STUDIES 
 
 ^ 
 
 FROM 
 
 THE BENDER. 
 
 Hygiervic Laboratory 
 
 ALBANY. N. Y. 
 
 REPRINTS 
 
 Volume I 
 
 1904 
 
 ay twicrosom 
 
This book was digitized by Microsoft Corporation in 
 
 cooperation witli Cornell University Libraries, 2007. 
 
 You may use and print this copy in limited quantity 
 
 for your personal purposes, but may not distribute or 
 
 provide access to it (or modified or partial versions of it) 
 
 for revenue-generating or other commercial purposes. 
 
 Digitized by Microsoft® 
 
_ . Cornell University Library 
 
 RA 422.S.B45 
 
 V.I 
 
 Studies ... reprints. 
 
 3 1924 003 256 678 
 
 Digitized by Microsoft® 
 
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.archiv^.9rg/^^ai|s^u31 924003256678 
 
Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
STUDIES 
 
 FROM 
 
 THE BENDER 
 
 Hygienic Laboratory 
 
 ALBANY, N. Y. 
 
 REPRINTS 
 
 VOLUME 
 1904 
 
 FORT ORANGE PRESS 
 
 BRANDOW PRINTING COMPANY 
 
 ALBANY, N. Y. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 

 __^,,>-i-^- . 
 
 
 
 Y ^ 
 
 
 J 
 
 
 .j/' ■■''■' 
 
 
 ''/ ' 
 
 
 
 
 •.:P 
 
 
 
 -■^ -^ — . 
 
 
 / 
 
 
 ^iJij£. 
 
 
 i 
 
 \ '^ 
 
 -/ 
 
 ■ -v '_'' 
 
 
 
 
 ' ^i?" 
 
 
 
 - ._,^---=^"' 
 
 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
The Organization ^ the Laboratory 
 
 Trustees 
 
 Dr. A. Vander Veer Dr. Henry Hun 
 
 Dr. George E. Gorham Hon. Robt. G. Scherer 
 
 Harry H. Bender 
 
 Director 
 
 Richard M. Pearce, M.D. 
 
 Assistants 
 
 E. MacD. Stanton, M.D. Charles K. Winne, Jr., M.D. 
 
 Assistant Chemist and Bacteriologist 
 
 (In charge of the Albany FiUration Plant) 
 
 Leonard M. Wachter 
 
 Voluntiuy Assistants 
 
 Elliqe McDonald, M.D., i903-'o4 
 Edwin F. Sibley, M.D., i904-'o5 
 
 In Connection with the Laboratory Courses of the Albany Medical College 
 
 Richard M. Pearce, M.D., Prof essor of Pathology and Bacteriology 
 E. MacD. Stanton, M. D., Lecturer in Normal Histology and in the 
 
 Anatomy of the Nervous System 
 Arthur W. Elting, M.D., Lecturer in Surgical Pathology 
 Arthur T. Laird, M.D., Lecturer in Clinical Microscopy 
 Herbert D. Pease, M.D., Lecturer upon Antitoxins and Immunity 
 Charles K. Winne, Jr., M.D., Instructor in Bacteriology 
 H. JuDSON Lipes, M.D., Instructor in Obstetrical Pathology 
 Harry W. Carey, M.D., Instructor in Surgical Pathology 
 Edwin F. Sibley, M.D., Instructor in Clinical Microscopy 
 Donald Boyd, M.D., Demonstrator of the Anatomy of the Nervous 
 
 System and Assistant in Normal Histology 
 
 Stenographer 
 
 Miss Maude Marshall 
 
 Preparator 
 
 Miss Helen R. King 
 
 Janitor 
 J. E. G. Roy 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
NOTE 
 
 ♦ * 
 
 Cms volume comprises the publications of the Bender Hy- 
 gienic Laboratory for the year i903-'o4. These have 
 appeared in various medical journals, but for the sake of uni- 
 formity in paper and press work, the reprints have been taken 
 from the Albany Medical Annals. The object of again present- 
 ing these studies is to bring them in collected form before those 
 especially interested in the laboratory or its work. As this is 
 the first attempt to establish an annual publication, which shall 
 adequately represent the work of this institution, a short histor- 
 ical sketch of the origin and development of the laboratory, 
 with a list of previous publications, has been included. 
 
 RICHARD M. PEARCE, 
 
 Director. 
 
 The Bender Hygienic Laboratory 
 Albany, N. Y., November, 1904. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
CONTENTS 
 
 Frontispiece — The Bender Hygienic Laboratory. 
 
 I. Historical Sketch (with portraits of Matthew W. Bender 
 and Dr. George Blumer). 
 
 George E. Gorham, M.D. 
 II. Annual Report of the Director. 
 
 III. Concerning the Specificity of the Somatogenic Cyto- 
 
 toxins. 
 
 Richard M. Pearce, M.D. 
 
 IV. The Experimental Production of Liver Necroses by 
 
 Haemagglutinins. 
 
 Richard M. Pearce, M.D. 
 
 V. Concerning Haemagglutinins of Bacterial Origin and 
 Their Relation to Hyalin Thrombi and Liver Necroses. 
 
 Richard M. Pearce, M.D., and 
 Charles K. Winne, Jr., M.D. 
 
 VI. The Complications and Degenerations of Uterine Fibro- 
 myomata. 
 
 Ellice McDonald, M.D. 
 
 VII. A Peculiar Hypertrophy of the Prostate Accompanied 
 by an Ascending Infection and Cysts in the Ureters, 
 with a General Discussion of Ureteritis Cystica (with 
 one illustration). 
 
 Harry W. Carey, M.D., and 
 Arthur T. Laird, M.D. 
 
 VIII. Infantile Inguinal Hernia of the Vermiform Appendix 
 with Adhesion to the Epididymis (with one illustra- 
 tion). 
 
 Edwin MacD. Stanton, M.D., and 
 Henry L. K. Shaw. M.D. 
 
 Digitized by Microsoft® 
 
IX. A Study of the Changes Occurring in the Endometrium 
 during the Menstrual Cycle (with one illustration). 
 
 H. JUDSON LiPES, M. D. 
 
 X. The Islands of Langerhans in Congenital Syphilitic 
 Pancreatitis (with one illustration). 
 
 Richard M. Pearck, M.D. 
 
 XI. Cancer of the Pancreas and Glycosuria; A Histological 
 Study of the Islands of Langerhans. 
 
 Richard M. Pearce, M.D. 
 
 XII. Experimental Pancreatitis Considered in its Relation to 
 Acute Lesions of the Pancreas in Man. 
 
 Richard M. Pearce, M.D. 
 
 XIII. The Pathology of Diabetes. 
 
 Richard M. Pearce, M.D. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
MATTHEW W. BENDER 
 
 Digitized by Microsoft® 
 
Keprinted from Albast Msdicai. Annalb, November, 1904 
 
 HISTORY OF THE BENDER HYGIENIC LABORATORY 
 By GEORGE E. GORHAM, M. D. 
 
 On June 27th, 1895, the corporation of the Bender Hygienic 
 Laboratory was , formed, the members being: Mr. Matthew W. 
 Bender, Dr. Albert Vander Veer, Dr. George E. Gorham, Dr. 
 Henry Hun and Hon. Robert G. Scherer. 
 
 The corporation was organized to take over from Mr. Bender 
 twenty thousand dollars, a sum which Mr. Bender had offered 
 to give for the erection of a laboratory building upon the condi- 
 tion that a suitable site be furnished. The board of trustees was 
 immediately elected, and to its care were entrusted the generous 
 gift of Mr. Bender and the carrying out of his wishes in securing a 
 suitable site and erecting the building. A grant from the city 
 of Albany for the spacious and beautiful grounds on Lake avenue 
 now occupied by the laboratory was promptly obtained. Messrs. 
 Fuller and Wheeler were employed as architects, and under the 
 special supervision of Mr. Harry H. Bender the building was 
 completed and dedicated on October 27th, 1896. 
 
 Dr. George Blumer, of the Johns Hopkins Hospital, was 
 appointed director and under his supervision the laboratory was 
 fully equipped with necessary apparatus. About one hundred 
 microscopes, necessary tables, lockers, etc., were placed in the 
 class rooms; steam and hot air sterilizers, microtomes, thermo- 
 stats, freezing apparatus, etc., were added for general pathological 
 and bacteriological work. 
 
 This generous outlay for the equipment of the laboratory was 
 furnished by the faculty of the Albany Medical College. Such, 
 briefly, is the history of the conception and birth of the Bender 
 Hygienic Laboratory. Sixteen months from the time the subject 
 was first presented to Mr. Bender for his consideration by his 
 family physician the building was completed, and opportunity for 
 one of the most important scientific works in the history of 
 Albany was given. 
 
 The first director of the laboratory. Dr. George Blumer, and 
 his assistants, Dr. Lartigau, Dr. Elting, Dr. Carey and Dr. Laird, 
 
 Digitized by Microsoft® 
 
under the direction of the Albany Medical College, arranged and 
 carried out a course of instruction in pathology, bacteriology, 
 histology and microscopy for the students of the Albany Medical 
 College. These valuable courses of instruction are continued 
 throughout each college year. They are, in fact, part of the 
 regular curriculum of the Albany Medical College, and classes 
 of forty or fifty students take the different courses annually. Dr. 
 Blumer resigned in 1903, and the work has been continued by 
 his successor. Dr. Richard M. Pearce, assisted by Drs. Stanton 
 and Winne. 
 
 The Albany filtration plant for filtering Albany's water supply 
 is under supervision of the laboratory, and daily examinations of 
 the water are made. 
 
 By special contract with the city of Albany cultures from all 
 cases of suspected diphtheria are examined and reported to the 
 attending physician. Sputum from suspected cases of tuberculosis 
 is also examined. Work of the same character is done for the 
 State Department of Health of New York. 
 
 The examination of milk and water supplies in case of epi- 
 demics throughout the State is a part of the work done at the 
 laboratory. 
 
 Hundreds of examinations of pathological specimens, blood, 
 sputum and urine are annually made for the hospitals of Albany, 
 while physicians in the surrounding country are constantly send- 
 ing specimens for examination. 
 
 Added to this general work of the laboratory is special research 
 work, some forty papers having been published during Dr. 
 Blumer's directorate, and eleven during the past year by the pres- 
 ent staff. 
 
 Thus may be sketched in a few words the scope of the work 
 of the Bender Laboratory. But these words indicate only inade- 
 quately the importance of this institution and the great value for 
 years to come of the gift of Mr. Bender in promoting the health 
 and welfare of the community. 
 
 List of Publications From the Bender Hygienic Laboratory, From 
 
 1896 to 1903. 
 George Blumer, M. D. 
 
 i-Diseases Which Can be Directly Traced to Contaminated Drinking 
 Water, Albany Medical Annals, March-April, 1897. 
 
 2— Primary Round Cell Sarcoma of the Epididymis, Ai;bany Medi- 
 cal Annals, December, 1897. 
 
 Digitized by Microsoft® 
 
3 — Adeno-myoma of the Round Ligament, American Journal of 
 
 Obstetrics, no. i, 1898. 
 4 — Tuberculosis of the Stomach, Albany Medical Annals, March, 
 
 1898. 
 S — A System of Classification of Pathological Material, Albany 
 
 Medical Annals, November, i8g8. 
 6 — A Case of Post-typhoid Bone Inflammation Due to the Colon 
 Bacillus, The Pacific Record of Medicine and Surgery, Novem- 
 ber IS, 1898. 
 7 — Tuberculosis of the Aorta, American Journal of the Medical Sci- 
 ences, January, 1899. 
 8 — A Case of Mixed Puerperal and Typhoid Infection, American 
 
 Journal of Obstetrics, no. i, 1899. 
 9 — The Present Status of the Widal Reaction, Albany Medical An- 
 nals, January, 1899. 
 10 — A Third Case of Tuberculosis of the Aorta, Albany Medical An- 
 nals, May, 1899. 
 II — The Relation of Trauma to Tuberculosis, Albany Medical An- 
 nals, November, 1899. 
 12 — The Infectious Character of Tuberculosis and the Prognosis of 
 Incipient Consumption, Albany Medical Annals, April, 1900. 
 13 — Hsmorrhagic Infection in an Infant Due to the Typhoid Bacillus, 
 Journal of the American Medical Association, December 29, 
 1900. 
 14 — Bilateral Cholesteatomatous Endotheliomata of the Choroid Plexus, 
 
 Welch's Festchrift, igoo. 
 IS — Further Notes on Abnormalities of the Urinary System, Albany 
 
 Medical Annals, January, 1901. 
 16 — The Bacteriology of Lobar Pneumonia Especially in Adults, Al- 
 bany Medical Annals, August, igoi. 
 17 — The Present Status of the Infectious Theory of Malignant Ne- 
 oplasms, New York State Journal of Medicine, November, 1901. 
 18 — Some Aspects of the Pathology of the Pancreas, Medical News, 
 
 August 9, 1902. 
 19 — A Note on the Embryonal Glandular Tumors of the Kidney, Al- 
 bany Medical Annals, August, 1903. 
 20 — The Relation of the Status Lymphaticus to Sudden Death, etc., 
 Johns Hopkins Hospital Bulletin, October, 1903. 
 With W. H. Happel, M. D. 
 
 21 — A Case of Pulmonary Tuberculosis Associated with Round Ulcer 
 of the Stomach, Albany Medical Annals, December, 1898. 
 With L. H. Neuman, M. D. 
 
 22 — Report of a Family Outbreak of Trichinosis, American Journal 
 of the Medical Sciences, January, 1900. 
 With A. J. Lartigau, M. D. 
 
 23 — A Report of Three Cases of Ascending Urinary Infection Due to 
 Bacillus Pyocyaneus and the Proteus Vulgaris, New York 
 Medical Journal, September 22, 1900. 
 
 Digitized by Microsoft® 
 
With A. T. Laird, M. D. ^ i * 
 
 24— Hsemorrhagic Infection Due to the Bacillus Mucosus Capsulatus, 
 lohns Hopkins Hospital Bulletin, February, 1901. 
 With W, Kirk, Jr., M. D. . 
 
 2S— A Case of Periostitis of the Seventh Rib Occurring During the 
 Course of Typhoid Fever, Albany Medical Annals, January, 
 1901. 
 With J. M. MosHER, M. D. 
 
 26— Hyperplasia of the Pituitary Body with Eburnation of the Skull. 
 Albany Medical Annals, March, 1901. 
 With A. MacFarlane, M. D. " 
 
 27 — Report of an Epidemic of Noma, American Journal of the Medi- 
 cal Sciences, November, 1901. 
 With H. L. K. Shaw, M. D. 
 
 28 — A Case of Appendicitis in an Infant Seven Weeks of Age, 
 Archives of Pediatrics, August, igoi. 
 With H. C. Gordinier, M. D. 
 
 29 — A Case of Chronic Lymphatic Leuksemia Without Enlargement of 
 the Lymph Nodes, Medical News, October 31, 1903. 
 August J. Lartigau, M. D. 
 
 30 — The Bacillus Pyocyaneus as a Pathogenic Factor in Human 
 Pathology, with Report of Three Cases, Philadelphia Medical 
 Journal, September 17, 1898. 
 31 — A Contribution to the Study of the Pathogenesis of B. Pyocyaneus, 
 with Especial Reference to its Relation to an Epidemic Dysen- 
 tery, Journal of Experimental Medicine, 1898, vol. iii, no. 6. 
 32 — A Contribution to the Study of the Micrococcus Tetragenus in 
 
 Acute Angina, Philadelphia Medical Journal, April 22, 1899. 
 33 — On Typhoid Septicaemia, with the Report of Two Cases, one of 
 Which was a Typhoid Infection Without Intestinal Lesions, 
 Johns Hopkins Hospital Bulletin, April, 1899. 
 34 — Typhoid Infection of the Uterus, New York Medical Journal, 
 June 16, 1900. 
 With A. MacFarlane, M. D. 
 
 35 — Acute Angina Due to Micrococcus Tetragenus, Journal of the 
 American Medical Association, June 24, 1899. 
 With H. C. Gordinier, M. D. 
 
 36— A Report of a Case of Typhoid Pleurisy, American Journal of the 
 Medical Sciences, January, 1901. 
 Arthur W. Elting, M. D. 
 
 37— The Acute Pneumonic Form of Tuberculosis, American Journal 
 
 of Medical Sciences, May, igoo. 
 38— The Bacteriology of Gonococcus Infections, Albany Medical 
 
 Annals, March, 1900. 
 39- Intermittent Gastric Hypersecretion, Boston Medical and Surgi- 
 cal Journal, May, 1900. 
 40- Primary Carcinoma of the Vermiform Appendix, with a Report 
 of Three Cases, Annals of Surgery, April, 1903. 
 
 Digitized by Microsoft® 
 
(Mt^ i/Hi^.^,,,^^ 
 
 'T 
 
 DIRECTOR OF THE BENDER HYOIENiC LABORATORY. 1S96-1B03 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
With A. Vander Veer, M. D. 
 
 41 — Resume of the Suhject of Actinomycosis with Report of a Case 
 of Abdominal Actinomycosis, Albany Medical Annals, 
 January, 1903. 
 Hakry W. Carey, M. D. 
 
 42 — Report of Two Testicular Teratomata with a Review of the Re- 
 cent Literature, Johns Hopkins Hospital Bulletin, November, 
 igo2. 
 43 — The Presence of the Shiga Variety of Dysentery Bacilli in an Ex- 
 tensive Epidemic of Dysentery, Journal of Medical Research, 
 March, 1903. 
 Arthur T. Laird, M. D. 
 
 44 — Recent Studies Regarding the Morphology of the Diphtheria 
 Bacillus, Albany Medical Annals, 1903, xxiv, 251. 
 With Charles E. Davis, M. D. 
 
 45 — A Case of Diphtheria, Albany Medical Annals, 1901, xxii, 499 
 Harris Moak, M. D. 
 
 46 — On the Occurrence of Carcinoma and Tuberculosis in the Same 
 Organ or Tissue, Journal of Medical Research, 1902, viii, 128. 
 Thomas G. Dickson, M. D. 
 
 47 — Concerning the Nature of the Small Cysts Frequently Found in 
 the Peritoneum Covering the Fallopian Tubes, American Jour- 
 nal of Obstetrics, 1903, xlviii, no. i. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
Reprintecl from Albant Mbpioal Annals, Novepiber, 1904, 
 
 Annual Report of the Director for the Year Ending 
 August 31, 1904. 
 
 To the Trustees of the Bender Hygienic Laboratory: 
 
 I have the honor to submit my report for the year ending August 
 31, 1904. 
 
 Organization — The equipment of the laboratory and the organ- 
 ization of its work by my predecessor, Dr. George Blumer, were 
 such that upon September ist, 1903, I was able to enter upon my 
 duties under the most favorable circumstances. The staff of the 
 laboratory was immediately completed by the appointment, as 
 assistants, of Edwin MacDonald Stanton (Iowa State College, 
 B. S., 1898; University of Pennsylvania, M. D., 1903) and of 
 Charles Knickerbacker Winne, Jr., (Johns Hopkins University, 
 A. B., 1897; M- D., 1901). Early in the year it was evident that 
 the clerical work had increased to such an extent that the full time 
 of a stenographer would be required ; and with your approval this 
 service has been provided. It was also found that much valuable 
 time was consumed by the members of the staff in semi-technical 
 work which could be done as well at a moderate cost by a non- 
 medical assistant. By arrangement with the faculty of the Albany 
 Medical College the additional salary was guaranteed and the 
 appointment was made. Both of these changes have aided greatly 
 in the routine work of the laboratory. 
 
 Mr. Leonard M. Wachter was reappointed for the year i904-'o5 
 as assistant chemist and bacteriologist, in charge of the Albany 
 filtration plant. 
 
 The Work of the Laboratory — This may be conveniently divided 
 into routine examinations, under-graduate teaching, instruction to 
 special students and original investigations. 
 
 The nature of the routine work of the laboratory is shown in 
 the accompanying table which gives the source of the material 
 and the number and character of the examinations. This table 
 does not include the bacteriological studies carried out in connec- 
 tion with the autopsy work, the tests made for the milk committee 
 
 Digitized by Microsoft® 
 
of the Albany County Medical Society, or the examinations of 
 "winter-killed" deer, made during the past wmter for the iNew 
 York State Forestry, Fish and Game Commission. Mr. Wachter s 
 report of his chemical and bacteriological examinations, made at 
 the filtration plant, have been included in his annual statement 
 to the Water Bureau of the City of Albany and do not appear m 
 this report. 
 
 Tabulation of the Routine Examinations made at the 
 Bender Laboratory from September i, 1903, to August 
 
 31. 1904 
 
 
 
 t 5 
 
 
 
 la 
 
 <! 
 M 
 
 Is 
 
 <u "5, 
 
 i2" 
 
 It 
 
 3 s 
 
 a 
 
 § < 
 
 & 
 
 
 
 
 a 
 1 
 
 1. ce 
 
 i 
 
 SI 
 
 
 Sputum 
 
 272 
 
 35 
 
 
 
 2 
 
 
 
 
 32 
 
 34 
 
 341 
 
 Widals 
 
 106 
 
 10 
 
 77 
 
 5 
 
 2 
 
 2 
 
 
 
 226 
 
 Throat cultures 
 
 729 
 
 713 
 
 
 
 
 83 
 
 
 
 142 
 
 1667 
 
 Water, milk 
 and ice 
 
 407 
 
 
 3 
 
 
 ] 
 
 I 
 
 
 
 46 
 
 458 
 
 Miscellaneous 
 exarainatioDS 
 
 5 
 
 
 52 
 
 2 
 
 2 
 
 
 
 
 69 
 
 130 
 
 Surgical speci- 
 mens 
 
 
 
 939 
 
 19 
 
 3 
 
 
 
 
 42 
 
 1003 
 
 Autopsies 
 
 
 
 42 
 
 I 
 
 10 
 
 6 
 
 8 
 
 31 
 
 25 
 
 123 
 
 
 1519 
 
 758 
 
 H13 
 
 27 
 
 20 
 
 92 
 
 8 
 
 31 
 
 390 
 
 3958 
 
 The examination of surgical specimens has been greatly facili- 
 tated by the general adoption of Wright's frozen section method 
 which ensures the surgeons of an early report and also lessens the 
 daily labor in the laboratory. The custom adopted by the Albany 
 Hospital, at our suggestion, of forwarding with each specimen a 
 printed form upon which is entered the important data concerning 
 the tissue to be examined allows a more intelligent examination, 
 and also enables us to keep more complete laboratory records' 
 As will be seen by the accompanying table, the examination of 
 
 Digitized by Microsoft® 
 
material from the surgical service of the Albany Hospital con- 
 stitute a large share of our routine work. For the performance 
 of this work the hospital contributes neither service nor money. 
 It has occurred to me that hospital and laboratory alike would be 
 greatly benefited by the appointment of a pathological house officer 
 whose duty it would be to assist in that part of the hospital work 
 carried on in the laboratory. In return for this service the hospital 
 should furnish him with room and board. This assistant should 
 be a graduate in medicine and preferably one with post-graduate 
 hospital training. Such an arrangement has been practiced for 
 years at the Boston City Hospital which indeed has two patholog- 
 ical internes and in that institution has proved most satisfactory. 
 I believe it might profitably be adopted by the Albany Hospital and 
 respectfully suggest it to your board. 
 
 I wish also to call your attention to the great increase in the 
 am.ount of work done for the State Department of Health. This 
 increase of work has not been accompanied by an increase of appro- 
 priation ; and the labor is out of all proportion to the compensation. 
 I think an effort should be made to secure a larger appropriation. 
 The rapid development of this work is shown in the following 
 table, which is based on examinations made from September i, to 
 August 31st. 
 
 
 Throat 
 
 
 Widal 
 
 Water, milk, 
 
 
 cultures 
 
 Sputum 
 
 reactions 
 
 ice aiiU sewage 
 
 I90I-'02 
 
 222 
 
 33 
 
 11 
 
 90 
 
 i902-'03 
 
 807 
 
 71 
 
 SO 
 
 184 
 
 1903-04 
 
 729 
 
 2S6 
 
 123 
 
 412 
 
 The laboratory courses given during the past year (term of 
 seven and a half months) to undergraduates of the Albany Medical 
 School are as follows : 
 
 Pathology and bacteriology — Drs. Pearce, Stanton and Winne, 
 eight and a half hours a week. 
 
 Normal histology — Drs. Stanton and Winne, six hours a week. 
 
 Clinical microscopy — Dr. Laird, two and a half hours a week. 
 
 Surgical pathology — Drs. Elting and Carey, two and a half 
 hours a week. 
 
 Histology and pathology, in connection with the course in 
 obstetrics — Dr. Lipes, two and a half hours a week. 
 
 Anatomy of the nervous system — Dr. Becker, one hour a week. 
 
 For the year i904-'os several changes have been made in these 
 courses. The need of more detailed instruction in the methods 
 
 Digitized by Microsoft® 
 
of clinical microscopy has led to the division of this class into 
 two sections and the appointment of another assistant, Dr. E. F. 
 Sibley. The course in the anatomy and pathology of the nervous 
 system has been placed under the direction of Dr. Stanton who 
 will be assisted in this work, as well as in the course in histology, 
 by Dr. Donald Boyd. 
 
 In connection with the course in pathology the most serious 
 need is some arrangement by which post-mortem examinations 
 may be made before the entire class. At present, this instruction 
 can be given only to small sections at irregular intervals. It would 
 appear possible, by combined action of the Albany Hospital, the 
 faculty of the Medical School and the trustees of the laboratory, 
 to arrange for proper facilities at the Albany Hospital. I respect- 
 fully suggest to the trustees the necessity of immediate action 
 upon this question. 
 
 Of those who have availed themselves of the opportunity to 
 work in the laboratory for considerable periods of time are Dr. 
 Ellice McDonald, of Philadelphia ; Mr. Leon K. Baldauf, of the 
 Johns Hopkins Medical School; Dr. J. F. Rooney, Dr, E. F. 
 Sibley and Dr. Donald Boyd, of Albany; Mr. Joseph L. Don- 
 hauser, Mr. Erastus Corning, Mr. C. W. H. Hacker and Mr. 
 T. A. Hull, of the Albany Medical School; Mr. Jerome Meyers, 
 of Dartmouth College ; and Mr. W. E. Wright, of Liberty, N. Y. 
 These gentlemen have assisted in the routine work of the labora- 
 tory, pursued special lines of study, or engaged in research work. 
 To Dr. Ellice McDonald, we are especially grateful for valuable 
 assistance during the first half of the year. 
 
 For the year i904-'os Dr. E. F. Sibley has agreed to serve as 
 voluntary assistant in all laboratory work. 
 
 Original investigations in various lines of work have been 
 carried out during the year. The expense of the experimental 
 investigations has been defrayed by a grant from the Rockefeller 
 Institute for Medical Research; a similar grant has been obtained 
 for the year i904-'o5. The following papers have been published 
 or are now in press : 
 
 '■ ^TxTns*^' ^" ^°"'"""'"^ '^^ Specificity of the Somatogenic Cyto- 
 ^^°Re"e?rchT''" ' ""''"' ^'°" *' Rockefeller Institute for Medical 
 
 Journal of Medical Research, 1904, xxi, i; also Albany Medical 
 Annals, 1904, xxv, 576. -"j^aw y iviedical 
 
 Digitized by Microsoft® 
 
2. Pearce^ R. M., The Experimental Production of Liver Necroses by 
 
 Haemagglutinins, (i» press). 
 
 (Conducted under a grant from the Rockefeller Institute for Medical 
 Research). 
 
 3. Peaece, R. M., and Winne, C. K., Jr., Concerning Haemagglutinins of 
 
 Bacterial Origin and their Relation to Hyaline Thrombi and Liver 
 Necroses, (m press). 
 
 (Conducted under a grant from the Rockefeller Institute for Medical 
 Research). 
 
 4. Pearce, R. M., The Islands of Langerhans in Congenital Syphilitic Pan- 
 
 creatitis. 
 American Medicine, 1903; vi, 1020; also, Albany Medical AnnalSj 
 1904, XXV. 88. 
 
 5. Pearce, R. M., The Pathology of Diabetes. 
 
 Albany Medical Annals, 1904, xxv, 329. 
 
 6. Pearce, R. M., Experimental Pancreatitis Considered in its Relation 
 
 to Acute Lesions of the Pancreas in Man. 
 American Medicine, 1904, vii, 853; also Albany Medical Annals, 
 1904, xxv, 389. 
 
 7. Pearce, R. M., Cancer of the Pancreas and Glycosuria; A Histological 
 
 Study of the Islands of Langerhans, (in press). 
 
 8. McDonald, E., Complications and Degenerations of Uterine Fibromyo- 
 
 mata. 
 Journal of the American Medical Association, 1904, xlii, 13^14 ; also, 
 Albany Medical Annals, 1904, xxv, 515. 
 
 9. Carey, H. W., and Laird, A. T., A Peculiar Hypertrophy of the Pros- 
 
 tate Accompanied by an Ascending Infection and Cysts in the Ureters, 
 with a General Discussion of Ureteritis Cystica. 
 
 Albany Medical Annals, 1904, xxv, 532. 
 
 10. Stanton, E. MacD., and Shaw, H. L. K., Infantile Inguinal Hernia 
 
 of the Vermiform Appendix with Adhesion to the Epididymis. 
 
 Albany Medical Annals, 1904, xxv, 589. 
 
 11. LiPES, H. J., A Study of the Changes Occurring in the Endometrium 
 
 during the Menstrual Cycle. 
 
 Albany Medical Annals, 1904, xxv, 94. 
 
 Other investigations not yet completed are : An Experimental 
 Study of Hydrophobia, by Dr. E. MacD. Stanton; A Study of 
 the Uric Acid Elimination in (jout, by Dr. A. T. Laird; Cancer 
 of the Appendix, by Mr. Leon K. Baldauf ; and An Investigation 
 of the Giant Cells occurring in the Musculature of the Pregnant 
 Uterus, by Mr. Erastus Corning. 
 
 Digitized by Microsoft® 
 
In pursuance of a plan already submitted to you it is my J^ten- 
 tion to prepare for general distribution a volume which will 
 contain the above papers. This volume will be known as Studies 
 from the Bender Laboratory," and with your financial support, I 
 hope to make it an annual publication. 
 
 Improvements and Repairs— These have not been extensive. The 
 need, however, of more space for the general work of the labora- 
 tory led me to convert the large room on the first floor, formerly 
 occupied by the janitor, into an office for the stenographer and 
 the proper storing of the laboratory records. The janitor has been 
 given the room on the third floor, formerly a cloak room for the 
 use of students. The latter have been provided for by placing 
 hooks in the large hall off the class room of this floor. Three 
 small private rooms intended for the use of persons conducting 
 special researches have been procured by partitions isolating the 
 unused portion of the large class room on the third floor. These 
 assure more or less privacy and add to the desk space available 
 for special work. Other improvements include an additional work 
 table in the director's room, a work bench in the basement with 
 vise and tools for the janitor's department, important changes in 
 the arrangement of the animal room and additional shelving in 
 various parts of the building. During the past summer all the 
 floors have been filled and oiled, the wood work of halls and 
 lecture room varnished, the walls of halls and office painted, and 
 various repairs to the windows made. To prevent a repetition of 
 the disastrous flooding of the basement which occurred during 
 the freshets of last spring, the ground at the rear of the building 
 has been raised and sodded and the walls and cover of the cellar- 
 way repaired. The appearance of the front of the building has 
 been improved by cleaning the brick work and repointing the stone 
 work and steps. 
 
 Gifts — A portrait of Mr. Matthew W. Bender, presented by Dr. 
 George E. Gorham, has been placed in the lower hall ; a series of 
 fifty-one portraits, the gift of Dr. George Blumer, representing 
 prominent workers in scientific medicine, in the large class room 
 of the third floor; and a portrait of Dr. George Blumer, in the 
 Director's room. 
 
 Richard M. Pearce. 
 
 Digitized by Microsoft® 
 
Reprinted from Albany Medical Annals, August, 19M. 
 
 CONCERNING THE SPECIFICITY OF THE SOMATO- 
 GENIC CYTOTOXINS.* 
 
 Presented at the Fourth Annual Meeting of the American Association of 
 
 Pathologists and Bacteriologists, held at New York, April 
 
 1st and 2nd, 1904. 
 
 By RICHARD M. PEARCE, M. D., 
 Director of tlie Bender Laboratory, Albany, N. Y. 
 
 CONTENTS 
 
 Introduction — ^Methods — Observation of immunized animals — On the 
 action of anti-sera obtained by the injection of (o) red blood cells, (6) 
 liver, pancreas, adrenal and kidney containing blood, (c) the same organs 
 free of blood, (rf) blood serum, (e) urine, and {f) bile — Controls with 
 normal serum — Relative toxicity of the various sera — Summary and Con- 
 clusions — References. 
 
 INTRODUCTION 
 
 This study was undertaken with the object of demonstrating, 
 by a comparative study of the somatogenic cytotoxins, that the 
 action of any given cytotoxin depends on the chemical (receptor) 
 organization of the cell and not upon morphological characteristics ; 
 that an artificial anti-serum may have a definite toxic action on 
 several types of cells of widely differing morphology but pre- 
 sumably having in part at least common receptors ; that specificity 
 in the sense of morphological affinity is of secondary importance 
 and in the case of some sera may not exist; that many of the 
 results supposed to be due to the specific action of one cytotoxin 
 
 * TbiB atndy bae been conducted nnder a grant from the Rockefeller Institute for Medical 
 Research. 
 
 Digitized by Microsoft® 
 
are due rather to other cytotoxins adventitiously elaborated as the 
 result of improper methods of immunization. 
 
 My attention was first attracted to these various phases of cyto- 
 toxic activity by some observations made in the course of an 
 experimental investigation of nephrotoxins.* I found that by the 
 usual method of immunization, that is, the injection of the kidney 
 with its contained blood, a serum strongly haemagglutinafive and 
 haemolytic was produced and that the haemoglobinuria resulting 
 from the injection of this serum entirely obscured any nephrolytic 
 action it might possess. The haemagglutinative element of the 
 serum also caused the agglutination of the red blood corpuscles; 
 the thrombi thus formed occluding the vessels of the liver with 
 resulting focal necroses of that organ. It was therefore evident 
 that many of the results previously reported as due to a specific 
 nephrotoxin or hepatotoxin were duetohaemagglutininsorhaemo- 
 lysins forming in animals improperly immunized. To obviate this 
 difficulty blood-free kidney was used in the immunization with 
 the surprising result that although the serum thus obtained was 
 powerfully nephrotoxic, it was also, in a moderate degree, haemag- 
 glutinative and haemolytic. It did not however cause haemo- 
 globinuria. Moreover hepatotoxin, so-called, prepared in like 
 manner possessed similar characteristics but exerted no special 
 action upon the liver. 
 
 At this time an extended investigation of the literature indicated 
 that immunization with "washed" or blood free organs had in no 
 instance been practiced. It was evident therefore that a study of 
 cytotoxins by this method of immunization might yield results 
 differing widely from those of previous investigations and possibly 
 of fundamental importance in connection with the question of the 
 specificity of these substances. Before taking up the detailed dis- 
 cussion of this work, it may be well to summarize the more impor- 
 tant communications upon the question of specificity. 
 
 In the vast amount of literature which has accumulated the 
 
 ♦Pbarob. Ad Experimental Stady of Nephrotorins, ITniv. (jT Penna. Med. BuUetin 1903 
 
 Digitized by Microsoft® 
 
predolminating feature is the effort to demonstrate a morphological 
 afBnity. This is especially evident in the early work which neglects 
 to a great extent the question of chemical (or receptorial) struc- 
 ture. Due consideration has not been given to the very clear 
 and definite statements of Ehrlich and Morgenroth concerning 
 specificity, v. Dungern and Mexter produced anti-sera for cili- 
 ated epithelium and testicular cells and found them to be also 
 haemolytic. On account of these results they denied a strict 
 specificity to the cytotoxins. Metchnikoff and Metalnikoff, on the 
 other hand, believing in this specificity, endeavored to explain 
 the double action on the supposition that sufficient red blood cor- 
 puscles were injected with the epithelial cells to produce an haemo- 
 lysin. 
 
 Of wider interest is the production of haemolysins by injecting 
 body fluids entirely free of red corpuscles, v. Dungern, Besredka, 
 Tschistovitsch, Miiller and Morgenroth by injecting blood serum, 
 and Shattenfroh, Buffer and Crendiropoulo by injecting urine have 
 obtained such. v. Dungern produced an haemolysin also by inject- 
 ing cow's milk, and recently Portis has obtained a like result by 
 injecting the "colloid matter" of the thyroid. These observations, 
 at first glance opposed to the theory of specificity if considered 
 from the point of view of morphology, are explained by Morgen- 
 roth by the presence of free receptors in the blood and urine. 
 These free receptors, as first announced by Ehrlich, appear in 
 these fluids, especially the blood, as the result of the cell changes 
 accompanying normal metabolism. Thus it is possible for recep- 
 tors of all cells of the body to be present in the blood serum. 
 Naturally, immunization with fluids containing these receptors 
 would result in the formation of specific anti-bodies. Morgenroth 
 therefore draws the conclusion that specificity is a matter not of 
 cells but of receptors, whence a given cell may be in chemical 
 conformity with cells of a different organ or perhaps animal. Recent 
 investigations of anti-thyroid serum (Yates, MacCallum, and Por- 
 tis) and of anti-adrenal serum (Abbott, Yates) indicate that these 
 
 Digitized by Microsoft® 
 
have little or no specific action. Unwashed adrenal serum, how- 
 ever, Abbott found to be possessed of a definite haemolytic power; 
 Portis found the same to be true of washed thyroid serum. 
 
 In order to study this much debated question of specificity in a 
 somewhat comprehensive manner a comparative investigation of 
 several cytotoxins, as outlined in the introductory summary, was 
 undertaken. 
 
 METHODS 
 
 The animals selected for this study were the dog and the 
 rabbit ; the latter being immunized with the cells and fluids of the 
 former. The method of immunization adopted was that followed 
 in the work on nephratoxins. All injections were made into the 
 peritoneal cavity at intervals of a week or ten days, the animal 
 being bled a like period after the third, fourth or fifth injection. 
 Preliminary bleedings were made from the jugular veins ; the final 
 bleeding from the carotid arteries. The serum thus obtained was 
 injected into the saphenous, occasionally the femoral vein of the 
 dog. In determining the specific effect of these various sera, syste- 
 matic examinations of the urine and detailed gross and histological 
 studies of the several organs have been made; their action upon 
 red blood corpuscles has been determined in vitro. Attempts were 
 also made to determine by the test-tube method the action of the 
 various sera upon emulsions of the several types of parenchyma- 
 tous cells ; but the results were so confusing and contradictory that 
 this method was early discarded in favor of intravenous injection. 
 This method, analogous to pharmacological methods, of testing the 
 action of a serum by its effect on the living cells is, I believe, the 
 only method free of error. Moreover, it is by this method alone 
 that the cytotoxins may be utilized to explain various pathological 
 processes, for while tests in vitro may establish the theoretical 
 principles of this method of investigation, substantial progress 
 depends on the experimental application to the living cell in its 
 normal environment. 
 
 Digitized by Microsoft® 
 
Some information concerning what may be called the more 
 general action of a serum has been determined by observing 
 fluctuations in body weight; and of the relative lethal action, by 
 a comparative study of dosage. 
 
 The histological methods include hardening in Zenkers, Miiller's 
 and Fleming's solutions ; imbedding in celloidin and paraffin ; and 
 staining by haemotaxylin and eosin, methylen blue and eosin, Mal- 
 lory's connective tissue stain and safifranin. The various fresh 
 methods have been utilized in the study of fatty and pigmentary 
 changes. 
 
 OBSERVATIONS OF IMMUNIZED ANIMALS 
 
 Contrary to general opinion the rabbit may readily be treated 
 with various animal cells and fluids if due attention is paid to 
 cleanliness and to the condition of the aninml as shown by fluctua- 
 tions in body weight. In this series the only deaths occurred 
 among rabbits receiving bile and unwashed liver. Fifty per cent, 
 of the animals treated with these substances succumbed before 
 the third injection. This result is apparently due to the toxicity 
 of bile, for no difficulty was encountered in animals receiving 
 washed liver, blood corpuscles or blood serum. The kidneys of 
 animals dying within a few days after the injection of bile exhibit 
 a well marked degeneration of the convoluted tubules with deposi- 
 tion of a yellowish pigment. 
 
 Necrosis and suppuration of the subcutaneous tissues may occa- 
 sionally occur after the injection of organ emulsions ; while encap- 
 sulated masses of the injected material in the peritoneal cavity are 
 not infrequent. In one instance after injection of pancreas 
 emulsions fat necroses of the mesenteric, perirenal and pelvic fat 
 were observed. 
 
 The changes in the urine are but slight ; a heightening of color 
 may occur not uncommonly after injections of blood and adrenal ; 
 slight traces of albumen are infrequently found ; glycosuria, tran- 
 sitory and of slight degree usually follows the injection of adrenal. 
 
 Digitized by Microsoft® 
 
The several organs of these various animals with the exception 
 of those receiving bile, exhibited no histological changes of any 
 moment. 
 
 ACTION OF SERUM RESULTING FROM THE INJECTION OF 
 ERYTHROCYTES 
 
 While the action of haemolytic sera in vitro is well known, little 
 attention has been paid to the changes resulting from its injection 
 into the living animal. Kraus and Fukuhara only, as far as I am 
 aware, have described the histological changes. As such observa- 
 tions are essential, as controls to the experiments to follow, I have 
 made a careful study of this phase of the subject. Rabbits were 
 immunized with defibrinated and washed blood respectively, receiv- 
 ing three to five doses of five centimetres each. Whether defibrin- 
 ated blood or washed corpuscles are used appears to make no dif- 
 ference in the potency of the serum. Seventeen dogs have received 
 this serum, intravenously, in doses varying from 1 13000 to 1 :450.* 
 The effect varied with the dosage. In large amounts ( i :6oo to 
 I 400) death usually occurs in from three to fifteen hours. Post- 
 mortem examination of dogs poisoned in this way show the blood 
 to be laked and soft gelatinous clots to be present in the heart and 
 large vessels. The internal organs especially the liver, are intensely 
 congested; the serous cavities may contain a small amount of 
 blood tinged fluid. If the bladder contains urine it is dark claret 
 colored, rich in albumen with a sediment composed of casts, coal- 
 esced masses of altered haemoglobin and granular and fatty renal 
 cells. Histological examination reveals agglutinated red Mood cor- 
 puscles in the various organs ; the vessels of the spleen and liver 
 are intensely congested, the cells of the latter often being com- 
 pressed and disorganized ; beginnmg necrosis of liver cells is not 
 infrequent. The kidney tubules present a wide-spread fatty meta- 
 
 • Tliis method of Indicating the dose refers to the relation between the amonnt of semm 
 Injected and the weight of the animal. Thns 1 :52B indicates that a dog weighing 6300 gramme* 
 received 12c. c. of eemm or I c. c. to 685 grammes of hody weight. 
 
 Digitized by Microsoft® 
 
morphosis; their lumina contain altered haemoglobin in the form 
 of casts or irregular granular material; the tissue spaces are 
 oedematous and an accumulation of coagulated serum may be seen 
 in the glomerular spaces. 
 
 The smaller doses of i :iooo to i 700 are usually not fatal; but 
 the dogs develop within eighteen to twenty-four hours a haemo- 
 globinuria which persists for three to five days. After the dis- 
 appearance of the haemoglobin the urine is still high colored and 
 deeply stained with bile. A yellow jaundiced condition of the con- 
 junctivae is a constant syinptom. The albuminuria which accom- 
 panies the acute condition rapidly diminishes as the urine becomes 
 clear of blood pigment although it may be detected by the heat 
 test for many days ; in one animal persisting for three weeks. The 
 urinary sediment contains the products of the destruction of red 
 blood corpuscles and usually many renal cells. Beginning with the 
 second or third day a few -casts of the finely granular type are 
 found, but not invariably. Hyaline liver necroses* due to agglu- 
 tinated red blood corpuscle thrombi are constantly present in these 
 animals. Equally constant is a fatty degeneration of the kidney, 
 which is most marked in the tubules of the medullary rays. 
 Hemorrhages into the adrenal occurred in two of the seventeen 
 animals. They were also found occasionally in the lung. 
 
 With doses as low as i :i200 to i :3000 haemoglobinuria does not 
 always result, but the urine is bile stained and contains albumen. 
 Changes similar to those described above but less marked may 
 occur in the liver and kidney. 
 
 The attempts to produce haemolytic sera with small amounts of 
 blood have been of great interest. These were undertaken for the 
 purpose of controlling sera prepared by injecting organs (adrenal) 
 which contain an amount of blood so small that it has heretofore 
 been considered negligible. A series of rabbits receiving o.i cubic 
 centimetre of defibrinated blood yielded, after the second injection, 
 a serum which was moderately agglutinative and haemolytic in 
 
 * This condition baa been the subject of especial investigation and will be considered in 
 detail in another communication. 
 
 Digitized by Microsoft® 
 
vitro; after j&ve injections the serum was immediately and power- 
 fully agglutinative and haemolytic. This serum injected into 
 dogs in doses of 1 1550 to 1 1345 causes haemoglobinuria, multiple 
 liver necroses, fatty degeneration of the kidney and other changes, 
 including adrenal haemorrhages, similar to those produced by sera 
 prepared from larger amounts of blood. One dog, receiving a 
 dose of 1 :350, died in four hours with evidence of extensive 
 thrombosis due to agglutinated red blood corpuscles. 
 
 serum from unwashed organs ( kidney, liver, pancreas and 
 
 adrenal) 
 
 In this series the serum was prepared in the usual manner by 
 injecting emulsions of the unwashed organ, that is, with the normal 
 post-mortem blood content. Four to twelve grammes were used 
 at each injection of kidney, liver and pancreas; while animals 
 immunized against adrenal cells received a pair of the organs at 
 each injection. The experience of Yates in the production of 
 adrenal serum indicates that a larger amount of adrenal tissue has 
 no appreciable effect on the character of the serum. 
 
 All these sera, naturally, contain haemagglutinin and haemoly- 
 sin. The liver, kidney and adrenal sera injected into the dog in doses 
 of 1 :900 to 1 :300 produce similar changes which, moreover, can- 
 not be distinguished from those due to a pure haemolytic serum. 
 Haemoglobinuria, liver necroses, fatty degeneration of the kidney 
 and deposition of blood pigment in spleen and lymph nodes are 
 constant. Only in the case of the kidney serum was a dual action 
 evident. Here, the persistence of albumen and a larger number of 
 casts (granular, epithelial and fatty) after the disappearance of 
 the haemoglobinuria indicated the action of the specific nephrotoxic 
 constituent. This serum saturated with washed red blood cor- 
 puscles at zero temperature lost all but a trace of its haemolytic 
 power but was still powerfully nephrotoxic. 
 
 No changes occurred in the liver or adrenal which could be 
 ascribed to their supposedly specific sera. The kidney serum in one 
 
 8 
 
 Digitized by Microsoft® 
 
instance produced focal hemorrhages in the adrenal; the adrenal 
 serum similar lesions in the lung. 
 
 The pancreas serum in doses of i :8oo to i :390 did not produce 
 haemoglobinuria but a persistent albuminuria with fatty degenera- 
 tion of the kidney. The strongest dose given ( i :39o) caused also 
 well marked focal necroses of the liver with evidence of red blood 
 cell thrombosis. No change in the pancreas cells could be demon- 
 strated. 
 
 Of these four sera that prepared from adrenal appears to be the 
 most toxic, causing death in less than twenty-four hours in dose of 
 I 400. 
 
 It will be seen from this summary that these sera, with the 
 possible exception of nephrotoxin, produce lesions which differ in 
 no way from those due to a serum prepared from erythrocytes 
 and that they may all readily be explained by the action of haemag- 
 glutinin or haemolysin. 
 
 SERA PREPARED FROM WASHED ORGANS (kIDNEY^ LIVER, PANCREAS 
 AND adrenal) 
 
 In this series the blood, lymph and the various secretions with 
 their free receptors were removed from the organs by prolonged 
 washing with warm, sterile salt solution. After this procedure 
 it was assumed that only the cells with their fixed receptors re- 
 mained. The kidney was readily washed through its arteries; the 
 adrenal, pancreas and liver at one time through the abdominal 
 aorta ; the clearing of the liver vessels was augumented by washing 
 also through the portal vein. The success of this procedure was 
 tested by immediate examination of the respective organ emulsions 
 and later by examination of hardened sections of the organs. Kid- 
 ney, adrenal and pancreas, absolutely blood free, were readily 
 obtained ; a very few red cells (one or two in three or four fields) 
 were always found in the liver emulsions, despite the most thor- 
 ough washings with fifteen to twenty litres of salt solution. These 
 cells, however, appear to have had little influence upon the char- 
 acter of the serum produced. 
 
 Digitized by Microsoft® 
 
AH sera resulting from the injection of these blood free emul- 
 sions have the power of agglutinating and haemolysing red Wood 
 corpuscles in vitro. The agglutination in all instances is imme- 
 diate and powerful. The haemolytic activity of the adrenal serum 
 is almost equal to that of serum prepared from red blood cor- 
 puscles; liver, pancreas and kidney sera are less powerful, their 
 relative strength being in the order named. None of these sera 
 on the other hand are sufficiently powerful to produce haemoglo- 
 binuria when injected into the circulation of the dog. My experi- 
 ence is different from that of Portis, who has recently described 
 a serum prepared from washed thyroid, which is haemolytic in 
 vitro and in animals produces haemoglobinuria. 
 
 The only one of these sera which has a definite action in the 
 sense of morphological specificity is nephrotoxin. The action 
 of this cytotoxin I Tiave recently described in detail, and only ^ 
 general summary is now necessary. Doses of i :i6oo to 1 1500 
 produce severe albuminuria and cast excretion. The casts are 
 finely granular, and less frequently fatty, epithelial and hyaline; 
 granular and fatty renal cells are usually abundant. The albumen 
 gradually diminishes in amount, but may persist in faint traces 
 for two weeks or more ; the casts usually disappear within twelve to 
 fifteen days. The kidney is swollen and oedematous with slightly 
 adherent capsule. On section the cortex is oedematous and pale 
 yellow in color with occasionally a few punctate haemorrhages ; the 
 glomeruli are distinct but pale. Histologically the cells of the 
 convoluted tubules are found to be swollen and granular; the 
 cells of the tubules in the medullary rays are vacuolated and osmic 
 acid preparations reveal extensive fatty metamorphosis; the 
 tubules especially those of the pyramids contain much granular 
 material and many casts. Exudation into the glomerular space 
 and hyaline thrombosis with degeneration of the glomerular tuft 
 may occur. With the exception of the liver, which in some 
 animals exhibited extensive granular degeneration no other organs 
 appear to be affected by this serum. 
 
 The action of the so-called hepatotoxin is variable. In doses of 
 
 10 
 
 Digitized by Microsoft® 
 
I :iooo to 1 :525 its chief action is to cause a more or less extensive 
 fatty metamorphosis of the kidney. This may or may not be 
 accompanied by albuminuria ; in two experiments the albuminuria 
 was equal to that resulting from the administration of nephrotoxin 
 and abundant excretion of casts was present. That definite 
 changes are produced in the liver by this serum is not easily 
 determined. The only change observed was an increased granu- 
 lation of the liver cells with occasionally slight fatty transforma- 
 tion. This appearance however does not differ from that fre- 
 quently seen in the liver of apparently normal dogs and is quite 
 constant after the administration of other anti-sera, especially 
 adrenal serum. Focal necroses of the liver due to this serum I 
 have seen in but one instance. These were few in number and 
 followed the administration of a powerful dose ( i :266) of a 
 strongly haemagglutinative serum. As agglutinated masses of red 
 blood corpuscles could be seen in the vessels of the necrotic areas, 
 I consider this action to be due to the agglutinin present and not 
 to a hepatotoxic constitutent of the serum. The necroses described 
 by Delezenne and others working with serum prepared from the 
 unwashed organ may, I believe, be explained in like manner. The 
 changes in the urine described by Delezenne I have not seen. 
 
 Adrenal serum in doses of 1 1700 to i :5SO produces no lesion of 
 the adrenal gland; but well marked granular and fatty trans- 
 formation of liver and kidney. One animal, killed forty-eight 
 hours after a dose of i :5S6, presented multiple minute haemor- 
 rhages of the lungs and liver with free haemorrhages into 
 the glomerular spaces of the kidney. A peculiar phenomenon 
 observed in all animals receiving adrenal serum was an 
 immediate blanching of the mucous membranes with cardiac and 
 respiratory disturbance. This condition, which lasted fifteen to 
 twenty minutes and necessitated the employment of artificial 
 respiration, was not observed in animals receiving other cytotoxic 
 sera. Unfortunately a kymographic study of this condition was 
 not made. 
 
 II 
 
 Digitized by Microsoft® 
 
Washed pancreas serum produces in doses as high as i :620 
 no recognizable changes in the pancreas and but moderate granular 
 and fatty metamorphosis of the liver and kidney. The transitory 
 glycosuria described by Surmont and by Camot and Gamier I 
 have not seen. Their injeotions were made, however, directly 
 into the pancreas; the glycosuria therefore may have been the 
 result of trauma and analogous to that observed by Flexner and 
 Pearce to follow various injuries to the pancreas. 
 
 SERA PREPARED BY INJECTING BLOOD SERUM, BILE AND URINE 
 
 As has been stated above, several investigaitors have produced 
 haemolytic sera by immunizing animals with blood serum, urine 
 and other fluids free of red blood corpuscles. This result Mor- 
 genroth explains is due to the presence of free receptors normally 
 constitutents of the body fluids. In order to determine if sera 
 thus prepared contain cytotoxins for cells other than red blood 
 corpuscles ; also to control other experiments of this series, I have 
 injected rabbits three to five times with the blood serum, bile and 
 urine of the dog in doses of five to ten cubic centimetres. 
 
 All three sera thus prepared are strongly haemagglutinative. 
 The serum produced by injecting urine is powerfully haemolytic, 
 that by injecting serum, moderately haemolytic, while the "anti- 
 bile" serum has no haemolytic power whatever. This peculiarity 
 of "anti-bile" serum has been observed recently by RuflEer and 
 Crendiropoulo, and is apparently to be explained by the produc- 
 tion of anti-haemolysin, the bile being naturally haemolytic. 
 
 "Anti-bile" serum in doses of i :62o and i 700 causes fatty 
 degeneration of the kidney and focal necroses of the liver. These 
 latter are accompanied by agglutinative red-blood corpuscle throm- 
 bosis and apparently due to a mechanical disturbance of the liver 
 circulation. No changes are found in other organs. 
 
 The serum of rabbits receiving dog's serum injected in doses 
 of I :iooo to 1 :39o causes extensive fatty degeneration of the 
 kidney with sometimes slight traces of albumen in the urine. 
 
 The serum of rabbits receiving dog's urine produces in the dog 
 
 12 
 
 Digitized by Microsoft® 
 
a distinct congestion of the liver; otherwise no appreciable 
 changes of either tissues or excretions follow doses as high as 
 1 :674. 
 
 CONTROLS WITH NORMAL RABBIx's SERUM 
 
 Normal rabbit's serum injected in doses of i •.1320, i rgoo, i :6oo, 
 I :Soo and i :300 was without influence on the urine or histological 
 structure of the various organs examined (kidney, liver, pancreas 
 and adrenal). 
 
 RELATIVE TOXICITY OF THE VARIOUS SERA 
 
 An attempt has been made in addition to the above described 
 methods of study to gain some idea of the relative toxicity of 
 the various sera by a comparison of the changes in body weight 
 and of the varying periods of time elapsing before fatal termina- 
 tion. Little information has been obtained by the study of 
 changes in weight, for while a loss of weight is more or less 
 constant, there is little variation according to the kind of serum 
 injected, except that an appreciably greater loss of weight occurs 
 in animals receiving sera which produce albuminuria (nephro- 
 toxin) and haemoglobinuria (haemolysins produced by defibrin- 
 ated blood and unwashed organs). 
 
 Death followed only after the use of sera prepared from red 
 blood corpuscles and from blood containing organs. With the 
 exception of the powerful nephrotoxin, prepared from washed 
 kidney cortex, described in a previous communication,* death has 
 never resulted from the injection of sera prepared from washed 
 parenchymatous cells or from blood-free body fluids. There is 
 apparently considerable difference in the toxicity of the haemolysin 
 produced by the injection of red corpuscles and those produced 
 by washed parenchymatous cells; washed adrenal serum, for 
 example, the haemolytic power of which in vitro is almost as 
 great as that of true haemolysin has never caused death. On the 
 other hand, it is of interest that the lethal power of a serum pro- 
 
 * Loc cU., page 2S8. 
 
 »3 
 
 Digitized by Microsoft® 
 
duced by blood injected in o.i cubic centimetre doses is almost as 
 great as that following the injection of five cubic centimetre doses 
 The former has caused death in four hours, in dose of 1 1350 ; the 
 latter in three hours, in dose of i :45o. With these sera may be 
 compared the unwashed adrenal serum which caused death in 
 less than twenty-four hours in dose of i :400. These results indi- 
 cate that as a general rule the power of a serum is not increased 
 by large immunizing doses. 
 
 SUMMARY AND CONCLUSIONS 
 
 From these experiments it is evident that the cells of the various 
 organs of the body while differing in morphology and function 
 have certain receptor characteristics in common and that one 
 type of cell may therefore produce anti-bodies affecting several 
 cells of differing morphology but with like receptor groups. This 
 is shown by the sera prepared from washed liver, kidney, pancreas 
 and adrenal, all of which agglutinate and haemolyse red blood 
 corpuscles, and may cause degenerative changes also in the liver 
 and kidney. Some of these cytotoxic sera have no effect upon 
 organs for which they are supposed to have a morphological affin- 
 ity, but exert a powerful lytic influence upon other cells. Aside 
 from nephrotoxin, which has a distinct inj urious action upon renal 
 epithelium the various cytotoxins studied have no specific action 
 in the morphological sense. The action of the kidney serum itself 
 cannot be said to be specific, for in addition to the nephrotoxic 
 element another constitutent (haemolysin) is present. The action 
 of this serum might perhaps be described as "special." Pancreas 
 and adrenal serum on the other hand are devoid even of special 
 action, while the lesions of so-called hepatotoxin are doubtful and 
 may be produced by other sera. The striking characteristic of 
 adrenal serum is its powerful haemolysin. 
 
 The presence of free receptors in the various body fluids is 
 shown by the appearance of anti-bodies in the serum of animals 
 
 14 
 
 Digitized by Microsoft® 
 
receiving serum, bile and urine; the diverse character of these 
 .receptors by the variety of cytotoxins thus formed. These obser- 
 vations offer strong support of the view that specificity is a func- 
 tion of receptors and not of cells. 
 
 A comparison of the results obtained by sera prepared from 
 washed organs with those of sera prepared from blood and blood 
 containing organs indicates that many of the positive results de- 
 scribed as specific and characteristic of the action of certain cyto- 
 toxins are due rather to haemagglutinin and haemdysin adventi- 
 tiously formed as the result of impure methods of immunization. 
 
 In connection with the activity of haemolysin it is also necessary 
 to differentiate sharply between the primary specific action upon 
 the red blood cells and the secondary action upon other cells, as 
 those of the kidney, due to the toxic effect of haemoglobin deriv- 
 atives and possibly other substances set free by the destruction 
 of the red blood cells. 
 
 Another phase of the question, heretofore entirely disregarded, 
 is the part played by haemagglutinin. This substance was present 
 in all the sera studied, and in some, when sufficiently powerful, 
 caused the formation of red blood corpuscle thrombi and a result- 
 ing mechanical disturbance of the circulation leading to degenera- 
 tion, necrosis and haemorrhage. This action is best illustrated by 
 "anti-bile" serum which is without haemolytic power, but through 
 its powerful agglutinin causes the formation of red blood corpuscle 
 thrombi and thus the production of widespread liver necrosis. 
 Such lesions have heretofore been confounded with the action of 
 the cytolytic constituent of a serum, and have been considered 
 specific while they are in truth due to the much simpler agglutin- 
 ating body. 
 
 REFERENCES 
 A general enmmary of the literature of the cytotoxins with brief abstracts of the more 
 Important contribntions by German and French writers may be found in the recent commnni- 
 catlon by Saohs, Biochem. CmtrdO), 1903, 1, 673, 613, 653, 693. 
 Abbott. Jowmalof Ueilcal Sesearch, 1903, ix, 329. 
 
 Bbsbsdka. Cited by Asohoiv, " Ghrlich's Seitenkettentheorie," Jena, 1902, 
 
 p76. 
 Cashot AMD Gaxnub. C. B. de la Soc. de Biol., 1901, liil, 446. 
 
 Digitized by Microsoft® 
 
DltUEZXHNK. 
 T. DlraSXHH. 
 
 Ehbuoh. 
 
 Flbxnbb and Fiabcs. 
 
 FuExraiiu. 
 
 Ebaus. 
 
 MacCaixux. 
 
 Metalnikoff 
 
 MSTOHtnEOFF. 
 MOBSSHBOTH. 
 MOXTIB. 
 MSI.I.BB. 
 
 Peabcx. 
 
 P0BTI3. 
 
 RirFFlIB AND Cbbndibofoui.0, 
 
 Shattbhfbob 
 
 SflBUONT. 
 
 tsohistovitboh. 
 
 Tatbb. 
 
 C. B. at r Acad, aes Science!, HOO, cxxsl, «7; cited by Saohb 
 
 BiocTiem. Centralb., 1903, 1, 616. 
 Munch, med. Wochemchr., 1899, jclvi, 40B, «9, 1228; 1900 xlvU, 
 
 968; Die Antikarper, Jena, 1903. 
 SchJussbetraclitnngen, Sbthnai/el^s tpecUBt Path. u. Therapie, 
 
 1901, vlil 166. 
 
 miv. of Penna. Med. SuU., 1901, xiv, 193. 
 
 Zitgier's Beitrdge, 1901, xxxv, 434. 
 
 Verhandl. d. Deutech. path. GeaeUsch., 1902, v, 58. 
 
 Median Newt, 1908, Ixxxiii, 820. 
 
 Ann. de V Inetitut Pasteur, 1900, xiv, 67T. 
 
 Ann. de I' Inetitut Pasteur, 1900,3dv, 1, 869. 
 
 Munch, med. Wochenschr., 1902, xlix, 1033. 
 
 Deutsch. med. Wochenschr., 1900, xxvi, 61. 
 
 MUnch. med. Wochenschr., 1903, xlix, 1830. 
 
 Univ. of Penna. Med. Bull., 1903, xvl, 217, Trans. Assoc. Amer. 
 
 Physicians, 1903. 
 Jour, of Infectious Diseases, 1904, 1, 1S7. 
 British Med. Jour., 1903, ex, 130; C. B. de la Soc. de Biol., 1908, 
 
 Iv, 954; Jour, of Path, and Baet., 1904, Ix, 278. 
 Miinch. med. Wochenschr., 1901, xlviii, 1239; Arch. f. Hygiene, 
 
 1902, xliv, 339. 
 
 C. B. de la Soc. de Biol., 1901, liii, 445. 
 
 Arch. Buss, de Path., 1899, viii, 68; cited by Mokqbnboth 
 
 AsoHOFF and v. DuHesBN. 
 Univ. of Penna. Med. Bull., 1903, xri, 195. 
 
 16 
 
 Digitized by Microsoft® 
 
Reprinted from Albany Medical Annals, October, 1904. 
 
 THE EXPERIMENTAL PRODUCTION OF LIVER 
 NECROSES BY THE INTRAVENOUS INJECTION 
 OF HAEMAGGLUTININS.* 
 
 Presented before the American Association of Pathologists and Bacteri- 
 ologists at the fourth annual meeting held in New York, April I and 2, 1904. 
 
 By RICHARD M. PEARCE, M. D., 
 
 Director of the Bender Laboratory, Albany, N. Y. 
 
 Introduction — This investigation was undertaken with the 
 object of demonstrating experimentally that red blood cor- 
 puscles through the agency of agglutinins become fused to- 
 gether and form thrombi which occluding the vessels of the 
 liver may lead to hyaline necrosis of liver cells. That these 
 lesions might thus be produced, was indicated by observations 
 made during a study of various cytolytic sera. In the course 
 of an experimental study of nephrotoxins/ it was noted that 
 focal necrotic lesions of the liver with associated hyaline 
 thrombi frequently followed the intravenous injection of sera 
 possessing considerable hsemagglutinative power, but did not 
 follow the injection of sera devoid of this characteristic. Flex- 
 ner's^ communication concerning agglutinated red blood cor- 
 puscle thrombi, which had appeared but a short time before, 
 suggested the possibility of these hyaline thrombi being com- 
 posed of red blood corpuscles. A careful study of the lesions 
 proved this to be the case, and a series of experiments was 
 therefore at once begun for the purpose of establishing defi- 
 nitely the power of agglutinin to produce red cell thrombi and 
 also to establish a relation between the thrombi and the 
 necrotic lesions. Later, my interest in the subject was stimu- 
 lated by the appearance of Boxmeyer's' work on the liver 
 necrosis of hog cholera. Boxmeyer found that while one type 
 
 * This Investigation was condncted under a grant from the Bocliefeller Institute for 
 Medical Besearch. 
 1 Peabois. An Experimental Study of the Nephrotoxins, Universit)/ of Perm. Med, Bull. 
 
 1903, xvi, 217. 
 ' Plexneb. UttiverMyof Perm. Med. BmK., )902, xv, SS4. 
 ' BoxuETEB. Jowrrtalof Medical Research, 1903, ix, 146. 
 
 Digitized by Microsoft® 
 
of lesion was due to a plugging of the capillaries by endothelial 
 cells, aided probably also by a direct action of the toxins upon 
 the liver cells, another type could be explained only by the 
 occlusion of capillaries by fused masses of red blood corpuscles 
 (hyaline thrombi). Although he believed the fusion cf these 
 cells to be due to an agglutinin, he made no tests to demon- 
 strate its presence. 
 
 As far as I am aware no efforts have heretofore been made 
 to demonstrate experimentally a relation between agglutinins 
 and liver necroses. Focal necrotic lesions of the liver following 
 the injection of cytotoxic sera have been described by Dele- 
 zenne,* Kraus"* and Fukuhara," but in no instance has it been 
 suggested that agglutinins were concerned in their production. 
 
 Methods — Hasmagglutinative sera have been prepared by 
 injecting the rabbit with various organs and fluids of the dog. 
 As the investigation was carried on simultaneously with an- 
 other which had for its object the determination of the relative 
 specificity of the somatogenic cytotoxins,' it has been possible 
 to work with a large number of artificial sera. These include 
 sera prepared by injecting defibrinated blood, washed blood 
 corpuscles, blood serum, bile, urine and washed and unwashed 
 liver, pancreas, kidney and adrenal. All these sera possess the 
 power of agglutinating the red blood corpuscles of the dog in 
 vitro; all, however, as will be seen are not sufficiently power- 
 ful to cause red blood corpuscle thrombi in the living animal. 
 The sera was injected under ether anaesthesia, into the sap- 
 henous vein of the dog, and after varying periods of time the 
 animals were killed by chloroform. The histological methods 
 of examination include hardening in Zenker's, Muller's and 
 Fleming's solutions ; imbedding in celloidin and paraffin ; and 
 staining by hsematoxylin and eosin, methylen blue and eosin, 
 Mallory's connective tissue stain, Weigert's fibrin stain and 
 safranin. 
 
 General Summary of Results — Liver necroses were readily 
 produced by sera obtained by injecting bile, blood and blood 
 containing organs; they did not occur after the use of sera 
 
 'Dklezhnnk. C. B. del' Acad. de» Scietieee, 1900, cxxxi, 437, cited bv Sach Biochem 
 
 Centrall)., IWS, i, en. 
 » Kratjb. Verhandl. d. Deutsch. Path. Geselkch. , 1902, v, 58. 
 ' Fdkuhara. Zieffler's Beitrdge, 1904, xxxv, 434. 
 ' PEAncE. Concerning the Specificity of tlie Somatogenic Cytotoxins, Jour.Qf Med. Besearch, 
 
 1904) Xil, 1. 
 
 Digitized by Microsoft® 
 
prepared by injecting urine, serum and washed organs, except 
 in one instance in which an exceedingly high dose of washed 
 liver serum was administered. The power of these several 
 sera to produce conglutinative ■ thrombi and consequently 
 necroses varies naturally according to the degree of agglu- 
 tinating activity which they possess. While the most potent 
 sera judging by the reaction in vitro and by the extent and 
 severity of the lesions produced, are those prepared by inject- 
 ing blood and organs containing a large amount o± blood, as 
 the liver and kidney, it is worthy of note that the agglutinative 
 power does not depend entirely on this element, for sera almost 
 as powerful were obtained by injecting bile which contains 
 at best only blood derivatives. On the other hand the various 
 sera prepared by injecting blood free fluids and organs as 
 urine, serum and washed organs, while without power to 
 cause sufficient agglutination in the animal body to bring 
 about the formation of conglutinative thrombi, do cause dis- 
 tinct agglutination in vitro. The production of hasmagglutinins 
 therefore would appear to depend on the same principles 
 which govern the production of cytotoxins; that is, upon the 
 similarity in part at least, of the receptor (chemical) composi- 
 tion of the various cells and fluids.* 
 
 The more powerful sera produce multiple liver lesions 
 almost uniformly when administered in doses of i :iooo to 
 1:500;' smaller doses (less than 1:1000) may produce 
 necroses, but are usually without effect; large doses (1:500 
 to 1 :30o) usually cause death within a few hours. This last 
 result, however, has occurred only with the use of sera pro- 
 duced by injecting blood and unwashed adrenal. In this 
 regard it is of interest that the serum of an animal receiving 
 blood in doses of o.i cubic centimeter is as powerful as that 
 of animals receiving doses of five cubic centimeters. The 
 serum of animals receiving washed blood is as potent as that 
 of those receiving defibrinated blood. The power of unwashed 
 adrenal serum to produce liver necroses was but slightly less 
 
 ' For a consideration of this subject and also for a description of the cytolytic action of the 
 Tarions sera here mentioned see Journal of Medical Eeacarch 1904, xii, 1. Only the aggluti- 
 native action of the sera and the lesions produced in the liver will be considered in this 
 commnnicatlon. 
 
 » This method of indicating the dose refers to the relation between the amount of scrum 
 injected and the weight of the animal. Thus 1 :1000 indicates that an animal weighing 6000 
 grammes received 6 c. c. of serum or 1 c. c. to 1000 grammes of body weight. 
 
 Digitized by Microsoft® 
 
than that of sera prepared from blood. Unwashed hver and 
 kidney sera and serum prepared by injections of bde caused 
 liver necroses in the same doses as did the above sera, but 
 had no lethal eftect. On the other hand, washed liver serum, 
 the only one of this group (sera prepared from washed organs) 
 to cause necroses, did so only when administered in very large 
 doses (1 :266), and had no lethal action. Heated sera suffered 
 no loss of agglutinating power. 
 
 The necroses produced by these various sera develop in from 
 twenty-four to forty-eight hours. They are usually wide- 
 spread, closely massed and often confluent. The surface of 
 the liver presents a mottled appearance ; small, irregular, yel- 
 lowish brown or greyish yellow non-elevated areas being 
 sharply outlined by the deeply congested normal substance of 
 the liver. When the congestion is intense the necrotic areas 
 appear depressed owing to the swelling of the congested liver 
 surrounding them, and fine points of haemorrhage may be 
 seen. On section the brownish yellow areas have a distinct 
 hyaline appearance. The distribution of these lesions is almost 
 always superficial, few necroses being found more than half 
 a centimeter below the surface. When the distribution is not 
 so diffuse, the more prominent lesions are at or near the free 
 edges of the lobes and occasionally limited to the smaller 
 lobes about the gall-bladder. All sera produce uniformly the 
 same type of lesion, though not always the same degree of 
 destruction. 
 
 Histology — The necrotic lesion involves the greater portion 
 of a lobule, often the adjoining portions of several lobules. 
 Most frequently the necrosis is in the peripheral portion of the 
 lobule, but occasionally the central zone is involved. The rela- 
 tion to the portal vessels is striking, the altered areas fre- 
 quently entirely surrounding the portal vessels. The necrosis 
 is hyaline in character, the cells stain poorly with eosin, thus 
 being sharply marked off from adjacent unaltered cells. The 
 nuclei may be fragmented, but more commonly are shrunken 
 and irregular, and often fail entirely to take the stain. Leucocy- 
 tic infiltration seldom occurs. The capillaries in the midst of 
 these areas are tightly packed with red blood corpuscles. In 
 some capillaries the cells are loosely arranged, but in others 
 
 Digitized by Microsoft® 
 
homogeneous more or less cylindrical cast-like masses of 
 coalesced cells are seen. The endothelium of the capillaries 
 is usually unaltered and accumulations of endothelial cells do 
 not occur. An occasional endothelial cell or polymorphonu- 
 clear leucocyte may be found in the midst of these masses, 
 but are never present in considerable numbers. In the smaller 
 portal veins adjacent to the necrotic areas larger ovoid or 
 coalesced masses are seen. These may entirely fill the vessel, 
 though usually a few loose red cells may be found at the 
 periphery. The appearance of these masses is not constant. 
 All contain granular and hyaline areas; in the majority the 
 centre is granular and the periphery hyaline, but frequently 
 this arrangement is reversed. In some, the transition from 
 normal corpuscles to irregular, distorted or varicose corpuscles 
 and finally to hyaline and granular masses may be seen. The 
 hyaline areas upon close examination may be resolved into a 
 tangled mass of irregular varicose threads, representing appar- 
 ently greatly distorted and elongated corpuscles. Adherent 
 to the walls of the larger vessels distinctly hyaline masses 
 with stalactite-like processes may occasionally be seen. These 
 are apparently of the same origin as the smaller thrombi. 
 Fibrin does not enter into the formation of these thrombi, 
 though the detection of finer threads of fibrin is rendered diffi- 
 cult by the tendency of the altered corpuscles, especially after 
 Zenker hardening, to take the fibrin stain. Fibrin threads 
 may occasionally be found at the periphery of the larger 
 thrombi, but I have never seen them in connection with the 
 smaller plugs. 
 
 Taking advantage of the fact that an interlobular vein sup- 
 plies adjacent portions of several lobules, it has been possible, 
 especially in serial sections, to demonstrate a relation be- 
 tween a thrombus of such a vein and necrosis of the peripheral 
 portions of the lobules in direct relation to it. Also in areas 
 of extensive necrosis involving all but the hepatic vein zone 
 of many lobules, the portal terminals were found uniformly 
 occluded. While it is possible for the capillary thrombi within 
 the lobule to cause necroses the more important plugging 
 occurs, I believe, in the perilobular veins or the branches from 
 which they originate. The anatomical relation of the agglu- 
 tinated corpuscles to the lesion, in the absence of bacteria, 
 
 5 
 
 Digitized by Microsoft® 
 
phagocytic cell or fibrin thrombi, conclusively indicates that 
 the necroses are due to this obstruction of the smaller yems. 
 
 In some animals dying within a few hours after the injection 
 of large amounts of serum (i :400 to i :50o) an extraordmary 
 condition consisting of extreme congestion of the vessels and 
 disorganization of the liver cells was found. Vessels of all 
 sizes were widely dilated and filled with agglutinated cor- 
 puscles, while the pressure exerted caused a narrowing of the 
 columns of liver cells with more or less displacement. In one 
 instance the pressure was so great that complete disorganiza- 
 tion of the liver substance had occurred. This destruction was 
 so extreme that the histological picture, large sinus-like spaces 
 separated from one another by this compressed and irregularly 
 arranged columns of hepatic cells and filled with closely packed 
 agglutinated and more or less distorted red blood corpuscles, 
 was most suggestive of a cavernous angioma with thin tra- 
 beculae. In this early lesion the liver cells stained poorly, but 
 no distinct necroses were found in animals dying within eigh- 
 teen hours. The smaller doses, less than r aooo, on the other 
 hand, never produced changes in the liver other than an occa- 
 sional slight congestion. 
 
 In connection with these lesions the following questions 
 naturally occur. Why do not similar extensive thromboses 
 occur in other organs? Why do necroses occur almost exclu- 
 sively at the periphery of the liver and why are they limited 
 practically to the portal area? Capillary thrombi do occur 
 in other organs, especially in the glomeruli of the kidney, the 
 periphery of the lung, in the adrenal and in all these locations 
 they may be accompanied by minute haemorrhages. Their 
 occurrence is, however, infrequent and unaccompanied by 
 necrosis, except possibly in the glomerulus of the kidney. 
 That they do not occur more frequently is probably due to 
 peculiarities in the circulation in the various organs. It is 
 possible that the slow circulation of the portal system allows 
 the formation of larger masses of agglutinated corpuscles than 
 could form in the rapid flow of the other organs ; while on the 
 other hand the narrow capillaries of the liver are more readily 
 occluded. The first of these factors would also explain the 
 greater frequency of the lesion in the portal vein area of the 
 lobule. The more frequent involvement of the superficial 
 
 6 
 
 Digitized by Microsoft® 
 
portions of the liver can only be explained by Boxmeyer's 
 assumption that the blood supply of this portion of the liver 
 is less complete, and the collateral circulation less perfect than 
 in the deeper portions of the liver. 
 
 Attempts to Demonstrate the Mechanical Nature of the 
 'Thrombosis. — As the sera, with one exception, used in these 
 experiments was hasmolytic, and as several possessed other 
 cytotoxic constituents, it became necessary to rule out the 
 direct action of these cytotoxins upon liver cells and also the 
 possibility of an obstruction of the circulation by the remains 
 of hsemolysed red blood corpuscles. Several observations bear 
 on this point and support the theory of mechanical obstruction 
 by agglutinated corpuscles. In order to definitely determine 
 that an agglutination similar to that seen in vitro, takes place in 
 the circulating blood of an animal receiving an agglutinative 
 serum, the blood of a dog was examined four minutes after the 
 injection of serum in the dose of 1 1350. Blood taken quickly 
 from the heart, the superior vena cava, and the cut surface of 
 the liver, and diluted with salt solution showed abundant 
 coalesced masses of red blood corpuscles. Upon standing, a 
 light jelly-like clot formed with a supernatant fluid containing 
 distinct agglutinated masses not readily disintegrated by agita- 
 tion. Microscopically these masses, which varied in size, were 
 seen to be composed of from twenty to seventy or more red cor- 
 puscles many of which were distorted or elongated and often 
 so fused together as to be indistinguishable one from the other. 
 In another instance twenty-four hours after injection (dose 
 1 :985), masses of the same character were found in blood from 
 the cut surface of the liver. Here the cells were so fused to- 
 gether that the outlines of individual cells could not be recog- 
 nized. 
 
 That agglutinin alone is the important factor in producing 
 these necroses is shown by experiments with serum prepared 
 by injecting bile. This serum, while powerfully agglutinative, 
 unlike the other sera used, has no hsemolytic power. It causes, 
 however, in doses of i r/oo widespread lesions of the liver 
 similar to those described above. Attempts to render other 
 sera hsemolytically inactive by heating have been unsuccessful. 
 Animals receiving such sera develop hsemoglobinuria and the 
 
 Digitized by Microsoft® 
 
renal changes which usually follow the administration of a 
 hsemolytic serum, as well as typical liver lesions, the result of 
 the action of the agglutinin. 
 
 As a somewhat imperfect test" of the mechanical nature of 
 the thrombosis, attempts were made to produce necroses by in- 
 jecting agglutinated masses of red corpuscles freed oi the 
 excess of immune serum by repeated washing. These masses 
 were prepared by agglutinating dog's blood in vitro, and wash- 
 ing hastily in several changes of salt solution. Agglutinated 
 masses, thus prepared, injected into the mesenteric vein in 
 doses of five to ten cubic centimeters seldom produce necroses 
 and those which do occur are not of the same type or so numer- 
 ous as those produced by the injection of agglutinative serum. 
 They represent the type of granular necrosis with nuclear 
 fragmentation rather than the hyaline type and are very few 
 in number. Associated hyaline thrombi were found only occa- 
 sionally. In all these experiments the spleen contained a very 
 large amount of altered blood pigment, indicative of the fate 
 of the injected cells. I consider the results of these experi- 
 ments to be of little significance one way or the other, for the 
 injection of five or ten cubic centimeters of agglutinated cor- 
 puscles, an amount readily destroyed in the animail, is in no 
 way comparable to the extensive agglutination resulting from 
 the cumulative action of ten to twenty cubic centimeters of 
 agglutinative serum. 
 
 Various controls as the injection of defibrinated normal 
 dogs blood, washed hsemolyzed dogs corpuscles, the blood 
 and the clear serum of dogs in which the liver lesions had 
 been produced, and normal rabbits serum have been uniformly 
 negative. 
 
 Effect of Repeated Injections. — The severity and extent of the 
 lesions produced by the injection of agglutinative serum sug- 
 gested the possibility of producing, by repeated injections, 
 conditions analagous to acute yellow atrophy or the more acute 
 forms of cirrhosis. With this object several animals have 
 received serum intravenously at intervals of one to three 
 weeks. Thus far no permanent lesions have been found in ani- 
 mals which have been killed at the end of seven weeks after 
 
 1. Thie condition of this teBt Is not perfect, for the haemoIyUc amboceptor of the semm 
 are of course fixed by the red corpuscles. 
 
 Digitized by Microsoft® 
 
receiving three injections. Indeed, from a study of the liver 
 necroses in a small number of animals which have been allowed 
 to live more than a few days, there is very distinct evidence 
 that these lesions are completely repaired. The observations 
 upon this phase of the subject are, however, too few to warrant 
 a definite statement at this time. 
 
 May Agglutinins be a Factor in the Production of Liver 
 Necroses in Man? — In considering this question it is not neces- 
 sary to go into the general literature bearing upon the etiology 
 of liver necroses. That these lesions may be produced by the 
 direct action of bacteria (B. tuberculosis, B. pestis, and the 
 pyogenic cocci) is well known. Mallory^^ has shown very con- 
 clusively that they may be due also, especially in typhoid fever,- 
 to occlusion of the capillaries by clumps of phagocytic endo- 
 thelial cells with or without associated fibrin coagula. That a 
 strong toxin in the circulating blood may produce the so-called 
 central necroses is also indicated by his investigations. There 
 remain, however, many focal necroses which cannot be satis- 
 factorily explained. Associated with such are frequently 
 found capillary thrombi described as hyaline or fibrinous. It 
 seems probable that these, in view of the facts set forth in this 
 communication, may be agglutinative red blood corpuscle 
 thrombi and the essential element in causing the necroses with 
 which they are associated. Flexner and Boxmeyer working 
 independently came to the conclusion that the so-called hyaline 
 thrombi were composed of fused red blood corpuscles. Flex- 
 ner found such thrombi in the intestine of typhoid, the stomach 
 of carbolic acid poisoning and in broncho-pneumonia; and in 
 the rabbit produced them experimentally by the injection of 
 agglutinating substances as ricin, ether and dog's serum ; Box- 
 meyer found them in the liver necroses of experimental hog 
 cholera. It is significant that the latter could readily dis- 
 tinguish necroses due to such thrombi from those due to 
 phagocytic cell thrombi. Many of those who have studied hyaline 
 thrombi, state that they are composed of fused and altered red 
 blood corpuscles, but do not explain the mechanism of the fusion. 
 Ziegler's^^ opinion that they are formed largely from the colorless 
 
 •1 Mallobt. JowiuU of Exper. Medicine, 1898, Ui, 611; Jownal cf Medical Setearch, 
 
 1901, vl, 364. 
 " ZaousB. ffmeral Pathology, Eng. Ed., 1899, p. 122. 
 
 Digitized by Microsoft® 
 
elements of the red blood cells, is of interest in view of our newly 
 acquired knowledge of agglutination and haemolysis. 
 
 In attempting to bring agglutinative thrombi into relation 
 with liver necroses in man it is necessary to explain the origin 
 of the agglutinin. The possibility of its development as a 
 bacterial product in the course of infectious diseases, naturally 
 suggests itself. It is well known that a considerable number 
 of bacteria produce hsemolysins, and it has been demonstrated 
 that a few produce also hsemagglutinins. This agglutinating 
 power is possessed also by several poisons not of bacterial 
 origin. In this connection it is significant that most of the 
 conditions in which hyaline thrombi have been described are 
 of bacterial etiology as hog cholera, typhoid fever, broncho- 
 pneumonia, diphtheria and infections with the pyogenic cocci; 
 or intoxications as poisoning by ergotin, corrosive sublimate, 
 carbolic acid, ricin and abrin. Of considerable interest in this 
 connection is Flexner's reference to the occurrence of agglu- 
 tinative thrombi in the liver lesions of eclampsia. The occur- 
 rence of hyaline thrombi m the vessels of the rabbit's ear, after 
 freezing, points also to an injury of red cells leading to agglu- 
 tination. 
 
 It seems probable that the application of experimental 
 methods may demonstrate that agglutinins may be formed in 
 various infectious and toxic conditions in sufficient quantity 
 to cause agglutinative (hyaline?) thrombi with the develop- 
 ment of liver necroses and possibly other lesions. Experi- 
 ments, now in progress, with the products of antolyzed bacteria 
 indicate the presence of .such a body in sufficient amount to 
 produce agglutinative thrombi. 
 
 10 
 
 Digitized by Microsoft® 
 
Beprinted from Albany Mbdioal Anhals, November, 1904 
 
 CONCERNING HEMAGGLUTININS OF BACTERIAL 
 
 ORIGIN AND THEIR RELATION TO HYALINE 
 
 THROMBI AND LIVER NECROSES.* 
 
 By RICHARD M. PEARCE, M. D., : 
 
 AND 
 
 CHARLES K. WINNE, Jr., M. D. 
 (From the Bender Laboratory, Albany, N. Y.) 
 
 Elsewhere, one of us^ has described the hemagglutinative 
 power of various cytolytic immune sera and the relation of the 
 red blood corpuscle thrombi (hyaline thrombi) resulting from the 
 injection of such sera to focal hyaline necroses of the liver. The 
 distinct character of the liver lesions thus produced, and the fact 
 that they could be explained only by the presence of agglutinated 
 red blood corpuscle thrombi, suggested the possibility that similar 
 lesions occurring in various infectious diseases might be due to the 
 development of hemagglutinins of bacterial origin. In order to 
 test this supposition we have attempted to demonstrate the presence 
 of hemagglutinins in filtrates of bacteria treated in various ways, 
 and, by the injection of the same into animals, to produce aggluti- 
 nated red blood corpuscle thrombi and associated liver necroses. 
 
 It is not our intention to go into the general literature of these 
 subjects, but merely to touch upon those points which relate to 
 infectious conditions. The more important views concerning liver 
 
 *Thi9 study has been conducted under a grant from the Rockefeller Institute for 
 Medical Research. Published also in the American Journal of the Medical Sciences, 
 October, igo4 
 
 Digitized by Microsoft® 
 
necroses may be found in the papers of Flexner,^ Mallory,^ Box- 
 meyer* and Opie/ That the so-called hyaline thrombi may be 
 composed of altered and fused red blood corpuscles we consider 
 to be definitely established. The recent studies of Flexner," Box- 
 meyer, and Pearce offer conclusive evidence of such origin and 
 support the earlier observations of Klebs^ and Ziegler.^ The occur- 
 rence of hyaline thrombi in connection with various infectious and 
 toxic conditions has been emphasized by Welch,' who, in his sum- 
 mary of the literature, mentions particularly infections with S, 
 pyogenes aureus and B. cholerae suis, and offers a suggestion con- 
 cerning their origin from red blood corpuscles. In a later dis- 
 cussion^" of the subject he brings forth the hitherto unpublished 
 observations of Fisher concerning the plugging of small vessels by 
 agglutinated red cells as the result of injecting typhoid bacilli, and 
 calls attention to the evidence in favor of the interpretation of 
 many of the hyaline thrombi as derived from fused red blood cor- 
 puscles. Other important observations are those of Flexner, who 
 found similar thrombi in the intestine of typhoid fever, in broncho- 
 pneumonia and in experimental diphtheria ; and also of Boxmeyer, 
 who could distinguish, in the livers of rabbits dying of experi- 
 mental hog cholera, between necroses due to endothelial cell 
 thrombi and those due to agglutinated red blood corpuscle thrombi. 
 The hyaline thrombi occurring in various intoxications, due to 
 either organic or inorganic poisons, and in toxic conditions of 
 doubtful origin, as eclampsia, are undoubtedly of similar nature; 
 but as they do not come within the scope of this study, will not 
 be further considered. 
 
 Of considerable historical interest in connection with this ques- 
 tion of the occurrence of red blood corpuscle thrombi in the course 
 of infectious diseases are the observations of Hueter,^^ made in 
 1873. As the result of his studies, to which attention has recently 
 been called by Welch,^° Hueter came to the conclusion that clump- 
 ing of red blood corpuscles, causing permanent or transient occlu- 
 sion of capillaries and small vessels ("globulose Stase und 
 Embolie") occurred not uncommonly in septic and other fevers. 
 His theory that this phenomenon constituted the underlying basis 
 of fever has never gained acceptance, but the observations are of 
 interest in connection with our recently acquired knowledge of 
 bacterial hemagglutinins, v. Recklinghausen" considers this 
 globular stasis of Hueter to be analogous to a condition found in 
 
 Digitized by Microsoft® 
 
the capillaries after transfusion of alien blood, and which he 
 describes as an "agglutination" of red blood cells. 
 
 In planning this investigation it was evident that as necroses of 
 the liver are most frequently associated with infections by the bacil- 
 lus of typhoid fever and the bacillus of hog cholera, the bulk of 
 the work should be done with these organisms. In order, however, 
 that no opportunity might be lost, the literature dealing with bac- 
 terial hemolysins was examined in the hope of finding reference 
 to organisms which produced also hemagglutinins ; of twenty-nine 
 such investigations^^ but six^* refer to the production of hemagglu- 
 tinins by bacteria. Kraus and Ludwig found that bouillon cultures 
 of S. pyogenes aureus and Vibrio Paris had the power of agglu- 
 tinating red blood corpuscles in vitro in from ten to twenty-four 
 hours at room temperature. The mass which settled at the bottom 
 of the tube was seen upon agitation to be composed of small clumps 
 of red blood corpuscles which could not easily be disintegrated. 
 The agglutinin they considered to be independent of the hemolysin ; 
 but, on the other hand, they state that organisms which do not 
 hemolyse also do not agglutinate. Small amounts of the cultures 
 caused agglutination without hemolysis. Volk and Lipschiitz, 
 working with staphylolysin and vibriolysin, describe agglutination 
 in connection with incomplete hemolysis of rabbits' blood, but did 
 not observe it with other bloods. Volk, as a result of his obser- 
 vations, comes to the conclusion that agglutination is due not to a 
 specific body, but to non-isotonicity of the fluid. Calamida states 
 that no agglutination precedes the hemolysis caused by bouillon 
 cultures of the chicken-cholera bacillus. Flexner mentions agglu- 
 tinins of low activity in filtrates of B. pyocyaneus, B. typhosus, 
 and S. pyogenes aureus. Kayser, in his study of colilysin, ob- 
 served a clumping of red cells in connection with incomplete 
 hemolysis, but not when complete solution occurred. Of greatest 
 interest, however, and very suggestive, is the observation of Eisen- 
 berg," who found the blood of an individual with a generalized 
 pyocyaneus infection to be both auto-agglutinative and iso-agglu- 
 tinative. 
 
 As the result of this survey, it seemed advisable to utilize B. 
 pyocyaneus and S. pyogenes aureus in addition to B. typhosus 
 and B. cholerae suis; and these, with B. dysenterias and S. pyo- 
 genes, constituted the group of organisms finally selected for study. 
 
 Digitized by Microsoft® 
 
Methods. In previous attempts (Reed/" Flexner, Mallory, 
 and Boxmeyer) to produce liver necroses by the injection of micro- 
 organisms, either the living culture, grown in the ordinary way, 
 or its filtrate has been used. With the use of the former it has 
 been difficult to distinguish between the direct action of the organ- 
 ism and the effect produced by its toxin ; with the latter the action 
 of intracellular toxins has been excluded. In our experiments the 
 direct action of the microorganisms has been eliminated by the use 
 of filtered fluids, and an attempt has been made to obtain the intra- 
 cellular constituents of the bacterial cells by the utilization of 
 methods destined to cause its disintegration. Various writers have 
 described methods by which the bacterial cell is destroyed and its 
 receptors liberated. Of these methods we have tried, with certain 
 modifications, the processes of autolysis described by Conradi,^^ 
 Neisser" and Shiga, and Brieger and Mayer.^" The various 
 organisms have been grown for twenty-four to seventy-two hours 
 on agar in large flat flasks having a surface of twenty by eight 
 centimetres. The growth on several flasks, usually three or four, 
 has then been scraped off, subjected to 55° C. for forty-five min- 
 utes, and placed in tubes containing 0.85 per cent, salt solution 
 at a temperature of 37.5° C. for two to three days. In some 
 instances the fluid has stood for longer periods, varying from one 
 to seven weeks. The fluid has then been filtered (Chamberland 
 B. and F.) and injected intravenously into animals after testing 
 its agglutinative power in vitro. At various stages of the process 
 suitable controls were made to demonstrate absence of contamina- 
 tion and the efficiency of the filters. 
 
 In addition to this method, of growing the organism, an attempt 
 was also made to stimulate the production of hemagglutinin by pre- 
 viously growing the organisms in media containing red blood cells. 
 This method was suggested by the theory of "bacteriogenic cyto- 
 toxins," advanced by Welch in his Huxley Lecture: "Certain 
 substances of the host of cellular origin, assimilated by the para- 
 sites through the possession of haptophore groups with the proper 
 affinities, become anchored to receptors of the parasitic cell, which, 
 if not too much damaged, is thereby simulated to tlie overproduc- 
 tion of like receptors; these excessive receptors of the parasite, if 
 cast off into the fluids or the cells of the host, then constitute 
 intermediary bodies, etc." The writer further states : "by a similar 
 mechanism it is possible that parasites may be stimulated to the 
 
 Digitized by Microsoft® 
 
production also of antibodies other than cytotoxins, such as various 
 agglutinins, precipitins, antienzymes — all adjusted against the 
 host." 
 
 Assuming that if this hypothesis might apply to organisms grow- 
 ing in living cells or fluids, it might hold also for those grown on 
 cells outside the body, typhoid and hog cholera bacilli have been 
 grown in .various blood preparations. These include both defi- 
 brinated and washed blood added in various proportions to salt 
 solution, and fresh sterile rabbits' organs. The period of growth 
 in the various blood mixtures varied from twenty-four to seventy- 
 two hours ; in some instances inoculations into fresh blood solutions 
 or upon fresh organs being made on each of three successive days. 
 The sterile organs were prepared according to the technique de- 
 scribed by Flexner.^" By these various methods it was hoped that 
 the organism would be stimulated to produce agglutinins and that 
 this power would be retained during the subsequent growth on 
 agar, giving in the finished autolyzed product, a fluid of appreciably 
 augmented agglutinating power. Observations supporting this 
 theory have been made by Marmorek and by Simon, who found 
 that streptococci grown in media composed of leukocytes alone, 
 or with bouillon added, secrete a toxin of considerable power. 
 Simon^^ believes that the streptococcus in order to secrete toxin 
 must meet a certain resistance to its development, and that this 
 resistance is furnished by something in the leukocytes. This ex- 
 planation, put in the terms of Ehrlich, corresponds to Welch's 
 hypothesis. Our attempts to produce free agglutinin receptors by 
 growing the organisms in blood we consider analogous to Simon's 
 method of producing toxin by growth in leukocytes.* 
 
 Another method employed was the growth of an organism in 
 heated rabbit serum. This was suggested by the observation of 
 Besredka,^^ that the streptococcus produces an hemolysin when 
 thus grown. 
 
 Results. The tests for agglutination were made by adding 
 nine parts of the various filtrates to one part of blood. Observa- 
 tions were made at room temperature and at 37.5° C. at varying 
 
 *Another point in favor of this theory of bacteriogenic antibodies is the observation of 
 Theobald Smith (verbal communication) that recently isolated cultures of the hog- 
 cholera bacillus have a greater power to produce liver necroses than have old cultures. 
 Accepting Boxmeyer's work on the relation of agglutinative red blood corpuscle thrombi 
 to such lesions, the explanation naturally suggested is, that residence in the animal has 
 stimulated the power of the bacillus to produce agglutinin. 
 
 Digitized by Microsoft® 
 
periods up to twenty-four hours. The clumping of red blood cor- 
 puscles by bacterial hemagglutinins is of slight degree as compared 
 with that produced by various cytolytic immune sera. The power- 
 ful agglutination of the latter is never seen, and for this reason 
 more careful comparison with controls is necessary in order to 
 detect slight differences in reaction. This feeble activity renders it 
 difficult to make definite statements concerning some of the organ- 
 isms, studied. While all (B. typhosus, B. cholerae suis, B. pyocy- 
 aneus, B. dysenterias, S. pyogenes aureus and S. pyogenes) in the 
 majority of tests, in vitro, showed a slight agglutination as com- 
 pared with controls the effect of filtrates of B. dysenterise, B. 
 cholerae suis and B. pyocyaneus was so slight as to be considered 
 doubtful. With other organisms the agglutination was more evi- 
 dent, being most distinct with filtrates of S. pyogenes. Upon 
 microscopic examination (hanging drop) the clumping caused by 
 the streptococcus was always in the form of massive, more or less 
 branching, rouleaux; while with all other cultures an irregular 
 agglutination of a small number of corpuscles was seen. The 
 agglutination due to the streptococcus could be seen after fifteen 
 to twenty minutes, as the corpuscles settled, while with the other 
 organisms it could be seen only upon agitation after a considerable 
 period of time. More definite results were obtained with rabbits' 
 blood than with dogs' blood. 
 
 Of unusual interest and very suggestive are the agglutination 
 tests with filtrates of typhoid bacilli previously grown in blood 
 suspensions and on fresh sterile organs ; though carried out in but 
 a small number of tests, they appear to be confirmatory of Welch's 
 hypothesis mentioned above. These filtrates, especially those of 
 bacilli grown on sterile organs, exhibit an appreciably greater 
 power of agglutination than the filtrates of organisms grown on 
 agar only. 
 
 While no effort was made to study especially bacterial hemol- 
 ysins, it is of interest that ,the autolytic products of bacteria as 
 here prepared have never caused hemolysis. This result is in 
 harmony with the observations of Jordan,^^ Williamson''* and 
 others who hold that the hemolysis produced by many bacterial 
 filtrates (B. typhosus, B. pyocyaneus) is to be explained by the 
 presence of an alkali elaborated in the growth of the organism. 
 The filtrates used in our tests have all been faintly acid to phenol- 
 phthalein. 
 
 Digitized by Microsoft® 
 
The animals selected for the injection experiments were dogs 
 and rabbits ; the former receiving five to twenty cubic centimetres 
 (dose I :i 5 14 to i :235) of the filtrates in either the saphenous or 
 mesenteric vein; the latter one to twenty cubic centimetres (dose 
 I :i790 to 1 :64) in either the mesenteric or marginal ear vein. In 
 all, twenty-nine injections have been made; in no instance did 
 death occur as the immediate result of the mjection. Necroses 
 of the liver occurred in ten of the animals. Two of these we 
 have thought best to exclude because of associated coccidiosis, 
 although the necroses appeared to be entirely independent of the 
 parasitic invasion ; another has been excluded also because of the 
 chronic character of the lesion. The remaining seven cases ex- 
 hibited typical necroses ; two occurring in dogs and five in rabbits. 
 These lesions followed the injection of filtrates of the typhoid 
 bacillus in three cases, of the hog-cholera bacillus in three cases, 
 and of the dysentery bacillus in one case. Filtrates of the B. pyo- 
 cyaneus, S. pyogenes aureus, and S. pyogenes did not cause 
 necroses. These necroses are not always visible to the naked 
 eye, but when seen macroscopically they appear as miliary points, 
 or occasionally as round or oval areas or linear streaks of grayish 
 color, and distinctly separated from the normal liver tissue. They 
 are very few in number compared to those produced by the 
 agglutinins of cytolytic immune sera, and are usually limited to 
 the superficial portions of the liver. Portions of the liver in 
 contact with other lobes or adjacent organs are more frequently 
 involved than the free surfaces. Histologically, the liver cells 
 exhibit a hyalin necrosis, the nuclei being entirely destroyed, or 
 evident merely as faint rings ; fragmented or pyknotic nuclei were 
 seldom seen. In many of these areas the vessels are filled with 
 red blood corpuscles more or less altered and fused, and sugges- 
 tive of hyaline thrombi. This condition, however, is not constant, 
 and is not so evident as in the n xerotic lesions produced by cytolytic 
 immune sera. On the other hand, larger vessels, especially the 
 smaller branches of the portal vein, contain typical agglutinative 
 red corpuscle thrombi more or less completely occluding the ves- 
 sel. Similar thrombi of small vessels have been seen in other 
 organs, as the kidney, spleen, and lymph nodes, but without defi- 
 nitely associated lesions. Other thrombi, as those composed of 
 endothelial cells, fibrin, etc., have not been seen in the liver. 
 
 Fatty changes in the liver and kidney, intense congestion of 
 
 Digitized by Microsoft® 
 
the spleen and lymph nodes, and a proliferation of the endothelial 
 cells of the lymph sinuses have been found not infrequently after 
 the injection of typhoid and hog cholera filtrates. A relation 
 between these lesions and the injected fluids cannot, however, be 
 readily demonstrated; the proliferation of endothelial cells, espe- 
 cially, is not infrequent in apparently normal animals. That some 
 serious injury to the kidney is brought about is indicated by the ap- 
 pearance of albumen in the urine after the injection of filtrates 
 of B. typhosus and B. dysentriae. 
 
 The results of the injection of filtrates of typhoid and hog 
 cholera bacilli grown upon blood media, while very suggestive, 
 are not conclusive, owing to the small number of observations. 
 But six filtrates so grown were injected. Filtrates of the typhoid 
 bacillus grown in blood suspensions, while possessing an increased 
 agglutinating power in vitro, did not cause necroses upon injection ; 
 on the other hand, two different filtrates of the same culture grown 
 for several days on sterile rabbits' organs caused the most extensive 
 lesions observed in the course of this investigation. The bacillus 
 of hog cholera grown in blood suspensions caused, in one instance, 
 lesions in the liver. The attempt to produce necroses with such 
 filtrates has therefore been successful in three of six experiments ; 
 while with filtrates of the sam.e organism grown on plain agar 
 necroses resulted in three of twelve attempts. These results are 
 suggestive in connection with the increased agglutinating power 
 observed in vitro, but the small number of experiments does not 
 warrant definite conclusions. 
 
 Filtrates of B. typhosus grown in heated rabbit serum produced 
 no lesions upon injection into rabbits. 
 
 Conclusions. The filtered autolyzed products of a number of 
 bacteria possess hemagglutinins of low activity; the formation of 
 these may be stimulated by previously growing the microorgan- 
 isms on media containing fresh red blood cells. The injection of 
 these filtrates into dogs and rabbits is followed in some instances 
 by the occurrence of liver necroses with associated fused red blood 
 corpuscle thrombi ; like thrombi may occur in other organs. These 
 thrombi are similar in structure to those generally described as 
 hyaline thrombi. The demonstration of agglutination in vitro and 
 the association of the fused red cells with the liver lesions suggest 
 a relation of the agglutinin to the lesions. This relation, however, 
 cannot be considered as proven, for it is impossible to exclude the 
 
 Digitized by Microsoft® 
 
effect of toxins acting directly upon the liver cells ; on the other 
 hand, the similarity of the lesion to the very definite necroses 
 caused by he powerful agglutinins of cytolytic immune sera 
 strongly supports the possibility of such a relation. 
 
 The results observed by us are in accord with the slight potency 
 of the agglutinin obtained. It is probable that methods which 
 cause a more complete disintegration of the bacterial cells, as 
 those employed by Vaughan^" and by MacFadyan and Rowland,"' 
 would yield an agglutinin as powerful, comparatively, as the 
 toxins so derived, and capable of producing lesions more analogous 
 to those caused by the agglutinins of cytolytic immune sera. 
 
 We present this communication, therefore, not as an attempt to 
 prove that all hyaline thrombi are composed of fused red blood 
 corpuscles, or that all liver necroses are caused by such thrombi, 
 but rather to indicate the evidence in favor of the view that these 
 lesions, in some instances, may be due to bacterial hemagglutinins. 
 
 REFERENCES 
 
 1. Pearce. Concerning the Specificity of the Somatogenic Cytotoxins, Journal of 
 Medical Research, 1904, vol. xii. p. i; The Experimental Production of Liver Necroses 
 by the Intravenous Injections of Hemagglutinins, ibid., 1904, vol. xii., No. 3; also 
 Albany Medical Annals, 1904, xxv. 576, 696. 
 
 2. Flexner. Johns Hopkins Hospital Reports, 1897, vol. vi. p. 259. 
 
 3. Mallory. Journal of Experimental Medicine, 1898, vol. iii. p. 611; Journal of 
 Medical Research, igoi, vol. vi. p. 264. 
 
 4. Boxmeyer. Journal of Medical Research, 1903, vol. ix. p. 146. 
 5- Gpie, Ibid., 1904, vol. xii. p. 147. 
 
 6. Flexner. University of Pennsylvania Bulletin, 1902, vol. xv. p. 324. 
 
 7. Klebs. Die allgemeine Pathologie, 18S7, Bd. ii. p. 113. 
 
 8. Ziegler. General Pathology, 1899, Eng. ed., p. 122. 
 
 9. Welch. Thrombosis and Embolism, Allbutt's System of Medicine, 1898, vol. vii, p. 
 160. 
 
 10. Welch. Huxley Lecture, Science, 1902, vol. xvi. pp. 804, 850. 
 
 11. Hueter. Deutsche Zeitschr. f. Chirurgie, 1873-74, vol. iv. pp. 105, 330. 
 
 12. V. Recklinghausen. Handb. d. allg. Path, des Kreislaufs u. d. Ernahrung, 1882, p. 
 I3S- 
 
 13. Ehrlich. Berl. klin. Wochenschr,, i8g8, Bd, xxxv. p. 273; Madsen, Zeitschr. f. 
 Hygiene, 1899, Bd. xxxii. p. 214; Neisser and Wechsburg, ibid., igoi, Bd. xxxvi. p. 
 299; Todd, Lancet, 1901, vol. clxi. p. 1663; Castellani, ibid., vol. clxii. p. 440; 
 Bullock and Hunter, Centralb. f. Bakt., 1900, Bd. xxviii. p. 865; Weingeroff, ibid., 
 1901, Bd. xxix. p. "JT]', Lubenau, ibid., 1901, Bd. xxx. pp. 356, 402; Levy and Levy, 
 ibid., 1901, Bd. xxx. p. 405; MacFadyan and Rowland, ibid., 1901, Bd. xxx. p. 753; 
 Breyman, ibid., 1902, Bd. xxxi. p 481; Otto, ibid., 1903, Bd. xxxiv. p. 44; Simon, ibid., 
 1904, Bd. xxxv. p. 440; Schw'oner, ibid., 1904, Bd. xxxv. p. 608; Kraus and Clairmont, 
 Wiener klin, Wochenschr., 1900, Bd, xiii. p. 49; Jordan, Journal of Medical Research, 
 1903, vol. X. p. 31; Abbott and Gildersleeve, ibid., 1003, vol. x. p. 42; Ruediger, 
 Journal of the American Medical Association, 1903, vol. xii. p. 962; Maclntyre, ibid., 
 1904 vol. xlii. p. 1073; Bordet, Ann. de I'Tnstitut Pasteur, 1897, tome xl, p. 177; 
 Besredka, ibid., 1901, tome xv. p. 886, Williamson, Transactions of the Chicago Patho- 
 logical Society, 1904, vol, vi. p. 71; Breton, C. R. de la Soc. de Biol., 1903, tome Iv. p. 886. 
 
 14. Kraus and Ludwig. Wiener klin. Wochenschr., 1902, Bd. xx. pp. 120, 382; Volk 
 und Lipschiitz, ibid., 1903, Bd. xvi. p. T394; Volk. Centralb. f. Bakt., igoS, Bd. xxxiv. 
 
 Digitized by Microsoft® 
 
p. 843J Calamida, ibid., 1904, Bd. xxxv. p. 618; Flexner, University of Pennsylvania 
 Bulletin, 1902, vol. xv. p. 324; Kayser. Zeitschr, f. Hygiene 1903, vol. xlii. p. 118. 
 
 15. Eisenberg. Centralb. f. Bakt., 1903, Bd. xxxiv. p. 739- 
 
 16. Reed. Johns Hopkins Hospital Reports, 1S95, vol. v. p. 379. 
 
 17. Conradi. Deutsche med. Wochenschr. , 1903, Bd. xxix. p. 25. 
 
 18. Neisser and Shiga. Ibid., p. 61'. 
 
 19. Brieger and Mayer. Tbid., p. 309. 
 
 20. Flexner. Journal of Experimental Medicine, 1896, vol. i. p. 211. 
 
 21. Simon. Centralb. f. Bakt., 1904, Bd. xxxv. p. 440. 
 
 22. Besredka. Ann. de I'lnstitut Pasteur, 1900, tome xv. p. 880. 
 
 23. Jordan. Journal of Medical Research, 1903, vol. x. p. 31. 
 
 24. Williamson. Transactions of the Chicago Pathological Society, 1904, vol. vi. p. 71. 
 25- Vaughan. Transactions of the Association of American Physicians, 1901, vol. 
 
 xvi. p. 217. 
 
 26. MacFadyan and Rowland. Centralb. f. Bakt., 1903, Bd. xxxiv. pp. 618, 765. 
 
 10 
 
 Digitized by Microsoft® 
 
Seprinted from the Aisaht Msdioal Ansaj.b, July 1804 
 
 COMPLICATIONS AND DEGENERATIONS OF 
 UTERINE FIBROMYOMATA. 
 
 A PATHOLOGICAL AND STATISTICAL STUDY OF TWO HUNDRED 
 AND EIGHTY CASES. 
 
 Read before the Philadelphia Obstetrical Society. Published in the Jour- 
 nal of the American Medical Association, 1904, XLII, 
 1344, {May 2i). 
 
 By ELLICE M'DONALD, M. D., 
 
 Assistant Besident Physician, Lying-in Hospital of the City of New York; Late Voluntary As- 
 sistant, Bender Hygienic Laboratory, Albany, N. Y. 
 
 (From the Bender Hygienic Laboratory, Albany, N. Y.) 
 INTRODUCTION 
 
 Recent discussions in the surgical and gynecologic branches 
 of the American, British and French medical organizations, 
 indicate a renewed interest in the question of treatment of 
 uterine fibromyomata. This is due largely to facts brought 
 forth as the result of the systematic routine pathologic 
 examination of all "fibroid" tumors. The frequency with 
 which fibroid tumors present degenerations and are accom. 
 panied by complications has upset the classical teaching in 
 regard to prognosis and treatment of these growths. More 
 particularly has interest been stimulated in regard to the 
 so-called "malignant degenerations," as is evidenced by the 
 prominent place which carcinoma and sarcoma occupy in 
 recent publications on myoma. At present, however, the 
 results of pathologic studies are not definitely collected, and 
 there is a lack of statistical comparison. The importance of 
 bringing together such results is evident, for the question of 
 operative interference and extent of operation must depend 
 
 Digitized by Microsoft® 
 
on information derived from combined clinical and pathologic 
 reports. In regard especially to the extent of operation are 
 the results of pathologic investigation of value. While in 
 the removal of a benign tumor it is without doubt ideal that 
 the patient should suffer as little mutilation as possible, still 
 that operation must be done which offers the least danger, 
 and at the same time takes into consideration the various 
 attendant complications. On this point recent investigations 
 indicate that the complications of myomata are by no means 
 to be lightly considered. 
 
 Aside from the tumor itself, the adjacent organs must be 
 taken into account. The question of the condition of the 
 ovaries is important from the fact that in supravaginal 
 hysteromyomectomy the trend of opinion among gynecolo- 
 gists is that an ovary should be left in order to ameliorate 
 the severe menopausal symptoms sometimes produced by 
 complete ablations. For this reason the microscopic study 
 of these organs is important in order to determine the fre- 
 quency of disease of the ovary in connection with myoma. 
 
 Also, after hysteromyomectomies and myomectomies, the 
 possibility of infection from a diseased tube is of importance, 
 not only as concerns the union of the wound, but also as 
 regards the mortality of these operations. Hence the 
 examination of the tubes for salpingitis, pyosalpinx, etc., is 
 of no slight value. 
 
 In the study which is here presented, an attempt has been 
 made to determine the frequency of the various degenera- 
 tions and complications and to base thereon certain general 
 conclusions concerning treatment, 
 
 GENERAL LITERATURE 
 
 Since Martin^ in 1888 made his study of degenerations in 
 205 cases of fibromyomata, others have reported similar 
 series, notably Noble,^ 258 cases; Cullingworth,* 100 cases; 
 Mary Scharlieb,* 100 cases, and Frederick,^ 125 cases. 
 Combined, these make 788 cases available for the classifica- 
 
 Digitized by Microsoft® 
 
tion as to degenerations. There is some slight difference in 
 the classifications of the five writers, but by taking certain 
 liberties with their tables,* the degeneration may be sum- 
 marized as follows: 
 
 Fatty degenerations 7 
 
 Calcareous infiltration ii 
 
 Necrosis SS 
 
 Cystic degeneration 34 
 
 Telangiectasis 46 
 
 Sarcoma 17 
 
 Associated carcinoma cervicis 15 
 
 Associated carcinoma corporis 17 
 
 Myxomatous (edematous) 8 
 
 Associated syncytioma ^ i 
 
 (Total number of cases, 788.) 
 
 Martin has not estimated his complications; so for the 
 purpose of that classification only the series of Noble, 
 Cullingworth, Mary Scharlieb and Frederick are available 
 They are as follows : 
 
 Appendicitis 13 
 
 Hydrosalpinx 41 
 
 Hematosalpinx 2 
 
 Ovarian cyst 32 
 
 Dermoid cyst 5 
 
 Ectopic pregnancy 4 
 
 Papillary carcinoma of ovary 2 
 
 Pyosalpinx 18 
 
 Salpingitis 10 
 
 Abscess of ovary 7 
 
 Tubo-ovarian abscess I 
 
 Dilated ureters and hydronephrosis I 
 
 Dilated ureter, hydronephrosis and pyelitis i 
 
 Parovarian cyst 3 
 
 Cystic ovaries 18 
 
 Hematoma of ovary I 
 
 *In this table associatefl tnmors are arbitrarily included. Martin's ten cases of enppn- 
 ration are placed under necrosis. Noble's case of syncytioma is retained, although there is some 
 donbt from the pathologic report as to whether there was a fibroid associated with the new 
 growth. This tumor was twice reported. The following extract is from the second publication, 
 pathologic report by IDr. Babconfc : "In the early description of the growth it was described 
 as a sarcomatous or endothcliomatouB degeneration of a fibroid tumor; but restndy gives no 
 evidence that flbromnscnlar tissue is present more than wonld accompany the growth of the 
 syncytiAl tumor. It is believed that thiB ie a primaiy syncytioma mftliemiro in a uterus not the 
 iiaX dS any otSbBt tumor." 
 
 3 
 
 Digitized by Microsoft® 
 
Tubercular peritonitis 
 
 22 
 
 Universal adhesions / 
 
 (Total number of cases, 583-) 
 
 In making this compilation a great difference in classifica- 
 tion was apparent. Frederick and Mary Scharlieb alone 
 mention cystic ovaries. Culling worth in his lOO cases 
 does not mention salpingitis. Many complications of a 
 purely clinical nature, as retroversions of the uterus, etc., 
 have been eliminated in order to bring the classification in 
 accord with that of this paper. 
 
 In a more recent publication, Daniels' in 124 cases found 
 pathologic lesions in the appendages or their immediate neigh- 
 borhood in seventy-two; in seventy cases of complicated 
 fibromyomata, lesions were confined to the ovary in twenty- 
 eight, to the tubes in twelve, and involving both in thirty. 
 
 Marscher^ in twenty-two cases of myomata found ectopic 
 pregnancy in one and suppurative diseases of the appendages 
 in four. 
 
 Bond' reports fifty cases, in which eight had tubal disease, 
 which he claims was due mainly to infection from miscarriage. 
 Hunner'" in a series of 100 consecutive cases occurring in 
 Kelly's service found pyosalpinx or tubo-ovarian abscess in six 
 cases, hydrosalpinx or perisalpingitis in fifteen cases and ovarian 
 cysts in nine cases. There was found in association malig- 
 nant disease in seven cases; sarcoma in two, adenocarcinoma of 
 the fundus in three, and squamous carcinoma of cervix in two. 
 Other writers have reported smaller series of cases or series 
 devoted to one complication as appendicitis'^ or salpingitis'^ 
 As these reports do not include all complications they have 
 not been quoted. 
 
 From this summary it will be seen that the complications 
 of fibromyomata are many and varied. It will be noted that 
 tubal complications form a very large percentage in Daniel's 
 series, while in the combined series they are not prominent; 
 also in the latter series pyosalpinx is numerically greater than 
 galpiagitis. This asay, I think, bte explained by the fact that 
 
 4 
 
 Digitized by Microsoft® 
 
the earlier writers did not have a complete microscopic 
 examination of the condition of the tubes. Frederick men- 
 tions cystic ovaries which occurred in about fourteen per cent, 
 of his cases. Mary Scharlieb estimates ovarian cysts and 
 cystic ovaries together as twenty-three per cent. 
 
 In regard to associated malignant tumors in the combined 
 series, carcinoma occupies a relatively high place. No dis- 
 tinction is made between "sarcomatous degeneration" of a 
 myoma and of sarcoma associated with myoma except in 
 Martin's paper. Of the more recent publications on sarcoma- 
 tous degeneration of fibroids, Baker and Graves'* publish thirty- 
 three cases of fibroid tumors, in which three were sarcomatous. 
 The percentage is truly large; but this is similar to the find- 
 ings of Ulesko-Stroganowa," who reported before the Russian 
 Society for Obstetrics and Gynecology, March 22, 1902, 100 
 cases of fibromata with ten cases of malignant degeneration. 
 From the description these are evidently divided into two 
 groups, leiomyoma malignum and true sarcoma. It is not 
 quite clear what is meant by the former. In the discussion 
 which followed, her views were generally opposed. From 
 the abstract of her paper it is impossible to say how many of 
 these were really sarcomatous. 
 
 Tate" and Bertha Van Hoosen" each report a case of 
 sarcomatous degeneration of fibromyoma, which will be 
 referred to later under the special heading. 
 
 From these papers it will be seen that "sarcomatous 
 degeneration," while not common, is sufficiently so to merit 
 further study; first, in regard to its frequency, and second, 
 as to whether it is a true transformation of a myoma or 
 merely an associated but independent growth. 
 
 MATERIAL 
 
 The material on which this study is based* represents 
 seven years' service of a general hospital (Albany Hospital), 
 
 *I wish to express my thanks to the operators from whom these cases came, more particu- 
 larly to.tlie snreeons of the Albany Hospital. I also wish to express my gratitude to Dr. Blnmer, 
 late director of the Bender Laboratory, and to his assistants, whose conscientions routine exami- 
 nations and exhaustive records nave made this report possible. I also wish to express my sincere 
 thanks to Dr. R. M. Pearce, present director of the laboratory, for constant assistance and 
 encooragement 
 
 Digitized by Microsoft® 
 
and includes all myomata removed during that time. Com- 
 plete gross and microscopic examinations have been made in 
 all cases. The series also includes 26 cases in which 
 myomata were found at autopsy. 
 
 Summary 01; 280 Cases of Fibeomyomata. 
 Character: — 
 
 Single "O 
 
 Multiple 170 
 
 Small ,• 79 
 
 Medium 59 
 
 Large ' 99 
 
 Very large 43 
 
 Subserous 67 
 
 Interstitial 85 
 
 Submucous 23 
 
 Combined lOS 
 
 Degenerations — 
 
 Slight hyaline 18 
 
 Moderate hyaline 32 
 
 Marked hyaline 19 
 
 Hyaline and calcareous 8 
 
 Calcareous 14 
 
 Calcareous, necrotic, hemorrhagic 2 
 
 Cystic 8 
 
 Myxomatous 4 
 
 Hemorrhagic 6 
 
 Necrotic 4 
 
 Necrotic and hyaline 3 
 
 Necrotic, cystic, infected i 
 
 Adenomyoma 6 
 
 Associated with adenocarcinoma , 6 
 
 Associated with squamous cell carcinoma of cervix 2 
 
 Sarcoma 3 
 
 Deciduoma malignum i 
 
 Complications — 
 
 Ovarian cysts jc 
 
 Parovarian cysts 2 
 
 Cystic degeneration of ovaries ee 
 
 Tubo-ovarian cyst j 
 
 Calcification of ovary ^ 
 
 Ovarian fibroma 2 
 
 6 
 
 Digitized by Microsoft® 
 
Ovarian carcinoma 2 
 
 Salpingitis 38 
 
 Salpingitis and hydrosalpinx 3 
 
 Salpingitis and tubo-ovarian abscesses 3 
 
 Pyosalpinx S 
 
 Pyosalpinx and tubo-ovarian abscesses 2 
 
 Extra uterine pregnancy, salpingitis and hydrosalpinx i 
 
 Hydrosalpinx 6 
 
 Hydrosalpinx and tubo-ovarian abscess i 
 
 Autopsies — 
 
 Total number . < 26 
 
 Deaths due to. tumor 2 
 
 PATHOLOGY 
 
 Character. — All these variations in fibroid tumors are fa- 
 miliar to operators and pathologists. " Multiple" in this 
 series is taken to mean more than one tumor. The division 
 as to size has been based on the following scale: Small, up 
 to four centimetres; medium, four to eight centimetres; 
 large, eight to sixteen centimetres; very large, above six- 
 teen centimetres in diameter. In multiple tumors the 
 diameter of the largest tumor was taken as the determining 
 factor. The largest tumor noted weighed ninety pounds, and 
 measured roughly 50x40x20 centimetres. The tumor was sin- 
 gle, showed considerable necrosis and calcareous degeneration, 
 and caused pressure symptoms which brought on the death 
 of the patient. It was removed postmortem. Several other 
 tumors were noted between forty and fifty pounds. It will 
 be noted that 142 tumors measured more than eight centi- 
 metres in diameter. 
 
 The analysis as regards position was not easy, as it is not 
 always plain into which of the classes a tumor should go. I 
 regret that there is excluded a most important class of broad 
 ligament fibromyomata, of which the data at hand would not 
 allow a proper estimation. 
 
 By "combined" is meant multiple tumors, one each of 
 which may be classed in one of the divisions. Thus it is 
 
 7 
 
 Digitized by Microsoft® 
 
seen that there are sixty-five multiple tumors which fall into 
 one class (always interstitial or subserous), and 105 multiple 
 which fall into at least two classes. 
 
 In one case of multiple fibromyomata a medium-sized 
 fibroid had become detached from its pedicle and lay retro- 
 peritoneally, being nourished by the mesenteric vessels. Simi- 
 lar cases have been reported by Alban Doran" and others.^* 
 
 The comparative infrequency of submucous myomata is 
 noticeable. 
 
 Simple Degenerations. — In tumors of such size (142 above 8 
 centimetres diameter) it is remarkable that a large number 
 with degeneration, particularly with hyaline degeneration, 
 was not met with. The combination of hyaline and calcare- 
 ous degeneration in eight cases is of interest. It would seem 
 that the tendency of these tumors is to undergo progres- 
 sive hyaline change going on to calcareous degeneration. 
 
 The twenty-three examples of calcareous degeneration 
 existed in old women, but one being below fifty. These 
 varied in consistence from those of the "Wombstone" type, 
 ten centimetres across, to those with discrete, moderate 
 calcareous deposits. 
 
 In the consideration of cystic degeneration, as Alban 
 Doran*° shows, a distinction should be made between true 
 cystic degeneration and necrosis going on to cavity forma- 
 tion. In this series, however, his standard of judgment that 
 a "fibroid is not truly cystic unless it includes a cavity con- 
 taining half a pint of fluid," has not been adhered to, but 
 tumors with smaller cavities have been judged as truly cystic. 
 There were no cysts of large size found, the largest containing 
 approximately 1,000 cubic centimetres of fluid. Cystic 
 degeneration is more especially interesting from the fact that 
 of Graves' four cases of cystic degeneration three were sarco- 
 matous, evidently a necrosis going on to cyst formation. 
 Mary Scharlieb's series shows six cases of sarcomatous 
 degeneration, and of these three were cystic. 
 
 8 
 
 Digitized by Microsoft® 
 
Necrosis and Inflammation. — All stages of inflammation 
 and necrosis were noted. Necrosis occurred mainly in large 
 subserous tumors, and was evidenced, on section, by brownish- 
 yellow areas. There was usually considerable associated 
 odema. In four cases this went on to liquefaction of the cen- 
 tral area. It is interesting to note in regard to the possibility 
 of infection that two cases were associated with salpingitis and 
 two with hydrosalpinx. One large submucous tumor with 
 multiple interstitial nodules was necrotic and sloughing. In 
 all cases of sarcomatous degeneration, hemorrhage and 
 necrosis were noted. 
 
 Adenomyoma. — The percentage of these tumors is about 
 that usually found. They offered in this series nothing 
 of special remark in the gross appearance. One showed 
 marked hyaline change and one calcareous degeneration. 
 
 Malignant Changes. — -Sarcomatous degeneration was found 
 in three cases. There was no definite appearance by 
 which these sarcomatous changes could be recognized. The 
 main characteristics were necrosis, with broken-down areas 
 and discolored spots from recent and old hemorrhages. A 
 yellowish-white glistening growth could be distinguished in 
 one case infiltrating the muscle and surrounding the necrotic 
 areas. During the period from which these reports were 
 taken six cases of sarcoma of the uterus occurred, in addition 
 to these three cases of sarcomatous degeneration of uterine 
 fibroids. 
 
 In case one it was possible to make a diagnosis in the 
 gross specimen, as the disease had infiltrated the uterine 
 tissue and gave a semi-necrotic appearance with irregular, 
 hemorrhagic, pinkish areas devoid of fibrous arrangement, 
 and a shell of malignant tissue infiltrating the tumor. 
 
 Cases two and three were also necrotic, soft and friable, 
 and in the macroscopic examination were taken for necrotic 
 fibroids. 
 
 Chorio-epithelioma infiltrated one case of fibroid tumor. 
 
 Digitized by Microsoft® 
 
This is, so far as is known, the first authentic instance of 
 such invasion reported. Complete reports of these four 
 specimens from the records of the Bender Laboratory follow : 
 
 Case i.— Mrs. M. N. Specimen consists of uterus with tubes and ovaries. 
 One tube and ovary are separate from the rest of the specimen, the tube 
 having been amputated at the junction with the uterus. 
 
 Uterus.— The uterus is greatly enlarged and measures 19 cm. long and 
 about 14 cm. in diameter. Except for an irregular area about 7 cm. in 
 diameter corresponding to the fundus, the peritoneal surface is everywhere 
 covered with firm adhesions. Near the superior surface of the fundus are 
 two globular tumor masses the size of a cherry, which are attached to the 
 body of the uterus by a slender pedicle of peritoneum and fibrous connective 
 tissue. Near the above tumor can be seen a pea-sized submucous tumor 
 similar to the above. On section these tumors are pearly white with fine 
 whorls of connective tissue. Adjacent to the origin of the attached tube 
 projects a tumor mass about 3^ cm. in diameter which is soft and on 
 section shows a thin capsule with central contents similar to that of the 
 cavity of the uterus. On section the cavity of the uterus is found to be 
 filled with a necrotic mass, the core of which is the size of an orange, 
 and of a pinkish, yellow color with the consistency of liver tissue. This 
 central mass shows whorl-like markings suggesting the cross-section of a 
 large fibromyoma. Extending from the central core and infiltrating into 
 the muscular coat is a stratum of extremely friable tissue from 2 to 4 cm. 
 in thickness. This stratum is throughout semi-necrotic and shows ir- 
 regular hemorrhagic areas, while at the outer border is an irregular zone 
 which on section presents a jelly-like whitish glistening surface. The 
 muscular coat of the uterus is under considerable tension, retracting 
 rapidly when cut. This muscular coat has an average thickness of about 
 S mm., but in many places is not over 2 mm. in thickness. The boundary 
 between the muscular and semi-necrotic mass is nowhere sharply drawn, 
 so that the muscular coat appears to be infiltrated by the mass lying 
 beneath it. Near the fundus of the uterus, the muscular wall is much 
 thicker than elsewhere. On section there is disclosed a definitely encap- 
 sulated tumor mass the size of a hen's egg which readily separates from 
 the capsule and presents on section a soft, dark red, irregularly mottled 
 surface. It is connected by a broad isthmus to the mass occupying the 
 cavity of the uterus. 
 
 Tubes.— The attached tube is markedly congested, but otherwise appar- 
 ently normal. Opposite tube is the same. 
 
 10 
 
 Digitized by Microsoft® 
 
Ovaries. — The attached ovary is cystic throughout, having numerous 
 cysts ranging from the size of a pigeon's egg to that of a pea, the sub- 
 stance of the ovary being crowded to one side. Opposite ovary shows 
 very many cysts ranging from 2 to 10 cm. in diameter. These cysts con- 
 tain clear straw-colored fluid. 
 
 Anatomic Diagnosis. — Sarcoma of uterus with necrosis and hemorrhage. 
 
 Microscopic Description. — Tissues, uterine wall, ovary, polypoid tumor 
 described in gross description. Uterine wall: Sections taken through 
 portion described as capsule in gross description show smooth muscle 
 infiltrated with masses of spindle-shaped cells having large- clear staining 
 vesicular nuclei and little protoplasm. Among these cells are seen large 
 irregular cells, the nuclei of which have a lobulated outline. Indistinct 
 mitotic figures are seen in some of the cells. These cells have an utterly 
 lawless arrangement, and deeper in the sections they form a continuous 
 mass containing numerous thin walled blood' vessels. Many of the sections 
 show hemorrhagic areas with red blood corpuscles in various stages of 
 degeneration. Some sections show simply masses of necrotic cells. 
 
 Polypoid Tumor. — Body of tumor shows smooth muscle largely replaced 
 by fibrous tissue, while the pedicle of the tumor is seen to be infiltrated 
 with cells similar to those described in tumor of the uterine wall. 
 
 Ovary. — Shows large follicular cysts and corpora lutea undergoing 
 organization. 
 
 Microscopic Diagnosis. — Uterus with fibromyomata invaded by spindle- 
 celled sarcoma; necrosis; hemorrhage; follicular cysts of ovary. 
 
 Case 2. — Miss S. A. T. Specimen consists of an irregular mass of tissue 
 composed of uterus and appendages and measures 18 x 13 x 7 cm. At 
 one portion is attached a normal Fallopian tube, which measures 11 cm. 
 in length, and an ovary i x 2 x 2^ cm. which has undergone fibrous 
 change. On section the irregular surface of the mass is found to be 
 due to the presence of large, roughly spherical tumors, varying in size 
 from 2 to II cm. in diameter. The original form of the uterus can not 
 be made out. The cavity of the uterus is pale and covered with a some- 
 what sticky secretion. The tumor masses which lie in the myometrium, 
 on section, have the characteristic appearance of myomata; one of the 
 large tumor masses has undergone degeneration, is soft and friable, and 
 contains a considerable amount of blood. Immediately beneath the serous 
 coat are a number of small pedunculated tumor masses which average 
 I cm. in diameter. On section these are similar to the large tumor masses. 
 
 II 
 
 Digitized by Microsoft® 
 
Anatomic Diagnosis.— M.altip\e interstitial and subserous myofibromata 
 of the uterus. Normal Fallopian tube. Fibroid ovary. 
 
 Microscopic Description.— In places a definite picture of fibromyoma 
 may be seen. Adjacent to such areas and differentiated from it only by 
 the difference of structure are areas of typical spindle-celled sarcoma. 
 The sarcomatous tissue is extensively necrotic, the necrosis in places being 
 universal, and in other places sparing the cells immediately about the 
 vascular channels. The necrotic areas are infiltrated with leucocytes. 
 Necrosis is seen as well in the myomatous portion, but is not so prominent. 
 
 Microscopic Diagnosis. — Fibromyoma invaded by spindle-celled sarcoma. 
 
 Case 3. — Mrs. S. E. Specimen consists of about twenty-five separate 
 pieces, which in aggregate make a mass about the size of a man's hand. 
 One large globular mass measuring about 18 cm. in diameter makes 
 up the bulk of the specimen. This mass presents a convex smooth sur- 
 face covering about three-fifths of the mass. This surface is smooth, 
 pale pink in color, part of surface being glistening and part of it show- 
 ing firm, blood-stained connective tissue bands. The remaining periphery 
 of this mass is exceedingly irregular and consists of various sized strips 
 of tissue which are soft, friable and deeply stained and show numerous 
 large blood vessels. The contents of many of the blood vessels consist 
 of a firm coagulum. These shaggy strips are distinctly necrotic and are 
 undoubtedly further represented by the numerous small pieces of semi- 
 necrotic material which go to make up the remainder of the specimen. 
 The comparatively smooth surface of this mass is irregularly lobulated 
 and shows beneath the surface a number of small cysts filled with clear 
 amber-colored fluid, also numerous pea-sized to cherry-sized spherical 
 tumor masses, some of which present on cross-section a pearly white 
 glistening surface with a whorl-like arrangement of the tissue elements. 
 On section of the large tumor mass, the cut surface presents a thin cortex 
 varying from 2 to 10 cm. in thickness composed of a firm, pinkish-white 
 tissue beneath which is a whitish, coarsely granular surface showing 
 the remains of numerous small cysts, most of which are filled with a 
 blood-stained fluid. As one gets further away from smooth surface the 
 secton becomes considerably blood stained and takes on a yellowish 
 necrotic look and the cysts become larger in size until finally the semi- 
 necrotic shredded mass described above is reached. The small pieces 
 accompanying this specimen consists of stringy masses of distinctly 
 necrotic tissue similar in character to that described as occurring in the 
 center of the large tumor mass. One of the larger pieces accompanying 
 
 12 
 
 Digitized by Microsoft® 
 
the specimen is thin, averaging about i cm. in thickness, and has on its 
 smooth side considerable adipose tissue and resembles in consistency, on 
 cross-section, the firm tissue described as forming the capsule in the main 
 tumor mass. No evidence of tubes and ovaries. 
 Anatomic Diagnosis. — Degenerated myoma of uterus. 
 
 Microscopic Description. — Sections through portion described as capsule 
 of large tumor show thin layer of smooth muscle tissue with invading 
 masses of medium-sized cells having rather darkly staining oval or spindle- 
 shaped nuclei and little protoplasm. Many irregular mitotic figures are seen 
 among these cells. The deeper, softer portions of the tumor are seen to 
 be composed of solid masses of these cells. Numerous thin-walled blood 
 vessels are seen in the cell masses. No evidences of necrosis are seen. 
 Many blood vessels at the periphery of the tumor are completely plugged 
 with the tumor cells. Sections through nodular masses at surface of 
 tumor show smooth muscle tissue rich in cells and a moderate amount of 
 fibrous tissue. 
 
 Microscopic Diagnosis. — Myoma of uterus invaded by spindle-celled 
 sarcoma. 
 
 Case 4. — Mrs. C. R. Specimen consists of an enlarged uterus measuring 
 11x6x5 cm. Peritoneal surface is smooth. The walls are for the most 
 part normal. At one point there is a typical myoma one centimeter in 
 diameter, with a hemorrhagic center. Springing from the mucous mem- 
 brane are a number of cauliflower-like excrescences over areas vary- 
 ing from I to 2 cm. in diameter. These project .5 cm. above the surface 
 of the mucoso, some of them ulcerated, covered by little papillary pro- 
 jections and distinctly hemorrhagic. On section they appear to be part of 
 the new growth which extends down into the uterine musculature. This 
 growth is for the most part distinctly hemorrhagic, though in one or two 
 places of a homogenous white color. In the uterine musculature of the 
 fundus is a small cavity, about a centimeter in diameter, into which pro- 
 jects a number of pinhead-sized, congested, papillary excrescences. 
 
 Anatomic Diagnosis. — Adenocarcinoma of the uterus (?). Small inter- 
 stitial myoma. 
 
 Microscopic Description. — Section presents a small piece of typical fibro- 
 myoma, at one side of which is a large hemorrhagic mass. The line of 
 separation from blood clot is not distinct. The adjacent portion of tumor 
 shows some lymphocytic infiltration. This hemorrhagic mass shows 
 throughout some necrotic tissue with irregular strands and masses of 
 syncytial tissue composed of cells with a large amount of protoplasm and 
 
 13 
 
 Digitized by Microsoft® 
 
large deeply staining nuclei. The outlines of these cells are not always 
 distinct owing to- the fusing of their protoplasm. The syncytial cells are 
 found everywhere in the midst of hemorrhage. The necrotic tissue is 
 moderately infiltrated with leucocytes. The nuclei of the large cells are 
 often irregularly stained, sometimes are so arranged as to give the appear- 
 ance of giant cells. These masses are in direct continuity with the myoma; 
 in several places a distinct invasion is evident. 
 
 Microscopic Diagnosis. — Fibromyoma of the uterus infiltrated with 
 chorio-epitheloma. 
 
 Relation to Cancer. — Adenocarcinoma of the fundus was 
 associated with fibroid tumor in six cases; in two of 
 which the fibroid was infiltrated by the carcinomatous tissue. 
 
 In five of these cases the uterus was the host of interstitial 
 fibroids of small or medium size. When the relative 
 frequency of adenocarcinoma and squamous cell carcinoma 
 is considered it will be seen that we have here a much higher 
 percentage of adenocarcinoma than is found in the usual run 
 of cancer cases complicated and uncomplicated by fibroids. 
 This agrees with Martin's estimation. 
 
 Squamous cell carcinoma of the cervix occurred in two 
 cases. In both cases the fibromata were interstitial and very 
 small. 
 
 In view of the relative frequency of adenocarcinoma and 
 fibromata of the uterus, it would seem that the association 
 of the two are worthy of consideration. If the percentage 
 of fibroids at autopsies, 14.8 per cent. (v. infra), be taken as 
 an approximate ratio of their occurrence, it seems at least 
 that while the association of squamous carcinoma with 
 fibroids is accidental, the occurrence of adenocarcinoma may 
 be more or less influenced by the presence of fibroid tumors. 
 In this connection statistics compiled from the records of 
 institutions having large series of cases and autopsy reports 
 would be of value. 
 
 Complica£ons. — Ovarian cysts occurred in thirteen 
 cases and varied in size from six to twelve cm. in diameter; 
 
 14 
 
 Digitized by Microsoft® 
 
but one was of the papillomatous type. Seven of these 
 cases were additionally complicated by salpingitis. 
 
 The two parovarian cysts were of moderate size, about six 
 to eight centimetres in diameter. One tubo-ovarian cyst 
 existed, and was of moderate size. 
 
 The ovarian fibromata measured 8x7x4.5 and 20x11x10 
 centimetres, respectively. The latter was associated with 
 a very small interstitial fibroma of the uterus. Of the two 
 cases of carcinoma of the ovary, one was bilateral, and in both 
 instances the growths were due to metastases from a tumor 
 of the breast. 
 
 Cystic ovaries (one or both) in fifty-five cases (twenty per 
 cent.) may seem to be a large percentage, but compares very 
 closely with Scharlieb's* estimation of this complication. 
 These are due, it is thought, to the undue thickness or 
 toughness of the ovarian boundaries. There is in almost all 
 these ovaries an increase in the amount of fibrous tissue. 
 
 Tubed Complications,. — The tubes were diseased in sixty 
 cases; of these thirty-eight were salpingitis alone. In several 
 cases the gross change was slight and the diagnosis made 
 microscopically. In twelve cases there was pus in the 
 appendages in the form of pyosalpinx or tubo-ovarian abscess. 
 There were also eleven cases of hydrosalpinx, of which five 
 were additionally complicated by salpingitis, tubo-ovarian 
 abscess or ectopic pregnancy. One case combines all three — 
 ectopic pregnancy, salpingitis and hydrosalpinx. 
 
 It will thus be seen that the ovaries, tubes and adjacent 
 neighborhood were diseased in a large number of cases. 
 Alban Doran states that his experience is more like that of 
 Cullingworth, who found no disease in the appendages save 
 five cases of hydrosalpinx in 100 fibroids. Lawson Tate and 
 _Twambly,^° on the other hand, imply that this condition exists 
 in fifty per cent, of all cases of fibroids. Daniels found 
 disease of the appendages in fifty-eight per cent, of his cases. 
 In this series (including fifty-five cases of cystic ovaries) the 
 
 IS 
 
 Digitized by Microsoft® 
 
appendages showed disease in 137 cases, or forty-nine per 
 cent. In addition to these cases there were many cases in 
 which the remains of adhesions were noted, but as it was 
 obviously impossible to estimate this with exactness, without 
 description of the appearance in situ at the time of operation 
 this complication has not been noted. 
 
 Myomata in Routine Autopsy Work. — On examination the 
 records (Bender Laboratory) of autopsies on women show 
 that fibromyoma of the uterus was found in twenty-six. In 
 two cases apparently death was indirectly caused by the 
 tumor. To what extent the inter-current disease, which was 
 the cause of death in other cases, was hastened in its course 
 by the associated fibroids, can not be deducted. The tumors 
 were large in three (greater than eight centimetres diameter); 
 small in twenty cases (less than four centimetres diameter), of 
 these last ten were single and ten multiple. The age at which 
 these patients died in decennial periods were as follows: Under 
 thirty-five, three; between thirty-five and forty-five, six; 
 between forty-five and fifty-five, six; between fifty-five and 
 sixty-five, six; between sixty-five and seventy-five, four; 
 above seventy-five, one. The youngest was twenty-eight; 
 the eldest seventy-eight. The influence of the tumor was 
 apparent in one case in which death was due to anemia and 
 nephritis; a large submucous tumor by pressure had caused 
 hydroureter and uterine hemorrhages. In a second case a 
 large fibroid tumor of ninety pounds by pressure had caused 
 hydroureter and venous congestion with edema of the lungs. 
 In this case the weight of the woman, less the tumor, was 
 sixty-five pounds. The other cases, which were probably 
 influenced by the fibroid tumor, are deaths from pyonephrosis, 
 pyelitis, rupture of bladder and acute peritonitis. Bishop^^ 
 has collected thirty-seven fatal cases of fibromyomata. 
 
 Of the twenty small tumors, one was complicated by tubo- 
 ovarian cyst and hydrosalpinx. Four were associated with 
 adhesions, and one with carcinoma of the ovary secondary to 
 mammary cancer. 
 
 16 
 
 Digitized by Microsoft® 
 
Of the medium sized tumors, of which there were three 
 one was associated with bilateral pyosalpinx which had caused 
 a general peritonitis. The tumors in this case were three in 
 number and interstitial. 
 
 The total number of autopsies done on women over twenty- 
 years of age, during the seven years in which these autopsy 
 reports were collected, amounted to 175. So that of all adult 
 women coming to autopsy, 14.8 per cent , or one in every 
 seven, had uterine fibroids. From this it may be seen that 
 small fibroids of the uterus may cause little or no disturbance, 
 and that many of these tumors cause no symptoms and are 
 not discoverable during life. The danger of these growths, 
 apart from submucous tumors, is due to their complications, 
 and while these may not always endanger life, it is evident 
 that they frequently cause sufficient disturbance to produce 
 alterations in the surrounding tissues, as is shown by the 
 presence of adhesions, etc. 
 
 RELATION OF PATHOLOGIC CHANGES TO TREATMENT. 
 
 In conclusion, what deductions can be drawn from this 
 series? So much has been written by men of wider experience 
 that it would be a work of supererogation to add to their 
 well-debated advice. But there are a few questions to be 
 considered: 
 
 First, in regard to the advice we must give to a patient 
 suffering from fibroid. It is well understood that the opera- 
 tion for fibroids, in the hands of capable operators, is in 
 uncomplicated cases practically devoid of danger. It is to 
 complications and degenerations that mortality in fibroid 
 tumors is due, as we know from experience and from pub- 
 lished reports. Delay in an uncomplicated fibroid exposes 
 the patient to the possibility of the development of compli- 
 cations. And with the large percentage of complications in 
 this series it is evident that with delay the liability to com- 
 plication is increased. While this is so, it is evident that 
 there is one class of fibrtwds which way cause few or no 
 
 Digitized by Microsoft® 
 
symptoms, viz., the small subserous and interstitial tumors. 
 We may, therefore, say that delay is justifiable only in these 
 latter cases and not justifiable in even these when complica- 
 tions exist. 
 
 In reference to the extent of operation, while it is obvious 
 that myomectomy is the ideal operation, there is no doubt 
 that it is only applicable to a small class of cases — those in 
 which the tumors are easily accessible and in which it is 
 possible to eradicate them completely. If eradication is not 
 complete, there is the possibility of renewed growth, and 
 should pregnancy occur, with its increased blood supply, this 
 possibility is the greater. This should be kept in mind, as 
 those cases in which myomectomy is desirable are usually 
 childless women. 
 
 It is at present accepted that the operation of supravaginal 
 hysteromyomectomy''^ is accompanied by the lowest mor- 
 tality^^. But this operation is obviously not applicable to 
 those cases with malignant degeneration and associated with 
 malignant disease of the uterus. Such cases demand com- 
 plete extirpation of the cervix. Should, however, the com- 
 monly accepted operation of supravaginal hysteromyomectomy 
 be attempted, and on opening the uterus adenocarcinoma or 
 suspicious necrotic change be seen, the cervix may be 
 removed and a complete ablation done.^* 
 
 Adenocarcinoma and sarcoma are the forms of malignancy 
 which, in connection with fibroids, are most likely to come to 
 operation unrecognized. If, however, the case be uncompli- 
 cated by malignancy, the choice must lie between amputation 
 through the cervix and myomectomy. It is rarely advisable 
 to do myomectomy when the appendages are adherent or in 
 the presence of old or recent infection. The site of enuclea- 
 tion may become the starting point of a peritonitis.^' In view 
 of the large percentage (21.4) of diseased tubes, it is evident 
 that myomectomy is limited to a small class of cases. 
 
 Should amputation through the cervix be the operation of 
 
 18 
 
 Digitized by Microsoft® 
 
choice, one or both ovaries may be left to ameliorate the 
 symptoms of the surgical menopause. In such cases not only 
 must the question of infection from the tubes be considered, 
 but also the condition of the ovary. In a large percentage 
 of this series cystic ovaries were found. A cystic ovary 
 should not be left; if necessary, resection may be done and 
 the cysts removed, 
 
 CONCLUSIONS 
 
 First. Fibroids which are producing symptoms, and all 
 fibroids except those designated as small (especially when 
 subserous or intramural), should be removed, because the risk 
 to the patient from operation is far less than that from the 
 tumors themselves. 
 
 Second. In view of the sarcomatous changes, carcinoma- 
 tous associations and complications of uterine fibromyomata, 
 early removal is indicated when they are of sufficient size to 
 produce symptoms. Small uncomplicated tumors not pro- 
 ducing symptoms do not require early treatment. 
 
 Third. Thorough pathologic examination should be made 
 of all fibroids for evidence of malignancy. Particular study 
 should be devoted to those tumors which are necrotic, cystic 
 or both, as among these are found the largest proportion of 
 malignant changes. 
 
 References 
 
 1 . Mabtik. a. : CentrcUUattfur Chjnakologie. 1888, p. 389. 
 
 2. NoBLB, Chakles p.: The Complications and Degenerations of Fibroid Tmnors of the 
 TJterna as Bearing on the Treatment of These Growths, BHtish dytuc. Jour., 1901, vol. xvii, p. 
 170 Also: A Study of the Degenerations and Complications of Fihroid Tumors of the Uterus, 
 etc., American Qynecology, 1903, April. 
 
 3. Ottllingworth, C. J. : An Analysis of One Hondred Cases of Uterine Fibromyoma, 
 etc., Jour. Obst. and Gynec. British Empire, 1902. 
 
 4. SoHARLiEB, Mart: An Analysis of One Hundred Cases of Fibromyomata Uteri, Jour, 
 Obet. and Oynec. British Empire, 1902, October, p. 330. 
 
 5. Prbbbbick, C. C. : A Study of Uterine Fibromyomata with ^special Reference to Their 
 Degeneration and Complications, etc., American Gynecology, 1902, vol. i, p. 255. 
 
 6. Noble, Charles P.: Two Cases of Deciduoma Malignnm, American Jour. Obst., 1892, 
 vol. xlvi. No. 3. 
 
 7. Daniels: Rev. Oyn. Chlr.Abd., 1903, No. 112. 
 
 8. Marsoheb. Ueber eine nnnnterbrochene Beihe von 125 glflcklich verlaufene Laparoto- 
 mien. CmtraXblatt far Oynak., 1903, No. 19, p. 576. 
 
 9. Bond, C. J. : Lancet, Jan. IT, 1903. 
 
 19 
 
 Digitized by Microsoft® 
 
10. HnNNBR, Gut: One Hundred Caees of Fibroid Tumors, paper read before the Southern 
 Surgical and Gynecological Society. 
 
 11. Becasens, S. : Dos caeos de miomas nterinos complicados de apendicitis. Bee. de. med. 
 V drug, pract., Madrid, 1903, Iviil, p. lal. 
 
 12. Selage : Coesistence d'un flbrome de I'uteras avec cancer du col et une double Balpin- 
 gite suppur^e, BvU. et Man. Soc. Anat. de Paris, 1901, 63, iii, p. 341. 
 
 13. Bakeb AifD Gkavbs: Amer. Jour. Olisi., 1903, September. 
 
 14. Ulesko-Stroganowa : TJeber bSsartiges Myom der Geharmntter. Mmatsachrift fSr 
 GetnirtsMUfe und Gynalcologie, 1902, Sept., p. 427, Oct., p. 826. 
 
 15. Tate: A Case of Sarcoma of the Uterus Associated with Fibromyomata, JoKf. Oist. 
 and Bynec. British Empire. 1902, ii, 258. 
 
 16. Van Hoosbn, Bebtha : Sarcomatous Degeneration ill Uterine Myoma, American 
 Jour. Obst., 1903, bdl, pp. 233-236. 
 
 17. DoBAN, AxBAN H. G. : Fibroid of the Broad Ligament Associated with Ovarian Cyst, 
 Trans. Obst. Soc, London, vol. xiii, p. 260. 
 
 18. Stbeeteb: .4m«r. JiiMr. Odsi., vol. xxi. p. 211. Also Shields Marmaduke: Trans, of 
 Boyal Chir. Soc, vol. Ixxx, p. 205. 
 
 19. DoRAK, Alban, H. G. : Fibroids of the Uterus and Its Ligaments, Lecture iii, The 
 Lancet, 1903, vol. i. No. 8, p. 497. 
 
 20. Bishop, Stakhobe : Uterine Fibromyomata, p. 49. 
 
 21. Bishop, Stanmobe : Uterine Fibromyomata, 1901, pp. 27-31. 
 
 22. WBIJ.S : Surgical Treatment of Fibroids, Amer. Jonr. Obst., vol. xlvli, p. 306. 
 
 23. Hehbicius : ArchivfUr Gynakologie, vol. Ixvii, No. 2. 
 
 24. Bland-Sdtton : Trans. Obet. Soc, 1901, vol. xliii. p. 200. 
 
 25. CuLLEH, I. E.; Sarcomatous Transformation of Myoma, Discussion, The Journal 
 A. M. A., 1903, Oct. 24, p. 1015. 
 
 20 
 
 Digitized by Microsoft® 
 
Reprinted from the Albany Medical Annals, July, 1904 
 
 A PECULIAR HYPERTROPHY OF THE PROSTATE, 
 ACCOMPANIED BY AN ASCENDING INFEC- 
 TION AND CYSTS IN THE URETERS, 
 WITH A GENERAL DISCUSSION 
 OF URETERITIS CYSTICA. 
 
 By harry W. CAREY, M. D., 
 
 Patliolo0Bt to the Samaritan Hospital, Troy, N. T. ; ABsistant in Surgical Pathology, Albany 
 
 Medioal College, 
 
 AND 
 
 ARTHUR T. LAIRD, M. D., 
 
 Lecturer in Clinical Microscopy, Albany Medical College. 
 
 (Prom The Bender Laboratory, Albany, N. Y.) 
 
 While hypertrophy of the prostate is of common occur- 
 rence, the peculiar form to be described is unusual and 
 explains some of the difficulties of catheterization and 
 cystoscopic examination met with in this condition. The 
 formation of cysts in the mucous membrane of the ureter is, 
 on the other hand, much less frequent, being noted but twice 
 in the records of over 600 post-mortem examinations at the 
 Bender Laboratory. Previous to 1894 Lubarsch^ cites but 
 ten cases in the literature and found the condition himself in 
 four out of 3,000 sections. 
 
 Following is a brief summary of the two cases we wish to 
 report : 
 
 Case I. — (Bender Laboratory Records, autopsy No. 0-604). Clinical 
 History: E. S., male, age 80, admitted to the Albany County Hospital, 
 February 18, 1903. Patient complains of diarrhoea and difficulty in voiding 
 his urine. Family history — negative. Past history — Has used alcohol 
 
 Digitized by Microsoft® 
 
freely. Denies any venereal infection. Present condition— The patient's 
 mental condition is defective and he gives untrustworthy replies to ques- 
 tions relating to his illness. He has used a metal catheter for some 
 time, passing it himself without antiseptic precautions. 
 
 Physical Exammation^-The. heart, lungs and abdomen negative. The 
 rectum is prolapsed. The left testicle is apparently swollen, tense and 
 transmits light and is about 8 centimetres in diameter. The prostate per 
 rectum is found to be enlarged. There is a purulent discharge from the 
 urethra. When the urine is withdrawn by a rubber catheter the first por- 
 tion is pure pus and is followed by ammoniacal cloudy urine. Death ten 
 days after admission, February 28, 1904. 
 
 Autopsy seventy-two hours after death by Dr. A. T. Laird. 
 
 Anatomical diagnosis. — ^Hypertrophy (fibro-myoma) of the prostate. 
 Diphtheritic cystitis. Ascending infection (pyelo-nephritis) of both kid- 
 neys. Bilateral ureteritis cystica. Acute vesiculitis. Chronic interstitial 
 nephritis. Chronic interstitial orchitis. Tuberculosis of epididymis and 
 prostate. 
 
 Prostate — Shows hypertrophy, particularly in the lateral lobes, which 
 cause obstruction to the outflow of urine, not, however, by pressure up- 
 wards as is seen with hypertrophy of the middle lobe, but rather by pres- 
 sure downward. The chief hypertrophy is in the right lateral lobe, which 
 extends across the upper wall of the urethra. It is peculiar in being 
 attached to the lower portion of the gland by a thin pedicle which allows 
 considerable mobility of the tumor. Extending outward in a lateral direc- 
 tion from the prostatic portion of the urethra is a pocket which measures 
 3x2 centimetres and which communicates with the bladder by a narrow 
 orifice. Into this pocket the pedunculated lateral lobe fits, completely clos- 
 ing the communication between the bladder and the pocket, thus pre- 
 venting the passage of urine and acting as a "drop latch." When a catheter 
 is introduced it passes into the pocket and drains the pus which has accumu- 
 lated there, at the same time it lifts the pedunculated lateral lobe allowing 
 the urine to escape. The mucosa lining the pocket has a greenish grey 
 color and is studded with fine miliary abscesses. The left lobe of the 
 prostate is firm in consistency; the right lobe is softer than normal and 
 nodular and on section the stroma predominates; small yellowish areas 
 of glandular tissue are present. In the left lobe no glandular tissue can 
 be found. 
 
 The Bladder is externally roughened and congested and attached to 
 the adjacent coils of the small intestines by fresh adhesions. There are 
 also extensive adhesions between the rectum and bladder. The bladder is 
 dilated, its walls average 0.5 centimetre in thickness and it is infiltrated 
 with pus throughout all its coats. The mucosa is covered with a sero- 
 purulent exudate and is of greenish color with here and there streaks 
 of red. 
 
 The Right Kidney — Measures 9x5x3 centimetres. The capsule strips 
 readily leaving a smooth surface. Beneath this are found a number of 
 minute yellowish areas about 0.5 centimetre in diameter, which when 
 
 Digitized by Microsoft® 
 
sectioned appear to be made up of purulent material. The substance of the 
 kidney appears normal, although the markings are somewhat indistinct. 
 The pelvis of the kidney is dilated and the mucosa is congested. 
 
 Right Ureter — The mucosa is congested and about fifty small cysts 
 project from it into the lumen. They have a glistening yellowish color 
 and are filled with serous fluid which is under considerable pressure, spurt- 
 ing out when the cysts are punctured. In some cysts the contents are 
 slightly tinged with blood. 
 
 The Left Kidney — Measures 7x4x3 centimetres. It has been converted 
 into an abscess cavity and but very little of the normal kidney substance 
 remains. The abscess cavity is lined by a purulent membrane. The mucosa 
 of the pelvis is swollen and is covered with a purulent exudate. 
 
 Left Ureter — Is much dilated, measuring 1.5 centimetres in diameter. 
 About a half centimeter from the ureteral orifice in the bladder several 
 cysts are found, smaller, however, than in the ureter of the opposite side. 
 
 Seminal vesicles — Contain pus and yellow mucoid material. 
 Right testicle — Shows extensive atrophy. There is also a hydrocele 
 measuring 10x4 centimetres. 
 
 Left testicle — The globus major of the epididymis is caseous and the 
 interstitial tissue of the testicle is increased. 
 
 Urethra — Stricture of large calibre in the bulbous portion. 
 
 Histological Description — The cysts of the ureter are surrounded by a 
 distinct membrana propria derived from the submucosa and are lined by 
 transitional epithelium. The cyst cavities are filled with exudate often 
 hemorrhagic, and containing desquamated epithelial cells. Among these 
 are oval, pale, faintly staining bodies frequently showing nuclear shadows. 
 The epithelium of the ureter is absent for the most part over the cysts ; only 
 a few strands of connective tissue separating the cysts from the lumen. 
 Several of v. Brunn's cell nests are present in the submucosa; no lumen 
 connecting them with the ureter can be found. 
 
 Microscopic examination pf the cyst contents showed many epithelial 
 and pus cells and bacteria. Peculiar round or oval highly refractive bodies 
 containing vacuoles are also present and with Sudan III give the char^ 
 acteristic reaction for fat. Cultures made from the cysts showed the pres- 
 ence of the B. coli communis. 
 
 Case II — (Bender Laboratory Records, Autopsy No. 0-323). L. G. R., 
 female, aged 54. Autopsy, October 20, 1900, by Dr. G. Blumer. 
 
 Anatomical Diagnosis— 'Ltft renal calculus. Right ureteral calculus. 
 Pyelitis on right side with atrophy of kidiiey. Cystic ureteritis. Pyelone- 
 phrosis with multiple miliary abscesses of kidney substance. Cystitis. 
 Brown atrophy of heart muscle. Emphysema of lungs with hypostatic 
 congestion and oedema. Cloudy swelling of the liver. Acute splenic tumor. 
 
 Left Kidney is greatly dilated and on section the pelvis is found to be 
 enormously distended with thick greenish pus and contains a single yel- 
 
 Digitized by Microsoft® 
 
lowish white stone, granulaT in appearance (phosphatic). From Jt at one 
 point a brownish stone projects, probably uratic (?) in origin. The re- 
 moval of the capsule tears away the kidney substance which moreover is 
 greatly diminished in amount. The surface of the kidney is congested 
 and dotted with miliary abscesses. 
 
 Left ureter is markedly dilated and the mucosa is covered with little 
 polypoid excrescences, perfectly round and translucent and average I centi- 
 meter in diameter. 
 
 Right kidney is greatly contracted and on removal of the capsule the 
 surface is finely granular. On section the cortex is considerably diminished 
 in amount. The pelvis of the kidney is dilated and contains thick pus. 
 
 Right ureter is blocked about 2 to 3 centimetres above its vesical orifice 
 by several small stones. 
 
 Bladder contains thick muco-pus greenish in color. It is markedly con- 
 tracted and the mucosa is congested. 
 
 Bacterioscopic — In the kidney abscesses was found a bacillus of the 
 mucosus capsulatus group. 
 
 Histological Examination of cysts in the ureters. The cysts are located 
 in the submucosa but project prominently into the lumen of the ureter, 
 fully five-sixths of the cysts are above the level of the mucosa. The 
 epithelium of the ureter has desquamated over the cysts leaving only a 
 few strands of connective tissue. This connective tissue, derived from the 
 submucosa, surrounds the cysts and also separates them from the lumen 
 of the ureter. The lining of the cysts is composed of epithelium similar 
 to that lining the ureter, that is transitional epithelium. At the base of 
 the cyst are several layers of epithelial cells, but at the periphery 
 where the pressure is greatest there are fewer layers and the cells are 
 quite flattened. The contents of the cysts are in some cases colloid 
 material similar to that found in the thyroid; in others, cellular debris. 
 In the contents oval and refractive bodies, three to ten times the size of 
 a leucocyte, are constantly present and in some of these faint bodies can 
 be seen (nuclei). These are probably identical with the so-called sporozoa 
 that have been described in association with cysts of the ureter. 
 
 Several of V. Brunn's cell nests are present in the submucosa. In one 
 instance a connection between a nest and the lumen of the ureter exists. 
 
 Formation of Ureteral Cysts 
 
 Cysts of the ureter, pelvis of the kidney and bladder were 
 first described by Rayer'' and later by Rokitansky* and 
 Morgagni*. Virchow^ thought their formation in the ureter 
 was due to a blocking of the crypts in the mucosa of the 
 ureter followed by a retention of secretion ; and in the blad- 
 der, on account of the presence of concentric bodies in the 
 
 Digitized by Microsoft® 
 
To Illustrate Dr. Carey's and Dr. Laird's Article on "A Peculiar Hyper- 
 trophy of the Prostate." 
 
 Albany Medical Annals^ July^ 1904. 
 
 Fig. I —Shows the cysts in the mucosa of the nreter (actual size). 
 
 
 t^^' 
 
 
 ':i- 
 
 avYC.y 
 
 Fig. II.— Shows the strnctnre of one of the cysts; the membrana propria extending 
 
 entirely around the cyst wall. (A) are the bodies described as 
 
 degenerated epithelial cells (so-called sporozoa); (B) a 
 
 cell nest (v. Brunn) ; (objective, 1-6; ocular, I.). 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
cyst contents, attributed them to misplaced prostatic glands. 
 Hamburger* described glands in the mucosa of the ureter 
 which simulated the appearance of sebaceous glands and he 
 thought the cysts originated from these. In a histological 
 study of the ureter v. Brunn found nests of epithelial cells 
 surrounded by a membrana propria and called attention for 
 the first time to the peculiar arrangement of the connective 
 tissue beneath the epithelium of the ureter. This arrange- 
 ment is that of a meshwork with thickening at the nodal 
 points forming septa which extend upward into the epithelial 
 layers. The height of the septa varies inversely with the 
 intra-ureteral pressure, being higher when the pressure is low. 
 To variations in pressure v. Brunn'' attributed the formation 
 of these cell nests, small portions of the epithelium being 
 enclosed and surrounded by the septa of connective tissue. 
 
 To determine whether cell nests were normally present 
 Aschoff* studied the pelvis of the kidney, the ureters and the 
 bladder in the new born and adult, but in no instance were 
 they present, hence they are not constant elements. He 
 discredits the influence of pressure in the formation of the 
 cell nests. 
 
 Later investigations of Lubarsch®, Marckwald'", Stoerk", 
 ZuckerhandP and Paschkis'^ support the theory that the 
 cysts originate from the cell nests of v. Brunn. 
 
 The method of formation of the cysts is described by 
 Aschoff as follows: 
 
 (a) Enclosure of a portion of epithelium by septa of con- 
 nective tissue with the formation of a cell nest. This causes 
 small macroscopic nodules or granules in the mucosa of the 
 ureter. 
 
 (b) Central degeneration of these cells accompanied by a 
 transudation of fluid into cysts cavity. 
 
 The Location of Cysts is fairly constant and corresponds to 
 the points of constriction in the ureter. They are almost 
 constantly found in the pelvis of the kidney, upper third of 
 the ureter and at the uretero-vesical orifice on the trigonum. 
 They are frequently bilateral. In the two cases above 
 reported the distribution of the cysts was diffuse and not 
 limited to any portion of the ureter. 
 
 5 
 
 Digitized by Microsoft® 
 
Etiology. — The exact cause of the formation of these cysts 
 in the ureter is not known, but in the large majority of the 
 cases reported they developed in association with inflamma- 
 tory and suppurative conditions of the bladder, ureter and 
 pelvis of the kidney (ascending infections). 
 
 In the twenty-four cases, including our two, a cystitis, 
 calculus or pyelitis existed in all but two ; in the latter no 
 lesion of the urinary tract could be found. 
 
 Mention has been made above of the oval translucent 
 bodies found in the cysts contents. These have been 
 described as sporozoa by v. Kahlden", Pisenti^^ Silcock", 
 Eve", Clark" and Sutton", and were believed to be the cause 
 of the cyst formation. This view has not been supported by 
 later investigations since the parasites could not be cultivated 
 from the cysts and moreover it is improbable that any 
 infection of this nature should always be limited to such a 
 prescribed portion of the urinary tract. 
 
 The so-called sporozoa in our sections, while resembling 
 parasites in some respects, were by no means constantly 
 present in the cysts and never very numerous. They were 
 never found outside of the cysts, in the lumen of the ureter 
 or in the epithelium ; where they were surrounded by cellular 
 debris a gradual transition from epithelial cells to these 
 so-called parasites could be traced. In some, several faint 
 nuclei could be seen. It is our opinion that these parasitic 
 like structures are an early stage in the degeneration of the 
 epithelial cells lining the cysts and later become colloid. 
 The interest in this condition of Ureteritis Cystica does not, 
 however, relate entirely to its pathology. Aschoff has called 
 attention to the fact that in the early stages of the formation 
 of the cysts they may be mistaken for miliary tubercles in 
 cystoscopic examinations. 
 
 In one of the cases reported by Marckwald the develop- 
 ment of the cysts had been so extensive as to block the 
 ureter completely, giving rise to great hydronephrosis. 
 Healton and Gamgee^" have reported a case in which 
 haematuria had existed for many years and at autopsy a 
 cystic villous papilloma of the renal pelvis was found. The 
 bleeding, in part, was due to the rupture of the cysts. 
 
 Digitized by Microsoft® 
 
Resume 
 
 1. Hypertrophic lateral lobes of the prostate may be 
 pedunculated and by their downward pressure on the urethra 
 obstruct the outflow of urine, acting like a "drop latch." 
 
 2. The origin of cysts in the ureter is attributed to the 
 central degeneration and later transudation of fluid into 
 V. Brunn's cell nests. 
 
 3. Ascending infection of the urinary tract is the 
 important factor in the etiology of these cysts. 
 
 4. The peculiar translucent oval bodies resembling 
 sporozoa are in some instances the product of epithelial 
 degeneration. 
 
 5. The cysts may, in their early stages, simulate miliary 
 tubercles. 
 
 6. Cysts of the ureter may cause obstruction to the 
 passage of urine from the kidney to the bladder with conse- 
 quent hydronephrosis. 
 
 7. Rupture of the cysts can give rise to haematuria. 
 
 
 
 LITERATURE 
 
 1. 
 
 LUBABBCH. 
 
 Centralbl.f. AUg. Path., 1894, v. 468. 
 
 2. 
 
 Eatbr. 
 
 Traite des maladies dcB reins, 1841, iii., 660. 
 
 3. 
 
 BOKITAS9KT. 
 
 Lehrbuch der path. Anat., 1841, iU., 354. 
 
 i. 
 
 MORQAGNI. 
 
 De sedlbns et causis morborum. 
 
 5. 
 
 VlBOHOW. 
 
 Die krankhaJten QescliwfllBte, 1863, 1., 233. 
 
 6. 
 
 Hahbubgek. 
 
 Archiv fur Mkr. Anal., 1879, xtU., 14. 
 
 7. 
 
 VOH Brumk. 
 
 lUd, 1898, xli., 294 
 
 8. 
 
 A^OHOTF. 
 
 nrehow. Archiv., 1894, xxxviii, 119. 
 
 9. 
 
 LltBAIlBCH. 
 
 Archiv far mkr. Anat., 1893, xli. 303. 
 
 10. 
 
 Marckwals. 
 
 Munch. Med. TTocA^resAr., Aug., 1898. 
 
 U. 
 
 Stoerk. 
 
 Ziegler's BMr'age, 1899, xxxvi., 397. 
 
 12. 
 
 ZUCKBRHANDL. 
 
 Monatsbtricht Vrologie, 1903, vii. 
 
 13. 
 
 Pasohkib. 
 
 Monatsiericht Urologie, 1903, viii. 
 
 14. 
 
 T. Kahldbn. 
 
 Ziegler's Beitrdge, 1893, xlv. 
 
 IB. 
 
 PlSEUTI. 
 
 Cmtr. f. AUg. Path., 1898, v. 
 
 16. 
 
 SiLCOCK. 
 
 Tram. London Path. 3oc., 1899, xl. 
 
 17. 
 
 BVB. 
 
 ZVoM. London Path. Soc., xl., p. 441. 
 
 18. 
 
 Cl.AItK. 
 
 Ibid, xliii. 
 
 19. 
 
 SUTTOH. 
 
 Lancet, 1888. 
 
 20. 
 
 Healton asd Qamgee. 
 
 Ibid, 1902, p. 746. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
Reprinted from the Albany Medical Annals, August, 1904 
 
 INFANTILE INGUINAL HERNIA OF THE VERMI- 
 FORM APPENDIX WITH ADHESION 
 TO THE EPIDIDYMIS.* 
 
 By E. MacD. STANTON, M. D., 
 AsBistant, Bender Laboratory, Albany, N. T., 
 
 Ain> 
 
 HENRY L. K. SHAW, M. D., 
 
 Lecturer in Pediatrics, Albany Medical College. 
 
 Hernia of the appendix is not rare, but it is uncommon. Mor- 
 gagni, in 1750, was, perhaps, the first to make reference to it. 
 Scarpa reported several cases in 1806, after which we have series 
 of cases reported by Klein, Brieger, Bajardi, Rivet, Gibbon, and 
 others, until, in 1899, Peterson was able to collect 116 published 
 cases, while Piersol, in 1901, estimated the number of reported 
 cases as 125. That it is more frequent than the number of reported 
 cases would lead one to believe appears from the fact that Coley 
 found hernia of the appendix, without the presence of other por- 
 tions of the intestinal tract, five times in 351 operations for femoral 
 or inguinal hernia. 
 
 Hutchinson classifies hernia of the appendix into two divisions, 
 congenital and acquired. The congenital form is always inguinal, 
 and almost invariably occurs in the male sex and on the right side. 
 In this form it is the rule to find adhesions between the appendix 
 and some of the remains of the processus vaginalis — either to the 
 wall of the sac, the cord or to the tunica vaginalis. The origin 
 and significance of these adhesions have been the subject of much 
 speculation. They have been ascribed to the persistence of folds 
 of peritoneum normally present in foetal life; or as the result of 
 
 * Previously published in briefer form in the Archives of Pediatrics (March, 1904) to which 
 journal we are indebted for the ase of the accompanying cut. 
 
 Digitized by Microsoft® 
 
foetal peritonitis, the frequency of which is upheld by Virchow, 
 B. Schmidt, Orth and others ; to the persistence of certain fibres 
 of the gubernaculum ; and they have also been considered as of 
 post-hemial origin. There seems to be little doubt that in some 
 of the cases these adhesions, by uniting the appendix to the 
 migrating structures, have enabled the same forces which bring 
 about the descent of the testicle to also play an active part in 
 causing the descent of the appendix into the inguinal canal. 
 
 In support of the view that these attachments may originate in 
 foetal folds of the peritoneum may be cited the fact that previous 
 to the descent of the testicle into the inguinal canal the peritoneum 
 overlying the testicle encloses it and forms a mesentery, the 
 mesorchium, which also encloses the gubernaculum and forms two 
 folds, one above the testicle and the other below it. The one 
 below the testicle encloses the lower part of th gubernaculum, 
 and is known as the plica guhernatrix ; the one above the testicle 
 is not constant, but when present usually covers the spermatic 
 vessels and is known as the plica vascularis. Hutchinson states 
 that at the time the testicle passes into the inguinal canal this 
 ascending band or ridge of peritoneum can be found in about ten 
 per cent, of all cases : its upper attachment varies ; usually it is 
 to the under surface of the mesentery of the lower end of the 
 ilium, less commonly to the caecum or vermiform appendix. A 
 similar band is much more frequently found in females (fifty 
 per cent, of all cases, according to Hutchinson), in whom it per- 
 sists as the well-known appendiculo-ovarian ligament. The pres- 
 ence of this band with rigid superior attachments has been given 
 as a cause of incomplete descent of the testicle. 
 
 As is well known, the testicle develops opposite the second or 
 third lumbar vertebra, from which point it descends during the 
 third month, by the end of which time it has reached the internal 
 ring. It remains in this position until near the close of gestation 
 or, according to some authors, enters the processus vaginalis as 
 early as the middle of the seventh month. The position of the 
 appendix is high in the abnominal cavity until the descent of the 
 caecum, which occurs about the sixth month, but, owing to the 
 relatively great length of the appendix in the foetus (eleven milli- 
 metres in each of two twelve centimetres foetuses examined by 
 us), it is at all times capable of coming in contact with the peri- 
 toneum covering the testicle. It is thus seen that inflammatory 
 
 Digitized by Microsoft® 
 
Xo Illustrate the Article by Drs. Stanton and Shaw on " Infantile Inguinal 
 Hernia of the Vermiform Appendix." 
 
 Albany Medical Annals, August, iqo4. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
conditions occurring in either the appendix or the testicle could 
 result in the formation of adhesions between the appendix and 
 the migratory peritoneum covering the testicle, and thus cause 
 the descent of the appendix along with the testicle. 
 
 The following case, occurring in the service of one of us (Shaw) 
 at the Albany Orphan Asylum, is of interest in that it represents 
 a case of congenital hernia of the appendix, in which the descent 
 of the appendix into the inguinal canjd was undoubtedly caused 
 by an early inflammation which involved both testicles, together 
 with their peritoneal covering, and which resulted, we believe, 
 in the formation of an adhesion between the distal end of the 
 appendix and the tunica albuginea overlying the head of the right 
 epididymis. 
 
 Sidney P., born November lo, 1902; admitted to the Albany Orphan 
 Asylum January 31, 1903 ; died October 29, 1903. While under observa- 
 tion in the Asylum there were no symptoms referable to the appendix 
 or genital organs; the patient dying of broncho-pneumonia following 
 measles. 
 
 The following notes are from the autopsy protocol (Stanton) : — 
 
 Abdomen — Abdomen is free from fluid. Omentum contains very little 
 adipose tissue. Serosa smooth and glistening. Appendix is eight centi- 
 metres long and extends directly downward with free meso-appendix from 
 caecum through internal ring and inguinal canal and is so united by 
 adhesions to the testicle that the latter can be drawn up into the inguinal 
 canal by traction on the appendix. 
 
 When the testicle is in normal position there is a slight evagination of the 
 peritoneum around the appendix at the right of the internal ring; other- 
 wise the funicular process is completely obliterated and there is no evidence 
 of a hernia sac. 
 
 On dissecting out the appendix it is found that the portion lying within 
 the canal is covered with a loose areolar tissue while a thick band of 
 firm fibrous tissue unites the tip of the appendix to the tunica albuginea 
 overlying the head of the epididymis. 
 
 Testicles — The parietal and visceral layers of the tunica vaginalis of 
 both testes are distinctly thickened. 
 
 Anatomical diagnosis — Bilateral broncho-pneumonia. Cloudy swelling of 
 liver and kidney. Acute congestion of spleen. Congenital hernia of 
 appendix with adhesions between epididymis and appendix. Chronic 
 bilateral peri-orchitis. 
 
 Microscopic examination — Sections from the left testicle show extensive 
 thickening of the capsule and a growth of fibrous tissue between the 
 tubules. 
 
 Microscopic diagnosis — Broncho-pneumonia; endothelial hyperplasia in 
 spleen and retroperitoneal hemolymph node. Albumino us and fatty degen- 
 
 Digitized by Microsoft® 
 
eration of liver. Slight albuminous degeneration of kidney. Atrophied 
 thymus. Cloudy swelling of heart muscle. Chronic orchitis. Chronic 
 periorchitis. 
 
 Explanation of illustration — The larger drawing shows the parietal peri- 
 toneum becoming continuous with the peritoneum of the appendix at the 
 point where the appendix entered the inguinal canal and the band of 
 fibrous tissue uniting the tip of the appendix with the tunica vaginalis. 
 The smaller drawing is from a section through the tunica vaginalis show- 
 ing the thickening of the parietal portion and the relation of the adhesion 
 to the tunica vaginalis and the head of the epididymis. 
 
 REFERENCES 
 
 Beamamn. Arch. f. Anat. «. MitieicMungsgesch., 1884; S. 310. 
 
 Bribgbb. LangenbecWs Archlv, 1892, xIt, 892. 
 
 COLET. Annals of Surg., 1896, xxi. 389; 1897, xxv, 070. 
 
 Gibbon. Jour. Am. Med. Assoc., 1888, xxx, 1385. 
 
 Hdtohihbon. Brit. Med Jour., 1899, cii, 1092. 
 
 JoPBON. Vmv. Med. Mag., 1900, xlii, 94. 
 
 MiNOT. Human Embryology, 1892. 
 
 Pbtebbon. Jour. Am. Med. Assoc., 1899, xxxiii, 1407. 
 
 PiBRsoL. Univ. of Penna. Med. Bull., 1901, xiv, «78. 
 
 Digitized by Microsoft® 
 
Reprinted from Albany Mbmoai, Annals, January, 1904. 
 
 A STUDY OF THE CHANGES OCCURRING IN THE 
 ENDOMETRIUM DURING THE MENSTRUAL CYCLE. 
 
 By H. JUDSON LIPES, M. D., 
 
 Clinical Professor of Obstetrics, Albany Medical College; Head Obstetrician, Albany 
 Guild for the Care of the Sick Poor. 
 
 (.From the Bender Laboratory, Albany, N. Y.) 
 
 Probably the first recorded observations on the condition of the 
 uterine mucosa during menstruation were pubhshed by Kundrat 
 and Engelmann in 1873. Since that time numerous investigators 
 have studied the subject but the results of their work show the 
 greatest differences of opinion. 
 
 Without doubt, as we shall see, these varied views were due to 
 several causes, principally the source of the material and the 
 more accurate results obtained through better preserving fluids 
 and more satisfactory fixing methods as well as to the recent rapid 
 progress in microscopical studies. 
 
 In the study of the changes which take place in the uterine 
 mucosa during the menstrual cycle we have several methods to 
 pursue. The first method employed was that of the examination 
 of the mucosa in uteri removed at autopsy and consequently, as 
 this was the most natural source for the securing of material 
 for microscopical study, we find that the earlier and the greater 
 number of observers, Kundrat, Englemann, Sir John Williams, 
 Sinety, Leopold, Wyder, Von Kohlden, Christ, and Wendeln 
 conducted their studies on the uteri removed at autopsy from 
 patients who had died during menstruation. 
 
 That the earlier observations were faulty is not to be wondered 
 at since it is a well-known fact that autopsy material unless 
 secured very early post mortem is more or less changed. Gebhard 
 referring to the conclusion of Wendeln obtained by the study of 
 autopsy specimens, claimed that it is never allowable to show 
 cadaveric preparations as proof of the condition of the mucosa 
 where the material has been taken thirty-four hours post mortem. 
 We believe that even this is rather late to secure proper material. 
 
 Digitized by Microsoft® 
 
In the first place, as pointed out particularly by Leopold and 
 Wyder, it is absolutely necessary that an exact anamnesis be 
 secured in each case, whatever the source of the material, in order 
 to be assured that one is dealing with a true menstrual hemor- 
 rhage. That this has not been followed in many of the earlier 
 observations is certain. 
 
 Another source of error in the earlier observations, and ohe 
 that is liable to be experienced by every laboratory worker unless 
 the greatest care is taken, is that in passing the material through 
 the fixing solutions there may be more or less loss of substance. 
 Inasmuch as the first important question to be decided is whether 
 there is a loss of substance resulting from menstruation or not, it 
 is necessarily important to use such fixing solutions as will pYe- 
 serve the specimens intaot. 
 
 A number of authors have not stated their methods of pre- 
 serving and fixing their material. A number employed Miiller's 
 fluid. Leopold preserved his specimens first in weak then in 
 stronger alcohol. He was opposed to Muller's fluid because the 
 superficial layers which are saturated with blood are easily broken 
 off. This is certainly true unless the specimens are most care- 
 fully handled, and it is due to the fact that Miiller's fluid must 
 be changed and the specimens washed so often before the tissue 
 is fixed. After hardening has taken place the pieces of tissue may 
 then be moved about and the fluids changed frequently without 
 loss of substance. Mandl finds that the corrosive sublimate solu- 
 tion proves the best for the study of the condition of the epithe^ 
 lium and that Flemming's fluid was of the greatest service for the 
 study of the process of regeneration. 
 
 Finley and other recent observers used Zenker's solution arid 
 since this fixes the specimens quickly without changing the solu- 
 tion and without handling them frequently it has proved very 
 satisfactory. Could this solution have been used in earlier obser- 
 vations there would have been less disagreement as to the changes 
 which take place especially as regards the loss of substance. 
 
 Moricke was the first to examine scrapings from the normal 
 menstruating uterus and thus excluded the possibility of post^ 
 mortem changes. Lehlein, Gebhard, Westphalen and others sooti 
 followed his example and these investigators arrived at diflfererlt 
 conclusions than did the earlier oibservers who used autopsy 
 material. Wendeln had some doubt of the value of the examina- 
 tion of curettage specimens on the grounds that the whole mU6dUs 
 
 Digitized by Microsoft® 
 
membrane is not obtainable and consequently conclusions based on 
 the examination of only a part should not be made. It is pos- 
 sible, too, that curetted material may appear hemorrhagic on 
 account of the numerous blood cells which are intimately asso- 
 ciated with the material in its removal and in the process of 
 blocking. 
 
 Sinety examined the menstrual flow and the mucus adhering to 
 the superficial layer of the mucosa in order to draw conclusions 
 as to the condition of the deeper layers. In such a method the 
 possibility of arriving at erroneous conclusions is apparent. 
 Wyder aspirated the menstrual discharge from the cavity of the 
 uterus by means of a uterine syringe in order to avoid the glands 
 of the cervix and to reduce the negative pressure to a minimum. 
 
 In order to determine the probable loss of substance of the 
 mucosa during menstruation the writer undertook to examine the 
 napkins of thirty patients through the menstrual period. The 
 patients were instructed to change the napkins frequently and 
 to soak in cold water, being careful to save any clots and at the 
 end of menstruation to take a vaginal douche of plain cold sterile 
 water so that all the discharge might be collected. This douche 
 water and that in which the napkins were soaked was centrifuged 
 so as to obtain all the sediment possible. This sediment was then 
 hardened in alcohol and passed through in the routine manner. 
 Great care was taken to secure patients who had a normal men- 
 strual history and in whom there was no apparent diseased con- 
 dition of the endometrium. 
 
 The small amount of material thus obtained was cut and stained 
 with hsematoxylin and eosin. The result of such examination 
 will be mentioned later on. 
 
 Another method which is the most important of all is that of 
 the examination of sections of uteri removed during the menstrual 
 period for lesions not involving the endometrium. This method 
 of examination was only lately carried out since surgeons prefer 
 not to remove the uterus during active menstruation. 
 
 Probably the most elaborate investigations were carried on by 
 Gebhard, in Berlin, and Westphalen, in Keil, who not only studied 
 the mucosa of menstruating uteri obtained by curettment and of 
 uteri obtained at autopsy but also sections of uteri removed by 
 operation during the menstrual period. Williams, Findley, Edgar 
 and others in this country also recorded their observations on the 
 examination of the mucosa of uteri removed during operation. 
 
 Digitized by Microsoft® 
 
In women, who must lose diseased adnexia and in whom the 
 anamnese and the examination of the uterus led one to expect no 
 particularly abnormal condition of the uterine mucous membrane, 
 Mandl and other operators awaited menstruation and removed the 
 uteri during this period. No intra-uterine manipulation was 
 carried out and in order to protect the organ from unnecessary 
 handling extirpation was performed as quickly as possible by 
 clamping off the parametrium and ligaments and cutting rapidly. 
 In order to avoid the slightest possible injury to the mucosa the 
 uterus was opened on the anterior wall cutting from the cervix to 
 the fundus and then a portion was removed from the posterior 
 wall and placed immediately in fixing fluids. Findley placed the 
 uterus in salt solution immediately after removal and then placed 
 in Zenker's solution for twenty-four hours. 
 
 The views of the numerous investigators are so contradictory 
 on several points that it will be necessary to consider them more 
 in detail. 
 
 Leopold divides the menstrual cycle into four stages : ( i ) active 
 menstruation; (2) regeneration; (3) resting stage; (4) pre- 
 menstrual swelling. 
 
 In as much as the stage last mentioned is really the preparatory 
 one, according to my observations, and the beginning of the phe- 
 nomena exhibited in the change of the uterine mucosa due to men- 
 struation, we will discuss this division of our topic first. 
 
 Westphalen, Leopold, Doleris, Pompe, Van Meerderwordt, Her- 
 zog, and others are practically agreed with the conclusions of 
 Gebhard based on his observations of the scrapings and sections 
 of uteri removed during the menstrual period for lesions not 
 involving the endometrium. 
 
 From a study of the literature on the subject, and from the 
 examination of such material as I have been able to secure, the 
 following information has been gleaned. 
 
 In this stage, which begins to develop gradually as soon as the 
 process of regeneration is complete, which is about the eighteenth 
 day after the flow ceases, we find that the capillaries of the mucosa 
 are congested resulting in a serous or serosanguineous exudate 
 infiltrating the stroma of the mucosa. During the state of regen- 
 eration the cells of the stroma lay over each other rather thickly 
 but now become pressed apdrt particularly in the outer third of 
 the mucosa. 
 
 The protoplasm of these cells become compressed and the pro- 
 
 Digitized by Microsoft® 
 
jections by which. they are bound together are either greatly 
 lengthened or completely separated. In the meshes lies a light 
 opaque mass, which seems similar to blood serum and which is 
 often separated in the larger vessels from the chromocytes. 
 
 Later the blood leaves the capillaries and infiltrates the stroma, 
 gravitating in the direction of least resistance, i. e., toward the 
 uterine cavity, and, just before the active flow begins, forming a 
 collection of blood beneath the surface epithelium — ^the sub-epithe- 
 lial haematoma of Gebhard. 
 
 It is quite impossible to say whether the bloody condition of 
 the effusion along the borders of the mucosa belongs entirely to 
 the normal or not, for it is apparent that in curettage hemorrhage 
 easily occurs in the mucosa while in extirpation an engorgement 
 of the vessels occurs very easily, especially when the broad liga- 
 ment is not clamped. Even traction on the uterus from below, 
 according to Westphalen, results in venous stasis with consequent 
 small hemorrhages. 
 
 The vessels become larger in the loosened swollen stroma and 
 are for the most part more filled than in the period of regeneration 
 but the greater increase in size occurs only shortly before the 
 beginning of menstruation proper. 
 
 Many observers agree that there is also an increase in the size 
 of the glands of the mucous membrane, since these appear widened 
 and winding and frequently on cross section the walls seem to 
 be thrown into iolds in the lumen. The increase in size of the 
 glands is due, no doubt, to the collection of the secretion of the 
 gland cells. It is this mucous-like product of the gland cells 
 which gives them quite a granular appearance. The gland cells 
 become uniformly swollen and take stains more evenly, and their 
 nuclei are more widely separated as a result of the increase in 
 the volume of the protoplasm, and are uniformly more round in 
 comparison to the oval-nuclei which are seen in the regeneration 
 period. In all the stages the cell nuclei are near the base as a 
 rule, although Westphalen has often noticed that they stand in 
 the middle of the cell in the beginning of the state of pre-menstrual 
 swelling. Consequently the mucosa is increased in thickness from 
 two to three millimetres up to six and seven millimetres, — the 
 thickness varying greatly in the same uterus owing to the unequal 
 distribution of the muscle layers directly beneath the mucosa. 
 
 Kundrat and Englemann found a fatty degeneration of the 
 mucous epithelium as well as of the gland cells and interglandular 
 
 Digitized by Microsoft® 
 
tissue and the hemorrhage and oedema of the mucosa according 
 to these early investigators, were produced by a fatty degenera- 
 tion of the walls of the vessels which were now increased in size 
 and completely filled with blood — these conditions resulting in a 
 laceration of their walls and a consequent hemorrhage into the 
 tissue. They noticed that the greater part of the hemorrhage was 
 superficial and that fatty degeneration is most noticeable in the 
 same portion of the mucosa. 
 
 Sir John Williams also believed that fatty degeneration occurred 
 and consequent degenerative changes took place. 
 
 Leopold claimed, and rightly, that the herriorrhage is a super- 
 ficial capillary one due largely to the peculiar condition of the 
 vascular structure in the mucosa (numerous arteries and capillar- 
 ies and proportionally a less number of veins). What he described 
 as a. fatty degeneration of the superficial elements, was assigned, 
 not as the previously mentioned observers assigned as the cause, 
 but as the result of the hemorrhage. 
 
 Moricke and other observers maintain that fatty degeneration 
 of even the slightest grades is not to be found. 
 
 Kundrat, Sinety, Leopold, and others have described an increase 
 of the round cells and small celled infiltration but this is not con- 
 firmed by subsequent observers and their presence is denied by 
 others. 
 
 Leopold described the appearance of giant cells (with eight to 
 ten and even more nuclei) especially in the projections of the 
 mucosa into the muscular tissue. Other investigators have con- 
 firmed this observation. 
 
 At the end of the pre^1lenstrual stage we find, therefore, the 
 mucosa is swollen greatly and infiltrated with a serous or sero- 
 sanguineous exudate and that collections of blood are formed 
 beneath the surface epitheUum ready to burst out. 
 
 The stage of active hemorrhage now appears — ^being described 
 as the first stage by Leopold and as the second by Gebhard. The 
 question now arises as to how the blood reaches the uterine cavity 
 and how much mucous tissue is destroyed by this process. 
 
 Gebhard states that the blood reaches the uterine cavity by 
 being forced between the epithelial cells of the surface or of the 
 glands, or the continuity of the surface is broken by the lifting of 
 small masses of cells from their bed these providing an exit for 
 the collections of blood beneath. 
 
 Christ believed that in the weaker menstrual congestion the 
 
 Digitized by Microsoft® 
 
To Illustrate Br Lipes! Article on "A Study of the Changes Occurring in 
 the Endometrium during the Menstrual Cycle " 
 
 Albany Medical Annals, January, iqo-i. 
 
 
 (%^T' 
 
 
 •I 
 
 PCGUItE 1. 
 
 I-'HiCRE 3. 
 
 Digitized by Microsoft® ^ 
 
Digitized by Microsoft® 
 
blood escaped between the epithelial cells of the surface of the 
 mucosa and of the glands. Where the congestion was more 
 marked slight defects occurred in the surface through which the 
 blood poured out and where the menstrual flow was very pro- 
 fuse there was a more sudden and a greater loss of the surface 
 epithelium, which allowed the outflow of blood. This observer 
 claimed to have seen in- many places the passage of red cells 
 between the superficial epithelium as well as between the epithe- 
 lium of the glands but he has given us no drawings of this con- 
 dition. Others agree with this view but do not make the statement 
 as a result of their own observations. 
 
 Mandl made careful examinations of a number of sections but 
 did not observe in any case the passage of red blood cells between 
 the epithelium although these cells are easily observed when speci- 
 mens are fixed in sublimate solution and stained in haemotoxylin 
 and eosin — ^the method is adapted. He repeatedly found leuco- 
 cytes, however, which had become fixed in their wanderings 
 between the epithelium of the surface and of the glands as well, 
 and believed that by the wandering of these leucocytes the union 
 of the epithelial cells is broken so that the blood follows more 
 easily in such places. 
 
 But perhaps the most important question, and the one concern- 
 ing which there is the greatest diversity of opinion, is whether or 
 not any portion of the mucosa is destroyed by the process of men- 
 struation. 
 
 If we examine the results of the studies of the various authors 
 already mentioned we will see that the earlier observers particu- 
 larly held the opinion that not only is the whole superficial epithe- 
 lial layer destroyed but also a considerable part of the mucosa 
 disappears. 
 
 Another group of observers believe that the whole mucosa and 
 even the superficial epithelial layer remains intact ; while a third 
 group takes the middle ground and describes a loss sometimes of 
 a less, sometimes of a greater part of the epithelium and mucous 
 membrane. 
 
 In the first group we find Engelmann, Kundrat, Sir John Will- 
 iams and Von Kahlden. The latter, who undertook a through 
 study of the published communications on this subject up to the 
 appearance of his work in 1889, came to the conclusion that in all 
 probability in menstruation the whole surface epithelium with a 
 very considerable part of the mucosa is thrown off. But these 
 
 Digitized by Microsoft® 
 
observations were made on cadaveric material which had been 
 obtained when post-mortem changes had probably taken place and 
 the fixing solutions used necessitated frequent changes and con- 
 sequently much handling. As already pointed out these condi- 
 tions would result in a loss of substance in such frail material as 
 the uterine mucous membrane, especially during menstruation. 
 
 Of the group of men who believe that at no time during men- 
 struation is any part of the mucosa lost, De Sinety is the only 
 early worker on autopsy material. Moricke, Strassmann, Find- 
 ley, J. Whitbridge Williams, C. Ruge, Veit, Gebhard, Mandl, 
 Leopold, J. Clarence Webster, Meerdervoort and others believe 
 that only a small portion of the superficial epithelium is lost and 
 that this is purely accidental. 
 
 Underbill, Wendeler, Heape, and others, believe that only the 
 superficial portion of the mucosa is destroyed. 
 
 There is no doubt but that the character of the hemorrhage has 
 much to do with the amount of destruction of the superficial epithe- 
 lium. If the congestion is of rapid development and the collection 
 of blood beneath the surface epithelium is extensive there must 
 necessarily follow a destruction of at least a portion of the over- 
 lying surface epithelium. If the hemorrhage is slight and takes 
 place largely by diapedesis the loss in the epithelium would be 
 almost nil. 
 
 This I have proved, at least to my own satisfaction, by the 
 examination of the menstrual discharges as already mentioned. 
 Lack of space will not allow a detailed description of these cases. 
 
 In the thirty cases examined the period of menstruation ex- 
 tended from one to eight days. There were two cases of one 
 day's duration, one of two days, twelve of three days, seven of 
 four days, three of five days, two of six days, two of seven days, 
 and one of eight days' duration, all of which were carefully 
 examined. 
 
 In the first three cases mentioned there was practically no sedi- 
 ment and this was composed principally, of red-blood cells and 
 some vaginal epithelium. In two of the twelve cases where the 
 flow lasted three days the onset was severe and the flow con- 
 siderable, but with little or no pain experienced by the patient. In 
 these two cases there was considerable sediment which contained 
 a large amount of blood and numerous epithlial cells from the 
 mucosa, as well as vaginal epithelium and other debris. 
 
 In one of the seven cases menstruating four days, the flow was 
 
 8 
 
 Digitized by Microsoft® 
 
slight the first day but increased in severity the last three days. 
 In this case, however, there were not many epithelial cells found 
 in the discharge. In the two cases of six days, the two of seven 
 days, and the one of eight days' duration the flow was consider- 
 able in amount as it extended over so many days but it occurred 
 gradually. Proportionately to the amount of sediment there was 
 a less number of epithelial cells than in the two specimens obtained 
 from the patients whose periods lasted three days. 
 
 In as much as only practica:lly normal cases had been selected 
 the results show that there is some loss of epithelium during 
 normal menstruation and that the amount of loss is dependent 
 upon the severity and not necessarily upon the length of the 
 menstrual period. There were certainly not enough epithelia in 
 the discharge in any of the cases to suggest a complete loss of the 
 superficial layer of the mucosa. 
 
 With the disappearance of the mentrual flow involution begins. 
 We find that the blood vessels decrease rapidly in size as the 
 engorgemnt lessens and that blood is no longer extravasated 
 into the connective tissue spaces. What blood is left in the 
 stroma is rapidly absorbed. 
 
 Granting now that there is a partial loss of the surface epithe- 
 lium the question arises as to how this loss is repaired. 
 
 According to Leopold the surface epithelium which has been 
 lifted from its bed, if the continuity has not been broken on all 
 sides, resumes its former place and the lost epithelium is regen- 
 erated from adjacent cells. This is the consensus of opinion. 
 
 Sir John Williams, who claimed that the whole mucosa was 
 destroyed, believed that the mucous membrane was regenerated 
 from the epithelium of the glands which penetrated the muscle 
 layer. Wyder, who taught the partial destruction of the mucosa, 
 believed that there was a. cell hyperplasia of the interglandular 
 tissue in the deeper layer of the mucosa, which provides for the 
 restoration of the loss of cells and that ^ a regeneration of the 
 superficial epithelium takes place from the gland epithelium and 
 from the remaining " cell-islands." 
 
 Wendeler believed in a partial loss of the epithelium and that 
 the necessary regeneration was accomplished as previously from 
 a cell membrane situated under the base of the old epithelium con- 
 taining flat, tender epithelium. No other investigator has observed 
 such a membrane, however. 
 
 Mandl was the first, I believe, to deimonstrate the various stages 
 
 Digitized by Microsoft® 
 
of well marked mitosis in the epithelium of the covering layer and 
 showed that the lost cells were replaced by indirect cell division 
 of the remaining cells. He believed that the gland cells also 
 assisted in this process but did not agree with Von Kahlden who 
 claimed that the formation of the new cells resulted entirely from 
 those of the gland whose openings became widely extended and 
 whose cells spread out in all directions. 
 
 The length of time necessary for the complete regeneration of 
 the mucosa varies necessarily according to the amount of tissue 
 lost. Williams believed that the entire mucous membrane was 
 restored in a few days. Leopold states that the regeneration of 
 the lost epithelium is performed likewise in a few days after the 
 close of the menstrual flow. Westphalen, however, thinks that 
 this process is a longer one consuming from six to eighteen days 
 and believes that this stage is prolonged through the third or 
 resting period as described by Leopold. In any event the interval 
 between the complete restoration of the mucosa and the pre- 
 menstrual congestion state is very short. 
 
 Whether the mucous layer of the Fallopian tubes undergoes a 
 change during menstruation is a question. Personally I have not 
 been able to examine the mucosa of the tubes removed during 
 menstruation. 
 
 Mandl in one case, where he had removed the uterus and 
 adnexia on the fourth day of menstruation, found that the mucous 
 membrane of the tubes was swollen and hemorrhagic although the 
 adnexia were normal. 
 
 Findley is the only observer who has exhibited a drawing of 
 the mucous membrane of the tube during menstruation. He 
 reports two cases in which the changes in the mucosa of the tubes 
 were similar to those in the uterus, i. e., there was a marked 
 engorgement with blood, and free blood was found in the mucosa 
 and in the lumen of the tube. The surface epithelium was found 
 intact. 
 
 In conclusion we would offer the following resume : 
 
 The menstrual cycle is divided into three periods; the first, 
 preceding the menstrual flow and characterized by gradual devel- 
 opment of a capillary congestion resulting in a serous or sero- 
 sanguineous infiltration of the stroma of the mucosa, separating 
 the cellular elements and gravitating toward the periphery form- 
 ing collections of blood beneath the surface epithelium. Slight 
 glandular hypertrophy as well as a swelling of the gland cells 
 
 ID 
 
 Digitized by Microsoft® 
 
takes place, with a consequent hypertrophy of the whole mucosa. 
 This stage is well exhibited in figure I, taken from a drawing by 
 one of my students, Mr. Harry Rullison, to whom I would express 
 my indebtedness, of a section of uterus removed immediately pre- 
 ceding the menstrual flow. A capillary engorgement and infiltra- 
 tion of the stroma, separating the cellular elements is especially 
 noticeable in the outer portion. The enlarged cork-screw-like 
 glands suggest a mild form of glandular hypertrophy. 
 
 The second stage of active hemorrhage is characterized by an 
 outflow of blood through a diapedesis or through a rupture of 
 the sub-epithelial hematoma, with more or less loss of the super^ 
 ficial epithelium depending largely upon the severity of the process. 
 Figure II, a drawing of a section of a uterus removed the first 
 day of the menstrual hemorrhage shows the infiltrated area near 
 the surface with a breaking in the continuity of the surface epithe- 
 lium allowing the blood an exit. In this plate the mucosa is 
 drawn relatively too norrow. 
 
 In the third stage involution takes place. The engorgement 
 of the blood vessels becomes quickly lessened and a rapid absorp- 
 tion of the blood in the stroma takes place. The epithelium which 
 has been lost is regenerated by indirect cell division of the adjacent 
 cells. This stage is prolonged until the premenstrual swelling 
 begins or there is a slight interval during which the mucosa is 
 
 inactive. 
 
 BIBLIOGRAPHY. 
 
 Christ. 
 
 DOLERIS, 
 
 Edgar. 
 
 FiNDLEY. 
 
 Gerhard. 
 
 KUNDRAT AND EnGELMANN. 
 
 Leopold. 
 
 LOHLEIN. 
 
 Mandl. 
 
 MORICKE. 
 
 Das Verhalten der Uterusschleimhaut wahrend der Men- 
 struation. Inaugural Dissertation, Giessen, 1892. 
 
 Noui'elles Archives d' Obstetrigue et de Gynaecologie, 1894, 
 p. 200. 
 
 Text-Book of Obstetrics, 1903, p. 21. 
 
 Gynaecological Diagnosis, 1903; also The American Jour- 
 nal of Obstetrics, April, 1902. 
 
 Ueber das Verhalten der Uterusschleimhaut bei der Men- 
 struation. Verhandlung der Gesselschaft fur Geburt- 
 shiilfe und Gynaecologie zu Berlin, Zeitschrift fiir 
 Geburtshiilfe und Gynaecologie, Bd. xxxii., s. 296. 
 
 Untersuchungen iiber Uterusschleimhaut. Strieker's Medi- 
 cinische Jahrbuch, 1873. 
 
 Studien iiber dis Uterusschleimhaut wahrend die Men- 
 struation, Schwangerschaft und Wochenbett. Archiv 
 far Gyndkologie, Bd. xi. 
 
 Gynakologische Tagesfragen, 2 H., 1891. 
 
 Beitrag zur Frage des Verhalten der Uterusmucosa 
 wahrend der Menstruation, Archiv fur Gynaecologie, 
 1896, s. 557. 
 
 Die Uterusschleimhaut in der Verschiedenen Altersperio- 
 den und zur zeit der Menstruation. Zeitschrift far 
 Geburtshiilfe und Gynakologie, Bd. vii. 
 
 II 
 
 Digitized by Microsoft® 
 
SlNETY. 
 
 Undekhill. 
 
 Van Meerdervort. 
 Van Kahlden. 
 
 •Wendeler. 
 
 Westphalen. 
 
 Williams, J. W. 
 Williams, Sir John. 
 
 Wyder. 
 
 Recherches sur la Muqueuse uterine pendant la Menstrua- 
 tion, Annals de Gynecologie, 1881, p. 295. 
 
 Note on the Uterine Mucous Membrane of a Woman 
 who Died Immediately after Menstruation. The Edin- 
 burgh Medical Journal, August, 1875. 
 
 Die normale und menstruirende Gebarmuterschleimhaut, 
 Inaugural Dissertation. Freiburg, 189s, Quoted. 
 
 Ueber das Verhalten der Uterusschleimhaut wahrend und 
 nach der Menstruation, Beitrdge su Geburtshulfe und 
 Gyndkologie, 1889. 
 
 Verhandlung der Gesellschaft fiir Geburtshiilfe u. Gyna- 
 kologie zu Berlin, Zeitschrift fUr Geburtshiilfe und 
 Gynackologie. Bd. xxxii., s. 317. 
 
 Sur physiologie der Menstruation. Archiv fiir Gynae- 
 cologie, 1896. s. 35. 
 
 Text-Book of Obstetrics, 1903, p. 73. 
 
 The Mucous Membrane of the Body of the Uterus. 
 Obstetrical Journal of Great Britain and Ireland. Lon- 
 don, 1875 and 1877. 
 
 Beitrage zur normalen und patbologischen Histologic der 
 menschlichen Uterusschleimhaut. Archiv fur GynS- 
 kologie, Bd. xiii. Das Verhalten der Mucosa Uteri 
 wahrend und nach der Menstruation. Zeitschrift fUr 
 Geburtshiilfe und Gyndkologie, Bd. ix. 
 
 12 
 
 Digitized by Microsoft® 
 
Reprinted from Albany Medical Annals, January, 1904. 
 
 THE ISLANDS OF LANGERHANS IN CONGENITAL 
 SYPHILITIC PANCREATITIS. 
 
 By RICHARD MILLS PEARCE, M. D., 
 
 Adjunct Professor of Pathology and Bacteriology, Albany Medical Collage. 
 
 {From the Bender Hygienic Laboratory, Albany, N, y.) 
 
 It is not my purpose in the present communication to give a 
 detailed study of the lesions of congenital syphilitic pancreatitis. 
 The condition is not important clinically, for it is found usually 
 only in the still-born or in those dying shortly after birth; the 
 pathology of the condition on the other hand, has been thoroughly 
 studied and little can be added to the original description of 
 Birch-Hirschfeld and the later systematic study of Schlesinger. 
 The histological picture, however, shows a peculiar relation between 
 the islands of Langerhans and the newly formed tissue which, 
 as it has not been generally recognized and as it offers evidence 
 in confirmation of our newer ideas concerning the anatomical and 
 physiological independence of these structures, is I believe, worthy 
 of special consideration. The study of the islands in this con- 
 dition is also important, because occurring as it does in the foetus, 
 it offers an opportunity to study the structures under adverse 
 conditions of development, thus filling the gap between the normal 
 development and the pathological changes seen in the adult organ. 
 
 Our recently acquired knowledge of the islands is almost 
 entirely due to the work of Opie,^ who has shown that a certain 
 proportion of the cases of pancreatic diabetes have a definite 
 relation to lesions of the islands of Langerhans. Upon these 
 observations, which have been confirmed by several American 
 and German investigators, is based the assumption that the islands 
 
 Digitized by Microsoft® 
 
have a physiology distinct from the remainder of the pancreas, 
 and concerned especially with carbo-hydrate metabolism, there- 
 fore of prime importance in the pathology of diabetes. Very 
 substantial support of this analtomical and physiological inde- 
 pendence is offered by the experiments of Ssobolew' and Schulze.^ 
 These investigators, working independently, have shown that after 
 ligation of the pancreatic duct in animals, the glandular portion 
 of the organ atrophies and is replaced by fibrous tissue while the 
 islands are unchanged and the urine remains free from sugar. 
 Opie has observed a similar condition in the pancreas of man 
 as the result of partial or complete duct obstruction by calculus 
 or cancer. 
 
 It is this same picture of a chronic sclerotic inflammation with 
 disappearance of the greater portion of the glandular structure 
 and the persistence of the islands that is seen in the congenital 
 luetic lesion of the pancreas. This persistence of the islands has 
 not, however, been generally recognized, owing probably to the 
 infrequency of the post-mortem examination of newly born chil- 
 dren and in part to the fact that our interest in the pathology of 
 the island has only recently been aroused. Previous to this 
 awakening the only references are those of Schlesinger* and Con- 
 don° who merely mention the presence of the islands ; while, since, 
 Opie and Ssobolew the chief supporters of the theory of anat- 
 tomical independence have offered this persistence of the islands 
 as evidence in support of their view. At the time of their pub- 
 lications, however, our knowledge of the normal development of 
 the islands was incomplete. This gap has been filled by my 
 recent study* of the development of the island in the human 
 embryo, and it is in connection with the results of this study that 
 I wish to consider the changes in the syphilitic organ. 
 
 Literature of Congenital Syphilitic Pancreatitis. The first 
 complete study of luetic pancreatitis of the new bom was made 
 by Birch-Hirschfeld' in 1875. In the series then reported, lesions 
 of the pancreas were found in thirteen of twenty-three cases 
 examined, but in a later series, in but twenty-nine' of one hundred 
 
 Digitized by Microsoft® 
 
twenty-four. The frequency of the lesion in the second series 
 agrees with the observations of subsequent writers. 
 
 The gross appearance of the organ is characteristic, the most 
 striking feature being the increase in consistence, which in 
 advanced cases approaches that of cartilage. An increase in size 
 and weight is a,lso quite constant. On section the surface of the 
 organ is hard, white and glistening, the glandular appearance 
 being entirely effaced. Birch-Hirschfeld compares the appear- 
 ance on section to that of a fibroid; Cruveilhier, to that of a 
 scirrhus cancer. Previous to Birch-Hirschfeld's study, this 
 induration of the pancreas in congenital lues had been noted in 
 isolated instances by Osterloh,' Cruveilhier," Hecker,^" CEdmans- 
 son" and Wegner." The changes are most marked in the head 
 of the pancreas, a fact which led Mraczek^' and Beck" to consider 
 the lesion to be an extension from the duodenum. Schlesinger 
 states that there is no constant relation to syphilis of other organs, 
 though the pancreas may be adherent to such. 
 
 Birch-Hirschfeld (1875) describes the lesion, histologically as 
 an increase of connective tissue which compresses and causes 
 atrophy of the glandular elements, without evidence of degenera- 
 tion. Schlesinger (1898) gives a most complete histological 
 study and decides that the increase of connective tissue is of peri- 
 vascular origin. Contributions to the pathology of the lesion 
 have been made also, by Huber," Friedreich,^' Mraczek, Miiller,^'' 
 Baumgarten,^* Beck, Heubner,^" Ssobolew, Opie and Condon. 
 Gummata have been found only by Klebs,^" Beck and Birch- 
 Hirschfeld which is in striking contrast to the syphilitic lesion of 
 the adult pancreas. Baumgarten and Cruveilhier describe a lar- 
 daceous appearance, but whether they use this term in the descrip- 
 tive sense or to specifically refer to amyloid infiltration, is not clear 
 from the text. Some difference of opinion exists concerning the 
 period at which the lesion develops. Birch-Hirschfeld places it 
 late in pregnancy; Schlesinger much earlier and quotes cases 
 described by Muller and Mraczek in which it existed at the fifth 
 month. Opie's statement that the almost complete absence of 
 
 Digitized by Microsoft® 
 
glandular tissue is due to arrested development rather than to 
 degeneration indicates that he favors Schlesinger's view. It may 
 be mentioned that the condition has been found five months 
 (Birch-Hirschfeld) and three and a half years (Heubner) after 
 birth. The involvement of the pancreas occurs in order of fre- 
 quency after that of bone, lung, spleen and liver. 
 
 Material. My interest in congenital syphilitic pancreatitis was 
 first aroused by Prof. Marchand of Leipzig, at whose suggestion 
 I undertook an investigation of several minor points concerning 
 the development and pathology of the islands of Langerhans. 
 From the material of the Leipzig Institute, six specimens were 
 obtained. Later while completing a study of the embryology of 
 the islands, with the aid of material placed at my disposal by Prof. 
 Mall of Baltimore, two additional examples were found. A single 
 specimen was obtained from Dr. Flexner's collection of pancreas 
 material and one also from the collection of Dr. George Blumer. 
 I take this opportunity to thank these gentlemen for placing this 
 unusual material at my disposal. 
 
 In all of these ten specimens the lesion is of the advanced sclero- 
 tic type with almost complete disappearance of the glandulair 
 elements. The early stages with infiltration of plasma and 
 lymphoid cells are not seen. A careful study of the lesion adds 
 practically nothing new to our knowledge of the general process. 
 Selective stains prove the new tissue to be mainly of the white 
 fibrous variety; the elastic tissue to be unchanged, or indeed in 
 some instances, less in amount than in the normal pancreas. In one 
 case very widespread mucoid degeneration of the new connective 
 tissue was seen. Miliary gummata (syphilome miliare) were 
 found in but one of the ten specimens. 
 
 The most striking feature is the persistence of the islands of 
 Langerhans, despite the general destruction of the glandular 
 acini. Throughout the dense fibrous tissue, with here and there 
 the fragmentary remains of gland tissue, the islands stand out 
 prominently, often surrounded by denser whorls of investing 
 stroma, but apparently uninjured. (See illustration.) This 
 
 Digitized by Microsoft® 
 
To Illustrate Dr Pearce's Article on " The Islands of Langerhans in Congenital 
 Syphilitic Pancreatitis." 
 
 Albany Medical Annalx, January, 1904. 
 
 s^^i^ •; 
 
 I*!? v V , ' ' 
 
 ;''-l., V -.' ■ ^ 
 
 
 -,-/t 
 
 t .- I\ •, '.•■■ "■.. '■-■ ■ ■■:■■ .■■■'■ / - "' . 
 
 - ■■„■-■ %---,-■- .• ■• ." • , ■ " .-~^^ 
 
 *-.*(&-"' =iV«t ' 
 
 
 i^i'; 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
appearance is present constantly in all specimens and in all grades 
 of the lesion, and the most careful search fails to reveal, in the 
 islands, evidence of degeneration or atrophy, or the ingrowth of 
 connective tissue. The absolute resistance of the islands to the 
 process so destructive to the secreting portion of the gland is very 
 strong evidence in favor of their anatomical independence and this 
 condition, if considered in connection with our knowledge of the 
 development of the islands, is very readily explained. While in 
 most instances the islands appear entirely isolated, close examina- 
 tion shows that occasionally an island is connected With a glandu- 
 lar acinus by a stalk-like process of cells. This connecting band 
 is in most instances a solid mass of epithelial cells ; though, some- 
 times, a semblance of a lumen may be discerned. Opie alone, as 
 far as I am aware, notes this band of cells connecting island and 
 acinus. Upon this observation and the fact that no evidence of 
 degeneration is seen in the island he bases the assumption that 
 the island in its early development is in connection with the acinus 
 and that its development is interrupted, in syphilitic pancreatitis, 
 by the ingrowth of connective tissue which causes atrophy of the 
 glandular tissue and leaves the more resistant island in its unde- 
 veloped state. At the time of Opie's communication our knowl- 
 edge of the development of the islands was incomplete and 
 especially so the manner in which the island became independent 
 of the acinus from which it originated. That his assumption is 
 correct, however, is shown by the results of my recent study of 
 the development of these structures in the pancreas of the human 
 embryo. These results may be briefly summarized as follows : 
 
 "The islands of Langerhans (embryo of fifty-four millimetres), 
 originate through a proliferation and differentiation of the cells 
 of the primitive secreting tubules. The differentiated cells char- 
 acterized by a rich, finely granular, eosinophilic protoplasm lie 
 as small round or oval masses in direct continuity with the cells 
 of the tubule. Later, (embryo of about the third month) the 
 attached portion becomes constricted, and lengthening, forms a 
 a stalk-like, solid process of cells connecting the island with the 
 
 Digitized by Microsoft® 
 
acinus. At this period, a few entirely isolated islands are present. 
 A separation takes place apparently brought about by the pressure 
 of the investing connecting tissue. Vascularization has now 
 occurred. In still later stages a progressive vascularization, 
 increase of cells, arrangement of the cells in columns, and appear- 
 ance of a fine reticulum is observed. The rapidly forming glan- 
 dular structures finally surround the islands which then occupy 
 the centres of the lobules." 
 
 From this summary it will be seen that in the pancreas of the 
 third month the islands are still connected with the acini by solid 
 processes of cells. The proportion of connective tissue at this 
 time is very great, as the rapid glandular growth does not occur 
 until the fifth and sixth months. Any disease of the pancreas 
 occurring at this time would, therefore, influence the develop- 
 ment of the organ, and evidence of this faulty development would 
 be present at any subsequent stage of the lesion. 
 
 This seems to be the case in the lesion under consideration. 
 Syphilis of the pancreas has been observed as early as the fifth 
 month (Miiller) and although there are no definite observations 
 of its earlier occurrence, the appearance of the islands in the luetic 
 pancreas when compared with the normal development of these 
 structures indicates the possibility of the lesion occurring as early 
 as tlie third month. The general microscopic picture of the normal 
 pancreas at the end of the third month is very similar to the syphi- 
 litic organ at the sixth or seventh month, except that the latter has a 
 larger proportion of connective tissue : the same stage of develop- 
 ment of the islands is seen in both. In other words, at the latter 
 period we have the development of the third month plus the connec- 
 tive tissue produced by the syphilitic lesion. It seems justifiable, 
 therefore, to conclude that the syphilitic infection occurs at the 
 period (third to fourth month) when the islands are still connected 
 with the gland tubules, and that the newly formed tissue causes 
 atrophy of the gland acini, but spares the more resistant islands. 
 Thus is explained the similarity of the islands in the two conditions. 
 The ultimate factor in this resistance of the islands is their abundant 
 
 Digitized by Microsoft® 
 
blood supply. That this vascularization of the island is not altered 
 in syphilitic pancreatitis, I have demonstrated by the study of the 
 diseased organ injected with Berlin Blue. 
 
 The evidence obtained by experiment and by the study of the 
 pancreas of diabetics has already placed the theory of anatomical 
 independence of the islands upon a substantial basis. The study 
 of their development, has demonstrated that although they orig- 
 inate from the glandular tissue they eventually become wholly 
 independent structures anatomically. It is with the object of adding 
 still further confirmatory evidence, that this communication which 
 adds nothing essentially new, but attempts to correlate known 
 facts, is made. 
 
 REFERENCES. 
 
 1 Opie. 
 
 2 SaoBOLEw. 
 
 3 SCHULZE. 
 
 4 SCHLEStNGEE. 
 
 5 CONDOK. 
 
 6 Peabce. 
 
 7 Birch - Hirschfeld. 
 
 8 Obterloh. 
 
 9 Crdveilhub. 
 
 . 10 HecKER. 
 
 11 Oedhansson. 
 
 12 WSBIfEB. 
 
 13 Mraczek. 
 
 14 Beck. 
 
 15 Hdber. 
 
 IS Friedreich. 
 
 17 MSllbb. 
 
 18 Bauhoabten. 
 
 19 Heitbner. 
 
 20 Elebb. 
 
 Journal of the Boston Society of the Medical Sciences, 1900, iv, 251 . 
 
 Journal of Experimental Medicine, 1901, v, 397, 527. 
 
 Centralhlattfiir allgemeine Pathologle undpatholegiecTte Anatomie, 
 1900, xl, 202 ; yirehcw's Archiv, 1902, clxviii, 91. 
 
 Archiv fiir mikrosJaypiscJie Anatomie, 1900, Ivl, 491. 
 
 Virehote's Archiv, 1898, cliv, 501. 
 
 Transactions of lh,e Chicago Pathological Society, 1899-'01, iv, ZOO. 
 
 American Journal of Anatomy, 1903,11,445; University of Penn- 
 sylvania Medical Bulletin, 1903, xvl. 341. 
 
 Archiv der BeUkunde, 1875, xvl, 166 ; Oerliardt's Handbach der 
 Kinderkrankheiten, iv, Abth. ii, 753, Tflbingen, 1878. 
 
 Mittheilungen aus der kdnigliche sachsische Entbindungsinstiiut 
 zu Dresden. Qaoted by Biroh-Hirschfeld. 
 
 Atlas d'anatomie pathologigue, (Observat. 6 & 7). Quoted by 
 
 BlRCH'HlRBCHrSLD. 
 
 Virchow''s Archiv, 1859, xvii, 190 
 
 Niyrd. Med. Arch., i, 4. Kef. in Virchow-Birseh Jahresberieht fiir 
 
 1869, 2 abth., 561. 
 7irefiow''s Archiv, 1870, 1, 305. 
 Archiv far Dermatologieund SyphUis, iea3,x, 209; 1887, xiv, 117; 
 
 1893, 2 Erg, 279. 
 Pragermedicinische Wochenschrift, 1884, ix, 257, 266. 
 Archiv der Eeilkunde, 1878, xix, 425. 
 Pankreagkrankhelten, v. Zietnssen't Handbuch, 1878, vili, Sabtli., 
 
 254. 
 Virehow^s Archiv, 1883, xcii, 532. 
 Tirchow's Archiv, 1884, xcvii, 89. 
 
 Syphilis Hereditaria, 1886, p. 33. Quoted by Schlebingbr. 
 Bandbuch der pathologisehe Anatomie, Berlin, 1870, Bd. i, 2 abth. 
 581. 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
Reprinted from Albany Medical Annals, October, 1904. 
 
 CANCER OF THE PANCREAS AND GLYCOSURIA.* 
 
 {A Histological Study of the Changes in the Islands of Langerhans) 
 
 By RICHARD M. PEARCE, M. D., 
 
 Director of the Bender Laboratory, Albany, N. Y. 
 
 This investigation was undertaken for the purpose of deter- 
 mining what relation, if any, the islands of Langerhans may- 
 have to the glycosuria occasionally seen in connection with 
 cancer of the pancreas. That pathological changes in these 
 structures are frequently associated with diabetes has been 
 satisfactorily demonstrated by the investigations of Opie^ and 
 others, and has led to a more definite knowledge of a form of 
 diabetes which may properly be called pancreatic diabetes. 
 No attempt has been made on the other hand, as far as I am 
 aware, to demonstrate the presence of similar changes in con- 
 nection with the glycosuria occasionally associated with cancer 
 of the pancreas. That some of these changes — hyaline de- 
 generation, chronic interacinar pancreatitis involving the 
 islands and, according to a few investigators, diminution in the 
 number of the islands — might result from the involvement of 
 the organ by cancer is evident from our knowledge of the 
 secondary lesions frequently accompanying such invasion. It 
 is well known that duct obstruction, frequently the result of 
 cancer of the head of the pancreas, readily causes a chronic 
 pancreatitis; and it is possible also that the destruction of a 
 large portion of the pancreas parenchyma by the new growth 
 may destroy the islands to such an extent that they are no 
 longer equal to the demands made upon them. 
 
 On the other hand we have no adequate explanation of those 
 cases in which, though the pancreas is extensively or com- 
 pletely replaced by the new growth, diabetes does not exist. 
 
 •Previonaly pablished In the American Jmimal of the Medical Sciences, September, 1904, 
 CZZTIU, 478. 
 
 Digitized by Microsoft® 
 
Such complete destruction, as in the cases reported by Litten* 
 and V. Hansemann,' is analagous to the extirpation experi- 
 ments by V. Mering and Minkowski,* but the absence of gly- 
 cosuria is opposed to their results, v. Hansemann has at- 
 tempted to explain the discrepancy by assuming that in 
 primary cancer of the pancreas the tumor cells originating 
 from pancreas parenchyma inherit the function of internal 
 secretion peculiar to the normal cell. This explanation, how- 
 ever, is entirely hypothetical and in the light of our present 
 knowledge of pancreatic diabetes loses its force unless we as- 
 sume that all primary cancers of the pancreas arise from the 
 islands of Langerhans, which, despite the recent publication 
 of Fabozzi'^ supporting such origin, appears most improbable. 
 If our newer theories concerning the relations of the islands 
 to carbohydrate metabolism are firmly grounded, we should 
 expect to find evidence of the persistence of the islands in such 
 cases. With these various possibilities in mind, I have made 
 a careful histological study of thirty cases of cancer of the 
 pancreas, in three of which glycosuria was present. 
 
 The literature ofifers but scanty information concerning the 
 frequency with which glycosuria occurs in connection with 
 cancer of the pancreas. Such cases are not uncommonly re- 
 ported but without reference to the occurrence of cases not 
 complicated by glycosuria. From a hasty review of the sub- 
 ject, I believe the proportion (one in ten) shown by my cases 
 to be high. We have, however, definite information concerning 
 the reverse condition, that is, the frequency of cancer of the 
 pancreas in connection with all forms of diabetes. In Oser's' 
 series of i88 cases of diabetes, cancer occurred in 24; in Wil- 
 liamson's' series of 23, once; and in 139 cases reported by 
 Windle,^ in three, v. Hansemann's" tabulation of 72 cases of 
 diabetes with pancreatic lesion contains five cases of cancer. 
 I have examined many of the original descriptions of the cases 
 quoted in these series with the object of determining if any 
 peculiar macroscopic or microscopic conditions existed. The 
 task has been most unsatisfactory for many of the reports are 
 of a clinical nature and gross and histological description are 
 meagre or entirely lacking. This analysis, however, confirms 
 the well known clinical fact that in the majority of cases the 
 new growth is in the head of the pancreas and by duct obstruc- 
 
 Digitized by Microsoft® 
 
tion causes an atrophy (chronic pancreatitis?) of the distal 
 portion of the organ. This point is especially emphasized by 
 V. Hansemann, Dieckhoflf" and Windle. 
 
 The material upon which this study is based consists of the 
 tissues from thirty cases of cancer selected from a group of 
 thirty-five examples of new growth in the pancreas. As the 
 material has been gathered from different sources, often from 
 routine collections of autopsy material, it has not always been 
 possible to procure tissues from the different portions of the 
 uninvolved pancreas as well as from the cancerous portion, nor 
 for the same reason, have essential data concerning clinical his- 
 tory and pathology been available in all cases.* Thus while all 
 cases have been utilized in the study of the histological changes 
 but twenty-one have been available, on account of incomplete 
 examination of the urine, in arriving at conclusions concerning 
 the occurrence of glycosuria. The thirty cancers studied may 
 be classified as follows: primary in the head of the pancreas, 
 nine; in other portions of the organ, 4; secondary to gall 
 bladder or ducts, five ; to stomach, eight ; to breast, one, and 
 to uterus, one. Glycosuria occurred in but three cases; in 
 two of which the tumor was primary in the head of the pan- 
 creas ; concerning the third, no clinical or pathological notes 
 were available. 
 
 The results of the histological examinations, especially as 
 concerns the islands of Langerhans are more readily described 
 under the following headings: changes in (a) the immediate 
 neighborhood of the new growth, (b) at a distance from the 
 tumor, (c) in cases with extensive involvement, but no glyco- 
 suria, and (d) in cases accompanied by glycosuria. 
 
 (a) The important changes at the edge of the advancing 
 cancer are the persistence and increase in size of the islands 
 of Langerhans. While the glandular elements are invaded and 
 replaced by the new growth and the newly formed connective 
 tissue, the islands remain unaltered in the midst of the stroma, 
 and occasionally, are even enclosed by a tnass of tumor cells. 
 Not only is this true of the edge of the tumor, but in the well 
 infiltrated parts the islands may be seen in the stroma without 
 
 •I take this opportunity to acknowledge my indebtedneee to Prof. P. Marchand of Leipzig 
 Dr. Lndwig Hektoen of Chicago, Dr. F. B. Mallory of Boston, Dr. James Ewing of New York, 
 Dr. W. G. MacCallnm of Baltimore, Dr. W. T. Longcope of Philadelphia, Dr. M. Herzog of 
 Chicago, and Dr. George BInmer, who have kindly placed material at my disposal. 
 
 Digitized by Microsoft® 
 
remnants of acini. The increase in size of some of these struc- 
 tures is four or five times that of the normal islands. These 
 giant islands are typical in structure and readily distinguished 
 from small alveoli of cancer cells. They represent a true 
 hypertrophy of the islands and not altered pancreas acini. 
 
 (b) The changes at a distance from the tumor depend appar- 
 ently upon the location and size of the same. With a cancer 
 of the head of the pancreas causing pressure upon the duct 
 a chronic interstitial pancreatitis of severe grade always ex- 
 isted, which, with the exception of one case to be described 
 later, was always of the interlobular type. The islands per- 
 sist in the midst of dense connective tissue and are not invaded 
 by it. This condition has been observed by Opie in several 
 similar cases without glycosuria. 
 
 A tumor at the head of the pancreas not obstructing the duct 
 does not cause chronic interstitial changes. Small tumors of 
 any portion of the pancreas are unaccompanied by changes in 
 the islands of the normal portion of the organ; on the other 
 hand, if the new growth is of considerable size, definite hyper- 
 trophy of these structures is apparent. 
 
 (c) But two cases of diffuse involvement of the organ 
 occurred in this series; in neither was glycosuria present. 
 These are, however, not analagous to those discussed by 
 Hansemann, for areas of normal pancreas could be found here 
 and there, lying in the midst of the new growth. In such 
 areas were large islands, several times the normal size. 
 
 (d) Three cases of cancer of the pancreas accompanied by 
 glycosuria are included in this series. One of these could not 
 be utilized in the present study because of lack of data; the 
 gross description of the pancreas not being available and the 
 histological study incomplete inasmuch as only the tumor had 
 been preserved and a study of the uninvolved portion of the 
 pancreas was therefore impossible. The second case, a primary 
 cancer of the pancreas involving the head only and accom- 
 panied by a slight intermittent glycosuria with free fat in the 
 stools, showed no changes in the uninvolved portion of the 
 organ. The islands were few in number and the increase in 
 size usually seen in the immediate vicinity of the new growth 
 was not evident. The third case was one of cancer of the 
 pancreas with advanced chronic pancreatitis of the interacinar 
 
 Digitized by Microsoft® 
 
type. Sugar had been present in the urine for several months 
 before death. The entire length of the pancreas was fifteen 
 centimetres ; the proximal five centimetres occupied by the 
 tumor, the remainder of the organ atrophied and indurated. 
 Histological examination revealed a severe grade of chronic 
 pancreatitis with small scattered groups of more or less altered 
 acini. Within these, smaller strands of connective tissue had 
 penetrated, involving to a varying extent, the islands of Langer- 
 hans. 
 
 While it is impossible to draw any definite conclusions from 
 such a small number of cases, these observations are very sug- 
 gestive and offer some support of the newer ideas concerning 
 the anatomy and physiology of the islands. The persistence 
 of the islands in the midst of the tumor stroma supports the 
 theory of anatomical independence, an independence which 
 I have recently definitely demonstrated by a study of their 
 development.^^ This persistence is analagous to that described 
 by Schlesinger,'^ Opie, Condon^^ and Pearce" in congenital 
 syphilitic pancreatitis, and by Opie in various other sclerotic 
 conditions of the organ. The increase in size of the islands 
 suggests a purposeful enlargement due to increased function 
 and therefore of the nature of a compensatory hypertrophy. 
 A similar hypertrophy of these structures has been described 
 by Warthin^° and by Ohlmacher^" in connection with cancer 
 of the liver, and is considered by them, also, to be compensatory 
 in nature, compensating, however, for the lost or limited liver 
 function. The livers of my cases have not been available in 
 all instances for histological examination, and I have not 
 therefore been able to arrive at any conclusion concerning the 
 relation of enlarged islands to lesions of that organ. It is possi- 
 ble that a relation of a compensatory nature may exist between 
 the carbohydrate metabolism of the liver and the islands of 
 Langerhans ; but the hypertrophy of the latter structures when 
 associated with cancer of the pancreas is, I believe, an attempt 
 to compensate for the function of the lost islands. 
 
 The cases of diffuse cancer of the pancreas without glyco- 
 suria can be explained only on the supposition that the amount 
 of normal pancreas remaining is sufficient to furnish the secre- 
 tion necessary to carbohydrate metabolism, v. Mering and 
 Minkowski found in their extirpation experiments that if one- 
 
 5 
 
 Digitized by Microsoft® 
 
fourth to one-fifth of the pancreas remained, diabetes did not 
 result ; a destruction of the pancreas by cancer we may con- 
 sider analagous to partial extirpation. In both of my cases 
 of diffuse cancer, small areas of pancreas containing islands 
 could be found in the midst of the new growth in a sufficient 
 amount to correspond to the unextirpated fifth. In this con- 
 nection the possibility of an accessory pancreas (Thorel," 
 Sweet^'*) or other glands maintaining the balance of function 
 must also be considered, v. Hansemann's explanation of the 
 absence of glycosuria is not applicable unless we assume that 
 all cancers of the pancreas arise from the cells of the islands 
 oi Langerhans. 
 
 Finally we have the cases with glycosuria. In one of these 
 a true diabetes existed, in the other an intermittent glycosuria. 
 The first is readily explained by the presence of an interacinar 
 pancreatitis with involvement of the islands, a type of pan- 
 creatitis originally described by Opie in connection with diabetes 
 and since observed by many investigators. It therefore appears 
 that while cancer of the head of the organ with duct obstruction 
 may frequently cause a chronic pancreatitis, glycosuria does 
 not result unless the interacinar form with involvement of the 
 islands develops. This naturally brings cancer diabetes into 
 a common relation with pancreatic diabetes in general. 
 
 The case of intermittent glycosuria brings up the vexed 
 question of alimentary glycosuria. - It is well known that a 
 normal individual may digest without disturbance from 150 to 
 250 grammes of sugar, but that an increase over this amount 
 may lead to temporary excretion of glucose in the urine. This 
 excretion, due to hyperglycemia, is the inevitable result of a 
 sudden flooding of the circulation with an amount of glucose 
 beyond that which the body is able at once to consume. On 
 the other hand, in a number of abnormal conditions where 
 there is no actual diabetes, an alimentary glycosuria may be 
 produced by an amount of sugar which in the economy of the 
 normal individual is easily consumed. Among the prominent 
 conditions in which this occurs may be mentioned neuroses, 
 traumatic or otherwise, organic nervous diseases, alcoholism, 
 infectious diseases, exophthalmic goitre, hepatic and pan- 
 creatic disease. Is it not possible that this temporary excretion 
 of sugar is due to some transient disturbance of the function 
 
 Digitized by Microsoft® 
 
of the Islands brought about by vascular disturbances of the 
 islands or by some toxic substance, perhaps of the nature of an 
 autotoxin, affecting the secretion of these structures or tem- 
 porarily throwing them out of function ? In a destructive dis- 
 ease of the pancreas, as cancer, a certain number of the islands 
 are necessarily destroyed; the remainder, however, may be 
 sufficient under normal circumstances to furnish the necessary 
 secretion, whatever it may be, essential to carbohydrate meta- 
 bolism; but if overtaxed by an unusual injection of glucose, 
 hyperglycemia and a consequent glycosuria might result. This 
 explanation has occurred to me in connection with my case 
 of cancer accompanied by intermittent glycosuria; and that it 
 is not entirely hypothetical is indicated by the observations of 
 Ssobolew,^" confirmed by Herzog,'"* that in some cases of per- 
 manent diabetes the pancreas exhibits a great diminution in 
 the number of islands, and in some cases, an entire disappear- 
 ance. This deficiency, however, is far from proven, and in 
 mentioning it I do not wish to be placed in the position of 
 advocating a pancreatic origin for all forms of diabetes. 
 
 Conclusions — We have a definite basis for ascribing a certain 
 portion of the cases of diabetes to lesions of the pancreas and ulti- 
 mately to lesions of the island of Langerhans. As a result of the 
 study here briefly outlined we can, I believe, include the permanent 
 diabetes of cancer of the pancreas in this group. Further 
 observations must be made before the intermittent glycosuria 
 can be attributed to destruction of large numbers of islands. The 
 absence of diabetes in cases of extensive involvement of 
 the pancreas would appear to be due to the survival of small 
 fragments of tiie pancreas including a sufiicient number of islands 
 to carry on their peculiar function. The persistence of the 
 islands under the adverse conditions described supports the 
 observation heretofore made concerning their anatomical inde- 
 pendence, while their enlargement, suggesting as it does a com- 
 pensatory hypertrophy, is corroborative of the theory that they 
 have an independent special function. 
 
 REFERENCES. 
 
 1. CPU. Diaease of the Pancreas, its Cause and Nature, Philadelphia 
 
 and London, 1903. 
 
 2. LiTTEK. Charite Annalen, 1878, v. 181. 
 
 S. T. HANaxHAHH. Specljleltat, Altrulsmus, und Anaplasle der Zellen, Berlin 
 
 1888; Be/. Zeilschr.,/. Jclln. Med., 1891, zxvi, 198. 
 
 Digitized by Microsoft® 
 
4. 
 
 V. Mbriks uitd Minkowski. 
 
 5. 
 
 Fabozzi. 
 
 6. 
 
 Obbb. 
 
 7. 
 
 Williamson. 
 
 8. 
 
 WiNDLE. 
 
 9. 
 
 V. Hansemann. 
 
 10. 
 
 DlKCKHOFP. 
 
 11. 
 
 Fbarce. 
 
 18. 
 
 SCHLESINGSB. 
 
 13. 
 
 Condon. 
 
 14. 
 
 Pkabce. 
 
 IS. 
 
 Warthin. 
 
 16. 
 
 Ohlmaoheb. 
 
 17. 
 
 Thokel. 
 
 18. 
 
 Sweet, 
 
 19. 
 
 SSOBOLBW. 
 
 30. 
 
 Hbbzog. 
 
 Arch.f. exper. Path. u. PTMrmakol-, 1889, xxvi, 871. 
 
 ZieglerU Beitrage, 1903, xxxiv, 199. 
 
 Nothmgd's apecielle Path. «. Therapie, 1898, xviii, 67. 
 
 Diabetes Mellltue, Bdmbiirgh and London, 1898. 
 
 Dublin Journal of Medical Sciences, ,1883, Ixxvl, 118. 
 
 Zellschr.f. hlin. Med., 1894, xxvi, 191. 
 
 Beitrage zur. Path. Anat. des Pankreas, Leipzig, 1895. 
 
 American Journal of Anatomy, 1903, 11, 445. 
 
 nrchmo's ArcMv., 1898, cliv, 501. 
 
 Trans. Chicago Path. Soc, 1899-'01. iv., SOO. 
 
 American Medicine, 1903, vl, 1II20. 
 
 Philadelphia Medical Journal, 1900, vi, 124. 
 
 Trans. Chicago Path. Soc, 1904, vi, 83. 
 
 Virchow's Archiv, 1903, clxxiii, 231. 
 
 Journal of Medical Besearch, 1903, x, 855. 
 
 Ceniralb.f. AUg. Path. u. Path. Anat, 1900, xi, 202; Tirehow'i 
 
 Archiv., 1902, cxlviii, 91. 
 Virchow's Archiv., 1902, clxriii, 83. 
 
 Digitized by Microsoft® 
 
Reprinted from the Ai-baht Mkdioai. Ahkals, May, 1904., 
 
 EXPERIMENTAL PANCREATITIS CONSIDERED IN 
 
 ITS RELATION TO THE ACUTE FORMS OF 
 
 PANCREATITIS IN MAN. 
 
 Presented before the Medical Society of the Greater City of New York, 
 February 8, 1904, in a discussion of "Diseases of the Pancreas." 
 
 By RICHARD M. PEARCE, M. D, 
 Director ot the Bender Laboratory, Albany, N. Y. 
 
 The short space of time allotted me in this discussion will 
 not allow a comprehensive statement of the many fruitful 
 experimental researches having for their object the elucida- 
 tion of the physiology and pathology of the pancreas. 
 Various phases of the experimental investigations will, doubt- 
 less, be taken up by others who address you this evening; 
 and I have therefore, with the approval of your president, 
 decided to limit my remarks to a consideration of the acute 
 lesions which may be produced experimentally. What we 
 have gained by experiment in this field is practically a 
 knowledge of etiology, and it is therefore more particularly 
 the question of etiology that I will discuss. I will not hesi- 
 tate moreover to compare the experimental lesions with the 
 acute lesions seen in man, with the object as far as possible 
 of correlating the knowledge gained by experiment and that 
 gained by pathological study. 
 
 In any consideration* of the acute forms of pancreatitis we 
 are confronted at the outset with the question of classifica- 
 tion. The division into haemorrhagic, suppurative and 
 gangrenous forms as made by Fitz would appear to satisfy 
 all demands. Some writers, however, speak of a necrotic 
 form. This is of doubtful value, for all forms show more or 
 less necrosis; indeed, in the haemorrhagic form, especially 
 that produced experimentally, necrosis is a constant and 
 essential part of the process, though not always evident 
 macroscopically. The possibility of pancreatic haemorrhage 
 without inflammation — the so-called pancreatic apoplexy — 
 
 * In the preparation of this brief summary I haye freely used the recently published work 
 of Opie (Dleeaies of the Pancreas, Its Cause and Nature, J. B. Llppincott Co., 1903) and Flex- 
 ner's description of the pathological anatomy of the pancreas (Transactions of the Congress of 
 American Physicians and Surgeons, 1903, vi, 18); and I take this opportunity to acknowledge 
 my indebtedness. I have not attached a bibliography. All references may be found in the 
 bibliography of Ople's work. 
 
 Digitized by Microsoft® 
 
is another debatable question which, in the present state of 
 our knowledge cannot be definitely settled. It seems proba- 
 ble from our present point of view, that thorough histological 
 examinations of such cases would furnish evidence of inflam- 
 matory changes and bring this lesion under the head of 
 haemorrhagic pancreatitis. Seitz, the last writer (1892) to 
 make a careful study of this lesion is, however, opposed to 
 this view and believes that some cases, at least, may be 
 attributed to arterio-sclerosis. 
 
 hcemorrhagic Pancreatitis. — The gross appearance of this 
 form of pancreatitis — an enlarged organ with diffuse or 
 mottled haemorrhagic areas of varying color and greyish- 
 white spots or streaks of fat necrosis — is so well known that 
 we may turn at once to the question of etiology. Our 
 knowledge of the causes of pancreatitis is entirely the result 
 of experiments having for their object the reproduction of 
 the condition. Clinical observations have contributed nothing, 
 and with the exception of Opie's observations upon the 
 relation of cholelithiasis, the same may be said of general 
 pathological examinations. The simplest form of experi- 
 mentation is the injection into the pancreatic duct of various 
 irritating or supposedly injurious substances. A great variety 
 of substances thus introduced have been found to produce 
 typical haemorrhagic pancreatitis, usually with fat necrosis. 
 These substances are deliquescent zinc chloride (Thiroloix), 
 artificial gastric juice (Hlava, Flexner and Pearce) bile 
 (Opie, Flexner and Pearce), papaine (Carnot), and bacteria 
 and their toxins (Hlava, Carnot, Flexner). Various other 
 substances as formalin, nitric and chromic acids, and solution 
 of caustic alkalies, Flexner found would produce the same 
 results. The disease produced by these various agents 
 usually runs a rapidly fatal course, though some animals 
 may live for weeks and present at death extensive suppura- 
 tive or sclerotic changes. In some instances putrefactive 
 micro-organisms invade the injured organ and cause a 
 gangrenous pancreatitis. If we study these various methods 
 with the view of determining if any of them may operate to 
 cause the haemorrhagic pancreatitis seen in man, we have 
 four possibilities: action of (a) ferments originating in the 
 pancreas, (^) bacteria, (c) gastric juice, and {d) bile. 
 
 The possibility of the first is suggested by Carnot's experi- 
 ments with papaine, and some support is offered by Chjari's 
 
 Digitized by Microsoft® 
 
observations upon autodigestion of the pancreas. According 
 to Chiari this change occurs shortly before death and may be 
 accompanied by multiple small haemorrhages. The destruc- 
 tion of the parenchyma thus brought about affords an 
 opportunity for the diffusion of the pancreatic ferments. 
 There is some evidence that fat necrosis may be caused by 
 the fat splitting ferment thus liberated, but as yet we have no 
 definite proof that a true hemorrhagic pancreatitis may result 
 from the action of this or other ferments liberated at the 
 same time. 
 
 Various micro-organisms, as the bacillus coli, the pneu- 
 mococcus and the pyogenic cocci have been found, not 
 infrequently, associated with hsemorrhagic pancreatitis. 
 There has been no uniformity in these findings and most of 
 the writers reporting the results of bacteriological examina- 
 tions agree that the organisms represent a secondary invasion 
 of injured tissue and are in no way concerned in the etiology 
 of the lesion. 
 
 That the gastric juice, as suggested by Hlava may enter 
 the pancreatic duct and produce an inflammation of the 
 pancreas presupposes a powerful anli-peristaltic, or other 
 action, forcing the secretion into the duct. Of such action 
 we have no definite evidence. Under normal conditions it is 
 practically impossible to force fluids from the intestine into 
 either bile or pancreatic ducts. I have recently made some 
 experiments upon dogs in order to determine this point. The 
 intestine was ligated at varying distances above and below 
 the orifice of the ducts and distended by injecting colored 
 solutions and suspensions. After varying periods of time, 
 the animals were killed, but no staining of the ducts could be 
 demonstrated. In other experiments upon the dead animal, 
 forcible manual pressure was exerted upon the similarly 
 ligated intestine with negative results. On the other hand 
 it cannot be denied that under certain pathological conditions 
 of the duct orifice, intestinal fluids may possibly be forced 
 into the duct. Concerning the negative action of the natural 
 gastric juice upon pancreas parenchyma the following case 
 would appear conclusive ; 
 
 (Records of the Pathological Department of the University of Penn- 
 sylvania; Autopsy No. 0.148.) The clinical history of this case is brief 
 and incomplete. The patient, a negro, aged forty-one, was admitted 
 
 Digitized by Microsoft® 
 
to the Insane Department of the Philadelphia Hospital on July 19, 
 1900, with a diagnosis of delusional insanity. For the next six months 
 his condition, apparently, was not noteworthy. "On January 28, 1901, 
 patient seemed to be rather dull and was put to bed." On February 
 nth, regurgitation of blood mixed with milk curds was observed. 
 This continued until his death, on February 13th. Physical examina- 
 tions revealed a rather firm mass in right hypochondriac region, which, 
 upon palpation caused patient to complain of pain. At no other time 
 did he complain of pain. Temperature normal, pulse about 100 and 
 of poor volume; no distention of abdomen; marked constipation. 
 Albumen was found in the urine; and microscopically, numerous hya- 
 lin and granular casts. The clinical diagnosis was cancer of the 
 stomach and the liver with chronic nephritis. 
 
 Autopsy — Pathological diagnosis: Recent haemorrhage in and about 
 pancreas and duodenum, chronic gastro-pancreatic fistula; chronic 
 interstitial pancreatitis; chronic pulmonary tuberculosis; tubercolosis 
 of bronchial, gastro-hepatic and retro-peritoneal lymph nodes; chronic 
 mitral endocarditis; general chronic passive congestion; chronic dif- 
 fuse nephritis (parenchymatous type). 
 
 Abdominal cavity: Stomach greatly dilated, the lower margin ex- 
 tending below the umbilicus ; pylorus touching right costal margin. 
 Beneath the pylorus and entirely surrounding the duodenum is a 
 globular, firm, haemorrhagic mass, distinctly encapsulated and resting 
 on the upper end of the right kidney. The subperitoneal tissues 
 between the stomach and the transverse colon and in the neighbor- 
 hood of pylorus and adjacent abdominal wall are infiltrated with 
 clotted blood. Stomach, pancreas, duodenum and transverse colon 
 with haemorrhagic mass removed entire. Upon opening the stomach 
 it is found to be normal except for the presence of a small orifice 
 with smooth edges four centimeters from the oesophageal end. The 
 outer wall of the stomach about orifice is bound to the pancreas 
 nine centimeters from its splenic end by dense fibrous adhesions. 
 A proBe passed through the orifice in the stomach enters a sinus 
 leading to the pancreas. This sinus passes directly through the pan- 
 creatic tissue. Its walls are lined by dense fibrous tissue, and it con- 
 tains blood-stained mucus-like fluid. The canal can be followed for 
 a distance of five centimeters. It then approaches the surface of the 
 pancreas where it is continuous with the blood clot about the duo- 
 denum. No foreign body is found in the sinus. The splenic end of 
 the pancreas for a distance of eight centimeters appears normal; the 
 middle portion for a distance of four and a half centimeters is 
 paler and firmer than normal, and the lobulations are not so distinct; 
 through the remaining five centimeters constituting the duodenal end, 
 the pancreas is extensively infiltrated with blood. Surrounding the 
 duodenal end of the pancreas and completely encircling the duode- 
 num is an encapsulated mass of black clotted blood fourteen centi- 
 meters in length and seven centimeters in diameter. The long diam- 
 eter of this clot corresponds to the course of the duodenum. The 
 source of the haemorrhage cannot be discovered. The pancreatic and 
 common bile ducts are patulous. The mucous membrane of the duo- 
 
 Digitized by Microsoft® 
 
denum is very thin, blood-stained, and at one point presents a small 
 adherent blood clot two centimeters in diameter. No <at necrosis. 
 Histological examination: The wall of sinus is composed of dense 
 fibrous tissue. On its surface is a thin layer of granular material with 
 a few pus cells. Sections stained by the Gram-Weigert method show a 
 few cocci and bacilli. Scattered through superficial portion of sinus 
 wall are many lymphoid and plasma cells and a few polymorpho- 
 nuclear leucocytes. New blood vessels are seen throughout, and here 
 and there are small areas of haemorrhage. In the deeper portions 
 the haemorrhage is more diffuse, and here are seen small islands of 
 atrophic panr.reatic tissue infiltrated by lymphoid cells. From the 
 sinus wall coarse bands of fibrous tissue run into the substance of the 
 pancreas, and divide into finer lines which penetrate the pancreatic 
 acini. Sections from various portions of the pancreas, at a distance 
 from the sinus, show a moderate increase of perilobular connective 
 tissue. The cells of the parenchyma are very granular and stain 
 poorly. The islands of Langerhans are prominent and unaltered. 
 Throughout the duodenal portion extensive haemorrhage is seen in 
 the interstitial tissue, penetrating in places the parenchyma which is 
 oedematous. No necrosis or exudation is evident. The splenic por- 
 tion is free from haemorrhage. 
 
 From this description it is evident that in the early 
 development of the sinus the gastric juice must have been in 
 intimate contact with the pancreas parenchyma but failed to 
 produce a haemorrhagic pancreatitis. The hsemorrhage 
 into the duodenal portion, was unaccompanied by acute 
 changes and was due apparently to extension from the large 
 clot about duodenum. The chronic pancreatitis was merely 
 a localized process about the sinus. The case offers strong 
 evidence against the possibility of gastric juice exciting a 
 typical haemorrhagic pancreatitis. 
 
 The relation of bile to acute pancreatitis has received very 
 substantial support from experimental observations and also 
 from the not unusual occurrence of the lesion with 
 cholelithiasis. The most important contributions to this 
 phase of the subject have been made by Opie, who has 
 worked out the pathogenesis of pancreatitis associated with 
 gall-stones and by injecting bile into the pancreatic duct of 
 the dog has produced a typical inflammation of the gland 
 with associated fat necrosis. In one particularly fortunate 
 autopsy he found at the apex of the diverticulum of Vater 
 a small calculus completely blocking the duodenal orifice but 
 too small to obstruct the orifices of the common bile and the 
 pancreatic duct. These ducts were thus transformed into a 
 continuous closed channel from which it was impossible for 
 
 Digitized by Microsoft® 
 
their respective contents to escape. Upon dissecting the 
 pancreatic duct it was found to be stained light green 
 with bile, indicating that the slightly greater pressure in 
 the bile duct, probably due to the contraction of the 
 gall-bladder, had forced bile into it. An acute haemor- 
 rhagic pancreatitis with disseminated abdominal fat necrosis 
 existed. 
 
 The same writer has also analyzed the reported cases of 
 acute pancreatitis and finds that cholelithiasis is very con- 
 stantly associated with the haemorrhagic and gangrenous 
 forms of the lesion. On the other hand, acute pancreatitis 
 is a rare condition as compared with the relative frequency 
 of cholelithiasis. In this connection it is important to point 
 out, as Opie has done, that in only a small proportion of 
 individuals (in about three out of ten) are the anatomical 
 features such as to allow favorable conditions for bile to 
 penetrate the pancreatic duct. The ducts, in a considerable 
 number of individuals (ten per cent.) open separately and 
 the diverticulum does not exist. Also, the size of the 
 diverticulum bears an important relation to the production 
 of the condition; its variation in length is from zero to 
 eleven millimeters; the average diameter of the orifice being 
 two and a half millimeters. In only thirty of Opie's lOO 
 examinations did the length reach five millimeters. Thus 
 it is evident that a stone to produce obstruction of the 
 duodenal orifice only, must be relatively smaller than the 
 diverticulum, for otherwise the orifices of both ducts would 
 also be obstructed and the secretions forced back upon their 
 respective organs. 
 
 While cholelithiasis is thus definitely established to be an 
 important factor in the etiology of acute pancreatitis, it does 
 not explain all cases. Many cases have been reported in 
 which neither clinical history nor autopsy affords evidence of 
 cholelithiasis, though it is noteworthy that since attention has 
 been called to this association these cases are not so numer- 
 ous. It should also be noted in this connection that the 
 stone after obstructing the papilla orifice and causing a 
 pancreatitis may pass into the intestine and escape ob- 
 servation, or perhaps pass from the intestine entirely 
 A dilatation of the diverticulum may in such cases 
 indicate the previous resting place of the stone. The 
 following case would appear to be an example of this kind: 
 
 6 
 
 Digitized by Microsoft® 
 
Clinical abstract: Patient of Dr. W. E. Fay, of Boston, Mass.— Robust 
 woman, aged thirty-six, with history of occasional sharp attacks of gastro- 
 enteritis after excessive indulgence in wine. "The mortal attack began with 
 a 'French' dinner on Thanksgiving Day. From claret she changed to whis- 
 key, drinking considerable between November 29th and December 8th. When 
 first seen on the latter of these dates she was tremulous and sleepless, with 
 wild look in the eyes, intense burning in epigastrium and violent and 
 continuous emesis. Neither water nor cracked ice. could be retained. 
 The vomitus was green, thin and acid. After hypodermic use of 
 morphine and careful administration of chloroform, almost steadily 
 for an hour, sleep was obtained lasting three hours. Restlessness 
 then returned, with frequent hiccough, and ceaseless cry for water, 
 which was rejected in whatever quantity or temperature given. Wash- 
 ing of stomach, and paraldehyde administered by enema afforded tem- 
 porary relief. Abdomen was distended. Obstinate constipation was 
 overcome by high enemata. Persistent flatulence was present in both 
 stomach and bowel, and considerable pain centering in the epigastrium 
 and extending throughout the abdomen. This condition continued 
 with temporary alleviations for nine days after she was first seen. 
 Then, December 17th, slight jaundice was noticed. The patient 
 desired nothing but frequent drinks of milk and limewater. Decem- 
 ber 20th the patient sitting up and begins to eat gruel, but has no energy. 
 December 24th, sudden enlargement of right parotild and adjoining glands 
 with redness extending below angle of jaw; considerable emaciation 
 has taken place. After sitting up parts of two or three days, which 
 was agreeable to her, there was increase of tenderness throughout 
 the abdomen. Temperature rose to loi degrees. Distention was 
 constant, free fat appeared in the urine; sugar was present. The patient 
 sank rapidly and died on the 30th." 
 
 Autopsy, performed six and one-half hours after death. 
 
 Anatomical diagnosis: Hsemorrhagic and gangrenous pancreatitis 
 with fibrinous peritonitis and accumulation of grumous fluid in peri- 
 toneal cavity. Multiple fat necroses of omentum, mesentery, epiploic 
 appendages and subperitoneal fat generally. Acute duodenitis with 
 dilation of orifice of duodenal papilla. Localized dilatation of com- 
 mon bile duct. Dilatation of pancreatic duct. Fatty degeneration of 
 heart, liver and kidneys. Abdomen — Excessive amount of subcu- 
 taneous fat. Omentum, mesentery, epiploic appendages and abdominal 
 fat generally closely studded with round or oval greyish-white areas 
 two to five millimeters in diameter and distinctly circumscribed. Many 
 of the greatly enlarged epiploic appendages (3 x 1.5 cm.) are entirely 
 necrotic. In pelvis and both flanks is a thin greyish-red greasy fluid 
 containing small black shreds of friable material. Omentum is glued 
 to anterior abdominal wall by a blood-stained fibrinous exudate. A 
 similar exudate covers the coils of the small intestine. The position 
 of the pancreas is occupied by a soft friable, irregular reddish-black 
 mass 4 X i.S centimeters, freely moveable. The cavity in which this 
 mass lies is lined by reddish-black necrotic tissue, which adhere in 
 places to colon and stomach. The mass on section shows at splenic 
 end irregular areas of greyish pancreatic tissue separated by dark red 
 
 Digitized by Microsoft® 
 
lines of haemorrhage and necrosis. Elsewhere it is uniformly necrotic. 
 Duodenum — Contains material similar to that about pancreas and its mucosa 
 is injected. The orifice of papilla is dilated and in the duct one centimeter 
 from the orifice occurs a circular dilatation with smooth walls, one 
 and five-tenths centimeters in circumference. No calculi are found in 
 gall-bladder, ducts or intestine. The pancreatic duct is dilated and 
 separated from the pancreas; a probe from duodenal end passing 
 directly into the cavity about the gangrenous pancreas. Bacillus coli was 
 found in the necrotic material about the pancreas. 
 
 While this case is not conclusive, in as much as the stone 
 was not found, the anatomical condition indicates that a 
 calculus, at some time, was lodged in the common duct and 
 the possibility of later occlusion of the papilla with final dis- 
 charge of the stone is evident. 
 
 The Sequence of Histological Changes in Heemorrhagic Pan- 
 creatitis. — Much discussion has arisen concerning the prece- 
 dence and succession of the three principal phenomena — 
 hemorrhage, necrosis and exudation. Katz and Winkler 
 believe that the necrosis follows haemorrhage and is the 
 direct result of injury produced by the extravasated blood. 
 Hlava found extensive occlusion of the blood vessels after 
 acid injection and considered it to be the cause of both 
 necrosis and haemorrhage. Opie found haemorrhage inti- 
 mately associated with necrosis and followed only secondarily 
 by inflammatory changes. From his earlier experiments 
 Flexner drew the conclusion that haemorrhage and necrosis 
 are the result of the simultaneous action of the injurious 
 agent upon blood vessels and tissue, and that they occur 
 coincidentally. In a later series of experiments by Dr. 
 Flexner and myself an attempt was made to definitely settle 
 these mooted points. Dogs under choretone narcosis 
 received injections of one to two per cent, solutions of 
 hydrochloric acid into the duct and directly into the paren- 
 chyma of the pancreas. Small portions of the organ were 
 removed after periods varying from thirty-five minutes to 
 eight hours and examined histologically. It was found that 
 haemorrhage, hyalin necrosis and exudation of fluid and 
 leucocytes could be readily demonstrated in tissue removed 
 one hour after injection. Sections taken at the end of eight 
 hours represented the pathological process at its height. 
 That these lesions are correlated in the pancreas may there- 
 fore be considered as demonstrated. 
 
 At the same time in order to determine the effect of red 
 
 8 
 
 Digitized by Microsoft® 
 
blood corpuscles upon the pancreatic tissue, a condition 
 analagous to haemorrhage was brought about by the injection 
 of sterile defibrinated dog's blood through the duct and 
 directly into the parenchyma. Such injections did not set up 
 an acute pancreatitis but were followed by chronic proliferative 
 changes. Similar injections of dog's serum were negative. 
 
 Gangrenous Pancreatitis. — Experimental and post-mortem 
 observations both indicate that in the majority of cases 
 gangrenous pancreatitis is secondary to an acute haemor- 
 rhaglc inflammation of the organ. The tissue, primarily the 
 seat of haemorrhage, necrosis and inflammation readily 
 becomes gangrenous under the influence of putrefactive 
 micro-organisms invading it from the intestinal canal. The 
 association of hemorrhage and gangrene frequently seen at 
 autopsy was first emphasized by Fitz, who also recognized 
 however the possibility of the lesion originating from per- 
 forating inflammation of the gastro-intestinal canal or of the 
 biliary passages. 
 
 In the series of experiments by Dr. Flexner and myself, 
 two examples of gangrenous pancreatitis following acid and 
 bile injections respectively, occurred. Death occurred in 
 each instance in less than twenty-four hours, and the gan- 
 grenous portions of the pancreas were studded with 
 emphysematous blebs. Films showed bacilli having the 
 morphology of B. aerogenes capsulatus. Futhermore, gan- 
 grenous pancreatitis has the same relation to cholelithiasis as 
 has the haemorrhagic form. In Opie's table of thirty-seven 
 cases of acute pancreatitis associated with cholelithiasis, eight 
 are described as haemorrhagic and gangrenous and thirteen 
 as gangrenous. Disseminated fat necrosis is a striking 
 feature of this lesion. 
 
 Suppurative Pancreatitis. — Suppurative inflammation of the 
 pancreas, unlike the haemorrhagic and gangrenous forms, 
 does not differ essentially from that of other organs. It may 
 be primary or secondary. The relation of the pancreas to 
 the intestine allows the possibility of infection through the 
 duct, while lesions of adjacent hollow viscera may extend 
 directly or as the result of perforation. In general infection 
 of the blood, pyaemic abscesses may develop in the pancreas 
 as in other organs. Direct extension of an infectious 
 pylephlebitis may also occur. These lesions are seldom 
 accompanied by extensive necrosis, haemorrhage or dissemi- 
 
 9 • . 
 
 Digitized by Microsoft® 
 
nated fat necrosis. The accumulation of pus is not diffuse 
 but occurs usually as localized abscesses. Haemorrhage 
 and necrosis if present are generally focal in character, sug- 
 gesting the local action of an irritant. Rarely the pancreas 
 may be entirely destroyed and appear as a sequestrum in a 
 pus cavity. Experiments, especially those of Flexner, show 
 that abscesses of the pancreas may follow the injection of 
 various fluids, as weak solutions of hydrochloric, sulphuric 
 and nitric acids, sodium hydrate and suspensions of the B. 
 pyocyaneus. In these experiments we have evidently a 
 primary necrosis due to the irritant and a secondary invasion 
 by micro-organisms. Korte, in a few instances, produced 
 abscesses by injecting pus containing a bacillus of the colon 
 group. These experimental lesions are seldom accompanied 
 by fat necrosis. A study of the lesion in man and of 
 the experimental findings indicates that this suppurative form 
 of pancreatitis is due to some local influence causing limited 
 necrosis and suppuration. There is, however, some evidence 
 that duct obstruction may rarely lead to the condition 
 probably in that it favors the entrance of bacteria. In 
 Opie's table, previously quoted, "purulent pancreatitis" 
 occurred in four of forty-one cases complicated by 
 cholelithaisis. The common duct in only one of these, how- 
 ever, contained a stone. The following case illustrates 
 suppuration of the pancreas with fat necrosis, accompanying 
 cancer of the diverticulum and cholelithiasis: 
 
 Records of the Bender Laboratory, Autopsy No. 0.676. Clinical abstract: 
 The individual, a male aged seventy-three, patient of Dr. A. W. 
 Elting, was admitted to the Albany Hospital on August 24, 1903, with 
 a history of hepatic colic which began on August isth. Since May ist 
 he had suffered from gastric distress, nausea and dull pain in region 
 of stomach, and had lost in strength and weight. Physical examination 
 reveals a jaundiced, emaciated, feeble individual with evidence of 
 arterio-sclerosis, enlarged heart and emphysema of the lungs. Upon pal- 
 pation of abdomen a distinct irregular tumor mass was felt in the 
 epigastrium; deep pressure elicited some tenderness. During patient's 
 sojourn in the hospital there was slight change in his condition. The 
 jaundice continued and nausea after the taking of food persisted. 
 The pain subsided. Physical examination revealed nothing new. 
 Emaciation and weakness' progressed rapidly, and death occurred on 
 September nth. 
 
 Autopsy, by Dr. A. T. Laird, twelve hours after death. 
 
 Anatomical diagnosis: Cancer of the diverticulum of Vater with 
 ulceration of duodenum and metastases in liver; dilatation of common 
 duct and gall-bladder; cholangitis, gallstones free in colon; jaundice; 
 
 10 
 
 Digitized by Microsoft® 
 
multiple abscesses of head of pancreas ; peripancreatic fat necrosis ; haem- 
 orrhage into intestine; arterio-sclerosis. 
 
 The individual is extremely emaciated and jaundiced. In duodenum, 
 about one and a half centimeters above duodenal papilla, is a cir- 
 cular ulcer three .centimeters in diameter with irregular edges and 
 firm indurated bile-stained base. The base of the ulcer, five milli- 
 meters in thickness, rests upon the head of the pancreas to which it 
 is intimately adherent and upon indurated lymph nodes. The tissues 
 beneath the papilla are infiltrated by a dense greyish-white diffuse 
 growth of new tissue. The duodenal papilla is prominent and rests 
 on the indurated mass below it; bile readily flows from its orifice on 
 forcible pressure of gall-bladder. Common duct is slightly dilated. 
 Gall-bladder measures ten centimeters in length and is distended with 
 thick fluid bile, no calculi present. Liver is of a dark uniform green- 
 ish color, and bile ducts are distended with thick bile. Scattered 
 throughout the organ are numerous greyish-white firm areas, one- 
 half to two centimeters in diameter, of fairly definite outline. 
 Pancreas measures fourteen centimeters in length. The duodenal 
 end for a distance of five centimeters is irregularly swollen. On sec- 
 tion a yellowish purulent fluid escapes and the cut surface presents 
 numerous abscesses of varying size with thin trabeculated walls. The 
 middle portion of the organ is firmer than normal. Splenic end is 
 unaltered. Duct patent though its proximal portion is surrounded by 
 the new growth. Peripancreatic fat presents numerous distinctly 
 circumscribed yellowish opaque areas. The small intestine is blood- 
 stained throughout its length and contains an accumulation of blood, 
 fluid and clotted. A lesser amount of blood is found in the large 
 intestine. The source of the haemorrhage is apparently the ulcer in 
 the duodenum. In the lower portion of the discending colon are 
 found two smooth facetted calculi composed of cholesterin and inspis- 
 stated bile. Histological examination shows a cancer about and 
 involving the diverticulum and the common duct with extension to 
 the duodenum and to the head of the pancreas and with metastases 
 in the liver. The head of the pancreas is the seat of advanced peri- 
 lobular pancreatitis, in the midst of which are numerous abscesses. 
 These abscesses are chronic in character, and contain much necrotic 
 material and zoogloea of large thick bacilli. The tail of the pancreas 
 is the seat of a slight perilobular pancreatitis. Typical fat necroses 
 are numerous. The ducts are dilated and contain an accumulation of 
 mucous and desquamated epithelial cells. 
 
 Fai Necrosis. — This condition is so constantly a comcomi- 
 t^nt of acute pancreatitis and its elucidation so entirely the 
 result of experimentation that a consideration of the experi- 
 mental lesions of the pancreas would be incomplete without 
 reference to it. Haemorrhagic and gangrenous pancreatitis, 
 as they occur in man or as the result of experiment, 
 are generally accompanied by this peculiar lesion "of the 
 fat tissue. Rarely it occurs in connection with suppura- 
 
 II 
 
 Digitized by Microsoft® 
 
tive lesions, and occasionally, as will be pointed out later, 
 with chronic interstitial pancreatitis. The characteristic 
 appearance is that of small round or oval areas white 
 or greyish-white in color and usually sharply circum- 
 scribed. Microscopic examination shows these areas to 
 be composed of necrotic fat cells. In early necroses fatty 
 acid crystals may be demonstrated. These areas are found 
 most abundantly in the neighborhood of the pancreas and 
 are usually limited to the abdominal fat, but they may be 
 widely disseminated and appear in the fat beneath the peri- 
 cardium and the pleura, in the subcutaneous fat and in the 
 marrow of long bones. Fitz first suggested that fat necrosis 
 is the result of a lesion of the pancreas, but to Langerhans 
 belongs the credit of first demonstrating the essential nature 
 of the process. By chemical and histological studies he 
 showed that the changes within the fat cells are associated 
 with the splitting of the fat molecule into its fatty acid and 
 glycerine. The glycerine is absorbed and the acids are 
 deposited in the cells as needle-like crystals. The cell loses 
 its nucleus and is apparently necrotic. Later, chronic pro- 
 liferative changes occur about the areas. Leucocytes may 
 enter the necrotic fat but when numerous, indicate invasion 
 by bacteria. Although bacteria have been frequently found 
 in these necroses they are not concerned in the production 
 of the lesion and must be considered as secondary invaders. 
 Instances in which the injection of bacteria into the pan- 
 creatic duct has resulted in the necrosis of fat are to be 
 explained by the setting free of pancreatic ferments as the 
 result of injury to or inflammation of the gland. 
 
 The first experimental proof of the relation of fat necrosis 
 to pancreatic ferments was offered by Langerhans, who 
 injected suspensions of the comminuted pancreas of the rabbit 
 and dog into the perineal fat of the rabbit. One of his 
 thirteen experiments was successful and he concluded that 
 these necroses are due to the action of the fat splitting fer- 
 ment of the pancreatic juice. Jung obtained similar results 
 by introducing portions of the fresh pancreas of one animal 
 into the adbominal cavity of another; Hildebrand and 
 Dettmer, Williams, Whitney, Flexner, Milisch, Korte and 
 Opie have obtained similar results by various injuries allow- 
 ing the escape of the pancreatic juice into the adjacent fat 
 and the peritoneum. Their methods include ligation of 
 
 Digitized by Microsoft® 
 
ducts and vessels; ligation, laceration and incision of the 
 parenchyma; and excision of portions of the organ. The 
 experiments of Hlava, Korte, Oser, Opie, Flexner and 
 Pearce, in which an acute pancreatitis, usually haemorrhagic, 
 has followed the injection of various substances as acids, 
 alkalies, gastric juice, bile, bacteria, etc., have also caused 
 the appearance of fat necrosis. In these experiments the 
 fat necroses are not due to the substance injected but to 
 lesions of the pancreas which allow an escape of the pan- 
 creatic juice. The final link in this chain of experimental 
 evidence was Flexner's demonstration, by chemical test, of 
 the presence of a fat splitting ferment in the necrotic foci. 
 Portions of the necrosed foci, from both human and experi- 
 mental cases, \yhen brought in contact with neutral butter fat 
 caused a splitting of the fat with the freeing of fatty acids 
 recognizable by their odor and reaction. 
 
 It has been claimed by some writers, especially by Oser, 
 that the presence of a fat splitting ferment will not explain 
 the widely disseminated necroses, such as occur in the sub- 
 peric^rdial, sub-pleural, and sub-cutaneous fat. Opie has, 
 however, in a series of duct ligations demonstrated by Flex- 
 ner's method the presence of the fat splitting ferment in the 
 distant necroses of the subcutaneous and subpericardial fat 
 as well as in the peri-pancreatic foci. The diffusion of the 
 ferment in these experiments he also found was hastened 
 when the secretory activity of the pancreas was stimulated by 
 injection of pilocarpine. 
 
 Wells {Journal of Medical Research, iQOj, ix., 70) has 
 found that commercial "pancreatin" will produce typical fat 
 necrosis, but that this power is lost after heating for five min- 
 utes at some point between sixty-five and seventy-one degrees 
 centigrade. Its power to produce fat necrosis decreases 
 proportionally as the temperature is raised above fifty-five 
 degrees Centigrade. These observations indicate an enzyme 
 action as the source of the lesion. He was unable, however, 
 to determine which of the pancreatic enzymes causes fat 
 necrosis, and could not produce fat necroses with the lipolytic 
 extracts of hog's liver or cat's serum used either alone or in 
 combination with trypsin. He thinks it possible that the 
 lipase causes fat splitting after some other ingredient of the 
 pancreatic juice has injured the cell; fat splitting is subse- 
 quent to the necrosis and not its cause. The lipase normally 
 
 13 
 
 Digitized by Microsoft® < 
 
in fat tissues exerts no fat splitting effect upon necrotic fat 
 if the latter be preserved from external influences. Foci of 
 fat necrosis became visible to the naked eye three to five 
 hours after the application of "pancreatin" and were found out- 
 side the abdominal cavity as early as twelve hours after intra- 
 peritoneal injection. 
 
 In considering the application of the results of experi- 
 mental fat necrosis to the lesion as seen in man, it must be 
 borne in mind that fat necrosis bears the same relation to 
 obstruction of the pancreatic ducts as jaundice does to 
 hepatic obstruction. Obstruction to the outflow of the 
 pancreatic secretion causes its diffusion into the surrounding 
 tissues (Opie). The escape of the ferment in the acute 
 lesions of the pancreas is readily explained by obstruction or 
 destruction of the finer ducts or by injury to the parenchyma. 
 On the other hand either temporary or permanent obstruction 
 of the duct of Wirsung by calculi or by cancer may be 
 accompanied by fat necrosis. Usually other lesions exist as 
 suppurative or chronic pancreatitis but in a few cases, as 
 those reported by Fraenkel (stone in diverticulum) and by 
 Flexner (stones in common duct) no lesion of the pancreas 
 was found. There can be little doubt that in these cases the 
 calculus causes compression of the pancreatic duct and an 
 accumulation of secretion which finally becomes diffused into 
 the adjacent tissues. 
 
 The following case would appear to belong in this group. 
 A chronic al:rophic cholecystitis with atrophy of the bile 
 ducts and adhesions to head of pancreas apparently caused 
 pressure upon the pancreatic duct. 
 
 Records of the Bender Laboratory, Autopsy No. 0.614. Autopsy by 
 Dr. A. T. Laird, seven and one-half hours after death; female, aged 
 forty-two. 
 
 Anatomical diagnosis: Extensive syphilitic ulceration of nose and 
 palate, oedema of lungs, hydrothorax, chronic cholecystitis with almost 
 complete obliteration of gall-bladder and ducts, fat necrosis of pan- 
 creas. The gall-bladder is greatly atrophied and lies in a mass of 
 dense fibrous adhesions which bind it to the duodenum and the head 
 of the pancreas, and which also surround and penetrate the foramen of 
 Winslow. The cavity of the gall-bladder measures but two centi- 
 meters in diameter. The pancreas measures 17 x 4 x 2 centimeters. 
 At its head, the capsule is much thickened by adhesions binding it 
 to the gall-bladder and to the duodenum. Over its entire surface 
 numerous greyish-yellow opaque areas, not extending into pancreatic 
 tissue, are seen. The consistence of the pancreas is normal. 
 
 ^4 
 
 Digitized by Microsoft® 
 
Histological examination of the pancreas reveals numerous fat necro- 
 ses, and dilated ducts filled with hyaline and granular material, desqua- 
 mated cells and a few polymorphonuclear leucocytes. The latter also, 
 to a moderate degree, infiltrate the walls of the larger ducts. No inflamma- 
 tion of the parenchyma proper is evident. The islands of Langerhans 
 are unaltered. 
 
 The following accidental finding in a dog indicates that fat 
 necrosis may follow a simple thickening of the orifice of the 
 duct: 
 
 "Dog 84— weight 8650 gramms.— Isolated February 6th. — Urine nor- 
 mal on February 7th and 8th. Killed on February 12th for purpose 
 of obtaining organs for injections in connection with cytolytic work. 
 All organs normal except pancreas, the capsule of which is studded 
 with numerous small greyish-yellow areas of definite outline and of 
 firmer consistence than surrounding pancreatic tissue. No evidence 
 of acute or chronic inflammation. No fat necroses of abdominal fat 
 elsewhere. No obstruction of the duct is found. The papilla of duo- 
 denum, however, appears as a distinctly elevated, swollen, greyish- 
 yellow ring; a sharp contrast to the normal appearance. A fine probe 
 passes readily into the duct. The second pancreatic duct is not found. 
 The orifice of bile duct, lying one and a half centimeters above 
 pancreatic duct, can be located only by forcing bile from the gall- 
 bladder. Histological examination shows fat necroses, chronic in 
 character; pancreas otherwise normal." 
 
 Finally, as pointed out by Opie, we have fat necroses asso- 
 ciated with cases of chronic intestitial pancreatitis in which 
 duct obstruction by calculi is not present. Here it is proba- 
 ble that the newly formed fibrous tissues contracts about 
 the smaller ducts, constricts them, and allows an accumula- 
 tion of secretion which finally diffuses into the adjacent 
 tissues. These necroses are most apt to be parapancreatic. 
 I have found two examples of this condition in the autopsy 
 reports of the Bender Laboratory; in one associated with 
 localized chronic interstitial pancreatitis, in the other with 
 atrophy and lipomatosis. Macroscopic evidence of duct 
 obstruction did not exist. These reports are as follows: 
 
 No. 0.223. Autopsy twelve hours after death. Male, aged forty nine. 
 
 Anatomical diagnosis: Acute follicular duodenitis and ileitis; acute 
 swelling of spleen and mesenteric lymph nodes; fatty degeneration of 
 liver; arterio-sclerosis; atrophy of pancreas with lipomatosis, auto- 
 digestion and foci of fat necrosis; general infection with the bacillus 
 of Friedlander. The pancreas appears to be of normal size, but much 
 of the parenchyma is replaced by diffuse masses of fat tissue. On 
 section small islands of pancreas are seen widely separated by fat 
 tissue. This change is most marked at splenic end where little pan- 
 creatic tissue remains. The parenchyma in some areas is whiter 
 
 •5 
 
 Digitized by Microsoft® 
 
and softer than normal, and in other places presents a reddish-brown 
 appearance, suggesting minute haemorrhages. On the surface and on 
 section several, minute, opaque white points suggesting fat necrosis 
 are seen. No calculi are found in the ducts; the vessels present a few 
 small areas of atheroma. 
 
 Histological examination shows small islands of pancreas widely 
 separated by adipose tissue, slight self-digestion of these areas, and 
 typical foci of fat necrosis with minute haemorrhages. The inter- 
 lobular connective tissue is increased in amount; the small ducts 
 dilated and some acini appear as duct-like spaces with flattened epi- 
 thelium and contain a finely granular material. In some islands these 
 altered acini are very abundant and apparently indicate atrophy from 
 pressure within the lumen. The islands of Langerhans are unaltered. 
 
 No. 0.382. Autopsy ten hours aftet death. Male, aged sixty three. 
 
 Anatomical diagnosis: Arterio sclerosis, hypertrophy and dilatation 
 of heart; general passive congestion; oedema; double hydrothorax; 
 thrombus of right auricular appendage; multiple infarcts of lungs; 
 fat necrosis of pancreas. Pancreas twenty centimeters long. Twelve 
 centimeters from the tail, over an area two and a half centimeters 
 in diameter, the capsule is greatly thickened; beneath the capsule the 
 pancreas presents a congested area containing many small yellowish- 
 white areas of soft cheese-like consistence. A few similar areas are found 
 elsewhere in the pancreas. 
 
 Histological examination reveals a slight localized interlobular pan- 
 creatitis, with typical fat necrosis. 
 
 It is possible that the first of these cases belongs to the 
 group described by Chiari, in which, in his opinion, self diges- 
 tion of the pancreas occurs shortly before death as an agonal 
 change and allows the escape of the fat splitting ferment and 
 the development of fat necrosis with associated minute 
 haemorrhages. That fat necroses may develop very quickly 
 is shown in one of the experiments of the series carried out 
 by Dr. Flexner and myself, in which necroses were evident 
 eight hours after injection of a one per cent, solution of 
 hydrochloric acid into the pancreatic duct. Wells has 
 observed their occurrence three to five hours after the applica- 
 tion of "pancreatin." 
 
 The theory that postmortem digestion of the pancreas 
 may set free the pancreatic secretion and allow the 
 development of fat necrosis has little evidence to support it. 
 In connection with these agonal necroses Opie believes it 
 is possible that shortly before death, owing to alterations in 
 the character of the pancreatic juice, it may become thick 
 and viscid, in consequence of which it 'flows slowly. Stagna- 
 tion and slight diffusion of the secretion may result and cause 
 the insignificant necroses occasionally seen. 
 
 16 
 
 Digitized by Microsoft® 
 
Beprinted from the Albany Medical Asmam, AprU 1904, 
 
 SYMPOSIUM ON DIABETES * 
 
 I. THE PATHOLOGY OF DIABETES. 
 
 By RICHARD M. PEARCE, M. D., 
 Director of the Bender Hygienic Laboratory, Albany, N. T. 
 
 As it will be impossible in the short time at my disposal 
 to cover all phases, many of them little understood, of the 
 pathology of diabetes ; and as the questions of chemistry and 
 metabolism are to be discussed by Dr. Edsall who has 
 specialized in these fields, I have determined to limit my 
 remarks to a consideration of the pathology of that form of 
 permanent diabetes, which, in the light of recent investiga- 
 tions, appears to be dependent upon lesion of the pancreas. 
 
 We may for the moment disregard the various forms of 
 temporary glycosuria, variously known as alimentary, nervous, 
 renal, and hepatic glycosuria and devote our attention 
 to the permanent form, which, owing to its obscure 
 etiology and the severe nutritional disturbances by which it 
 is accompanied, the difficulties of treatment and the univers- 
 ally fatal termination has been the object of constant obser- 
 vation and investigation. 
 
 The literature of diabetes is replete with hypotheses and 
 theories of the most divergent and conflicting character; but 
 throughout this literature may be seen the tendency to 
 associate diabetes with an altered physiology of the pancreas. 
 This is due undoubtedly to the important function which this 
 organ plays in the metabolism of carbohydrates, but also in 
 no small degree to the frequency with which gross lesions 
 of the pancreas are found in persons dying of the disease. 
 Thus of the early English cases, Cawleyi" (1788), reports 
 pancreatic calculus with atrophy of the gland; Bright (1833), 
 cancer of the head of the pancreas with duct obstruction and 
 
 * Part of a Symposium on Diabetes, read before the New York State Medical Society a* 
 Albany, Jannary 26, 1904. 
 
 t As the literature of diabetes has been so recently collected by Opie in hia book. Diseases 
 of the Pancreas, a work which has been freely need in the preparation of this summary, I have 
 not attached a bibliography. 
 
 Digitized by Microsoft® 
 
consequent atrophy. The first definite statement of such a 
 relation appears to be that of Bouchardat (1785), which 
 received considerable support from both French and English 
 writers. General support, however, was lacking, and the 
 German medical world especially, did not concur until the 
 publication of the convincing extirpation experiments of 
 V. Mering and Minkowski. In these experiments it was 
 found that complete removal of the pancreas is followed 
 always by a definite diabetes of severe grade. Glycosuria 
 appears as early as six hours after the operation ; the sugar 
 gradually increasing until the third day when it is usually 
 about eight or ten per cent., though twenty-two per cent. 
 (Hedon) has been observed. Towards the end, these high 
 percentages fall, and as the sugar reaches its minimum, 
 diacetic acid, acetone and oxybuturic acid appear in the 
 urine. The other symptoms of diabetes are present, as 
 hunger, thirst, and loss of weight. The animals ultimately 
 succumb after varying periods of progressive emaciation and 
 weakness. A repetition of these experiments upon a variety 
 of animals has yielded uniformly concordant results. In 
 birds glycosuria does not occur but the blood shows a con- 
 siderable increase in its sugar content (hyperglycemia). 
 Diet has but a slight modifying effect. While the maximum 
 sugar elimination naturally occurs with a carbohydrate diet, 
 an albuminous diet does not reduce it below five or six per 
 cent, while fasting for seven days does not cause its entire dis- 
 appearance. Moreover, it was found that if one-fourth or one- 
 fifth of the pancreas was allowed to remain diabetes did not re- 
 sult ; nor did it result, if after complete extirpation, a portion of 
 the pancreas was transplanted subcutaneously. These experi- 
 ments demonstrated conclusively a specific relation between 
 some function of the pancreas and carbohydrate metabolism. 
 What this function might be, or what changes in the 
 pancreas might disturb its equilibrium was not clear. It 
 was evident, however, that it could be explained only by an. 
 internal secretion, for no disturbance followed ligation of the 
 duct. Furthermore it was evident that no other organ could 
 vicariously assume this function of the pancreas. 
 
 Stimulated by these experiments, interest in the pancreas 
 was renewed and as a result much data concerning the 
 pathology of the pancreas of diabetes came to hand. Seegen, 
 Hansemann, Dieckhoff and Williamson reported the results 
 of large series of cases. In these, as well as in those previ- 
 
 Digitized by Microsoft® 
 
ously reported by Windle, Frerichs and White, definite 
 lesions of the pancreas occurred in a very large proportion of 
 the cases. This relation is strikingly shown in the tabula- 
 tions of Oser. Calculus of the duct with secondary atrophy 
 stands out as a prominent condition. Cancer of the head, 
 with like atrophy was not uncommon. In Hansemann's 
 series simple atrophy was present in thirty-six of forty cases. 
 Other conditions found were I chronic pancreatitis and lipo- 
 matosis. 
 
 Still many cases of diabetes constantly came to notice 
 without lesion of the pancreas. Naunyn, indeed, in a series 
 of forty autopsies found only one case in which the disease 
 could be ascribed to disease of the pancreas. On the other 
 hand, many cases of grave disease of the pancreas without 
 diabetes were observed; Hansemann reports such a case in 
 which complete destruction of the pancreas existed. This 
 anomalous condition he attempts to explain by the hypothe- 
 sis that in primary cancer of the pancreas the cells of the 
 new growth retain, in part at least, the function of the cells 
 from which they are derived and that therefore the internal 
 secretion is not disturbed. 
 
 At the same time various attempts were made to associate 
 with diabetes some distinct histological picture; as the com- 
 bined chronic pancreatitis and arterio-sclerosis of Hoppe- 
 Seyler and Fleiner, and the perivascular sclerosis of Lemoine 
 and Lannois. An important condition observed by the latter 
 was the penetration of the fibrous strands into the acini, thus 
 separating the cells and presenting a condition which they 
 designated unicellular sclerosis. Hansemann particularly 
 associated with diabetes a condition of the pancreas which he 
 calls granular atrophy but which from his description appears 
 to be a primary chronic pancreatitis with atrophy. 
 
 The most important contributions to the histology of the 
 pancreas of diabetes are those made during the past few 
 years by Opie. He distinguishes two forms of chronic 
 pancreatitis; (i) interlobular pancreatitis, characterized by 
 the formation of fibrous tissue between the lobules, which are 
 often invaded from the periphery and (2) interacinar pan- 
 creatitis, in which the newly formed fibrous tissue is diffusely 
 distributed within the lobule, separating the acini and invad- 
 ing the islands of Langerhans. With the first type, frequently 
 associated with duct obstruction, diabetes occurs only in 
 those relatively uncommon cases in which an advanced 
 
 3 
 
 Digitized by Microsoft® 
 
sclerosis involves the islands. With the interacinar type, 
 however, diabetes is much more frequent. Another important 
 change observed by the same writer is a hyaline transforma- 
 tion of the islands with little or no alteration of the secreting 
 parenchyma. He therefore looks upon involvement of the 
 islands of Langerhans as the important point in the 
 pathology of diabetes. His figures support this view; 
 in twenty-one cases of interlobular pancreatitis, diabetes 
 occurred but once and in this case the advanced sclerosis 
 had produced alterations of the islands; while in seven of 
 nine cases of the interacinar type, diabetes coexisted; and in 
 six of nineteen cases of diabetes, hyaline transformation of 
 the islands was present. 
 
 Such observations imply that the islands are in some way 
 concerned in carbohydrate metabolism and therefore of 
 fundamental importance in any consideration of the etiology 
 of diabetes. The occurrence of these lesions in diabetes has 
 been confirmed by Weichselbaum and Stangl, Herzog, Gentes, 
 Wright and Joslin, Curtis^, Finney^ Schmidt, Hansemann 
 and Herxheimer^ The last three writers, however, are not 
 prepared to admit that these changes are of etiological im- 
 portance. 
 
 In this connection the following notes of five cases of 
 diabetes and one of acute glycosuria, studied during the past 
 two years, may be of interest: 
 
 /. Records of the Pathological Laboratory, University of Pennsylvania, 
 Autopsy No. 01.212. Female, aged sixty-two years. Clinical diagnosis — 
 Diabetes, moist gangrene of leg, tetanus. Anatomical diagnosis — Throm- 
 bosis of left iliac artery and its branches ; softened thrombi in left auricular 
 appendix; moist emphysematous gangrene of left leg; infarction of spleen 
 and left kidney; chronic obliterative pericarditis; chronic mitral endocar^ 
 ditis; hypertrophy of left ventricle; chronic diffuse nephritis; chronic in- 
 terstitial pancreatitis; chronic adhesive pleuritis. Histological examination 
 — Interacinar pancreatitis with involvement of many of the islands of 
 Langerhans by the invasion of dense bands of fibrous tissue. 
 
 //. Records of the Pathological Laboratory, University of Pennsylvania, 
 Autopsy No. 01.309. Male, twenty-seven years. (Service of Dr. J. H. 
 Musser, University Hospital.) Clinical diagnosis — Phthisis pulmonalis, 
 diabetes mellitus, Banti's disease. History of diabetes of two years' dura- 
 tion; sugar varied in amount from 1.25 to 7.5 per cent.; specific gravity 
 of urine 1028 to 1040. Anatomical diagnosis — Splenomegaly ; slight chronic 
 interstitial hepatitis; general tuberculous peritonitis; tuberculosis of lung 
 (caseous and gelatinous pneumonia with cavity formation) ; tuberculous 
 
 1. VBeho Med. du Nord, 1903, vi, 140. 
 
 2. 2%« Medical Chronicle, 1903, v, 137. 
 
 3. Orth-B'esttehrift, 1908, a. 38. 
 
 Digitized by Microsoft® 
 
pleuritis; slight chronic mitral endocarditis; fatty degeneration of heart 
 and kidneys. Pancreas — The surface is more nodular than normal, the de- 
 pressed portions oedematous. On section the nodular appearance is very 
 marked. The consistence is about that of the normal pancreas. Histo- 
 logical examinAtion reveals a moderate perivascular fibrosis and. a hyaline 
 degeneration of many of the islands; approximately one of every five 
 is hyaline. 
 
 ///. — Records of the Pathological Laboratory, University of Pennsylvania, 
 Autopsy 03.i6. Male, ?ged thirty-eight years. (Service of Dr. Alfred Sten- 
 gel, University Hospital.) Clinical diagnosis — Diabetes, tuberculosis. 
 Anatomical diagnosis — Gangrene and multiple abscesses of lungs; acute 
 fibrinous pleuritis; fatty degeneration of heart and liver; cloudy swelling 
 of kidneys ; slight atrophy of thyroid gland ; chronic interstitikl pancreatitis ; 
 fibrosis and caseation of right adrenal; arterio-sclerosis ; acute splenic 
 tumor; multiple superficial erosions of the oesophagus. Pancreas — Weight 
 seventy grammes, smaller than normal and firmer. The shape is more dis- 
 tinctly triangular than usual. The surface of the pancreas is grossly 
 nodular. Histological examination of the pancreas reveals a slight diffuse 
 increase of connective tissue; the islands of Langerhans are normal. 
 The right adrenal is almost completely destroyed by tuberculous casea- 
 tion , and newly formed fibrous tissue. A small ganglion adjacent 
 to the adrenal is not involved in this process. Left adrenal is normal. The 
 alveoli of thyroid gland are moderately dilated. 
 
 IV. — Case of Dr. Wm. Pepper, of Philadelphia, presenting a history of 
 nephritis, uraemia and acute (terminal) glycosuria. The autopsy report is 
 not available. For sections of the pancreas and for the laboratory report 
 which latter, owing to the peculiar terminal character df the glycosuria is 
 given in full, I am indebted to Dr. Pepper. Clinical history — Female, aged 
 fifty-eight, admitted to the University Hospital, September 4, 1902. 
 " Patient in a semi-delirious condition and cannot give a satisfactory history 
 of herself. From her daughter it was learned that the patient had been 
 complaining of headache, dizziness and defective vision for the last two 
 weeks, and that during the past week she has been at times quite delirious. 
 On admission catheterized ; urine slightly cloudy, dark amber; no precipi- 
 tate; specific gravity 1025; albumen i-iS by bulk (centrifugated) ; no 
 sugar; no bile, indican negative ; diazo not made; a few flat round epithelial 
 cells, a few pale and dark granular casts, and a few white blood cells. 
 Sept. 5th — Mental condition better. Bowels moved freely. Urine dark 
 amber; slightly cloudy; very light flocculent precipitate; specific gravity 
 1030; distinctly acid; albumen 1-15 by bulk (centrifugated); no sugar; 
 bile and indican negative ; a few flat and round epithelial cells, few white blood 
 cells, many pale and dark granular casts. Sept 6th — Blood — Haemoglobin, 
 eighty-nine per cent.; red cells 5,010,000; white cells 10,860; Widal nega- 
 tive ; no malaria parasites ; nothing abnormal in stained films. Urine 9 a. m. 
 dark amber; slightly cloudy; light, flocculent precipitate; specific gravity 
 1029; distinctly acid; albumen i-is by bulk (centrifugated); no sugar; 
 bile and indican negative. Many flat and round epithelial cells, a good 
 many white blood cells, very many very short, pale and dark granular 
 casts, a few waxy casts. S p. m. urine color changed to a very pale amber ; 
 slightly cloudy; no precipitate; specific gravity 1029; distinctly acid; a 
 
 Digitized by Microsoft® 
 
decided trace of albumen ; sugar 0.83 per cent, by Fehling's test ; bile and 
 indican negative; sediment shows many short pale and dark granular casts, 
 a few waxy casts, very many flat round and spindle epithelial cells, very 
 many white blood cells, many mixed urates. Sept. 7th — Urine 11 a. m. 
 pale amber, slightly cloudy; light flocculent precipitate; specific gravity 
 1031 ; distinctly acid ; decided trace of albumen ; sugar 0.83 per cent. ; bile 
 and indican negative. ' Many short pale and dark granular casts, few waxy 
 casts, many flat spindle and round epithelial cells, many white blood cor- 
 puscles, many mixed urates. Sept. 8th — Urine 2 p. m. pale amber; hazy; 
 specific gravity 1030 ; acid ; sugar present in large quantities ; decided trace 
 of albumen; epithehal cells, urates, no casts. Histological examination — 
 of the pancreas shows irregular foci of slight interacinar pancreatitis with 
 distinct involvement of the islands of Langerhans. The islands are fre- 
 quently enlarged; the fibrous tissue abundant and often somewhat hyaline 
 in appearance; the cells compressed and indistinctly outlined. 
 
 V — Bender Laboratory Records, Autopsy 0.726. Case of Dr. Andrew 
 MacFarlane, Albany, N. Y. Clinical abstract — Male, aged twenty-six; 
 laborer, duration of diabetes about eight months. Sugar constantly present 
 in urine rising at times as high as four per cent. ; diacetic acid and 
 acetone also present. Autopsy — The abdominal viscera only were 
 examined through a limited incision. Pancreas — Apparently normal in its 
 gross appearance. Histologically, scattered areas of interacinar pancre- 
 atitis are found. The islands present a slight but definite increase of con- 
 nective tissue with moderate atrophy of the intervening cells. A few small 
 areas of fat necrosis exist. 
 
 VI — Records of the Pathological Institute, University of Leipzig, 
 Autopsy 98* (i8p8). Female, aged twenty-six years. Clinical diagnosis — 
 Diabetes, pneumonia. Note of post-mortem examination — Extreme sclerosis 
 and atrophy of the pancreas with presence of calculi (weight of pancreas, 
 fifty-four grammes) ; diffuse lobular pneumonia with fibrinous pleuritis ; 
 hydro-pericardium ; cloudy swelling of liver ; fatty degeneration of kidney ; 
 calcareous bronchial lymph nodes. Histology of pancreas — Extensive 
 atrophy and chronic pancreatitis of the interlobular type. In the more 
 sclerotic portions the growth of connective tissue is interacinar. The 
 islands are few in number, small, often isolated and frequently invaded 
 by fibrous strands. 
 
 Lesions of the islands occurred in five of these six cases; 
 in one the change was of the nature of a hyaline transforma- 
 tion; in the other four an interacinar pancreatitis with 
 involvement of the islands existed. Whether or not the 
 glycosuria in case III, in which the pancreas was normal, was 
 dependent upon the lesion of the adrenal it is impossible to 
 say. Glycosuria does not usually accompany such lesions. 
 On the other hand the experiments of Blum, Herter and 
 Richards indicate that the adrenal is in some way concerned 
 in carbohydrate metabolism. These investigators found that 
 
 * For the privilege of reporting this case, I am indebted to Prof. ¥. Marchand of the 
 University of Leipzig. 
 
 6 
 
 Digitized by Microsoft® 
 
after the injection into animals of the adrenal gland or 
 adrenal extract, a temporary glycosuria results. Herter and 
 Richards produced the same result by painting the surface of 
 the pancreas with the extract. These experiments, however, 
 are not analogous to destruction of the adrenal. 
 
 I can at this time also give a preliminary report of an 
 investigation, as yet incomplete, of the diabetes associated 
 with cancer of the pancreas. It is well known that cancer of 
 the pancreas may or may not be accompanied by glycosuria. 
 We have no definite explanation of the condition under 
 which it occurs, though a resume of the literature shows that 
 it is most common when the tumor is in the head of the pan- 
 creas and by duct obstruction causes atrophy (chronic pan- 
 creatitis) of the distal portion of the organ. The reverse 
 condition, complete destruction of the gland by cancer with- 
 out diabetes, is likewise difficult of explanation. It is true 
 that Hansemann has suggested that the function of internal 
 secretion is preserved by the cells of the new growth; but 
 this is not applicable to secondary tumors and if we accept 
 the theory that the islands alone furnish the internal secre- 
 tion it is not tenable, as pointed out by Ssobolew and Opie, 
 unless we assume that all primary tumors of the pancreas 
 have their origin in the islands, and in support of this last 
 possibility there is little evidence. NichoUs* has reported a 
 primary adenoma of an islands, and Fabozzi* has recently 
 described several primary cancers of the pancreas which he 
 believes to have arisen from a like source. Fabozzi's obser- 
 vations, however, are not conclusive and are entirely opposed 
 to the great mass of work on this subject. 
 
 I have been able to collect for study thirty cancers of the 
 pancreas. Three of these were accompanied by glycosuria, 
 which in one case was a true diabetes, and in the second an 
 intermittent glycosuria. I have no data concerning the char- 
 acter of the glycosuria in the third case. In the first case the 
 tumor was in the head of the pancreas, causing obstruction of 
 the duct with chronic pancreatitis of the distal portion of the 
 organ. This chronic inflammation was for the most part 
 interlobular, but in places was of the interacinar type with 
 involvement of the islands, that is, presented the lesion 
 which, according to our new ideas is supposed to be essen- 
 tial to the occurrence of diabetes. The third case was 
 
 i . Journal of Medical Beiearch, 1908, viii , 386. 
 6. ZUgler's SeitrSge, 1808, xxxiv., 19». 
 
 Digitized by Microsoft® 
 
unsatisfactory for study inasmuch as only the tumor had 
 been preserved and observations concerning the condition 
 of the uninvolved pancreas could not be made. In the 
 second case, accompanied by intermittent glycosuria, the pan- 
 creas showed no noteworthy change in the uninvolved portion 
 of the organ. This case, similar to those reported by Wille, 
 brings up the vexed question of alimentary glycosuria. It 
 is known that a normal individual may digest without dis- 
 turbance from 150 to 250 grammes of sugar, but that an 
 increase over this amount may lead to temporary excretion 
 of glucose in the urine. This excretion is due to hypergly- 
 cemia, the inevitable result of a sudden flooding of the cir- 
 culation with an amount of glucose beyond that which the 
 body is able at once to consume. On the other hand, in a 
 number of abnormal conditions where there is no actual dia- 
 betes, an alimentary glycosuria may be produced by an 
 amount of sugar which in the economy of the normal indi- 
 vidual is easily consumed. Among the prominent conditions 
 in which this occurs may be mentioned neuroses (traumatic 
 and other forms), organic nervous diseases, alcoholism, infec- 
 tious diseases, exophthalmic goitre, hepatic and pancreatic 
 disease. Is it not possible that this temporary excretion of 
 sugar is due to some transient disturbance of the function of 
 the islands brought about by vascular disturbances of the 
 islands or by some toxic substance, perhaps of the nature of 
 an autotoxin, affecting the secretion of these structures or 
 temporarily throwing them out of function? In a destructive 
 disease of the pancreas, as cancer, a certain number of the 
 islands are necessarily destroyed; the remainder, however, 
 may be sufficient under normal circumstances to furnish the 
 necessary secretion, whatever it may be, essential to carbohy- 
 drate metabolism; but if overtaxed by an unusual ingestion 
 of glucose, hyperglycemia and a consequent, glycosuria 
 might result. This explanation has occurred to me in con- 
 nection with my case of cancer accompanied by intermittent 
 glycosuria and that it is not entirely hypothetical is indi- 
 cated by the observations of Ssobolew, confirmed by Herzog, 
 that in some cases of permanent diabetes the pancreas 
 exhibits a great diminution in the number of islands and in 
 some cases an entire disappearance. 
 
 This, functional deficiency, however, is far from proven, 
 and in mentioning it I do not wish to be placed in the 
 position of advocating pancreatic origin for all forms of 
 
 8 
 
 Digitized by Microsoft® 
 
diabetes. We have a definite basis for ascribing a certain 
 portion of the cases of diabetes to lesions of the pancreas; but 
 here we must rest until an increase in our knowledge allows 
 further conclusions. It is probable as suggested by Naunyn 
 that with the pancreas other organs preside over carbohydrate 
 metabolism, though perhaps in a less degree, and that the 
 pancreas may act vicariously for them, although their function 
 may not be sufficient to be substituted for that of the pancreas. 
 
 Another point brought out by this investigation of cancer 
 of the pancreas is the resistance of the islands to invasion or 
 obliteration by the new growth and the accompanying 
 sclerotizing inflammation. Numerous unaltered islands were 
 frequently seen in the fibrous tissue at the margin of the 
 growth where the acini had been completely destroyed; 
 moreover, these structures, both at the edge of the tumor 
 and in distant portions of the organ, were often three to five 
 times larger than normal. This persistence indicates the 
 resistance of the islands and the increase in size suggests a 
 hypertrophy possibly compensatory in character. No cases 
 of complete involvement of the organ by cancer similar to 
 those of Litten and Hansemann were seen; in the few 
 instances of extensive invasion a sufficient amount of the 
 organ, by analogy with the partial extirpation of v. Mering 
 and Minkowski, remained to carry on the internal function 
 of the gland. 
 
 Whatever the ultimate explanation of diabetes may be, it 
 must be admitted that ouy knowledge of pancreatic diabetes 
 has been largely extended by the recent studies of the islands 
 of Langerhans. And here it may be pertinent to inquire 
 what evidence have we that the islands from their anatomy, 
 development and appearance in diseases other than diabetes 
 are independent structures concerned in carbohydrate 
 metabolism. 
 
 From the discovery of these structures by Langerhans in 
 i86g, until within a few years, many and varied efforts have 
 been made to explain their origin and function. They have 
 been regarded from time to time as nervous elements, 
 primary ducts, embryonic remains and as immature or wasted 
 acini. The study of the ductless glands is largely responsible 
 for the recognition of the similarity of structure of the islands 
 and these organs. The structure of the islands — columns of 
 cells rich in protoplasm, lying in an abundant vascular net- 
 work of a sinusoidal character and with no excretory duct — 
 
 Digitized by Microsoft® 
 
naturally suggests as their function the elaboration of some 
 substance destined to be introduced into the circulation. This 
 theory of the independence of the islands, first advanced by 
 Shafer, received considerable support from the experiments of 
 Ssobolew. He found that after feeding animals on carbohy- 
 drates the islands became granular, indicating, according to 
 his interpretation, a greater functional activity due to the 
 increased demand for a ferment essential to carbohydrate 
 metabolism. Furthermore he found that in the sclerotic 
 atrophy following ligation of the ducts in animals the islands 
 are not implicated and glycosuria does not result. These 
 observations have been confirmed by Schulze. A similar 
 persistence is frequently seen in the human pancreas in the 
 sclerosis following duct obstruction; and has also been 
 observed by Schlesinger, Opie, Condon* and myself' in the 
 peculiar fibrous induration of the organ which occurs, not 
 infrequently, in congenital syphilis. 
 
 Again, in the embryonic development of the islands we 
 have evidence that they are destined for some special func- 
 tion. As I have recently demonstrated*, the islands at an 
 early stage of the growth of the pancreas arise from the 
 primary acini and for some time remain connected to these 
 structures by a solid process of cells; but eventually become 
 entirely separated, and with the development of a blood sup- 
 ply proportionally greater than that of the acinus, become 
 anatomically, independent. Such anatomical differentiation 
 argues for a special function, and taking into consideration 
 the recently accumulated knowledge, the result of pathological 
 investigation, we are forced to the conclusion that this 
 function is an internal secretion concerned in carbohydrate 
 metabolism and therefore of fundamental importance in any 
 v;onsideration of diabetes. This, however, is not equivalent 
 to saying that all forms of diabetes may be traced to func- 
 tional or structural changes of these structures. Fully one- 
 half of the reported cases studied with regard to lesions of 
 the islands show no alteration of these structures. That 
 several forms of diabetes exist and that other organs than 
 the pancreas are concerned in their etiology is evident; but 
 this does not detract from the value of our recently acquired 
 knowledge of pancreatic diabetes. The manner in which the 
 
 6. Tram. Chicago Path. Society, 1899-'01, iv., 200. 
 
 7. ATnerican Medicine, 1903, vi, 1030. 
 
 8. American Journal (if Anatomy, 1903, ii, 145. 
 
 lO 
 
 Digitized by Microsoft® 
 
hypothetical internal secretion of the islands is supposed to. 
 act comes under the head of chemistry and metabolism and 
 will doubtless be considered by Dr. Edsall. 
 
 A consideration of diabetes would be incomplete without a 
 consideration of the forms of glycosuria apparently dependent 
 upon changes in organs other than the pancreas. As little 
 however has been added in recent years to our knowledge of 
 these conditions, I will speak of them but briefly. The forms 
 of glycosuria associated with acromegaly, with cirrhosis of 
 the liver, with arterio-sclerosis, and especially the diabetes of 
 hsemochromatosis, in the light of recent investigations would 
 appear to have a relation to disease of the pancreas. 
 
 Thus in acromegaly chronic inflammation of the pancreas 
 has frequently been found at autopsy and it is probable as 
 suggested by Opie, that "induration presumably analogous to 
 that which acromegaly causes elsewhere occurs also in the 
 pancreas and where sufficiently advanced produces glycosuria 
 or even permanent diabetes." 
 
 The frequency with which arterio-sclerosis is associated with 
 diabetes renders it probable that this condition by affecting the 
 riutrition of the pancreas may be the cause of changes, inflam- 
 matory or sclerotic in nature, leading eventually to diabetes. 
 
 That the diabetes accompanying cirrhosis of the liver may, 
 in many cases, be due to a coincident chronic pancreatitis, 
 the result of a common etiological agent is indicated by the 
 studies of Lefas, Opie and Steinhaus*. A more striking 
 example of the dependence of chronic lesions of the liver 
 and the pancreas upon a common etiological factor is 
 hsemochromatosis [diahete bronzd'). This condition, first 
 described by v. Recklinghausen, is characterized by the 
 extensive deposition of an iron-containing pigment in the 
 cells of various organs and tissues, and is associated with 
 a necrosis of the parenchymatous cells and a proliferation of 
 the stroma. The liver and pancreas suffer most, the former 
 being the seat of a hypertrophic cirrhosis, the latter of a 
 chronic pancreatitis. The diabetes occurring in this condi- 
 tion has been explained by some writers as the result of the 
 blood destruction, by others as the result of the pancreas 
 lesion. Opie, in one case, found the inflammation to be of 
 the intracinar type with involvement of the islands of 
 Langerhans; and therefore concludes that this lesion is 
 responsible for the occurrence of diabetes. In his case, 
 
 9. DmUeh. Arch.f. klin. Med,. 1902, Ixxiv. 
 
 II 
 
 Digitized by Microsoft® 
 
however, diabetes did not exist owing to the incomplete 
 destruction of the inlands; the patient dying of typhoid fever 
 before the disease had reached its height. 
 
 It is well established that glycosuria, temporary or other- 
 wise, may follow lesions of the central nervous system and 
 perhaps also of the sympathetic and larger peripheral nerves. 
 Claude Bernard's classical piqilre of the floor of the fourth 
 ventricle in animals, lesions (haemorrhage, tumor and soften- 
 ing) of the same location (medulla) and various pathological 
 changes of the brain and cord in man illustrate this relation. 
 We have, however, no definite information and Naunyn's 
 theory that through the secretory or nutritive nerves we have 
 a disturbance of function in the organs presiding over carbo- 
 hydrate metabolism, is as satisfactory as any other. 
 
 In experimental phloridzin diabetes it has been demon- 
 strated that a true hyperglycemia does not exist and that the 
 glycosuria is due to an injury of the renal epithelium which 
 allows an escape of the sugar normally present in the blood. 
 An analogous condition (renal diabetes) in man has been sus- 
 pected but not proven. 
 
 The influence of the adrenal gland upon carbohydrate 
 metabolism has already been mentioned. 
 
 The relation of the thyroid gland to glycosuria is unex- 
 plained. Hansemann has collected fifteen cases in which 
 diabetes and exophthalmic goitre were associated. Alimentary 
 glycosuria is also frequently seen in this condition, according 
 to Chvostek in sixty-nine per cent, of all cases; and it has 
 also been observed after the therapeutic use of thyroid 
 extract. It is worthy of note, however, that in none of the 
 cases of diabetes associated with exophthalmic goitre has the 
 condition of the pancreas been studied (Opie). 
 
 I have attempted to place before you in this brief summary 
 the relation of the pancreas and especially the islands of 
 Langerhans to diabetes. If I have dwelt too strongly upon 
 this point it is because the knowledge of this relation enables 
 us to sharply separate at least one form of diabetes from all 
 others, and thus offers a definite point of departure for the 
 study of those forms of glycosuria which, from our knowledge 
 of the complexity of carbohydrate metabolism, we assume 
 may be due to disturbances of other organs. 
 
 12 
 
 Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
Digitized by Microsoft® 
 
Digitized by Microsoft®