissors. There was some hoarseness for several days, after which it gradually disappeared. No other effects were noted. The patient made a good recovery. Case 3 ; See report under closing the carotid artery. In this patient the vagus was resected. One 1/100 grain of atropin \vas administered prior to Ihc oiperation. No immediate effects were noted. The voice could not, for obvious reasons, be observed. Case 4 (Abstract) : A full charge from a shot gun, whose muzzle was within a few inches of the patient's neck when dis- Crile — Surgical Operations 5 charged, entered the neck in the upper part of the carotid triangle. The artery was torn off and the nerve lacerated. The wadding and shot were firmly packed upon and driven into the torn nerve and other structures of the neck. The pulse was reduced to 42 beats per minute. Respirations were slow, exhibiting quick- ened respiratory action with lengthened pause and prolonged expiratory action. The slow pulse continued more than two hours, after which the "vagal" mechanism went into resolution and an extremely rapid cardiac action followed. Case 5 (Abstract) : In removing a mixed tumor of the parotid, the dissection was carried for some distance upon the carotid and the vagus. While freeing the nerve near the level of the tip of the styloid process the pulse dropped from 90 to 5G. The wound was almost bloodless, exhibiting the various struc- tures in plain sight. A 4% solution of cocain was applied upon a piece of cotton and the dissection carried elsewhere. After three minutes the dissection of the vagus was again resumed. Meanwhile, the pulse had returned to 86. During the remainder of the resection of the vagus there was no appreciable alteration in the heart's action, in spite of a more severe manipulation than had before been given. The depth of the tumor not having been anticipated no atropin had been administered. This case illus- ti'ated the effect of stimulation of the upper portion of the vagus, notably near the point at which the superior laryngeal is given off. This nerve trunk, though not seen at the time, might have con- tributed to the cardiac phenomena from indirect violence. Case 6 (Abstract) : Emergency removal, without anes- thesia, of a large goitre in unconsciousness from asphyxia. Vagus clamped. Recovery. Preliminary administration of 1/100 grain of atropin ; inclu- sion of vagus nerve in large forceps ; no cardiac inhibition ; artificial respiration. Female, aged 64; had had a goitre 12 years, which during the six months preceding the operation had rapidly developed. The tumor was large, quite firm, deeply situated in the neck, extending well down behind the sternum and clavicle, displacing the apex of the lung and producing a compression of the trachea against the vertebral column. The tumor was so firmly fixed as to bn scarcely movable. During two months she had been obliged to make use of the extraordinay muscles of respiration. The devel- opment of the platysma was remarkable. In each inspirator} effort the contraction of the muscle was so powerful as to draw- down the angles of the mouth, lower lip, and the integument of 6 Crile — Surgical Operations the lower part of her face, to throw in vertical folds the entire surface of her neck and upper thorax, and to entirely change the aspect of the neck and face. The lower jaw, meanwhile, was carried fixedly forward and upward. By this means she was able to so slightly relieve the pressure upon the trachea as to accom- plish a stridulous though scant exchange of air. It was planned to expose the tumor under local anesthesia, then try to relieve the obstruction by elevating the tumor while administering general anesthesia. She was placed upon the table, but was absolutely unable to breathe in a lying posture. She struggled for breath, sprang up in a sitting posture, then gasping, turned cyanotic, became unconscious, and fell apparently lifeless on the table. At a single stroke with a scalpel the tumor was laid bare, then literally torn out, while the blood from the many torn vessels flooded her face and neck. Artificial respiration was in the meantime begun. The hemorrhage was first controlled by firm gauze packing, then with a large forceps the common carotid, jugular vein, and vagus nerve were grasped en masse below and likewise above, com- pletely controlling the hemorrhage. The operation was performed in about 45 seconds. The patient had been given 1/100 grain of atropin previous to the operation, which prevented any cardiac inhibition, although the vagus was crushed at the point of clasp- ing. The operation was completed without her knowledge. She said afterwards that she believed she was dying and was con- scious that something was done, but had no pain. Patient made a good recovery. ON THE EFFECT OF INTRAVENOUS INFUSION OF SALINE SOLUTION SUMMARY OP SIXTY-ONE EXPERIMENTS Cir dilation: The intravenous injection of saline solution at or near normal temperature into the circulation of the animal from a height producing a pressure greater than that of the blood usually caused a rise in the blood-pressure. The beginning of this rise appeared as soon as the force of the stream was added to that of the circulation. The rise was sometimes abrupt, some- times gradual, appearing at once or after a lapse of a short time. In almost every instance a point was soon reached above which no amount of saline solution, irrespective of the height to which the bottle was elevated, -could raise it. As a rule the increase in the height of the blood-pressure was but slight, varying from two to six or eight millimetres of mercury. The cases in which there was no immediate rise after the beginning of the infusion were mostly those in which there has been some previous alter- Crile — Surgical Operations 7 ation in the blood-pressure. In a small number of the experi- ments there was immediate temporary fall. This was usually rather abrupt, and in every instance a recovery to the previous level was made or the pressure went even higher. The character of the heart-strokes in many instances was not altered ; but in cases in which it was, an increase in the length of the stroke occurred. The frequency of the heart beats was usually diminished. The general characteristics of the blood- pressure appeared from the time at which it reached its maximum height, after the beginning of the flow, to the time when the beginning of the final decline occurred. The curve was remark- ably even. If at the beginning of the flow the curve was irregular the saline injection usually prevented its continuation, establishing a mean level. When, as a result of the continuous flow, death was produced by an excessive amount, the beginning of the death phenomena was marked by the gradual decline in the blood- pressure and in no instance was it possible to stop the downward tendency. The heart-beats composing this curve were character- ized by their becoming for a time increasingly longer, then gradually shorter, until the last beat. The frequency was gener- ally diminished from the beginning of the final decline, and if, during this descending curve, the animal executed a respiratory movement, even though it were but a gasp, a very marked altera- tion in the blood-pressure curve was produced, that is to say, the final descending curve presented the essential characteristics of death from asphyxia. The preceding remarks apply to the experiments in which normal saline solution at the temperature of the body was given to normal animals in a continuous injection until death. If at any time after a sufiFicient amount of saline had been administered to raise the blood-pressure to the maximum the animal's foot was burned, or other injury inflicted, an additional rise would follow. This rise, however, was not so high as in the cases in which no infusion had been previously given. It was observed that chloro- form, even the inhalation of a few drops, caused an immediate but gradual fall in the blood-pressure, in many instances the fall being marked. Ether, as a rule, produced no change upon the blood-pressure. The effect of chloroform was the more marked the greater the amount of saline that had been previously given. As to the effect of administeriiig the saline at different rates of flow, the more rapidly it was introduced, the more quickh the blood-pressure reached the highest point in that particular case, but the final height would be the same. When introduced 8 Crile — Surgical Operations with great rapidity the animal, though under full surgical anes- thesia, would respond in a peculiar subconscious way, showing a tendency to struggle. The Effect upon the Blood Itself: On making blood-counts by means of a Thoma-Zeiss instrument before and during the saline infusion a decrease on the number of red cells was usually shown. The blood-counts, however, exhibited a great variation in the different experiments. In two they showed an actual increase in the number of red cells, but, as a rule, the numbei fell, generally about 1/4 to 1/5. The blood-counts, though very carefully made, in the same experiment at different stages, showed varying results, and in some instances after considerable diminu- tion there would be a secondary increase. The blood-count was not proportional to the amount of infusion. The tendency to clot seemed to increase with the experiment. This was especially marked in the blood from the liver. The color of the blood grew darker from the time of the beginning of the final descent of the blood-pressure until the end. Earlier its color was a lighter red. There was a very marked increase in the tendency to hemorrhage, especially from the small vessels. Wounds made before the intro- duction of the saline and which had become dry began to ooze soon after the beginning of the infusion. The temperature in many instances was slightly raised. The foregoing applies to a dog under surgical anesthesin in which normal saline solution at the temperature of the body was allowed to flow until death occurred. On Respiration: The respirations were increased or dimin- ished according to the circumstances. If saline was introduced rapidly the alteration, both as to increase in frequency and in the amplitude of the stroke, was more decided. This increase did not continue, but after a period of time, corresponding fairly well with that allowed for the circulatory changes to reach their maxi - mum, the respiratory rhythm returned nearly to the normal. A; the animal became increasingly under the effect of the infusion the respirations decreased in frequency while the amplitude of th.- excursions increased. The abdominal factor of respiration grad- ually diminished until it was lost, and the costal factor alone assumed the burden. This factor soon began to fail — death ensu- ing. When once this tendency was inaugurated, it continued until the end. In every instance after the appearance of these phe- nomena death ensued. There was not even a temporary improve- ment. During the latter part of this period the extraordinary muscles of respiration were brought into action, the respiration Crile — Surgical Operations ' 9 becoming gasping. The respirations always failed before the heart. All the animals died of respiratory failure. The character- istic decline of the blood-pressure curve, referred to under the heading "Effects upon Circulation," were inaugurated after res- piration had ceased, pari passu with the development of the later saline phenomena. Coarse, moist rales of varying intensity devel- oped. They were first heard most distinctly over the pulmonary bases, then over the entire chest, and later could be heard when standing near the dog-board. During this stage of the experiment dulness over the pulmonary bases developed, extending later over other portions. In the greater number of the experiments the tracheal cannula became filled with ' fluid having a consistency varying from a tenacious to a watery fluid. Sometimes it seemed to consist of bloody serum, at other times of a frothy mucus. In some instances this fluid collected in such quantities as to materially interfere with the respiratory action, making it neces- sary to tilt the dog-board and turn out the fluid. The Effect upon Tissues and Organs: There was an increased amount of fluid in the skin. The muscles of the extrem- ities contained a slightly increased amount of fluid and those of the trunk and neck were decidedly more edematous. The con- nective tissue was more moist than normal and occasionally was emphysematous. The heart was always in diastole, the chambers widely distended and usually filled with clots. No fluid was in the pericardium. The heart muscles were somewhat edematous. The venous trunks were everywhere distended, imparting to the touch the sense of a decidedly increased tension. The smaller veins also were distended, especially the mesenteric and subcutaneous. The capillaries and the smaller vessels of the gastro-intestinal tract almost disappeared. In the lungs, underneath the capsule of the kidney, and in the walls of the stomach there was in many cases capillary hemorrhage. The same was noticed, in some instances, under the coverings of the brain. The arterial system was not distended. The portal circulation was greatly engorged. Respiratory Tract: Nose. The mucous membrane of the nose was edematous and usually bathed in mucus. During the latter stages mucus discharged freely from the nose. Trachea. The mucous membrane of the trachea was edema- tous, being more or less filled with clear, though more frequently a bloody fluid. Lungs. In almost every instance the lungs were edematous and ecchymosed. The edema and ecchvmosis were most marked ,10 Crile — Surgical Operations in the bases, diminishing over other portions. The ecchymosis varied from small points of a diffused color up to an ecchymosis so intense as to resemble the liver. In some instances the bases were extremely dark, soggy, heavy, and when fragments were thrown into the water they floated very low. On incising them, quantities of frothy, bloody fluid escaped, but in some the fluid was quite clear. Frequently fluid was found in the thoracic cavity. This in many instances was bloody. The Alimentary Tract: Mouth. The mucous membrane was thickened and edematous. There was free discharge from the mouth consisting of a mixture of a watery fluid and mucus. In some instances there was a free discharge of fluid, probably from the stomach. The pharynx, esophagus and stomach were edematous and the mucous membrane was thickened. The stomach in every instance was much distended with watery fluid. The wall of the stomach was considerably thickened, and at times both the mucous membrane and the serous surfaces were pale, though more fre- quently white. The wall was so edematous that on incising it its histologic layers were to a considerable extent separated and watery fluid oozed from its cut surface. Its cavity was more or less filled with watery fluid. Intestines. The small intestines were white, their walls much thickened and edematous. On incising them the histologic layers were well separated and fluid oozed from the cut surface. Usually their lumen was filled with watery fluid. The large intes- tine presented like conditions, decreasing toward the anus. There was free fluid in the peritoneal cavity and in some instances a large quantity. In many of the experiments the intestines were so filled with fluid that it escaped from both the anus and the mouth. Liver. The liver in many instances was hard and greatlv enlarged. On making incisions, large quantities of diluted blood escaped, at times spurting. Even when the incision was made at one point the entire liver decreased in size pari passu with the flow, and the hardness disappeared. Gall Bladder. The gall bladder was usually filled with bile. Spleen. The spleen usually contained more fluid than normal. Pancreas. The pancreas was in most instances enlarged and edematous. Urinary Tract: The kidneys were but slightly enlarged, and on incising them considerable pale fluid escaped, especially from Crile — Surgical Operations 11 the pelvis. Occasionally there was ecchymosis underneath the capsule. The ureters were somewhat enlarged. When the experi- ment was continued for a length of time the urinary bladder was usually extremely distended, but its walls were not thickened. In shorter experiments there was no distention. experiments in which the abdominal aorta, including THE splanchnic ARTERIES, OR IN WHICH THE latter alone WERE CLOSED In this series of experiments, in which either the abdominal aorta or at least some of the splanchnic vessels were closed before the saline solution was allowed to flow, death ensued before an equal relative amount of the solution had been given, that is to say, the normal dog could take much more saline than an animal whose splanchnic area had been excluded by closing the supplying vessels. Pulmonary edema developed. The animals died of respiratory failure, the blood usually became cyanotic. In most cases the heart showed the effect of asphyxia by beating more slowly but very strongly. The circulatory phenomena were vir- tually the same as in the experiments in -which the aorta of the splanclmic vessels had not been clamped. The respiratory changes occurred first. On making blood-counts, at the beginning of an experiment, it was found that in the cases in which the splanchnic circulation had been modified by clamping one or more of the supplying vessels, the number of red blood cells was strikingly diminished. This was in marked contrast with the comparatively slight changes occurring in the experiments in which such exclu- sion had not been made. At the autopsy in this series of experiments no alterations were found in the abdominal viscera. There was no free fluid in the abdominal cavity. Usually the intestines were cyanotic. The stomach, when its blood supply had been excluded, remained normal, but it was noted in those cases that there was in the stomach and in the intestines a peculiar dark, bloody, gelatinous fluid which adhered closely to the mucous membrane. The walls of the hollow viscera were not thickened as in the other experi- ments. In comparison with the fore extremities, the hind were quite dry. The heart was pale, the pericardium containing some fluid. The heart stopped in diastole; the chambers were filled with diluted blood and there was considerable pressure in the aorta. In almost every instance the lungs were extremely edema- tous, much more so than in the experiments in which the splanch- nic blood supply had not been excluded. Death in every instance was due to asphyxia, as in the first series. 12 Crile — Surgical Operations Effect of the Varying Height of the Saline Column: A series of experiments was also made upon the effect of allowing: the solution to flow from different heights. It was found that increas- ing the height of flow usually raised the blood-pressure. Great irregularity was noted in the change of the . blood-pressure. In some instances there was but a slight rise, in others considerable. In the same experiment, at different times, the rise in the blood- pressure might not be equal to that produced by an elevation made earlier or later; that is to say, there was no direct ratio between the elevation and the rise in the blood-pressure, neither was there any ratio between the rise and the amount of saline the dog had received previous to the elevation of the bottle. Sonic Drug and Other Effects: After a considerable amount of saline had been infused, the inhalation of chloroform, even in small dosage, produced a very marked depression on the cir- culation. Upon the administration of strychnin or nitroglycerin, their usual physiologic actions were noted, though to a lesser degree. Thermal, electrical and mechanical stimulation of the tissues produced the usual effects, e. g., burning or crushing the paw caused a rise in the blood-pressure and an increased respiratory rhythm, manipulation of the larynx, a reflex inhibition of the respiration and the heart, etc. Effect of Varying TevAperature, etc.: No matter what the temperature, the effects were eventually about the same. When the solution was cold, the heart-beats were diminished and the strokes became longer. Frequently at the beginning there would be a fall in the blood-pressure, after which in many cases the visual rise observed in all saline infusions would occur. If the lost pressure was not regained during the flow, it was immediately on cessation of flow, with a return to the previous rapidity and length of strokes. Hot saline produced the opposite effect, vis., an increase in the rapidity and a diminution of the length of the strokes, with a rise in the pressure. On cessation of the infusion the strokes would resume their former rate and length and the blood-pressure would fall to its previous level. The results were then almost ultimately alike. Even extreme variation in the tem- peratures of the solution produced but minor alterations in the temperature of the body. The effects of saline infusion were almost wholly mechanical and physical, — within a reasonable range they were independent of variation in the temperature, the rate of flow, the height above the animal, and the vein into which it was introduced. The mechanical factor consists in adding the force Crile — Surgical Operations 13 of the infusion to the force of the venous blood-stream which raises the venous blood-pressure, thus increasing the output of the heart. The amount of the output has been proved to be dependent upon the venous blood-pressure. One of the reasons why the infusion did not raise the pres- sure indefinitely was the escape of the solution from the circulation at a rate corresponding to the rate of the infusion. The escape occurred principally through the structures that normally absorb fluids — vi.z., gastro-intestinal tract, to a much less extent through the mouth and respiratory tract, and still less into the tissues of the somatic area. The rate of escape of the saline solution through these channels almost equalled any rate of introduction we were able to devise. This was substantiated by numerous blood-counts, showing that after a certain dilution had been reached the count remained about the same. A secondary increase of blood- corpuscles was shown, although an enormous amount of saline was introduced during the observations, but in the experiments i-n which the splanchnic area had been previously excluded, by closing the splanchnic vessels, the dilution increased pari passu with the flow, strikingly diminishing the number of red blood- corpuscles. In these cases, the broncho-pulmonary tract elimi- nated larger quantities than it did in the experiments in which the "leaky" gastro-intestinal tract had not been excluded. It might be supposed that in the experiments in which the. great channel of elimination — viz., the gastro-intestinal — had been excluded by closing the supplying vessels the blood-pressure would be raised higher than in the other experiments, but this was not observed. Neither was it found that the administration of an excessive amount of saline materially interfered with the response of the heart or the vasomotor mechanism to stimuli, as proved by the reflex inhibition from laryngeal manipulation, or crushing and burning the paws, by the administration of strychnin and nitro- glycerin, as well as by the compensation that occurred on tilting the board. Conclusions: The foregoing applies tO' the normal dog under surgical anesthesia. In experiments in which the blood-pressure had been lowered by a reasonable hemorrhage alone, saline infu- sion promptly restored the lost pressure. If the pressure had been lowered by the exhaustion of the vasomotor nervous system by afferent impulses set up by injury of the cerebro spinal or the sympathetic nervous system^, the infusion would restore the pres- sure in proportion to the vasomotor exhaustion ; that is to say, normal saline solution is effectual in shock in proportion to the 14 Crile — Surgical Operations impairment of the vasomotor mechanism. If this mechanism has gone into resolution, infusion is without curative effect. If the impairment is considerable, the infusion will partially restore the pressure, etc. Taking into consideration all the facts, the reason why the blood-pressure is raised but little, if at all, higher than the normal is due to the rapid escape from the vessels and the action of the automatic mechanism in the medulla, which when the pressure rises above the normal diminishes the force and the frequency of the heart-beats and lessens the vaso-constriction in the area of peripheral resistance to reduce the pressure to the normal level. The peripheral resistance determines the height of the blood-pressure, no matter how swift the stream nor how great the volume of blood. The limitations of the effect of normal saline infusion must now be apparent. If the peripheral resistance is lost (break down of the vasomotor mechanism; that is to say, fatal "shock"), no amount of infusion can do more than tempor- arily or partially restore the blood-pressure, and death is inevit- able. If the shock is much increased by regional accumulation of blood ( so-called intravascular hemorrhage) as in operations on the splanchnic area infusion may be effective because the peripheral resistance is still present ; that is to say, the vasomotor mechanism has not gone into resolution. If hemorrhage complicates shock and the vasomotor mechanism is still intact, infusion is effectual. Such proportions may be multiplied. The foregoing deductions explain why injuries of the somatic area, such as mangling of limbs in railway accidents, are fre- quently but little, if at all, benefited by saline infusion. It is true that in almost every case an artificial pulse may be produced, even a pulse of considerable volume, but it is without resistance. It will disappear almost as quickly as it came, and no amount of infusion will sustain the circulation in such a case because the vasomotor mechanism has gone into resolution, destroying peri- pheral resistance, hence no blood-pressure can be created. Illustrative Case (Abstract) : Brakeman, aged 25, in previous normal health, as proved at autopsy, was thrown under the trucks of a railway carriage, losing both legs above the knees. There was considerable hemorrhage. Circnlaiion: Pulse 162, small volume, rhythmic, slight ten- sion. Nails, bluish ; small superficial veins of chest and abdomen prominent; blood vessels of the lips and face distended and of venous color, giving a cyanotic pallor; pulsation in die neck marked, indicating toneless vessels ; surface moist and cold, pre-- -senting bluish tinge. Crile — Surgical Operations 15 Respiration: Rate 39; inspiratory phase, quickened and shortened; expiratory phase, relatively lengthened; pause increased ; extraordinary muscles in light action ; slight rhythmic movement of the larynx and alae nasae. Nervous System: Mind clear and alert ; special senses acute : complains but little of pain ; is restless and begs for water. Discussion of the Physiologic State: The massive mechan- ical irritation and exposure of the nerve endings and nerve trunks produced an excessive action of the vasomotor and cardiac cen- ters, especially the first, leading the exhaustion. Proportionately to the degree of exhaustion the "peripheral resistance" is dimin- ished. Proportionately to the diminution of peripheral resistance the general blood-pressure is lowered and the venous return diminished. Proportionately to the diminution of the venous return the output of the heart is diminished, and proportionately to the diminished output of the heart the volume of the pulse and the general blood-pressure is diminished. This "toneless" state of the vascular system, then, is due to the loss of the "peripheral resistance," which in turn is due to the exhaustion of the vaso- motor mechanism, which is due to the excessive stimulation, owing to massive mechanical irritation of the nerve supply of the lower extremities, by the wheel of a car. In this toneless state the larger arterial trunks are relatively empty, so that the blood thrown into the arterial trunks with each contraction of the heart produced a high but short wave, accounting for the marked pulsa- tion in the arteries of the neck ; the wave then, when it reaches the extremities {e. g., radial pulse) is proportionately diminished or lost, is abrupt in ascent, is not sustained, and has proportion- ately lost its resistance. The capillary circulation is correspond- ingly diminished, producing the pallor which, together with the loss of heat by perspiration, produces the cooling of skin. The diminution of the capillary pressure proportionately lessens the venous flow, thereby causing an accumulation of blood upon the venous side, which accounts for the prominence of the small superficial veins and the bluish tinge of the skin — "cyanotic pallor." The acceleration of the blood-pressure is reduced, the heart-beat is increased in frequency and in force; and, other things being equal, when its intake is diminished it beats with increased force and frequency. (There are many important ques- tions that this discussion has opened, but as they are not pertinent to our present inquiry they will not be discussed.) The respiration is increased in frequency in accordance with a well known law governing its automatic center — ms., the dimin- 16 Crile — Surgical Operations ution of oxygen in the blood causes an increased respiratory action, and when the blood reaches a certain degree of cyanosis the extraordinary muscles of respiration are brought into action whether obstruction is or is not present. There being no obstruc- tion present, and the extraordinary muscles having no resistance to overcome, their work is so light as to give but little evidence of their action — the inauguration of their action being rather an indication of the degree of cyanosis, which in turn indicates the degree of circulatory failure. The amount of blood under these conditions actually circulating through the lungs is pro- portionately diminished, which would be in effect a hemorrhage, an intravascular hemorrhage. The effect upon the respiratory mechanism is essentially the same as in an actual external hemorrhage. Owing to the diminished nutrition and excessive action the respiratory mechanism becomes fatigued, the earlier indications being an increased pause and a quickened inspiratory phase. During 20 minutes, 3,000 cubic centimetres of normal saline solution at 100° F. was infused into the median basilic vein (other features of the case will not be discussed). The pulse was reduced in frequency to 134. Tlie volume was fully as large as normal, the ascent of the wave was abrupt, the fall equally so. The tension remained low. The patient perspired freely and was less restless. Cyanosis diminished. Superficial veins remained about the same. Pulsations of the neck were more marked than before. The patient was not so restless. There was an improve- ment in most of the symptoms, but what of the vasomotor mechan- ism? The ascent and descent of the wave was as sharp as before. There was an enormous increase in the volume, but only slight in tension. The arterial trunks in the neck oscillated even more than before. The superficial veins were even still more prominent. The skin continued to have a cyanotic pallor. These several phenomena show that the vasomotor impair- ment and impending break-down still existed. Though a large pulse was artificially created, the patient's chances for recovery were virtually what they were before. The effects of the saline solution gradually passed away, and an hour later another saline infusion of 1,500 c.c. was given. This time the effect was not so marked and were less sustained than in the preceding. All of the symptoms rapidly grew worse and the patient died at the end of four hours. After the first infusion had been given oxygen was administered. During the administration respirations were dimin- ished in frequency and the cyanosis was lessened. The participa- tion of the extraordinary muscles of respiration was diminished. Crile — Surgical Operations 17 This discussion has been extended for the purpose of defining a well-marked and fairly characteristic group in which saline infu- sion is of but temporary aid and in which it does not alter the essential conditions present. The vasomotor break-down is an impairment that cannot be relieved by saline infusion. Drugs are equally inefifective in these cases for the same reason that, although the heart may be stimulated, there being no peripheral resistance, no blood-pressure (pulse) can be created, and death is inevitable. Drugs having action upon the vasomotor mechanism are equally powerless, because this mechanism having become exhausted, it cannot respond. This also gives a clear reason for the benefits of infusion in cases of hemorrhage alone or hemorrhage with shock, by restoring the normal volume of fluid and adding force to the venous stream. In sliock (vasomotor impairment — i. e., lowered peripheral resistance) the benefit is due to the force added to the venous circulation. The venous pressure falls pari passu with the diminu- tion of the peripheral resistance (vasomotor impairment shock), so that the saline infusion supplies to the venous blood-pressure force which the decreasing peripheral resistance does not ordinar- ily supply. It must be borne in mind that the output of the heart is wholly dependent upon the venous pressure ; that is to say, the infusion may merely tide over a circulatory crisis, during which other means for restoring the circulatory equilibrium and tone must be employed. In cases of dangeous hemorrhage the combi- nation of oxygen inhalation with infusion is beneficial, because the reduction in the number of corpuscles so diminishes the amount of oxygen carried that it should be supplied in concen- trated form, which,' by virtue of the law of diffusion of gases, increases the volume's percent of oxygen in the blood. COCAIN AND EUCAIN SUMMARY OF EIGHTY-NINE EXPERIMENTS Histologic Examination of Cords: In the recovery experi- ments in the cases in which the solution was injected into the cord very great difficulty was experienced in locating the point of injec- tion. Only in those cords in which no asceptic or antiseptic pre- cautions had been taken at the time of injection could the track of the needle be traced with the unaided eye. Careful inspection of the cords in which aseptic precautions had been taken failed to reveal the point of injection. The cords were hardened and sectioned throughout the approximate area of injection. In but few instances was there any degeneration, infiltration, or hem- 18 Crile — Surgical Operations orrhage found. In the non-aseptic cords the track of the needle was readily traced by the slight congestion present. Sections through this area after the cords were hardened showed some small celled infiltration with some degeneration, and in one instance a hemorrhage into the right anterior cornu. The needle had traversed the right half of the cord to within one m.m. of the anterior surface in this case. Effect Upon Peripheral Nerve Trunks: The injection of eucain or cocain into a nerve trunk so as to place all its structures in contact with the drug produces an effectual physiologic "block." By the word "block" is meant such condition of the nerve that neither afferent nor efferent impulses can pass, the conductivity being as completely interrupted as if the nerve were divided. While general anesthesia prevents the appreciation of pain and the production of voluntary motion, it does not prevent such other afferent impulses as those caused by mechanical, thermal, or elec- trical stimulation of the nerve endings or trunks, which produce changes in the frequency and the amplitude of the respirations, in the frequency and force of the heart-beats, and in vasomotor action. Either eucain or cocain injected into a nerve trunk as above described prevents the passing of such afferent impulses, thereby preventing effects upon the respiration, the heart, or the- vasomotor mechanism — i. e., shock. Under general anesthesia, if the paw of an animal is subjected to the flame of a Bunsen burner, after the lapse of a short time the leg is drawn up by the contraction of groups of muscles in a deliberate but rather forcible manner, removing the foot from the flamie. General anesthesia, no matter how deep nor what anesthetic employed, does not pre- vent such action of the muscles. It seems, if the expression may be allowed, to be an "unconscious purposive" action. Either eucain or cocain injected into the path of these afferent impulses prevents this phenomenon. If it is intended to produce an imme- diate effect it is necessary to make a thorough injection. If a little time is allowed to elapse, the solution need not be directh injected into all the parts of the nerve-trunk. Even if injected underneath the sheath, without penetrating the substance of the nerve-trunk, a "physiologic" block may be produced. No unfav- orable later effects were noted. In a number of cases in which the nerves were thus blocked, and the animal allowed to recover, there was but temporary functional impairment, and in no instance was there evidence of neuritis or of degeneration following. The effect of the eucain and cocain upon nerve structures is apparently Ihe same as their well-known general effects upon the protoplasm ; Crile — Surgical Operations 19 that is, they temporarily suspend its functional activity. They form no chemical combination and cause no destruction either of its physiologic properties or of its substance. As to afferent impulses, it was found that the cortical discharges of the brain were blocked, either when they originated as a voluntary action, or when they originated as an artificial convulsion produced by the administration of the essential oil of absinthe. Even powerful electrical currents applied to the nerve trunk, near the block, were found to be incapable of forcing their impulses through the "block." That which has been said of the effect of cocain thus applied to the nerve trunks may be said of like injections into the spinal cord. The effect upon the optic nerve is that of blocking the impulses of the light waves through this nerve, at least par- tially, and were the injections given directly into this nerve the "block" would probably be complete. Either eucain or cocain when applied up>on the medulla or fourth ventricle within a few seconds suspends the action of the respiratory center. This sus- pension is characterized by a gradually increasing slowness of respiration, together with gradually decreasing amplitude, so that within 30 seconds respirations cease. The blood-pressure, in nearly every instance, suffers a profound depression, the nature of which is a gradual decline such as is observed on making h cross section of the cervical spinal cord ; that is to say, the vaso- motor center or paths are anesthetized. Another effect of the application of eucain or cocain upon the medulla or the floor of the fourth ventricle is immediate com- plete general anesthesia and immediate total loss of all voluntar}' action. The corneal reflexes are at once abolished and the pupils are dilated. When the paws of the dog are exposed to the flame of a Bunsen flame the legs are not drawn up, the blood-pressure is not altered, and the heart's action is not affected. It is needless to say that there are no respiratory changes. In other words, application of these anesthetics upon the medulla or the floor of the fourth ventricle suspends temporarily all the manifest func- tions of that organ excepting the heart's action, and that is modi- fied. General anesthesia may be indefinitely prolonged by repeated applications. Upon the vagi the effect of an injection of these drugs is to suspend their inhibitory action. The action of cocain is probably a little more prompt than that of eucain, the latter, however, seemed to be quite as effective as the former. Cocain and eucain block the impulses set up by electrical stimulation in nerve-trunks even after death; that is, if after the death of an animal a nerve-trunk is stimulated, within a certain time the 20 CRIL&— Surgical Operations muscles supplied will be thrown into contraction; but if cocain or eucain is injected into the nerve-trunk and a stimulus applied above it, no contraction will occur. The physiologic action of cocain and eucain, both local and general, are so nearly alike that one description may serve for both. The first effect, observed after the intravenous injection, is a temporary increase in the blood-pressure. This increase appears almost immediately, and continues for a brief period of from five to 20 seconds, when the blood-pressure returns to or near its former level. The heart-strokes forming the curve are usually a little shorter, and the rapidity of the heart's action some- what increased. No definite vasomotor change was indicated by the peripheral venous or peripheral arterial manometers. A water manometer recording the splanchnic blood-pressure indicated a rise out of proportion to the rise in the general blood-pressure. In rare instances there was a fall in the blood-pressure, but com- pensation was immediately inaugurated and the lost pressure was quickly regained. In overwhelming doses with lethal effect the general blood-pressure, in fact all the pressures, rapidly sinking to the abscissa line. It was noted that when the animal was under the systematic effect of these drugs the blood-pressure curve was, as a rule, not so regular as under normal conditions. It was also noted that in a number of experiments the length of the stroke of the writing style, expressing the heart's action, was shortened under the systematic effects of these drugs. This irregularity of the blood-pressure curve was similar to the irregular curve when the animal was under physiologic dosage of atropin or when both vagi had previously been severed. It was also found that when animals were under the effect of these drugs, stimulation, by applying the electrodes upon the vagi, did not produce normal characteristic effect; that is to say, that while in normal condi- tions the application of- a Dubois-Reymond electrode upon the vagi causes slowing or arrest of the heart, in animals under the systematic influence of cocain or eucain the application of such stimulation to the vagi in most instances produced little or no effect. In a number of experiments it was observed that if, after having secured a control tracing of the inhibitory effect of intra- laryngeal manipulation, the animal was subjected to a physiologic dosage of cocain or eucain, a like manipulation of the lar3'nx usually produced no inhibition. In the experiments in which inhibition was noted it was in most instances less than normal. The same may be said of other experiments upon the superior laryngeal nerve. The physiologic effect oi cocain and eucain in Crile — Surgical Operations 21 this respect is quite analogous to that of atropin, though the effect is not so marked. The increase in the blood-pressure after the administration of cocain is in a measure similar to that which follows section of both vagi. Taking into consideration all of the evidence, it would seem that cocain and eucain partially or wholly suspend the inhibitory function of the vagi, whether pro- duced by direct or indirect stimulation. While not prepared to make a positive statement on the subject, it appears that the increased rapidity of the heart's action under the influence of these drugs was due to the removal of the vagal influence and not to stimulation of the accelerators. The splanchnic area, especially the veins, when the abdominal viscera were subjected to exposure or irritation, or both, was dilated, the intestines became red, extremely congested, and often livid. When the animal had been given a physiologic dose of cocain or eucain and exposed to like experiment the splanchnic vessels did not dilate, excepting those at the bases of the intestines. The arteries became decidedly smaller and the intestines a peculiar palish red. In a large series of control experiments it was found that, with but rare exceptions, such irritation or exposure of the splanchnic area caused a fail in the general blood-pressure proportional to the exposure or irritation and the condition of the animal. In some instances the fall was extremely rapid and the animal soon died, but in a series of experiments in which cocain was systematically administered there was but a slight, if any, fall in the general blood-pressure. There was a striking difference between the results in the control experiments and the "cocain" ones. In order to make the com- parisons more reliable double experiments were performed. Two animals of as nearly the same size and under as nearly the same conditions as possible were placed side by side on similar dog- boards, and precisely the same experiments were performed simul- taneously upon each. In every instance the benefit of any doubt was allowed to control the dog. The writing style recording the blood-pressure and the respiratory action of each was placed in a vertical line, so that direct comparisons could be accurately made. The result of these double experiments may be summar- ized as follows : In the control dogs exposure and manipulation of the intestines produced a fall in the blood-pressure; in the cocain and eucain dogs, as a rule, no fall occurred. The cocain and eucain dogs endured more mechanical injury than the control dog. The latter in every experiment died first. In burning the hind feet in thertional to the completeness and the anatomical parts involved. The curve in the descent of the blood-pressure was gradual and even after which a regular line was maintained for some distance, which indicated the loss of vasomotor control. The efifect upon the blood-pressure when the medulla or the fourth ventricle was cocainized was the greatest of all, the pressure falling almost to the abscissa line. When all the cord had been subjected to the_ influence of cocain and the pressure had fallen as above described, if any part of the body below the level cocainized was subjected to burning, crushing, or any other mechanical, thermal, or elec- trical stimulation, no rise in the blood-pressure occurred. There was usually but a trifling amount, if any, of compensation after the fall of the blood-pressure until the cocain effects had passed off. Effects on Respiration. The immediate effects on respira- tion, of a subarachnoid injection of a comparatively small amount of cocain in any part of the spinal cord, not involving the medulla, is acceleration. The application of a 1% solution of cocain upon the medulla or the floor of the fourth ventricle produced 24 Crile — Surgical Operations within a period of time ranging from a few seconds to a few minutes, complete respiratory paralysis. There is first loss of the intercostal and extraordinary muscles, then the abdominal muscles and lastly the diaphragm. The action of the diaphragm becomes shallower at each contraction until it is entirely paralyzed. The membranes of the cord are so nearly inelastic that for the present purpose they may be regarded as being so, while the cord itself is so nearly incompressible that it may be considered so. The subarachnoid space is always filled with its own fluid. If additional fluid is added, it must cause a displacement similar to that of fluid in a capillary glass tube. The rapid and uncon- trollable ascent of the anesthesia of the cord was most striking. In order to better study this a series of injections was made with cocain solution colored with methylen blue. It was found that an ordinary injection of the lumbar region of one-half dram of this solution stained the entire cord and the under surface of the brain within 30 seconds. All the various localized functions of the cord and medulla were with rapidity anesthetized. The respir- atory center in the medulla, for example, could be anesthetized by lumbar subarachnoid injection within a few seconds, so rapidly did the fluid pass up the cord. Marked fall in the blood-pressure and cessation of the respiration occurred within a few seconds, after a rather forcible injection in the lumbar subarachnoid space. The fluid ascended about as readily in the vertical posture as in a horizontal. There can be but little doubt that the effect is due to the local contact of the nerve structure and not to absorption. This view is in full accord with the action of cocain on other nerve tissue. A solution injected with considerable force into the lumbar subarachnoid space was attended immediately by convul- sions. The convulsions were due to the stimulation of the con- vulsive center in the medulla. The dosage used in these experi- ments was purposely made large to determine the control, or rather the want of control, the operator could have upon the extent of the anesthesia. In control experiments in which normal saline solution was injected into the spinal cord an immediate fall in the blood-pressure occurred, but compensation quickly followed. The respirations were but slightly afifected. There was the most striking difference between the overwhelming par- alysis in the one case and the want of it in the other. The experi- ments showed that the operator has but little control over the extent of the anesthesia produced under the subarachnoid injec- tion. While direct injection into the cord gave a complete control of the extent, it produced a distinct mechanical lesion. In the Crile — Surgical Operations 25 clinical reports of the subarachnoid anesthesia the experimental data have been corroborated. This was most strikingly exhibited in a case described by Fowler in which the anesthesia during three minutes extended up to the level of the clavicle^ at which time the patient became cyanotic and artificial respiration was necessary. Other observers have noted the marked effect on res- piration, the lowered blood-pressure, and the rapid pulse, the latter Indicating that the cocain solution was affecting the centers of the, medulla. In 692 cases there were six deaths that were attrib- utable to the anesthesia, a mortality rate at least fifty times greater than that of chloroform. ON THE CLINICAL APPLICATION OF THE EXPERIMENTAL EVIDENCE Operations on the Extremities: Leg. The "blocking" method was employed independently by Dr Rudolph Matas, of New Orleans, of which his brilliant monogram on anesthesia gives a full account. Applying the so-called physiologic "blocking" properties of cocain or eucain to surgical practice, we have been enabled to perform certain operations upon the extremities with- out causing pain and without shock by injecting a 1% solution of cocain into the supplying nerve-trunks. The external cutane- ous nerve is so superficial that it is readily accessible. The anterior crural is readily exposed in its relations with the artery and the sciatic at the margin of the gluteal fold along the inner border of the biceps muscle. In operations performed upon the area supplied by the "blocked" nerve-trunks the afferent impulses cannot reach the central nervous system. There is, therefore, neither pain or shock. This method is of the greatest possible importance in operations in which general anesthesia is for one reason contraindicated. The operations under these circumstances cannot cause any more shock than if the member had no connec- tion with the body, as the "block" for all such purposes is equal to a physiologic amputation. In this manner I have five times performed amputation of the leg below the knee, and in all but one the patient was not aware that the operation was performed until told of it afterwards. It is necessary to control the patient well. After preliminary preparations have been made, the patient's attention should be diverted. I have usually said that an examination and a dressing would be made requiring considerable time and that the operation would be performed next day. In the meantime the eyes were covered. In the one case the patient became aware of the pro- gress of the operation by hearing the noise of the saw while divid- 26 Crile — Surgical Operations ing- the bone. The "block" continues from 25 to 30 minutes. The clinical observations are in entire accord with the experimental evidence. Operations in the Area of the Distribution of the Ulnar Nerve: Remarks. The superficial position of this nerve at the elbow joint enables the surgeon to apply a cocain or a eucain "block" almost painlessly by inserting a hypodermic needle, first into its close vicinity, then into the trunk itself, injecting the solu- tion on its way. After the lapse of ten minutes the entire area supplied will be rendered anesthetized, and if the patient's atten- tion is diverted operative procedures, such as amputations, resec- tions, may be performed painlessly and without the patient's knowledge. REPORT OF CASES Case 1 : A railway employee whose hand and little finger were severely crushed within the area of the distribution of the ulnar nerve required amputation and revision. Bending the elbow, a wheal was produced by injecting a 1/12% solution of beta eucain, thereby creating a painless path to the nerve-trunk which was then anesthetized. In a few seconds there was complete anesthesia, and the finger and the corresponding metacarpal bones Avere removed while the patient was an interested spectator. The night following this operation the patient complained of a burn- ing sensation over the distribution of this nerve. There was some local tenderness at the point of injection, but this disappeared after several days. Case 2 : A small boy discharged a pistol, which took effect in the uFnar side of the hand, tearing away the soft parts and a portion of the fifth metacarpal bone. By "blocking" the ulnar nerve at the elbow the wound was revised, and the fragments of the bone removed without pain. In this case there was no com- plaint of the burning sensation described in the preceding. The wound healed readily. Case 3: In a tubercular patient a local focus appeared in the metacarpo-phalangeal joint. In performing an operation for the removal of this focus, the ulner nerve was "blocked." At first an attempt was made to secure anesthesia by injecting the solution around the nerve, but after waiting five minutes it was found that anesthesia was only partial and that it was necessary to inject the nerve itself. In performing the operation it is best to fix the nerve well against the bone and insert the needle grad- ually, as anesthesia occurs in advance of the needle. After such Crile — Surgical Operations 27 an injection the anesthesia was complete and the operation was perfornKd painlessly. The patient complained of some burning the first nig-ht, but the second day it decreased and was not again experienced. Observations: In two other cases this nerve was in a similar manner "blocked." In each the anesthesia was complete in five to 10 minutes and no after effects were noted. In no case was there any interference with the function of this nerve in conse- quence of this injection. Neither did the points at which the injection was made remain tender. No neuritis followed. Amputations at the Shoulder Joint: Amputations at the shoulder joint are usually indicated on account of a serious acci- dent or disease, and in consequence such operations are frequently performed under unfavorable circumstances. There has been a considerable mortality following this operation, even under the more favored conditions. In operations for malignant disease in the aged, and in operations in the presence of profound depression or shock, general anesthesia adds seriously to the danger. There are many instances of contraindication to the use of general anes- thesia. A method by which this operation may be performed without general anesthesia, without shock and without hemor- rhage, was devised in accordance with the experimental evidence set forth in the preceding pages, and put into practice in June, 1898. Technique: The technique is based upon the fact that nerve- trunks may be safely and efifectually subjected to a physiologic "block" by injecting cocain or eucain in a comparative weak solu- tion, and that arteries may be, with entire safety, temporarily closed without injuring their walls. Fortunately in the application of these principles, in amputation of the shoulder joint the sub- clavian artery is in close surgical relation with the brachial plexus so that the same incision may be utilized for exposing the nerve and the blood supply. REPORT OF CASES Case 1 : Female, aged 74, was suffering from sarcoma of the arm, situated in the lower third and extending well down to the elbow. There was a metastatic growth in the axilla. She was suffering great pain and the tumor was growing rapidly. Owing to her extreme age, an amputation at the shoulder joint by the methods hitherto in vogue, giving a general anesthesia without "blocking" the nerve-trunks to protect her against the afferent impulses caused by the mechanical irritation of the amputation. 28 Crile — Surgical Operations thereby producing shock, would have been a risk too great to assume. It was decided to perform the operation by employing the technique above described. An incision was made along the outer border of the sternomastoid muscle under 1/10% mfil- tration cocain anesthesia. The incision was carried through the superficial and the deep fascia, exposing in the first part of the incision the external jugular vein. The lower part of the incision was carried well down on the clavicle. The omohyoid muscle was retracted downward, the anterior angle of the trapezius backward and the posterior margin of the scalenus anticus forward, thereby exposing the trunks of the brachial plexus, and by extending the dissection a trifle farther downward and inward the arching sub- clavian artery was brought into the field. In making this dissec- tion it is important to keep the field of the operation entirely free from blood, so that the translucency of the tissue will permit the ready recognition of the anatomical structures in their minutest detail. It will then be possible to detect small nerve twigs before they are encountered and enable the operator to subject them to local anesthesia in advance. In this way the area supplied by these branches may be rendered anesthetized. The small vessels may be caught with narrow bladed forceps between which the incision may be carried. The smaller nerve twigs are usually found running along the blood vessels or in the connective tissue planes. It was observed in this dissection that in the deeper struc- tures the sensory nerve supply is not so abundant as in the more superficial. After exposing the trunks of the brachial plexus, there being but a slight amount of pain in the dissection, they were subjected to a physiologic blacking by injecting first on their outer covering, then into the substance, a J^% solution of cocain, just sufficient to cause a localized swelling. It required but a small amount of solution to accomplish this. After injecting each trunk there was a total loss of sensation and of motion in all the parts supplied by the brachial plexus. The subclavian artery was then closed by means of a special clamp, over the blades of which rubber tubing was drawn. The blades were then approximated by adjusting the screw sufficiently to close the lumen of the vessel. The patient was then told that the operation would not be performed at that time, but would be deferred until the next day. A towel was thrown over her eyes, and under the pretext of making a careful examination of her arm the amputa- tion was made without her knowledge. The flap on the outer and posterior aspects over the deltoid was made rather low because of the subcutaneous distribution of the branches of nerves Crile — Surgical Operations 29 from the cervical plexus, which of course had not been included in the physiologic block. She experienced no pain except a slight one as the incision was carried around the posterior surface of the upper portion of the arm supplied by the supra-acromial nerve. The pain was, however, comparatively slight, and was felt only during the incision of the skin. During the disarticulation the patient was not aware that she was being touched. After the operation had been completed it was found that there was abso- lutely no shock and that the operation had made no appreciable impression on her. The vessels were all picked up and tied before releasing the clamp from the subclavian. The total amount of cocain used in the operation was about one-eighth of a grain. A portion of this amount was recovered by sponging away the free solution in the wound. When the patient was returned to her bed the patient was not aware that her arm had been removed. She soon missed it and manifesting some excitement was informed by the nurse. She experienced some pain a few hours after the operation and vomited several times the first night. She made a good recovery from the operation and there was nothing in the after-progress of the case different from operations performed in the usual way. Case 2 : Amputation of the arm at the middle was per- formed by the same technique as the preceding without producing any pain and without the slightest shock. Patient made a good recovery. The operation was performed on account of moist gangrene of the forearm in a patient having advanced pulmonary tuberculosis. Amputation of Half the Shoulder Girdle: Remarks. This operation has been performed a number of times by various methods. The purpose of discussing it is to point out a technique by means of which hemorrhage and shock may be wholly avoided. Under general anesthesia an incision is made over the clavicle and the inner half of this bone is resected, after which the subclavian vein and the trunks of the brachial plexus are exposed. The trunks are then subjected to a physiologic block of cocain and eucain in comparatively weak solution — say J4%i. The brachial plexus is next severed and the artery and vein closed by ligature. The incision for the further technique in removing the scapula will vary with the object for which the operation is done. The amount of shock will be limited to what will be produced by mak- ing the incision through the structures supplied by the nerves from the cervical plexus, which is almost nil. 30 Crile — Surgical Operations Observation of the Pharynx: Clinical experience, as well as physiologic experiments, have demonstrated that when the pharynx is subjected to a considerable manipulation, especially that portion nearest the glottis, reflex inhibition both of the respir- ation and of the heart may occur. The respiratory inhibition is the more frequently produced. In the cases in which manipulation required considerable force the heart may be inhibited, causing collapse. I have observed this reflex inhibition of both the heart and the respiration in removing a tumor of considerable size from the nasopharynx. In operations for removing adenoid growths from the nasopharynx these phenomena have also been observed. In extracting large foreign bodies collapse may be produced. Not infrequently, in performing difficult operations in this portion of the pharynx, reflex inhibition confuses the operator. The respira- tory inhibition is likely to give the impression that the patient is suffering from mechanical obstruction. The inclination might be to clear out the upper respiratory passage, but this additional irritation would increase the symptoms. In the experiments it was found that refle.x inhibition in this area may be prevented by the local application of a 2% solution of cocain. The solution may be as weak as 1%, or even jAfo, and be effectual A hypo- dermic injection of atropin prevents reflex inhibition of the heart. In cases necessitating the removal of adenoid growths and tumors of the pharynx the efficiency of these drugs was proved. It is advisable, before beginning the technique of an operation involv- ing this area, to make a local application of a solution of eucain or cocain, and a hypodermic injection of atropin, to prevent reflex inhibition. If during an operation inhibition does occur, the dis- tinction between inhibition and obstruction must be borne in mind, for if the case is one of obstruction there will be increased respira- tory efforts, but if it is a reflex inhibition respirations instantly cease. In obstructions the pulse continues unaltered for some time before it becomes markedly slower. In reflex inhibition the pulse is instantly and markedly slowed or arrested. Laryngotomy: Remarks. Not infrequently in this opera- tion at the moment the larynx is opened the patient goes into a state of collapse from which he may never recover. This opera- lion is more frequently performed on children, oftentimes in great haste, under the stress of circumstances. If the operation is performed through the cricoid, collapse at the moment of enter- ing does not occur. If made higher, it is very likely to occur the reason being that in the higher operations the inhibition area of the larynx is mechanically stimulated. This causes a reflex Chile — Surgical Operations 31 inhibition, as in operations upon the pharynx. The superior laryngeal nerves are endowed with very strong inhibitory func- tions, which are more active in the upper part of the larynx. The clinical observations are in entire accord with the experimental evidence. REPORT OF CASES Case 1 : Dr M. called in a colleague to aid in performing a laryngotomy upon a child who had a grain of corn in the larynx. The operation was successful until the larynx was opened, when suddenly collapse occurred, resuscitation seemed impossible. Dur- ing the first stages of the collapse the corn was removed. Arti- ficial respiration was maintained for a time, though life seemed extinct, when suddenly respirations began and there was an uneventful recovery. Case 2 : I was called to see a child three years old having a large bean lodged in the larynx. The history of the case was that while the child was playing with the bean, in a fit of laughter, inspired it into the larynx. Paroxysms of coughing followed. These occurred at intervals. Each time the child stopj>ed breath- ing, became cyanotic and apparently dead. After a brief interval respirations returned, another paroxysm soon followed with a repetition of the collapse. From these symptoms alone the loca- tion of the bean was diagnosed as being in the upper part of the larynx. An operation under local anesthesia was performed. The incision was made through the cricoid cartilage, below the so-called inhibition area, and the laryngeal mucosa was treated with a 2% solution of cocain, after which the larynx could be readily explored, the bean located and removed without inducing reflex inhibition. Observations: Cases might be multiplied, but the foregoing are typical. The difference between reflex inhibition and obstruc- tion is very marked. Reflex inhibition cannot be produced by a foreign body at any point below the so-called inhibition area of the larynx. The importance of the use of local anesthesia, to pre- vent reflex inhibition in laryngeal operations, cannot be over- stated. In all the operations upon the larynx, especially in laryn- gectomy and intralaryngeal procedures, the use of cocain and eucain is of the greatest importance. In laryngectomy especial attention has been called to the collapse that not infrequently appears while removing the larynx from its attachments. Bardenhauer encountered this three times in one case while 32 Crile — Surgical Operations inflating the Trendelenburg apparatus. In intubation sudden death frequently occurs, the collapse being due to reflex inhibition of either the respiration or the heart, or both. Cocain or eucain applied on the mucous membrane wholly prevents such reflex inhibition. If such local application cannot be applied, all the necessary arrangements for the maintenance of the artificial res- piration may be made in advance. A hypodermic injection of atropin will prevent the cardiac inhibition, so that, without the use of local anesthesia, atropin, with artificial respiration, may be depended upon to carry the patient over the inhibition crisis. In one hundred and fifty-six intubations I have encountered reflex inhibition six times, twice fatal, and they occurred before the nature of the inhibition was comprehended. Since making use of the experimental data, no case has been lost from reflex inhibition or "laryngeal collapse." Death cannot occur as a result of reflex inhibition if a pre- liminary hypodermic of a physiologic dose of atropin is given. The use of cocain is not practical in intubations for diphtheritic stenosis. Clinical Summary: In the clinical use of cocain and eucain particular attention is called to a most important feature — viz., that shock is almost wholly avoided as all afferent impulses are Hocked. It is now known that afferent impulses set up by injury or operation are the causes of shock. These impulses are but slightly modified by general anesthesia. The afferent impulse, constituting pain, is abolished by general anesthesia, but those affecting the vasomotor, the respiratory, and the cardiac mechan- isms are not controlled; but cocain or eucain absolutely blocks their passage, making a physiologic amputation of the part. These anesthetics wholly prevent reflex inhibition, the principal causes of collapse in operations and injuries — e. g., operations on the larynx and pharynx. Given hypodermically, the experimental evidence shows that they diminish shock in operations on the splanchnic area and absolutely alter this area in the processes of operation or exposure, as abundantly proved by the series of double experiments. I have had but two opportunities of testing this clinically, both in operations for gun-shot wounds of the intestines, and in each the experimental evidence seemed to be corroborated. Com- parative results require such a large number of observations that I prefer for the present to offer no more than the clinical suggestion. Crile — Surgical Operations 33 on the effect of temporary closure of the carotid arteries summary of nineteen experiments Histologic: The gross specimens presented an oval outline on section at the constricted portion. This flattening of the artery was more marked in those carotids which had been clamped for some hours and in those in which the clamps had been tightly adjusted. The histologic appearance of arteries clamped for short periods and examined at once showed but slight change. Arteries clamped for periods of from 15 minutes to half hour showed little effect other than a slight tearing of endothelium at the extremities of the oval. Those clamped for an hour showed a greater amount of distortion of the endothelium at the margins of the oval with some separation and endothelial cells, which were massed between folds of fenestrated membrane. The elastic layers were slightly distorted at the constricted portion. The elements of the middle layer were massed at the extremities and somewhat disarranged. The adventitia was unchanged. The histologic appearance of the carotids from the recovery experi- ments varied with several conditions. The amount of pressure exerted by the clamp, the presence or absence of wound infection, and the length of time the clamps were allowed to remain on the artery, modified the results. Sdme specimens clamped too tightly for four or six hours showed marked degeneration of the middle coats with edema and a thickening and disarrangement of the intima, with loss of endo- thelium and a very perceptible narrowing of the lumen ; others \vere thrombosed, some were necrotic; but in those carotids in which care was taken to so adjust the clamps as to exert only sufficient pressure on the artery to close its lumen, the histologic changes were unimportant. A clamp adjusted too tightly caused pressure necrosis in a few hours, while other carotids were clamped for from 24 to 48 hours without notable damage to the arterial walls. The intima and elastic membrane were but slightly affected though the media showed some evidence of degeneration. The adventitia was but slightly altered. The presence or absence of infection of the wound was of great importance. In those cases in which an infection appeared the arteries showed the greatest changes. In many instances the artery was necrotic at the clamped portion, and in some instances it was severed. The media and adventitia, both above and below the constricted portion, showed round celled infiltration and in some areas necrosis. The intima and the inner elastic membrane were disorganized and distorted. In those thrombosed, the lumen was narrowed owing to edema and thickening of the walls. 34 Crile — Surgical Operations The experiments showed that a properly adjusted damp could be left in position, closing the artery for from 24 to 48 hours, without serious injury to the walls. Physiologic: The immediate effect on the circulation of temporarily closing one carotid artery was to increase the blood- pressure, but usually a compensation followed, and the pressure returned to its normal level. No effect upon the respiration was observed. Simultaneously closing both carotid arteries produced a greater rise in the blood-pressure, which by physiologic compen- sation usually soon returned to the normal level. In many of the experiments there was a decrease in the respiratory action, although the effect was very slight. In no instance were there any striking results noted. In the recovery experiments in the cases in which the clamps were allowed to remain on the arteries, clos- ing them and the wound pursuing an aseptic course, no effect upon the animal was observed beyond that attributable to the anesthesia and the operation. The animals seemed playful and strong. Even after 24 hours of complete closure there was not much microscopic evidence of injury to the vessel wall. The circulation through the clamped portion was readily re-established. However, in cases in which, during the application of clamps for a considerable length of time, say two days, the animal in the meantime had suffered infective inflammation of the wound, the damage of the vessel walls was very considerable and the lumen was in some instances occluded. As to the after effects, in no case was there clotting ; the aseptic cases made good recoveries ; the circulation was re-established ; and no impairment of conse- quence was observed. The circulation of the brain was carefully observed at the post mortem, and in no case was either emboli or thrombi found, or any effect on the brain noted. After consider- ing several devices, the most accurate, efficient, and safe one seemed to be that of applying a clamp, so constructed that its blades could be adjusted by means of a set screw, and when they were approximated so as to close the vessel, but not compress its walls, the blades were parallel to each other. One blade was made longer than the other, and its end turned up so as to pre- vent the escape of the artery. Over these blades were stretched pieces of rubber tubing, thereby minimizing the effect of contact with the vessel wall. In applying the same, it is necessary to bear in mind that the walls need only be approximated, not compressed. The adjustable screw gives so perfect a mechanical control of the lumen of the vessel as to enable the surgeon to perform the operation and secure the bleeding vessels with a minimum loss of blood. Crile — Surgical Operations 35 clinical application a new method op controlling hemorrhage in certain opera- tions on the head and neck Technique: Twenty minutes previous to making the incision one one-hundreth of a grain of atropin sliould be injected, in cases in which the technique is Hkely to involve the trunks of the vagi or their superior laryngeal branches, for the purpose of preventing possible inhibitory action upon the heart. Each com- mon carotid artery is closed by means of a small clamp, whose blade is long and protected by a thin rubber tubing. The lower blade is slightly longer than the upper, and turns up at its free end so that its grasp upon the artery will be more secure. The spring end of the clamp is so arranged that when the blades are closed sufficiently to approximate the walls of the vessels they become parallel. The closing of the blade is accomplished by an adjust- able thumb screw, making definite closure. In operations in which blood may enter the pulmonary tract, the patient should be placed in a Trendelenburg pvosture. This partially compensates the low- ered cerebral blood-pressure resulting from closing the carotids. While this posture somewhat increases the venous and capillary hemorrhage, the increase of the venous pressure diminishes the danger of the entrance of air into the larger venous trunk, should they accidentally be injured. Fortunately venous and capillary hemorrhages, except in cases involving the larger veins, are rela- tively of little consequence. On completion of the operation, in cases in which the Trendelenburg posture is employed, it is safer to restore the patient to the horizontal position before releasing the carotids, as in the inclined posture the normal blood-pressure of the brain is increased by the mechanical factor, and releasing the clamps in this position would raise the pressure above the normal. The release of the clamp should be made slowly while inspecting the field of operation to detect any vessels that might have been overlooked. The control of the arterial hemorrhage is absolute, except in such vessels as received direct collateral pressure from the vertebral arteries. According to the researches of Bayliss and Starling, there are no vasomotor nerves supplied to the vessels of the brain. The circulation being mechanical, the blood-vessels of the brain should be more favorable to the employment of such technique than the vessels of almost any other organ of the body in which the blood supply is more or less regulated by the vasomotor mechanism. 36 Crile — Surgical Operations reports of cases Case 1 : Both common carotids closed. Recovery. Opera- tion was performed January, 1897. The patient upon whom this technique was employed was a colored mian, 46 years of age, admitted to St. Alexis Hospital on account of a large fibrosarcoma filling the mouth so as to render its complete closure impossible. Breathing was so obstructed as to threaten suffocation, and at night was so heavy and labored that it could be heard at a considerable distance. Under cocain, tracheotomy was performed for the double pur- pose of removing the danger of asphyxia and as a part of the technique to be employed. Aside from emaciation from the nec- essary liquid diet, the patient was in good condition. The tumor was first observed six years previous in the posterior part of the hard palate. A year later it was removed, but recurred, and had been growing since that time. On account of its size, the extent of its attachments, when I first saw him, could not be determined. Its translucent surface displayed a rich supply of blood-vessels, some of considerable size. In the operation the technique here described was employed. The tumor was deHmited by an incision in the healthy mucous membrane. The hard palate was divided along this line. The vomer was severed along the floor of the nose and the entire mass turned out. After the necessary revision of the wound the principal vessels were secured, and everywhere the wound was touched with the thermo-cautery. After the operation had been completed, the wound dressed, and perhaps 10 minutes had elapsed, respiration suddenly failed. Artificial respiration was maintained during 25 minutes. The application of ice, alternated with a brisk rubbing with a warm towel, proved an efficient stimulus to respiration. The entire mouth was packed with iodoform gauze, which was allowed to remain for 24 hours, after which boric acid solution with sufficient thymol to correct the disagreeable odor was used in a mouth wash. The tracheal tube was removed after two weeks, when it was thought that the danger of pulmonary infection had passed. The patient was soon able to leave his bed, and made an uneventful recovery. Four years later there was no recurrence. Case 2 : Removal of a large congenital tumor of the neck. Closure of both common carotid arteries. Recovery. Female, 21 years of age. At birth it was large and more developed on the right side than on the left, greatly increasing in size as she grew older. At the time of operation the tumor occu- Crile — Surgical Operations 37 pied the entire anterior and much of the latter portion of the neck. On the left side it extended past the line of the ear. On the right, over the border of the sternomastoid muscle. It extended from the sternum to the chin. The whole tumor was very large and pendulous. There was free discharge of a glairy mucus from several sinuses. In these sinuses a probe could be passed down to the level of the larynx. The tumor mass was of varied con- sistency, at places cystic and moderately fluctuating, at others giving the resistance of fleshy tissue. The sinuses did not com- municate with the interior of the larynx. A laryngoscopic exami- nation showed that the trachea was markedly flattened in its antero-posterior diameter. Both voice and respiration were impaired. The danger of hemorrhage was so great that the patient had previously been advised against operation. Operation: The carotid arteries were found pressed back against the vertebral column. They were closed by means of the clamps, after which the operation was carried out almost blood- lessly. The only blood-loss was in making the incisions in the portion supplied by the inferior thyroid arteries, which were not closed. When the common carotid arteries were ' clamped, the face became blanched and the pulse disappeared from all portions of the head. In experiments on animals we have been able to show that the intracranial pressure is kept sufficiently high for the functioning of the bulbar centers by the vertebral arteries alone. If, however, the closure of the common carotids is perma- nent, cerebral softening is likely to ensue ; if temporary, untoward effects follow. The patient made a good recovery, and there is but a minimum scar on the neck. Both of the external jugular veins were excised with the tumor. All of the deeper structures were laid bare. The titmor was in anatomical relation with the sheaths of the common carotid arteries, the trachea, the larynx, and all the deeper structures of the neck. There was no capsule. At first I was inclined to believe that one of the recurrent laryn- geal nerves was sacrificed. .She spoke in stridulous tones. Later the voice cleared. Case 3 : Clamping of both common carotids ; partial resec- tion of the tongue ; removal of the floor of the mouth ; excision of the submaxillary and sublingual glands ; resection of the paro- tid; excision of the superficial and deep cervical lymphatics on the left side ; excision of the jugular vein ;' resection of the buccal aspect of the inferior maxillary bone. Recovery. 38 Crile — Surgical Operations Diagnosis: Typical epithelioma situated on the floor of the mouth, extending- from the left side of the tongue to the inferior maxilla. Slight enlargement of several lymphatic glands could be palpated. Age 48 ; previous health good ; obese and plethoric : weight 336 pounds ; neck short, thick and fat. Operation: Chloroform-morphin anesthesia. One one- hundredth grain of atropin was given half an hour before the operation to prevent cardiac inhibition from probable mechanical irritation of the superior laryngeal or of the vagus. Both common carotids were closed by means of the rubber-tipped screw clamps. The incision on the right side being an inch long, fibres of the sternomastoid were separated. On the left side the vessel was secured in a like manner. The incision was carried upward, then outward, parallel with the jaw to the parotid gland, and an inner incision was carried to the median line. Reflecting the skin exposed the entire cervical field. The superficial chain of glands was first removed, then the deeper. The submaxillary gland was encroached upon by a metastasis of the adjacent lymphatics, and was accordingly removed. Metastases were found in the deep cervical, in the parotid region, and along the jugular vein. The jugular together with the glands was excised. While dissecting out the deeper glands the pulse increased rapidly to 162, due to increased stimulation of the sympathetic, while the vagal action was prevented by the atropin. The cause being recognized, no stimulation was given. The pulse soon returned to the previous rate. The extensive cervical dissection was then packed with gauze. The tongue was held well over and the mouth lightly packed with gauze. The base of the tongue, the entire floor of the mouth on the left side, and about half of the adjacent jaw was removed. The resection of the tongue included about one- third of its left half and base. There was free communication between the mouth and neck. The clamps were now gradually unscrewed and the circula- tion of the mouth, face and neck re-established. The absolute control by means of the screw clamps made it possible to secure all the bleeding points without appreciable blood-loss. There was but a trifling hemorrhage, mostly venous, and the operation was greatly facilitated by keeping a bloodless field. The patient made a rapid recovery. There were no unfavorable symptoms due to the closure of the carotids either during the operation or after it. Case 4 (Abstract) : Temporary closure of common caro- tid. Removal of sarcoma of parotid; vagus exposed; external carotid and jugular tied ; cardiac inhibition from vagal and laryn- Crile — Surgical Operations 39 ■geal irritation; application of 2% solution of cocain prevented further inhibition. Recovery. The common carotid was closed by means of special clamp ; the jugular vein and the external carotid was excised; the vagus was laid bare. On account of an insufiicient dose of atropin, irri- tation of the vagus while separating it reduced the heart-beats from 93 to 56. The nerve being exposed, cotton saturated with a 2% solution of cocain was packed around it. Although it became necessary to inflict greater mechanical irritation in the further dis- section than had been previously inflicted, the heart promptly returned to 90 and was not further affected. Cocain blocked the afferent impulses and protected the heart. Quick recovery followed. Case 5 (Abstract) : Excision of the tongue, left floor of the mouth, middle half of the jaw, glands of the neck, submaxilla, and a portion of the parotid glands en bloc for carcinoma. Recovery. Male, aged 58 ; preliminary tracheotomy one week before the operation. Both common carotids closed by means of the clamp. One one-hundredth grain of atropin given. Incision carried along the large vessels in the neck, exposing them, to the angle of the jaw, then upward to the base of the jaw ; another incision parallel to the jaw. Skin flaps directed in all directions. Large vessels exposed at the base of the neck. All the glands and fascia removed up to the jaw. Jaw sawed through at two points, after which the entire tongue, floor of the mouth, and left tonsil were removed with the scissors. The pharynx had been previously packed with gauze. There was but a trifling hemorrhage. The patient's pulse-rate never changed during the operation which was completed in 30 minutes. The closing of the common carotid arteries afforded a bloodless field so far as capillary and arterial hemorrhage was concerned, and but slight from the venous sources. The atropin paralyzing the terminals of the vagus pre- vented any inhibition. Patient made a rapid recovery. Case 6 (Abstract) : Carcinoma arising from the duct of the parotid gland. Excision; closure of the common carotid. Recovery. Female, aged 12 ; in good physical condition. During the operation the large vessels and the upper, portion of the vagus and the superior laryngeal were exposed. Although the patient had been given one one-hundred and twenty-fifth grain of atropin the pulse was reduced during the manipulation of these structures 40 Crile — Surgical Operations from 90 to 52, A piece of cotton saturated with a 3% solution of cocain was packed down upon the nerve, after which the pulse returned to its previous rate in less than a minute. The operation was completed, involving continuous manipulation of the vagus and the superior laryngeal, with no effect upon the heart's action. The respirations were shortened and deepened. Case 7 (Abstract) : Operation for tubercular glands of the neck. Dissection involved the jugular, deeper vessels, and the vagus nerve; cardio-inhibitory action from manipulating the vagus. Recovery. Patient eight years old. While separating the glands from the vagus in its upper portion near the parotid the heart-beats dropped from '92 to 62. The field was kept bloodless, so that all the structures could be seen. The vagus had been laid bare. A 2% solution of cocain was applied upon the nerve, and the pulse went up immediately to 90, after which, although the dissection involved the nerve more than before, no inhibition was noted. Case 8 (Abstract) : Closure of the common carotids. Phys- iologic dosage of atropin ; laryngeal application of cocain ; excision of the tongue, epiglottis, left tonsil, floor of the mouth, lower jaw, submaxillary and parotid glands, left jugular vein, left external carotid artery, and the vagus nerve ; but little shock. Easy immediate recovery. Death from secondary hemorrhage on the 13th day. Patient had had six operations for epithelioma ; the disease originated in the floor of the mouth. The operations and the disease had so far destroyed the jaw and the soft parts that the patient was with increasing difficulty able to eat. The cicatricial and carcinomatous contractions and hardening fixed the lower jaw and was progressingly closing the mouth. The procedure was recommended after consultation with distinguished surgeons. The preHminary preparations having been made, an incision was car- ried from each angle of the mouth outward and downward in relation to the growth. The jaw was disarticulated on the left side, while on the right it was severed in the upper portion of the ramus. The extent of the external part of the operation was indicated by a free skin incision. The jaw was severed first on the right side, then on the left. The tongue, tonsil, and the floor of the mouth were then severed latteraly and posteriorly, after which, by continuing the dissection along the vertical plane of the esophagus, larynx and trachea, all the structures, including the parotid, submaxillary and sublingual, and regional lymphatic glands, were removed. The dissection then passed through the Crile — Surgical Operations 41 plane of the carotid artery, jugular vein, and vagus on the left side, all of which were included in the parts removed. Especial care was taken in securing- the veins. The epiglottis showing a tendency to close, the larynx in a light valve-like manner was excised. The removal en bloc by carrying the dissection along the planes indicated not only facilitated dissection, but insured the removal of all the local carcinomatous tissue. The patient bore the operation well, exhibiting a pulse rate of 9G at the close of the operation and good respiratory rhythm. On opening the arteries the circulation of the head was quickly restored and consciousness almost immediately regained. The patient progressed favorably until the 13th day, when death from secondary hemorrhage occurred. The absolute control of the blood supply by means of the special device, thereby maintained a clear field for dissection ; the prevention of cardiac collapse through either direct or reflex inhibition by the administration of a physiologic dosage of atropin ; the prevention of reflex inhibition of respiration by the applica- tion of cocain upon the laryngeal mucosa, places even so extensive an operation on a safe basis so far as the immediate operative effects are concerned. Ca.se 9 (Abstract): Infant, seven months old; had an angio sarcoma of the cheek and neck of rapid growth and great vascularity. Common carotids closed by means of special mechan- ism. Removal of the entire growth. Recovery. The extensive development of the tumor and frailty of so young a subject almost precluded any operative procedure, but on account of the assurance of a safe and absolute control of the blood supply an excision was attempted. Both the common caro- tids were closed. The blood supply was absolutely controlled, and the dissection could be made in a clear field without loss of blood. The tumor had invaded the structures of the cheek, extending down to the mucous membrane, and in the neck extending well down below the angle of the jaw, involving the parotid region. After removing the entire growth it was impossible to bring the parts closely together. Repair was rapid and the child made an uneventful recovery, the operation having been well borne. Case 10 (Abstract) : Carcinoma of the septum of the nose. Excision. Closing both common carotid arteries. Operation wa's performed bloodlessly, involving the tempor- ary lateral resection of the entire nose, extending to the posterior nares, the base of the skull. Hemorrhage entirely controlled. Nose was replaced. Good recovery. 42 Crile — Surgical Operations Case 11 (Abstract) : Excision of one half the tongue, the floor of the mouth, submaxillary glands, entire chain of lymphatic glands, extending along the jugular and angle of the jaw. Excis- ion of the jugular vein. Recovery. Preliminary tracheotomy. Both carotid arteries closed ; fairly bloodless field excepting a small amount of venous hemorrhage. There was not even an appreciable alteration in the pulse and res- piration. Patient made a good recovery. CONCLUDING REMARKS The proper interpretation of a slowed or of an accelerated pulse, or of an inhibited respiration, the prevention of either direct or reflex inhibition of the heart from mechanical stimulation of the vagus or of its branches by the use of atropin and cocain, the safe and absolute control of hemorrhage by temporarily closing the carotid arteries render operative procedures of the head and neck so much safer as to greatly increase surgical possibilities. OBSEKVATIONS ON THE OEIGIN AND OCGUREENCE OF CELLS WITH E0SI:N0PHILE GRANULATIONS IN NORMAL AND PATHOLOGICAL TISSUES. By W. T. HOWARD, JR., M. O., AND R. G. PERKINS, M. D. (From the Pathological Laboratory of Lakeside Hospital, Cleveland, Ohio.) (Plate VIII.) In the routine examination of 825 specimens derived from operations, and of the organs from 120 autopsies from the various services of Lakeside Hospital, we have been struck with the fre- quency with which cells with eosiaophilic granulations have been found. It is our purpose in this article to classify and analyse these observations. By the terms " eosiuophile and eosinophilic cell," and " ceU with eosinophilic granulations " as used iu this article, we mean a cell with coarse or moderately coarse gTanulations which stain deeply and intensely with acid dyes, especially with eosin. These cells correspond to the eosinophilic cells of human blood of Ehr- lich, and to the coarsely granular oxyphile cell of Kanthack and Hardy. All cells with homogeneous non-granular protoplasm, as well as the polymorphonuclear neutrophilic cell of Ehrlich or the finely granular oxyphilic cell of Kanthack and Hardy, are ex- cluded from consideration. Methods. — Most of the tissues upon which this study is based were hardened in Zenker's fluid, some in Orth's fluid, and a few in 95 per cent alcohol. The best results were obtained with Zenker's fluid. Celloidiri was commonly used for embedding, though the parafiin method was used in some cases. As a matter of routine sections were stained in haematoxylin followed by a rather strong aqueous solution of eosin which was allowed to act for a short time, Overstaining with eosin was studiously avoided. In most cases there was no intention of staining for eosinophilic granulations, but the sections were stained in the usual manner by a number of different workers from the various services of the 250 W. T. Howard, Jr., and R. 0. Perkins. hospital. Sections from a number of cases were stained with eosin followed by methylene-blue, which gave good results. Besides eosin in special cases other acid dyes, such as acid fuchsin and picric acid combined (Van Gieson's stain) the Biondi-Heidenhain and the Ehrlich triacid stains, were tried and always with positive re- sults. Sections which showed large numbers of eosinophiles were treated with freshly-prepared ammonium sulphide for from six to twenty-four hours but in no case were any of the granules changed in color. Sections treated with fresh potassium ferrocyanide also failed to show the presence of iron. For convenience we shall classify our observations as follows: Observations on the presence of cells with eosinophilic granula- tions in (I) apparently normal organs (II), in pathological tissues in material derived from operations, (III) in organs showing patho- logical changes in cases coming to autopsy. I. Cells with Eosinophilic Geanulations in Apparently NOEMAL OeGANS. In ten normal appendices these cells were numerous; in five cases they were confined to the stroma of the mucosa, in two they were in both the stroma and in the lymphoid tissue, while in three they were numerous in the stroma with a few in the capillaries. A few eosinophiles were found in the blood-vessels of a normal Fallopian tube.' They were found in apparently normal organs in the following cases coming to autopsy: Case I. A few polymorphonuclear eosinophiles were present in the kidney of a five days old infant with omphalitis and Staphy- lococcus aureus bacteriaemia. Case II. A child one month old, dead of streptococcus infec- tion. A few polymorphonucler eosinophiles were present in the mucosa of the ileum. ' Since the above was written we have examined a large number of both normal and inflamed appendices, in nearly every one of which there were large numbers of eosinophiles. In the normal appendix they are confined to the interglandular stroma and the lymphoid tissue, while in inflammation they invade all coats. They are much more readily found after hardening in Orth's or Zenker's fluids than in alcohol. Eosinophile Granulations. 251 Case III. In a still-bom child, seven montlis in utero, the liver capillaries contained large numbers of polymorphonuclear eosiao- philes. Case IV. In a child with imperforate anus, dead three days after birth, the muscularis of the rectum at the point of atresia showed numerous round cells, and a few polymorphonuclear eosin- ophiles. Case V. In a case of epidemic cerebro-spinal meningitis, due to Diplococcus intracellularis meningitidis, a few polymorphonu- clear eosinophilic cells were seen in the stroma of the mucosa of the small intestine. Case VI. In a case of a child two days old, in whom death followed gastro'intestinal haemorrhage, due probably to haemo- philia, polymorphonuclear cells with eosinophilic granulations were found in numbers in the connective tissue of the portal spaces in the liver. Case VII. In a child two weeks old, dead of follicular dysen- tery and omphalitis, the latter due to Staphylococcus pyogenes aureus, there was marked congestion of all the organs, including the thymus. The lymphoid tissue of the latter organ contained large numbers of coarsely granular eosinophiles. A few of these cells had small deeply staining reniform or polymorphous nuclei. Most of them, however, had a single round, oval, or reniform slightly vesicular nucleus resembling in all respects the nucleus of the plasma cell. The nuclei of many of these cells were eccen- tric. In some plasma cells nuclear figures were seen. There was no special relation of the eosinophiles to Hassal's bodies. The eosinophiles were most numerous at the peripheral portions of the glands. In four of the seven eases there was infection, septicaemia oc- curring in two. In these four cases the infectious agent may have stimulated the formation of the eosinophiles. II. Cells with Eosestophilic Graktjlations in Pathological Tis- sues nsr Material deeived from Operation. In twelve cases of acute appendicitis, these cells were numerous in the tissues of five, in the tissues and vessels of one, and very numerous in the tissues and vessels of two, and very numerous in the tissues in four. 252 W. T. Howard, Jr., and B. G. Perkins. In six cases of chronic appendicitis, a few were found m the tissues and exudate in two, and in the other four they were numer- ous in the tissues. In six cases of salpingitis, one of which was tubercular, a few were found in the stroma of the mucosa and in the muscularis. In four cases a few occurred in both vessels and tissues, in six cases there were a few in the blood-vessels, and numbers in the tissues; in eight cases they were numerous in the tissues and absent from the blood-vessels; in eight cases they were numerous in both ves- sels and tissues; in one case they were numerous in the vessels and very numerous in the tissues; in two cases they were very numerous in the tissues and absent from the blood-vessels; in one case they were very numerous in both vessels and tissues, and in one case they were present in small numbers in the vessels and ia great numbers in the tissues, making thirty-seven cases of salpin- gitis in all. In one case of acute ovarian abscess they were present in small numbers in the tissues, while of eight cases of chronic ovarian ab- scess they were present in the tissues in small numbers in one case, in large numbers in two cases, and in great numbers in four cases, but were absent from the blood-vessels, while in two cases they were present in great numbers in both tissues and blood-vessels. In two cases of acute perioophoritis they were present in both vessels and tissues in large numbers. In four cases of chronic perioophoritis they were present in the tissues in small numbers in one case, in large numbers in another case, in very large numbers in a third case, while in the last case they occurred in small num- bers in the blood-vessels and in large numbers in the tissues. In two corpus luteum cysts they were numerous in the tissues in one case, present in small numbers in the vessels and very nu- merous in the tissues of the other. In one case of congestion of the Fallopian tube they were pres- ent in numbers in the blood-vessels. In a case of inflammatory thickening of the broad ligament they occurred in numbers in both blood-vessels and tissues. In one case of hypertrophic endometritis a few were found in the tissues, and in a case of acute interstitial endometritis after abortion, a few were present in the blood-vessels and large num- bers in the tissues. Eosinophik Granulations. 253 In three cases of carcinoma uteri they occurred in one case in numbers in both vessels and tissues, in the other two cases in great numbers in the tissues while none were found in the blood-vessels. In two cases of epithelioma of the vagina numbers were found in the tissues in one case, a few in the vessels and great numbers in the tissues in the other case. In one Staphylococcus aureus abscess of the gland of Bartholin numbers were found in the tissues but none in the blood-vessels. Two cases of proctitis showed very large numbers in the tissues. One case of pilonidal abscess showed numbers in the tissues, but none in the vessels. In three cases of omentitis one showed numbers in the tissues, but none in the vessels, the second, numbers in both blood-vessels and the tissues and in the third none in either the vessels or the tissues. In one case each of carcinoma of the arm, secondary epi- thelioma of a lypmh gland, epithelioma of the scrotum, and in two cases of epithelioma of the face they were found in the tissues in numbers. They were present in the tissues in varying numbers in cases of tuberculosis involving the following organs: striated muscle, anus, rectum, ureter, cervical lymph glands, and skin. In three nasal polyps numbers of eosinophiles were present in the tis- sues in two and in the tissues and blood-vessels in one. They were numerous in the lesions in six cases of osteomyelitis, three of which were tuberculous. They were found in considerable numbers in the tissues in one case of gonorrhoeal lymphangitis and in a gumma of the skin. For comparison with the foregoing some fifty cases of inflam- matory and other lesions of various tissues were noted in which eosinophiles were absent. Among these there were twenty cases of acute and chronic salpingitis, three cases of abortion remnants, three of endometritis, and several corpus luteum cysts. Altogether coarsely granular eosinophiles were present in either the tissues, exudates or blood-vessels in 108 out of 825 consecutive specimens derived from operation, or in 13.09 per cent of the cases; 80 of these cases were inflammatory, 24 being acute and 56 chronic. Appendicitis was the lesion in one-half of the acute cases. There were 9 cases of tuberculosis and 9 of syphilis; 9 of the 16 non-infectious cases were carcinomata and 3 were nasal 254 W. T. Howard, Jr., and B. Q. Perkins. polyps. We haYe never found eosinophiles in otlier tumors ex- cept in one case of lymphosarcoma'' (Case XX). Of the 108 cases 74 per cent were inflammatory and counting the cases of tuberculosis and syphilis 83.3 per cent were infectious processes. These observations show that these cells are frequently present and probably take an active part in inflammatory lesions. Eosino- philes were found in greatest numbers in appendicitis, pyosalpinx, chronic ovarian abscess, and carcinoma. They were almost always associated with plasma cells, which were often very numerous. Polymorphonuclear neutrophiles were not numerous as a rule, but in some cases, in salpingitis especially, they were abundant. There is a close relation between the occurrence of eosinophiles and plasma cells. In many cases (as will be hereinafter described), especially in appendicitis, the transition of plasma cells into eosino- philes was readily traced. In appendicitis the eosinophiles nor- mally present in the stroma of the mucosa and in the lymphoid tis- sue are markedly increased in number and wander into the in- flamed areas. In chronic appendicitis they are commonly seen in great numbers in the muscularis. In carcinoma they are some- times seen in the blood-vessels, but even here they are closely asso- ciated with plasma cells. In a large number of cases of salpin- gitis, on the other hand, numbers of eosinophiles were present in the blood-vessels as well as in the tissues. In salpingitis eosino- philes were found in greatest numbers in the mucosa and in the muscularis, and but rarely in the serosa or in the lumen of the tube. There is evidence, then, that in salpingitis some of the eosinophiles in the exudation are derived from the blood-vessels. "We are informed, however, by Dr. Weir, the Resident Gynaecol- ogist, that in a number of the cases of salpingitis operated on in Dr. Eobb's service showing a decided leucocytosis (of the polymor- phonuclear neutrophiles) before operation, a differential count of the leucocytes always failed to show eosinophilia, even in the cases in which eosinophiles were found in large numbers in the pelvic lesions. The consideration of the bacteriology of these cases vdll be postponed to the end of the article. ''In a recently examined case of giant cell sarcoma of the mediastinal glands invading the pectoral muscles there were great numbers of eosinophiles in the tumor and in the neighboring muscle. Eosinopliile Granulations. 255 Description of the Eosinophilic Cells met with in the above series. In the apparently normal tissues tlie stroma of the mucosa of the appendix, ileum, and stomach, for instance, by far the majority of the eosinophilic cells corresponded in all respects to those com- monly found in the circulating blood. They had horseshoe-shaped, or trilobate nuclei, surrounded usually with coarse granules which stained deeply with eosin. In some of these cells, however, the granules were finer than in others. The cell outlines were often irregular. Besides these cells a few were seen with single, usually round, deeply-staining nuclei, surrounded, some by finely granu- lar, others by coarsely granular protoplasm. Both varieties of gran- ules stained deeply with eosin. These cells were about the same size as the large mononuclear cells with hyaline protoplasm. The nuclei of the two varieties of cells were identical in appearance, with the exception that the nuclei of the eosinophilic cells com- monly stained more deeply than those of the large hyaline cells of the blood. In the second series of cases, in which eosinophilic cells were found in inflammatory and other lesions, these cells were in the main similar to those found in the apparently normal organs, with the exception that mononuclear cells were present in larger proportion in the former than in the latter. In some of the inflammatory cases, notably in ovarian abscesses and in pyosalpinx the eosinophilic cells were very numerous in the tissues. As a rule, few polymorphonuclear neutrophilic cells were present with the eosinophiles, except in acute cases, where the former were usually very numerous, and the latter few in number. Elsewhere these two varieties of cells were not commonly seen to- gether. In the chronic cases as a rule, and often in the acute cases, especially in ovarian abscess and pyosalpinx, the eosinophiles were associated with plasma cells, which were present in large num- bers. In only a few cases were the eosinophilic cells the most numerous in the exudation. III. Cells with Eosinophilic Geai^ulations in Oeoans showing Pathological Changes in Cases coming to Autopsy. In the ten following cases cells with eosinophilic granulations were found in the spleen, in association with chronic interstitial splenitis, and in various other organs. 256 W. T. Howard, Jr., and B. G. Perkins. Case VIII. Male, aged 60 years. Interstitial splenitis with polymorphonuclear eosinophiles in fairly large numbers in the spleen pulp; carcinoma of the pancreas and liver. Case IX. Male, aged 45 years. Congestion of the spleen with haematoidin in the trabeculae, eosinophiles in clumps of three and four in the splenic pulp. Death from intestinal obstruction. Case X. Female, aged 3 years. Pleuropneumonia with serous effusion, interstitial splenitis, persistent thymus, healed eczema, eosinophilic cells in the splenic pulp and in the thymus. The thymus measured 5x3x2 cm. and was of ordinary appear- ance on section. The spleen weighed 25 gm. and measured 7x5 X 2 cm. The capsule was adherent to the diaphragm over an area 1x1 cm. On section the consistency was increased, the Malpi- ghian bodies and trabeculae were both well marked. The outlines of sections of the spleen were serrated, the depressions correspond- ing to the very thick trabeculae. Beneath the capsule there were broad and narrow bands of connective-tissue thickening. In the pulp here and there, polymorphonuclear eosinophilic cells were found in considerable numbers. In many of the smaller arteries there was an obliterative endarteritis, while the media of many was hyaline. There was no round cell infiltration. In sections of the thymus the lobule as well as Hassal's bodies were well marked. Tha interlobular stroma was normal. Large numbers of polymorpho- and mononuclear cells with eosinophilic granulations were seen in the lobules, in the interlobular connec- tive-tissue and in the blood-vessels. Case XI. Male, aged 5 years. Bronchopneumonia with gen- eral infection with Bacillus mucosus capsulatus; eosinophiles in the spleen and thymus. Sections of the spleen showed marked congestion with numer- ous eosinophiles in the pulp. The thymus showed marked con- gestion with many eosinophiles (mostly mononuclear) in the lymphoid tissue; in many places they were very numerous about the bodies of Hassal. Case XII. A still-bom child at full term, with hydrocephalus and meningocele ; eosinophiles in the spleen and thymus. Sections of spleen showed marked congestion of the vessels of the pulp in which a few polymorphonuclear eosinophiles were seen. EosinopMle Granulations. 257 The thymus was much congested and showed a moderate number of eosinophiles in the lymphoid tissue. Case XIII. 'Male, aged 4 months. Bronchopneumonia; mal- nutrition; eosinophiles in the liver, spleen, ileum, colon, thymus and lymph glands. Sections of the liver showed a number of plasma cells and a few polymorphonuclear eosinophiles in the vessels; the spleen showed large numbers of the same cells in the pulp. The small and large intestines, the thymus and the mesenteric lymph glands showed large numbers of eosinophiles, both mononuclear and polymorpho- nuclear, in the lymphoid as well as in the interstitial tissues. Case XIV. Male, aged 5 years. Burns of the third degree; chronic interstitial splenitis; eosinophiles in the liver, spleen, adrenals, thymus, mesenteric lymph glands, and gastro-intestinal tract; persistent thymus. Sections of the spleen showed congestion and diffuse increase of fibrous tissue, with a large number of mono- and polymorphonu- clear eosinophiles scattered throughout the pulp, which contained large numbers of typical plasma cells. Every stage of transition between plasma cells and eosinophiles could be made out. The liver showed marked congestion of the capillaries with cloudy swell- ing of the liver cells; polymorphonuclear eosinophiles were found in the capillaries of the liver and adrenals, the lymphoid tissue of the thymus showed large numbers of the same cells; the mesen- teric glands showed congestion and hyperplasia with large numbers of eosinophiles in the lymphoid tissue. Case XV. Female, aged 40 years. Streptococcus bacteriae- mia following pelvic abscess; interstitial splenitis with eosinophiles in the spleen, liver and abscess wall. Sections of the spleen showed marked congestion and increase of fibrous tissue with a moderate number of mono- and polymorphonuclear cells in the pulp; a few of the same cells were present in the blood-vessels of the liver. In sections of the abscess wall there were large numbers of strep- tococci, numerous mononuclear and polymorphonuclear eosino- philes, plasma cells and neutrophiles. Case XVI. Male, aged 45 years. Empyaema and lobar pneu- monia; interstitial splenitis with eosinophiles in the lung and spleen. In sections of the pneumonic area the alveoli were filled with masses of fibrillated fibrin containing polymorphonuclear neutro- 258 W. T. Howard, Jr., and R. G. Perkins. philes. In some alveoli there were large globular polypoid masses of fibrin attached to the alveolar v^alls by narrow bases and con- taining fibroblasts and capillaries. In many of the alveoli and in some of the alveolar walls numerous polymorphonuclear eosino- philes were seen. The spleen showed diffuse connective-tissue in- crease with a few polymorphonuclear eosinophiles and many plas- ma cells in the pulp. Case XVII. Male, aged 35 years. Miliary tuberculosis of the lungs with some mononuclear eosinophiles in the granulation tissues about the tubercles; diffuse interstitial splenitis with poly- morphonuclear eosinophiles in the pulp; general chronic passive congestion; malaria (tertian form). Coverslips made from the spleen showed dark pigment in the leucocytes and a number of red blood-cells containing malarial parasites of the tertian type. Sections of the left lung showed chronic interstitial pneumonia, with fibrous tubercles. The tuberculous tissue was rich in plasma cells, polymorphonuclear neutrophiles and mononuclear and poly- morphonuclear eosinophiles. The spleen showed marked conges- tion, a large amount of haematoidin in the trabeculae, and inter- stitial splenitis with a large number of polymorphonuclear eosino- philes in the pulp. Five of the foregoing cases were young children (Cases X, XI, XII, XIII, XIV) and all showed large numbers of eosinophiles in the thymus as well as in the spleen. Case XIV is of special in- terest as the formation of eosinophiles was in active progress, and these cells were present in increased numbers in the various or- gans, as well as in the tissues of the spleen, thymus (persistent), lymph glands and intestines. Besides their occurrence in the spleen in the five remaining cases, they were also found in one case in the wall of a pelvic abscess (Case XV), in one case in the exudate of lobar pneumonia (Case XVI), and in one case in tuber- culous areas in the lungs (Case XVII). In two cases they were present in all the organs except the heart and lungs. Although circulatory eosinophilia has been described in malaria, as far as we have been able to learn. Case XVII is the first one in which these cells have been observed in the spleen in this disease. In the next four cases cells with eosinophilic granulations were present in large numbers in the gastro-intestinal tract. Eosinophile Granulations. 259 Case XVIII. Male, aged 51 years. Atrophic cirrhosis of the liver and gastric mucosa, with eosinophiles in the stroma of the latter. Sections of the stomach showed atrophy of the glands, and infiltration of the stroma with numerous mono- and polymorpho- nuclear eosinophilic cells. Case XIX. Male, aged 34 years. Arsenic poisoning; poly- morphonuclear eosinophiles in the intertubular tissue of the gas- tric mucosa. Sections of the stomach showed some loss of the surface epithelium and marked dilatation of the blood-vessels of the mucosa. There was no special destruction of the tubules. The intertubular tissue contained numbers of plasma cells and poly- morphonuclear eosinophiles. The epithelial cells of the tubules were often separated from the basement membrane, and there was marked congestion of the submucosa. Case XX. Male, aged 64 years. General lymphosarcoma with numerous mono- and polymorphonuclear eosinophiles in the stroma and muscularis of the small intestine. Sections of the spleen, liver, kidneys, intestines and mesenteric glands showed lymphosarcoma. Sections of the duodenum and jejunum showed loss of the superficial epithelium of the mucous membrane of some of the glands, with numbers of large bacilli {B. aerogenes capsu- latus) in the degenerated areas. In the muscularis and to a less degree in the stroma of the mucosa, a large number of cells with eosinophilic granulations were seen. These cells were of several types, and showed distinct transition forms. In the stroma of the mucosa the cells were aU of one type. They had rather large, pale nuclei, and were similar in every way to plasma cells, except for the eosinophilic granulations. There were a moderate number of plasma cells in association with the eosinophiles. In the muscu- laris the picture was materially different. Here the prevailing type of cell had a small, very dense eccentric nucleus, round, oval or elongated, and an irregular cell body, which conformed itself to the clefts between the muscle fibres, and consequently showed a great variety of shapes; the granulations were extremely coarse, and crowded closely together. Besides these there were a number of cells similar to the above except that the nuclei were somewhat vesicular, and contained small clumps of chromatin. These forms were the most common ones, but there were also a fairly large number of ordinary polymorphonucler eosinophiles with double, 260 W. T. Howard, Jr., and E. G. Perkins. horse-shoe, or trilobate nuclei. Besides these there was a great variety of forms which were apparently transitional between the smooth muscle cell of the intestine, and the typical, fully-formed polymorphonuclear eosinophils. There were <3ells with double or even trilobate nuclei, but without a trace of nuclear detail, and others whose nuclei were markedly vesicular. The eosinophiles apparently had a double origin, arising on the one hand from the plasma cells which were present in both the stroma of the mucosa and in the muscularis, and on the other, from the muscle cells. Eosinophiles were present also in the capillaries and in the small veins, but were infrequent in the larger vessels. Only two types occurred, the ordinary polymorphonuclear form, and the mononu- clear form with a vesicular nucleus. The forms with dense nuclei were entirely absent from the vessels. No polymorphonuclear neutrophiles were seen in the tissues. Case XXI. Male, aged 34 years. Extensive ulceration of the rectum, sigmoid flexure and descending colon, with polymorpho- nuclear eosinophilic leucocytes in the large intestine and in the spleen; taenia saginata in small intestine. There was an extensive suppurating wound on the left thigh, communicating with the sig- moid. The rectum, sigmoid and colon up to the splenic flexure were the seat of larger and smaller ulcers with granulating bases and margins. Over large areas of the intestine the mucosa was lost, the surface being covered with granulation tissue. The spleen weighed 180 gms., and measured 12x6x4 cm. The capsule was smooth; on section the organ was dark red in color, the consistency was normal. The trabeculae were prominent, the Malpighian bodies obscure. Microscopically the trabeculae were thickened, and there was a diffuse increase of connective- tissue in the pulp, with numbers of polymorphonuclear cells with eosinophile granulations. The capillaries contained a large num- ber of polymorphonuclear neutrophilic leucocytes. In the granu- lation tissue of the rectum and sigmoid, in addition to fibroblasts, plasma cells and polymorphonuclear neutrophilic leucocytes, there were numbers of cells with eosinophilic granulations. In these four cases there were numbers of eosinophiles in gas- tro-intestinal lesions, and in only one case (Case XIV) were those cells found in other organs. In the case of chronic dysentery they were numerous in the very much thickened splenic pulp. In con- Eosinophile Granulations. 261 nection with Case XXI it is interesting to note that in a case of chronic salpingitis and ovarian abscess, occurring in a woman with tapeworm, eosinophiles were present in great numbers in the in- flammatory lesions and in the blood-vessels. Unfortunately a dif- ferential count of the leucocytes of the blood was not made. Case XXII. Female, aged 65 years. Carcinoma of bladder and uterus with numerous eosinophiles in the stroma of the tumor; interstitial splenitis with polymorphonuclear eosinophiles in the pulp. A few polymorphonuclear eosinophiles were scattered through the splenic pulp. Sections made from various portions of the uterine tumor showed typical adeno-carcinoma with large alveoli. In the scanty stroma and in the surrounding tissues there were large numbers of mononuclear and polymorphonuclear cells vrith eosinophilic granulations. These cells were especially numerous in the media of small arteries, where they lay between the muscle cells and were often much elongated. The following case is of special interest as being the first case of chronic interstitial nephritis in the round cell infiltration of which we have been able to find cells with eosinophilic granula- tions; it showed transition forms between the plasma cell and the eosinophile cell. Case XXIII. Female, aged 44 years. Chronic interstitial nephritis and splenitis with eosinophilic mono- and polymorphonu- clear cells in the kidneys. Sections of the kidneys showed wide- spread interstitial nephritis with marked atrophy of the renal tis- sue, both tubular and glomerular, with fibroid interstitial growth. Among the cells in the areas of round cell-infiltration there were some large mononuclear cells with non-granular cytoplasm which stained deeply and diffusely with eosin. Besides these there were many cells with granular cytoplasm which stained deeply with eosin. The liver and spleen showed nothing of special interest. The next case is an instance of acute suppurative nephritis en- grafted on chronic arteriosclerotic nephritis. The exudation in this case was marked by the presence of large numbers of poly- morphonuclear eosinophiles in association with plasma cells and polymorphonuclear neutrophiles. Case XXIV. Male, aged 62 years. Haemorrhagic cystitis, abscess of prostate, pyelo-nephritis ; chronic interstitial nephritis; 262 W. T. Howard, Jr., and B. G. PerUns. polymorphonuclear neutrophilic and eosinophilic leucocytes and plasma cells in the exudation. Sections of the kidneys showed much the same changes. There was well-marked arteriosclerotic nephritis with atrophied hyaline and fibrous glomeruli, tubular atrophy and fibrous vessel increase. Scattered through the sections there were larger and smaller areas of abscess formation with diffuse infiltration with great areas of plasma cells and polymorphonuclear neutrophilic and eosinophilic leucocytes. The most numerous cells were the neutrophiles. The capillaries were dilated, the tubules were disorganized, and the epithelial cells were in various stages of degeneration. In the deep cortical and medullary portions there were larger and smaller areas of diffuse infiltration with plasma cells, polymorphonuclear neutrophiles and eosinophiles. These cells were seen both within and without the tubules, some tubules being enormously distended with them. Cocci, short, thin bacilli, and long and stout bacilli were found. The long thick bacilli were present in great num- bers. ISTo eosinophilic cells were found in sections of the bladder and prostate. In the next three cases" eosinophilic cells were found in acute interstitial nephritis. Case XXV. Male, aged 57 years. Acute and chronic inter- stitial nephritis with the presence of mono- and polymorphonuclear eosinophiles and plasma cells in the interstitial tissue, epithelioma of the bladder. Cultures were negative. Urotropin had been administered sev- eral days before death. Sections of the kidneys showed thicken- ing of the capsules, with thinning of the cortices, which were the seat of widespread interstitial nephritis. Most of the glomeruli had disappeared, and of those which were left some were mark- edly thickened, while others were hyaline. The tubules were extensively atrophied, and there was a diffuse growth of fibrous tissue, rich in cells, most of which were polymorphonuclear neu- trophiles, though a number of eosinophiles were seen among them, larger and smaller collections of eosinophiles were present in the tissue. The blood-vessels were dilated, and many of the capil- 3 These three cases were reported In full in the American Journal of the Medi- cal Sciences, December, 1900. Eosino-phile Granulations. 263 laries and small veins contained numbers of leucocytes. In many places the process was acute ratker than chronic, the tissues being filled with leucocytes. A few small scattered abscesses were seen in the tissue. The medullary portion showed marked congestion, the tissue was compressed and atrophied, and here and there larger and smaller collections of small round cells were seen. In the adventitia of many of the small arteries there was marked cellular infiltration. These cells were for the most part small round cells and plasma cells, but a number of large mononuclear and smaller polymorphonuclear eosinophilic cells were seen. In the adven- titia of one of the small arteries, near some typical plasma cells, and some large mononuclear cells, two very interesting cells were seen. One was the size of a large mononuclear leucocyte, which had a distinct kidney-shaped nucleus. The cell had the appearance of a typical transition cell of the blood. Next to it was a cell slightly smaller in size, which was crowded with rather fine granulations, much smaller than those of a typical eosinophile; these granula- tions stained very deeply with eosin. The nucleus of this cell was slightly oval, and marked by larger and smaller deeply staining chromatic masses. This cell was apparently a transition of a large mononuclear into an eosinophilic leucocyte. Large numbers of eosinophiles were scattered through the kidney and in the areas of round cell infiltration. Case XXVI. Female, aged 30 years. Acute interstitial ne- phritis with mono- and polymorphonuclear eosinophilic leucocytes and plasma cells in the intertubular exudation. Tuberculosis of bronchial glands with eosinophiles in the tuberculous tissue. Pro- lapsus uteri, abortion, vaginal hysterectomy. Small stitch abscess containing Staphylococcus aureus. In the bronchial glands there were a number of areas containing giant and epithelioid cells, while in the lymph spaces and stroma there were numerous mono- and polymorphonuclear eosinophiles. In the kidneys there were numerous areas of cellular infiltra- tion, in many of which the glomerular capillaries were thick and congested and showed an increase of cells, some of which had eosinophilic granulations. The tubules were compressed by an accumulation of cells in the intertubular tissue. These cells were for the most part plasma cells, but a considerable number of eosino- philes were present. The latter were seen in the capillaries as 264 W. T. Howard, Jr., and B. G. Perkins. well as in the tissues. Tlie epithelial cells of the tubules were compressed and many were swollen and granular. Case XXVII. Female, aged 35 years. Acute interstitial ne- phritis with numbers of eosinophiles and plasma cells in the exu- dation. Anatomical Diagnosis. — Streptococcus septicaemia following abortion; stomatitis, pharyngitis, bronchitis, bronchopneumonia, and oedema and congestion of the lungs; acute interstitial nephri- tis, acute splenic tumor, retained placenta. The kidneys were large; the right weighed 370 and the left 330 grammes. The capsules stripped off readily, the surfaces were smooth, pale and stained with bile pigment. On section the organs were hyperaemic and markedly oedematous and opaque. The consistency was in- creased. The cortices averaged 1 cm. in thickness; the glomeruli were inconspicuous; both organs were bile stained. On micro- scopic examination the kidneys showed marked oedema and con- gestion. The glomerular capillaries were dilated and contained a few polymorphonuclear eosinophiles. The capsular epithelium was swollen and granular as was the epithelium of the convoluted tubules. There were larger and smaller areas of cellular infil- tration in the intertubular tissues of the cortex both near the sur- face and near the medulla. In many places the tubules were markedly compressed by this exudation, in which three main va- rieties of cells could be distinguished, viz., lymphocytes, plasma cells and polymorphonuclear cells with eosinophilic granulations. Some tubules were filled with desquamated epithelial cells and leu- cocytes in varying numbers. Throughout the kidneys the blood- vessels, especially the capillaries, contained large numbers of poly- morphonuclear eosinophilic cells, which were most numerous near the areas of cellular infiltration. In these areas of interstitial ne- phritis the most numerous cells were plasma cells ; next in number to these ranked the eosinophilic cells. In some places the eosino- philic cells were the most numerous cells in the exudation. There were five cases of nephritis with eosinophiles in the renal lesions. The first (Case XXIII) was a typical case of chronic in- terstitial nephritis, in the cellular infiltration of which eosino- philes were found. In Case XXIV, the kidneys were the seat of a moderate degree of arteriosclerotic nephritis, and there was a double-sided acute suppurative nephritis, secondary to abscess of Eosinophile Granulations. 265 the prostate, in the exudation of which besides polymorphonuclear neutrophiles there were many plasma cells and eosinophiles. Case XXV was also an example of moderate chronic nephritis with acute suppurative nephritis, secondary to chronic cystitis and epi- thelioma of the bladder. The eosinophiles and plasma cells were very numerous. In Cases XXVI and XXVII, there was typical acute interstitial nephritis with large numbers of eosinophiles and plasma cells in the exudation. In both cases there had been abor- tion and in the last case streptococcus septicaemia. The eosino- philes were so numerous that the process might with propriety be called acute eosinophilous nephritis. Case XXVIII. Male, aged 19 years. Caseous and miliary tuberculosis with oedema of the lymph glands with the presence of large numbers of eosinophiles. Recent tubercular pleurisy and pericarditis. General infection with Staphylococcus aureus. Sections of various lymph glands showed marked oedema with miliary and caseous tuberculosis. Sections of one mesenteric gland showed as follows: Larger and smaller areas of caseation were surrounded by a reticular fibrous tissue in the meshes of which there were many cells, among which there were giant cells, epithelioid cells, plasma cells and lymphocytes. There was enormous oedema which dis- tended the spaces of the reticular tissue as well as the lymph chan- nels. Many of the lymph channels, especially in the centre of the gland, were crowded with cells, most of which were plasma cells, hyaline leucocytes and lymphocytes, with relatively few polymor- phous nuclear cells. In many plasma cells nuclear figures were seen. In the widely distended reticulum, there were relatively few cells, most of which were very large oval cells, some with densely staining and some with vesicular nuclei, surrounded by a large amount of finely granular protoplasm. The nuclei were often eccentric and sometimes reniform. The granules were not eosinophilic, but the cells were markedly phagocytic, and often contained plasma cells, lymphocytes and even eosinophiles. The germinal centres were increased in size, the most numerous cell being the plasma cell. There were also many eosinophiles. !N^early everywhere in the tissue there were variable and often large num- bers of coarsely granular eosinophiles, some with polymorphous nuclei but most with a single, rather deeply-staining nucleus. 266 W. T. Howard, Jr., and R. Q. Perkins. Many cells with nuclei similar to those of plasma cells surrounded with fine eosinophilic granulations were seen. The blood-vessels were not congested and showed no increase of leucocytes. Case XXIX. Male, aged 17 years. General miliary tubercu- losis with mono- and polymorphonuclear eosinophiles in the mes- enteric and retroperitoneal lymph glands, which were large, soft and oedematous. Sections from these glands showed typical miliary and conglom- erate tubercles containing many giant cells and often showing case- ation. In many places there was marked oedema, increase of the cells of the germinal centres and dilatation of the lymph channels, with the presence of large numbers of cells in the stroma and in the channels. These cells were chiefly plasma cells but there were also present a number of mononuclear eosinophiles with ec- centric nuclei, closely resembling plasma cells in all respects ex- cept eosinophilic granulations. A few ordinary polymorphonu- clear eosinophiles were also present. These two cases are of great interest, for with the eight cases of tuberculosis with eosinophilic infiltration met with in our operative material, and the lymph gland tuberculosis of Case XXVI and the lung tuberculosis in Case XVII, they make twelve instances in which we have observed eosinophiles in tubercular lesions. The development of eosinophiles from plasma cells and hyaline leuco- cytes could be readily traced in the tuberculous lymph glands. Case XXX. Male, aged 49 years. Eosinophiles in a wall of a liver abscess secondary to pneumonia and empyaema. General streptococcus infection. Sections of the wall of the hepatic ab- scess showed marked congestion of the blood-vessels with compres- sion of the liver cells. The abscess wall was infiltrated with poly- morphonuclear neutrophiles, plasma cells and coarsely granular eosinophiles. This is the third case in which we have found eosinophiles in the liver. Case XIV showed them in the portal spaces in a child dead of bums ; while in Case XV they were found in the vessels of an- gioma of the liver of a woman with pelvic abscess. Case XXXI. This case has already been published by one of us (Howard).* Tor the clinical history and autopsy we are in- •"Howard, W. T., Jr. Philadelphia Medical Journal, 1899, vol. IV, p. 1085. ^osinophile Granulations. 2(57 debted to the kindness of Dr. Hoover. Female, aged 35 years, was ill for three weeks with diarrhoea and vomiting and com- plained of pain in the abdomen and the muscles of the extremities. Coverslip preparations of the blood stained with Ehrlich's triacid stain, on examination by Dr. Hoover, showed a slight leucocy- tosis, without however an increase in the number of eosinophiles. Although Dr. Hoover did not make a differential count, he is confi- dent that there was no circulatory eosinophilia. A piece of the extensor quadratus of this case was excised and showed large num- bers of Trichina spiralis. Death occurred at the end of three weeks from the onset. Anatomical Diagnosis. — General infection of the voluntary muscles with Trichina spiralis; acute yellow atrophy of the liver; infarction of the spleen; fatty degeneration of the heart and kid- neys; congestion of the lungs and intestines. Cultures and coverslip preparations made from the various or- gans showed no bacteria. Teased preparations of the tongue, the diaphragm, quadratus lumborum, psoas, and quadriceps femoris muscles showed large numbers of actively motile trichinae. No trichinae were found in the intestinal contents. Sections of the heart showed congestion of the capillaries and small veins — in some of the latter there were small accumulations of large and small mononuclear leucocytes; no cells with eosino- philic granulations were found. Larger and smaller areas of cellu- lar infiltration were seen between some of the muscle fibres, the cells consisting of lymphocytes, plasma cells, a few polymorphonu- clear neutrophilic leucocytes, and a number of both polymorpho- and mononuclear cells with marked eosinophilic granulations. In the lungs there was found marked oedema with congestion. In the peribronchial tissue and in the lymph spaces, but never in the small blood-vessels a few polymorpho- and mononuclear eosino- philes were found. The larger pulmonary veins, however, con- tained numbers of these cells. In sections of the liver, it was often impossible to recognize the tissue. There was widespread fatty degeneration and nuclear fragmentation. In some places the liver cells were small, shrunk- en, very granular, and stained intensely with eosin; in other places they were swollen, and contained large and small fat drops. The 268 W. T. Howard, Jr., and B. 0. Perkins. blood-vessels showed no special changes, but many capillaries con- tained numerous neutrophiles, but no eosinophiles. The spleen showed congestion, with thickening of capsule, trabe- culae and pulp, the latter containing many polymorphonuclear eosiaophiles. The kidneys showed congestion and oedema, with cloudy swelling of the epithelium of the convoluted tubules. In the stomach and large and small intestines there were a num- ber of polymorpho- and mononuclear eosinophiles in the inter- glandular stroma. In sections of the tongue and diaphragm there were large num- bers of trichinae of varying size situated in the muscle fibres. The trichinae were usually single, but in many places two and even three were seen in the same cyst. The trichinae were situated in the muscle fibres, and were surrounded by a hyaline capsule of varying thickness. In many places, especially at the poles of the cysts, there was very active multiplication of muscle nuclei, which were for the most part long, vesicular, and often irregularly con- stricted. Muscle fibres containing trichinae, both near to and at a distance from the latter were swollen, had lost their striations and showed large and small finely granular amorphous areas of de- generation which stained deeply with eosin. About some trich- inae there was little cellular reaction, while about others great numbers of cells were seen. In many places there was marked proliferation of muscle nuclei in fibres in which no trichinae were to be found. The process was evidently an acute one. There was no evidence of calcification. In the cellular reaction about the trichinous areas the following varieties of cells could be distinguished: I. Cells in every way like plasma cells, having a single, round, or oval, somewhat pale nucleus, containing larger and smaller chromatin masses, surrounded by a varying amount of homogene- ous cytoplasm which stained faintly with eosin. In some of these cells the nucleus was centrally placed, but in most it was eccen- tric. In shape these cells were round or long oval; occasionally they were very much elongated (amoeboid movement?). They were the most numerous cells present. II. Cells with nuclei of the same size and appearance as No. I, surrounded by homogeneous cytoplasm which stained deeply with eosin. These cells were fairly numerous. Eo^inophile Granulations. 269 III. Cells with, nuclei similar to l^os. I and II, surrounded bj cytoplasm containing very fine eosinophilic granulations. IV. Cells of the size of large mononuclear leucocytes, with rather deeply staining nuclei, surrounded by either finely granular or coarsely granular deeply staining eosinophilic granulations. These cells were quite numerous. Y. Polymorphonuclear cells, with reniform horseshoe-shaped, or trilobate nuclei, and cytoplasm containing either finely or trilobate nuclei, and cytoplasm containing either finely or coarsely granular eosinophilic granulations. These cells were numerous, and were the exact counterpart of the polymorphonuclear eosino- philic cell of the blood. VI. A variable number of small round cells with single deeply- staining nuclei, surrounded by a narrow rim of granular proto- plasm. These were typical lymphocytes. VII. Cells of varying size, usually very large, containing large, elongated, vesicular nuclei. These cells were very numerous in places, and were evidently proliferating muscle cells. VTII. A few typical fibro-blasts. IX. A few giant cells with from four to six nuclei, and homo- geneous non-granular protoplasm. X. A few neutrophilic polymorphonuclear leucocytes. In some places, especially at the margins of the areas of cellular infiltration, there were seen rows of polymorpho- and mononuclear cells with eccentric nuclei, and many with fine or coarse eosino- philic granulations. The blood-vessels were dilated but showed no increase of leu- cocytes in the small veins, but in some capillaries there were a considerable number of mononuclear cells. Careful search failed to discover transition of polymorphonu- clear neutrophilic cells into eosinophiles. There was absolutely no evidence of this mode of origin for the eosinophilic cells. The most numerous and apparently the most important cell in the reaction in this ease was the plasma cell. The cells with eosino- philic granulations were much less numerous than the plasma cells. The development of the former from the latter could be readily and easily traced as follows: The homogeneous proto- plasm of the plasma cell becomes first faintly and then strongly eosinophilic- In the next stage, fine, and later coarser eosino- 270 W. T. Howard, Jr., and R. G. Perkins. philic granulations develop, producing typical mononuclear eosino- philes. Some of these latter develop polymorphous nuclei. With the development of granulations the plasma cells often increase in size. The size of the eosinophilic cells varies, hov^ever, within rather wide limits. In this case at least, it seems certain that the eosinophiles were formed in the areas of cellular reaction about the miiscle trichinae, and were not brought there from a distance by the blood-vessels being drawn by chemotactic substances thrown out by the parasites, or furnished by the degenerating muscle. The substances present, whatever their nature, exerted a chemo- tactic action on the plasma cells. The presence of increased numbers of eosinophiles in the stroma jf the mucosa of the stomach and intestine is probably to be ex- plained by the recent presence of trichinae in the intestine and intestinal wall. Here too the eosinophiles probably had their origin in plasma cells. The course of the disease in this case was n-ore acute than in the cases of Brown,' Gwyn,° Atkinson' and Stump ° and this fact probably explains the absence of circulatory eosinophilia. Had the individual lived longer, or the infection been less severe, it is probable that eosinophiles would have made their way into the blood. It is recognized that the presence of a circulatory eosinophilia during life cannot be positively denied in the absence of a differential count, on the strength of Dr. Hoover's impression from the examination of stained preparations. If, however, the eosinophiles were much increased in the circula- ting blood during life, it would be strange to find the percentage i.ormal in differential counts of preparations made from the blood of the organs after they reached the laboratory. This case demonstrates among other things that trichinosis may exist and run a fatal course apparently without increase of eosino- philes in the circulating blood, and that a high grade of eosino- philia, such as was present in the cases of Brown, Gwyn, Atkinson and Stump (42, 49, 68.2, 72, 58.5 and 52 per cent) speaks for a 5 Brown, T. E. Bulletin of the Johns Hopkins Hospital, 1887, vol. VIII. Journal of Experimental Medicine, 1898. «Gwyn, N. B. Centralbl. f. Bakteriologie, 1899, Band xxv, S. 746. ' Atkinson. Philadelphia Medical Journal, June 3, 1899. 8 Stump. Philadelphia Medical Journal, June 17, 1899. Eosinophile Oranulations. 271 good prognosis and indicates that the organism is reacting well to the poison of the trichinae. The marked degeneration of the liver and kidneys of our case shows that the tissues were acted upon by a powerful poison. Brown's observation that the eosinophiles in the muscle in trichinosis are formed in the lesions and are not at- tracted there by chemotactic substances, is supported by the find- ings in this case. We believe that trichinae or a poison or poisons secreted by them have a positive chemotaxis for hyaline leucocytes, plasma cells, and polymorphonuclear leucocytes and that for some imknown reason certain of the first two varieties of cells develop eosinophilic granulations. HiSTOGElfESIS OF EOSINOPHILIC CeLLS. There are several theories in regard to the histogenesis of eosinophilic cells and in support of each theory a certain amount of evidence has been brought forward. These theories and their main supporters are as follows : I. From polymorphonuclear neutrophilic leucocytes by a kind of ripening process. Max Schultze,' GuUand," Zappert," Brown" and others. II. From eosinophilic myelocytes in the bone-marrow only, Ehrlich." III. Division of pre-existing cells of the same kind, by mitosis, Miiller und Eieder," and Van der Stricht;" by amitosis, Eenaut," Denys" and Arnold; GuUand," who has exhaustively studied the leucocytes of many species of animals thinks that all forms of leu- cocytes give rise on division to lymphocytes, but that occasionally on the division of eosinophiles the daughter-cells may contain a few eosinophilic granulations. 9 Schultze, Max. Arch. f. Mikroskop. Anat. 1865, I. M Gulland. Journal of Physiology, 1895-6, Vol. xix. " Zappert. Zeitschrift 1 Klin. Med. 1893, Bd. xxiii. "Brown. Loc. cit. "Ehrlioh and Lazarus. Die Anemie. Wlen, 1898. » Miiller und Rieder. Deutsch. Arch. f. Klin. Med., 1891-3, Vol. xlviii. 15 Van der Stricht. Quoted by Brown, loc. cit. " Renaut. Quoted by Brown, loc. cit. " Denys. Quoted by Brown, loo. cit. 1* Gulland. Loc. cit. 272 W. T. Howard, Jr., and B. 0. Perkins. IV. From the connective-tissue of various parts of the body, Gol- lasch," ISTeusser/" Weiss'' and others. Kanthack and Hardy found these cells in numbers in the caelomie cavity and connective- tissues of animals. V. Occurring as incompletely developed erythrocytes, derived from haematoblasts, Przewaski.'' VI. From the engulfing of haemoglobin of haematoblasts by leucocytes, Sacharoff,"' who thinks that the granules are connected with the dispersion of haematin and that the intense coloration of eosinophilic granulations is due to the greater thickness of their iron-bearing paranuclein. According to Sacharoff, bacteria taken up by leucocytes may furnish eosinophilic material, their nuclein being changed into eosinophilic paranuclein. Klein"" thinks that eosinophiles are formed from other leucocytes by the taking up of iron derived from red blood cells. VII. From lymphocytes by transition, Pappenheim.''' It is neither impossible nor indeed improbable that these cells may have their origin in any of these several ways. Ehrlich and Lazarus strongly combat the idea of the local for- mation of eosinophiles. While not denying that these cells may have their origin in many ways, in the study of our rather large material, we have never been able to trace any one of these modes of origin (except from eosinophilic myelocytes). We have been struck on the other hand with the great frequency vwth which the origin of these cells can be traced from the hyaline leucocyte, and espe- cially from the plasma cell. In case after case, in organ after organ, and especially in the lymphoid tissues (the thymus, lymph glands, the spleen, and the bone-marrow), in the gastro-intestinal mucosa (gastritis, lympho-sarcoma, colitis, and appendicitis), in iSGollasch. Fortschritte d. Med., 1S99, Bd. Tii. MNeusser. Wien. klin. Wochenschr., 1892, Bd. xli, p. 64. «i Weiss. Quoted by Brown, also by H. F. Miiller. Centralbl. f. allgem. Path, u. path. Anatomie. 22 Kanthack and Hardy. Journal of Physiology, 1894-5, Vol. xvii. 2iiPrzewaski. Centralbl. f. allg. Path, und path. Anat., 1896, Bd. fii, S. 177. 2* Sacharoff. Centralbl. f. Bakteriologie, 1897, Bd. xxi, S. 365. 26 Klein, 8. Centralbl. f. Innere Med., 1899, Bd. L, S. 28. 2«Pappenheim. Virchow's Archiv., 1899, Bd. 157, p. 71. Eosinofhile Granulations. 273 exudations rich in lympliocytes, hyaline leucocytes and plasma cells (acute and chronic nephritis, salpingitis and ovarian abscess, etc.) and in various other processes, we have observed the transition of both these varieties of cells into eosinophiles. As already pointed out, the hyaline, or better as Kanthack and Hardy aptly term it, the " ground glass " cytoplasm of the hyaline cell or the plasma cell (for the latter often comes from the former) stains more deeply "vvith eosin than its neighbors. In other cells with clear nuclei, which are often eccentric, there are fine gran- ules which stain deeply with eosin. Other cells with similar nuclei have more coarsely granular and others still very coarsely gTanular cytoplasm. With the development of the eosinophilic granules the cells usually, but by no means always, increase in size. With the development of the granules the nuclei become less vesicular, and the nuclear chromatin more diffuse. Many coarsely granular as well as finely granular mononuclear eosinophiles were seen, but always a large proportion of the mononuclear cells de- velop polymorphous nuclei. We have been struck with the large proportion of mononuclear eosinophiles present in some cases. If the idea of Gulland and others that the change in the shape of the nucleus of the leucocyte is dependent largely upon the motion of the cell is correct, the above fact is another argument in favor of the origin of eosinophiles from plasma cells. In size the eosino- phile in the tissues varies within rather vsdde limits, from the size of a small plasma cell, scarcely larger than a lymphocyte, to that of the typical eosinophile of the blood." In Case XX we observed appearances which we cannot inter- pret otherwise than that they represent the development of coarsely granular eosinophiles from smooth muscle cells. In the same case- there were also many examples of the plasma cell origin of these cells. Cbaeactee of Eosinophilic Ge adulations. We have nothing to add to the knowledge of the character and composition of these granules. For the literature of the subject "A good description with » full account of the plasma cell will be found in Councilman's article on Acute Interstitial Nephritis, Journal of Experimental Medicine, 1898, vol. Ill, p. 393. See also Marschalko's Zur Plasmazellenfrage, Centralbl. f. AUgem. Pathologie und path. Anatomic. 1899, Bd. x, S. 851. 274 W. T. Howard, Jr., and B. G. Perkins. the reader is referred to Brown's recent article in the Journal of Experimental Medicine (loc. cit.). First mistaken for fat, which was disproved by Ehrlich,'' who first suggested that they represented a nutritive reserve formed by the cell, these granules have been considered in turn non-proteid (Schwarze) ;'" proteid (Zappert);™ haemoglobin or its derivatives (Przewaski" and others); nucleo-albumin (Sherrington);"' and se- cretory granules of a defensive nature (Hankin),"° Kanthack and Hardy." Gulland,°° however, insists that they are not products of the metabolic activity of the cell imbedded in the protoplasm, but that they represent an altered condition of the microsomes, form a part of the cytomitoma, and are therefore, plastic and not para- plastic, as is generally believed. GuUand thinks that they are concerned in the amoeboid activity of the cell. As previously stated we were unable to demonstrate the presence of iron in the granules of eosinophiles in hardened tissues. Presence of Eosinophiles in IsToemal Oegans. Beside the blood, coarsely granular eosinophiles are normally found in various tissues of the human body, the spleen, thymus (earliest mention probably by J. Sehaffer*'), the lymph glands, bone-marrow, the gastro-intestinal tract (especially the stroma of the mucosa), in the kidneys and liver, and in various connective tissues. They are found in the same locations in various lower animals. They form from 25 to 40 per cent of the wandering cells in the peritoneal cavity in the guinea-pig, and from 30 to 50 per cent in rabbits (Kanthack and Hardy). In these animals they are also numerous in the subcutaneous tissue. We have found these cells but rarely in apparently normal organs of adults — in ^'Ehrlioh. Loc. cit. S. 86, and in other writings. ''Schwarze. Ueber eosinopliile Zellen. Inaug. Diss. Berlin, 1880. ^ Zappert. Loc. cit. " Przewaski. Loc. cit. 32 Sherrington. Proceed. Royal. Soc. Lond., 1894, vol. It. Also quoted by Brown. as Hankin. Centralbl. f. Bakteriologie, 1892, Bd, xi. "Kanthack and Hardy. Loc. cit. •'SGulland. Loc. cit. 'BSchafler J. Centralbl. f. d. Med. Wissensch., 1891, nos. 22 u. 3.S. JSosinophile Granulations. 275 ten cases in the appendix, once in tlie Fallopian tube, and once in the gastro-intestinal mucosa." Heidenhain found them constantly present in the intestinal mucosa of dogs, in which their numbers were greatly increased after fasting and a dose of magnesium sul- phate. Eosinophiles were not uncommonly found in various or- gans in children dying both with and without septicaemia. "We have also been able to confirm the observations of those who have described eosinophiles in numbers in the blood and tissues of the foetus. Presence of Eosinophiles in Inflammatoey and othee Lesions. With a few notable exceptions coarsely granular eosinophiles have not been described in inflammatory reactions and in infec- tious granulomata. They have been found in gonorrhoeal pus in numbers by Gollasch°' and others, in the sputum in bronchial asthma by Gollasch,°° Leyden " and others, in the sputum in bron- chitis by Leyden," Teichmiiller, Grunwald*" and others; in the skin in lymphoderma perniciosa by Weiss *" and in leprosy by Ja- dassohn;" in the blisters of pemphigus by Neusser" and others; in Diihring's disease by Leredde and Perrin;" in blisters produced artificially by Bettman" and by Kanthack and Hardy; in hemor- rhagic pleurisy by Klein*' and Harmasen." Gronven"" found 3' On further study we find eosinophiles constantly present in the stroma of the mucosa and often in the lymphoid tissue of both the normal and inflamed appendix. ^^Gollasch. Loc. cit. 39 Gollasch. Loc. cit. « Leyden. Deutsch. Med. Wochenschr., 1881, Bd. XTii, S. 1085. ,t , On ., external "examination it; was evei;y\yhere }ia.rd,j.^rni andj^ip-: elastic... .iMid\yay between ths cardia;a,nd pylorus the organ ,wa,s , markedly constricted, roughly resembling an hour-glass in shape. On section its wall was everywhere greatly thickened, being from 1 to 3 or 4 cm. in thickness. At the point of con- striction the lumen was 3 cm. in diameter. The mucosa, which was of a brownisli-red color, was unbroken. The thick- ening of the stomach wall was due to the presence of a dense and firm grayish-white tissue, which had a rather homogeneous appearance. This tissue apparently did not encroach upon the mucosa, but involved the submucosa and muscularis, both of which coats were extensively infiltrated. The gastro-hepatic, mesenteric and retro-peritoneal lymph glands were enlarged and liad the same appearance as the growth in the wall of the stomach. The other organs showed nothing of present interest. Sections of the stomach made of various portions showed in some places disintegration of the mucosa. In some places it had entirely disappeared. In all the sections examined, the submucosa was transformed into a fibro-cellular tissue composed of newly-formed fibrous tissue containing variable numbers of small round and oval cells. The cells showed no special ar- rangement, but often occurred in groups and masses and were always supported by fibrous tissue or by a fibrillar intercellular substance. The individual cells varied from 8 to 15 mikrons in diameter. They had a single round or oval nucleus which usually stained deeply and uniformly with hematoxylin. Some cells had paler nuclei. The cell bodies were usually round or slightly oval. A few spindle-shaped cells were seen. In all sections the muscularis was markedly invaded, in many places diffusely. All grades of invasion of this coat occurred from small islands of tumor cells to complete transformation into tumor tissue. The muscle tissue was pushed aside and in- filtrated by the new growth. There was no evidence of pro- liferation of the muscle cells, which, so far as could be deter- mined, played no part in the tumor formation. The origin ot the tumor from the fibrous tissue of the submucosa was clear. A few blood vessels were found in the new growth. In no places was invasion of blood vessels made out. The amount of fibrous tissue present in some parts of the tumor warrants the term fibrosarcoma. Sections of the lymph glands showed a growth in all respects similar to that of the stomach. Anatomic Diagnosis: Small round and oval cell sarcoma of the stomach, primary in the submucosa, and infiltrating the muscularis. Constriction producing hour-glass shape of the stomach. Secondary sarcomatosis of the gastro-hepatic mesen- teric and retro-peritoneal lymph glands. Anemia. Case 2. — Mixed cell sarcoma of the stomach with abscess formation in the tumor, with metastases in the neighboring lymph glands. R. F., female, aged 46 years, was admitted to the Lakeside Hospital, Sept. 7, 1898, service of Dr. Dudley P. Allen (Dr. Nevison acting), complaining of anorexia, headache, pain, and swelling in the region of the spleen. There was no vomiting and no pain in the epigastric region. A large mass could be felt in the region of the spleen. Examination was otherwise negative. The urine contained a trace of albumin but no casts. The red blood cells were normal in number and ap- pearance. A considerable leucocytosis was noted. Clinical Diagnosis: Tumor of the spleen. The patient died 12 days after admission. Autopsy a few hours after death. Anatomic Diagnosis: Primary mixed cell sarcoma of the stomach with abscess formation and peritonitis. No metas- tases. ICdema and congestion of the lungs. Adenoma of the duodenum. Streptococcus pyogenes and Staphylococcus pyo- genes aureus, and a short liquefying bacillus in the peritoneal exudate and abscess of the tumor of the stomach. The body was 172 cm. long. The general nutrition was good. There were no scars, wounds or growths upon the sur- face of the body. There was bilateral enlargement of the thy- roid. The chest was well shaped, the pleurae negative; the lungs showed congestion, edema and emphysema, but were free from tumors. The pericardium, heart and aorta are without present interest. The abdomen was distended and very tym- panitic. The abdominal wall contained a thick layer of fat; the muscles were pale. The parietal peritoneum was covered with a thin fibrino-purulent exudation. The coils of the in- testine were bound together by recent fibrous adhesions, con- taining larger and smaller collections of creamy pus. The liver projected 4 cm. below the costal border in the mammary line. The under surface was adherent to the stomach, and the colon. The stomach was intimately bound to the surrounding structures, and was displaced downwards towards the left, so that its posterior surface lay over the left kidney. Its an- terior wall was of ordinary size and appearance. The poste- rior wall was thickened. On the right side there was a ragged opening, 8x4 cm., between the peritoneal cavity and with a large cystic area in the posterior wall of the stomach. Stomach: On removal and section the stomach contained foul- smelling yellowish material. The cubic capacity of the organ was reduced, while its bulk was markedly increased. The an- terior wall was normal in appearance. Near the lesser curva- ture, and aflecting nearly the whole of the posterior wall, there was a tumor mass 20x16 cm. which varied from 0.5 to 8 cm. in thickness. Near the lesser curvature there was a fungus- like mass, 0x8 em. in breadth and length, which protruded into 10 the iumen of the stomach. This piotuheranee was irregular in outline and ulcerated, and covered wjth a foul-smelling yel- lowish pus. The tumor did not extend to or involve the pylorus, which was quite soft. The tumor lay beneath the mucosa everywhere except at the ulcerated area. At the cardiac end of the tumor there was a cyst 7x4 cm., and at the pyloric end a, similar cyst 6x5 cm.; between these there was a larger cyst, communicating with the other two. These cysts contained fluid and necrotic material. The larger cyst communicated with the peritoneal cavity. The tumor varied very much in thickness, the cyst walls being from 1 to 3 cm. thick. The tumor involved the submucosa and invaded the muscularis and mucosa irregularly. On section it varied much in consistency, in some places it was soft, while in others it was dense and firm. On the right side of the stomach near the tail of the pancreas, there was a nodular mass, apparently an enlarged lymph gland, measuring 8x4 em., which on -section had the same appearance as the growth in the stomach, and contained a cyst 3x2 em. in size. The neighboring lymph glands were swollen. The posterior wall of the stomach was bound to the pancreas, which was not, however, invaded by the tumor. The duodenum was markedly congested; 8 cm. below the pylorus there was a small nodule, 1.5 cm. in diameter, situated in the submucosa. The rest of the intestines showed nothing abnor- mal.. The other organs showed nothing of present interest. Sections of the stomach made at a distance from the tumor showed nothing abnormal. Sections from various portions of the tumor all showed much the same structure, the tissue being com- posed of rather large round, oval and spindle-shaped cells of the connective-tissue type, and with relatively large nuclei. The cells were not arranged in alveoli, but a supporting connective- tissue could be made out. In some places, especially at the borders, the tumor was fibro-cellular, while in others it was markedly cellular in structure. Larger and smaller areas of necrosis were found. Sections taken from the margins of the cysts showed marked necrosis, with nuclear fragmentation. In this material there were large and small clumps of large bacilli. In a few places there was an intimate relation between the tumor cells and the adventitia of arteries and veins, suggestive of the vascular origin of the former. These areas were always isolated. In other places, at a considerable distance from the main tumor, there were small irregular islands of typical sarcoma tissue and apparently springing from the connective tissue of the submucosa. The tumor was situated in the sub- mucosa and invaded the muscularis and the mucosa in only a few places. Its origin was without doubt in the submucosa. Sections of the smaller tumor showed the same structure met with in the larger. No trace of lymphatic tissue, however, re- 11 mained. There was markedly little inflammatory reaction about and in the tumors of the stomach; no groups of tumor cells were found in either blood or lymph vessels. Some of the veins, however, contained recent thrombi. Sections of the small tumor situated in the submucosa of the duodenum showed a typical adenoma, composed of a large num- ber of glands divided into lobules by connective tissue bands. The glands were lined with a single row of epithelial cells, which were columnar in some and euboidal in other glands. The cytoplasm of these cells was finely granular and stained poorly; the nuclei were vesicular and placed at the base of the cells, where the latter were supported by a delicate membrana propria. The epithelial cells formed a single row in the glands and never broke through the membrana propria. No ducts were to be found. The tumor lay almost entirely in the sub- mucosa, but in some places had push'ed through the inuscularis mucosEE and encroached upon the mucosa, which was normal in appearance. Sections of the tumor of the thyroid showed a papillary adenoma. The other organs are without present in- terest. Case 3. — Sarcoma of the pylorus. G. H., white, a laborer, aged 48 years, was admitted to St. Alexis' Hospitaf, service ot Dr. Cogan, Oct. 26, 1900, complaining of vomiting and yire&k- ness. His family history was negative. The patient stated that he had always enjoyed good health until December, 1899, when his appetite failed and he began to lose weight. In October, 1900, he had to give up work on account of weakness and discomiort in his stomach. At this time he began to vomit after each meal. Aftei' admission to the hospital he vomited almost continuously until his death, which occurred on Novem- ber 3. No food could be retained by the stomach. Examination of the chest was negative. No tumor could be felt in the abdomen. On testing the stomach contents the acidity \Aas normal. Motility and absorption were both dimin- ished. The red blood cells numbered 1,000,000 per cubic milli- meter. The leucocytes were not increased. Anatomic Diagnosis : Sarcoma of the pylorus and neighboring portion of the stomach. Broncho-pneumonia of both lungs. Chronic adhesive pleuritis. Anemia. The body was poorly nourished. Rigor mortis was present. The other organs showed nothing of present interest; only the description of the stomach is taken from the autopsy protocol. Stomach: This was somewhat dilated. On opening the organ the pylorus and the wall of the stomach on all sides for a dis- tance of 10 cm. from the pylorus, were found to be thirckened. The mucosa was unbroken. The gastric wall at the pylorus measut'ed from 1 to 2 em. in thickness. The tumor on section Has at nearly every point beneatli the mucosa, •and its chief 12 seat was the submucosa, but in many places it could be traced into the muscularis. The tumor was homogeneous in appear- ance and firm in consistency. The lymph glands in the neigh- borhood of the stomach were not enlarged. No metastases were found in any organ. In sections made from various portions of the tumor the mucosa was atrophied, the tubules having been compressed and the interglandular stroma replaced by fibrous tissue contain- ing, round, oval or fvisiform cells. The muscularis mueosse was absent in many places. The submucosa was everywhere thick- ened, due to a new growth of fibrous tissue, containing a vary- ing number of fusiform, round and oval cells, supported by a variable amount of intercellular substance. At many places this tissue extended into the muscularis. In some places the sarcomatous tissue was markedly fibrous. Throughout the tumor the capillaries were numerous. In many places the lymph vessels were dilated and contained tumor cells. In some places the endothelial cells of the lymphatics showed marked proliferation and nearly filled the lumina of these vessels. In some lymphatics the lining cells had assumed a cuboidal shape and the structure had the appearance of a tubular gland. The new gro\vth in some places invaded the nerves and even the serosa. There was no peritonitis. Case 4. — A saloonkeeper, aged 39 years, married and a moder- ate drinker, 'suffered with occasional attacks of indigestion and discomfort in the epigastrium for a year before consulting Dr. Robert H. Sunkle, to whose kindness I am indebted for the history and autop.sy in this case. When first seen by Dr. Sunkle, the man complained of pain in the abdomen, loss of weight, strength and appetite. The liver was greatly and uni- formly enlarged, but no nodules were palpated. No tumor could be made out in connection with the stomach. The heart and lungs were negative. There was occasional vomiting, but no special pain referable to the stomach. A few days before death, which occurred one month after Dr. Sunkle first saw the case, dulness was made out in the flanks. Dr. Sunkle's diagnosis was hypertrophic cirrhosis of the liver, in which he was upheld by Dr. Sihler, who. saw the case in con- sultation. Later the possibility of carcinoma of the liver was considered. ^ ^ The autopsy was made by Dr. Sunkle, who brought the organs to my laboratory for examination. Anatomic Diagnosis: Primary angio-sarcoma of the lesser curvature of the stomach, with metastases in the neighboring lymph glands and in the liver. * As the stomach and liver were the only organs showing changes of present interest, a description of the other organs is omitted. 13 Stomach: The organ was of ordinary size. The lesser curv- ature was thickened. On section, the cardiac and pyloric ends were normal, as were the greater curvature and the anterior and posterior walls. The mucosa was somewhat thickened. The whole length of the lesser curvature was thickened and from it a mass of tissue protruded into the cavity of the organ. The mass was from 1.5 to 2 cm. in thickness and 4x5 cm. in outline. The surface was irregularly ulcerated and it was difficult to make out the mucosa. On i=eeUon the tumor was soft, readily broken down, and lacked the pearly translucency of carcinoma. The growth extended through the coats of the stomach, and in- cluded and surrounded several enlarged bmph glands, which were virtually involved in the growth. On section these glands were soft and had the same appearance as the gastric tumor. The peritoneal cavity contained a large amount of clear fluid. There was no peritonitis. The liver was markedly enlarged, its edges rounded, the capsule mottled red and yellow. Scattered over the surface, but lying well beneath the capsule, there were a number of soft yellowish areas. On section both the right and left lobes of the organ were the seat of numerous grayish-yellow areas of soft consistency, sharply marked off' from the surrounding liver tissue, and varying from 0.5 to 3 or 4 cm. in diameter. Many of the large branches of the portal vein were occluded by grayish-yellow masses, which were here and there mixed with blood and coagula. Portal vessels could be traced into many of the larger masses, which often contained central hemorrhagic areas. The liver tissue between the masses was congested and in places compressed. The external surface of the gall bladder showed a small nodule similar to those found in the liver. The gall bladder and bile ducts on section were normal. The mesenteric glands and other organs were free from metastases. Histologic E anamination: Stomach. In sections of the gastric wall made at a distance from the tumor, there was disintegra- tion of the superficial layer of the mucosa. In sections cut through the tumor at various places the mucosa showed disintegration to a varying depth. In some sections the deeper layers of the mucosa were invaded by tumor tissue which com- pressed and replaced the tubules and interglandular tissue. The submucosa was markedly thickened, due to the presence of large areas of tumor tissue which was diffuse in some places and in others sharply circumscribed. This latter appearance was usually due to the extension of tumor tissue into large and small veins. The tumor evidently had its origin in the submucosa and varied in structure considerably. In many places the structure was typical of a mixed cell sarcoma, being com- posed of large and small round and oval cells, large spindle- shaped and numerous giant cells, and a variable amount of 14 intercellular tissue, which was conspicuous in some places and scanty in others. In general, the tumor was very cellular. In ^11 parts of the growth, capillaries, most of which contained blood, were numerous. In most places the relation of the tumor cells to the capillaries -was intimate, viz., masses of round, oval, flattened and often fusiform cells surrounded blood capillaries, and were placed directly in relation with their walls. In some places, especially in the metastases in the liver, tumor cells were arranged in a regular row about a small lumen, in such a manner as to resemble a cross-section of a tubule. Blood cells and granular material were present in some of these spaces. In still other places the tumor cells were arranged in large alveoli, supported by a variable amount of intercellular tissue and capillaries. The muscularis was only slightly invaded, and the mucosa was markedly invaded in places. The invasion of the large and small veins of the sub- mucosa was marked. Most of these vessels in the tumor area weve completely occluded by tumor tissue, similar to that above described. The tvimor cells varied much in size and shape. Their nuclei were large, round or oval, some staining densely, while in others fine cliromatin masses could be made out. In many nuclei there was a rather large oval or round hyaline body, which stained pink with eosin. Nuclear figures were common. The giant cells had very large single nuclei. Many of the tumor cells showed marked fatty degeneration. The growths in the lymph glands and the liver were identical with that of the stomach. The relation of the tumor cells to the capillaries was especially well marked in the liver metastases. Some of the branches of the portal vein contained both tumor cells and fibrinous thrombi. Many were completely occluded. In many places the new growth in the liver had passed through the veins and was infiltrating the liver tissue. In sections some distance from metastases, there was marked necrosis of the central portions of the liver lobules with hemorrhage. There could be no doubt in regard to the sarcomatous nature of the neoplasm, which evidently sprung from the blood vessels of the submucosa of the stomach. Frequency. — The disease is certainly rare, but prob- ably occurs more commonly than is generally supposed. There can be no doubt but that in the absence of routine microscopic examination of tumors found at operation and autopsy, many cases must have been mistaken for carcinoma, which, indeed, was the clinical diagnosis in a considerable number of the recorded cases. So far there are sixty-one known cases. Next to carcinoma it is eer- pu^-^ ^^^ ^^^ whooping cough, diphtheria, in- '°^'^' A tvnhoid fever, but none of these recently. Six fluenza, and typnom , , , , j • • j weeks previous to admission she had had a misearriage, and rtTeVhad been discharge from the vagina ever since, though she was not conscious of any fever. Physical examination of the chest and abdomen revealed nothing of importance. Vaginal examination showed a soft, somewhat enlarged uterus, with a foul-smelling discharge from the cervix. The urine contained a trace of albumen, with a few casts, and gave a well-marked sugar test with Fehling's solution. The presence of sugar persisted till death. The blood on admission showed 72 per cent, of haemoglobin, and a leucocytosis of 14,000, which increased to 32,000 the day before death. The temperature on admission was 100.5° C. and varied from 96° to 103.2°, the last taken being 103°. The pulse on admission was 90, but it rose steadily, reaching 180 the day before death. The patient complained of much pain, and had marked opisthotonos and inequality of the pupils. She became rapidly worse, sank into a comatose condition, and died eight days after admission. Her condi- tion was never such as to admit of any operation. Clinical diagnosis : meningitis, probably tubercular. The autopsy was held 19 hours after death, the body having been in cold storage at 0° C. Pathological Brief. — Anatomical diagnosis. Fibrino- purulent endometritis after abortion. Fibrino-purulent cerebro-spinal meningitis and ependymitis, with the presence of B. pyocyaneus in the uterine and cerebro-spinal exudates. Congestion, oedema, emphysema, and atalectasis of the lungs. Cloudy swelling of the liver and kidneys. Corpus luteum of pregnancy in the left ovary. General chronic passive con- gestion. The abdominal and thoracic organs showed nothing of present interest. There was no peritonitis. The left ovary showed a corpus luteum of pregnancy. The tubes were J parently normal. The uterus was rather large, soft and B. PYOCYANEUS. 285 filled with an exceedingly foul-smelling brownish semi-fluid material. At the upper right-hand side of the uterine cavity just below the cornu there was an adherent mass i X i X .5 cm. in size, soft and necrotic, suggesting placental remains. Examination of the head showed no abnormality in scalp or skull. The vertex and the ventral portions of the brain were negative, the base was covered with a thick yellowish-white purulent exudate, rather firm in consistency, filling up all the crevices of the lower surface. It extended from the ventral border of the pons, dorsally to the medulla and cord, and down the cord to its extreme lower end. Examination of the ventricles showed purulent fluid with a heavy sediment of pus, but no adherent exudate, the process being apparently of later date than that on the surface of the brain. All the ventricles were involved. The sinuses showed no thrombi or other changes. None of the exudates, either uterine or cerebral, showed any blue or green color. Cultures were taken as is usual in our autopsy routine from all the organs and all the pathological exudates. Coverslips were made from the uterus, the meninges, and the ventricles. Coverslips from the uterus showed a large variety of organ- isms including cocci and numerous slender bacilli, staining irregularly with methylene blue ; a few large stout bacilli were also present. Coverslips from the meningeal exudate and from the ventricles showed numerous long, rather slender bacilli which showed a marked tendency to irregular staining, after the manner of the Klebs-Loeffler bacillus, to which they bore a close resemblance. Cultures from the heart, spleen, and kidneys, kept at 37° C. for 48 hours, showed no growth. Cultures from the lung showed fairly numerous colonies of a large, unidentified coc- cus, not pathogenic for rabbits. Cultures from the base of the brain, from the ventricles, and from the spinal cord showed very numerous colonies, those on the surface large, spreading, blue in color, many with a rather concentric ap- pearance. Homogeneous under the low power. The col- onies in the depth were opaque and yellowish. The plates 286 PERKINS. had a strong, sickly, sweet odor, resembling that of the pyocyaneus group. Cultures from the liver showed a few colonies of the same type. Cultures from the uterus showed a variety of colonies, most numerous of which were colonies like those in the brain and liver plates, associated with colonies of the yellow and white staphylococci. The plate had an odor similar to the plates from the brain and cord. Coverslips from these colonies showed rather long bacilli, staining irregularly with Loeffler's methylene blue, many of them with slightly clubbed ends. Examination by the hang- ing-drop method showed no motility. Transplantations were made into various media, with the result that the growths and characteristics changed materially, until the third and fourth generations, since when no change has been noted. The first generation showed no color, no motility, and very slight liquefactive powers, but with successive transplanta- tions the color became a dark green, the motility became marked, and the power of liquefaction increased, until now the cultures answer well to the usual descriptions of B. pyocyaneus. Throughout the cerebro-spinal tract and in the uterus, two varieties were isolated, both forming green pigment, and only differing from one another in the vigor of their growth, and in their chemical activities. The liver colonies all be- longed to one variety. The essential cultural characteristics of both organisms were as follows, in 24-hour cultures : Glycerin agar. Grayish-green, moist, translucent growth with wavy edges ; water of condensation cloudy, greenish, medium uniformly light green. Growth to bottom of stab, no gas. Glucose. Same. No gas. Bouillon. I-ight green color with well marked fluores- cence, distinct viscidity, grayish scum on surface, medium cloudy throughout ; no sediment. B. PYOCYANEUS. 287 Litmus milk. Reaction unchanged. Human blood serum. Medium greenish. The growth lies in a trench made by beginning liquefaction. Gelatin. No liquefaction. Potato. Light brown growth; water of condensation cloudy, potato not discolored ; sickly, sweet odor well marked. In cultures two to twelve days old, both milk and blood serum were peptonized without change of reaction, the filtered solution giving the Biuret test for peptone ; «gelatin tubes were liquefied completely in 10 to 12 days, the lique- faction early reaching the sides of the tube, the growth on potato became more markedly green, and the odor on all media was more distinct. Coverslips from the cultures, after the second generation, showed short rather oval bacilli i.o to 1.5 /u, in length, by 0.33 to 0.50 /w, in width, which were actively motile and decolorized by Gram's method. Owing to the depth of the color, and to the change of the blue to red when acidulated, the indol test was difficult of demonstration, but it was thought that a posi- tive reaction was obtained in the earlier generations, though not later, agreeing with Jordan's observations.^ Extraction with CHCI3 revealed the presence of pyocyanin in large amount. Microscopical Descriptions. — The liver and spleen showed well-marked congestion and cloudy swelling. The pancreas showed no changes, either in the areas of Langerhans, or elsewhere. The other abdominal and thoracic organs showed nothing of present interest. The uterus was markedly con- gested, the mucosa was lost in places, and covered with a thick, fibrinous exudate, containing numerous polymorpho- nuclear leucocytes and plasma cells. Large numbers of short, slender bacilli were seen together with a smaller number of cocci and some few large, stout bacill^i with rounded ends. The infiltration with cells, and the presence of the slendtr bacilli, extended to a considerable depth below the mucosa. Sections through the meninges and through the under- lying tissues of the brain and cord showed a thick fibrino- 288 PERKINS. purulent exudate, apparently firmly adherent to the tissue be- neath. The exudate next to the nervous tissue was composed of large numbers of plasma cells, with some small round cells, and occasional eosinophiles and mast cells. Further out the exudate was more acute in character and was composed chiefly of polymorphonuclear leucocytes. Fibrin was pres- ent only in small amount. Staining with Weigert's fibrin stain showed no bacteria, but eosin and methylene blue brought out a small number of slender bacilli, with a rather irregular stain. The brain and cord beneath the exudate showed a well-marked chronic inflammatory reaction, the vessels were surrounded with an infiltration of small round cells, and no polymorphonuclear leucocytes or bacteria were noted in the tissues. Sections through the walls of the lateral ventricle showed a purulent exudate without fibrin, and a well-marked infiltration of the tissue in the neighborhood, with fibrinous thrombi in the small veins. A moderate num- ber of bacteria was found both in the exudate and in the brain substance, closely resembling the organisms seen in the meningeal exudate. Animal Experiments. — Inoculations of guinea pigs gave no result. A rabbit inoculated at the root of the ear with J c.c. of a 24-hour bouillon culture died in two weeks with exten- sive, subcutaneous abscess formation, from which B. pyocy- aneus was recovered. Case II. — Puerperal septicaemia, with purulent endome- tritis, broncho-pneumonia with abscess formation in the con- solidated areas. B. pyocyaneus was found in pure culture in the uterus, and associated with a small number of large, un- identified bacilli in the lungs. Clinical Brief . — Negress, 21 years of age. Died Sept. 28, 1899, in the service of Dr. Hoover at the City Hospital of Cleveland, a few days after admission. The only history ob- tainable was that of a full term delivery six weeks previous to- admission, and the onset of chills and fever a few days after the birth of the child. Clinical Diagnosis. — Puerperal infection. Thrombosis of pulmonary arteries, broncho-pneumonia and multiple ab- B. PYOCYANEUS. 28c scesses of both lungs. Acute splenic tumor. Cloudy swell- ing and fatty degeneration of the liver and kidneys, witt acute parenchymatous nephritis. Lungs. — There was marked oedema with multiple areaj of recent broncho-pneumonia, varying considerably in size, Some of these, from 2-3 cm. in diameter, showed central soft- ening with the formation of small cavities containing foul-smell- ing, grayish-yellow pus. This condition involved the entire right lung, but only the lower lobe of the left lung. Both pulmonary arteries showed recent, non-adherent, grayish-red thrombi in their branches. Uterus. — Somewhat enlarged, endometrium dark red in color, but without fibrinous exudation. The other organs showed nothing of present interest. No gas was found in any organ. Coverslips from the lungs in the consolidated areas, and from the uterus showed numerous small, rather slender bacilli, and some large, stout bacilli. Aerobic cultures were made from the various organs, but all remained sterile save those from the uterus and lungs. Cultures from both of these showed a slender bacillus which stained by Gram's method and was actively motile. It peptonized milk without forma- tion of acid, and was otherwise similar to the organism found in Case I. Microscopical Description. — Uterus. The mucosa was lost for the most part. The underlying tissues were deeply in- filtrated with polymorphonuclear leucocytes, and with a large number of small, rather oval bacilli. On the surface of the mucosa these bacilli were associated with a few large, stout bacilli with rounded ends. Sections from the consolidated areas of the lungs showed well-marked broncho-pneumonia. The alveoli were filled with an exudate composed of polymorphonuclear leucocytes, with here and there a small amount of fibrin. In numerous areas the alveolar walls were broken down, with formation of small abscesses, filled with polymorphonuclear leucocytes, many of which showed marked nuclear fragmentation. Sec- tions stained with eosin and methylene blue showed very 290 PERKINS. large numbers of bacteria, of which a few were large and stout with rounded ends, but most of which were slender, some much longer than others, many showing a sHght curve. They stained irregularly with methylene blue, and sharply by Gram's method. They occurred in greatest numbers in and about the abscesses, though present everywhere in the consolidated areas. Case III. — Scirrhus carcinoma of the pylorus, gastroen- terostomy, with subsequent foecal fibrino-purulent peritonitis. B. pyocyaneus was present in pure culture in the peritoneal exudate. Clinical Brief. — H. S. Male, 30 yrs. old. Admitted to Lakeside Hospital for gastro-enterostomy operation for car- cinoma of pylorus. Post-operative peritonitis developed, with death as a result. Anatomical Diagnosis. — Scirrhus carcinoma of stomach with metastases in liver and in mesenteric lymph glands, fibrino-purulent peritonitis, sero-fibrinous pleurisy, and peri- carditis. Chronic adhesive apical pleurisy with chronic fibroid apical tuberculosis and apical tubercular pneumonia; congestion and emphysema of lungs with unresolved pneu- monia ; congestion of liver, spleen, and kidneys ; fatty liver, interstitial splenitis, cloudy swelling of kidneys. Healed gastro-enterostomy and recent gastro-enterostomy, with gas- tro-peritoneal fistula. The organs showed the changes noted in the anatomical diagnosis. The peritoneum contained 1,500 c.c. of yellow fluid, containing many fibrin flakes, and hfiving a markedly fecal odor. Coverslips and cultures from the pneumonic areas in the lungs and from the pleura and pericardium, showed diplo- coccus lanceolatus in pure culture. Coverslips from the peritoneal cavity showed large, stout capsulated bacilli and small, slender bacilli, staining irregularly with methylene blue. Cultures showed a short, slender bacillus, actively motile, which decolorized by Gram. It peptonized milk with acid reaction and coagulation, and bore a close resem- blance to the organisms previously described. B. PYOCYANEUS. 29 1 Case IV. — Acute orchitis in typhoid ( ?) fever with the presence of B. pyocyaneus in pure culture in the testicle. Clinical Brief. — Male, white, 27 years old. Admitted to the City Hospital with a mild case of supposed typhoid fever. A few days after admission, one testicle became large and tender, and the patient was transferred to the ser- vice of Dr. C. A. Hamann, who did an orchidectomy. There was no actual purulent exudate, but the tissue was markedly softened. Pathological Brief. — Cultures were made and studied at the Pathological Laboratory of Western Reserve University, and showed a pure growth of an actively motile, slender bacillus which decolorized by Gram's method. Litmus milk was rapidly peptonized without change in reaction, and the general cultural characteristics closely resembled those of the organisms described in the previous cases. Microscop- ical sections of the testicle and epididymis showed well- marked purulent infiltration, with the presence of large numbers of eosinophiles and a few mast-cells. Case V. — Pustular eruption of an ecthymatous type on the inner surfaces of the thighs, with the presence of B. pyocyaneus in pure culture. Negro, 40 years old. Admitted to the City Hospital in 1898, with enlarged glands in the neck, groins, and axillae, and a pustular ecthymatous eruption on the inner surfaces of the thighs. A clinical diagnosis of glanders was made and cultures were taken from the pustular lesions. These cultures were studied at the Pathological Laboratory of West- ern Reserve University; they showed a pure culture of a short bacillus, rather slender, and often occurring in thread- like forms, — by Gram's method, and actively motile. The growths on the various media closely resembled those of the organisms in the previous cases. Diagnosis B. pyocyaneus. Case VI. — Abscesses about seventh rib of the left side and in the left elbow joint, with the presence of B. pyocyaneus in pure culture in both abscesses. Clinical Brief. — Male, 30 years. Admitted to the medi- cal wards of Lakeside Hospital, Oct. 10, 1899. He had just 292 PERKINS. had pneumonia, and physical examination showed the pres- ence of fluid in his chest. Shortly after admission a fluctu- ating area was noted in the left axilla about the seventh rib, and he soon developed an acute arthritis of the left elbow joint. He was transferred to the surgical service for opera- tion, and both abscesses were opened. They contained large quantities of greenish, foul-smelling pus, and no con- nection between the abscess about the rib, and the pleural cavity could be found. The patient's temperature continued at about 102°, with a pulse of 112, and upward, and he died a few days later. Autopsy was refused. Pathological Brief. — Coverslips from both abscesses showed short, slender bacilli. Cultures showed a short, slender bacillus, motile, decolorizing by Gram's method and closely resembling the organisms described in the previous cases. Resection of a rib in the floor of the abscess showed that there was no empyema, but a sub-acute fibrinous pleu- risy. The mode of examination forced upon us made cult- ures from the pleural cavity impossible. Case VII. — Acute diarrhoea. Broncho-pneumonia' Gen- eral congestion of organs. Presence of B. pyocyaneus .in small numbers in all the organs. Clinical Brief. — Male, black, 4 months old. The patient had had diarrhoea for a week, with numerous green dejec- tions. He also had frequent spasms, almost amounting to convulsions, since the onset of the disease. The abdomen was tender and the peristalsis was violent. The child became rapidly emaciated and died of asthenia. Pathological Brief. — Anatomical Diagnosis. Congestion, atalectasis, broncho-pneumonia of lungs; congestion and fatty degeneration of liver ; congestion of spleen and kidneys and of gastro-intestinal tract. Cultures from the various organs showed small numbers of colonies of a bacillus which was motile, decolorized by Gram's method, and otherwise resembled the bacilli found in previous cases. Unfortunately no cultures were made from the intestines. The organs showed cloudy swelling and congestion, and B. PYOCYANEUS. 293 the mucosa of the stomach and intestines showed well marked congestion, but no ulcerations. Besides these seven cases, B. pyocyaneus was found twice where its association with existent pathological lesions is rather suggested than definitely proven. Case VIII. — OEdema and congestion of the lungs, with the presence of B. pyocyaneus in large numbers in pure culture. Congestion and cloudy swelling of the liver and kidneys; lymphoid hyperplasia in the gastro-intestinai tract. Miliary abscesses of the brain, with the presence of a bacillus similar in morphology and staining reactions to that found in the lung, but not cultivated. Clinical Brief. — Boy two months old. Admitted to the children's ward of Lakeside Hospital. He had had diarrhoea for the last three weeks with marked progressive emaciation, and died under the clinical diagnosis of malnutrition. Pathological Brief. — Anatomical diagnosis : Miliary ab- scesses of the brain. Congestion, fatty degeneration, and cloudy swelling of the liver and kidneys ; emphysema, cedem'a, and congestion of the lungs. Cultures from all the organs except the lung were negative. Cultures from the lung showed numerous colonies of short, slender bacilli, actively motile, and decolorizing by Gram's method. The growth on culture media and the chemical characteristics were similar to those of the organisms in the previously described cases. Sections from the brain about the abscess cavities showed necrosis of tissue with very slight leucocytic infiltration. At the edge of the necrotic area, and throughout the vessels on the section, were large numbers of bacteria. These were of two sizes, one small, slender, and staining irregularly, the other large and stout, resembling B. aerogenes capsulatus. Both these bacilli stained well with methylene blue, but sec- tions stained byWeigert's method show no bacteria whatever, thus excluding B. aerogenes capsulatus, which stains by Gram's method. The fact that the organisms decolorize by Gram, together with consideration of their morphology, suggests their identity 294 PERKINS. either with B. mucosus or B. pyocyaneus brain abscess. It is not improbable that the organisms in the brain and in the lung are one and the same. Case IX. — Burns of the third degree, with congestion and cloudy swelling of the organ's, and the presence of B. pyocy- aneus in the heart's blood. Clinical Brief. — F. F. Male, five years. Entered the Lakeside Hospital Aug. i, 1899, with extensive burns of the third degree, and died the next day. Pathological Brief. — Anatomical diagnosis : Infected burns of third degree involving about two-thirds of the pkin surface. General congestion of organs. Cloudy swelling of liver and kidneys. Chronic interstitial splenitis and hyperplasia of mesenteric lymph glands. The organs showed nothing but cloudy swelling. Coverslips from the pus on the surface of the body showed cocci, singly and in chains, and numerous short, slender bacilli. Cultures from the organs were negative, with the exception of the plate from the heart's blood, which showed a few colonies of small, slender bacilli. Cultivation of these showed a bacillus which was actively motile and decolorized by Gram's method. The cultural characteristics closely resembled those of the organisms in the eight previous cases. The organisms from these nine cases were carefully studied, as well in their morphological and cultural characteristics as in their relations to one another and to stock cultures of B. pyocyaneus obtained from the Johns Hopkins Hospital. The variations from this culture and from each other were of degree rather than kind, but some general points seem worthy of note. The irregular staining with methylene blue, noted by Jordan,^'' was observed in every case, both in sections from hardened tissues, and in coverslips from exudates or artificial media, and in one case, in which the organisms were of un- usual length, was so marked as to bear a close resemblance to the Klebs-Loefifler bacillus. In the reactions of the various organisms to Gram's stain, B. PYOCYANEUS. 295 my results did not coincide with those of Jordan '^* and Ru- zicka,^ who note complete decolorization in every case. Of the ten varieties studied in this present series, seven decolor- ized uniformly and regularly when treated by this method, but three retained their color at every trial. Of these three only one was associated with tissue lesions studied micro- scopically; in this case (Case II.), as noted above, Weigert's fibrin stain showed the bacilli in undiminished numbers in the tissues, while in those cases in which the organisms decolor- ized by Gram's method, no bacteria could be found in sections treated according to Weigert. The cultures on agar and on potato varied in their color production and in the profusion of their growth, but in every case the green color, the staining of the potato, and the odor were characteristic of B. pyocyaneus. Ernst's chameleon phenomenon was not seen in any of the cases. All the varieties liquefied blood serum and peptonized lit- mus milk, though only three cases (III., V., and VI.) caused acid reaction on coagulation of the casein, the media in the other seven cases remaining blue or violet and fluid until the process was complete and the litmus discharged. The fil- trates from milk and serum cultures, kept ten days or more at 37° C, gave a sharp, positive reaction to the Biuret test for peptone. The pigment formation was studied in six of the cases with the result that all of them fell within variety a of Jordan's classification ; namely, that variety which produces both pyocyanin and a fluorescent pigment. Cultures from the other four cases were less carefully studied, but one of them (Case VII.) was so much lighter in color than the rest, even on old cultures on agar, that it is probable it would have been found to coincide with variety 7, fluorescigenic only. Summary. In nine cases of pyocyaneus infection observed in the last three years, two found their portal of entry in the uterine mucosa, in puerperal septic endometritis, and gave rise sec- ondarily, in Case I. to cerebro-spinal meningitis and epen- 296 PERKINS. dymitis, and in Case II. to broncho-pneumonia with abscess formation. Case III. adds a third to the previously reported cases of peritonitis, and was secondary to a ruptured gastro- enterostomy for carcinoma of the pylorus. Case IV. records a lesion hitherto undescribed in connection with B. pyocy- aneus, an acute orchitis in late typhoid (?) fever. Case V. adds one to the cases already on record of B. pyocyaneus in ' pustular skin lesions. Case VI. brings out the third recorded case of infection of the synovial membranes, the organism having been found in the elbow joint, as well as in a peri- costal abscess. Case VII. shows the fourth instance of broncho-pneumonia, as well as an additional case of general bacterisemia in children. In Case VIII. the organism was present in the lungs, with oedema and congestion, and organ- isms resembling closely in morphology and staining reactions those cultivated were found in association with miliary ab- scesses of the brain. Case IX. is one of wide-spread super- ficial burns, with infection of the burned areas, and the presence of the organism in pure culture in the heart's blood. B. PYOCYANEUS. 297 REFERENCES. 1. Phila. Med. Journ., Sept. 17, 1898. Journ. Exp. Med., 189S, Vol. III., No. 6. N.Y. Med. Journ., Sept. 22, 19CX). Journ. Am. Med. Ass., July 31, 1897. 2. Journ. Boston Soc. Med. Sci., Vol. V., p. 385. 3. Arch. f. Derm. u. Syph., Vol. L., p. 71. (Ref. Centr. f. Allg. Path. u. path. Anat.) 1890. 4. Monatschr. f. Ohrenheilk., No. 7, p. 887. 5. Centr. f. Allg. path. Anat, Vol. X., p. 651. 6. Zeitschr. f. Ohrenheilk., i89i,p. 44. 7. Centr. f. Allg. Path. u. path. Anat., Vol. XIV., p. 26. 8. Centr. f. Allg. Path. u. path. Anat., Vol. XXVIII., p. 729. 9. Orvosi Hetilap., 1889, No. 50. (Ref. Centr. f. Allg. Path. u. path. Anat, 1890.) 10. Centr. f. Allg. Path. u. path. Anat., Vol. XXIII., p. 967. 11. Arch, of Pediatrics, Vol. XIV., p. 11. 12. Zeitsch. f. Hygiene, Vol. XXII., p. 140. 13. Phila. Med. Journ., Vol. IV., p. 928. 14. Zeitsch. f. Hyg., Vol. XXL, p. 281. 15. Ann. de I'Inst. Pasteur, Vol. X., p. 52&. 16. Centr. f. Allg. Path. u. path. Anat, Vol. XXV., p. 116. 17. Arch. f. Derm. u. Syph., Vol. LII., p. 349. 18. Johns Hopkins Hosp. Bull, January, 1901. 19. Brit. Med. Journ., September, 1894. 20. Am. Journ. Med. Sci., October, 189S. 21. Am. Journ. Med. Sci., 1898, p. 252. 22. Journ. Exp. Med., 1900-1901. 23. Jour, of Exp. Med., 1899. 24. Op. cit. 25. Cited by Jordan. Op. cit Extracted from The American Journal of the Medical Sciences, February, 1903. THE PATHOLOGY OF LABIAL AND NASAL HERPES AND OF HERPES OF THE BODY OCCURRING IN ACUTE CROUPOUS PNEUMONIA, AND THEIR RELATION TO THE SO-CALLED HERPES ZOSTER.' By William Travis Howard, Jr., M.D., PROFESSOK OF PATHOLOGY, WESTERN RESERVE UNIVERSITY. (From the Pathological Laboratory of the Lakeside Hospital, Cleveland.) Our knowledge of herpes zoster was very much extended by the recent contribution of Head and Carpenter.^ These authors showed, among other things: 1. That herpes zoster (in 18 cases, at least) is associated with destructive, and usually inflammatory, changes in the sensory ganglia (posterior root or Gasserian) corresponding to the nerve supply of the part affected with herpes. 2. That in recent cases (before the eleventh day) degenerative changes are not demonstrable in either the peripheral nerves, the posterior root fibres, or the central nervous system. 3. That in cases examined after the lapse of this period degenerative changes are present in these structures and apparently correspond to the affected areas in the ganglia. Their work was based upon the study of 21 cases in all stages of the eruption. They state that before their own work there were only two well-reported autopsies on eases of herpes ophthalmicus (Wys and Sattler) and five satisfactory reports on zoster of the trunk (Lesser, Chandelux and Dubler). Von Barensprung's oft-quoted case is classified as unsatis- factory, because, though he noted the occurrence of hemorrhage into a ganglion, he failed to mention to which of the three ganglia removed he alluded. While this valuable communication has placed our knowledge of herpes zoster on a substantial basis, it is clear that more light is needed upon the etiological factors concerned, as well as upon the identification aad classification of the various types of herpes. 1 Read at the Seventeenth Annual Meeting of the Association of American Physicians, held in Washington, D. C, 1902. 2 Brain, Autumn, 1900. ^ HOWARD: HBEPES IN CEOUPODS PNEUMONIA. In order to clearly understand our present knowledge on the subject, the following summary of the main facts may be helpful. There are a number of clinical and apparently several anatomical types of herpes : herpes zoster involving the neck, trunk, and extremities (fol- lowing the distribution of the spinal posterior root ganglia) ; herpes ophthalmicus (following the distribution of the first division of the fifth or trigeminal nerve, Gasserian ganglion) ; herpes facialis, herpes labialis et nasalis (probably following the distribution of the second and third divisions of the trigeminal nerve, but unsupported at this time by anatomical observations), and herpes genitalis (commonly believed to be due to local irritation and not known to be associated with any lesion of the nervous system). The herpes facialis, labialis, nasalis, and genitalis are classed by Hartzell' as herpes simplex, in contradistinction to herpes zoster of the neck, trunk, extremities, and ophthalmic area (herpes zoster oph- thalmicus). These latter forms of herpes have been shown to be iden- tical in that they are associated with the same lesions of the nervous system (especially those of the sensory ganglion), and the changes in the skin are believed to be the same. It seems further to have been conclusively demonstrated that the lesions of the skin are secondary to the primary changes in the corre- sponding sensory ganglia, which in turn may be brought about by a variety of causes. The cases, however, may be classified into the primary or spontaneous herpes zoster, in which the affection is appar- ently primary and not due to any evident preceding infection or injury, and the secondary form, in which there is more or less clear evidence that the disease occurs as the sequel or complication of some antecedent affection, as pneumonia, cerebro-spinal meningitis, etc., or to injury. Our knowledge of the changes in the skin in the various clinical forms of herpes is far from satisfactory. Not until a sufficient number of careful observations upon the histological changes in all the stages of each variety of herpes is available will we be in a position to com- pare critically the processes, and, perhaps, distinguish clearly the different forms. Suffice it to say for the present, that after comparing the histological appearances described by various authors of the changes in herpes zoster, herpes labialis and genitalis (all available), with some of my own cases herewith reported, and making due allow- ances for the different stages of the processes studied, I am not able to find any constant distinguishing characters by which the skin lesions of the various clinical forms of herpes may be separated. I have recently had the opf)ortuntty of studying two cases which 1 Reference Handbook of the Medical Sciences, 1902, vol. iv. HOWARD: HERPES IN CROUPOUS PNEUMONIA. 3 throw some light upon several points of importance in the pathology of herpes. The first case occurred in the private practice of my friend Dr. Edward F. Gushing, who recognized its interest and importance and obtained the autopsy. I take this opportunity of expressing my thanks to Dr. Gushing for the autopsy and for the use of the clinical history of the case. This case illustrates the causal relation between acute infections and herpes zoster, and emphasizes the identity of the lesions of the skin and nervous system in primary and secondary herpes of the trunk. Gase I. Summary. Bronchitis, acute croupous pneumonia of both lungs, herpes zoster of the left side, distributed on the back and side fsixth dorsal), and the abdomen (eleventh dorsal) ; congestion and hem- orrhage into the capsular and interstitial tissue at one side of the eleventh dorsal ganglion, with slight cellular infiltration and destruc- tion of a few ganglion cells ; amylaceous and hyaloid bodies in another portion of the same ganglion. Slight chronic and acute interstitial nephritis, acute parenchymatous and interstitial hepatitis, acute splenic tumor, old pericardial adhesions, hypertrophy and dilatation of the heart ; fibromyoma uteri. Clinical History. Miss T., aged sixty-three years, had not been well for about one year. On October 10, 1901, she was taken sick with a cold in the head, which a few days later was followed by bronchitis. She. then developed pneumonia of the lower lobe of the left lung. Later the whole of the left lung became consolidated. On the morning of October 24th there were the physical signs of consolidation of the lower lobe of the right lung. Coincidently with the consolidation of the lower lobe of the left lung there developed well-marked herpes zoster in the mid-dorsal region of the left side, extending from the middle line in the back to about the anterior axillary line, correspond- ing to Head's sixth dorsal area. During the last two days of life a similar eruption occurred on the abdomen, in the left umbilical and iliac regions (eleventh dorsal area ; see autopsy protocol). Albumin was present in the urine. Death took place at 2.30 p.m., October 24, 1901. Autopsy six hours later. The autopsy was done hurriedly and under difficulties, at the home of the patient. The body measured 160 cm. long, was sparely built, except over the abdomen and lower portion of the back, where there was a considerable amount of fat. The body was cold, rigor mortis moderate. There were a few reddish discolored areas on the outer aspects of the legs. On the left side of the abdomen, situated in the outer and lower portion of the left umbilical and the upper portion of the left iliac regions there was an area, the size of the palm of the hand, studded with scattered pale-red or gray areas from 3 to 6 or 10 mm. in diameter. Some of these areas were elevated and covered with dry grayish-red material, while others showed a distinct loss of epidermis. No typical vesicles containing fluid were to be found, but the areas had the appearance of dried vesicles. In distribu- tion this area corresponded with Head's eleventh dorsal area. A similar but narrow area extended from the middle line in the mid- dorsal region as far as the anterior axillary line under the nipple. 4 HOWABD: HEKPBS IN CROtJPOUS PNEUMONIA. corresponding to Head's sixth dorsal area. The areas here were larger and the process appeared older than that on the abdomen. No other lesions were found upon the surface of the body. There was no facial herpes. Chest. The chest was fairly well shaped, the muscles and fat rather wasted. Both pleurae were free from adhesions, and the pleural cavities con- tained no fluid. The visceral pleurae over considerable areas were covered with a thin layer of fibrin, and were red and lustreless. Right Lang. The apex was free from puckering and thickening. The upper lobe on section was congested and markedly (edematous. The whole of the middle and a large part of the lower lobe were con- solidated, airless, and of a grayish-red color, and granular appearance. The mucous membrane of the bronchi was congested. Left Lung. The whole lung was consolidated, airless, of a gray or grayish-red color, and granular appearance. The bronchial mucosa was congested. The bronchial glands were pigmented, but free from tubercle. Heart. The pericardium was adherent to the diaphragm over a large area. Firm fibrous adhesions obliterated the pericardial sac. The walls of both ventricles were thickened. The valves, the coronary arteries and veins, as well as the aorta and large arteries, were normal. The myocardium was pale, otherwise normal ; the heart weighed 400 grammes. Abdomen. The muscles and fat of the abdominal wall were of ordi- nary appearance. The left lumbar region was prominent. The stomach was distended with gas, the large and small intestines contracted. Liver. The liver was of ordinary appearance and size. The edges were rounded, the surfaces smooth, consistency not increased. On section the lobules are visible, the surface pale. The gall-bladder and bile-ducts were negative. Spleen. The spleen was twice the ordinary size. On section it was of a grayish-red color ; the trabeculse and Malpighian bodies visible ; consistency decreased. Kidneys. The kidneys were of about the same size and appearance. The capsules were adherent in places, the surfaces were somewhat granular and marked by a few old scars. On section the consistency was somewhat increased, the surfaces pale, the cortices thinner than ordinary. The arteries, pelvis, ureters, bladder, and adrenals appeared normal, as did the stomach and intestines. The ovaries were small and tough to the touch. The uterus was considerably enlarged ; the body was the seat of a globular fibroma the size of an orange. Nervous System. The brain could not be examined, and the examina- tion was confined to the lower portion of the spinal cord ; from the tenth dorsal segment downward with the cord the eleventh and twelfth dorsal and first lumbar posterior ganglia of both sides were removed. It was not possible to go higher in the time available for the examination. There was an excess of cerebro-spinal fluid, which, however, was quite clear. The meninges were normal in appearance. The cord showed no lesions on section. The ganglia showed no macroscopic lesions, and no hemorrhages were found about the posterior roots or the nerves going from the ganglia. HOWARD: HERPES IN CROUPOUS PNEUMONIA. 5 Portions of the various organs, including the cord and the ganglia, were hardened in Orth's fluid. The ganglia were transferred to 96 per cent, alcohol after a few hours. Portions of various organs, including the skin, were also hardened in Zenker's fluid. Bacteriological Examination. Cover-slip preparations from the lungs and pleurae showed lanceolate diplococci in capsules. Cultures on blood serum and on glycerin agar from the lungs, heart's blood, liver, spleen, kidneys, and cerebro-spinal fluid remained sterile. Histological Examination. Lungs. Sections from both lungs showed well-marked croupous pneumonia, with lanceolate diplococci in the alveolar exudate. Liver. There was well-marked fatty degeneration of the liver cells, congestion of the capillaries, with marked round-cell infiltration about many of the portal systems. No focal necrosis and no bacteria were found in suitably stained sections. Kidneys. There were scattered areas of chronic glomerulitis, with thickening of the capillaries and of Bowman's capsule, with but little atrophy of the neighboring tubules. The epithelial cells of the con- voluted tubules showed cloudy swelling. With no apparent relation with changes in the glomeruli or tubules, there were numerous areas of round-cell infiltration ; most of these cells were of the plasma-cell type, but some were lymphocytes. There was no necrosis of the renal epithelium and no abscess formation. No bacteria were to be found. Skin. Sections were made through both areas of herpes, and, with the exception that the lesions were more extensive in the older lesion (sixth dorsal region), the changes were the same. There were few unbroken vesicles, but in most places the surface of the lesion was covered with amorphous and fibrillated fibrin containing a larger or smaller number of desquamated, swollen, and often disintegrating and necrotic epithelial cells. Some of these cells had undergone coagula- tion, and others liquefactive necrosis. In some places the cells of the deeper layer of the epidermis were markedly swollen, vacuolated, and .evidently dropsical. In the fibrin and among the epithelial cells on the surface there were variable, often great, numbers of polymorphonu- clear neutrophilic leucocytes. In some places there were cavities formed by the elevation of the superficial or subcorneal layer of epithelium. These were of considerable size in places, but were not numerous. Other small vesicles occurred in the exudate, their walls being formed by necrotic cells. Still other vesicles were between the epidermis and the underlying papillary layer. These were always small. In many places the papillary layer was exposed by the total destruction of the epidermis, and was covered by the exudation. In the papillae and in the upper layer of the corium the veins and capillaries were markedly dilated, and in places there was hemorrhage, either into the tissue or upon the surface. The papillae and the upper layers of the corium showed other intense changes, coagulation necrosis, with hyaline and fibrinoid changes, nuclear fragmentation, and infiltration, with large numbers of poly- morphonuclear neutrophiles, lymphocytes, and some plasma cells. Here and there a few eosinophiles were to be seen. This cellular infil- tration was both diffuse and along the bloodvessels. Many of the sweat glands and hair follicles were swollen and distended with leuco- cytie and serous exudate. The epithelial cells were commonly swollen, 6 HOWARD: HEKPES IN CROUPOUS PNEUMONIA. granular, and often desquamated and necrotic. About many of these glands there was well-marked leucocytic infiltration, a few of the cells being eosinophiles. Scattered collections of leucocytes were also seen in the deeper layers of the skin. No bacteria were to be found in any of the lesions in sections stained by Weigert's fibrin method or with eosin and methylene blue. The whole appearance of the lesion points to rapidly necrotic and exudative processes. Posterior Root Ganglia. Sections were cut of the eleventh and twelfth dorsal and first lumbar ganglia of both sides and stained by Nissl's, Weigert's, and Marchi's methods, as well as with methylene blue and eosin, and hfematoxylin and eosin. Sections from all the ganglia studied, except the eleventh dorsal of the left side, were entirely normal. In sections of the latter ganglion changes were found in two places : (1) at one side of the ganglion about midway of its length, and (2) at the peripheral end. 1. In the sections cut transversely through about the middle of the ganglion at one side there was considerable hemorrhage, with pigmen- tation, into the capsule. In some of the sections corresponding to the area of capsular hemorrhage there was just beneath the capsule marked congestion of the capillaries and interstitial hemorrhage between the ganglion cells. In several sections small areas of cellular infiltration about destroyed ganglion cells were found. The ganglion cells else- where stained well and appeared normal. 2. At the peripheral end of the ganglion there were a number of amylaceous and hyaloid bodies, both singly and in groups, of the size and shape of ganglion cells, and lying in spaces similar to ganglion cell spaces. Many of these bodies were finely granular, some of the granules staining pink with eosin, and others blue with toluidin blue. Many bodies stained diffusely pink with eosin, and some diffusely blue with blue dyes. In some a central nucleus was apparent, and many showed a concentric arrangement. In a few places hyaline masses were made out in lymph spaces and in blood capillaries. Hyaloid and, amylaceous bodies were also found in unmistakable lymph and blood- vessels. The question whether or not some at least of these bodies represented degenerated nerve cells is difficult to answer. From careful study I am convinced that some of them were formed and lie in nerve cell spaces. No lesions of the peripheral nerve, the posterior nerve roots, or in the spinal cord, could be made out. The interpretation of the pathological changes of the above case appear to be as follows : Acute croupous pneumonia following bronchitis, with the ordinary parenchymatous changes of the liver and kidneys common in such cases, with the addition of acute interstitial hepatitis and nephritis ; an her- petic eruption of certain cutaneous areas, the spinal root ganglion corresponding to one of which areas was shown to be the seat of definite lesions (congestion and hemorrhage of both the capsule and a small area of the interstitial tissue of the ganglion, with cellular infiltration about a few degenerated ganglion cells, with hyaloid and amylaceous bodies, in another portion of the ganglion). HOWARD: HERPES IN CROUPOUS PNEUMONIA. 7 It seems fair, in the light of our present knowledge, to attribute these various lesions to the toxins of the pneumococcus. The failure to find degenerative changes in the cord, the posterior roots, and the peripheral nerve is in accord with observations of Head and Carpenter, who did not find changes in these organs in a case of herpes dying on the eighth day. They were present in a case dying on the thirteenth day. Including the 5 cases of Lesser, Chaudelux, and Dubler with the 16 cases of Head and Carpenter, there are 21 well-studied cases of herpes zoster of the trunk in which definite destructive lesions have been demonstrated in the posterior root ganglia, corresponding to the dis- tribution of the herpetic eruption. Two of the 21 cases of Head and Carpenter involved the head, in 1 case the ganglion was not removed, and in 2 others, in which death took place in 139 and 240 days, respec- tively, no lesions of the nervous system were found. In all 9 of the acute cases (three to sixteen days after the appearance of the eruption), definite and unmistakable ganglionic lesions were found. Of these cases 2 had had acute infections (cystitis, with retention of urine in paraplegia, and acute bronchitis and tubercular pneumonia). In the 12 chronic cases in which death occurred in from 57 to 790 days after the herpes the sensory ganglia corresponding to the distribution of the eruption showed well-marked chronic changes. Our next case shows that the changes in the skin and Gasserian ganglion in herpes of the lips and nose in pneumonia are identical with those occurring in herpes ophthalmicus, and in the skin and posterior root ganglia in both primary and secondary herpes zoster of the trunk. It is also, curiously enough, apparently the first case of herpes of this region in which changes have been sought for in the Gasserian ganglion. Case II. Summary. Acute croupous pneumonia in a bartender, aged forty-one years. Death on the sixth day of the disease. Two days before death well-marked herpes of the upper lip and the nose, being much more extensive on the left side. Autopsy showed acute croupous pneumonia (gray hepatization) involving the whole of the right lung, with fibrinopurulent pleurisy of the left side, oedema and congestion of the left lung. Chronic fibrous obliterative pleurisy of the right side. Herpes of the upper lip and nose, most marked on the left side. Congestion of the veins about the origins of the superior maxillary branches of both Gasserian ganglia. Hemorrhage into the capsule and tissue, with interstitial cellular infil- tration and compression and degeneration of the ganglion cells near the origin of the superior maxillary branch of the left Gasserian gan- glion. A few small areas of cellular infiltration in the same part of the right Gasserian ganglion. Marked congestion of the veins of the neck and brain and of the cerebral sinuses. Pneumococcus in the right lung and left pleura. 8 HOWARD: HEKPES IN CROUPOUS PNEUMONIA. Clinical History. For permission to use the clinical history of the following case I am indebted to the courtesy of my colleague, Dr. J. H. Lowman, in whose service in the Lakeside Hospital the patient was admitted February 6, 1902, complaining of pain in the right side of the chest, cough, and fever. The patient was a male, aged forty-one years, white, and a barkeeper by occupation. His family history was without present interest. Personal History. He did not recall having ever been sick before, and denied having had diphtheria, scarlatina, rheumatism, pneumonia, pleurisy, syphilis, gonorrhoea, tvphoid and malarial fevers. Some time ago he injured his right temple by a fall, but the wound healed promptly and was followed by no clinical symptoms. He drank beer and whiskey to excess at times, and also used tobacco. He had been a barkeeper for twelve years. Present Illness. He got his feet wet on the night of February 8th, and awoke the next morning with severe pain in the right side of the chest, but did not have a chill. On admission he had the physical signs of consolidation of the lower portion of the right lung. The left lung was clear. The spleen was enlarged. Examination of the other organs was negative. The leuco- cyte count on February 8th was 19,000, and on February 9th 13,000 per cubic millimetre. On February 9th, the fourth day of the disease, there appeared a well-marked vesicular eruption over the upper lip and the outer surface, and in the vestibule of the left side of the nose. This eruption on the lip and the outside of the nose extended some- what to the right side of the median line. The sputum contained lanceolate diplococci. By February 11th the whole right lung was consolidated, and numerous moist rales were to be heard over the left lung. The patient was delirious. Death occurred suddenly at 1.15 p.m., February 11, 1902, on the fifth day of the pneumonia and the begin- ning of the third day of the herpetic eruption. Autopsy One Hour after Death. Body warm, no rigor mortis. Body 165 centimetres long, well built, and well nourished. There were no marks, scars, or wounds on the body. On the upper lip, involving both mucous and skin surfaces, there were a number of dried crusts surmounting slightly elevated areas, with red margins. These areas varied from 2 to 5 millimetres in diameter. Similar papules and dried vesicles were present on the skin of the vestibule and external surface of the left side of the nose. Altogether there were from fifteen to twenty such areas. Several similar lesions were also found to the right side of the median line of the lip and nose. No other herpetic erup- tions were to be found on the body. Nervous System. The veins of the dura and pia arachnoid and all the sinuses were markedly congested ; the pia arachnoid was smooth and glistening. The structures at the base of the brain were normal. The brain on section showed congestion, but was otherwise normal in appearance. The Oasserian Ganglia. The veins near the ganglion were engorged with blood. The ganglion was of ordinary size. On the upper surface just below the entrance of the sensory root and between and about the exit of the ophthalmic and superior maxillary branches the veins were congested and stood out prominently. The same appearance was seen on the posterior surface, but the congestion was less well marked. HOWAED: HERPES IN CEOUPOTJS PNEUMONIA. 9 The Left Oanglion. The vessels about the ganglion were engorged, as were the veins about and between the origin of the ophthalmic and superior maxillary branches of the ganglion. About the root of the superior maxillary branch this was very marked, and there was also hemorrhage into the capsule at this part on the posterior aspect of the ganglion. The fifth nerve and the branches from both ganglia ap- peared normal. The cord and the posterior root ganglia were not removed. Chest. The chest was well formed ; the sternum, ribs, and costal cartilages were normal. The right lung was bound down by old but thin adhesions ; the pleural cavity was obliterated. The right lung was consolidated throughout. On section it was firm and airless, of a grayish color, and quite granular in appearance. The mucosa of the bronchi was congested, otherwise normal. The left pleural cavity contained about 100 CO. of seropurulent fluid. The surface of the lower lobe was covered with a thin fibrinous exudate. The lung crepitated throughout. On section the lung was markedly congested and cedematous, but was free from consolidation. The description of the other organs is without present interest, and will, therefore, be omitted. Cover-slips from the right lung and left pleura showed encapsulated lanceolate diplococci. Plate cultures from both lungs, the left pleura, and heart's blood and pericardium showed pneumococcus in pure culture. Similar cul- tures from the liver, spleen, kidneys, and gall-bladder remained sterile. Histological Examination. Lungs. Sections of the consolidated area showed well-marked croupous pneumonia, with lanceolate diplo- cocci in the exudate. Sections of the other organs, with the exception of the skin of the lip and nose and of the Gasserian ganglia, showed nothing of present interest. Shin of Lip and Nose. Bits of tissue were cut out and hardened in formalin and 95 per cent, alcohol, and sections were stained in carbol- toluidin blue and eosin, hsematoxylin and eosin, and by Weigert's fibrin method. The changes found were practically identical with those from the herpetic areas of Case I. As in Case I., few well-developed vesicles remained, the surface of the lesions, as a rule, being covered with a crust of amorphous and fibrillated fibrin and fibrinoid material containing desquamated and often necrotic epithelial cells, polymorphonuclear and mononuclear leucocytes, and cellular and nuclear fragments. Here and there rather small vesicles could be made out. As in Case I., some were elongated cavities in the exudate, or in the superficial layer of the skin, just beneath the horny layer, while others were between the underlying inflamed papillae and the epidermis. Most of the latter cavities were small, and their walls were formed of necrotic epithelium and their floors of the naked papilla;. In some places ballooning of the cells of the epidermis could be made out. In many places the margin between the necrotic and the unchanged epidermis was rather clear cut, but in some places near the necrotic portions the small vesicles could be made out in the neighboring otherwise normal epidermis. In most places at the site of the lesion the whole thickness of the epidermis was necrotic. 2 10 HOWARD: HERPES IS CROUPOUS PNEUMONIA. Beneath the fibrinous layer there were, in all sections examined, changes in the epithelium, marked coagulation and liquefactive necro- sis, and hyaline transformation of the cells. The papillary layer and outer part of the corium were similarly affected. There was, in many places, marked nuclear fragmentation. The epithelium of the sweat glands and some of the hair follicles, when present, showed changes — swelling and liquefaction, and often a reticular appearance. Some cells were markedly swollen, many cells showed coagulation necrosis. Many nuclear fragments were to be seen. The bloodvessels, especially the cap- illaries and small veins of the papillse and corium, were markedly dil- ated. Some were crowded with leucocytes. In some of the superficial vessels there were thrombi composed of pink-staining material. The papillae and the superficial portions of the corium were diffusely infil- trated with great numbers of polymorphonuclear, neutrophilic, and large and small mononuclear leucocytes, and some plasma cells. The leucocytic infiltration could also be traced along some of the small vessels into the deeper layers of the corium. About some of the super- ficial veins and capillaries often large numbers of red blood cells had escaped into the tissue. Some of the lymph spaces were dilated. The lesions here, as in the herpes of Case I., were both degenerative and exudative, and involved both epidermis and corium. No inflammatory changes were found in nerves in the skin. No bacteria were found in the lesions of the skin, but in the superficial part of the fibrinous exudate, in sections taken from the vestibule of the nose, diplococci and some bacilli, morphologically like B. mucosus capsulatus, were found. Gasserian Ganglia. The ganglia with the fifth nerve roots and the origins of the ganglionic branches were hardened in formalin (10 per cent, solution). The two ganglia were mounted in celloidin, side by side, on the same block and cut in serial sections, every other section being stained and mounted in order. Many of the alternate sections were also stained. In mounting on the block the left ganglion was at a somewhat higher level than the right, so in the parallel sections as mounted for histo- logical study the two gauglia were not identical in structure, but as the series was followed to the end the right ganglion was found to be simi- lar to the left in structural detail. Left Gasserian Ganglion. In sections 1 to 8 there were no ganglion cells near the superior maxillary branch. The veins about this branch were enormously dilated and filled with red blood cells. In seme places in the largest vein there were collections of leucocytes and larger and smaller granular masses, suggesting fibrin. No changes were found in the opposite end of the ganglion. About the dilated veins there was heniorrhage into the pia or capsule of theganglion. In section 9 a small oval bit of ganglion tissue, containing ganglion cells, became apparent near the superior maxillary branch. In sections 9, 10, 11, and 12, corresponding to this, there" were capsular and subcapsular hemorrhage, with compression of the superficial ganglion cells. In sec- tions 15 to 35 the small strip of ganglion tissue in relation with the superior maxillary branch became very much larger, and formed an elongated strip of tissue traversing the thin portion of the ganglion, reaching from the side of the superior maxillary branch to, but not HOWARD: HERPES IN CROITPOU8 PNEUMONIA. 11 uniting with, the larger ganglionic mass corresponding to the origin of the inferior or third branch. About the middle of this strip there was marked congestion of the veins about the superior maxillary branch, into and, here and there, just beneath the capsule of the ganglion, This congestion and hemorrhage must have compressed the ganglion cells in this region. Sections 36 and 37 showed the first well-marked changes in ganglion cells and interstitial tissue. About the central portion of this strip of ganglion tissue some of the ganglion cells were vacuolated, others were very pale, non-granular, homogeneous, and stained poorly. The nuclei showed no special changes. Some ganglion cells were evidently necrotic, the cytoplasm staining diffusely, and the nuclei being displaced to one side. Here, especially about the changed ganglion cells, there was marked cellular infiltration, with plasma cells, lymphocytes, and polymorphonuclear neutrophiles. These cells were both in the interstitial tissue and grouped about changed ganglion cells, which were evidently compressed thereby. In places there had evidently been a proliferation of the endothelial cells about the ganglion cells. In sections 40 to 60 the congestion and hemorrhage were noticeable, the cellular infiltration being inconspicuous. Beginning in section 73, cellular infiltration became marked again. In section 77, in the above-mentioned strip of ganglion tissue involving an area of from twenty to twenty-five ganglion cells, there was marked cellular infiltration with interstitial hemorrhage. Some of the ganglion cells were necrotic. Smaller areas of interstitial cellular infiltration were found near this area. Section 78, showed, at the border of the same hemorrhagic area a group of degenerated ganglion cells surrounded by large numbers of lymphocytes, plasma cells, and polymorphonuclear leucocytes, red blood cells, aud proliferated endothelial cells of the pericellular spaces. The ganglion cells of the affected area showed the same changes described above; many have entirely disappeared. They seemed to have been compressed and destroyed by the interstitial exudate and hemorrhage. A few ganglion cells in this area were the seat of large vacuoles. Nuclear figures could be made out here and there in the cells of the interstitial infiltration. In most of the sections from 60 to 85 these changes were more or less well marked. The area of hemorrhage could be traced in them all. The changes in the left ganglion were lost after the 95th section. Right Oasserian Ganqlion. All the sections cut showed congestion of the veins and capillaries about the second or superior maxillary branch, as described for the right ganglion. Hemorrhage was, how- ever, absent. In sections 80 to 95 a few areas of cellular infiltration were found in the area corresponding to that showing similar changes in the left ganglion. These areas of cellular infiltration were small and scattered, and nothing like so well marked as those found in the left ganglion. In both ganglia a considerable number of amylaceous and hyaloid bodies were found in the lymph spaces. In preparations of the fifth nerve, and of the superior maxillary branches of both sides, made by Marchi's method, no degenerations were to be found. No bacteria were found in the ganglia or in the dilated veins. 12 HOWAED: HEEPES IN CEOUPOUS PNEUMONIA. That herpes zoster, following some part of the distribution of the trigeminal nerve is due to changes in the Gasserian ganglion is sup- ported by observations on four previous cases. Wys, in 1871 (quoted by Head and Carpenter), reported a case of herpes zoster involving the distribution of the whole of the first division of the trigeminal nerve ; the patient died on the seventh day of the eruption. There were thrombosis of the ophthalmic vein, abscesses of the eye muscles, and purulent infiltration of the connective tissue of the eyeball. There was hemorrhage about the origin of the ophthalmic division and about the inner side of the ganglion. Microscopically, there was extravasation of blood into the ganglion and into the first division of the nerve, with purulent inflammation of the ganglion. Sattler, in 1875 (quoted by Head and Carpenter), reported a case of herpes ophthalmicus of the right side, in a man, aged eighty-five years, occurring some days after carbonic oxide gas poisoning. Death oc- curred fourteen days afterward. In the corresponding Gasserian ganglion there was destruction of ganglion cells and small-cell infiltra- tion. The ophthalmic branch of the nerve showed degeneration, the other branches being normal. Head and Carpenter reported a case of herpes of the right frontal region in which, one hundred and ninety days later, at autopsy, they found hemorrhage and destruction of ganglion cells in the right Gas- serian ganglion. In another case these authors studied the right Gasserian ganglion thirty days after an attack of herpes of the chin and cheek of the right side of the face (third division). " In that portion of the right gan- glion which receives the fibres of the inferior maxillary branch of the nerve " there was congestion of the bloodvessels, destruction of ganglion cells, and cellular infiltration. Degenerated fibres were found in the fifth nerve central to the ganglion, in the pons and medulla, as well as in the inferior maxillary division of the fifth nerve. Both of the last two cases were subjects of general paralysis. Wys' case was apparently spontaneous, while that of Sattler followed carbonic oxide gas poisoning. As far as I am able to find, these comprise all the observations on the relation between changes in the Gasserian ganglia and herpes distributed along the course of the trigeminus. Of these four cases, three were examples of ophthalmic herpes, and one of inferior maxillary herpes. I am not aware that the relation of the changes in the Gasserian ganglia to herpes of the upper lip and nose, occurring either spontaneously or in the course of pneumonia, malaria, cerebro-spinal meningitis, gastro-intestinal disorders, and other infec- tions and intoxications, has been investigated. I wish to call attention to the practical identity of both the lesions of the skin and of the nervous system in these two cases of herpes, the HOWARD: HERPES IN CROUPOUS PNEUMONIA. 13 one of the trunk, and the other of the upper lip and nose, both occur- ring during the course of acute croupous pneumonia. Reference to the detailed descriptions will, I think, show that the lesions agree in all important particulars, and only differ in severity. The extent and severity of the skin lesions are also in definite relation to those of the corresponding ganglia. The skin and ganglion changes in Case II. (trigeminal distribution) were more severe than in Case I. (dorsal 11). One is forced to the conclusion that the herpes of the trunk and the herpes of the lip and nose occurring in these two cases are identical clinically, histologically, and probably etiologically. They further demonstrate that the ganglionic lesions of herpes zoster of the trunk and herpes of the lip and nose occurring in pneumonia are identical with those described by previous observers for herpes zoster of the trunk and extremities and herpes distributed along both the first and third divisions of the trigeminus, occurring both spontaneously (idio- pathic herpes), and secondarily to acute infections, as cystitis, bronchitis, acute tubercular pneumonia (Head and Carpenter), after poisons, as carbonic oxide gas (Sattler), and the invasion of tumors (Head and Carpenter). As far as I have been able to learn, no anatomical studies have been made on the lesions of the nervous system in herpes of the lip and nose occurring in malaria and other continued fevers, in pneu- monia, cerebro-spinal meningitis, coryza, or gastro -intestinal disorders. We are also without similar observations in genital herpes. In the case of the latter, however, Unna's description of the skin lesions, based on observations in three cases, is practically identical with his findings in a case of herpes labialis in an individual dead of a continued fever, with these lesions of our cases and with those found by Head and Carpenter and others in herpes zoster. Unna, however, found marked ballooning of the epithelial cells of herpes zoster (trunk ?), a condi- tion not found by him in herpes genitalis and his one case of herpes labialis. In our case of labial and nasal herpes, in some places the swelling and liquefactioa of the epithelial cells, especially in the hair follicles, was better marked than in the case of herpes of the trunk. It seems likely that in a necrotic and exudative process passing rapidly through dif- ferent stages the changes would not be uniform in the lesions of each and every case. All that can be said at present is that the skin lesions of our two cases are identical and agree with those described by most authors. Though there are no previous observations on the relation of changes in the nervous system in herpes labialis and nasalis occurring in pneu- monia, and none at all in malaria, for the third acute infectious process iu which herpes is common there are some most interesting and sugges- tive data. I refer to epidemic cerebro-spinal meningitis. In their 14 HOWARD: HEKPES IN CEOUPOUS PNEUMONIA. monograph on this disease, Councilman, Mallory, and Wright' call at- tention to the frequent occurrence of herpes of both the lips and nose, and to the fact that it may appear on other parts of the face and even elsewhere on the body; and in the summary on page 163 they state that " in its (the infection) extension along the fifth nerve it produces an acute inflammation of the Gasserian ganglion, with destruction and degeneration of the nerve cells composing it." On page 114 they state that sections were made of the Gasserian ganglia in five cases and of the spinal ganglia in two. The Gasserian ganglia in acute cases were " infiltrated with pus, and masses of ganglion cells were often separated from their connection." There was more or less hemorrhage ; some ganglion cells were small and some devoid of nuclei ; some cells contained large vacuolar spaces. Diplococci were found in some sec- tions. In the more chronic cases the amount of leucocytic infiltration was less, but there was marked proliferation of the interstitial tissue, and in one case there was oedema, with marked cellular infiltration, especially about the ganglion cells, lymphoid, epithelioid, and plasma cells being found. There was atrophy, often going on to complete necrosis of the cells (ganglion cells). The spinal ganglia were not equally affected, but all seemed somewhat swollen and oedematous. On microscopic examination much the same changes were found in these as in the Gasserian ganglia. Degenerative changes were, of course, to be found in the nerves. They also showed that these changes may occur rapidly, for purulent infiltration was observed in the ganglia of a goat twelve hours after inoculation of the diplococcus intracellularis meningitidis into the spinal cord. On page 125 they state that in one case the herpetic vesicles on the lip were examined. In them there was extensive infiltration, with pus cells in the tissue around the vessels, and proliferation of the fixed cells of the tissue. No mention is made of the epithelium ; no diplo- cocci were found in the pus cells. Unfortunately no statement is made concerning the presence or absence of herpes of either the head or the trunk in the cases showing lesions of the Gasserian and spinal ganglia. In the one case showing skin lesions (small, dark purplish spots on the trunk and extremities) coming to autopsy they found congestion and dilatation of the blood- vessels of the skin, with hemorrhage in places, and in some places prolifer- ation of the cells about the vessels. In the centre of the hemorrhagic area there was some infiltration with pus cells, and in one place the upper layers of the epithelium were slightly elevated by the accumulation of pus cells beneath. Tliese observations furnish at least ample explanation for the occur- ' Report oi the Stale Board of Health of Massachusetts, 1898. HOWARD: HERPES IN CROUPOtTS PNEUMONIA. 15 rence of herpes of both the head and trunk in meningitis. The influ- ence of the changes in the ganglia upon the other changes in the skin are, of course, only conjectural. Etiology of Herpes. From what has gone before, it is evident that the lesions of the skin and nervous system of primary, or idiopathic, or spontaneous herpes zoster of the head, neck, trunk, and extremities are the same, as are also the same lesions in certain cases of herpes of the trunk and face (upper lip and nose) occurring in pneumonia (Howard), and similar lesions of the nervous system in herpes of the trunk occurring in acute cystitis in a man with paraplegia, acute bronchitis ; pernicious anaemia in a patient who had taken arsenic, and after the invasion of a lymphosarcoma (Head and Carpenter), and finally in ophthalmic herpes after poisoning with carbonic oxide gas (Sattler). Primary herpes zoster is regarded by Head and Carpenter, and by Van Harlingen' and others, as an acute specific infectious disease due to an unknown cause, having a selective affinity for certain sensory ganglia in which certain degenerative, destructive and often inflam- matory lesions are produced. These lesions in some unknown way cause acute degenerative, exudative, and proliferative changes in the skin and mucous, and possibly serous, membranes corresponding to the nerve distribution of the affected ganglia. There is a considerable amount of evidence in favor of the hypothesis that there is an acute specific infectious process of this nature, but no convincing proof has been so far adduced in its support. It must be further borne in mind that there is a large group of cases of herpes either following the action of known toxins (carbonic oxide gas and arsenic), the involvement of the ganglia by tumors, or occur- ring in the course of certain infections, as pneumonia, malarial and typhoid fevers, and cerebro-spinal meningitis (in which, as has been shown, there are often intense lesions of the sensory ganglia), in all of which the lesions, as far as is known, are identical. Herpes, like pneu- monia, meningitis, and inflammation in general, is a pathological con- dition, with definite lesions, capable, however, of being excited by a variety of causes, and is not always produced by the same causes ; an efficient cause being one producing compression, degeneration, or destruction of ganglion tissue. It is not improbable that the primary lesions are often due directly or indirectly to the soluble toxins of various micro-organisms. The special 'frequency of labial and nasal herpes, and of herpes of the head and face in general, in malaria, pneu- monia, and cerebro-spinal meningitis, is readily explained. In the case of malaria the organisms in the circulating blood often accumulate in the brain, where the Gasserian ganglia may be readily affected by con- ' The American Journal of the Medical Sciences, January, 1902. 16 HOWARD: HERPES IN CROUPOUS PNEUMONIA. gestion, the action of toxines, or even by capillary thrombi of parasites. In pneumonia, the organisms or their toxines may locate in the ganglia. What, however, to my mind seems the most reasonable predisposing cause of the frequent occurrence of labial and nasal herpes in this dis- ease is the marked passive congestion which is so often present, ihis was especially noticeable in Case II., in which there was marked con- gestion of not only the bloodvessels and sinuses of the brain and meninges, but of the vessels of and near the Gasserian ganglia them- selves, and especially about the second division. This marked con- gestion, which favors hemorrhage, the action of the toxines, poisons, and organisms in the blood, and thrombosis, may probably of itself, by pressure and by interfering with nutrition of the ganglion cells, cause degeneration sufficient to excite the herpetic changes. As Councilman, Mallory, and Wright have shown, in epidemic cere- bro-spinal meningitis, the spinal sensory ganglia, and especially the Gasserian ganglia, are often the seat of extensive changes due to the extension of the infection along nerve sheaths to these ganglia. There is, a priori, no reason why changes in these ganglia should not occur in bronchitis and pleurisy. There is also no reason why the peripheral lesion in herpes zoster should not affect the pleura, as sug- gested by Curtin, just as the same process is known to affect the mouth, cheek, and pharynx. We know nothing in regard to the relation of lesions of the nervous system to the so-called simple herpes (with which labial iind nasal herpes of pneumonia, malaria, and epidemic cerebro- spinal meningitis have hitherto been included) of the face, especially of the lips and nose, occurring in coryza, gastro-intestinal disorders, and the herpes genitalis. It is to be hoped that observations will be made on this point, and it is not impossible that some cases, at least, will fall in the class we have been considering. The (question why should the peripheral changes occur after the lesions of the nervous system found in herpes is a difficult one to answer. There is no evidence that the skin lesions are due to the local presence of parasites, infection with which is predisposed by the nerve lesions. Head and Carpenter are " inclined to think that the trophic disturb- ances of the skin are an extreme form of activity of the same cells, dis- turbance of which by afferent impulses along the white ramus produces the hyperalgesia that accompanies visceral referred pain ;" but they " do not imagine that the eruption of herpes zoster is produced by disturbance of special trophic nerves, but by intense irritation of cells in the ganglion which normally subserves the function of pain, and more particularly that form of pain produced by afferent visceral im- pulses." HOWARD: HERPES IN CROUPOUS PNEUMONIA. 17 This is, however, far from explaining why irritation of these ganglion cells and nerve iibres causes a severe inflammatory process marked by intense vascular changes, cell destruction, exudation, and proliferation The presence of hyaline, granular, and apparently red blood cell thrombi in many of the dilated bloodvessels of the herpetic areas in my cases is of interest, and may account for the hemorrhage, but were hardly extensive enough to explain the necrosis. What was the origin of these thrombi ? Were they associated with changes in the vascular endothelium due to nerve changes ? Were the hemorrhage and exudation due, in part, to the increased permeability of the vessel walls, the result of loss of nerve control or nerve irritability ? These are some of the questions awaiting solution. Conclusions. 1. Herpes zoster is a pathological condition, like pneumonia, for instance, with definite lesions of certain sensory ganglia, sensory nerves, and the skin, capable of being excited by a variety of causes. It is probable that the primary ganglionic lesions are com- monly due directly or indirectly to the soluble toxines of various micro-organisms. The skin lesions may be on the head, neck, trunk, or extremities, corresponding to the Gasserian and posterior root ganglia affected. 2. Various forms can be distinguished, a. Spontaneous or primary herpes, thought by Head and Carpenter, and others, to be a specific infectious disease, the specific causal agent of which has a special afiinity for certain sensory ganglia (posterior spinal and Gasserian). h. Herpes occurring after certain definite toxic agents, as arsenic and carbonic oxide gas, etc. c. Herpes occurring in the course of certain acute infectious diseases, as pneumonia, cerebro-spinal meningitis, and probably of malarial and typhoid fevers. The lesions of the gan- glia and of the skin in the above three forms are the same, and the processes, therefore, presumably identical, d. Herpes simplex, so-called, affecting the lips and nose in coryza, gastro-intestinal intoxications, etc., and genitals (herpes genitalis) has not been sufiiciently investi- gated to be classified ; no evidence exists for or against its connection with changes in the nervous system. 3. As far as changes in the skin in herpes are concerned, they are illustrations of particular forms of necrosis and inflammatory reaction, and, as in similar lesions in other organs, can probably be excited in a variety of ways. 4. Herpes should be classified according to its relation to changes in the nervous system, and to this end every possible opportunity should be embraced for extending our knowledge in this direction. ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM, WITH THE REPORT OK A CASE DUE TO AN UNIDENTIFIED MEMBER OF THE ACTINOMYCES GROUP.* William Tkavis Howard, Jr., M.D. {^Professor of Pathology, Western Reserve University^ {From the Pathological Laboratory of the Lakeside Hospital^ Cleveland.) The observation of an apparently primary abcess of tiie brain caused by a branching, intertwining, thread-like organ- ism similar to those recently found in a variety of lesions, and commonly called streptothrix, but properly classed under the genus Actinomyces, led to a reviev^r of the litera- ture of actinomycosis of the central nervous system. Including mine, eighteen cases were found, and of these the nervous system was apparently affected primarily in five, and secondarily in thirteen. On comparison many of these are very similar to each other and to our own. It is my intention in the present article to state the salient features of the cases in the literature, report my own case, call attention to the identity of so-called Streptothrix with Actinomyces and, finally, summarize our knowledge on the subject. I. PRIMARY ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. The first recorded is that of O. Bollinger^ (1887), who reported the case of a woman aged twenty-six years, whose illness began about one year before death with headache, soon after which there was paralysis of left N. abducens. Six months later there was paralysis, with eczema, of the forearms, and about this time she gave birth to a vigorous boy. Headache persisted, and nineteen days before death there was double vision and changes were made out in the left retina. Death occurred in coma after convulsions. Diagnosis, * Received for publication March 24, 1903. 301 302 HOWARD. brain tumor. The lateral ventricles were enormou.sly dilated, and the foramen of Monro and the third ventricle were distended by a growth the size of a hazelnut, with smooth surfaces, and of a bluish gray color. On section a mucoid, sticky fluid escaped. Microscopicallythe mass was composed of granulation tissue containing numerous typical colonies of actinomyces. No other focus of actinomycosis was to be found in the body. The woman is said to have had bad teeth. Almquist^ (1890) reported the case of an artilleryman who died after a two weeks' illness with cerebro-spinal meningitis. The autopsy was made twenty hours after death. Plate cultures made from the base of the brain and from the lateral ventricles gave colonies of a large micro- coccus, B. proteus, and a single colony of a streptothrix. This organism grew well on gelatine, forming rather large convex hemispherical colonies and causing liquefaction of the medium. On agar there was a rapid luxuriant growth, with larger and smaller white crusts. In broth there was a flaky growth, with a thin surface crust. Microscopically there was a mycelium of unjoined threads with true and frequent branching. The threads averaged 0.5 /tt. in diame- ter and sometimes reached 100 fi. in length. Air hyphae which broke up into small cubical or oval cells appeared on the white crusts. Both cells and threads stained lightly with the aniline dyes. No mention is made of finding this organ- ism in the pus or sections of the meninges, or of animal experiments. Absence of microscopical proof of the pres- ence of this organism in the lesions and the fact that a single colony was obtained in the cultures lend strong probability to the natural inference of air~contamination. I am, there- fore, very strongly inclined 'to question the propriety of including this case. In 1 895 Ferre et Foquet^ reported the case of abscess of the brain of a man who during life had epileptiform crises. . From the description the brain abscess was probably primary. From the pus they obtained a " streptothrix " in pure culture. The organism grew well on the various media, but best on ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 303 potato. Microscopically it was made up of ramifying fila- ments with knob-like terminations. The organism stained by Gram and was not pathogenic for guinea-pigs. On sub- dural inoculation of a rabbit there was general diffusion of the parasite in the organs without reaction or the formation of pseudo tubercles. Musser, Pearce, and Gwyn* in 1901 reported a case of probably primary brain abscess in a man twenty-four years old. " At a point just posterior to the upper portion of the fissure of Rolando is an area three centimeters in diameter which is fluctuant, grayish yellow in color." The side of the brain affected is not mentioned. On section at this point there was found an abscess 3.5 cm. in diameter, containing thick partly chocolate-colored and partly grayish yellow fluid of the consistency of oil. No granules were made out. The walls of the cavity were greenish gray in color and soft and neurotic. The brain tissue about the abscess was softened and of a yellow color. The pus had a foul odor. (The striking resemblance to the abscess in my case will be noted.) The rest of the brain was normal. The ears and the bones of the skull appeared normal. The other organs could not be examined. In the pus of the brain abscess after search there was found a large ball-like tangled mass made up of long irregularly staining filaments. There was some suggestion of branching of the periphery of the mass. His- tological examination of the abscess wall showed a large mass of radiating threads, which at the periphery were long and thin and stained irregularly. The whole had the appear- ance of an actinomyces colony. Clubs and rosettes could not be demonstrated. No tubercle bacilli could be found. Aerobic and anaerobic cultures and animal inoculations (guinea-pigs) were negative. J. C. — Male, white, aged fifty-two years, was admitted to the medical wards, service of Dr. Edward F. Gushing, of the Lakeside Hospital, April 21, 1902. The patient had been an orderly in the hospital for a number of years and had always been well until his present illness. Several days before his illness he had complained of not feeling v/e!I, and 304 HOWARD. for two days before admission to the ward he had not re- ported for work. On the morning of April twenty-first he was found in his room in a condition of stupor. He was very drowsy, but on being aroused could answer questions and complained of pain in his right eye. The conjunctiva of this eye was reddened, the other pupil was markedly dilated. The left eye and pupil appeared normal. The patient could move his hands and arms and there was no evidence of paralysis. His eyes were examined by Dr. Millikin, who found nothing abnormal. Physical examination disclosed nothing further except varicosity of the veins of the left leg. Ulceration of the leg is not mentioned in the clinical history and was therefore probably not present. On April twenty- second his condition was much the same, but questions were answered with difficulty and stupidly. The right pupil was dilated and did not react to light. Breathing was regular and somewhat accelerated. The pulse was full sixty-four to the minute. Physical examination of the thorax and abdo- men was negative. The knee jerks on both sides were much increased. On the left side there was a marked Babinski re- flex. The patient could move his upper and lower extremi- ties, but there was marked rigidity of the muscles of the left arm and to a less degree those of the left leg. The tempera- ture ranged between lOO" to 101.5° F. There was involun- tary micturition and defecation. On April twenty-third the note made records that the patient continued throughout the preceding day in a comatose state, and during the night the coma deepened. At times he was quite restless and com- plained of pain in the right side of the head. To-day there was deep coma from which he could not be aroused. The right pupil was still widely dilated ; the reflexes were the same as yesterday. There was now rigidity of the right arm and leg, with paralysis of the left arm and leg. Breathing was stertorous, but not especially rapid. The pulse was strong and eighty to the minute. The leucocyte count showed twelve hundred per cubic millimeter. There was no discharge from the ears, the membranes appeared normal. He was transferred to the surgical service of Dr. Dudley P. ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 305 Allen, who trephined over the motor area of the right side. There was distinct bulging of the brain when the dura was cut through. The meningeal vessels were markedly con- gested, but no hemorrhage and meningitis, and no evidence of brain abscess were found. The rigidity of the arms be- came less marked after incision of the dura. The patient died at five-forty the afternoon of April twenty-third. Autopsy six hours after death. The body was fairly well built and nourished, and 170 cm. in length. Rigor mortis was present. The right side of the head was shaved and over the right parietal region there was a semicircular wound closed with sutures. The right pupil was somewhat larger than the left, which was of ordinary size. The external surface of the left leg was the seat of a large varicose- ulcer. Near this there were a number of scars of former ulcers. No- other scars were to be seen. The head was of ordinary size and shape. The wound above referred to was healthy in appearance and free from pus. The skull was of ordinary thickness and ap- pearance. There was no softening. The dura at and near the seat of the operation was markedly congested. The brain and pia-arachnoid bulged slightly at the opening. The pia-arachnoid over a considerable area was congested and in a large vein just beneath the wound contained a dark red thrombus. On removal of the brain, the superior tem- pero- sphenoidal and the superior marginal convolutions of the right side were somewhat softer than usual. About the middle of the superior tempero-sphenoidal convolution and 15 mm. below the trephine hole through the skull there was an area of fluctuation just beneath the surface. On handling the brain it ruptured at this point with the escape of a con- siderable amount of dark, grayish green, thick pus with a foul putrid odor like that of butyric acid fermentation. On in- cision in this region there was an abscess the size of a walnut, surrounded by several smaller abscesses varying from 3 to 15 or 20 mm. in diameter. They all had soft necrotic walls and were evidently recent. The brain tissue for some dis- tance, 4 cm. in some places, on all sides of the main abscess was soft, edematous, and of a peculiar gelatinous consistence 306 HOWARD. and appearance. The abscesses with the area of softening extended antero-posteriorly from the ascending frontal to t e angular convolution and internally nearly to the right latera ventricle. The pia-arachnoid over the abscess was markedly congested and cloudy in appearance. On section of the bram no other abscesses were to be found. The structures at the base and all the vessels appeared normal. The sinuses were free from thrombi and the middle ears and mastoid cells were normal. The spinal cord was not removed. Chest. The muscles were well developed; the costal cartilages were calcified. Both lungs were bound to the chest-wall by dense fibrous adhesions ; both pleural cavities were almost completely obliterated. Both lungs were volu- minous and on section markedly congested and edematous. No areas of consolidation were to be found. The bronchi contained frothy, blood-stained serum. The bronchial glands were pigmented, but otherwise were normal. The pulmonary vessels were normal. The trachea, tongue, and thyroid were negative. The heart, liver, kidneys, spleen, stomach, intes- tines, and other organs showed nothing of present interest and their description is therefore omitted. The thick, creamy, grayish green appearance and the butyric acid odor of the pus of the brain abscess reminded me at once of pus with similar characteristics obtained some years ago from two cases of empyema from which an organ- ism apparently belonging to the ray fungus group was obtained, so cover-slips were examined immediately during the progress of the autopsy. The pus was rich in poly- morphonuclear neutrophilic leucocytes and showed a few scattered bacillus and thread-like forms. In several prepara- tions stained by Gram's method larger and smaller clumps of branching threads were found. A few leucocytes con- taining bacillus and long thread-like forms were seen in some preparations. The threads varied very much in length and to some degree in thickness. The bacillus and thread- like forms varied from three to thirty or forty microns in length. As a rule they stained deeply and uniformly with ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 307 fuchsin and gentian violet in either aqueous or aniline solu- tion, but in many, especially the shorter forms, i.e., those of four to eight microns in length, showe'd larger or smaller unstained areas which usually involved the whole thickness of the organism. In some of these forms with unstained areas a thin cell membrane was discernible. The organisms varied from 0.3 to 0.5 in thickness, the former being the average. Some forms showed irregular slight swellings, which were usually situated at one or the other end. No typical knob or club-like forms could, however, be found. The ends were usually square, were often rounded, especially in the shorter forms. Many of the longer forms were wavy in outline. Some of the shorter forms were bent, forming almost comma shapes. Some threads and rods showed transverse division into short bacillus and even coccus forms, and thus chains of bacilli and cocci were not uncommon. The latter forms were always larger than streptococcus pyogenes. The mass of intertwining branching threads showed no special spore-bearing structures ; careful search failed to demonstrate the presence of clubs. The forms were non-septate, and the branching was true, the protoplasm of the main stem always being directly traceable into the branches. In the larger clumps the growth was always radiating, and the mycelium formed a distinct felt work. No capsules could be made out. Numerous cultures, both aerobic and anaerobic on slants and on Petri plates were made on various media, including plain glycerine and glucose agar and potato with glucose bouillon and chest serum, inoculated from the pus of the brain abscess. Care was taken to use a large amount of pus for sowing the cultures. In many of the cultures strepto- cocci, which reacted in every way to the tests for strepto- coccus pyogenes, grew. In none of the cultures were there forms suggestive of those found on the coverslips. Follow- ing the plan of Norris and Larkin, attempts were made to grow the organism on the organs of newly killed rabbits, but always without success. Aerobic plate cultures from the lungs gave streptococcus pyogenes and staphylococcus 308 HOWARD. pyogenes aureus (a few), from the liver and kidneys strepto- coccus pyogenes. A rabbit and a guinea-pig were inoculated at once intro- peritoneally. The guinea-pig remained well and was killed six weeks later and showed no lesions. The rabbit died twelve days after inoculation, and showed a wide-spread purulent peritonitis, the peritoneal cavity being distended with a large amount of thick, white, sticky pus of the consistency of thick cream, with a well marked butyric acid odor. The surfaces of the liver and spleen, as well as those of the intestines, were covered with similar material. Coverslips showed no branching forms, but numbers of long and short threads, bacillus and coccus forms, with leucocytes. Cultures from the peritoneum showed a few colonies of streptococcus pyogenes, but no other organisms. Pus from the peritoneal cavity was inoculated into the pleural cavity of a rabbit (rabbit II.), which died in three days with extensive purulent pleuritis and pericarditis, and pus from this inoculated into a third rabbit (rabbit III.) caused its death in two days with identical lesions. Other rabbits were inoculated intravenously with pus from the peritoneal cavity of rabbit II., died within a few minutes with extensive thrombosis of the right ventricle, and one inoculated into the trachea died during the night with puru- lent pleuritis. (Rabbit IV.) Histological examination and sections from various por- tions of the brain showed as follows : Brain abscess : A number of sections cut in both celloidin and paraffin and stained in hematoxylin and eosin, toloidin blue and eosin, aniline oil gentian violet, followed by ani- line oil and xylol, Weigert's gentian violet method, aniline oil fuchsin followed or not with iodine, decolorized with aniline oil, and Mallory's stain for clubs ; Weigert's stain followed by aniline fuchsin, and the Ziel stain for'tubercle bacilli. The abscess was made up of a thick mass of cells, most of which were polymorphonuclear neutrophilic leucocytes, among ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 309 which there were both large and small mononuclear cells. About the margins of the abscess the tissue was infiltrated with the same varieties of cells, and often contained red blood cells. Many of the cells were necrotic ; nuclear fragments were numerous. The brain tissue for a considerable distance from the abscess proper was edematous, and contained col- lections of leucocytes. The blood vessels were markedly dilated, and in many places there was marked perivascular cellular infiltration, with proliferation of the endothelium of the perivascular lymphatics. No micro-organisms were found in the blood or lymph vessels. The only micro-organ- isms to be found were clumps of long threads always lying near the centre of small abscesses or about the margin of a large abscess. These organisms were sometimes seen singly or in small clumps, but nearly always formed larger or smaller colonies from twenty to one hundred microns in diameter. These colonies were composed of branching nonseptate threads, which formed a meshwork and radiated from the centre of the colony. These colonies when cut through the centre were nearly round in outline and had often rather regular but somewhat serrated margins. In sections stained with hematoxylin or methylene blue the colonies are finely granular and stain homogeneously with the blue stain. No organisms were made out even at the periphery. In very thin sections stained with eosin followed by Wei- gert's fibrin stain, very beautiful pictures were given. Here the colonies appeared composed of a finely granular pink ground, traversed by radiating, intertwining, branching dark blue threads, which projected from the margin of the colony in an irregular wavy line. The finely granular structureless pink staining mass appeared to support the mycelium of branching threads. The threads were usually more thickly placed at the border of the colonies, and while on first exam- ination the outlines of the colonies were uniform, closer study showed that many threads projected for some distance from the colonies between the surrounding leucocytes. Hence the relation between the colonies and the surrounding leucocytes was an intimate one. No leucocytes were ever, however, 3IO HOWARD. found penetrating the colonies, although they approached individual threads. These projecting threads never formed clubs, and showed no special tendency to break up into short bacillus and coccus forms. Near the colonies both long and short free threads were to be found. Reverting to the threads in the colonies, they stained well with aniline gentian violet and both aniline and carbol fuchsin, and retained the stain tenaciously whether treated or not with Lugol's solu- tion, or exposured to aniline oil or alcohol. The threads were of about the same size as those described on coverslips made from the fresh pus. They often showed unstained spaces and irregular lateral swellings or bulging of the cell membrane. Many were wavy and tortuous in out- line. The threads took the stain very differently — some deeply and others very faintly. No clubs were found. The threads were not septate. No spore-containing bodies were seefn. Here and there among the leucocytes were larger and smaller masses of threads, some of which were evidently parts of colonies, while others represented detached masses. The threads readily yielded the fuchsin and gentian violet stains on treatment with acids in either weak aqueous or alco- holic solution, and therefore do not belong to the acid fast fungi. Sections, including the pia-arachnoid, over the abscess and near the seat of operation, showed congestion of the blood vessels and leucocytic infiltration. No threads were found here. Histological examination of the other organs showed no acute or chronic lesions in any way suggestive of reaction to this fungus. Animal experiments and sections of the peritoneum over the abdominal wall, liver, and intestines of rabbit I. showed swelling, desquanjation, and necrosis of the endothelium in places, with an exudate of leucocytes. The leucocytic accu- mulation varied markedly in extent in different places There was a notable absence of fibrin. Most of the leuco- cytes were of the polymorphonuclear neutrophilic type. Among the leucocytes there were in many places great ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 3 1 I numbers of bacillus and streptococcus forms. In some places long threads were seen, but no branching forms could be found. The threads were never in the form of colonies, as in the brain abscess, but were met with singly or in scattered clumps. They were of the same size, general appearance, and had the same staining characters as those described in the brain abscess. The same applies for the bacillus and coccus forms. Many of the former were curved or bent, and the latter were on the whole larger than strepto- coccus pyogenes. No nodular formations were present in the peritoneum or in the organs. Sections of the lungs and pleura and pericardium of rabbits II. and III. showed wide-spread purulent exudation, without recognizable fibrin, or the formation of definite nodules. The micro-organisms were even more numerous in these exudations than that of rabbit I. They had the same characters, with the ex- ception that here clumps of definite branching forms could be found. There were no well-marked colonies, but some- what elongated tissues of interlacing threads, some of which showed true branching. The shorter non-branching bacillus and coccus forms were, however, much more numerous. In both animals, as in the brain abscess, there was a certain amount of phagocytosis on the part of the polymorpho- nuclear leucocytes, some of which contained numbers of short threads or forms. The parenchyma of the lungs of rabbit II. was not appre- ciably affected, and no organisms were found in other organs. The organisms had the same staining reaction noted for those in the brain abscess.* Including Almquist's rather doubtful case, there are five known cases of primary cerebral actinomycosis, three of which have been reported as instances of " streptothrix" in- fections. In all except Bollinger's case, the only one to show the changes of " typical actinomycosis," the course of the disease was acute. ' In this case the diagnosis was brain tumor. In one case the diagnosis was tuberculous meningitis, * My thanks are due to Drs. Gushing and Allen for the use of the clinical history of the case. 312 HOWARD. in another, cerebro-spinal meningitis, and in one cerebral hemorrhage. Four cases were males and one female ; all were adults. The organisms were cultivated in two cases (Almquist and Ferre et Fouquet), and not classified as typical actinomyces, but as " streptothrices." Pathogenesis for animals was tried in four cases, and was positive in only one (mine). II. SECONDARY ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. The earliest reported case of secondary cerebral actinomy- cosis is that of Ponfick," described in his work on actinomy- cosis in 1882, and quoted by O. Bollinger. A woman who had actinomycosis of the neck and prevertebral tissue de- veloped actinomycotic masses of the left cerebral hemi- sphere, as well as of the pericardium, lungs, and spleen. In a second of Ponfick's cases, an actinomycotic process extended directly through the skull and invaded the pia and the right temporal and frontal lobes. Kohler,^ in 1884, reported the case of a man with a num- ber of fluctuating nodules on the surface of the body which were mistaken for glanders. At autopsy there were found ulceration of the upper jaw and sternum, and abscesses of the brain, neck, tongue, spleen, liver, and kidneys, and colon. In the tenacious green pus of the various abscesses there were small greenish masses, which on microscopical examination proved to be colonies of a ray fungus. No mention is made of cultures or animal inoculations. O. Israel,' in 1884, described the case of a laborer thirty- one years old, who entered the hospital with a number of ulcers and abscesses on the surface of the body. At autopsy there were found ulcers of the skin of the chest, pericardial adhesions, abscesses of the heart, spleen, liver, kidney, and intestine, and brain. Numbers of actinomyces were found in the pus from the skin lesions. Naunyn,' in 1888, reported the case of a girl of sixteen years who died with the symptoms of chorea. At autopsy ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 313 there were small excrescences on the mitral valve and red- dish-brown masses on the cerebral pia-mater. In both lesions a branching thread-like organism was found. Cult- ures and animal experiments were not mentioned and were probably not made. It is interesting to recall that Zopf, to whom specimens were submitted, classified the organism as a species between Cladothrix and Leptothrix. Keller,^ in 1890, reported the case of a female forty years old who had pleurisy in 1885, and who in 1886 developed an abscess over the sixth and seventh costal cartilages. Acti- nomyces were found in the pus of these abscesses. Two years later she was taken ill with convulsions, paralysis of the left side, and coma. At operation, and later at autopsy, chronic meningitis and a small abscess at the right side of the brain were found. In 1 89 1 Eppinger^" described, under the title " Ueber eine neue pathogene Cladothrix und eine durch sie hervorgeru- fene Pseudotuberculosis (Cladothrichica)," a case of brain abscess and meningitis associated with changes in the bron- chial and supra-clavicular glands. The patient was a glass- worker, fifty-two years old, who developed what was appar- ently joint rheumatism. Eight days later he was unable to work. Later there was paralysis of the left side, tenderness of the right side of the head, delirium, and coma. Death occurred after an illness of twelve days. At the autopsy the skull was healthy. There was well-marked cortical, basal, and spinal meningitis, the exudate at the base of the brain being thick and sticky and of a yellow color. The brain tissue was congested and soft, the lateral ventricles were dis- tended with yellowish pus. Communicating with the right lateral ventricle through an opening in the optic thalmus there was a large abscess of the right hemisphere extending from the middle of the frontal to the occipital lobe. The abscess was filled with pasty pus ; its wall was thick and rather firm. There were firm nodules in the lungs and pleura (pseudo-tuberculosis). The bronchial and mediasti- nal lymph glands contained firm grayish-white nodules. The right supra-clavicular glands were very much enlarged and soft. 314 HOWARD. In coverslips, and in cultures from the brain abscess, the meningeal exudate, and the bronchial and supra-clavicular lymph glands, Eppinger obtained a branching, thread-like organism growing in star-shaped colonies which produced pseudo-tuberculosis in both guinea-pigs and in rabbits. This organism, which grew readily on ordinary culture media, has been classed by Kruse, Lehmann, and Neumann and others as Streptothrix Eppingeri. MacCullum, who has recently isolated the same organism from a case of peritonitis, cor- rectly classes it under the actinomyces, as Actinomyces asteroides. Sabrazes et Riviere" (1894) described the case of a man thirty-one years old, an alcoholic, who was said to have been ill for three weeks before admission to the hospital. Ten days before admission he showed marked mental disturbance. The day of his admission he was aphasic, had great pain in the head, vomiting, and obstinate constipation. He soon became comatose and died. Diagnosis, tuberculous meningitis. At the autopsy the meninges were found normal. The left cerebral hemisphere was markedly edematous, soft, and almost diffluent. There was an abscess the size of a small orange in front of the centrum ovale in the prefrontal region and involving the first and second frontal convolutions. The apices of both lungs were consolidated and one contained a small abscess. The right kidney was the seat of a large in- farct. In the green pus of the cerebral abscess there were long branching filaments, some showing unstained areas. Chains of irregular micrococcus-like forms were also seen. The same appearances were met with in the renal infarct, where coccus and rod forms as well as masses or colonies of interlacing filaments were also plentiful. The colonies were circular in shape and had a radiated appearance. The ends of the filaments were regular, and showed neither swelling nor diminution in size. Sections of the brain abscess showed edema and disintegration of the brain tissue, with inflamma- tion ; but no organisms were found. Sections of the renal infarct showed them, however. No filaments were found in ACTINOMYCOSIS OF THE CENTRAL NERVOUS 'SYSTEM. 3 I 5 the lesions of the lungs. The organism failed to grow aerobi- cally, but an anaerobic gelatine culture gave, after five days, a single colony of branching, filamentous organism like that found in the brain and kidney. Transplantation was unsuc- cessful. A guinea-pig inoculated with the pus from the brain died in two days of septicemia due to a micrococcus. As no organisms were found in the pulmonary lesions, the portal of entry is obscure and the primary seat of the affec- tion is uncertain. It is probable, however, that more careful investigation would have discovered the fungus in the lungs. The bronchial glands were not mentioned. The authors regard their organisms and that of Eppinger as different from that of actinomycosis. Dolore,''^ in 1896, recorded the case of cervical and facial actinomycosis with loss of consciousness, delirium, convul- sions, and episthotonus. At autopsy there was found marked cerebro-spinal pachymeningitis, in the exudate of which there were yellow granules, which on microscopical examination proved to be actinomyces. In the same year C. F. Martin '' reported from Chiari's laboratory two cases of secondary actinomycotic abscesses of the brain. In the first, a man thirty-eight years old, with actinomycosis of the lungs, sternum, and ribs, there was a metastatic abscess of the left occipital lobe. Actinomyces were found in the pulmonary lesions, and in the pus of the cerebral abscess. In the second, a boy of sixteen years, there was actinomycosis of the right lung and chest wall, with metastatic abscesses of the right cerebral hemisphere, the right lung, and the kidneys. The secondary abscesses were acute and contained numbers of actinomyces, clubbed forms being rare. Yellow granules were visible in the pus of both the skin and the cerebral abscesses. No mention is made of cultures or animal experiments. In 1900 Chiari,^* under the title of " Ueber Myelitis sup- purativa bei Bronchiektasie," recorded two cases of meningitis and abscesses of the central nervous system containing branching threads which he regards as identical with, or closely related to, actinomycosis. Both occurred in males 3l6 HOWARD. with bronchiectasis. Sections of the bronchiectatic cavities which were made in one of the cases showed the same organ- isms found in the lesions of the nervous system ; hence Chian regards the latter as secondary in both cases. In the first, a man forty-three years old, there was a large abscess of the cerebellum, multiple abscesses of the cord, with cerebro-spinal meningitis. In addition to Pneumococ- cus, bundles of branching threads without peripheral clubs were found in these lesions. There was no growth in cultures after two days, and no animal inoculations were made. The masses of threads were thin, showed true branching and took the various stains, including Gram's. Club formation was not observed. Chiari states that the organism of this case had the same microscopical appearances of those found in Mar- tin's above described cases of secondary cerebral actinomy- cosis, in which the " drusen " of the cerebral abscesses lacked peripheral clubs, though the latter were present in the lesions in the other organs. Chiari's second case was a man thirty-seven years of age who had chronic purulent bronchitis, with saccular bron- chiectasis of the right lower lobe, purulent cerebro-spinal meningitis, with an abscess of the left frontal lobe. The brain abscess and sections of the wall of the bronchiectatic cavities showed numerous bundles of branching threads similar to those found in the first case, with the exception that some threads showed peripheral clubs. No cultures or animal inoculations were made. III. CLASSIFICATION. It is quite evident that the micro-organism found in my case does not belong to the lower bacteria on the one hand, and lacks some of the characters usually regarded as distinc- tive of the typical actinomyces on the other. We have to deal with an organism which forms a mycelium'of elongated branching, intertwining, or interwoven threads, which break into beaded segments forming chains of coccus and rod-like forms, but which apparently does not form clubs or club- shaped hyphae. It caused in rabbits diffuse suppuration in ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 317 the subcutaneous tissues and diffuse peritonitis without nodules and without typical mycelia, only long and short threads which sometimes branched, and coccus and bacillus forms being present in the exudates. Failure of the organ- ism to grow in artificial media and on the organs of freshly- killed rabbits interferes seriously with its comparison with somewhat similar organisms described as actinomyces, strep- tothrices, and psuedo-actinomyces. The terms " cladothrix '' and "leptothrix" clearly do not apply here, and the term " streptothrix," though used by Eppinger,^" Ferre et Foquet,^ Sobrases et Riviere,^^ Flexner,^^ Norris and Larkin,^^ Musser, Pierce, and Gwynn,* and others for organisms more or less closely corresponding to mine is a misnomer, as Lachner- Sandoval " and Hektoen ^^ have clearly pointed out. Kruse" and Lubarsch^" are largely responsible for the application of the term " streptothrix " to the group. Lubarsch would include under the group name of strepto- thrix, placed midway between hyphomycetes and schizomy- cetes, all ray fungi, as well as those organisms usually classed as bacteria, which occasionally produce true branches in cultures and ray forms in the animal organism, regarding the various members of the group as transition stages. .As a distinct family of the streptothrix group, Lubarsch would separate those organisms which occur in ray form in the tissues, calling them ray fungi. Hektoen^* strenuously objects to the bringing together in a closed circle the most typical member of the group actinomyces albus, with the most atypical or imperfect, as the tuberculomyces, for instance, simply on the strength of what many in opposition to Lubarsch and others regard as a pathological condition, namely, club formation. Hektoen further points out the illogical and untenable position of those who use in this connection the term " strep- tothrix," a term used by Corda in 1839 for an entirely differ- ent family of hypomycetes. Hektoen, therefore, following the lead of Gasperini, Berestenew, and Lachner-Sandoval, insists, according to the principle of priority, upon the use of the term " actinomyces " for this group. MacCTallum ^' and Abbott and Gildersleeve ^ have recently emphasized this. 3l8 HOWARD. Lachner-Sandoval ^'' recognizes twenty-nine species of ray fungi, and as some of these forms have not been cultivated on artificial media the group may be smaller, owing to un- known duplicates. Many new species must be added if the term is to include all the branching, mycehum and club forming organisms now classed as bacteria (the acid resisting bacilli, B. diphtherise, etc., see Abbott and Gildersleeve ^). Berestenew^' concludes that the ray fungus disease, caused by parasites of genus actinomyces, to which various names such as streptothrix, oospora, nocardia, etc., have been CTiven, may be divided into («) typical actinomycosis (such as described by Bollinger and J. Israel Wolff, Bostrom, and others); (b) atypical actinomycosis without granules in the pus and without spherical masses of the parasites in the tis- sues, to which belong the cases of V. Eppinger, Sabrozes et Riviere, as well as certain cases of Berestenew and others ; (c) pseudo-actinomycosis, that is, cases with all the symp- toms of typical actinomycosis, and caused by different micro- organisms, (i) those which stain by Gram and which like B. tuberculosis possess the property of forming branched threads with clubs, and (2) those caused by bacteria which do not stain by Gram's method. He describes several cases of psuedo-actinomycosis. Silberschmidt, ^* writing in 1901, studied the organisms from seven cases of human and two of bovine actinomycosis. Of the human cases, one was unusual as involving the lungs, liver, and thigh, three were cases of infection of the lower lachrymal duct, and three were typical clinical cases of acti- nomycosis (jaw, mouth, and back). With the exception of two of the forms from the lachrymal duct, all showed great similarity in their cultural characters, but none corresponded morphologically with the form described by Bostrom as the only cause of human and bovine actinomycosis. There was great similarity to the forms described by Wolff and Israel, but there were many cultural differences. Silberschmidt regards as similar to his the organisms from two of Berestenew's cases of so-called psuedo-actinomycosis. He quotes Poncet and Berard's description of the points of difference between actinomycosis and pseudo-actinomycosis : ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 319 (i.) In actinomycosis the mycelium consists of hairlike, radiating threads with numerous peripheral clubs, while in pseudo-actinomycosis the mycelium is more interlaced or interwoven and not radially arranged, the branching less fre- quent, and the spores " lightly stained," the organism having more the appearance of leptothrix. (2.) The actinomycosis fungus is more difficult to culti- vate and may be either aerobic or anaerobic, while in pseudo- actinomycosis the growth is more luxuriant and microscopi- cally made up mostly of forms like B. diphtheria, but with neither granules nor long threads. After comparing the cultures from his own cases with each other and with others obtained from I'lnstitute Pasteur and Krai's laboratory, Silberschmidt makes the following groups : (i.) Growth aerobic and at room temperature, colonies on agar and blood serum are adherent to and send out numerous mycelial outgrowths into the media. (a.) Liquefies gelatine, threads long, interlaced, and un- broken, includes actinomycosis hominis and bovis (Bostrom and others), and actinmodurae. (3.) Does not liquefy gelatine, threads broken and often appearing in shprt forms, actinomyces asteroides (Eppinger), actinomyces Caprae (Silberschmidt). (2.) Colonies not adherent to media, no offshoots, gela- tine not liquefied, threads mostly short with many bacillary forms, examples of the group are actinomyces farcini C, and the organism obtained in Silberschmidt's case 3. (3.) Growth preferably anaerobic, colonies show no myce- lial projections, and on solid media are usually small and sharply circumscribed. No growth on gelatine or at room temperature ; the colonies are readily crushed and broken up. The organism does not survive as long as those of groups one and two on artificial media. To this group be- long the cultures from six of his human cases as well as those from actinomycosis bovis. Rethinks it too early, however, to attempt a definite division of the pathogenic varieties of the genus actinomyces, and he does not regard pathogenesis as a possible means of diagnosis. 320 HOWARD. Silberschmidt concludes that: (i.) Actinomycosis, contrary to the earlier idea, is not a specific infection caused by a single ray fungus, for a number of different micro-organisms may cause the typical disease picture. (2.) The " drusen," which are not always to be found microscopically, may or may not have peripheral clubs, and are colonies which various micro-organisms may form in tissues. (3.) The microscopical investigation of the " drusen," or of sections of tissue, permits the finding of the organisms, but not diagnosis of the same. (4.) Cultures which he obtained without difficulty must be resorted to in order to distinguish the cause of the disease. (5.) Mixed infection is not the rule in actinomycosis. (6.) It is not at present possible to distinguish by bac- teriological methods between actinomycosis and pseudo- actinomycosis. (7.) Most of the organisms found in actinomycosis be- long to the class of actinomyces, and of this class certain subdivisions are to be made. Abbott and Gildersleeve^^ have recently pointed out that the various acid-resisting bacilli, distinct from the so-called B. tuberculosis, may form mycelial threads with or without typical actinomycotic peripheral clubs, and may cause larger and smaller inflammatory nodules in the tissues of various animals. As is well known, it has been found by Schulze,^^ Lubarsch,^'' Babes et Levadili,^^ Abbott and Gil- dersleeve,^ Friedrich and Nosse,^'' and others that various members of the tubercle bacillus group and other acid-re- sisting bacilli, as well as streptothrix Eppingeri (Lubarsch), may form branching mycelial threads similar to actinomyces. Abbott and Gildersleeve urge that these organisms be classed, not as bacilli, but under genus actinomyces. Lehmann and Neumann, in the last edition of their work, classify as actinomycetes all the delicately threaded organisms free from chlorophyll, with true branching, in part very ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 32 1 abundantly ramifying mycelium, partly with the formation of conidia. This class is further divided into the coryne-bacte- ria group containing B. diphtherise, B. mallei, and B. pseudo- diphtherise, B. Xerosis ; and Mycobacteria, a group con- taining B. tuberculosis and the acid-resisting, so-called, bacilli in general ; and Actinomyces, a group made up of Actinomyces bovis, Actinomyces farcinicus, Actinomyces asteroides, Ac- tinomyces Madurse, and the various non-pathogenic acti- nomyces. It is certain that our organism belongs to the branch of the actinomyces family in which the formation of branched threads without clubs is more or less constant. Its exact position, on account of its failure to grow on artificial media, must remain unknown. Its marked pathogenicity for rabbits serves to separate it from certain forms most commonly met with in the typical actinomyces of man and cattle. Summary. Actinomycosis of the central nervous system is rare, we are able to find only five, including our own case, primary and thirteen secondary cases. The disease usually affects males in the prime of life ; in the primary cases the age being given as twenty-four, twenty-six, and fifty-two years, respectively, in three cases; not mentioned in two cases. In the secondary group, two were sixteen, five be- tween thirty and forty, one forty-three, and one fifty-two years ; age not mentioned in four. Two cases occurred in females. Of the primary, three were reported as in- stances of " streptothrix " infection and only one (Bollinger) was regarded as typical actinomycosis. Of the thirteen surely secondary cases, eight were regarded as typical actinomycosis, two as probable actinomycosis, two as cladothrix infection, and one as streptothrix infection. All of the primary cases except one (Bollinger) were acute, while of the secondary in at least ten the lesions of the nervous system were recent, and usually the cause of death. The organisms were cultivated in two primary (Almquist and Ferre et Fouquet), and in two secondary cases, in one of which (Eppinger's) aerobic and in the other (Sabrozes et Riviere) anaerobic growths Were obtained. Eppinger's 322 HOWARD. organism was pathogenic for rabbits and guinea-pigs- most of the cases neither cultures nor animal experiments were made, the diagnosis being made from the microscopical examination alone. No mention is made of any attempts to test the acid-resisting powers of any of the organisms found. The so-called actinomycotic granules were recognized in the pus of several cases. In one, instead of abscesses there were multiple nodular masses in the brain, and in Bollinger's case there was a firm mass the size of a hazelnut in the third ven- tricle and in the foramen of Monro. Primary lesions of the lungs and bronchial glands were the most common sources of the metastatic lesions of the central nervous system. In one case the actinomycotic process extended directly through the skull to the meninges and brain. It seems established that, in the present state of our knowl- edge, all processes caused by micro-organisms having a myce- lium of branching, interlacing, and sometimes radiating threads are to be considered as actinomycosis, and that or- ganisms with the above characteristics, whether or not they form " drusen " and clubbed hyphse in the tissues, are to be regarded as actinomyces. Further knowledge is necessary before a satisfactory clas- sification of the actinomyces and the processes caused by them is possible. To class tuberculosis under actinomycosis is of doubtful value. It is well to recall, however, that there are two broad divisions of genus actinomyces as used by Abbott and others, the acid-resisting (B. tuberculosis and the acid-resisting, so-called, bacilli) and the non acid-resisting or acid-bleaching, and that typical tuberculosis is always caused by the prominent member of the former, and typical actinomycosis by the prominent member of the latter group. It is better, with Lehmann and Neumann, to use the term " actinomycetes " for the large number of organisms of the latter group. This acid-bleaching group is apparently a large one, and our knowledge concerning it is best summar- ized in Silberschmidt's conclusions. The organism of our case belongs to the branch of this group in which club formation is inconstant or absent, ACTINOMYCOSIS OF THE CENTRAL NERVOUS SYSTEM. 323 pathogenesis for animals is marked, and cultures on artificial media negative. BIBLIOGRAPHY. 1. O. Bollinger. Munch. Med. Wochenschr., 1887, No. 41, S. 789. 2. Almquist, Zeitschr. f. Hygiene u. Infectionskrankheiten, 1890, Bd. viii, S. 193. 3. Ferr^ et Fouquet. Mercredi Medical, 1895, p. 441. 4. Musser, Pierce, and Gwynn. Transactions of Association of Amer- ican Physicians, 1901. 5. Ponfick. Die Actinomykose des Menschen eine neue Infections- krankheiten, Berlin, 1882. 6. Kdhler. Berlin. Klin. Wochenschr., 1884, Bd. xxi, p. 414. 7. O. Israel. Ibid., 1884, S. 360. 8. Naunym. Mittheilungen aus der Medizinischen Klinik zu Konigs- berg, 1888. 9. Keller. Brit. Med. Jour., March 29, 1890. 10. Eppinger. Zeigler's Beitrage zur path. Anatomie, etc., 1890, Bd. ix. 11. Sabrazes et Rivifere. La Presse Medicale, 1894, Sept. 22, p. 302. 12. Dolore. Gazette hebdomadaire, May 24, 1896. 13. Martin. Journal of Pathology and Bacteriology, 1896, Vol. iii. 14. Chiari. Zeitschr. f. Heilkunde, 1900. Abthiel f. path. Anatomie, :, S.351. 15. Flexner. Journal of Experimental Medicine, 1898, Vol. iii. 16. Norris and Larkin. Ibid., 1900, Vol. x. 17. Lachner-Sandoval. Inaugural Dissertation, Strassburg, 1898. i8. Hektoen. Philadelphia Monthly Medical Jour., 1899, p. 615. 19. Kruse. Fliigge's Micro-organismen. 20. Lubarsch. Zeitschr. f. Hygiene u. Infectionskrankheiten, 1899, Bd. xxxi, S. 187. 21. MacCallum. Centralblatt f. Bakteriologie, 1902, Bd. xxxi. 22. Abbott and Gildersleeve. Ibid, and Univ. of Penn. Med. Bull., June, 1902. 23. Berestenew. Zeitschr. f. Hygiene u. Infectionskrankheiten, i8g8. Bd. xxix, S. 94. * 24. Silberschmidt. Ibid., 1901, Bd. xxxvii, S. 345. 25. Schulze. Ibid., 1899, Bd. xxxi, p. 153. 26. Friedrich und Nosse. Zeigler's Beitrage zur path. Anatomie, etc., 1899, Bd. xxvi, S. 470. 27. Babes et Levadili, Arch, de Med. Exp. et d'Anat. Path., 1897, ix, p. 1041. The Journal of Medical Research, May, 1903, Vol. IX., No. 3. [Reprinted from The Journal of Medical Research, Vol. Vl., July, igoi.] STREPTOCOCCUS MUCOSUS (Nov. Spec?) PATHOGENIC FOR MAN AND ANIMALS. William Travis Howard, Jr., M.D. (^Professor of Pathology, Western Reserve University), AND Roger G. Perkins, M.D. (^Resident Pathologist, Lakeside Hospital. ) (From the Pathological Laboratory of the Lakeside Hospital, Cleveland.) From time to time streptococci have been described which show various points of difference from as well as of resem- blance to Streptococcus pyogenes and Streptococcus lanceo- latus or Pneumococcus. Among these may be mentioned Diplococcus der Brustseuche der Pferde of Schiitz ;' Strepto- coccus of contagious pneumonia of cattle of Poels and Nolen ;^ Streptococcus der meningitis cerebro-spinalis epidemica of Bonome ; ' and Streptococcus capsulatus found in spontaneous peribronchitis and multiple pulmonary abscesses in a guinea- pig by Binaghi. * We have lately obtained from the peri- toneum and various other organs at autopsy on a woman dead of peritonitis a capsulated streptococcus which cannot be classed with either Streptococcus pyogenes or Pneumo- coccus. The object of this paper is to record our observa- tions on this organism and to compare it with similar streptococci described by others. Clinical Stimmary. — L. D., female, aged thirty-four years, was admitted to Lakeside Hospital, service of Dr. E. F. Gush- ing, on May i, 1900, complaining of pain in the abdomen. Two years before her present illness she had a miscarriage at the seventh month followed by chills and fever. Since this time she had had frequent attacks of pain in the pelvic region. The day before admission she had intense pain in the hypo- gastrium which was followed by convulsions and coma. Examination of the urine showed albumin and casts. Clinical Diagnosis. — Chronic nephritis, uraemia, chronic pelvic peritonitis. (163) 1 64 HOWARD AND PERKINS. Anatomical Diagnosis. — Chronic tubo-ovarian abscess. Acute fibrino-purulent and chronic peritonitis. Chronic in- terstitial nephritis with contraction, pyelonephrosis, and pyoureter and chronic cystitis. CEdema and congestion of the lungs and brain. Acute splenic tumor. Enchondroma of the right lung, cyst of the left suprarenal capsule, multiple myofibromata uteri. The following description of the peritoneum and pelvic organs is taken from the autopsy protocol, the other organs being without present interest: " The abdomen is somewhat distended ; the abdominal and pectoral muscles are fairly well developed. The position of the intestines is normal. The omentum extends to the anterior brim of the pelvis on the right side. On the left side it passes under the coils of the small intestines and is bound by firm adhesions to the structures on the left side of the uterus. The surface of the small intestines, the caecum, and the descending colon are covered with a thin layer of opaque, grayish-white viscid fibrinous exudate. In the dependent parts of the abdomen there is a considerable quantity of creamy pus. " The contents of the pelvis are matted together, and on the left side are densely adherent to the pelvic brim. The uterus is displaced to the right. The right fallopian tube runs back- wards and to the left and, with the ovary, is bound to the uterus. On the left side of the uterus, occupying the position of the left tube and ovary, there is a fluctuating mass the size of a small orange and containing thick pus. Between the bladder and the vagina and communicating with the tubo- ovarian abscess there is an abscess filled with creamy pus containing gas bubbles. The walls of these abscesses are thick and lined with a thick yellowish-gray necrotic material. The left ovary cannot be made out. Besides the large abscess, the tubo-ovarian mass is the seat of a number of smaller abscesses. The uterus is enlarged and in its wall there are several fibroid tumors. The abdominal lymph glands are moderately enlarged." STREPTOCOCCUS MUCOSUS. 1 65 Histological Examination of the Tubo- Ovarian Abscess. — Sections of the tubo-ovarian mass showed little recognizable ovarian tissue. The abscess cavity contained necrotic ma- terial and a large number of polymorphonuclear neutrophilic leucocytes. Along* the borders of the abscess wall there was a large amount of fibrillated fibrin in the meshes of which there were variable numbers of polymorphonuclear neutro- philes. The abscess wall was composed of a granulation tissue rich in fibroblasts, but rather poor in blood-vessels. This tissue was infiltrated with large numbers of polymor- phonuclear eosinophilic leucocytes and with a few plasma cells and polymorphonuclear neutrophiles. In sections of the wall of the abscess stained with eosin and methylene blue there were scattered chains of streptococci and a few short bacilli. The wall of the left fallopian tube was much thickened and showed areas of infiltration with lymphocytes, plasma cells, and eosino- philes. The lumen of the tube contained desquamated epithe- lium, leucocytes, and necrotic material. The mucosa was nearly everywhere destroyed and the submucosa was infiltrated with plasma cells and polymorphonuclear neutrophiles and eosinophiles. The uterus showed nothing of special interest. The submucosa of the bladder was infiltrated with large num- bers of plasma cells. The pelvic lymph glands showed marked hyperplasia and contained numbers of eosinophiles. Bacteriological' Examination. — Coverslip preparations made from the tubo-ovarian abscess, the kidneys, bladder, and peritoneum showed large numbers of biscuit-shaped cocci in pairs, usually arranged in chains of four, six, eight, or twenty elements, and surrounded by a wide and sharply defined cap- sule which stained well with Welch's capsule stain. (See photograph.) In the abscess and bladder there were also short capsulated bacilli, and in the peritoneal exudate large, stout capsulated bacilli of the size and appearance of B. aero- genes capsulatus. Aerobic Petri-plate agar cultures were made from the heart's blood, lungs, pleurae, liver, spleen, brain, kidneys, tubo-ovarian abscess, and bladder. The cultures from the brain, pleurae, left kidney, and liver remained sterile. In the cultures from the lungs there were 1 66 HOWARD AND PERKINS. fifteen colonies of Staphylococcus pyogenes aureus and ten colonies of a bacillus which proved on further study to be B. mucosus capsulatus. In the cultures from the peritoneum there were a few colonies of Staphylococcus pyogenes aureus, many of B. mucosus capsulatus, and a number of fine col- onies ; in those from the tubo-ovarian abscess and bladder there were colonies of B. mucosus capsulatus and a small number of fine colonies. Cultures from the heart's blood gave a few, and those from the spleen a large number of fine pale colonies. Analysis of the above shows that fine pale colonies grew either in pure culture, or mixed with other bacteria (staphylococci and B. mucosus capsulatus) from the heart's blood, spleen, peritoneum, tubo-ovarian abscess, bladder, and right kidney. The fine colonies in all these cultures were similar in every way macroscopically and microscopically. Secondary cultures from all these cultures were made and studied in various media, and found to be identical. The growth on plain agar, glycerine agar, and glucose agar (neu- tral to phenolphthalein) was identical, and after twenty-four hours in the incubator at body temperature was made up of a thin, very moist, transparent pale-gray growth. The colonies were small at first, but after forty-eight hours reached the diameter of three millimetres. They were slightly raised, very moist, and had the appearance of dewdrops. Single colonies were round in outline, with regular edges. When magnified fifty times they were of a bluish color and rather coarsely granular, the granules lying in parallel lines. The water of condensation was cloudy and somewhat viscid. There was a growth of fine grayish colonies along the line of the stab. Under the No. 7 lens the young colonies were seen to be composed of long twisted colonies of cocci. No. gas was formed. On alkaline, plain and glucose, agar slants, there was a thin watery transparent grayish growth. Separate colonies showed a concentric appearance. Under the low power the colonies had coarse and fine granulations. There was a thin growth of fine colonies along the line of the stab, without gas forma- STREPTOCOCCUS MUCOSUS. 1 67 tion. The water of condensation showed a yellowish sedin ment. After forty-eight hours the colonies were larger and the growth very transparent and moist. On blood serum after twenty-four hours in the incubator there was a thin, moist, pale-gray transparent growth. The separate colonies were of pin-head size and were round with regular outlines. When magnified fifty times they were finely granular. On gelatine slants (neutral or slightly alkaline) after six days at room temperature there was a fairly well-marked pale-gray -transparent surface growth. Along the line of puncture there was a fairly profuse, feathery growth, without liquefaction. In neutral and in slightly alkaline bouillon after twenty- four hours at body temperature no growth was apparent; after forty-eight hours there was faint cloudiness, with a pale deposit on the sides of the tube. On potato (acid reaction) there was no visible growth and coverslip preparations were negative. Preparations made from the water below the potato showed streptococci. Plain and glycerine agar and bouillon -|-i.S* showed no growth; a slight growth was found on glucose agar -j-i-S. Litmus milk was unchanged after twenty-four hours in the incubator ; on the fourth day there was slight acidity, but no coagulation. Capsulated biscuit-shaped streptococci in long chains were present in small numbers. Optimum growth took place in media made according to the directions in the "Procedures recommended for the Study of Bacteria " (American Public Health Association Report, September, 1897) and rendered neutral to phenolphthalein and at a temperature of 37° C. It grew fairly well, but slowly, at room temperature. The organism grew both aerobically and anaerobically. No indol was produced in bouillon cultures. It was not motile. Morphology . — The morphology of the organism was very similar for the various media. It occurred as a coccus, in chains, the elements of which were arranged in pairs. The number of cells in each chain varied from two to one hun- 1 68 HOWARD AND PERKINS. dred or more. The pairs showed marked similarity to the gonococcus, their elements being biscuit-shaped, separated by a slight interval, and with the flat surfaces towards one another. The pairs, including the interval between the in- dividuals, measured 1.25 to 1.75 m. in length, and 0.5 to 0.75 m. in width. When the excess of light was cut off a definite halo was seen about each chain. Special capsule stains failed to stain any definite area, but numerous small, deeply- stained granules were to be seen within the halo, especially near its outer borders. The halo varied in width from 1.5 to 3.0 m., and the granulations were not constant. This halo was more marked about organisms grown on human blood serum than about those grown on plain glycerine or sugar agar, gelatine, or bouillon. Cultures in litmus milk showed the same type, but in this medium the capsules could be sharply stained by the usual methods. The refusal of the capsule to stain when the organism was grown on artificial media was in striking contrast to the action of capsule stains on the capsules of the streptococci obtained from the tubo- ovarian abscess, and the peritoneal and other exudates, as well as from the animal experiments. The streptococci from both the cultures and from the tissues stained deeply with Gram's and Weigert's stains, as well as with the ordinary aniline dyes. Animal Experiments. I. During the autopsy a guinea-pig was inoculated intra- peritoneally with 0.3 cc, of the peritoneal exudate. The animal became comatose in thirty-six hours, and died in forty-eight hours from the time of inoculation. The autopsy showed no lesion at the point of inoculation, and no subcutaneous infiltration was to be seen. The peri- toneal cavity contained between fifteen and twenty cubic centimetres of a cloudy grayish seropurulent fluid, with numerous flakes of fibrin. The most notable characteristic of this fluid was its extreme viscidity, which was such that it could be drawn out into threads several centimetres long. The abdominal viscera were heavily coated with grayish STREPTOCOCCUS MUCOSUS. 1 69 yellow fibrin, which could be readily stripped off, leaving a deeply congested surface beneath. The abdominal and tho- racic organs showed no macroscopic changes other than in- tense congestion. The inguinal and mesenteric glands were distinctly enlarged. Coverslips and cultures were made from the peritoneal ex- udate, and from the heart's blood, the liver, and the spleen. Coverslips from the organs were uniformly negative. Cover- slips from the peritoneum showed very numerous biscuit- shaped cocci in pairs, usually arranged in chains of four, six, or more elements up to twenty or thirty, and surrounded by a wide and sharply defined capsule. Cultures from the organs were negative, while those from the peritoneal exudate showed a pure growth of a streptococcus. The growth on various media was identical with that from the plates at the autopsy. Forty-eight hours later a series of inoculations was begun to test the virulence of the organism. The series included six guinea-pigs, four rabbits, and one white mouse. The following is an abstract of the results of these experi- ments : Guinea-Pig II. {a.) May 7. One cc. 24-hr. bouillon culture from the growth obtained from guinea-pig I., intra-perito- neal inoculation. (6.) May 12. One cc. 48-hr. glucose bouillon culture of same, (c.) May 18. Two cc. 24-hr. glucose bouillon culture of same. Died May 28. The autopsy findings were the same as- in the case of the first guinea-pig. Although three successive inoculations were necessary to cause the death of this animal, the virulence of the organism was reestablished. Guinea-pigs III., IV., V., and VI. were inoculated intra- peritoneaHy with fresh peritoneal exudate from previous cases, in decreasing amounts from i to 0.25 cc, and died in from twenty-four to seventy-two hours. Rabbits II. and IV. were each inoculated in the ear vein 170 HOWARD AND PERKINS. with 0.5 cc. of fresh exudate from the peritoneum or subcuta- neous tissue of guinea-pig and died in twenty-four to forty- eight hours. The autopsy findings were similar to those of guinea-pigs I. and II., except that in all subsequent cases the organism was recovered in pure culture from the organs as well as from the exudates. The following are the variations from the usual type : In guinea-pig IV. there was a marked exudate into the pleural cavities, which was of the same type as the peritoneal exudate. In guinea-pig V. there was marked thickening and oedema of the subcutaneous tissues of the abdomen. Bloody fluid, which contained large numbers of capsulated streptococci, ex- uded on section. Rabbit III., inoculated subcutaneously, showed a slight peritonitis and a marked serous pericarditis, from both of which pure cultures of the capsulated streptococcus were ob- tained. The seat of inoculation showed no especial lesion. In mouse I. none of the serous cavities showed exudate, and cultures from them were negative, while on the other hand cultures from the organs showed the wide distribution of the streptococcus. Rabbit I. was inoculated in the ear with 0.5 cc. of a twenty-four-hour glucose bouillon culture from guinea-pig I. The animal did not die, but extensive local inflammation and abscess formation took place at the base of the ear. A portion was excised under ether for microscopical examina- tion. The ear healed completely after some weeks. Portions of the various organs of these animals were hard- ened in Orth's fluid. Sections were stained with hjematoxy- lin and eosin, methylene blue and eosin, and Weigert's fibrin stain. Sections from the hearts, lungs, and kidneys showed marked congestion. Sections from the livers and spleens showed marked congestion, but although streptococci were found scattered through the tissues, no other changes of importance were noted in the organs themselves. The exter- nal surfaces of many of the organs were covered with an ex- STREPTOCOCCUS MUCOSUS. I /I udate of varying thickness up to two millimetres. In the cases which died in from twenty-four to forty-eight hours the exudate was composed almost entirely of streptococci with a variable amount of granular and fibrillated fibrin. Most of the few cells which were present were coarsely granular oxyphiles. In the animals living several days, the cellular exudate was much more marked, and in one case was begin- ning to organize, the process starting from the capsule of the liver. In no case was phagocytosis observed. The organisms did not stain with haematoxylin, but were brought out sharply with methylene blue and eosin, and Weigert's fibrin stain. Guinea-pigs and rabbits inoculated with one and two cubic centimetres of filtered bouillon cultures remained well, show- ing no signs of illness. Summary. — In this case, therefore, there was a mixed in- fection with a capsulated- streptococcus, B. mucosus capsu- latus and B. aerogenes capsulatus. The streptococcus was evidently the most numerous and most important of the organisms present. To it we are inclined to attribute not only the acute lesions, peritonitis, and splenic tumor, but the tubo-ovarian abscess and possibly the chronic nephritis aS well. Schiiltz's (loc. cit.) coccus was oval, and occurred in pairs and sometimes in chains. It did not stain by Gram's method, and according to Frosch and Kolle* should be classified mid- way between the Pneumococcus and the bacilli of the chicken cholera group. The coccus of Poels and Nolen grew in pairs and chains with a transparent capsule which stained with dif- ficulty. It grew like B. Friedlanderi in gelatine (nail-shaped growth) and did not stain by Gram. These two organisms probably belong to the same group, and are possibly identical ; the first was pathogenic for horses, rabbits, guinea-pigs, mice, and the second for cows, dogs, rabbits, and guinea-pigs. They are evidently quite different from the organism of our case. The streptococci of Bonome and Binaghi, however, present a number of points of likeness to our streptococcus and to each other. The streptococcus of Bonome was ob- 1/2 HOWARD AND PERKINS. tained from the meningeal exudate and lungs of six cases of epidemic meningitis cerebrospinalis occurring in Padua. It occurred as an oval coccus, isolated, in pairs, and in chains. The individual organisms varied in size and were always extracellular. It was encapsulated in cultures on certain media and in the exudates of animals, but not in those of man. It stained with the aniline dyes and by Gram's method. When grown in bouillon, the medium was rendered diffusely cloudy with a flocculent sediment, the organisms growing in long chains without capsules. On agar-agar the colonies were said to be characteristic, being round and sharply defined. When magnified fifty or sixty diameters the deep colonies were oval, markedly granular, of a light green color, and had sharp borders. The superficial col- onies were granular in the centre, but showed wavy concen- tric "layers" at the periphery, having the appearance of highly refractile filaments. In streak cultures, both on the surface and in the depth, the growth was in the form of a semitransparent coating bordering the line of inoculation like a halo for the distance of 0.5 to i mm. or more. In agar- agar stab cultures, there was a thin dirty gray growth along the line of the stab, without any surface growth. Coverslip preparations of this growth showed streptococci surrounded by a halo, which would not stain by any method tried. The growth was luxuriant on glycerine agar; the organisms forming long chains in which the individuals varied in size and shape, some forms being triangular (biscuit shaped?). The coccus failed to grow on blood serum, gelatine, and potato. On artificial media it died out in a few generations and its virulence was rapidly lost. White mice and rabbits were quite susceptible and showed fibrinous exudates upon serous membranes. In mice there was no invasion of the blood, but in rabbits capsulated streptococci were found in this fluid. Guinea-pigs and dogs were more resistant, but were killed and showed local fibrinous and gelatinous exuda- tions. Binaghi's capsulated streptococcus, found in spontaneous peribronchitis and multiple pulmonary abscesses of a guinea- STREPTOCOCCUS MUCOSUS. 1/3 pig, occurred in pus and pus cells, in pairs and short chains surrounded by a clear capsule which stained deeply (carbol- fuchsin). In bouillon cultures there was considerable growth, made up of streptococci. On agar at a temperature of 37° C. there was a growth of slowly developing dewdrop colonies. There was no growth on gelatine. The organism stained by Gram's method. Two guinea-pigs inoculated with the pus died in four days, and showed at the point of injection a mucoid and hemorrhagic diffuse oedema. The liver, kidneys, and spleen were enlarged. A guinea-pig , inoculated with cultures remained alive. It seems probable that the strepto- cocci of Bonome and Binaghi and the organism of our case belong to the same group, and are possibly varieties of Streptococcus lanceolatus. Their points'of resemblance are the formation of moist, transparent colonies on solid media, the growth in chains, which are capsulated in certain media and in animals, and their reaction to Gram's stain. They differ, however, in various ways. The colonies vary some- what in shape and appearance. The description of the coccus of Binaghi is too meagre for accurate comparison with the others. The deep agar colonies of Bonome's organ- ism mre said to be ball shaped. The peculiar wavy concentric " wrinkling " of the periphery of the superficial colonies of the organism were not present in ours. Unlike the other two streptococci, ours grew well on blood serum and on gelatine. Milk and potato cultures were not made by Bonome and Binaghi ; our organism grew in the former, and in the water about the latter. All three organisms are very much alike in their effects upon animals, in which they differ from those of Streptococcus pyogenes ; but do not vary markedly from those of Streptococcus lanceolatus. Here, however, they show certain differences, for two of them (Bonome's and our own) produce a hemorrhagic and gelatinous or mucoid oedema in the subcutaneous tissues. We append a table showing the main characteristics of the three organisms. From the foregoing it appears that there is a small group of pathogenic capsulated streptococci, characterized by the viscidity of their growth, and by the formation of gelatinous 174 HOWARD AND PERKINS. exudations in animals. For this group, composed of the streptococci of Bonome, Binaghi, and our case, we propose the name STREPTOCOCCUS MUCOSUS. REFERENCES. 1. Ref. in Centralb. f. Bakteriologie, Bd. I., s. 393. 2. Fortschritte d. Med., Bd. 86, s. 217. 3. Ziegler's Beitrage, 1890, Bd. VIII. , s. 377. 4. Centralbl. f. Bakteriologie, 1897, Bd. XXII., s. 273. 5. American Public Health Association Committee's Report, 1897. 6. Flugge. Die Microorganismen, 1898, II., s. 161. DESCRIPTION OF PLATE Vila. Streptococcus mucosus, from peritoneal exudate, X. 1200. Described by Bonome. Binaghi. Name. Streptococcus der meningitis cerebro-spinalis epidemica. Streptococcus capsulatus. Howard and Perkins. Streptococcus mucosus. Source. Meningitis. Spontaneous peribronchitis, and pulmonary abscesses of a guinea-pig. Morphology. Large, round, or oval cocci, isolated or in pairs or chains. Grown in agar they are surrounded by a halo. Encapsulated in rab- bit's blood. Cocci in pairs and in short chains, surrounded by a clear deeply staining cap- sule. Tubo-ovarian abscess, perito- nitis, spleen, .and heart's blood of a woman. In the lesions they occurred as biscuit-shaped cocci, in pairs, and in chains of from 4 to 20 or more ele- ments surrounded Jjy a dis- tinct capsule which stained readily. In cultures the organisms occurred in chains of from 4 to lOO elements, were biscuit-shaped, and were surrounded by a dis- tinct halo which did not stain, except when grown in milk. Growth in Agar. Round sharply defined colonies. The deep colonies when magnified 50 to 60 times were granular, light green in color, with sharp borders. Su- perficial colonies were granular in the centre, but showed concentric wavy layers at the per- iphery, giving the appearance of highly refractive filaments. Streak culture showed a transparent coating. Stab cultures showed a thin dirty gray growth along the inoculation line without devel- opment on the surface. The growth was luxuri- ant on glycerine agar. On agar in incubator at 37° C. there were slowly developing transparent dewdrop colonies. On neutral agar (plain, glycerine, and glucose) the growth after 24 hours at body temperature was thin, very moist and transparent, and grayish- white colonies reached 3 mm. in diameter. Microscopically separate colonies were bluish, coarsely granular, regular. In the stab there was a growth of fine grayish colonies. Water of condensation was viscid. Blood Serum. No growth. Not tried. Like agar growth well marked. Gelatine. No growth. No growth. Growth well marked after six days at room temper- ature. No nail-shaped growth in stab cultures. No liquefaction. Bouillon. Media ren- dered opaque throughout; flocculent sediment. Considerable growth which showed strep- tococci in long chains. No growth. Not tried. No apparent growth in 24 hours. In 48 hours faint cloudiness with a pale deposit on the sides of the tube. No growth on potato, but some multiplica- tion in the water of the bottom of the tube. Milk — in 4 days slight acidity, no ccagula- tion. No rec Nci rec Slii Gelatine. "No growth. No growth. Growth well marked after six days at room temper- ature. No nail -shaped growth in stab cuhures. No liquefaction. Bouillon. Media ren- dered opaque throughout; flocculent sediment. Potato. No growth. Acid Production. Not recorded. Indol. Not recorded. Considerable growth which showed strep- tococci in long chains. Not tried. Not recorded. Aerobiosis. Faculative an- aerobe. Not recorded. Gram's Stain. Positive. Aerobe. No apparent growth in 24 hours. In 48 hours faint cloudiness with a pale deposit on the sides of the tube. No growth on potato, but some multiplica- tion in the water of the bottom of the tube. Milk — in 4 days slight acidity, no coagula- tion. Slight. None. Optimum Tempera- ture. Body tem- perature. Pathogenesis. Positive. Faculative an- aerobe. Positive. Body tem- perature. In white mice and in rabbits it caused fibrinous exuda- tions. No organisms found in blood of former, but many in that of latter, in which they were encap- sulated. Guinea-pigs and dogs were more resistant, but showed local fibrinous and gelatinous exudations. Organisms extra cellular. The virulence was rapidly lost in artificial cultures. Body tem- perature. Two guinea-pigs inoculated with the pus from the pul- monary abscesses died in four days, showing enlarge- ment of the liver, spleen, and kidneys, and a mucoid hemorrhagic diffuse oede- ma at the point of inocula- tion. A guinea-pig inocu- lated with cultures did not die. Pathogenic for white mice, guinea-pigs, and rabbits, producing local fibrinous mucoid and hemorrhagic ffidema and fibrinous and mucoid exudations on the serous membranes. No phagocytosis. Large num- bers of capsulated strepto- cocci in the lesions. The virulence was rapidly lost in artificial cultures. When inoculated from animal to animal the virulence was markedly increased. JOURNAL OF MEDICAL RESEARCH. Vol. VL, Plate Vila. Fig. Howard and Perkins. S. Mucosus. [Reprinted from the Cleveland Journal of Medicine, June, 1901.] Report of Cwo Cases of Hngiosarcoma of the Brain. BY ROGER G. PERKINS', M. D., CLEVELAND, Resident Pathologist to the Lakeside Hospital ; Demonstrator in Pathology Western Reserve Medical School. [From the Pathological Laboratory of Lakeside Hospital.] NEXT to the localization of tumors of the brain and their relation to the tracts of fibers which connect the various parts of the brain with the different organs of the body, perhaps the most interesting study of these lesions is their origin. The greater part of the cases of brain tumor which are found in the literature are little more than clinical reports of the symptoms and ante- mortem localizations, with a brief statement as to whether the diagnosis was or was not confirmed at operation or autopsy. The pathologic report is ordinarily a mere statement that the tumor was sarcoma, or glioma, or, more frequently, gliosarcoma. This last term, as some of the later writers, potably Taylor (1), have emphasized, is rarely, if ever, an accurate one. The brain is essentially epiblastic in its origin, and if the classification of tumors according to their embryologic sources is to be adhered- to, the only tissues in the brain from which sarcoma may arise are the membranes on the outside and the vessels and their sheaths within the brain substance, as these include all the mesoblastic tissues which have to do with the structure of the organ. While the glia tissue subserves the same function to the brain as connective tissue does to the body, it is an integral part of the tissue of the cerebral vesicles, and so must be purely epiblastic. Furthermore it reacts differently from connective tissue, especially chemically, in relation to special stains. It is of course conceivable that a previously formed glioma might be invaded, whether by metastasis, direct invasion, or even possibly by a primary forma- tion about the vessels, by a definite and typical sarcoma. But if such cases have occurred, they have never been reported as such, and it is furthermore probable that the line of demarkation between the two types of new growth could be clearly made out by the microscope. The following case is of interest as one in which the diagnosis of angio- sarcoma can be definitely made, but in which no trace of special connection with the blood-vessels, either endothelial or perithelial in type, could be made out. Clinical History — L. R., female, white, 25 years old, married. She was admitted to the Lakeside Hospital, in the service of Dr. Hunter Robb, to whom I am indebted for the clinical report, on February 27, 1899, complain- ing of persistent vomiting. The family history was negative, and no record of tuberculosis, -tumors, or insanity was obtainable. The patient had always been healthy, with the 2 Angiosarcoma of the Brain. exception of the usual diseases of cliildhood, and she denied venereal disease. The present illness dated from about four months previous to admission, at which time she began to vomit, usually in the morning. The attacks bore no special relation to the taking of food, but became more and more frequent, until during the three weeks previous to admission she was unable to retain any food long enough to digest it. She had not menstruated since the onset of the vomiting, and believed herself to be pregnant. On admission, she was in poor physical condition, but beyond the determination of a three to four months' pregnancy, nothing of interest was Case I — Angiosarcoma of frontal lobe ; ^ natural size. found on examination. The presence of the fetus, associated with the vom- iting, kd to a diagnosis of vomiting of pregnancy and the patient was treated accordingly. For about a week she was much better, and only vomited once or twice, but on the seventh day after admission, she complained of severe headache and a boring pain in the left eye. External examination of the eyes showed nothing of interest, the pupils were equal, reacted well to light and accommodation, and the symptoms were relieved by the use of blue glasses. During the next week there was marked improvement ; the nausea disap- peared, and the patient was able to eat the ordinary ward diet without dis- comfort. On the fourteenth day after admission there was a sudden and Angiosarcoma of the Brain. 3 severe attack of vomiting, associated with marked signs of mental aberration. The patient was dull and listless, and on the sixteenth day the knee-jerk could not be obtained. She again complained of pain in the eyes, and an ophthalmoscopic examination was made by Dr. B. L. Millikin, who made a diagnosis of double optic neuritis. On the nineteenth day there was an attempt to induce abortion, in the hope of relieving the symptoms. The patient remained for some days in a partly unconscious condition, with invol- untary passage of urine^ and occasional twitchings of the arms. During this time the abortion was completed, and the uterus thoroughly cleaned out. There was slight improvement for a day or so, but the patient soon relapsed into a condition of partial, deepening to complete unconsciousness, and died on the thirty-fourth day after admission. Repeated examinations of the urine and the blood were negative. The temperature, pulse and respiration were about normal until the operation. Following this, they rose steadily until the temperature reached 104°. This was followed by a fall to 101°, coinciding with the temporary improvement. On the thirty-first day it rose to 104°, on the thirty-second to 104.5°, on the thirty-third to 105.8°, and on the thirty-fourth to 106°, taken one hour before death ; at this time the pulse was 154, and the respirations 56. The combination of headache, optic neuritis, and vomiting, together witli the fact that abortion failed to relieve the condition of the patient, led to the diagnosis of cerebral tumor. There were no motor disturbances until the very end of the disease, so the probable localization of the growth was thought to be one of the silent areas of the cerebrum. The patient was nor- mally rather dull, but so far as could be ascertained, showed no especial departure from her ordinary psychic condition. This led to a further locali- zation in the frontal lobe, probably the right frontal, as the patient had no afifection of the motor speech-center, and was right-handed. The question of operation was discussed, but she was not in a condition to give any hope of a successful result, and died without rallying. The autopsy was performed sixteen hours after death, the body being still sHghtly warm. Anatomic Diagnosis — Sarcoma of the right frontal lobe of the brain, without metastases; abortion in the fourth month; fibrinopurulent pleurisy with bronchopneumonia of the left lung; congestion of the liver, spleen, and kidneys ; cloudy swelling of the kidneys. Cultures from the left lung and pleura and from the uterus showed Bacillus mucosus capsulatus in pure culture. Cultures from other organs were negative. The description of the brain is copied from the autopsy protocol, the other details of which, being without present interest, are omitted. 4 Angiosarcoma of the Brain. External ExAMiNATiON-The skull was symmetric in shape, and of ordinary thickness. The dura was not adherent, either to the bone or to the pia. There was no excess of fluid at the base, and the cranial nerves and other structures at the base appeared normal. The brain was of ordmary size The blood-vessels of the pia-arachnoid were markedly congested, especially on the right side. The left side of the brain was normal on mspec- tion, but on the right side there was a marked flattening of the convolutions, which were almost obliterated over the anterior portion. On the orbital surface of the frontal lobe, 2 cm. from the inner edge of the hemisphere, and 3 cm. from the anterior tip of the inferior frontal lobe, there was a small dark spot, 1 cm. in diameter, firmer to the touch than the adjacent cortex. The consistency of the entire right cerebral hemisphere was distinctly greater than that of the left, and this condition was especially marked over the frontal and temporal lobes. Section — For various reasons it was necessary to cut up the brain at the time of autopsy. The convexities of the cerebral lobes were removed by a horizontal incision, and vertical sections were then made, about 1 cm. apart. The medulla, pons, cerebellum, and the left cerebral hemisphere showed nothing abnormal. On the right side, the occipital, parietal and temporal lobes showed nothing but marked congestion of the blood-vessels, both those of the pia-arachnoid and the branches of the cerebral arteries and veins in the brain substance. Section of the frontal lobes showed a tumor-mass involving the anterior portion, and extending back into the region of the Island of Reil. The mass was dark reddish-brown in color, very sharply marked off from the surrounding brain-tissue, and contained numerous vessels in some of which were seen grayish-red adherent plugs. The center was soft and apparently necrotic. In the anterior portion there was a smooth-walled cavity, apparently cystic, containing dark, thick fluid. The brain was hardened as a whole without further dissection. Examination of the frontal lobes of the right side, after hardening, showed the localization of the tumor to be as follows : The mass was roughly oval in shape, its long diameter lying dorso- ventrally. It measured 6 cm. in length by 3.5 cm. in breadth at its widest part. Ventrally, it extended to a point 1 cm. from the surface of the cortex, dorsally, to a point opposite the lower end of the precentral sulcus, mesially, to within 1 to 2 cm. from the mesial surface of the frontal lobe, and externally to within 1.5 to 2 cm. of the cortical surface. Its lowest point involved the cortex about the dark spot above mentioned, and the upper border reached the plane of the lower end of the fissure of Rolando, With the exception Angiosarcoma of tlic Brain. 5 of the small spot on the orbital surface of the frontal lobe, the tumor lay wholly within the white matter of the hemisphere. There was everywhere a sharp line of demarkation between the tumor and the surrounding tissues, and over much of the surface the mass could be readily peeled out ; there was, however, no definite capsule to be made out. The brain-tissue in the immediate neighborhood showed marked congestion, and seemed to be compressed by, rather than involved in the growth. The structures in the brain were much displaced by the tumor, and in many places the exact relations were difiicult to determine. Ventrally, the only gross change apparent was a marked compression both of cortex and of white matter, tending to obliterate the sulci between the convolutions. Dorsally, the mass lay in the neighborhood of the basal nuclei, and thus became relatively more important. The entire corpus striatum was lifted up toward the summit of the brain, and displaced backward as well, though it was not in any way involved in the tumor mass, being separated from it by a layer of white matter from .5 to 1 cm. thick. The right lateral ventricle was much compressed and displaced upward and backward, but had no connection with the cyst-cavity. The general impression given was that the tumor had begun ventrally, where the greater part of the degenerated area was situated, and had grown dorsally, pushing the various structures of the hemisphere before it. The dorsal border approached the motor area, but seemed to have pushed it away, rather than to have involved it, so that the tumor lay wholly within the anterior silent area, Flechsig's "anterior center of association." The ventral part of the tumor was occupied by a cyst-cavity, containing dark fluid and small pieces of necrotic material. This cyst occupied the upper ventral portion of the tumor, and measured 3 by 3 by 1 cm. in size. It extended ventrally almost to the tip of the tumor, and dorsally about half way to the end. Microscopic Examination — Pieces were removed from various portions of the tumor and from the cortex for microscopic examination. The methods of examination were as follows : 1. Tissues hardened in 4% formaldehyd ; a. Hematoxylin and eosin. b. Methylene blue and eosin. c. Van Giesen's picric acid and acid fuchsin. d. Weigert's fibrin stain. e. Weigert's stain for elastic tissue. f. Potassium ferro- and ferricyanid. 6 Angiosarcoma of the Brain. 2. Tissues hardened by special methods for, a. Weigert's m_yehn-sheath stain. b. Mallory's neuroglia stains. 1. Modified Weigert method. 2. Phosphotungstic-acid hematoxylin. 3. Phosphomolybdic-acid hematoxylin. General Description — The brain-tissue in the immediate neigh- borhood of the tumor showed marked compression, the cells being apparently more numerous than in sections taken from a distance. The blood-vessels were much distended, and many of them contained thrombi, some recent^ made up of red blood-corpuscles, leukocytes and fibrin, while others showed beginning organization. There was a uniformly sharp line of demarkation between the tumor and the adjacent brain tissue, and in two places the pia-arachnoid of the inferior frontal sulcus passed down between the two. The microscopic character of the tumor varied markedly at different points, but its general characteristics were the same throughout. Sections from the various portions showed a very large number of cells, with a scanty stroma, pierced by numerous blood-vessels, both arteries and veins. The cells were large, irregular in shape, with large, round or oval vesicular nuclei, in which num.ferous karyokinetic fip-ures could be seen. These cells had no particular arrangement, and showed no tendency either to an alveolar struc- ture, or to a perithelia! arrangement about the vessels. The stroma was somewhat fibrillar in structure, and contained few cells. The blood-vessels, both arteries and veins, showed many thrombi in all stages, from the most recent to those which had become completely hyaline. Here and there throughout the sections, small emboli broken off from these thrombi were seen, stopping up the course of the smaller arteries. The vessel-walls in many places were increased in thickness, but bore no special relation to the sarcomatous tissue. The central part of the tumor showed complete necro- sis, and took no nuclear stains ; the inner and deeper portions of the necrotic mass showed no structure whatever, but at the edges the outlines of the cells and the blood-vessels could be readily made out, stained diffusely with the protoplasmic stains; All the blood-vessels in the necrotic area showed thrombi, and the proportion of thrombi in the vessels throughout the tumor was much greater near this area than elsewhere. The necrosis is therefore apparently due to the cutting off of the circulation, causing necrosis en masse, or coagulation necrosis. In the degenerated areas, in some of the thrombi, and also in a few of the cells in the immediate vicinity, there was a varying amount of brownish- yellow pigment, amorphous in type and resembling hematoidin. Angiosarcoma of the Brain. 7 In sections from the dorsal portion of tlie tumor tliere were numerous giant-cells, with large vesicular nuclei of the same type as those in the pre- vailing mononuclear cells. Some of these giant-cells contained 8 to 10 nuclei. In sections from the roof of the cyst it was found that the tumor-cells lay only a few rows deep, immediately adjoining the cyst, and were enclosed by a zone of intense congestion, with thrombi in many of the blood-vessels ; beyond this the brain was comparatively normal. Sections from the floor of the cyst, on the other hand, showed this to be composed entirely of tumor. Pieces were hardened by Mallory's special method for neuroglia, and stained in the various ways recommended by him. Unfortunately the autopsy Case I — Showing line of deiuarkation between tumor and adjacent tissue — 120 diameters, was SO long after death (16 hours), that the results of these stains, even in the normal cortex, were entirely inconclusive. Sections of the formalin-hardened brain stained by Van Giesen's method, showed a small amount of rather fibrillar tissue, lying among and between the cells. It took the same stain as the connective tissue of the vessel sheaths, and no connections between cells and fibers could be demon- strated. Sections stained with Weigert's fibrin-stain showed the presence of large amounts of fibrin in the thrombi in the blood-vessels. The separate fibers were rather thicker than usual, and showed a marked tendency to lie in 8 Angiosarcoma of the Brain. bundles, rather than to form a network. In many places fibrm centers, emanating from leukocytes could be readily made out. No bacteria of any kind could be found. Sections from various portions were treated by Weigert's latest method for elastic fibers, in view of the results obtained by Alice Hamilton.^ The elastic tissue in the walls of the blood-vessels took the stain sharply, but no bundles of elastic tissue were found elsewhere. The use of Weigert's method for the staining of medullated fibers failed entirely to demonstrate the presence of nerves in the tumor at any point. In order to determine whether the pigment which occurred in and about the degenerated areas was a blood-derivative, or melanin, sections were treated for 6 to 8 hours with a 1% solution of potassium ferro-ferricyanid, and placed in acid alcohol for the same time. The pigment stained diffusely blue, which fact excluded melanin, and therefore melanosarcoma. In order to determine whether ihe slight fibrillar network seen with Van Giesen's stain had any connection \\'ith the cells, teased specimens were made and stained with phosphotungstic-acid hematoxylin ; the cells were found to be of the ordinary polymorphous type, with vesicular nuclei, and in no case were fibers found extending from them. Conclusions — -In view of the character and arrangement of the tumor-cellSj their lack of processes, the minimal amount of stroma, and the sharp line of demarkation between tumor and brain-substance, the weight of evidence seems to be in favor of the diagnosis of sarcoma ; the abundance of vessels throughout the mass, although no distinct relations between them and the tumor could be made out, suggests an angiosarcoma. On the other hand, the absence of the infiltration of the surrounding tissues, the complete lack of nerve-cells and fibers, and the apparent absence of glia fibers, are against a diagnosis of glioma. It is true that the failure to stain the glia fibers in portions of the brain where they are known to exist, prevents an absolutely final conclusion, but a view of all the points noted leads to the belief that the tumor is an angio- sarcoma, and bears no special relation to the neuroglia tissue of the brain. In connection with the case reported above it is of interest to consider another case of intracranial tumor somewhat similar to it in which the micro- scope places the diagnosis beyond doubt. The specimen was found in the pathologic museum of the Western Reserve Medical College where it had been placed a number of years ago by the late Dr. I. N. Himes. The history accompanying the specimen states that it was obtained from a woman 45 years old who had paralysis and spasmodic muscular twitchings of the left arm. Angiosarcoma of the Brain. 9 At autopsy a tumor was found lying under the center of the left cerebral hemisphere, about the shape of a flattened orange and measuring 9x6x5 cm. in its various diameters. The surface was irregularly lobulated, being apparently constricted by narrow fibrous bands. The consistency was firm, and the whole mass could be readily shelled out from the surrounding brain- tissue. There is no record of the exact relations, and the specimen preserved consists only of the tumor and a thin slice of brain-tissue overlying it. The hardened specimen was dense in consistency, and cut somewhat like fibrous tissue. It was surrounded by a definite fibrous capsule. The speci- men was hardened and preserved in alcohol. Pieces were removed from various portions of the tumor and brain-tissue and imbedded in celloidin. Sections were stained with hematoxylin and eosin, methylene blue and eosin, and \'an Giesen's picric-acid fuchsin. Microscopic Examination — Sections from the surface of the tumor mass show a fibrous capsule of varying thickness, usually about 1 mm, from which fibrous bands run down into the tissue at points and lines corre- sponding to the depressions and lobulations noted in the gross description. The growth is very cellular, and the most striking point about its appear- ance under the low power is the arangement of the cells. These lie in num- erous concentric whorls of various size, some almost circular, others oval, while still others are much elongated and elliptical in shape. Closer study with higher powers shows that a very large number of these whorls bear definite relations to blood-vessels of varying size which constitute the centers about which the cells are arranged. The angles at which the vessels are cut modify the shapes of the whorls, as the cells lie parallel to the lumina of the vessels. In some of these vessels the lumina persist, in others they are filled with cells lying in concentric layers ; in others they are filled with hyalin material often arranged in concentric layers and staining a deep pink with Van Giesen's stain. In some of these concentric hyalin masses a few cell- outlines can still be traced. In many cases in sections stained with hema- toxylon and eosin, or with methylene blue and eosin, no vascular center can be definitely made out even with the oil immersion, but in many of these areas the use of Van Giesen's stain shows some remnant of tissue staining pink, usually with no special form. It seems at least probable that these are rem.nants ofthe original central blood-vessel whose walls havebeen infiltrated and compressed by the tumor-growth until there is only this trace left behind to show the origin. In other words such places seem to be an intervening stage between the whorls with a definite vascular center and those in which no trace of vessel persists. The cell proliferation can be traced directly from the adventitia into the tumor-mass, though it is interesting to note that the 10 Angiosarcoma of the Brain. hyperplastic tissue at once ceases to take up selective connective-tissue stains such as Van Giesen's. The cells composing the tumor are mainly of one type or are slight variants from it. They are large, oval, or spindle-shaped with large oval vesicular nuclei, and lie in a minimal amount of stroma. Some of the cells, especially those in the immediate vicinity of the vessels, are much elongated and flattened, while the cells which lie at the meeting-point of several whorls are often almost round. Here and there are scattered a few cells of another type, smaller, round or oval, with nuclei containing much more chromatin than the other cells, but with no demonstrable processes. There is nowhere any inflammatory reaction, nor any cell-infiltration, nor is there any excess of leukocytes in the vessels. No thrombi are seen in any part. The growth appears to have been slow, as there are very few nuclear figures to be seen. No nervous-tissue elements, whether ganglion-cells, glia- cells and fibers, or medullated fibers, could be demonstrated. The encapsu- lation and sharp differentiation from the brain-tissue, the vascularity, the abundance of cells with minimal stroma between them, and the absence of nervous elements, point to the diagnosis of angiosarcoma, while the definite arrangement of the cells about the blood-vessels, and their evident origin from the adventitia, place the tumor under the head of the perithelial angio- sarcomata. Inasmuch as a very small amount of brain-tissue in addition to the tumor was preserved, it was not possible to follow out the lesions causing the clinical symptoms. The nomenclature of tumors of the brain has been the subject of much discussion for a long time. Until the embryology of the cerebral vesicles was carefully worked out there was no definite basis on which to go, and solid tumors of the brain-substance as well as those of the membranes were put together under the head of sarcomata. Even since the cerebral vesicles have been finally classed as entirely epiblastic in origin, there is much diversity in definition even among the best authorities. A sharp dififerentiation between the terms sarcoma and glioma is rarely made, and the combined name gliosarcoma is still in common use. A brief summary of views of some of the more important writers will show this clearly. Orth^ states that if fibers predominate we have a pure glioma, but if there are many free cells we can diagnose gliosarcoma. Thoma lays stress in his text-book on the epiblastic origin of glioma, but admits the possibility of gliomata undergoing a change and passing over into rapidly growing gliosarcomata. Birch-Hirschfeld also allows the term gliosar- coma, and makes as his only distinction between glioma and sarcoma the richer development of cellular elements and the greater size in the latter. Angiosarcoma of the Brain. 11 Oppenheim^ admits the possibility of mixed timiors, and Starr'' uses the term ghosarcoma in his text-book. Virchow** bases tlie diagnosis on the character of the cells. In a glioma the cells have processes, nerve-cells and fibers are absent, the cells are few and the ground-substance is permanent. In sarcoma on the other hand there are numerous polymorphous cells without processes, and the cement substance is scanty. Ziegler^ states in the last edition of his text-book that true gliosarcoma may occur when a perivascular, adventitial cell-growth whose product forms an integral part of the tumor enters a Case II— Showing concentric arrangement of ceUs — 800 diameters. glioma. Stroebe^" agrees with Ziegler. Thomas and Hamilton are of the opin- ion that all new growths of the brain are to be regarded as gliomata unless they can be proved to conform the vessel-walls. They use the term gliosar- coma however in describing a rather cellular tumor in which the cells pos- sess definite processes. Taylor in his extensive article on neuroglia in con- nection with which he reports two cases distinguishes between gliomata in which the fibers are still undifferentiated and older forms in which thev are more or less completely so. In view of the embryonic origin of the tissues involved, he excludes gliosarcoma as a definite tumor, although he also states that there is as yet no fundamental distinction between glioma and sar- coma. Von Lenhossek^i states that a sarcoma cannot be a glioma. 12 Angiosarcoina of the Brairi. LITERATURE. 1. Taylor— yowrna; of Experimental Medicine, 1890. 2. Thomas & B"amilton— JburnaZ of Erperimenidl Medicine, 1899. 3. Oi'th.— Pathologic und anatomische Diagnostic. Berlin, 1894. 4. Thoma — Lehrbuch d. paihologischen Anatomie, 1894. 5. Birch-Hirschfeld — Lehrbuchd. pathologischen Anatomie, Bd. II, S. 332- 6. Oppenheira — Die Geschwuelsie des Gehirns. Spezielle Pathologie u. Therapie — Nothnagel, Bd. IX, S. 1. 7. Starr— Brain Surgery ^ New York. 8. Virchovr— Die krankhaffe?i Geschivioelste. 0. Z\eg\ev—L€?irbuch d. paihologischen Anatomie, Bd. II, S. 363. 10. Stroebe — Ueber Entstehung von Band. Gehirnglioma. Ziegler's Beitraege z. path. A nat. u. allg. Path., 1895, S. 405. 11. Von 'Ijenhosse]s.-- Die feinere Hau des Nervensy stems. Berlin, 1895. To Hunter Rob^, M.D. Visiting Gynecologist to the Lakeside Hospital, Cleveland, 0.: I have the honor to submit the following brief resume of the work done in the Gynecological Department of the Pathological Laboratory of Lakeside Hospital during my term as Resident Gynecologist from January 1, 1898, to January 1, 1901. My hearty thanks are due the assistants connected with the Gyne- cological Service for their valuable aid in carrying out this work, and also to Dr. W. T. Howard, Jr., the Director of the Pathologi- cal Department of the Hospital, for kindly advice. William H. Weir, M.D., Resident Gynecologist, Lakeside Hospital. f( A REPORT IN GYNECOLOGICAL PATHOLOGY.^ ■WILLIAM H. WEIR, M.D., Late Resident Gynecologist to the Lalteside Hospital and Demonstrator of Gynecology, Western Reserve University, Cleveland, O. The routine pathological examination of all tissues removed at operation has been carried out in the gynecological depart- ment since the opening of the new hospital building in January, 1898. From this date to January 1, 1901, 523 operative cases supplied tissues for examination, a brief resume of the findings being given in this paper. Of course operations such as ventral suspensions of the uterus in which no tissues were removed, or vaginal punctures with the evacuation of pelvic abscesses in which merely cultures were made, are not included in these 523 cases. Although frequently the microscopic findings in a single instance may seem of little interest, the study of any extensive series of similar cases cannot fail to render the observer familiar with the various stages of similar pathological processes and thus lead him to a correct interpretation of the appearances in a given case, which might still remain obscure to one not familiar with the possible changes in these tissues. Valuable statistics may also result from this routine work when it has been carried out for a sufficient time, and occasionally special interest will be aroused by finding a totally unsuspected pathological condition. Moreover, the microscope sometimes affords us the means of making an early diagnosis, so that operative measures can be instituted at a period when complete removal of the disease can still be hoped for. Examination of Tissues. — The tissues were fixed as soon as pos- sible after operation. Curettings were hardened usually in alcohol, but formalin has been employed lately with good results. Larger specimens, such as ovaries, were sketched and the gross appearances described, the weight and measurements being also given. ""Cover-slips and cultures of pus, if present, were made and fluids from cysts examined. As a rule these specimens were ^From the service of Dr. Hunter Robb and the Pathological Labora- tory of the Lakeside Hospital, Cleveland, 0. Reprinted from the American Journal of Obstetrics, May, 1