AUTHORIZED ENGLISH EDITION TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS IN MAN AND ANIMALS / PROCESSES OF INFECTION AND RESISTANCE A BIOLOGICAL AND EXPERIMENTAL STUDY * / WITH THIRTY-ONE TEXT ILLUSTRATIONS AND TWENTY-FIVE • COLORED PLATES V BY ALBERT CALMETTE / Associate Director of the Pasteur Institute, Paris / AUTHORIZED TRANSLATION BY WILLARD B. SOPER', M.D. "and GEORGE H. SMITH, Ph.D. Saranac Lake, New York School of Medicine, Yale University PUBLISHED BY WILLIAMS & WILKINS COMPANY BALTIMORE, U. S. A. 1923 Copyright 1923 -WILLIAMS & WILKINS COMPANY Made in United States of America All rights reserved, including that of translation into foreign languages, including the Scandinavian COMPOSED AND PRINTED AT THE WAVERLY PRESS By the Williams & Wilkins Company Baltimore, Md., U. S; A. PREFACE Since the publication in 1895 of I. Straus’s splendid book,1 an immense number of studies on Tuberculosis have appeared in all languages. In writing this work, my object has been to sift out from the most important of these studies,—and also from my own research of several years,'—the scientific principles on which to base, in the present state of our knowledge, the campaign against the most terrible of human infectious diseases. I have intentionally omitted all discussion of doctrines or of theories, and have restricted the bibliographic indications to those which in my opinion are strictly necessary to any one seeking details of observations or experiments of which I state only the conclusions. This was indispensable if I were to hope to make clear the most impor- tant facts and the new ideas. I have attempted above all to write a work for the biologist and the investigator. This book then is addressed to physicians and to veterinarians and at the same time to laboratory workers, and it is my hope that the latter will derive from it fruitful suggestions for their researches. Many of my friends were formerly my pupils and have been asso- ciated with me in my work for almost twenty years. Two of them, alas, Lucien Bruyant, agrege professor, and Leon Massol, agricultural engineer, have passed away, one of them a victim of the disease which we were studying together, the other a modest and glorious hero of the great war for the liberation of humanity. To them first, and then to my dear co-workers Camille Guerin and Maurice Breton and all those best of comrades who have shared my laboratory life, I dedicate this book in token of my very deep affection and in memory of those days both joyous and sorrowful through which we have lived together. Also to M. Millot, in charge of the course in animal painting at the Museum of Natural History, I owe the warmest thanks for the care with which he has reproduced from my photographs the greater part of the numerous colored plates of this book. M. Masson has seen fit to make of the French edition a veritable work of art. A. Calmette. 1 La tuberculose et son bacille, Paris, 1895, Rueff & Cie III TRANSLATORS’ PREFACE Interest in the problem of tuberculosis has never been greater than today, when, in this country at least, the effort to combat it is appar- ently meeting with success. The casual observer might infer that the campaign is built upon a thoroughly solid foundation of knowledge and that our understanding of the disease is complete. On the contrary, numerous problems remain to be solved. Much of the pathology of the disease must be rewritten in the light of the phenom- ena of reinfection; our understanding of tuberculosis immunity is still chaotic; the cure continues so near and yet so far; and many there are who contend that reductions in mortality are deceptive and that the results do not justify the expenditures of money and effort. At such a time appears UInfection bacillaire et la Tuberculose chez Uhomme et chez les animaux, by Professor Albert Calmette. The author is one who stands out among a small internationally-known group to whom tuberculosis has been a life-long study and upon whose continued contributions year after year much of the present conception of the disease has been established. Never, probably, in the rush of this very work and of his varied services, would time have been found for such an undertaking had not Professor Calmette been forced to undergo the hardship of an internment at Lille throughout the Great War. It was under these circumstances that most of his book was compiled. Many who have studied this work in the orginal French have been struck by the need of bringing it before the English reading public. It is at once a resume of the ripe experience of a brilliant scientist, and a summary of our present scientific knowledge. It is a study which can be read and reread, not only for its facts, but also for its wealth of suggestion. The book is full of stimulating ideas, and quite as much so where the author is on debatable ground. The translators desire to gratefully acknowledge the encouragement received from Dr. Linsly R. Williams of the Commission for the Prevention of Tuberculosis in France, of the Rockefeller Foundation; also the very great assistance of Dr. E. F. Ducasse of Paris, who, through his thorough knowledge of both the French and English languages; was ever ready with indispensable aid. April, 1922. V TABLE OF CONTENTS INTRODUCTION Tuberculous Virus Some pages of history, Bayle, Laennec, Villemin and Robert Koch 1 PART ONE THE TUBERCLE BACILLUS AND PROCESSES OF INFECTION BY IT Chapter I MORPHOLOGY OF THE TUBERCLE BACILLUS.—METHODS OF EXAMINATION, OF STAINING AND OF DIFFERENTIATION A. Morphology 11 B. Technique of staining methods 14 C. Concentration of tuberculous material 21 D. The staining of sections 25 E. Differential staining methods 27 Chapter II CULTIVATION AND ISOLATION OF THE TUBERCLE BACILLUS A. Cultivation of the tubercle bacillus from pathological material free from other bacteria 31 a. Cultures on potato 33 b. Cultures on fluid media 34 c. Cultures on egg media 41 d. Cultures on tissue 43 e. Cultures in collodion bags and in filter bougies... • 43 f. Cultures on various organic media—-Bile media 43 B. Culture of the bacillus from pathological material which contains other bacteria 45 Chapter 111 INFLUENCE OF PHYSICAL AND CHEMICAL AGENTS UPON THE TUBERCLE BACILLUS A. Action of air and atmospheric pressure 49 B. Action of light and ultra-violet rays of the spectrum 49 C. Action of low temperatures 50 D. Action of heat 50 E. Effect of desiccation 51 VII VIII TABLE OF CONTENTS F. Influence of putrefaction 52 G. Action of electricity and ozone 53 H. Effect of age upon cultures 53 I. Action of various chemical agents 54 Chapter IV CHEMICAL COMPOSITION OF THE TUBERCLE BACILLUS A. Mineral composition 59 B. Substances extractable by fat and wax solvents 61 C. Carbohydrates 65 D. Protein substances 65 Chapter V TOXINS OF THE TUBERCLE BACILLUS—EXO- AND ENDO-TOXINS—TUBERCULINS A. Toxicity of dead tubercle bacilli.—Endo-tuberculins.—Volatile poisons of the tubercle bacillus 70 B. Tuberculin of Robert Koch.—Its preparation 75 C. Exo- and endo-toxins of the bacillus 76 D. Purified tuberculins 77 E. Chemical properties of the tuberculins 78 F. Estimation of the toxicity of the tuberculins 80 G. Products derived from the tuberculin of Koch 82 I. Tuberculin TR, of Koch 82 II. Tuberculin BE, of Koch 83 III. Tuberculin AF, (without albumose) of Koch 83 IV. Tuberculocidine of Klebs 84 V. Tuberculin of Maragliano 84 VI. Oxytoxine of Hirschfelder 85 VII. Tuberculol of Landmann 86 VIII. Tuberculin of Beranek 87 IX. Tubplytin of Siebert and Romer 88 X. Tuberculo-plasmin of H. Buchner and Hahn 88 XL Tuberculin of Rosenbach 89 XII. Tuberculin of Vaudremer 89 XIII. Neurine-tuberculin of Much 90 XIV. Tuberculo-mucin of Weleminsky 90 XV. Tuberculin bovine PTO of Spengler 91 XVI. Tuberculous endo-toxin of Baudran 91 XVII. Ferruginous tuberculin of Ditthorn and Schultz 92 XVI11. Teb6an of Levy and Kaenker 93 XIX. Tuberculo-toxoidine of Ishigami 93 XX. Tebesapin or mollimentNo. 8 (former Prosp6rol) of Zeuner, 93 H. Action of physical and chemical agents on the tuberculins 94 TABLE OF CONTENTS IX Chapter VI THE HISTOGENESIS AND EVOLUTION OF THE TUBERCLE AND OF BACILLARY LESIONS WITHOUT TUBERCLE FORMATION A. Histogenesis and evolution of the tubercle.—Anatomical processes of healing 97 B. Process of caseation of tubercles 108 C. Tubercle bacillus lesions without tubercle formation 112 Chapter VII PRINCIPAL PATHOLOGICO-ANATOMICAL TYPES OF INFECTION BY THE TUBERCLE BACILLUS A. Tubercle bacillus septicemia.—Miliary tuberculosis 114 B. Latent tuberculous infection 117 C. Progressive tuberculous infection—Predominance of pulmonary local- izations 119 D. Localizations in the pleura 121 E. Other localizations 121 Chapter VIII MECHANISM OF TUBERCULOUS INFECTION—PENETRATION OF THE VIRUS INTO THE BODY BY WAY OF THE SKIN AND MUCOUS MEMBRANES A. Lymphatic circulation.—Lymph.—-Glands.—Role of the Leucocytes in tuberculous infection 123 B. Portals of entry of the virus in latent tuberculous conditions 127 C. Infection through the skin -.. 129 D. Tuberculous infection through the mucous membranes 131 a. Ocular mucous membrane 132 b. Nasal mucous membrane 133 c. Bucco-pharyngeal mucous membranes.—Tonsils 134 d. Genito-urinary mucous membranes 136 Chapter IX TUBERCULOUS INFECTION BY THE RESPIRATORY PASSAGES A. Mechanism of primary tuberculous infection of the respiratory passages 138 B. Experimental pulmonary infection with dried bacilli and with dried tuberculous material 142 C. Experimental infection of the lungs with fresh bacilli and with sprays of tuberculous material 145 D. Conditions and relative frequency of primary infection of the lung by the inspired air 151 X TABLE OF CONTENTS Chapter X TUBERCULOUS INFECTION BY ABSORPTION FROM THE DIGESTIVE TRACT A. The mechanism of digestive absorption of tubercle bacilli.—Their migrations in the body 157 B. Experimental demonstration of the passage of tubercle bacilli through normal mucous membrane.—The course which they follow in infecting the lungs or other organs 160 Chapter XI FREQUENCY AND PATHOLOGICO-ANATOMIC CHARACTERISTICS OF TUBERCULOUS INFECTION IN CHILDREN A. Tuberculosis of the lungs and tracheo-bronchial glands (tuberculose ganglio-pulmonaire) 171 B. Tuberculosis of the meninges 179 C. Abdominal tuberculosis 180 D. Relative frequency of abdominal tuberculosis and tuberculous menin- gitis in children 181 E. Tuberculosis of the glands of the neck 183 Chapter XII PATHOLOGICO-ANATOMIC CHARACTERISTICS OF TUBERCULOUS LUNG INFECTION IN ADULTS AND IN THE AGED A. Acute miliary tuberculosis of the lungs 186 B. Acute pneumonic tuberculosis 187 C. Chronic pulmonary phthisis 189 D. Pulmonary tuberculosis in the aged 193 Chapter XIII TUBERCULOSIS OF THE SEROUS MEMBRANES A. Tuberculous pleurisies 196 1. Acute sero-fibrinous pleurisy 196 II. Suppurative pleurisy 198 B. Tuberculous peritonitis 199 C. Tuberculous pericarditis 201 Chapter XIV TUBERCULOUS MENINGITIS AND TUBERCULOSIS OF THE NERVOUS CENTERS A. Tuberculous menipgitis 202 B. Tuberculosis of the nervous centers 207 TABLE OF CONTENTS XI Chapter XV TUBERCULOUS INFECTION OF THE LIVER, SPLEEN, KIDNEYS AND INTESTINE A. Tuberculous infection of the liver 208 I. Degenerations and atrophies 209 II. Hepatic tubercles 209 III. Tuberculous cirrhoses 210 IV. Tuberculous parenchymatous hepatitis 210 B. Tuberculous infection of the spleen 211 C. Tuberculous infection of the kidneys 212 D. Tuberculosis of the suprarenal capsules 216 E. Pathologico-anatomic characteristics of tuberculous lesions of the intestine 217 Chapter XVI LOCALIZATIONS OF TUBERCLE BACILLI IN BONES AND JOINTS A. Principal forms of bone tuberculosis; their genesis 221 B. Influence of mixed infections in tuberculosis of the bones 224 C. Pathologico-anatomic characteristics of bone and joint tuberculosis.. 225 Chapter XVII CUTANEOUS LOCALIZATIONS IN TUBERCULOUS INFECTION.—THEIR PATHOLOCICO- ANATOMIC CHARACTERISTICS A. Lupus 228 B. Tuberculous ulcers 232 C. Tuberculous gummata 233 D. Tuberculous lymphangitis 234 E. Tuberculides 234 Chapter XVIII TUBERCULOUS BACILLEMIA Tubercle bacillus bacillemia 237 Chapter XIX THE RbLE OF HEREDITY IN TUBERCULOUS INFECTION.—TRANSMISSION OF THE BACILLUS BY THE PARENTS.—HEREDITARY DYSTROPHIES.— SPECIFIC PREDISPOSITION AND FAMILY CONTAGION A. Infection of the ovum 251 B. Transplacental infection 253 C. Hereditary dystrophies.—Specific predisposition 255 D. Contamination after birth—Family contagion 258 XII TABLE OF CONTENTS PART TWO EXPERIMENTAL TUBERCULOSIS AND TUBERCLE BACILLUS INFECTION IN ANIMALS Chapter XX DIFFERENT MODES OF INOCULATION AND OF EXPERIMENTAL TUBERCULOUS INFECTION A. Experimental tuberculous infection by subcutaneous inoculation of virulent material.—Conditions of experimental infection.— Influence of the number and virulence of the infecting bacilli 263 B. Different modes of experimental inoculation and infection of the guinea pig and rabbit 268 a. lntraperitoneal inoculation 271 b. Intravascular and intracardiac inoculations 271 c. Intracranial and intraspinal inoculations 274 d. Inoculation by the eye 275 e. Infection by the digestive tract 276 f. Infection by the rectum 277 g. Infection by the bladder 278 h. Infection by the respiratory passages 278 i. Transcutaneous infection 280 k. Intramammary infection 281 l. Inoculation into gall bladder or peritoneum after laparotomy.. 281 m. Intrapleural and intra-articular inoculations, etc 282 C. Choice of experimental animals 282 Chapter XXI TUBERCLE BACILLI OF MAMMALS.—DIFFERENTIAL CHARACTERS OF HUMAN AND BOVINE TYPES A. Differential characters of human and bovine types as regards morphol- ogy and culture 285 B. Differential characters of human and bovine types in experimental inoculation 287 C. Virulence of human type bacilli for cattle.—Attempts to transform the human type into the bovine type 291 Chapter XXII BOVINE TUBERCULOSIS.—ITS PATHOLOGICO-ANATOMICAL CHARACTERS A. Pulmonary localizations 303 B. Digestive tract and abdominal viscera 305 C. Serous membranes 305 D. Lymphatic glands 307 E. Skin.,.,,, 307 TABLE OF CONTENTS XIII F. Tuberculosis of the udder 308 G. Genital organs 309 H. Other localizations 309 I. Tuberculosis in the calf 311 Chapter XXIII FREQUENCY AND GEOGRAPHIC DISTRIBUTION OF SPONTANEOUS TUBERCULOUS INFECTION IN CATTLE Occurrence of bovine tuberculous infection 312 Chapter XXIV THE SPECIFIC DIAGNOSTIC REACTIONS OF BOVINE TUBERCULOSIS A. Methods of using tuberculin 319 B. Technique of subcutaneous inoculation of tuberculin 320 C. Interpretation of results of inoculation of tuberculin 322 D. Acquired tolerance to tuberculin.—"Doping:” Its detection 323 E. Local reactions to tuberculin 324 F. Sero-diagnosis 327 Chapter XXV r6lE OF BOVINE TUBERCULOSIS IN THE INFECTION OF MAN.—-THE QUESTION OF MILK A. Tuberculous meats 338 B. The question of milk 340 Chapter XXVI SPONTANEOUS TUBERCULOUS INFECTION IN VARIOUS MAMMALS OTHER THAN MAN AND CATTLE A. Spontaneous tuberculous infection in wild animals 346 B. Tuberculosis of domestic ruminants other than cattle 351 C. Tuberculosis in the horse and in the ass 351 D. Tuberculosis of swine 354 E. Tuberculosis of the cat and dog 357 Chapter XXVII TUBERCULOUS INFECTION IN BIRDS A. Physiological characters of the avian tubercle bacillus 359 B. Pathologico-anatomic characters of tuberculosis in birds 361 C. Symptomatology and pathogenesis of tuberculous infection in birds . 363 D. Virulence of mammalian tubercle bacilli for birds 364 E. Virulence of avian bacilli for mammals 368 F. Avian bacillus tuberculin 370 XIV TABLE OF CONTENTS Chapter XXV111 ACID-FAST BACILLI OF COLD-BLOODED ANIMALS.—THEIR RELATION TO THE TUBERCLE BACILLUS A. The piscine bacillus.—Its characters 371 B. Acid-fast bacilli of reptiles and fish 373 C. Attempts at transformation of tubercle bacilli of warm-blooded animals into bacilli of the piscine type 374 D. Non-identity of tuberculous virus of warm-blooded animals and acid- fast bacilli of cold-blooded animals 376 Chapter XXIX THE ACID-FAST PSEUDO OR PARATUBERCLE BACILLI A. Special characters of the principal varieties of paratubercle bacilli. ... 379 1. Acid-fast bacilli encountered in normal man 379 2. Acid-fast bacilli of soil, sewage and excrement 380 3. Acid-fast bacilli of milk and of butter 381 B. Differential diagnosis of paratubercle acid-fasts from true tubercle bacilli 383 C. Action of paratubercle bacilli upon the evolution of tuberculosis.. .. 385 D. Experiments of J. Ferran on the transmutation of the tubercle bacillus into a saprophyte 386 PART THREE PROCESSES OF DEFENSE AND THE DIAGNOSIS OF TUBERCULOUS INFECTION Chapter XXX . REACTIONS OF DEFENSE OF THE BODY AGAINST TUBERCULOUS INFECTION.— CELLULAR ENZYMES A. Cellular protease 392 B. Cellular lipolytic ferments 393 C. Coagulins 395 D. Lysins 396 E. The phenomenon of agglutination and its practical applications 399 F. The phenomenon of precipitation.—Tuberculous precipitins and pre- cipitin diagnosis 403 TABLE OF CONTENTS XV Chapter XXXI REACTIONS OF DEFENSE OF THE BODY AGAINST TUBERCULOUS INFECTION (continued) A. Complement and antibodies. Titration of complement in blood serum. 407 B. Opsonins.—Determination of the opsonic index 410 I. Preparation of leucocytes 411 11. Preparation of the bacillus suspension 412 III. Preparation of serum to be studied 413 IV. Technique of the reaction 413 C. Clinical value of the opsonic index 416 D. Cytology of sero-fibrinous exudates and effusions in tuberculosis.— Cyto-diagnostic methods 421 Chapter XXXII REACTIONS OF DEFENSE OF THE BODY AGAINST TUBERCULOUS INFECTION (CONTINUED).—THE BLOOD AND ITS CELLULAR ELEMENTS A. The red cells.—’Resistance to hemolysis.—Anemia of the tuberculous. 424 B. The leucocytes.—The formula of Arneth 428 Chapter XXX1I1 ELIMINATION OF TUBERCLE BACILLI BY THE VARIOUS EXCRETORY PATHS A. Expectoration 433 1. Microscopic and macroscopic products of expectoration 433 2. Technique of examination for tubercle bacilli 435 3. Morphological characters of the bacilli 438 4. Mixed infections of tuberculous sputa 439 5. Control by experimental inoculation 441 6. Determination of human or bovine origin of bacilli contained in sputum 443 B. Excretion by the intestine and bile ducts 444 C. Excretion in the urine 450 D. Excretion by the mammary glands 453 Chapter XXXIV ACCESSORY REACTIONS FOR THE DIAGNOSIS OF TUBERCULOUS INFECTIONS A. Albumin reaction.—Its diagnostic value 456 B. The Abderhalden reaction 461 1. The dialyzation method.—Technique of the reaction 461 11. The optical method 463 III. Application of the Abderhalden reaction to the diagnosis of tuberculosis 464 XVI TABLE OP CONTENTS C. The reaction of activation of cobra venom (quantity of free lipoids in the serum) 465 D. Potassium iodide reaction 472 E. M6iostagmine reaction 473 F. Hypophysis test 474 G. Urinary eliminations and diagnostic reactions of urine in tuberculous infection 475 I. The uro-reaction of Malmejac 477 II. The reaction of Moriz-Weiss 479 III. The examination of the urine for tuberculin 480 Chapter XXXV PHYSIOLOGICAL ACTION OF TUBERCULINS.—MECHANISM OF TUBERCULIN REACTIONS A. Comparative toxicity of tuberculins for normal and tuberculous subjects 481 B. Relationship between susceptibility to tuberculins and the degree of infection 483 C. Mechanism of the specific action of tuberculins. Its relationship to the phenomena of anaphylaxis and anaphylatoxin action 484 D. Action of tuberculins upon the cellular elements 495 E. Resistance of the tuberculous body to tuberculin and tolerance 496 F. The influence of tuberculin upon the mobilization of bacilli in the organism 498 Chapter XXXVI DIAGNOSIS OF TUBERCULOUS INFECTION BY TUBERCULIN REACTIONS A. General or subcutaneous tuberculin reaction 500 B. Tuberculin reaction by rectal absorption 502 C. Cuti-reaction of Von Pirquet 503 D. Modification of the cuti-reaction 507 a. The procedure of Lignieres 507 b. The procedure of Lautier 508 c. Transcutaneous reaction of Moro 508 d. The rhino-reaction 508 e. Urethral and vaginal reactions 509 f. Subcutaneous local reaction (the stichreaktion) of Escherich and Hamburger 509 E. Intradermic reaction of Ch. Mantoux 510 F. The ophthalmic reaction of Wolff-Eisner-Calmette 512 G. Specificity of local tuberculin reactions 515 H. Simultaneous or successive application of the various tuberculin reactions.—Local sensitization to tuberculin 518 I. Comparison of clinical results obtained with the local tuberculin reactions.—Proportion of positive reactions in apparently normal subjects/ 521 TABLE OF CONTENTS XVII Chapter XXXVII TUBERCULOUS ANTIBODIES AND THEIR R6LE IN THE DEFENSE OF THE BODY AGAINST INFECTION A. Antigenic functions of tubercle bacilli and their secretory products.— Tuberculous antibodies.—Their demonstration by the fixation reaction of Bordet-Gengou 526 B. Preparation of washed red cells.—Preparation and titration of hemo- lytic sera.—Choice, titration and preservation of complement 528 I. Washed red cells 528 II. Hemolytic serum. Its titration 528 III. Complement. Its titration 530 C. Preparation of sera to be tested and to be titrated for antibodies 531 D. Choice and preparation of antigens.—-Measure of their value.—Their relative stability and specificity 532 E. Antigenic properties of organs, exudates, pus and glandular excretions of tuberculous subjects 540 F. Examination for antibodies or amboceptors in tuberculous sera.— Their titration 541 G. Method of obtaining sera rich in antibodies 542 H. Preventive or inhibitory effect of certain sera of tuberculous or hyper- vaccinated animals upon the fixation reaction 545 I. Nature and functions of the inhibitory substance (inhibitrice) 549 K. Examination for antibodies in extracts of organs and in tuberculous exudates 553 L. Hereditary transmission of tuberculous antibodies 554 M. The diagnostic and prognostic importance of the detection and titra- tion of antibodies in tuberculous infection 555 N. The role of antibodies in the defense of the body against tuberculous infection 559 PART FOUR NATURAL IMMUNITY AND PROCESSES OF IMMUNIZATION AGAINST TUBERCULOUS INFECTION Chapter XXXVIII NATURAL VARIANCE IN THE VIRULENCE OF THE TUBERCLE BACILLUS Variable virulence of the tubercle bacillus 565 Chapter XXXIX NATURAL IMMUNITY.—"PHENOMENON OF KOCIl” AND RESISTANCE OF TUBER- CULOUS INDIVIDUALS TO FURTHER INFECTIONS BY THE BACILLUS A. Natural immunity 570 B. The “phenomenon of Koch”—Resistance to superinfections 571 XVIII TABLE OF CONTENTS Chapter XL FREQUENCY AND GEOGRAPHIC DISTRIBUTION OF TUBERCULOUS INFECTION.— RELATIVE SUSCEPTIBILITY OF THE VARIOUS HUMAN RACES A. Mortality and morbidity from tuberculosis in Europe 584 B. Asia 591 C. Africa 593 D. Oceania 595 E. American Continent 596 F. The relative susceptibility of the various human races to tuberculous infection 598 Chapter XL1 PASSIVE IMMUNITY.—ATTEMPTS AT ANTI-TUBERCULOSIS SERO-THERAPY A. Attempts at anti-tuberculosis sero-therapy.—Mode of preparation of sera 604 I. Serum of Maragliano 607 II. Serum of Marmoreck 609 III. Serum of Vall6e 611 IV. Serum of Ruppel and Rickmann 612 V. Serum of Bruschettini 613 VI. Serum of Jousset 614 VII. Immunizing bodies (IK) of C. Spengler 614 B. The properties of anti-tuberculosis sera 617 a. Agglutinating power 617 b. Power to precipitate tuberculins 618 c. Antibody content 619 C. Mode of action and therapeutic value of the anti-tuberculosis sera 622 Chapter XL1I ACTIVE IMMUNITY.—ATTEMPTS AT VACCINATION WITH TOXINS AND TUBERCLE BACILLI A. Attempts at vaccination with tuberculin and extracts of tubercle bacilli 625 B. Attempts at vaccination with bacilli killed or modified by various chemical and physical agents 627 1. Bacilli killed by heating 627 2. Bacillary lipoids and bacilli treated with lipoid solvents 629 3. Bacilli treated with various chemical reagents 631 4. Bacilli killed or modified by light rays 634 C. Attempts at vaccination with virulent and attenuated bacilli 635 1. Bovo-vaccination of Von Behring 636 2. Tauruman of R. Koch and Schutz 641 3. The method of Heymans 643 TABLE OF CONTENTS XIX 4. The method of Klimmer 643 5. The vaccine of S. Arloing 644 6. The method of Theobald Smith 645 7. Vaccination of cattle with the avian bacillus 646 8. The method of Friedmann 646 9. The vaccine of J. Ferran 647 10. Vaccination with sensitized tubercle bacilli 648 11. Attempts at attenuation of the virulence of tubercle bacilli in the digestive tube of the leech 649 12. Attempts at vaccination with emulsion of tuberculous glands.. 649 13. The method of Bruschettini 650 14. Attempts at vaccination by the inoculation of increasing doses of virulent bacilli 651 15. Attempts at immunization by the digestive tract 652 16. Vaccination with biliated (bilies) bacilli of Calmette and Guerin 654 Chapter XLIII CHEMOTHERAPY OF TUBERCULOSIS A. Calcium salts 659 B. Gold, silver and bismuth salts 660 C. Copper salts 661 D. Rare earths of the ceric group 663 E. Radioactive substances 664 F. Arsenical compounds 665 G. Benzyl alcohol, xylol, the creosote series 665 H. Iodine and iodized compounds 666 I. Dyestuffs 667 J. Fatty acids of chaulmoogra and cod-liver oils 667 Chapter XLIV SCIENTIFIC PRINCIPLES WHICH SHOULD SERVE AS A BASIS OF PROPHYLAXIS AGAINST TUBERCULOSIS A. The tuberculizable soil (terrain tuberculisable) 672 B. How to dry up the sources of tuberculous infection 674 PLATES AND FIGURES COLORED PLATES I. Cultures of the tubercle bacillus upon solid media 32 II. Cultures of the tubercle bacillus upon glycerin broth 32 III. Genesis of giant cells and early stage tubercles 102 IV. Phagocytosis and modifications undergone by the tubercle bacil- lus within the giant cells of the gerbille (from Metchnikoff).. 104 V. Generalized tuberculous infection in the guinea pig.—Lymphatic stage of the infection produced by instillation upon the ocular conjunctiva 130 VI. Primary glandular tuberculosis in an infant of six months.— Caseous glandular tuberculosis of childhood with adherent pleura.—Caseous tuberculosis of childhood.—Cavities and disseminated tuberculous foci (massive primary infection).. 172 VII. Pulmonary tuberculosis with cavity formation and tracheo-bron- chial adenopathy in an infant of 13 months.—-Acute miliary tuberculosis in a child 178 VIII. Intestinal tuberculosis in the child (primary infection).—Intes- tinal tuberculosis in the adult (annular ulcerations).—Tuber- culosis of the large intestine in the adult (ulcers in plaques). 182 IX. Pulmonary tuberculosis (suspendue). Tuberculous nodules and multiple small cavities. Anthracotic sclerosis.—Renal tuberculosis. Cavities in the lower pole of the kidney.— Pleural symphysis. Multiple partitioned cavities of the apex. Disseminated tuberculous nodules 188 X. Chronic pulmonary tuberculosis with large and small cavities and anthracosis.—-Chronic pulmonary tuberculosis with large partitioned cavity of the apex 192 XI. Tuberculo-gummatous lupus of the nose and lips.—Cutaneous tuberculosis: tuberculous folliculitis of the back of the hand.—Papillomatous cutaneous tuberculosis of the hand.— Scrofulo-tuberculous dactylitis; fungous synovitis of the right index finger 230 XII. Generalized tuberculous infection in the guinea pig.—Tubercu- lous spleen and liver of guinea pig, in a state of cirrho-fatty degeneration 264 XIII. Tuberculous cow.—Vegetative tuberculosis of the pleura in the cow 298 JQV. Sub-mucous tuberculosis of the tongue of an ox.—Tuberculous ljver of an ox.—Tuberculous spleen of the same animal 306 XXI XXII PLATES AND FIGURES XV. Fibro-caseo-calcareous tuberculosis of the lung in the ox.—Mili- ary tuberculosis in a heifer.—Intestinal tuberculosis in the ox (ulcerations of the small intestine) 306 XVI. Fibro-caseo-calcareous tuberculosis of the lung in the ox.— Tuberculosis of the rumen.—Tuberculosis of the ovary in the cow 306 XVII. Tuberculosis of the mammary gland in the cow.—Longitudinal section of udder containing a tuberculous cavity.—Trans- verse section of the same.—Transverse section of tubercu- lous esophageal glands in the cow 310 XVIII. lntradermic reaction to tuberculin at the site of election in the cow.—lntradermic reaction to tuberculin at the site of elec- tion in swine 328 XIX. Spontaneous generalized tuberculosis of digestive origin, in a monkey 348 XX. Pulmonary tuberculosis in the horse {sarcomatous form).—'Tuber- culosis of the spleen in the horse 352 XXI. Lung of healthy swine.—Spleen of healthy swine.—Lung of tuber- culous swine.—Spleen of tuberculous swine 356 XXII. Generalized avian tuberculosis in a hen 362 XXIII. Parrot with tuberculous lesions of the crest.—Foot of a goshawk with scaly verrucous tubercles.—Tuberculous liver of a goose. 362 XXIV. Human blood cells 430 XXV. Cuti-reaction.—-Ophthalmo-reaction.—lntradermic reaction to tuberculin.—Tuberculous lupus of the cheek, in a child 504 FIGURES 1. Laennec 2 2. Villemin 4 3. Robert Koch 6 4. Tubercle bacilli in the sputum of a case of phthisis 12 5. Staining jar 15 6. Invasion of the air passages in acute miliary tuberculosis 106 7. Lesion of the pulmonary vein. Tuberculous nodules and conglom- erate miliary granulations. Tuberculous focus breaking through a vein wall 109 8. Nodular tuberculosis. Destruction of a bronchiolar wall by a caseous tubercular nodule Ill 9. Schema of apparatus used to infect the guinea pig by inhalation. ... 148 10. Miliary tuberculosis of the lung. Giant cell 187 11. Miliary tuberculosis of the lung. Two fused giant cells in a tubercu- lous nodule 189 12. Transfusion of blood from the carotid of a tuberculous guinea pig to the jugular vein of a normal guinea pig, for the study of tuber- culous bacillemia at different stages of infection 246 PLATES AND FIGURES XXIII 13. Technique of subcutaneous inoculation into the thigh of the guinea pig 265 14. Schema of the lymphatic gland system in the rabbit 269 15. Schema of the lymphatic gland system in the guinea pig 270 16. Technique of intravenous inoculation into the guinea pig 272 17. Technique of lymphatic tuberculous infection, in the guinea pig, by ocular instillation 276 18. Technique of tuberculous infection of the guinea pig by ingestion through an esophageal catheter 277 19. Apparatus for tuberculous infection of the guinea pig through inhala- tion 279 20. Schema of the abdominal lymphatic gland system in cattle 299 21. Schema of the intestinal lymphatic gland system in cattle 300 22. Schema of the lymphatic gland system of the tongue and maxillary region in cattle 301 23. Type of tuberculin reaction in tuberculous cattle 321 24. Pessary thermometer of C. Guerin for controlling tuberculin reactions in the cow 321 25. Acid-fast bacilli in fecal matter of cattle 379 26. Determination of opsonic index, by the method of Sir A. Wright.... 414 27. Phagocytosis of tubercle bacilli by polynuclear leucocytes 415 28. Permanent biliary fistula in a heifer, for the study of the elimination of tubercle bacilli 446 29. Annual tuberculosis mortality, per million inhabitants, in England and Wales 587 30. Mortality from tuberculosis in France in 1911 588 31. Intravenous inoculation of young calves with virus-vaccine 656 INTRODUCTION THE VIRUS OF TUBERCULOSIS Some Pages of History—Bayle, Laennec, Villemin and Robert Koch The origin of the virus of tuberculosis probably dates back to the days of long ago when men were beginning to live in compact social groups. Elliott Smith and Armand Ruffer, Foquet, Wood Jones and Derry,1 through the study of Egyptian mummies, have disclosed how it wrought havoc among subjects of the Rameses and the Pharaohs. In ancient times, the Veda of India, the Zend-Avesta, sacred book of the Parsees, the writings of Hippocrates, those of Celsus, of Aretaeus of Cappadocia (70 B. C.) and of Avicenna abound in documents relative to the history of phthisis. But the human disease, which was finally to receive this name was not really given a definite character until the end of the 18th century when two English physicians, Th. Reid2 (1782) and Baillie3 (1793), called attention for the first time to granulations and tubercles which increase in size and whose centers become purulent to the point of forming large abscesses in the lung substance. A little later (1810). G. L. Bayle,4 thought that he could differ- entiate the military tubercle encountered in certain cases of phthisis from some other cartilage-like granular forms which, in his opinion, produced tuberculous phthisis. To him belongs the great credit of being the first to point out that miliary tuberculosis is not a local lesion confined to the lung, but a general disease “probably identical with scrofula.” It remained for Laennec (1781-1826) to lay the real foundation 1 Bull. Archeological Survey of Nubia; 1907; Bull. Soc. Archeol. d’Alex- andrie, 1907-12. 2 Essay on the nature and cure of 'phthisis pulmonalis. Lond., 1782, Cadell. 3 The morbid anatomy of some of the most important parts of the human body. Lond., 1793, Johnson; French transl., Paris, 1803, Sampson. 4 Recherches sur la phthisie pulmonaire. Paris, 1810, Gabon. 1 2 INTRODUCTION of our knowledge of the pathological anatomy of tuberculosis. This medical genius who at the age of 35, was himself to fall a victim to the terrible malady whose study had made him illustrious, demon- strated clearly the oneness or unity of tuberculous matter, at first grey and translucent (gray granulation), then yellow and opaque, then purulent. Laennec said, “Tuberculous matter can develop in the lungs and other organs in two principal forms: as isolated bodies (granulation, Fig. 1. Laennec miliary tubercle, non-caseous tubercle, caseous tubercle, ulceration or cavity), and as an infiltration.” He used thus to differentiate the two chief anatomical types of tuberculosis which we to-day call follicular and non-follicular. Thanks to the method of mediate auscultation, of which he was the brilliant originator, he learned to detect the development of tubercles in the living subject. Humanity will be forever grateful to him for having thus created the first scientific means of diagnosing phthisis. “There is perhaps no organ,” wrote Laennec,5 “which is immune 5 De l’auscultation mediate ou traite du diagnostic des maladies des poumons et du coeur. Paris, 1819, Brosson & Chaude. THE VIRUS OF TUBERCULOSIS 3 against the development of tubercles. I shall here indicate those in which I have found them, and approximately in the order of frequency: bronchial and mediastinal glands, cervical glands, mesenteric glands, glands of all other parts of the body . . . the surface of the peritoneum and of the pleura, where very many small tubercles are ordinarily to be found in the grey and translucent, or non-caseous stage . . ., the spleen . . ., the brain . . ., the bodies of the vertebrae or the spaces between their ligaments, the thick portion of the ribs, all other bones . . . Tubercles develop more rarely in voluntary muscles than in any other part of the body. Occasion- ally, but very seldom, tubercle formation is primary in the organs just enumerated, especially in the intestinal mucous membrane and lymphatic glands, and the development of tubercles in the lungs is the result of a secondary extension.” The infectious nature of the disease therefore seemed obvious to Laennec. Furthermore he believed in the close relationship between tubercles in the lungs and tubercles in the glands, to which the name of scrofula is given, “and whose softening, as is well known, is very often followed by complete cure.” A little later, Cruveilhier6 was to go still further in this conception. To his mind, tubercles of the lungs are really scrofula of the lungs. Laennec regards the tubercle as a small tumor and Virchow,7 applying to his study the then new method of examination by the microscope, shows it to be made up of a mass of small round cells with their nuclei extending almost to the periphery as in the case of lymphoid cells of the glands or spleen. Thereafter he regards the tubercle as a lymphoid follicle, a lymphoma, whose evolution ends at times in caseation, again in calcification or fibrosis, or yet again with complete resorption resulting in cure. But, according to Virchow, the caseous infiltrations of the bung (caseous pneumonias or broncho- pneumonias) have nothing in common with the genuine tubercle, although they too produce phthisis. The latter may be due therefore either to an invasion of tubercles in the Laennec sense, or to a catarrhal or exudative inflammation bringing on obstruction of the bronchi and pulmonary alveoli. This dualistic conception had gained many followers toward the middle of the last century. In France it was supported by Ch. 6 Bull. Soc. Anatomique, 1826, p. 171. 7 Die krankhaften Geschwiilste, Bert, 1865. 4 INTRODUCTION Robin, Lorian and Empis, and Jaccoud, the latter bringing to it for a long period the weight of his great authority as a clinician. Herard and Cornil on the other hand opposed it on pathologico-ana- tomic grounds. The triumph of Laennec’s idea of the unity (Unicisme) of tubercle became certain only when Villemin8 furnished experimental proof of the invculability of the tubercle and of caseous material. Fig. 2. Villemin The date of this discovery (1865), contemporary with the celebrated work of Pasteur on so-called spontaneous generation and his first researches into silk worm diseases, marks the beginning of a brilliant era during which our knowledge of the etiology and the pathogenesis of tuberculosis was to make rapid and decisive progress. The first report of Villemin, presented on the 5th of December, 1865, 8 Etudes sur la tuberculose, preuves rationelles et experimentales de sa sp£ci- ficite et de son inoculabilite. Paris, 1868. THE VIRUS OF TUBERCULOSIS 5 at the Academy of Medicine, told of his results in inoculating human tuberculous matter into rabbits. He drew the following conclusions: “ Tubercu!osis is a specific affection. Its cause lies in an inoculable agent. It therefore belongs among the diseases of virus origin and should be classified by the side of syphilis, nearer however to glanders.’ ’ A few months later, checking himself always by the experimental method, he brought proof that the virus of pommeliere9 (tuberculosis) of cows produces in the rabbit a disease identical with that which develops when the latter animal is inoculated with the virus of human phthisis. Furthermore he showed that this virus is inoculable not only into rabbits but also into guinea pigs, and with more difficulty into dogs and cats. He did not succeed in communicating it to the sheep. Hens and pigeons were found equally refractory. During the years which followed, the facts announced by Villemin provoked the most violent controversies on all sides, and particularly at the Academy of Medicine at Paris. Colin,10 Chauffard, Piorry, Pidoux, vainly attempted to mitigate the effects of these discoveries which were tending to nothing less than the destruction of time honored doctrines. “Experiments on animals,” cried out Pidoux, ‘‘give such and such results and you, instead of controlling them by clinical experiments and by all the accepted facts of human physiology construct upon them a general theory of human tuberculosis and all diseases! For it you upset all the ideas already acquired. We must accept over night that phthisis falls from the clouds and that in its pathogenesis, the subject himself, habitus, hygienic surroundings, heredity and diatheses are of no account; that all depends upon an impossible tuberculous virus originating without doubt in a tuber- culous individual who had it from some other individual and so on to the first man, who however had it from nobody at all and must have created it himself out of nothing.” It was not long before the echo of such tirades was silenced in the brilliant proofs which came from all sides to confirm the researches of Villemin. Herard first, Gueneau de Mussy, Hardy, H. Bouley, then particularly Chauveau in France, Klebs, Cohnheim in Germany, Clark in England, all brought new facts which no one dared longer 9 Formerly used because of the resemblance of some bovine tubercles to a small apple (pomme). Bull. Acad, med., 1866, 32, 897; 1868, 33, 550; 595; 599; 603; 645. 6 INTRODUCTION dispute. In 1868, Chauveau11 was able to write, “it is now proven that the identity of tuberculosis and the diseases recognized as of virus origin is so complete and so absolute that one must either grant this property of virus origin to tuberculosis or deny that such origin exists. The conclusions which Villemin has drawn from his inoculation results have therefore the value which he has attrib- uted to them. ” Fig. 3. Robert Koch The case was settled. It remained to demonstrate the virulent agent by the methods orginated by Pasteur and perfected by Robert Koch for the isolation and for the study of pathogenic bacteria. The credit of discovering the bacillus was to fall to Robert Koch,12 whose name remains gloriously associated with it. The first publication on the discovery of the bacillus is a master- 11 Gaz. hebdomadaire, 1868, p. 753. 12 Berl. klin. Wchnschr., 1882, 8, No. 15; Mitteil. a. d. k. Gsndhtsamte, Vol. 2. The various studies published by Robert Koch on tuberculosis will be found assembled in two volumes edited by J. Schwalbe; Gesammelte Werke von Robert Koch, Leipzig, 1912, George Thieme. THE VIRUS OF TUBERCULOSIS 7 piece still fresh despite the lapse of years. It established in a definite and an irrefutable manner the bacillary etiology of tuberculosis. It demonstrated that the specific bacillus exists in the sputum of all phthisical subjects, in all tuberculous matter from man or animal, in scrofulous glands, in white swellings and in spontaneous as well as in experimental disease. And Robert Koch furnished the proof that the microbe can be readily revealed wherever it exists, thanks to the special staining measures which Weigert had introduced into histolog- ical technique; that it can be cultivated on artificial media and that these cultures reproduce in susceptible animals the same lesions which characterize spontaneous tuberculosis. “ For the future, ” concluded Robert Koch, “ in the campaign against the terrible scourge of tuberculosis, we no longer have to deal with something vague and undetermined; we are in the presence of a visi- ble and tangible parasite, the conditions of whose existence we already know in part and can now study more closely. We know that the germ finds the conditions necessary for its existence only in the body of man and animals and that it cannot, like the bacillus of anthrax, develop in matter outside the animal economy: this fact is very consoling from the point of view of combatting the disease. From it we reason that we must devote ourselves, before anything else, to drying up the sources of infection. One of the sources, and certainly the chief one, is the sputum of tuberculous individuals, which must be disinfected and rendered innocuous. In this manner the largest element in the contagion of tuberculosis will be eliminated.” The publication of this memorable communication by Robert Koch, soon rendered more precise and complete by his further researches, was destined necessarily to have a most fortunate influence upon the development of opinion in favor of the experimental method. Thanks to the rapid progress of the latter, workers in all countries, clinicians, bacteriologists, hygienists, veterinarians, passionately attacked the study of tuberculosis, and so many are the works written on the subject during the last 30 years that their mere enumer- ation would fill several volumes. The reader will pardon me then if I cite only those to which he should refer in undertaking or control- ling investigations. The initial work by Villemin and the discovery by Koch are the scientific bases of our present knowledge of tuberculous infection. I shall not here go further into their story since all that the following chapters contain is but its development and amplification. PART ONE The Tubercle Bacillus and the Processes of Infection by It CHAPTER I MORPHOLOGY OF THE TUBERCLE BACILLUS Methods of Examination, of Staining and of Differentiation Tubercle bacilli are present in every tuberculous lesion. They are found in clumps in the center of miliary tubercles, in more or less considerable number in the pus of tuberculous abscesses, in the sputum of phthisical patients, in scrofulous glands and in some serous effusions (pleura, joints, peritoneum, etc.), in the skin alterations of lupus and at times in the circulating blood. But it is not always easy to demonstrate them, especially in old calcified or fibrotic lesions. They can then he disclosed only by inoculating the ground ujp contents or walls of lesions into a susceptible animal such as the guinea pig. The bacilli are almost always enclosed within the cell elements; but when the latter die and disintegrate, the bacilli are set free and may then be excreted from the body by various normal or acci- dental excretory paths. By direct microscopic examination, even with the highest mag- nification, only the very experienced observer can determine,—and then with uncertainty—whether he has to do with the tubercle bacillus or with other microbes which have the same appearance when in the fresh state. Fortunately the nature of the tubercle bacillus can be settled by taking advantage of its property of fixing certain anilin dyes in a manner to differentiate it from coexisting bacteria or cells in tuberculous lesions or products. A. MORPHOLOGY In the sputa of phthisical patients (fig. 4), tubercle bacilli have the form of slender non-motile rods, whose average length is from a quarter to half the diameter of a human red blood cell, 1.5 to 3.5 microns (at times 0.5 micron up to 8 microns according to Eastwood1) and whose thickness is about 0.3 micron. Usually they are found 1 Rep. Royal Commission on Tuberculosis, 1907-1911. 11 12 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS singly or in groups of 2 or 3, or at times in small irregular clumps. Some are free, others are contained within polynuclear leucocytes. Only by staining can their form be well brought out, when they often appear slightly curved or beaded. Staining causes them to appear more thick, since the ectoplasm fixes the dye very intensely, although irregularly, with the result that some portions of the bacillus remain transparent while others become completely opaque. The transparent portions, from their resemblance to spores were regarded as such (G. Spengler2), but we know today that they are nothing but Fig. 4. Tubercle Bacilli in the Sputum of a Case of Phthisis Stained with Ziehl and methylene blue. Imm. ; oc. comp. 6, Reichert small masses of protoplasmic substance having the characteristics of lipoids (the Gram-positive granules of Much). The tubercle bacillus, at least during its parasitic existence, and also in cultures on artificial media, does not reproduce itself by sporulation but by elongation and transverse division of the rods. The ectoplasm is made up in part of fatty substance and a sort of wax, to which are due the double property of taking up certain anilin dyes only with difficulty when unheated and of retaining them, once taken up, in spite of the action decolorizing agents. 2 Deutsch. med. Wchnschr., 1907, 33, 337. MORPHOLOGY OF THE TUBERCLE BACILLUS 13 Cultures on artificial media always show a certain number of bacilli unusually long and thick and at times branched and with terminal clubbings, (Metchnikoff,3 later I. Klein, Fischel, Hueppe). This applies especially to old cultures on solid media and more partic- ularly to those of avian origin. Their appearance is then not unlike that of cultures from actinomycosis {Actinomyces bovis). Young cultures are both more readily stained and more readily decolorized (Nocard and Roux). Microscopic study shows that the films formed on fluid media are composed of three sorts or elements dissimilar in their dye affinities (F. Bezan§on and A. Philibert),4 but we are totally ignorant of their respective roles as regards virulence. These elements are: 1. A substance forming a sort of frame-work both membranous and fibrillar, cyanoph.il and non-acid-fast; 2. A fuchsinophil substance (bacilli properly speaking, more or less long, containing small chromophil violet granules (acid-fast); 3. Gentianophil granules, staining a violet black by the method of Gram and found either within the bacilli, or free and increasingly abundant in proportion to the age of the culture. In tuberculous lesions it is impossible to distinguish the cyanophil substance; only bacillary forms and the chromophil bodies are to be seen in the sections. In the case of certain rodents which are particularly resistant to tuberculosis, such as the Gerbille {Meriones shawii) which is found in the Sahara in Algeria and Tunis, tubercle bacilli, when introduced experimentally into the tissues, assume quite unusual forms. They have been described by Metchnikoff and will be studied later in this book (Chapter YI). These forms, which J. E. Magroux calls actinophytes, are found occasionally in naturally developed tuberculosis, but they are very rare. Coppen Jones5 has found them in the sputum of phthisical cases with cavity. Babes and Levaditi,6 and Lubarsch obtained them by inoculating human or avian bacilli into the brain of the rabbit. Friedrich, and Otto Schulze7 found them in the kidneys, in 3 Virchow’s Arch., 1888, 113, 63. 4 Bull. Soc. d’etudes scient. sur la tuberc., 1914, March 12, 32. 8 Centralbl. f. Bakt., 1895, 17, 1; 70. 6 Arch, de med. exper., 1897, 9, 1041. 7 Ztschr. f. Hyg., 1899, 31, 153. 14 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS the lungs and in the brains of rabbits which had received inoculations of tubercle bacilli into the carotid artery. Examination in the fresh state of very young cultures on fluid media shows that the bacilli possess a real motility and that they are provided with a variable number of flagella at each pole. Soon the flagella become entangled rendering the bacilli non-motile and apparently playing a capital role in the facility with which the organ- isms adhere to one another to form compact clumps. The fatty wraxy ectoplasm keeps them agglutinated and floating in dry and wrinkled, more or less thick films, on the surface of the media. The morphological characteristics of tubercle bacilli are greatly influenced by the chemical composition of the artificial media on which they are grown. They are at times shorter and more slender, or again longer and thicker. Similar variations occur also according as the bacilli are derived from old or recent lesions, from tuberculosis in active evolution or in process of cure, and furthermore according to the host which harbors them. Thus it is that bacilli of bovine origin are, generally speaking, shorter and thicker than those of human origin {Plate I). B. TECHNIQUE OF STAINING METHODS The technique for staining the tubercle bacillus has been the ob- ject of a large number of investigations. The original method employed by Robert Koch was as follows: Slides bearing tuberculous material (either sputum or smears from organs) were left immersed for 24 hours in a staining bath com- posed of: cc. Concentrated alcoholic solution of methylene blue 1.0 Distilled water 200.0 10 per cent potassium solution 0.2 They were then washed in distilled water and left in a concentrated aqueous solution of Bismarck brown a few minutes, until they had taken on a frankly brownish tint. They were again washed in water, dried and mounted in Canada balsam. The tubercle bacilli retained their blue color and stood out quite clearly against the uniform brown color of the other elements on the slides. MORPHOLOGY OF THE TUBERCLE BACILLUS 15 But this method was unreliable. Ehrlich8 soon greatly improved it by staining the bacillus with aniline methyl violet instead of alka- line methylene blue and next decolorizing the slides in a solution of nitric acid one part and water two parts. The tubercle bacilli then retain the violet color and appear almost black, while the other bac- teria and cells take on a greyish tint. But Ehrlich9 himself soon showed that aniline fuchsin (chlorhy- drate of rosanilin) gives even better results than methyl violet. The staining methods of Koch-Ehrlich are no longer in use. The Ziehl-Neelsen10 method is now preferred, and rightly so, as being quicker and more trustworthy. The staining bath of Ziehl-Neelsen is prepared as follows: One gram of fuchsin rubine is ground up in a mortar in 10 cc. of absolute alcohol. To this 5 gms. of white phenol crystals are Fig. 5. Staining Jar added and then 60 cc. of distilled water in small quantities, with constant stirring, and the whole poured into a flask. The mortar is rinsed with 40 cc. of distilled water, which is also added to the flask. The solution is allowed to stand 24 hours and is then filtered. The preparations of tissue smears or of sputum are immersed in this bath in a large Borrel tube or in any other suitable receptacle, which is then left for two hours in the incubator at 37°C. (fig. 5). Or again, if hurried, a few drops of the staining fluid may be poured directly on the slide, which is then heated in a Bunsen flame or over an alcohol lamp until the fluid steams. This heating is continued at least three minutes. The slide is next washed an instant in cold water to remove the excess of stain, and a solution of nitric acid (one part of acid to three parts of distilled water), or of sulphuric acid (one part to four parts of water), or again a solution of acetic acid (one part of acid to two 8 Deutsch. med. Wchnschr., 1882, 8, 269. 9 Charite Annalen, Berlin, vol. II, p. 123. 10 Deutsch. med. Wchnschr., 1882, 8, 451; 1883, 9, 247. 16 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS parts of 95 per cent alcohol) is poured on. This is allowed to act for 20 seconds. The slides are washed in 60 per cent alcohol until the preparation is well decolorized, and are then rinsed in water. Counterstaining is done with an aqueous solution of methylene blue, or better with the blue of Kiihne applied for 20 seconds. Methylene blue 1.5 gm. Absolute alcohol 10.0 cc. Phenol solution 5 per cent 100.0 cc. Finally the slides are washed in running water for two or three seconds, then dried and examined with the oil immersion lens. If the slide is to be preserved, the oil should be removed with xylol and the preparation dried. It can then later be re-stained if neces- sary, since the bacilli lose their color with time, after prolonged exposure to daylight. By this method of Ziehl-Neelsen, the tubercle bacilli stand out sharply red against a blue background. However, instead of employing a one-quarter strength solution of nitric acid or a one-fifth solution of sulphuric or acetic acid as a decolorizing agent after staining with carbol-fuchsin, it will be found preferable to use a 2 per cent solution of aniline chlorhydrate in water, as recommended by Kiihne.11 This reagent is much less harsh, and less destructive to the histological elements which accompany the tubercle bacillus in the preparations. It is allowed to act about 30 seconds. Decolorization is then completed in 95 per cent alcohol and the slide washed in water and counter-stained with methylene blue as already described. A solution of 1 per cent sulphite of soda, allowed to remain on the slide for one to two minutes, may also be used to decolorize. Numerous other procedures have been proposed and various ad- vantages claimed for them. I am describing only a few, stating at the outset that the method of Ziehl-Neelsen wTith the Kiihne modification is infinitely superior to all the others (see Chapter XXXIIJ). Method of Weichselbaum.12 Stain with hot carbol-fuchsin as in the Ziehl method; 11 Praktische Anleitung zum mikroskopischen Nachweis der Bakterien im tierischen Gewebe, Leipz., 1888, Gunther; Centralbl. f. Bakt., 1890, 8 , 293. 12 Wien. med. Wchnschr., 1883, 33, 63. 17 MORPHOLOGY OF THE TUBERCLE BACILLUS Wash in water; counterstain 30 seconds with a saturated alcoholic solution of methylene blue; wash in water. Method of Rondelli and Buscalioni.13 Stain with carbol fuchsin according to Ziehl; Decolorize 2 to 3 minutes with Javel water. This is prepared by dissolving 6 gms. of calcium hypochlorite in 60 cc. of water; and 12 gms. of potassium carbonate in 40 cc. of water, filtering separately and mixing before using. The background of preparations decolorized with this fluid becomes brownish: it is therefore useless to counterstain. Method of Fraenkel-Gabbet.11 Stain with the Ziehl’s carbol-fuchsin. Decolorize and counterstain simultaneously with a saturated solu- tion of methylene blue in: CC. Absolute alcohol 50 Sulphuric acid 25 Distilled water 100 Method of Muller.15 Stain with ZiehPs carbol-fuchsin; Wash in water; Decolorize in a 10 'per cent solution of bicarbonate of soda in 70 per cent alcohol at least 15 minutes, or 5 to 10 minutes in hydrogen perox- ide (oxygen 12 volumes) rendered alkaline with sodium hydroxide; Counterstain with methylene blue. Method of Von Betegh.16 Pour upon the smears, previously fixed by heat, a few drops of 15 per cent nitric acid and heat a few seconds, this serving as a mordant. Wash, then stain a few minutes with a mixture of one to two drops of Loeffler’s alkaline methylene blue and two or three drops of carbol fuchsin; heat once more. Wash in water. Decolorize in 60 per cent alcohol and counterstain for one to two minutes with malachite green (saturated aqueous solution); wash, dry and mount in balsam or cedar oil. ]3 Centralbl. f. Bakt., 1897, 21, 70. 14 Berl. klin. Wchnschr., 1884, 21, 193; 214; Lancet 1887, i, 757. 15 Centralbl. f. Bakt., 1901, 29, 791. 16 Ibid., 1908, 47, 654; 1909, 49, 461; 1909, 52, 550. 18 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS The bacilli are colored red while the granules take on a dark blue tint against a green background. Method of Herman17 (of Mons). Stain for one minute over a low Bunsen flame with a mixture of one part of a 3 per cent solution of crystal violet in absolute alcohol (or in 95 per cent methyl alcohol) and three parts of a 1 per cent solution of ammonium carbonate in distilled water. Decolorize a few seconds in 10 per cent nitric acid, then in 95 per cent alcohol, to a pale blue color. Wash quickly in distilled water and counterstain either with a 1 per cent aqueous solution of eosin, an alcoholic solution of carmine, with Bismarck brown or with a 1 per cent aqueous solution of safranin. The bacilli and the granules derived from them are stained an intense homogeneous blue, while the background is either red or brown according to the counterstain employed. This is a very good method, but a little longer and more complicated than that of Ziehl. Methods of C. Spengler.18 There are several of them. The last one proposed, which is in fact but a modification of the others, is as follows: Stain with hot carbol-fuchsin; Wash; leave in a saturated alcoholic solution of picric acid for two or three minutes. Wash in 60 per cent alcohol and in 15 per cent nitric acid for 20 to 25 seconds. Wash again in alcohol until completely decolorized. Wash in water. Treat for the last time with the picric alcohol solution. The brilliant red bacilli contrast well against the yellow. Method of Spengler, modified by P. Spehl. Stain for 2 or 3 minutes with heat, or for 15 to 30 minutes cold in a freshly prepared mixture of 3 parts of Ziehl fuchsin solution and 2 parts of gentian violet; Treat with picric alcohol (saturated solution of picric acid in water, 60 parts; alcohol, 40 parts) during one minute; Alcohol 60 per cent strength; Decolorize with weak nitric acid (1 part to 6 parts of water), then with 60 per cent alcohol; 17 Ann. de l’lnst. Pasteur, 1889, 3,160; 1908, 22, 92. 18 Deutsch. med. Wchnschr., 1907, 33, 337. MORPHOLOGY OF THE TUBERCLE BACILLUS 19 Stain for one minute in the picric alcohol; Wash in water; dry. The bacilli are red, the granules of Much are black, the background is of a pale yellow color. Methods of Much.19 This author too has proposed several methods, which are modifications of the Gram stain. The best one is this: Stain for 24 to 48 hours in the following solution: CC. Concentrated alcoholic solution of methyl violet BN 10 2 per cent solution of carbolic acid 100 Immerse 12 minutes in the iodine-iodide solution of Gram-Lugol; Treat one minute with a 5 per cent solution of nitric acid; Treat 10 seconds with 8 per cent solution of hydrochloric acid; Wash in acetone alcohol (equal parts of each) until all color is eliminated; Counterstain with dilute fuchsin solution or with 1 per cent aqueous solution of safranin (5 to 10 seconds); Wash in water; dry. This method gives good results, but it is long and complicated and the preparations do not last as well as those stained by the Ziehl method. It brings out the lipoid granules and the degenerated bacilli. These granular forms pointed out by Much are not however to be regarded as special forms of tuberculous virus. R. Bittrolff and K. Momose20 have demonstrated that they are also stainable by the Ziehl method. A. Kirchenstein has modified Much’s method as follows: Stain the bacilli by the picric acid method of Spengler (differentia- tion with the picric alcohol solution is not indispensable); Wash carefully; Stain with dahlia-violet or with methyl-violet for two or three minutes, heating to the point of steaming; Wash and decolorize in a 5 per cent solution of iodide of potassium in 80 per cent alcohol. Decolonization takes place in 10 to 15 seconds. Successfully prepared slides should appear greyish blue or pale blue to the naked eye; 19 Berlin, klin. Wchnschr., 1908, 45, 691; Beitr. z. Klin. d. Tub. 1907, 8, 85; 357. 20 Veroffentl. d. R. Koch Stift. z. Bekampf. d. Tuberk., 1913, H. 4, 18. 20 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Wash for some time and dry. Prolonged washing is necessary if one wishes lasting preparations. Any traces of iodine not removed tend to favor decolorization. By this method the granules are colored a blackish blue. Young bacilli may show a granule at each extremity. Fully developed bacilli show 5 to 9 of them. The envelope of the bacillus is barely visible. Method of Gasis. Gasis21 was guided by the idea that the tubercle bacillus by virtue of its chemical composition is more resistant to alkalies than to acids. He used first as mordant a solution of eosin prepared as follows: Eosin crystals 1 gm. Absolute alcohol 5 cc. Distilled water 95 cc. This mixture is poured into a small Erlenmeyer flask to which is then added a quantity of bichloride of mercury the size of a lentil. The mixture is gently boiled, wTith shaking, until the sublimate is completely dissolved. The solution clears and throws down a precipitate. The supernatant fluid is applied hot to the preparation for two or three minutes as a mordant. Wash in water and treat with an alkaline decolorizing reagent made up as follows: Sodium hydrate 0.5 gm. Potassium iodide 1.0 gm. 50 per cent alcohol 100.0 cc. until the red color disappears and is replaced by a greenish tint. Wash carefully in absolute alcohol to eliminate the decolorant; then rinse in water and counterstain cold for two or three seconds with an acid solution of methylene blue made up as follows: Methylene blue crystals 1.0 gm. Absolute alcohol 10.0 cc. Hydrochloric acid 0.5 cc. Distilled water 90.0 cc. Wash in water and dry. 21 Centralbl. f. Bakt., 1909. 50, 111; Berl. klin. Wchnschr., 1909, 46; 830; 1910, 47, 1449. MORPHOLOGY OF THE TUBERCLE BACILLUS 21 This method gives very brilliant preparations. The bacilli have a beautiful red color against a blue background. Finally I shall mention a technique described by Fontes (of Rio de Janeiro), which lends itself readily to the study of the chromato- phil granules of tubercle bacilli. Method of Fontes.22 Stain with Ziehl fuchsin for 2 minutes, heating to vaporization. Wash in water; stain with carbol crystal violet for 2 minutes. Cover the slide (without washing) with Gram’s iodine-iodide solution, pouring off and renewing three times; treat with acetone-alcohol until completely decolorized; wash in water; counterstain rapidly with an aqueous solution of methylene blue. Dry and examine in immersion oil. The bacillus, in cultures as well as in sputum, now appears made up of two parts: a protoplasmic portion stained red and granules stained a dark violet. The older the bacilli the more numerous are the granules (up to about 6). Young bacilli ordinarily show only a single central chromatophil granule. I do not think it worth while to describe any other procedures such as those of Kronberger,23 Yamamoto,24 Giacomi,25 and C. Biot,26 etc. which have no advantages over those already given. As I said before, the method of Ziehl-Neelsen, properly employed, fills every need and still remains the best.27 C. CONCENTRATION (hOMOGENEISATION) OF TUBERCULOUS MATERIAL When tubercle bacilli are to be sought for in material where they are present in but small numbers,—and it is then that their discovery is of the more value in clearing up a diagnosis,—it is always advanta- geous to bring about their separation from other matter (such as masses of pus, cell detritus, etc.,) in which they may be scattered and often concealed. 22 Centralbl. f. Bakt., 1909, 49, 317. 23 Beitr. z. klin. d. Tuberk., 1910, 16, 157. 24 Centralbl. f. Bakt., 1908, 47, 570. 25 Fortschr. f. Med., 1883, No. 5. 26 Gaz. des hop., 1914,.87, 42. 27 Critical summaries of the different methods of staining the tubercle bacillus will be found in the articles of Berger (Centralbl. f. Bakt., 1910, 53, 174), Dold (Arb. a. d. k. Gsndtsamte, 1911, 36, 433), and Bohm (Centralbl. f. Bakt., 1912, 62, 497) 22 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS The methods of concentration enable us to do this very well. They consist in the digestion of sputum or feces, for exnmple by various enzymes or by chemical solvents. The bacilli are not dis- solved thanks to the waxy fatty capsule, and they can afterward be collected either by simple decantation, by centrifugation, or by causing the bacilli to adhere to particles of oil or similar substances. Biedert,28 then Muhlhauser and Czaplewski,29 were the first to recommend shaking the sputum with 3 or 4 times its volume of a 0.2 per cent solution of sodium hydroxide. The tube is stoppered with a rubber cork and shaken violently. The material is poured out into a conical jar, neutralized with a 5 per cent solution of acetic acid with a few drops of phenolphthalein as indicator, and the sediment allowed to settle out; or else the material is centrifuged after the addition of two parts of 90 per cent alcohol for each part of the fluid. The sediment is then spread upon slides and stained by Ziehl. Hammerl30 treats the sputum with a mixture of equal parts of sodium and ammonium hydroxide, then shakes with a small quantity of acetone and centrifuges. Spengler,31 and Von Philipp prefer to digest the sputum 24 hours in the incubator at 37° with a small quantity of pancreatin. Yon Ellerman and Erlandsen32 bring about auto-digestion by mixing the sputum with half its volume of 0.6 per cent solution of bicarbonate of soda and incubating for 24 hours at 37°C. The mixture is to be centrifugated in a graduated tube. To one volume of the sediment are added 4 volumes of a 0.25 per cent solution of sodium hydroxide. The mixture is heated to boiling, with careful shaking, and centrifugated a final time. Bezangon and Philibert33 attach great importance to the specific gravity of the liquid from which the bacilli are to be separated. They have worked on a method the express object of which is to lower this gravity. This technique is as follows: A given volume of sputum is measured out in a graduated glass. Ten times this quantity of water is measured out in another graduated 28 Berl. klin. Wchnschr., 1886, 23, 713. 29 Deutsch. med. Wchnschr., 1891, 17, 282. 30 Munch, med. Wchnschr., 1909, 56, 1955. 31 Deutsch. med. Wchnschr., 1895, 21, 244. 3i Ztschr. f. Hyg., 1908, 61, 219. 33 Compt. rend. Soc. de biol., 1903, 55, 35; 237; 259. MORPHOLOGY OF THE TUBERCLE BACILLUS 23 glass. The sputum and half of the water are then poured into a porcelain dish and as many drops of 0.2 per cent sodium hydroxide are added as there are cubic centimeters of sputum. For example: CC. Sputum 10 Water 100 0.2 per cent sodium hydroxide 10 drops The porcelain dish is then gently heated over a Bunsen flame with constant stirring. Little by little the remainder of the 100 cc. of water is added. The whole process of heating requires about 10 minutes. The liquid is allowed to cool and its specific gravity is determined. If the gravity exceeds 1004 (that of the bacillus varying from 1010 to 1080), a little 50 per cent alcohol is added until the density falls to 999 or 1000. Enough liquid is now taken to fill two to four tubes which are centrifugated for 45 minutes to one hour. The sediment is spread upon slides and is dried and stained by the method of Ziehl-Neelsen. Sputum is much more easily dissolved by the antiformin method of Uhlenhuth and Xylander,u (-mixture of calcium hypochlorite and sodium hydroxide)35 or by the still older method (1900) of Lannoise and Girard36 with Javel water37 one-third strength or pure, hot or cold, and with sodium hydroxide. The technique which I have adopted is the following: Several cubic centimeters of sputum are mixed in a centrifuge tube with an equal quantity of a 30 per cent solution of antiformin in water, or with 5 to 10 volumes of Javel water, then shaken vigorously for two to three minutes and left in the incubator at 37°C. over night or for a few hours. The mixture then centrifugated. The superna- tant liquid is poured off and replaced by a like quantity of physiologi- cal salt solution. The mixture is again shaken and centrifugated. 34 Arb. a. d. k. Gsndhtsamte, 1909, 32, 158. 35 Antiformin is a disinfectant introduced in 1900 by Victor Tornell and Axel Sjoo, of Stockholm, for sterilizing brewery utensils. It oxidizes strongly. Javel water or pure Labarraque’s fluid may be submitted for antiformin, but their action is less rapid. 36 Presse med., 1902, May 5. 37 Also called Javelle’s water. A solution of potassium hypochlorite (trans.). 24 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS The tubercle bacilli collect in the sediment, which is spread upo slides and fixed with heat. The antiformin should be diluted with distilled water in order that no acid-fast bacteria be introduced from outside. After dissolving the sputum in antiformin, Loeffier prefers to shake the mixture with a little chloroform and alcohol (one part of chloro- form and 9 of absolute alcohol) before centrifugating. Bernhardt,38 Haserodt,39 Kawai,40 Kinyoun, Jane L. Berry and Mary A. Smeaton41 use a mineral oil, liqroin instead of chloroform. This enables one to dispense with centrifugation. With chloroform the bacilli are carried to the bottom. With ligroin, on the contrary, they are creamed and are found in the zone of separation of the two liquids, underneath the ligroin which floats above. Many modifications of these procedures have been proposed. That of Lorentz gives good results. To 5 cc. of sputum add 15 cc. of 50 per cent antiformin; shake vigorously until completely fluidified; heat until steam is emitted; centrifuge 10 minutes and spread the sediment upon slides. To do away with the air bubbles after fluidifying the sputum, it is recom- mended to add 15 cc. of alcohol and then centrifuge. The antiformin method is very practical and trustworthy and can be applied not only to sputum but also to pus from cold abscesses, to fragments of ground up body tissue and to feces. In searching for tubercle bacilli in the latter, one should take about 5 grams of material and dilute it first with 10 cc. of water; a considerable quantity of antiformin (about 30 cc. of a 40 per cent solution) is then added. After shaking, the mixture is allowed to digest for a couple of hours at 37°C., and is then centrifugated. When tubercle bacilli are sought in exudates (pleural, spinal or joint fluid) antiformin should not be used, since the cellular elements must be preserved intact for cell count and differentiation. These factors are very important for diagnosis and the centrifuge alone should here be employed. To demonstrate the presence of tubercle bacilli in the circulat- ing blood, the best method consists in drawing off 10 or 20 cc. of 38 Deutsch. med. Wchnschr., 1909, 35, 1428. 39 Hyg. Rundschau, 1909, 19, 699. 40 Med. klin. 1911, 7, 142; 186. 41 J. Infect. Dis., 1914, 14, 159. MORPHOLOGY OF THE TUBERCLE BACILLUS 25 blood from an elbow vein, with a syringe, and adding it immediately to a tube containing 10 or 12 cc. of 2 per cent sodium citrate in physio- logical salt solution. The tube is stoppered with a sterile rubber cork; it is turned upside down two or three times and put in an ice box for 24 hours. The supernatant fluid is next carefully poured off and the sediment drawn up with a pipette, either for stained prep- arations or for inoculating animals. Leon Bernard, R. Debre and Baron42 prefer to treat the blood, immediately on its being drawn, with 20 cc. of 30 per cent alcohol (to 10 cc. of blood). The laking of the cells is then completed by the gradual addition of 30 cc. of Jfi per cent alcohol. The mixture is next shaken vigorously and centrifugated for a half hour. After the supernatant fluid is poured off, the sediment is taken up in 40 cc. of 40 per cent alcohol; the mixture is shaken and one or two drops of a 1 in 10 alcoholic sodium hydroxide solution are added. The centrifu- gation is repeated. The very small sediment thus obtained is spread upon a slide and stained. When blood clots are to be examined for bacilli the method of digestion in fluorated medium, as proposed by Jousset,43 may be employed. This method serves to digest the fibrin by means of a fluorated artificial gastric juice composed of: Crystallized pepsin (Titre 50 of the French Pharmacopoeia) 1 to 2 gms. Pure glycerine 1 g. Hydrochloric acid j ® Sodium fluoride 3 gms. Distilled water 1 liter Mix 10 to 20 volumes of this fluid with one volume of clot’ leave in the incubator at 37° for three hours; centrifuge. D. THE STAINING OF SECTIONS If one wishes to study the relation of tubercle bacilli to the cellular elements in tissue, the latter should be cut into small cubes 0.5 cm. in thickness and fixed first in 60 per cent alcohol. They are then passed successively, in the course of 24 hours, through a series of 70 per cent, 80 per cent, 90 per cent and absolute alcohol. The pieces may also be fixed from the start by being put in a 10 per cent solution 42 Bull. Soc. d’etudes scient. sur la tuberc., Nov. 1912. 43 Semaine Med., 1903, 23, 22. 26 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS of formol for 24 hours and then in absolute alcohol. They are afterwards embedded in paraffin according to the usual histological technique (alcohol-xylol, pure xylol, xylol-paraffin, paraffin). The sections should be as thin as possible and should be fixed on the slides by means of a little thymolated albumen. They are dipped in ether to dissolve out the paraffin, then in absolute, 80 per cent, 60 per cent and 40 per cent alcohol successively, and in distilled water. They are finally immersed in Ziehl solution where they should remain at least one hour at 37°C., or 24 hours at laboratory temperature. After being decolorized with an alcohol-hydrochloric acid solution (1 cc. of HC1 in 100 cc. of 70 per cent alcohol) or still better in chlorhydrate of aniline (in 2 per cent aqueous solu- tion) they are washed in 95 per cent alcohol; in water, and coun- terstained with the carbolaled blue of Kiihne for one minute, they are again washed in water, dehydrated rapidly in absolute alcohol, cleared in oil of cloves and, after a final and generous bath in xylol, mounted in Canada balsam. The bacilli stand out clearly red against a blue background. It may be desirable, for purposes of study, to do a differential stain of the cell nuclei. In such a case the section is first treated for two minutes with Delafield’s hematoxylin diluted to one third strength ; then washed carefully in water and stained with Ziehl as described above, one hour at 37°C., etc. The method of Herman-Caan44 is also to be highly recommended. The technique is as follows: First stain the section with chlorhydric carmine (Mayer’s carmine) for 10 minutes. Differentiate in 1 per cent hydrochloric alcohol (1 cc. CHI pure in 100 cc. of 70 per cent alcohol) until the nuclei are distinctly apparent; wash; stain about two hours in a solution of crystal-violet in ammonium carbonate (3 parts of a 1 per cent solution of ammonium carbonate in distilled water and one part of a 3 per cent solution of crystal-violet in 95 per cent alcohol); decolorize a few seconds with a 10 per cent nitric acid solution and next treat with 95 per cent alcohol until the carmine color reappears; Dry; mount in balsam. 44 Central*)!, f. Bakt., 1909, 49, 637. MORPHOLOGY OF THE TUBERCLE BACILLUS 27 E. DIFFERENTIAL STAINING METHODS One of the essential characteristics of the tubercle bacillus is its so-called acid-fastness, in other words the property of retaining the stain, once fixed, despite treatment with acids or other decolorizing agents. This property is, however, not restricted to the tubercle bacillus. Benian45 and Hope Sherman46 conclude, from studies of both ground up and unbroken bacilli, that this quality is of a purely physical nature and exists only when the bodies of the bacilli are intact. There is to day a whole series of bacteria which are known to be equally acid-fast. Such is the case with the leprosy bacillus, dis- covered by Hansen; the bacilli of the skin to which Lustgarten, in 1884, attributed the etiology of syphilis; the smegma bacillus of Alvarez and Tavel (1885); that found by Gottstein (1886) in cerumen from the ear; certain bacilli found fairly commonly in butter and milk (Koch, Petri, Korn, Binot, L. Rabinowitsch, Kayser- ling), and in manure (the mist bacillus of Moeller).47 Acid-fast bacilli have also been found in soil (Karlinski, Moeller), in sewage (Spina, Houston) and also on certain plants (Bacilli of the graminaceae: grassbacillus, Timothy bacillus of Moeller). Bacteriologists have long discussed the question whether the vari- ous acid-fast bacilli, some pathogenic, others saprophytes,—the latter grown easily and rapidly on artificial media,—have any com- mon source or origin. The problem is not solved, but the answer up to now seems to be in the negative for the following reasons: In spite of numerous attempts on the part of the most skilful investigators, no one has succeeded in producing a tuberculous in- fection, inoculable in series from one animal to another, with any one of the above acid-fast bacteria. These bacilli, when inoculated into non-tuberculous animals, do not render the latter specifically sensitive to subsequent inocula- tions with true tubercle bacilli, a sensitization which characterizes 45 J. Path, and Bact. 1912, 17, 199. 46 J. Infect. Dis., 1913, 12, 249. 47 See the monograph of Potet on this subject; Thesis, Lyons 1902; also Weber, Arb. a. d. k. Gsndhtsamte, vol. IX; L. Rabinowitsch, Centralbl. f. Bakt., 1899, 25 77; and further in this book, chapters XXVIII and XXIX. 28 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS what we shall later study under the name of the phenomenon of Koch (Chapter XXXIX). They are not agglutinated by sera of tuberculous individuals nor can they serve as antigens against tuberculous antibodies. Their secretory products have no such toxic effect on tuberculous animals as has tuberculin made from the tubercle bacillus of Koch. Finally, inoculation of these acid-fast bacilli into the healthy organism does not produce the slightest resistance to true tuberculous infection,48 Meanwhile there is nothing to prove that bacilli in their natural cycle and by slow adaptation to their parasitic life in the animal or human body, cannot become capable of producing tubercles or of transforming themselves little by little into tubercle bacilli. But this is only an hypothesis with no underlying experimental facts. Quite different in its importance is the question of knowing on what criteria should be based the differentiation of the authentic tubercle bacillus from the acid-fast, bacilli which resemble it and which are encountered so frequently in pathological material, in the excretions or in the secretions (sputum, feces, urine), on mucous mem- branes or on healthy or diseased skin. Indeed the staining reactions do not seem to permit of this differentiation. Several of the bacilli are less resistant to decolorization with acids than is the true tubercle bacillus, but such variations in acid-fastness are too slight to permit of assurance. I should add however that it is often possible to distinguish the acid-fast bacilli found so commonly on the genital mucous membranes, especially of women (according to Grunbaum, they are found in urine of women in 59 per cent of cases) and also those of the normal skin, from the tubercle bacillus, by utilizing the technique of Dahms. This consists in immersing the slides prepared from centrifuged urine sediment for example, in absolute alcohol for three hours before any fixation whatever. The slides are next placed in a bath of 8 per cent chromic acid during 15 minutes and then stained with Ziehl fuchsin. After decolorization with nitric acid diluted to one- fourth strength or with 2 per cent chlorhydrate of anilin and alcohol, the slides are counterstained rather longer than usual (about five 48 Statements contrary to this assertion, published by Bayon (Soc. of Tropical Medicine, 1912) and by Fritzsche (Dissertation, Zurich 1908) are absolutely incorrect, according to the results of numerous experimenters. 1 have been able to satisfy myself that they are erroneous. MORPHOLOGY OF THE TUBERCLE BACILLUS 29 minutes) with a concentrated alcoholic solution of methylene blue. The acid-fast smegma bacilli have a blue tint, while the tubercle bacilli are distinctly red. Dahms insists too on the fact that the smegma bacillus never shows the curved forms which are so fre- quently observed with the tubercle bacillus. Antiformin is also excellent for differentiating tubercle bacilli from the numerous varieties of para-tubercle and acid-fast bacilli, which will be discussed later. The bacilli of smegma for example, those of Moeller and of Tobler, the acid-fasts of fecal matter, are completely dissolved (as Anna Y. Spindler-Engelsen49 of Zurich, has shown) in a half hour in 15 per cent solution, while human and bovine tubercle bacilli, even after four days of soaking in a 50 per cent solution of antiformin, are still to be recognized morphologically. The slow-worm bacillus is a little more resistant that other paratubercle bacilli. It is however dissolved in 24 hours in the 50 per cent solution. But,—and this will be the lesson of this chapter,—in the present state of our knowledge, the surest means of making an exact diagnosis lies in experimental inoculation. One should regard as tubercle bacilli only those which, when intro- duced into the body of a susceptible animal, such as the guinea pig, produce tuberculous lesions further inoculable in series, that is to say capable of consecutively infecting susceptible animals, of the same or different species (guinea-pig, monkey), when made to pass from one to another. 49 Centralbl. f. Bakt., 1915, 76, 356. CHAPTER II CULTIVATION AND ISOLATION OF THE TUBERCLE BACILLUS The tubercle bacillus exists ordinarily as a parasite of the lym- phatic cells. Its cultivation on artificial media is slow and difficult, especially in the first generations. It can however be grown without too much difficulty from tuberculous tissue in which it exists in pure state. But its isolation requires quite delicate methods if it is to be accomplished from sputum or from open lesions where other micro- organisms—such as the usual bacteria of pus—are present in large number and grow much more easily on the same media. Robert Koch,1 after numerous failures, was the first to succeed in obtaining a pure culture by crushing newly softened tubercles and smearing them over beef or sheep serum coagulated and sterilized by several successive heatings at 68°C. The serum was contained in small crystal dishes with glass covers and, after inoculation, was kept in an incubator at 37°C. On examining with a magnifying glass, after ten to fifteen days, Koch perceived some tiny grayish scaly colonies on the surface of the translucent media. When these were transplanted to other similar dishes, they produced a new growth, and this time more quickly, of irregular clumps of elevated colonies, always dry and scaly. Microscopic examination of the latter, after staining, showed bacilli identical with those in the tuber- cles from which the cultures had been taken. Furthermore their inoculation into the guinea pig and rabbit reproduced exactly the same lesions as were obtained when the ground-up pulp of tubercu- lous organs was introduced into the body of these animals, according to Villemin’s technique. Robert Koch tried other culture media. He deposited upon the surface of sterile fluid serum some of these scaly flakes from an initial culture on coagulated serum and obtained a scanty growth in the form of a thin extremely fragile film. But he did not succeed in growing the bacillus on broth or on nutrient agar, and after having described all his efforts he declared “that it is not to be hoped that 1 Mitt- a, d, k. Gsndhtsamte, 1884, 2, 66. 30 CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 31 the cultivation of the tubercle bacillus is to play any very important part in the study of the disease.” Shortly afterward, Nocard and Roux2 published in the Annals of the Pasteur Institute an important article wherein was described an improvement of technique to which we are indebted for great ad- vancement in the study of tubercle bacilli and their secretory prod- ucts. These scientists brought out the fact that glycerin added in proper proportion (5 to 8 per cent) to broth, agar or serum, renders the media particularly favorable for the growth of the germ. A. CULTIVATION OF THE TUBERCLE BACILLUS FROM PATHO- LOGICAL MATERIAL FREE FROM OTHER BACTERIA Despite improvements in technique, it is always difficult to obtain an initial culture of the bacilli, even though the inoculation be made with tuberculous tissue in which the bacilli are present in a pure state. The most approved means is by killing a tuberculous animal (by chloroform for example). The autopsy is quickly done, with all usual precautions as to asepsis, in such a way as to expose first the spleen, and then the principal groups of lymph nodes, especially those of the mediastinum. The surface of the spleen or node is seared with a red hot spatula and, through the cauterized area, some small bits are excised with a sterile scalpel. The pieces are put into a sterile rather thick glass tube, plugged with cotton, and are then minced to as finely homogeneous a pulp as possible by means of a sterile glass mixing rod with smooth tip. This pulp is smeared with a platinum spatula over the surfaces of several culture tube slants of 4 per cent glycerinated beef serum coagulated by heating at 70°C. The tubes are sealed with sterile rubber caps and in- cubated at 38°C. in an inclined, almost horizontal, position. All of the tubes thus planted do not show growth. But if they are care- fully examined after 8 to 12 days, it is found that, on the surface of the serum of certain tubes, there are some small gray slightly ele- vated masses which are the colonies of bacilli. As soon as they are quite apparent their nature and purity are verified under the micro- scope. They should then be immediately transplanted to fresh culture media, where they develop more vigorously. To this end, 2 Ann. de l’lnst. Pasteur, 1887, 1, 19. 32 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS 1, 2, 3, 4. Cultures upon 4 per cent glycerin potato 1. Human tubercle bacilli. 2. Bovine tubercle bacilli. 3. Bovine tubercle bacilli (biliated). 4. Avian tubercle bacilli. 5, 6, 7. Cultures upon 4 per cent glycerin agar 5. Human tubercle bacilli. 6. Bovine tubercle bacilli. 7. Avian tubercle bacilli. PLATE I PLATE II 1. Culture of bovine tubercle bacilli upon glycerin broth. 1' Prepara- tion from same culture. Stained with Ziehl. 2. Culture of human tubercle bacilli upon glycerin broth. 2' Prepara- tion from same culture. Stained with Ziehl. A. CALMETTE. Tuberculosis. Plate I. A. CALMETTE. Tuberculosis. Plate II. CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 33 a generous amount of the growth is taken on a flamed platinum spatula and smeared over the whole surface of a new tube of glycerin- ated coagulated serum which is in turn put in the incubator at 37°C. Usually, by the end of three or four weeks, the development of colonies is sufficiently abundant to permit of their inoculation into other tubes of serum, glycerin broth, or glycerin potato, and future cultures in series are readily obtained (Plates I and II). The serial cultures, starting from an original good growth on coagu- lated glycerin serum, are best made on glycerin potato or glycerin broth. The material which is to serve for the initial inoculations can be made richer in bacilli if, as Wedensky3 showed, the piece of tissue (spleen or lymph node), aseptically excised, is suspended by a sterile wire in a large test tube or in an Erlenmayer flask containing a little glycerin broth, so that the tissue is only partly submerged. In one or two weeks the growth of bacilli is sufficient to permit of cultures being secured more easily. a. Cultures on potato The method of culture on glycerin potato was first described by A. D. Pawlowsky4 who studied it in 1888 at the Pasteur Institute, utilizing test tubes constricted at the junction of the middle and lower thirds;-—such tubes as Roux had already been using there since 1886. Semi-cylindrical sections are cut out of large potatoes by means of a punch. After the skin is removed from the ends, each piece should be 5 to 6 cm. long and have a diameter corresponding to that of the interior of the tubes. The freshly cut sections should be quickly immersed in a dish containing a 1 per cent solution of sodium car- bonate, where they are left to soak for one or two hours, then dried on a cloth and one piece put into each test tube. The latter should have been previously filled with broth, or 5 per cent glycerinated physiological salt solution, up to the constriction level. The tubes are plugged with cotton and sterilized once in the autoclave for 30 minutes at 120°C. after which they are sealed with sterilized rubber caps to prevent evaporation. A rather stiff platinum spatula is used to inoculate the surface of the 3 Centralbl. f. Bakt., 1913, 68, 429. 4 Ann. de l’lnst. Pasteur, 1888, 2, 303. 34 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS potato. On this medium the tubercle bacillus grows more rapidly and much more abundantly than on coagulated serum, so that this medium is very useful for procuring large quantities of bacteria. In four to five weeks the surface of the potato is entirely covered with a thick coating of heaped up granular colonies which stand out irreg- ularly and are of a grayish white color; at times they are dry or again moist, according to the origin of the bacilli, whether human, bovine or avian. On some varieties of potato, which contain a little glucose, the culture takes on a pink color approaching a brick red. Glycerinated potato can serve for the initial cultivation of tubercle bacilli from the pus of cold abscesses, or from the ground up pulp of tuberculous organs free from other bacteria; but without any question the primary cultures are obtained with more certainty on glycerinated coagulated serum. In return, bacilli grown first on potato develop with much more vigor afterward when replanted upon the same media or upon glycerin broth. b. Cultures on fluid media It is only from cultures on fluid media that tuberculin can be pre- pared and the soluble excretory poisons of the bacillus studied. Consequently these media are much used in laboratories. The most widely employed is well cleared ordinary veal or beef peptone broth, to which 4 to 5 per cent of glycerin is added and which is rendered weakly alkaline, litmus serving as indicator. Meat is not indispensable. Beck replaces it to advantage and very simply by 100 cc. of serum (from horse, ox or swine) which is added to 900 cc. of water and heated to boiling for one hour (in an unbolted autoclave). After filtration the following are added to the clear fluid: gms. Citrate of magnesia 2.5 Asparagin 2.0 Glycerin 20.0 The mixture is again autoclaved for 15 to 20 minutes at 112°C. and filtered. The broth is portioned out in flat bottom flasks, each containing fluid to a depth of 2 cm. The flasks are plugged with cotton and sterilized for a half hour in the autoclave at 120°C. After cooling, each one is inoculated by carefully depositing, on the surface of the fluid, a few scaly fragments from a potato culture or a CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 35 fragment of the film from another culture on broth. This is done by means of a platinum loop or spatula. The thin films of recent growth or young cultures on solid media serve best for these plant- ings on broth. The essential point is not to submerge- the piece transplanted since, if the latter is drowned in the mass of fluid, it will not grow. The flasks are put in an incubator which can be kept at a constant temperature of 38°C. and which is rarely opened. After a few days a very delicate film can be seen to form about the transplanted fragment. At the end of 3 to 4 weeks the film covers the whole surface of the liquid and tends even to ascend on the sides of the flask; it then thickens itself by wrinkling more and more. In six weeks the growth is complete. There is no further increase. The film takes on the appearance of a layer of candle wax wrinkled like leather; it then becomes split up and at the slightest movement of the flask the fragments fall to the bottom. The subjacent fluid always remains perfectly clear. If it becomes cloudy, it is because the culture is contaminated by some other organism. Repeated reinoculations on broth should always be made with very thin films of recent growth. Attempts have been made to prepare fluid media whose chemical composition would be more constant than that of a meat infusion which always varies to a considerable degree. Kuhne, Proskauer and Beck,5 Uschinski, Hipp, Martin, C. Fraenkel, Von Schweinitz,8 Lowenstein and Pick, have proposed complex synthetic formulae into which enter, along with various mineral salts, leucin, tryosin, asparagin, glycocoll, sarcosin, hippuric acid, certain sugars (glucose, galactose, maltose) and polyatomic alcohols (dulcite and mannite). The preparation of Kiihne’s7 media is begun by mixing the following: gms. Sodium chloride 16.0 Magnesium sulphate, cryst 3.5 Calcined calcium sulphate 1.5 Calcined magnesia 2.5 Anhydrous potassium 62.13 Sodium hydroxide 7.35 5 Ztschr. f. Hyg. 1894, 18, 128. 6 Centralbl. f. Bakt., 1893, 14, 330. 7 Ztschr. f. Biol., 1892, 30, 221. 36 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS gms. Reduced iron 6.2 Phosphoric acid (specific gravity 1.3) 95.0 Lactic acid (specific gravity 1.2) 50.0 to 60.0 Distilled water 600.0 Heat to boiling. Twelve cubic centimeters of this mixture represent about 10 grams of extract of meat, that is to say the quantity necessary to prepare one liter of nutrient solution. To the 12 cc. are added (for one liter). gms, Leucin 4.0 Tyrosin 1.0 Asparagin 2.0 Ammonium succinate 2.0 Taurin : 0.5 Glycerin 40.0 Sodium chloride 5.0 The mixture is rendered weakly alkaline with sodium hydrate, is sterilized and distributed into flasks. The medium of Proskauer and Beck contains the following elements: gms. Ammonium carbonate 0.35 Magnesium sulphate 0.25 Potassium phosphate, (mono) 0.15 Glycerin 1.50 Water 100.0 For the preparation of albumin-free tuberculin (tuberculin A. F., albumosefrei), which will be discussed later (Chapter V), this medium was modified in Koch’s laboratory as follows: gms. Potassium phosphate, (mono) 0.50 Magnesium sulphate 0.06 Magnesium citrate 0.25 Asparagin 0.50 Glycerin 2.0 Caustic soda 0.25 Distilled water 100.0 The medium which I have studied with L. Massol and M. Breton is very satisfactory and is easily prepared.8 8 Compt. rend. Soc. de biol., 1909, 67, 580. CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 37 Its composition, per liter, is as follows: gms. Sodium carbonate 1.0 Ferrous sulphate 0.040 Magnesium sulphate 0.050 Potassium phosphate 1.00 Sodium chloride 8.5 Glucose 10.00 Glycerin 40.00 Witte’s peptone 10.00 Distilled water 1000.0 The peptone may be replaced by 2 gms. per 1000 of succinimid or by 2 gms. per 1000 of asparagin, but with the latter the sodium car- bonate must be left out. On this medium growth is very abundant. L. Massol and M. Breton9 have shown that glucose and levulose may be used to replace glycerin for cultures on potato. But this does not apply to saccharose. However if one takes care to invert the latter, culture is then possible and the growth is as abundant as in the presence of glucose. The tubercle bacillus therefore does not produce any invertin. Potassium and magnesium appear to be the mineral elements most necessary for the growth of the bacillus on artificial media, since there is no growth when they are lacking (Bezangon, Philibert and Boudin).10 Baudran11 has prepared another medium on a basis of glycero- phosphates, taking advantage of the fact that the bacillus utilizes phosphorus-containing organic bodies in its growth. In glycerin cultures, the glycerin is oxidized and converted into glycero-phos- phoric acid, and the latter is the source of the lecithin which is a constituent of the body of the bacillus. The medium of Baudran has the following composition: gms. Glycero-phosphate of sodium 2.24 Glycero-phosphate of calcium 1.20 Glycero-phosphate of potassium 0.60 Glycero-phosphate of magnesium 1.76 9 Compt. rend. Soc. de biol., 1911, 71, 340. 10 Bull. Soc. d’etude scient. de la tuberc., 1913, Feb. 13 11 Compt. rend. Acad, des sci., 1910, 150, 1200. 38 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS gms. Byla’s albumose 10.00 Glycerin 50.00 Sodium citrate 4.00 Water 1000.0 The glycerin may be omitted if the quantity of sodium glycero- phosphate is increased to 10 gms. and that of the sodium citrate to 8 gms. Tiffeneau and Marie12 defined the conditions of bacillary growth in a mineral medium similar to that of Proskauer and Beck and of which the formula is as follows: gms. Monopotassium phosphate 5.0 Citrate (or sulphate) of magnesium 2.5 Mannite 6.0 Ammonium sulphate 2.0 Glycerin 15.0 Water q. s. ad 1000.0 These workers found that, while cultures in peptone broth require definite alkalinity, the opposite is true with regard to the above mineral glycerin medium. The optimal acidity titrated with sodium hydroxide solution (phenolphthalein as indicator) varies between 0.05 to 0.08 per 100. A. Frouin13 obtains growth on a medium based on a combination of glucosamin and sarcosin. This medium has the following com- position : gms. Sodium chloride 6.0 Potassium chloride 0.30 Disodium phosphate 0.50 Magnesium sulphate 0.30 Calcium chloride 0.15 Glycerin 40.00 Glucosamin 2.00 Sarcosin 2.00 Water 1000.0 This solution, after being neutralized, sterilized, filtered, portioned out in flasks and sterilized anew, is perfectly suited to the growth of the tubercle bacillus. 12 Compt. rend. Soc. de biol., 1912, 72, 48. 13 Compt. rend. Soc. de biol., 1910, 68, 915. CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 39 P. Armand-Delille, A. Mayer, G. Schaeffer and E. Terroine14 have tried to determine what elements entering into the composition of peptone broth are indispensable to the growth of the germ. From their experimental results, they conclude that the purin bases appear to be of no importance, while the diamino acids (arginin, histidin) have a markedly favorable effect on the cultures, and the true ex- tractive substances are of capital importance. Among the latter the most essential are creatin, carnosin and sarcosin. The first of these has the effect of rendering the growth more abundant, while sarcosin hastens it. Inosite and glucose, the latter especially, appear also to have a distinctly favorable influence. The medium which in the minds of these investigators combines the greatest number of advantages, has the following composition: gms. Water 250.00 Sodium chloride 1.25 Magnesium citrate 0.60 Monopotassium phosphate 1.25 Glyeocoll 0.50 Aspartic acid 0.50 Nitrate of carnosin 0.10 Creatin 0.10 Sarcosin 0.10 Glucose 0.50 Inosite . 0.10 Glycerin 10.0 1 per cent sodium hydroxide solution 1.0 cc. In a more recent publication,15 the same authors state that they have obtained growths still more quickly and more abundantly on the following medium which contains a mono-amino acid (glyeocoll) and a diamino acid (arginin): gms. Water 250.00 Sodium chloride 1.25 Monopotassium phosphate 1.25 Magnesium citrate 0.60 Glucose 100 Glycerin 10.00 14 Compt. rend. Acad, des sci., 1912, 154, 537. 15 Compt. rend. Soc. de biol., 1913, 74, 272. 40 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS gms. Glycocoll 1.00 Arginin 0.50 1 per cent solution of sodium hydroxide 1 cc. (added after previous neutralization) In twelve days, on this medium, the bacilli form a complete thick wrinkled pellicle reaching up on the sides of the flask. Virulence is well preserved and the fluid on evaporation furnishes an active tuberculin. B. Sauton16 showed, as Leon Massol and I had already discovered, that it is desirable to add a small quantity of iron to the mineral media in order to obtain abundant growth. The one which he com- posed contains, per liter: gms. Asparagin 4.0 Glycerin 60.0 Citric acid 2.0 Dipotassium phosphate 0.5 Magnesium sulphate 0.5 Iron citrate (ammoniacal) 0.05 After 20 days of growth, the weight of the bacilli obtained on this medium is about 1 grn. (in the dry state) per 100 cc. of the liquid, while on glycerin broth the weight is only 0.65 gm., and on the fluid of Proskauer and Beck only 0.35 gm. According to B. Sauton the presence of iron in a proportion of 1 to 100,000 is sufficient to triple the weight of the growth. Malm (of Christiania)17 finds that tubercle bacilli refuse to grow in media devoid of phosphorus and that the presence of a small quantity of silicate of potassium exerts a very favorable influence. In fluid media deprived of albumin, the bacillus forms it, and the tuberculin obtained can be precipitated with alcohol in the form of a white powder, toxic for tuberculous animals. The alcoholic filtrate con- tains no albumin and is non-toxic. From this it is evident that the tuberculin is contained in the substance of albuminoid nature formed by the bacillus. Malm considers tuberculin as chiefly an exchange product of the tubercle bacillus and not an extract of its protoplasm. With albumin-free artificial media which contain asparagin, glycerin, and 1 per cent of mannite or dextrose, for example, Kendall 16 Compt. rend. Acad, des sci., 1912, 155, 1860. 17 CentraJbJ, f. Bakt., 1913, 70, 141. CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 41 Day and Walker18 have shown that the tubercle bacillus forms the nitrogenous elements which enter into its composition at the expense of the asparagin, and the fats and waxes at the expense of the glyc- erin and sugars. The bacillus secretes a lijpase which splits ethyl butyrate and, although weakly, castor oil. This lipase is also found in cultures of other acid-fast bacilli, such as bovine and avian tubercle bacilli, the paratubercle bacilli of the skin, smegma, grass bacilli, etc. Its presence can even be demonstrated in culture filtrates. It is thermostabile and increases in amount in proportion to culture development; but after having reached a maximum, it decreases in amount and then disappears. It is not present in old cultures. If one does not propose to study the products of secretion of the bacillus and is only trying to cultivate it from tuberculous organs, it is always preferable to employ coagulated glycerinated serum, or glycerin potato. The egg medium of Dorset,19 the preparation of which is to be described, can be used very conveniently in certain cases. c. Cultures on egg media Hens eggs are carefully scrubbed with a brush in boiled water and then immersed for some minutes in a 5 per cent carbolic acid solution. They are next taken between two sheets of sterile blotting paper, the two ends flamed and a hole made in each with pointed flamed forceps. By means of a rubber tube containing sterile cotton as a filter, the contents of the egg are blown into a previously weighed sterile Erlenmayer flask. The egg should be blown from the upper or air chamber end. A quantity of water equal to 10 per cent of the weight of the egg is next added to the flask, which is shaken to assure a homogeneous mixture of the yolk and white and the water, care being taken to avoid air bubbles. The whole is then put through heavy sterile muslin in a funnel and the filtered mass portioned out aseptically in test tubes. The tubes are coagulated by being put in a thermostat, in an inclined position, at 70°C. for two hours. They are then left in the incubator at 37°C. for three days, in order to de- tect any contamination. Finally they are sealed with rubber caps and kept in a vertical position until used. 18 J. Infect. Dis., 1914, 15, 417. 19 Amer, Med., 1902, 3, 555. 42 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Lubenau20 also highly recommends this culture medium but modi- fies its preparation in that for each egg, he adds 30 per cent of its weight of alkaline broth glycerinated to 5 per cent. The use of the egg for cultivating tubercle bacilli had already been proposed in 1896 by Capaldi. Bezangon and Griffon21 utilize only the yolk which they mix raw with glycerin agar in the proportion of one part of yolk to two parts of agar. On this medium the human bacillus grows rapidly, the colonies being moist and greasy, instead of dry and scaly as on glycerin serum or ordinary glycerin agar. Besredka,22 with the collaboration of F. Jupille, made a very in- teresting study of a mixed medium composed of 100 gms. of meat broth without peptone, to which are added cold a mixture of 20 cc. of a 10 per cent emulsion of egg white and 5 to 20 cc. of another 10 per cent emulsion of egg yolk, each mixture having been separately filtered and sterilized at 115°C. for 20 minutes. Upon this liquid medium, which is devoid of glycerin and salt, and which can be hardened, if desirable, by adding 1 per cent of agar, growth is from the outset obtained deep in the fluid and as abundantly as that of the streptococcus in broth. In 2 or 3 weeks, the cultures form a whitish membrane which completely covers the bottom of the flask and can be converted into a fine powder with a little shaking. In this fluid bovine bacilli assume a peculiar appearance of glairy threads, adhering to the sides of the flask, and of a muco-membranous consistency. Human bacilli, on the other hand, after 4 to 6 weeks, form small scales which do not adhere to the glass. The characteristic odor developed by tubercle bacilli in ordinary media is totally lacking with this medium of Besredka. And yet a very active tuberculin is elaborated, since after incubation at 38°C. for 3 to 4 weeks, 1.5 to 2 cc. of the filtered non-concentrated fluid suffices to kill tuberculous guinea pigs in less than 24 hours after intraperitoneal injection. We shall see later that the egg media are of special use when it is necessary to determine whether tubercle bacilli are of human or bovine origin. 20 Hyg. Rundschau, 1907, 17, 1456. 21 Compt. rend. Soc. de biol., 1903, 55, 603. 22 Compt. rend. Acad, des sci., 1913, 156, 1633; Ann. de l’lnst. Pasteur, 1913, 27, 1009; 1914, 28, 576. CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 43 d. Cultures on tissue It was natural to attempt to grow bacilli on animal tissues. A. and L. Lumiere23 succeeded with pieces of liver and spleen from ox or calf. The tissues were cooked for 40 minutes in an autoclave, then cut up into square prisms, washed in distilled water, immersed for one hour in 6 per cent glycerin water and finally sterilized in potato tubes for 15 minutes at 120°C. On such a medium growth is very rapid. P. Gioelli (of Genoa)24 prefers pieces of human placenta in 6 per cent glycerin broth. Growth is observed after the fourth day, no matter whether the bacilli are from cultures or from tissue pulp. e. Cultures in collodion bags and in filter bougies I have finally to call attention to the procedure of growing the bacillus in vivo, as employed by G. Moussu25 (of Alfort). This con- sists in inoculating glycerin broth in very thin sterile collodion bags which are introduced aseptically into the peritoneal cavity of the guinea pig, rabbit, sheep or ox. Instead of collodion bags, small porcelain filter bougies (Chamberland filter L2) may be used. After being filled and inoculated they are sealed with rubber and Golaz wax. The bags or the candles may be left for several weeks in the peritoneal cavity without harm to the animal. While the bacilli are developing, their dialysable secretory products are diffused out and soluble substances are taken in from the body fluids. /. Cultures on various organic media—bile media Other nutritive substances of animal or vegetable nature, whether solid or liquid, can be used for the culture of the tubercle bacillus, for example milk, its fats for the most part centrifuged off, glycerin- ated up to 2 or 3 per cent and sterilized; or again glycerin agar or glycerinated turnip or carrot. But these different media are of no special interest. It is quite otherwise as regards a method which I have studied with C. Guerin26 and which possesses the double advantage of enabling 23 Comp. rend. Soc. de biol., 1906, 60, 568. 24 Policlinico, 1907, 14, sez. med., 118. 25 Compt. rend. Soc. de biol., 1906, 61, 96. 26 Compt. rend. Acad, des sci., 1908, 147, 1456. 44 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS one to determine at the start whether the bacillus is of human or bovine origin and to progressively modify its virulence as desired. The method consists in inoculating the tuberculous material upon potatoes cooked in 5 per cent glycerin bile and leaving the culture to grow in the presence of an excess of biliary fluid. According as one uses human or bovine bile, only the human or bovine bacillus can be grown on the media. Our technique is as follows: The contents of several gall bladders, as fresh as possible,27 are put into a large round flask which is sterilized at 120°C. and put aside undisturbed for 3 weeks at laboratory temperature. An abundant sediment of brick red pigments is formed and this should be filtered out with paper just before use. Pieces of potato, cut out with a punch, are completely immersed in this bile to which 5 per cent of glycerin is added. The whole is left in a water bath at 75°C. for 3 hours. The potatoes are next allowed to drain and are portioned out among the special potato tubes, which are filled up to the line of constriction with bile glycerin- ated to 5 per cent. The tubes are finally sterilized for 30 minutes at 120°C. When sown upon potatoes prepared in this way, tubercle bacilli increase very rapidly and the growth takes on an altogether peculiar appearance which is in no way like that presented by ordinary cul- tures on glycerin potato. By the end of 10 days the whole surface is covered with a thin creamy greenish-gray layer of growth which thickens little by little to reach its maximum at the end of 45 days. At that time the potato is covered with a uniform glossy coating, of a light buff color and resembling an old culture of glanders bacilli. a7 According to Daniel Brunet and C. Rolland (Compt. rend. Acad, des sci., 1911, 153, 900) bovine bile has the following composition: gms. per 1000 Ash 12.5 to 14.3 Chlorides (as NaCl) 2.38 to 2.68 Phosphates (as P206) 1.31 to 1.58 Total nitrogen 2.3 to 2.5 Iron 0.016 to 0.018 Fat residue 27.80 to 28.80 Bile salts (tauro- and glycocholate of sodium) 15.30 to 15.80 Biliary micleo-proteid 1.15 to 2.25 Lipoids 1.100 to 2.130 Of which /^^0^ester^ns 0.410 to 0.813 \Lecithins and neutral soaps 0.690 to 1.317 CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS 45 The quantity of bacteria obtained from a single tube of potato medium is about 0.5 gm. (bacteria weighed in moist state). Bacilli so cultivated are granular, slender and somewhat longer than those grown on the usual media. They preserve the same staining characteristics with Ziehl. Weight for weight, they give a higher proportion of fatty matter soluble in alcohol. Transplanted back upon ordinary media, they quickly reassume the appearance which they normally have on each of them. On glycerin broth, however, a first inoculation below the surface yields small grumous masses not unlike cultures of actinomyces. The second transplant to broth produces a film. Such are the media most serviceable for the culture and study of tubercle bacilli, whether the latter be derived from other cultures, from body tissues or from material in which the bacilli exist in pure state. B. CULTURE OF THE BACILLUS FROM PATHOLOGICAL MATERIAL WHICH CONTAINS OTHER BACTERIA • If the tubercle bacillus is to be isolated from material in which there are other pathogenic or saprophytic microorganisms, recourse must be had to methods which permit of eliminating the latter with- out harming the vitality of the bacillus of Koch. The material may be sputum, pus from a lung cavity or from any abscess communicating with the exterior, fecal matter or urine. The procedure most to be recommended is that with antiformin, as proposed by Uhlenhuth.28 A measured quantity of sputum, 5 to 10 gms., is mixed in a sterile centrifuge tube with an equal quantity of a 30 per cent solution of antiformin. The tube is corked with rubber, shaken vigorously and allowed to stand for one hour; then, after centrifugation, the supernatant fluid is poured off and the sediment washed twice with sterile water. After the final centrifugation, the sediment is spread with a spatula on tubes of glycerin coagulated serum or on a 5 per cent glycerin potato medium. The antiformin destroys almost all the contaminating bacteria and one usually succeeds in this way in obtaining pure cultures from the start. Just as soon as the colonies are visible they should be inoculated upon new media. Weber and Dieterlen29 recommend the still simpler procedure of 38 Arb. a. d. k. Gsndhtsamte, 1909, 32, 158. 39 Tub. Arb. a. d. k. Gsndhtsamte, 1912, H. 12, 1. 46 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS treating the sputum with 4 per cent antiformin. After one hour of contact the mixture is centrifugated and the sediment inoculated. Donges30 has correctly called attention to the fact that certain strains of tubercle bacilli, both bovine and human, are more resistant to antiformin than are others. He was able to study several in which vitality was destroyed only after 12 or even 24 hours in a 15 per cent solution. The medium of Hesse31 can also be utilized. It is composed of: Heyden nutrose 5 gms. Normal solution of sodium hydroxide (crystallized at 28.6 per hundred) 5 cc. Sodium chloride 5 gms. Glycerin 30 gms. Agar 10 gms. Distilled water 1000 cc. On this medium, poured into Petri dishes, a pure growth is fre- quently visible within 4 to 6 days, provided the sputum is inoculated immediately after being freshly collected in a sterile container. The mouth of the patient should be washed beforehand with dilute hydro- gen peroxide. The medium proposed by S. A. Petroff,32 at the 1915 meeting of the Society of the American Bacteriologists, is likewise to be recom- nended. According to Petroff, it enables one to quickly isolate tubercle bacilli from sputum by virtue of the inhibitive action of gentian violet on the other microorganisms. An infusion is first prepared from 500 gms. of beef or veal in 500 cc. of watfeir containing 15 per cent of glycerin. The meat is left to macerate for 24 hours; it is then put through a sterilized press and the fluid collected in a sterile vessel. Meanwhile the shells of several clean fresh eggs are sterilized by immersing them for 10 minutes in 70°C. alcohol, and the eggs are broken into a sterile vessel. The whites and the yolks, mixed by shaking, are filtered through sterile gauze on a funnel, and the egg mixture added part for part to the meat infusion. To this last combination a 1 per cent alcoholic solution of gentian violet is added in the proportion of 1 to 10,000. The medium is tubed and kept in an inclined position in an oven 30 Ztschr. f. Hyg., 1913, 75, 185. 31 Ztschr. f. Hyg.., 1899, 31, 502. 32 J. Exper. Med., 1915, 21, 38. 47 CULTIVATION AND ISOLATION OF TUBERCLE BACILLUS at 85°C. until completely hardened. On each of the two succeeding days it is kept at 75°C. for one hour. To grow the bacillus from sputum, the latter (as fresh as possible) is diluted with an equal quantity of a 3 per cent solution of sodium hydroxide, left for a half hour in the incubator at 37°, neutralized to litmus with hydrochloric acid, centrifugated and the sediment divided among the culture tubes with a drawn out pipette. C. Spengler (of Davos)33 published in 1903 an ingenious procedure based on the fact that formaldehyde exercises its antiseptic action on saprophytic bacteria much more rapidly than on tubercle bacilli. His method is as follows: In the bottom of a Petri dish, completely covered beforehand with a layer of filter paper, about 3 cubic centimeters of sputum are spread to a depth of two and a half millimeters. Pancreatin is sprinkled on the sputum to hasten digestion of the mucus. The inside of the cover of the dish is also fitted with a layer of filter paper which, after the action of the pancreatin (5 or 6 hours in the incubator at 37°C.), is moistened with 3 to 5 drops of commercial formalin. The dish is then kept at a temperature of 20 to 25°C. for two hours. This suffices to kill all the bacteria, except the tubercle bacilli, which cart be directly planted upon coagulated glycerin serum in order to obtain pure cultures. S. Piatkowski,34 inspired by this method, proceeds in the following manner: He emulsifies the bacterial mixture to be studied in 10 cc. of water or broth; to this he adds 2 or 3 drops of formalin and shakes vigor- ously. Half an hour later, the fluid (or better, one part of the sediment after centrifugation) is inoculated upon ordinary agar while another part is inoculated upon glycerin agar. The tube is shaken anew and the sowings repeated every quarter of an hour. Thus he obtains a series of tubes of which some may give a pure culture of acid-fast bacilli, all the other bacteria having been killed by the formol. Many acid-fast bacilli which do not produce tuberculosis grow rapidly on ordinary agar. They are thus readily differentiated. C. Spengler showed that sputa can be incompletely sterilized by heat in a manner to kill all bacteria except the tubercle bacillus. An amount of nummular sputum the size of a hazel nut is taken 33 Ztschr. f. Hyg., 1903, 42, 90. 34 Deutsch. med. Wchnschr., 1904, 30, 878. 48 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS up on a platinum loop and held to the flame in such a way that it is roasted over its whole surface without being detached from the loop. This procedure is repeated two or three times and the desired sterilization is obtained. The flamed mass is then put upon glycerinated serum and crushed. This method is possible but it requires practice. The antiformin method is certainly much more simple and more practical. I have still to mention the method studied by F. W. Twort who isolates the tubercle bacillus directly by means of ericoline, a gluco- side which kills the associated bacteria. A fragment of sputum is put into a 2 per cent aqueous solution of this substance, kept at 38°C. for one hour and inoculated directly upon Dorset’s egg medium. There is also the method of F. Ditthorn and W. Schultz35 which consists in diluting the sputum with an equal quantity of water, adding a 15 per cent caustic potash solution up to 10 per cent of the volume of the sputum water mixture, then heating for 10 to 20 minutes in a water bath at 47 to 50° in order to render the whole homogeneous. To 30 cc. of this emulsion are added 1.5 to 2 cc. of a 2 per cent solution of iron oxychloride. A precipitate is formed which is collected and spread on slides without the necessity of centrifugating. All of these methods, except that of Hesse and that of C. Spengler, are applicable to the isolation of tubercle bacilli from fecal matter, from urine, from the pus of lung cavities, from pneumothorax or from abscesses. For fecal matter the antiformin method is the most suitable. One should take about 10 gms. of material which is diluted in 20 cc. of sterile water and left in 30 cc. of a 20 per cent antiformin solution for about one hour. After being shaken vigorously, the mixture is centrifugated with aseptic precautions. The sediment is washed in sterile physiological salt solution, centrifugated a second time and (after pouring off the supernatant fluid) inoculated upon the different solid culture media (coagulated serum, potato, Dorset’s egg, Lubenau’s egg and potato with bile). For urine, a sufficiently large quantity (about 100 cc.) must be centrifugated to begin with. The sediment is taken up in several cubic centimeters of water to which is added an equal quantity of 20 per cent antiformin. The further procedure should be as already described. 35 Centralbl. f. Bakt. 1917, 79, 166. CHAPTER III INFLUENCE OF PHYSICAL AND CHEMICAL AGENTS UPON THE TUBERCLE BACILLUS A. ACTION OF AIR AND OF ATMOSPHERIC PRESSURE The tubercle bacillus is very largely aerobic and can be grown only in media and in containers which permit of plenty of air. Further- more, within the body it is in the very vascular organs, to which the blood stream brings a large supply of oxygen, that the bacillus de- velops most readily. However it is not a strict aerobe, since it vege- tates under relatively anaerobic conditions, for example in the tissues of certain viscera such as the spleen, the liver and the kidney. Hueppe,1 too, was able to cultivate it by direct planting upon the whole egg, the minute inoculation hole in the shell being carefully sealed with wax. By maintaining cultures during a series of successive generations at a temperature of 38°C. and under a pressure of two and one-half atmospheres, S. Arloing2 demonstrated that the bacilli assume elongated and irregular forms, frequently taking on a clubbed or conical shape and resolving themselves after a while into granules or spherical bodies. If such modified forms are again placed under ordinary atmospheric pressure and on new media, they reproduce normal bacilli. B. ACTION OF LIGHT AND ULTRA VIOLET RAYS As early as 1890, in a communication to the International Congress at Berlin, Robert Koch called attention to the fact that tubercle bacilli die rather quickly when directly exposed to the sun’s rays, and more slowly when exposed to diffuse light. Cultures are very sensitive to it. Two hours of summer sun are sufficient to render them sterile (I. Straus). Bacilli contained in sputum and spread thinly on glass slides are destroyed in 10 minutes under the same 1 Internat. Congr. Hyg. & Dem., 9th, Lond., 1891. 2 Compt. rend. Acad, des Sci., 1908, 146, 100. 49 50 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS conditions. According to Migneco’s3 experiments, linen and woolen cloths, soiled with tuberculous matter, then dried and cut into small squares and introduced under the skin of guinea pigs, are no longer capable of infecting the animals after 24 to 30 hours of exposure to full daylight. Ultra-violet rays are strongly bactericidal. According to M. and Mme. Victor Henri and Baroni,4 the rays emitted by the mercury vapor quartz lamp cause the bacilli to lose their acid-fastness in a few minutes and suffice to kill them in ten minutes: although the Gram-positive granules of Much remain and can still be stained (Rochaix and Colin).5 C. ACTION OF LOW TEMPERATURES A dry cold temperature, even liquid air itself, does not destroy either the vitality or the virulence of the tubercle bacillus. Accord- ing to Galtier, Cadeac and Malet, and Moussu,6 temperatures in the neighborhood of minus 180°C, (temperature of liquid air and of liquid nitrogen), prolonged from a few hours to 8 days, with alter- nate freezings and thawings as well, do not affect the vitality of bacilli contained in tuberculous lesions. D. ACTION OF HEAT The optimal temperature for cultivating the tubercle bacillus of mammals is 38°C. Above 42°C. and below 30°C. growth ceases. The avian bacillus on the other hand grows very actively between 40° and 42°C. Growth ceases at 45°C. The thermal requirements of these bacteria are therefore very rigorous. Consequently if one wishes abundant and uniform cultures, one must have a well regu- lated incubator the doors of which are opened as seldom as possible. High temperatures have very different effects depending on whether the bacilli are exposed in a dry or moist environment, or in cultures or tuberculous products. In order to definitely determine their resistance in the dry state, Krumwiede7 triturated cultures 3 Riforma med., 1895, No. 169, 227. 4 Compt. rend. Acad, des sci., 1910, 151, 724. 6 Ibid., 1911, 153, 1253; 1530. 6 Compt. rend, de la Caisse nat. d. recherches sci., 1912. 7 J. Infect. Dis., 1911, 9, 115. INFLUENCE OF PHYSICAL AND CHEMICAL AGENTS 51 from eight different sources, dried them in an incubator for 24 hours and then ground them anew. They were portioned out in narrow drawn out tubes, not sealed, but bent at one end. Each tube con- tained 25 mgms. of bacteria. The dried organisms were now heated at 100°C. on a water bath for varying periods, air being allowed to escape freely from the tubes. After cooling, suspensions in physiolog- ical salt solution were injected into guinea pigs. The animals inocu- lated with bacteria heated 20 minutes showed slightly active cir- cumscribed lesions. Those inoculated with bacilli heated 45 minutes remained definitely free from infection. Similar experiments, under a variety of conditions, had previously been performed by Schill and Fischer, Woelsch, Straus and Gamaleia, Yersin, J. Forster, Th. Smith, Grancher and Ledoux-Lebard, and others. The conclusion reached was that above 80°C. the vitality of the bacillus diminishes rapidly; the higher the temperature or the longer the exposure, the quicker the loss of vitality. In a moist environment, a prolonged exposure of 12 hours at 50°C. or 4 hours at 55° C., one hour at 60° C., 15 minutes at 65°C., 10 minutes at 70°C., 5 minutes at 80°C. or one minute at 95°C. is sufficient to destroy without fail the vitality of the bacillus in culture suspensions. Heat acts less rapidly on tuberculous products because of the pro- tection afforded by the albumin. At every temperature also, the duration of the heating is a factor of prime importance and the degree of heat must be maintained throughout the entire mass of fluid. As regards milk, Forster8 insists quite rightly on the unreliability of the heating contrivances used and on the necessity of maintaining a temperature of 70°C. for at least 30 minutes, if one wishes to be sure of destroying the bacilli. E. EFFECT OF DESICCATION Tuberculous sputum smeared on glass slides or on cloth and then dried at low temperatures in a dim light, retains its virulence from two to four months. If exposed to diffuse light, as in an apartment for example, it remains virulent for about 39 days, according to Twitchell. Beyond 60 days it can no longer infect guinea pigs. In the dry dust of offices or public places, streets, etc., in poor light, the bacilli survive scarcely longer than 10 days. This also 8 Centralbl. f. Bakt., 1909, 51,417. 52 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS applies to bacilli deposited upon the pages of books or on clothes (F. Kirstein9). P. Chausse10 smeared drops of tuberculous sputum over slides which he dried, some at room temperature, and others in an incuba- tor at 37°C. He then again took the material up in suspension and tested its comparative virulence for guinea pigs, both by injection and inhalation. He found that the two test methods give very differ- ent results. When the tests are made by injection, the bacilli dried at 15 to 20°C. in diffused light may infect the animal after 30 to 40 days; those dried in the dark remain virulent up to 60 days; those dried in the incubator are no longer virulent after 15 days. When the test is carried out by inhalation it is found that sputa dried at ordi- nary temperature are non-infective after 10 days and that those dried in the incubator at 37°C. are harmless after 4 days. Chausse concludes from his experiments that the vitality of tubercle bacilli in sputa persists long enough (seven days on the average) to make it necessary to destroy them. But since this persistence of vitality is not very protracted, it may be said that disinfection of the home, as a prophylactic measure,—which involves certain difficulties of execution, —can he replaced to advantage by the necessary regulations regarding spitting. F. INFLUENCE OF PUTREFACTION Galtier, Cadeac and Malet, and A. Gaertner performed numerous experiments with a view to determining the duration of the vitality of the bacilli in cadavers and in tuberculous organs in a state of putre- faction. After 167 days of burial under ground, virulence was not destroyed. Schottelius11 states that the bacilli remain alive several years in tuberculous corpses. The phenomena of disintegration of sewage, whether in septic tanks or spread as manure, destroy them only with extreme slowness. According to Musehold,12 and also to my own experiments, bacilli may continue virulent in wet mud for more than 4 months, and in garden soil for more than 7 months. A. Davos, F. Jessen, and Lydia Rabinowitsch13 were able to infect guinea pigs by injecting particles of slime collected 100 meters below sewer outlets. 9 Ztschr. f. Hyg., 1905, 50, 186. 10Compt. rend. Acad, des sci., 1912, 155, 486. 11 Deutsch. med. Wchnschr., 1890, 16, 226. 12 Arb. a. d. k. Gsndhtsamte, 1900, 17, 56. 13 Berl. klin. Wchnschr., 1910, 47, 878. INFLUENCE OF PHYSICAL AND CHEMICAL AGENTS 53 G. ACTION OF ELECTRICITY AND OF OZONE The passage of a continuous current of a few milliamperes through a solution of ammoniacal salts or of amino bodies containing tubercle bacilli in suspension has the effect of attracting the organisms in mass to the cathode. Ch. Russ14 thought to utilize this property for the detection of bacilli in suspected fluids where the bacilli were too few to be revealed by direct examination of slides. A little ethylamine is dissolved in the liquid (urine, milk, etc.), and the current passed through. Whatever gathers upon the cathode (the latter being enclosed in a glass tube) is collected, smeared and stained. Dry air charged with ozone to a concentration of 4 to 6 mgms. per liter, destroys the vitality of the bacillus within a few minutes (Marmier and Abraham). H. EFFECT OF AGE UPON CULTURES Long ago laboratory experience showed that the virulence of a tubercle bacillus culture falls off rather rapidly if the organisms are not frequently replanted upon new media. As a general rule the resowings should be made at least every month. The cultures, after being left in the incubator at 38°C. for 6 to 8 weeks, have already lost a large part of their vitality. Virulence persists for a variable period according to the source of the bacillus. Those of avian origin are more stable than bovine, which in turn are more stable than human bacilli. The susceptibility of the latter is probably due to the fact that they produce more acid (Th. Smith). Old cultures, of 6 or more months, reproduce themselves only exceptionally. However they can still infect susceptible animals, although the development of lesions may be very slow. In general, after 8 to 10 months, almost all the bacterial elements are either dead or so devitalized that they can no longer be revived. By regularly reinoculating a very virulent culture of bovine origin every three or four weeks on glycerin potato medium, we have pre- served it unaltered for more than 15 years. The same dose of this culture (3 mgm.), suspended and injected intravenously into calves aged from 6 to 12 months, produces constantly an acute miliary infection with death in 28 to 35 days. The solid culture media, particularly glycerin potato, sustain the 14 Brit. J. Tuberc., 1910, 5, 26. 54 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS virulence of the bacillus much better than do the fluid media. This is especially true if care is taken to remove the solid media cultures from the incubator after four weeks and to keep them in the cold. I. ACTION OF VARIOUS CHEMICAL AGENTS The more important chemical substances and the proportions which hinder or arrest the growth of pure cultures of the tubercle bacillus are: 1 in 100 potassium iodide; 1 in 900 arsenious acid, 1 in 900 boric acid, and the ammonia vapors (Villemin). Carbolic acid in 5 per cent strength kills the bacillus in 5 minutes; the same acid in 1 per cent concentration kills it almost as quickly; corrosive sublimate, 1 to 100 kills in one hour, 1 to 1,000 in 24 hours. Many other antiseptics have a bactericidal action upon the tubercle bacillus; for example 1 per cent tricresol; 2 per cent lysol; and formalin (1 per cent in one hour, 0.01 per cent in 24 hours). The same is true of several anilin dyes such as fuchsin, gentian violet, methylene blue and auramin. On the other hand iodoform, dusted upon cul- ture media, hardly disturbs the growth and becomes inhibitive only when massive doses (up to 5 per cent) are added. Indeed every one knows that surgeons have long remarked the favorable effects of iodoform dressings in the treatment of tuberculous abscesses, despite the fact that experimentally this substance is totally devoid of efficacy. Baumgarten, Rovsing in Salomonsen’s laboratory at Copenhagen, Troje and Tangl, Catrin, then Stchegoleff15 under the direction of I. Straus, have brought many proofs of this fact, by mixing large quantities of iodoform either with cultures or with tuberculous products which were then inoculated into animals subcu- taneously or intraperitoneally, or even into the anterior eye chamber of the rabbit, without the evolution of the infection being pre- vented; it was only, now and then, rather delayed. One must there- fore conclude that iodoform acts by altering the tissues immediately about the tuberculous focus being treated, and not as an antiseptic. Chloral in 1 per cent strength, oil of turpentine, benzin, terpin, creosote vapor, toluene and eucalyptus vapor scarcely retard cul- ture growth (Villemin16). The virulence of tuberculous products is destroyed only after 24 16 Arch, de med. exper., 1894, 6, 813. 18 Rev. de Verneuil, t. II, p. 237 (“Etude exp6rimentale et clinique sur la tuber culose”). INFLUENCE OF PHYSICAL AND CHEMICAL AGENTS 55 to 36 hours by a 1 to 500 solution of salicylic acid; after two days by 8 to 10 times their weight of hydrogen peroxide (oxygen 12 volumes); or after two days by 1 to 1000 brorpine water (Parrot and Hipp, Martin). Moreover the bacilli contained in mucous sputum, or in albuminous matter in general, are much more resistant to antiseptics than are those taken from cultures. They are killed beyond question only after at least 24 hours in 5 per cent carbolic solution or in a 1 to 1000 sublimate solution; after two hours in 2 per cent tricresol, in 4 per cent lysol, in 15 per cent formalin and in 2 per cent calcium chloride. The sulphur dioxide set free by burning 60 grams of sulphur to the cubic meter kills in 24 hours of contact (Thoinot), but in 6 hours with the Clayton furnace with a concentration of at least 6 per cent in the air of the infected room (Calmette). The most accurate work regarding the effect of different antisep- tics on avian tubercle bacilli in pure culture is that of Yersin.17 With a drawn out glass pipette he took up a little culture from the surface of a tube of glycerin agar of 15 days’ growth and transferred it quickly into a test tube containing the antiseptic solution. After varying intervals of time he drew up a bit of the sediment in a pipette and put it into a test'tube full of distilled water. A few hours later the material to be inoculated, well washed and freed from antiseptic, was transferred to a flask of glycerin broth kept ready in the incu- bator at 39°. Yersin then noted what flasks showed growth. By this method the following results were obtained: ANTISEPTICS PROPORTION IN THOUSANDTHS COMPLETE BACTERICI- DAL ACTION Carbolic acid 50 30 seconds Carbolic acid 10 1 minute Absolute alcohol 1000 5 minutes Iodoform ether 10 5 minutes Ether 1000 10 minutes Bichloride of mercury 1 10 minutes Thymol 3 2 hours Water saturated with creosote Incomplete Incomplete 6 hours Water saturated with naphthol Salicylic acid 2.5 Boric acid 40 Incomplete after twelve hours 17 Ann. de l’Inst. Pasteur, 1888, 2, 60. 56 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS According to the studies of Moussu and Goupil,18 of Uhlenhuth and Xylander,19 C. Fraenkel and E. Baumann,20 and F. M. Schmitt,21 chlorin, in the form of Javel water or of antiformin, has a very pronounced action on the tubercle bacillus. Not only does the chlorin quickly destroy its virulence, but it also combines with its component elements to modify its original state. After the mixing, hydrochloric acid is found to be formed in appreciable amount. The bacilli, stained by the Ziehl method, no longer resist the action of dilute acids; acid-fastness has disappeared. By making numerous tests at successive intervals from the beginning of the action of the chlorin, one can detect the gradual diminution and finally the loss of acid-fastness. Since this quality, as will be seen later, persists despite the action of fat solvents and even survives the action of potassium solutions raised to the boiling point, it appears that chlorin therefore by itself produces a profound alteration in the con- stitution of the bacillus. With iodin, the alteration is less marked. The bacilli treated with an excess of Lugol’s solution for 15 to 30 minutes, then washed, centrifugated and taken up in physiological salt solution, are still toxic for cellular elements (J. Nicolau22). Urea, when added to culture media in small proportions, has a definite inhibitive action, according to Rappin.23 On the other hand, sodium chloride, even in proportions up to 15 or 25 per cent, is said not to affect the vitality of the bacillus. M. Muller (of Strass- burg) proposed to utilize this property of salt to preserve pathologi- cal material, and also milk which was to be sent to the laboratory for examination for tubercle bacilli. Certain metals, especially in the colloidal state, appear to have an unfavorable effect on the growth of the bacillus on nutritive media. Robert Koch24 called attention long ago to the inhibitive effects of gold cyanide diluted to one in one million or even to one in two mil- 18 Compt. rend. Acad. des sci., 1907, 145, 1231. 19 Arb. a. d. k. Gsndhtsamte, 1909, 32, 158. 20 Ztschr. f. Hyg., 1906, 54, 247. 21 Ztschr. f. Infektionskr d. Haustiere, 1912, 11, 321; 401. 22 Compt. rend. Soc. de biol., 1914, 77, 178. 23 Ibid., 1901, 53, 691. 24 Internat. Congr. on Med., 10th, Berk, 1890. INFLUENCE OF PHYSICAL AND CHEMICAL AGENTS 57 lion parts. Bruck and Gluck,25 A. Feldt,26 Bettmann,27 Junker,28 Arthur Mayer,29 L. Hauck,30 and M. Breton,31 have studied experi- mentally, as well as clinically, the effects of different gold preparations (double cyanide of gold and potassium, colloidal gold, or arseniated colloidal gold as prepared by Fourneau at the Pasteur Institute). Results were at times favorable, but more often negative. The lack of stability of gold salts makes it impossible for them to penetrate into the organism without being decomposed. The same is true of copper salts (cyanide, sulphocyanide, sulphate, tribasic phosphate, acetate), which have been recently tried by different experimenters, especially by Finkler and the Countess Von Linden (of Bonn),32 A. Strauss, Meisen, S. Pekanowich. This holds like- wise for colloidal preparations of platinum, silver, palladium, rhodium, selenium, etc., studied by Paul Courmont and A. Dufourt.33 Renon,34 without any more success, has utilized the salts of nickel, yttrium, zirconium, as well as nickel, silicon and ruthenium in colloidal state. He has also tried certain anilin dyes, either pure or combined with iodin (iodized methylene blue). In Renon’s hands too, thorium (nitrate, sulphate, chloride) meso- ihorium, and the bromide and sulphate of radium had no inhibitive action on cultures in vitro, nor any therapeutic activity experimentally. A. Frouin has studied the influence of rare earths on the growth of tubercle bacilli. He finds that if, to the following medium, which is like that of Proskauer and Beck plus a little more lactose, gms. Distilled water 1000.0 Asparagin 5.0 Lactose 3.0 Glycerin... 40.0 Sodium citrate 1.5 Dipotassium phosphate 1.0 Magnesium sulphate 1.0 25 Miinchen. med. Wchnschr., 1913, 60, 57. 26 Deutsch. med. Wchnschr., 1913, 39, 549. 27 Miinchen. med. Wchnschr., 1913, 60, 798. 28 Miinchen. med. Wchnschr., 1913, 60, 1376. 29 Deutsch. med. Wchnschr., 1913, 39, 1678. 30 Miinchen. med. Wchnschr., 1913, 60, 1824. 31 Compt. rend. Soc. de biol., 1913, 74, 1200. 32 Internat. Congr. on Tuberculosis, 10th, Rome, 1912. 33 Compt. rend. Soc. de biol., 1913, 75, 454. 34 Rev. gen. de clin. et de therap., 1903, 17, 353; Compt. rend, de la Caisse nat. des recherches sci., 1913. 58 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS is added 0.4 gm. of sodium vanadate per liter, a much more abundant bacterial growth is obtained. The gain is less if ten times as much vanadate is added.35 The sulphates of cerium, lanthanum, neodymium, 'praseodymium, samarium, in doses of 0.05 gm. per liter of nutritive media, also favor the growth of tubercle bacilli. In a concentration of 1 to 1000 the sulphates of neodymium and praseodymium completely- prevented growth. In any case these salts cannot replace magnesium in so far as the latter is necessary for growth. V. Henri considers that the salts employed byFrouin act as catalys- ers and that, in large doses, they inhibit growth through the pro- duction of hydrogen peroxide or of other substances with very high oxidizing potential. This inhibiting action on the tubercle bacillus is lessened if easily oxidized bodies (glycerin, glucose) be added to the culture media. Knowing well the very marked influence of the radio-active salts of thorium and of uranium on the germination of grains and the later development of their organs, F. Becquerel36 thought that these same substances might have an analogous effect on tubercle bacilli. To settle the point he incorporated into the usual culture media variable quantities of solutions of uranium or thorium nitrate. He found that small doses, up to a maximum of 0.4 mgm. of the uranium salt or 0.04 mgm. of the thorium salt per cubic centimeter of liquid, affect favorably the rapidity and quantity of film growth. Beyond these doses, however, the toxic and inhibitive action becomes manifest. A. Frouin,37 who also investigated the same problem, did not secure exactly the same results. Working with thorium sulphate instead of the nitrate, and with uranium acetate he arrived at the conclusion that uranium does not favor the growth of tubercle bacilli, while thorium does favor it slightly. 35 Compt. rend. Soc. de biol., 1912, 72, 1034. 36 Compt. rend. Acad, des sci., 1913, 156, 164. 37 Compt. rend. Soc. de biol., 1913, 74, 282. CHAPTER IV CHEMICAL COMPOSITION OF THE TUBERCLE BACILLUS The average weight of one tubercle bacillus is about 2.5 one hundredth millionths of a milligram. There are approximately 40 million bacilli in 1 mgm. of growth on artificial media, whether solid or fluid, when weighed in the fresh state after incomplete drying between double thicknesses of blotting paper. According to Nebel,1 the specific, gravity of the tubercle bacillus varies between 1010 and 1080. The organisms have therefore a great tendency to fall to the bottom of culture fluids. They remain on the surface,—a condition essential for their growth on the majority of artificial media,—only when they exist in masses composed princi- pally of young organisms. The waxy fatty ectoplasm which com- prises in part each of the microorganisms prevents their becoming soaked and ensures their clumping together. The water content of the bacilli has been determined by Hammer- schlag2 as averaging 85.9 per cent. After being extracted with alcohol and ether, dried at 100°C. and calcined, the bacilli yield about 8 per cent of ash, which, accord- ing to the analyses of Schweinitz and Marion Dorset3 contains per 100 parts: A. MINERAL COMPOSITION Sodium 13.62 Potassium 6.35 Calcium 12.64 Magnesium 11.55 Silicon 0.57 Phosphoric acid 55.23 The absence of chlorin and sulphuric acid is probably due to the preliminary washing in hot water. Furthermore these figures are only approximate, since the chemical composition of the bacilli varies somewhat with their origin and the artificial culture media. 1 Arch. f. Hyg., 1903, 47, 57. 2 Centralbl. f. klin. Med., 1891, 12, 9. 3 Centralbl. f. Bakt., Ref., 1903, 33, 278. 59 60 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS whether solid or fluid, from which they are taken. This explains why the same investigators have found the proportion of phosphoric acid per 100 parts of ash to be as follows: For bovine bacilli 58.04 For bacilli of swine 56.48 For bacilli of the horse 55.40 For avian bacilli 55.63 For attenuated human bacilli 74.38 For virulent human bacilli 60.90 The phosphorus content parallels the variations in fats. The proportion of total ash to the weight of fresh bacilli varies between 2.31 and 3.96 per cent. Hammerschlag gives 2.55 per cent. Kressling gives the following composition for bacilli dried at 100° to 110°C. per cent Albuminoid substances 53.59 Fatty substances.. 38.95 Non-nitrogenous substances (by subtraction) 0.972 Nitrogen : 8.575 Ash. 2.55 Von Behring4 cites the results of analyses of bovine bacilli by Zincke. Ash content was to the weight of dry bacilli as 6.91-7.3 to 100. Combined calcium and magnesium phosphates made up 4.2 per cent. The insoluble fraction contained some carbonate and some phosphate of calcium, some phosphate of magnesium and traces of iron too small to be calculated. Siebert5 has made a comparative study of the ash content of peptonized meat broth without glycerin and of bacilli cultivated on this same broth to which glycerin was added. The results obtained were as follows: ASH OF THE BROTH total ash: 28.26 per CENT OF THE DRY RESIDUE ASH OF THE BACILLI TOTAL ash: 7.52 PER CENT OF THE BACILLI DRIED AT 110°C. Chlorin 43.98 6.60 Phosphoric acid 8.57 51.25 Sulphuric acid 2.25 0.84 Silicic acid 0.26 0.19 Sodium 29.69 9.18 Potassium 14.69 26.55 Magnesium 0.41 3.22 Calcium 0.15 2.17 4 Behringwerke Mitt., 1907. 5 Centralbl. f. Bakt., 1909, 61, 305. CHEMICAL COMPOSITION OF THE TUBERCLE BACILLUS 61 Bouveault6 also has made comparative analyses of glycerin broth before and after growth of avian bacilli. He found that the latter assimilates more readily the simplest nitrogen bodies, those approach- ing the amins and ammonia, while sarcosin and the complex albumi- noid compounds, such as gelatin and peptone, are recovered almost intact. As non-nitrogenous food they consume chiefly glycerin. The various sugars are scarcely touched. B. SUBSTANCES EXTRACTIBLE BY FAT AND WAX SOLVENTS Of all known bacteria, the tubercle bacillus yields the largest quan- tity of fatty and waxy bodies. To them, in large degree, it owes the property of acid-fastness and its chief biological characteristics. These substances, very complex in constitution, are difficult to extract completely, a fact which explains the variable results in the analyses made by those who have undertaken their study. The proportion of fats and waxes is considerably influenced, moreover, by the age of the cultures, the origin of the bacilli and the various media used for cultivation. Hammerschlag7 estimates that the substances extractible by fat solvents represent on the average 27.2 per cent of the weight of the dry bacilli. Aronson8 puts the proportion at 20 to 25 per cent, Giaxa9 at 35.2 to 40.4 per cent; Levene10 at 31.56 per cent; Krebs11 at 22 per cent; Yon Schweinitz and Dorset at 37 to 42 per cent; Kressling12 at 25 to 40 per cent; Ruppel13 at 8 to 26.5 per cent; Baudran at 36 to 44 per cent; and Nicolle and Alilaire at 39 per cent. It should be remarked that for extraction some of the investigators employed alcohol and ether, while others made use of ether, or alcohol and chloroform; and still others added benzol. Aronson was the first to use a mixture of ether, alcohol and hydrochloric acid, then recently trichlorethylene, the bacilli being treated in a shaking apparatus at a temperature of 37°C. 6 These, Paris, 1892. 7 Centralbl. f. klin. Med., 1891, 12, 9. 8 Berl. klin. Wchnschr., 1898, 35, 484; 1910, 47, 1617. 9 Centralbl. f. Bakt., Ref., 1901, 30, 670. 10 J. Med. Research, 1901, 6, 135. 11 Centralbl. f. Bakt., 1896, 20, 488. 12 Ibid., Ref., 1901, 30, 1200. 13 Ztschr. f. physiol. Chem., 1898, 26, 218. 62 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS E. Roux and A. Borrel, at the Pasteur Institute, extracted the fats from bacilli by first boiling them in a weak solution of hydro- chloric acid, then drying and treating them with hot xylol in an extraction apparatus. The preliminary treatment with acid can be replaced by heating in a drying oven at 140 to 150°C. The bacilli subjected to extraction with boiling xylol lose completely their prop- erty of acid-fastness. The total fatty substance, when so extracted and then spread upon slides, stains very intensely by Ziehl and is acid-fast. Auclair and Paris14 remove the fatty and waxy substance from the tubercle bacillus by treating successively with alcohol, ether and chloroform, allowing each solvent to act four days at 35°C. The alcohol dissolves out the stainable matter, a lipoid analogous to lecithin, some fatty acids and some alkaloidal substances yielding a chloroplatinate. The ether dissolves the neutral fats and a sub- stance analogous to cholesterol, while the chloroform takes up a part of the latter substance and some waxy matter. Petroleum ether, in a Soxhlet apparatus or in a ball shaker, does not dissolve these products. From 8.18 gms. of dry bacteria, it dissolves 11.552 per cent; the alcohol acting afterward dissolves 5.708 per cent; the ether 14.975 per cent; and the chloroform 1.594 per cent; a total of 33.826 per cent of the whole bacterial mass. If the petroleum ether be dispensed with, the alcohol dissolves 17.260 per cent, the total of the first two fractions. In the opinion of Auclair and Paris, acid-fastness is not a reaction limited to the waxy fatty matter; it persists in bacilli from which the fat has been removed and it belongs particularly to the cellulose frame- work and to the protoplasm of the microbe. The substance which holds together the clumps of bacteria (zooglea) is weakly acid-fast and gives the reactions of cellulose, since it resists boiling potassium and takes a blue color with iodine after treatment with sulphuric acid. With hydrochloric alcohol or with benzaldehyde Deycke15 ex- tracted a neutral fat which he named tuberculonastine. Von Schweinitz and Dorset saponified the fats, then distilled them in a medium acidified with sulphuric acid. In this manner they obtained some traces of volatile fatty acids and a larger quantity 14 Compt. rend. Acad, des sci., 1907, 144 , 278. 15 Mimchen. med. Wchnschr., 1910, 57, 633. CHEMICAL COMPOSITION OF THE TUBERCLE BACILLUS 63 of fixed fatty acids, some of them melting at 62°C. (those of pal- mitic acid); others with a melting point of 102°C. (those of arachidic acid); and finally others melting at 43°C. and which were identified as of lauric acid. According to G. Camus and Ph. Pagniez,16 the property of acid-fastness is due to these fatty acids and they are formed only gradually by the bacillus, since young organisms show either no acid-fastness at all or but little. In their opinion, tubercle bacilli, whether from cultures or from sputum, are more or less intensely stainable with blue, shading from, a pale blue to a black, when subjected to the test proposed by Benda as specific for fatty acids. The material if fixed on the slide with heat, is treated with a hot saturated solution of sub-acetate of copper and washed generously with water, then put into a saturated solution of hematoxylin and differentiated with a very weak solution of potas- sium ferrocyanide and borax.17 Robert Koch and Proskauer found that the bacilli contain an acid substance insoluble in cold alcohol, but soluble in boiling alcohol and ether and which, in their opinion, is a non-saturated fatty acid. Aronson regards this as a true wax. He found that after treating this substance with boiling potassic alcohol, as for the saponification of waxes, there still remains an insoluble portion. This residue seems to be a higher fatty alcohol not identical with cholesterol and which stains with Ziehl’s fuchsin. It is apparently a mixture of ceryl (C27H560) and myricyl (C3oH620) alcohols. According to Kressling,18 the fatty mass contains 14.38 per cent of free fatty acids, 77.25 per cent of a mixture of neutral fats with higher fatty alcohols, and 8.37 per cent of water-soluble substances. Their melting point is about 46°C. Their ash contains a fairly large quantity of phosphoric acid derived particularly from lecithin. There is also some cholesterol. The higher fatty alcohols have a melting point of from 43.5 to 44°C. and represent 39.1 per cent of the total fats. W. Bullock and J. R. Macleod19 used boiling methyl alcohol to extract the fats from several kilograms of dried bacilli. On being left to cool, this extract threw down a whitish precipitate; this is the 16 Presse Med., 1907, i, 65. 17 Compt. rend. Soc. de biol., 1905, 59, 386; 701; 703. 18 Centralbl. f. Bakt., 1901, 30, 897. 19 J. Hyg., 1904, 4, 1. 64 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS acid-fast substance, which can be saponified and thus broken down into fatty acids and a snowy powder which chemical analysis shows to be an alcohol. They were able to obtain 1 gm. of it in a pure state. To this alcohol, in the last analysis, they attribute the acid-fast properties of the tubercle bacillus. Dorset and Emery20 also, separated from the non-saponifiable portion of the ether extract, an acid-fast alcohol which they thought belonged to the aliphatic series. Moreover Fontes21 was able to extract some waxes by means of xylol, and among them he identi- fied various alcohols; cholesterol, isocholesterol and phytosterin. Aronson believes that the waxes are primarily a secretory product of the bacilli which unites them one to another, and that the bodies of the bacteria themselves do not contain any of it in their protoplasm. W. T. Ritchie22 thinks that the best solvents for the waxes of the tubercle bacillus are benzol with boiling for 48 hours, or benzol at ordinary temperature for 32 days or, still better, toluol with boiling for 8 hours. The waxes, as I have been able to convince myself, are easily emul- sified by heating gently near the melting point and rubbing up in an agate mortar (previously warmed in hot water) with a small quantity of egg yolk or beef bile, or simply with blood serum and physiological salt solution. The emulsion remains stable. It is agglutinated by neutral salts or by acids, as are the corresponding emulsions of beeswax. White and Gammon23 questioned whether any relationship existed between the fat of the tubercle bacillus and that of animals. It is known that the fat of adult man consists chiefly of palmitin and olein whereas that of the child is more rich in stearin, like that of cattle. These authors planted tubercle bacilli on 5 per cent glycerin agar in which they mixed, as thoroughly as possible, 6 to 20 per cent of various fats; control tubes were p.lso made. The following facts were brought out: human fat and butter improve media designed for the human bacillus. On tubes to which human fat has been added, the growth is a hundred times more abundant than on ordinary glycerin agar. Linseed and olive oil are rather unfavorable. For the bovine 20 Bur. Anim. Indust., Ann. Rep., Wash., 1904. 21 Centralbl. f. Bakt., 1909, 49, 317. 22 J. Path. & Bact., 1905, 10, 334. 23 J. Med. Research, 1912: 26, 257 CHEMICAL COMPOSITION OF THE TUBERCLE BACILLUS 65 bacillus, human fat, butter and olive oil are favorable, while linseed oil is rather inhibitive. Beef fat does not help either the human or the bovine bacillus. When the favorable fats are emulsified before- hand with an extract of liver, the advantage gained is said to be still greater. C. CARBOHYDRATES On treating the bodies of bacilli (previously freed of fat by alcohol and ether) with 1 per cent sodium hydroxide, and then again taking them up in concentrated sulphuric acid, a precipitate forms which dissolves in an alkaline cupric solution and which behaves like a cellulose in the opinion of Hammerschlag, like hemicellulose according to Dreyer and Marpmann, and Nishima. Auclair and Paris24 regard it as a hydrocellulose, since it assumes a blue color with iodine solu- tion. Bendix,25 by treating the bacterial bodies with a hot 5 per cent hydrochloric acid solution, obtained a substance which gives with osazone the characteristic reaction of the pentoses. In his opinion, the pentoses are derived from the decomposition of the bacillary nucleoproteins. Ruppel and Helbing regard the residue (about 8.3 per cent of the weight of the dry bacilli) obtained after treatment with strong mineral acids as a protein-like body analogous to keratin, chitin or to fibroin. Incidentally, the chitin which covers, the eggs of the tenia gives the same color reactions as does the tubercle bacillus. According to Baudran,26 the bacillus contains a cellulose which, on being treated with equal parts of chloride of zinc and hydrochloric acid, is dissolved and gives the blue reaction with iodine. Perman- ganate at a temperature of 36°C. breaks it down into acetic and butyric acids. It is said to make up 3.6 to 5.5 per cent of the weight of the bacilli. D. PROTEIN SUBSTANCES The complete solubility of tubercle bacilli in certain chemical substances makes possible the study of their protein constitution. Robert Koch used concentrated alkalies for this purpose, but they are too strong and too destructive. Hammerschlag, used a 1 per cent 24 Arch, de med. exper., 1907, 19, 129; 1908, 20, 737. 25 Deutsch. med. Wchnschr., 1901, 27, 18. 26 Compt. rend. Acad, des sci., 1906, 142, 657; 1910, 150, 1200. 66 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS solution of potassium hydroxide. The solution of bacterial cells saturated with acetic acid and precipitated with ammonium sulphate yielded an albuminoid substance which reacts like the xantho- proteins with Millon’s reagent, and gives a positive biuret test. Th. Weyl,27 by means of acetic acid, separated a kind of mucin from an alkaline extract of bacilli which had been cultivated on glycerin agar; this substance he named toxomucin. According to Klebs, the bacillus is made up for the most part of nuclein. If, after being rid of their fats with alcohol-ether, the bacil- lary bodies are treated with hydrochloric pepsin and then with alcohol a substance 'containing from 8 to 9 per cent of phosphorus is pre- cipitated. Ruppel28 washes the bacilli with water, then treats them with dilute alkali and thus separates out the nucleoproteins. Baudran macer- ates the bodies of the bacteria for 8 to 10 days in a solution of 1 per cent hydrochloric acid at 80°C. and finds that there is thus dissolved out an amount of albuminoid substances equal approximately to 50 or 60 per cent of the weight of the bacilli. Auclair and Paris, after extracting the fats, treat the bacilli with concentrated acetic acid; they thus obtain a paranucleoalbumin which they call bacillo-casein and which has especially toxic proper- ties. When injected under the skin of an animal, it produces within 24 hours a nodule ‘having the anatomical characteristics of the early gray tubercle; the corresponding glands become swollen and, later develop visceral lesions which reproduce, in the lung particularly, the appearance of grey tuberculous pneumonia. As for its general poisonous effect, it induces in the guinea pig a rapid emaciation, a progressive anemia, then cachexia and death, the latter intervening after about three months and after a single injection (see Chapter V). As early as 1897 Yon Schweinitz and Dorset had isolated from cultures on fluid media, a crystalline substance soluble in ether, alcohol and water, and possessed of very characteristic necrotizing properties for the liver, when injected into guinea pigs. They regarded it as a non-saturated acid of the fatty series. R. Koch29 attempted to extract the bacterial substance by grinding. After the mass had been finely triturated in an agate mortar and sub- 27 Deutsch. med. Wchnschr., 1891, 17, 256. 28 Ztschr. f. physiol. Chem., 1898, 26, 218. 29 Deutsch. med. Wchnschr., 1897, 23, 209; 1901, 27, 829. CHEMICAL COMPOSITION OF THE TUBEECLE BACILLUS 67 jected to extraction with distilled water, it furnished a special poison (tuberculin TR) which Koch employed in the treatment of tubercu- losis. This tuberculin will be studied later. According to Ruppel,30 the ground up bacilli on being extracted with water, furnish two distinct chemical substances: one of them, which can be precipitated by acetic acid, is a nucleoprotein which he called “ tuberculosamine.” The other is a nucleic acid which might be called tuberculinic acid. The latter is rich in phosphorus, no longer gives the reaction of albuminoids, contains no sulphur and behaves like nucleic acids from other sources. Its phosphorus content is from 9.2 to 9.4 per cent. On boiling on a water bath it throws down nucleic bases (probably guanin, a little xanthin, and adenin) and a phosphorated acid, tuberculo-thyminic acid. Yon Behring31 regarded tuberculinic acid as the true tuberculous toxin to which tuberculin owes its characteristic effects. Kitashima32 thought, to the contrary, that this effect is due to the thyminic acid group. Levene33 considers that the tubercle bacillus contains both free and combined nucleic acid. In order to obtain the free acid, he dries and pulverizes the bacilli and submits them to repeated extractions in solutions of 5 per cent sodium chloride and 5 per cent ammonium hydrochlorate. The extracts are then treated with picric acid and acidified with acetic acid. The precipitate collected by filtration, is washed with alcohol, redissolved in water and again precipitated with alcohol. In this manner there is obtained a perfectly white sediment which is taken up in water acidulated with acetic acid. When treated with a solution of copper chloride another precipitate forms which is washed with water until no more copper remains; with alcohol until it contains no more chlorine; and finally with ether after which it is dried in vacuo in the presence of sulphuric acid. Desiccation is car- ried out in an oven at 105°C. until its weight remains constant. The residue, after treatment with the sodium chloride, is treated for two hours with a 4 per cent solution of sodium hydroxide, and then neutralized with acetic acid. An excess of picric acid is added and acetic acid is used to acidify. After filtering, alcohol is added to 30 Ztschr. f. physiol. Chem., 1898, 26, 218; Beitr. z. exp. Therap., 1900, H. 4, 88. 31 Behringswerke Mitt., 1907. 32 Centralbl. f. Bakt., Ref., 1903, 33, 727. 33 J. Med. Research, 1901. 6, 135 68 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS the filtrate and a precipitate is formed which is redissolved and precipitated anew. This precipitate fails to give the biuret reaction and possesses all the properties of nucleic acid. When heated with mineral acids it does not reduce Fehling’s solution. By dissolving in water with a little alcohol and acidifying the solution with acetic acid a cuprous salt of nucleic acid is obtained. By this method, starting with different cultural strains, Levene has obtained nucleic acids in which the phosphorus content varies from 6.58 to 13.9 per cent. V. C. Vaughan and S. M. Wheeler used an alcoholic soda solution and extracted two substances from fat-free tubercle bacilli: one of them, toxic and soluble in alcohol; the other non-toxic and insoluble in alcohol. The toxic portion kills guinea pigs in doses of 75 to 100 mgm. The action of hydrogen peroxide was studied by Mine. Sieber- Choumov,34 because of its property of decomposing and reducing bodies of high molecular weight such as keratin and animal and vegetable pigments. This author found that heating at 143°C under a pressure of three atmospheres permits hydrogen peroxide to com- pletely dissolve bacilli previously dried and rid of fat. For one gram of bacilli it is necessary to use 300 to 350 cc. of hydrogen peroxide (oxygen 15 volumes) and to keep the mixture in an autoclave for about 2 hours. Such treatment gives a colorless liquid without residue and apparently of no toxicity. Much and Deycke35 completely dissolve the bacilli by means of cholin and neurin, and without previously extracting their fat. Lecithin and other lipoids (alkaline oleates) can also be utilized for this purpose. Ditthorn noted that this dissolving process is very slow at a temperature of 15 to 20°C., but it is almost complete in four days at 57°C., provided one uses concentrated solutions of neurin (25 per cent solution, Merck). Neurin is a strongly alkaline substance requiring 22 cc. of a decinormal solution or 1 cc. of an 8 per cent solution of hydrochloric acid to completely neutralize 1 cc. Salimbeni36 had better results with certain glycerin ethers, especially with monchlorhydrin. This solvent, which is soluble in water, enables one to treat the bacilli directly and without previously drying. With 34 Compt. rend. Soc. de biol., 1913, 74, 478. 35 Miinchen. med. Wchnschr., 1909, 56, 1825. 36 Compt. rend. Acad, des sci., 1912, 155, 368. CHEMICAL COMPOSITION OF THE TUBERCLE BACILLUS 69 it one can observe a series of modifications, more or less rapid and profound, according to the proportion between the bacterial mass, the ethers employed and the number of their acid radicals. If equal parts of bacilli and mono or dichlorhydrin (the latter is but slightly soluble in water) are rubbed up in a mortar, the bacterial mass becomes converted in a few seconds into an oily, homogeneous, more or less sticky paste. If a few more drops of reagent be added the paste is transformed into a very turbid fluid which clears itself more and more as the proportion of glyceride is increased. When trichlorhydrin is allowed to act upon the dried, finely-ground bacteria, a very fine emulsion is immediately obtained which is more transparent than those prepared with the mono and dichlorhydrin. Taking every- thing into account, the rapidity of the action, the appearance of the emulsions and the fineness and transparency of the bacterial clumps in suspension in the liquid, it is evident that the action of trichlorhydrin on tubercle bacilli is stronger and more complete than that of dichlorhydrin, and that the latter in turn is more active than the mono compound. When treated with trichlorhydrin, the bacilli lose their acid- fastness in a few minutes, become granular and are transformed into an amorphous mass which no longer takes the stain. If treated with water, this matter gives up a fairly large quantity of a soluble substance which can be precipitated by 3 volumes of absolute alcohol. The water-insoluble residue contains the nitrogenous elements of the bacteria, along with the fats and the waxes taken up by the glyceride water which had dissolved them. It gives the characteristic reac- tions of nitrogen and of albuminoid substances. In the water- soluble portion, on the contrary, no trace of nitrogenous substance can be demonstrated. In addition to this remarkable dissolving action, the glycerin ethers possess, for the tubercle bacillus, a bactericidal action so strong that only a few seconds of contact suffice to kill the microorganisms and to render them, after washing with water, innocuous for the guinea pig. It is seen then that our knowledge of the chemical composition of of the tubercle bacillus, although incomplete, is nevertheless sufficient to permit of the extraction of the essential elements, so that the study of the physiological action of each of them becomes possible. CHAPTER V TOXINS OF THE TUBERCLE BACILLUS.—EXO- AND ENDOTOXINS.—TUBERCULINS Tubercle bacilli contain toxic substances which are set free by prolonged maceration or dissociation of the bacterial elements, whether by grinding or by the aid of certain solvents (such as caustic alkalies, neurin, mono, di or trichlorhydrin). These toxic substances are closely connected with the protoplasm. They are altered, but not destroyed by boiling. When injected into normal animals, they cause the formation of smaller or larger abscesses, according to the dose inoculated. In doses sufficiently large, they produce a slow intoxication which can terminate in cachexia and in death. A. TOXICITY" OF DEAD TUBERCLE BACILLI—ENDOTUBERCULINS— VOLATILE POISONS OF THE TUBERCLE BACILLUS Whole bacilli, killed by heat, possess the same properties as the bacillary extracts. This fact was first demonstrated by Maffucci.1 By inoculating eggs with killed cultures of avian tubercle bacilli and then incubating, this author obtained chicks which were cachectic, but were free from tuberculosis. He also observed that if he injected dead cultures of human tubercle bacilli subcutaneously into guinea pigs, an abscess was formed at the point of inoculation and the animals died in a state of cachexia after a period varying from 15 days to 6 months, according to the quantity inoculated. At autopsy he found lesions of atrophic sclerosis of the liver and of the spleen, but no tubercles. The same lesions and death by cachexia could be produced by feeding guinea pigs with killed cultures. The accuracy of these facts has since been verified.2 , A little later, Prudden and E. Hodenpyl published the results of some interesting experiments showing that dead bacilli when intro- duced in capillary tubes under the skin of rabbits, exerted a positive chemotactic influence on the leucocytes, and that the same dead 1 Centralbl. f. allgem. Path. u. Path. Anat., 1890, 1, 825. 2 Compt. rend. Acad, des sci., 1906, 142, 441. 70 TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 71 bacilli, injected intravenously, provoked the formation of nodules which apparently had all the characteristics of tubercles in the lungs, but which did not go on to caseation. On microscopic examination stainable tubercle bacilli were readily demonstrated in the lesions. By injecting dead bacilli directly into the trachea, the same authors demonstrated the formation of foci of hepatization characterized by masses of leucocytes and epithelioid cells with giant cells. The foci never became caseous and ended by being absorbed after encapsula- tion with fibrous tissue. I. Straus and Gamaleia3 carried out a similar series of investiga- tions utilizing cultures of the human bacillus killed at 115°C. in an autoclave. They likewise observed the formation of real miliary granulomata in the rabbit lung after intravenous injection and in the peritoneum after intraperitoneal injection. On injecting a heavy suspension of the bacillus subcutaneously into the rabbit, an abscess was produced, which opened spontaneously after several weeks and discharged a creamy pus: but the corresponding lymph nodes showed no tumefaction, contrary to what is always found after subcutaneous inoculation of living bacilli. Cultures heated at 130°C. for one hour during each of ten succes- sive days, or submitted to several prolonged boilings in absolute alcohol, retain the same properties. This is also true of bacilli from which the fats have been removed with methyl alcohol and petro- leum ether (Cantacuzene) .4 The fundamental characteristic of the lesions produced by these dead bacilli, in animals free from any pre-existing tuberculous in- fection, is that they remain localized at the points where the bacilli were deposited in the body; they never become generalized. These lesions of localized tuberculosis are evidently due to intra-cellular poisons; that is to endotoxins of the tubercle bacillus. Grancher and Ledoux-Lebard,5 Vissmann,6 Kostenitsch,7 Masur and Kockel,8 Krompecher, Kelber, Engelhardt, and Baumgarten obtained similar results, both with avian bacilli and with cultures killed and stained with fuchsin. 3 Arch, de med. exper. et d’anat. path., 1891, 3, 705. 4 Ann. de l’lnst. Pasteur, 1905, 19, 699. 5 Arch, de m6d. exper., 1892, 4, 1. 6 Virchow’s Arch., 1892, 129, 163. 7 Arch, de med. exper., 1893, 5, 1. 8 Ann, de l’lnst. Pasteur, 1900, 14, 723. 72 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Kossel, and Weber and Heuss9 killed human bacilli with heat and caused them to be inhaled by the ox. They observed that, although this animal is not susceptible to the same bacillus alive, it reacts nevertheless to the endotoxin by forming local pulmonary lesions. To this tuberculous endotoxin Von Behring10 gives the name of Somatine. Protoplasmic endotoxins Auclair and Paris11 believe that they have isolated the true endo- tuberculin formed by the protoplasm of the bacillus. This substance behaves as a paranudeoalbumin and its chemical properties are indistinguishable from those of casein. They have named it bacillo- casein. Injected subcutaneously, it produces within 24 hours a nodule similar to the gray granulation; the corresponding glands become swollen and, some time afterward, visceral lesions appear which, in the lung particularly, simulate gray tuberculous pneu- monia. In addition there is-caused a rapid emaciation, a progressive anemia, cachexia and death. With a single dose death occurs in about 3 months. The technique employed by Auclair and Paris to separate the dif- ferent protoplasmic poisons from the bacilli was as follows: The bacteria are first washed in water and then allowed to macerate in distilled water for 24 hours in an incubator at 38°C. The filtrate now contains a very small quantity of albumins and some albumoses. The albumins are precipitated with difficulty by a saturated solution of ammonium sulphate; the albumoses remain in the liquid. The bacilli are next left to macerate under the same conditions and for the same period of time in a 1 per cent solution of sodium chloride. The filtrate then contains traces of globulins which are precipitated by saturated magnesium sulphate and collected by dialysis. The above steps must be conducted aseptically. These three classes of substances constitute the soluble poisons of the bacillus. When heated on a water bath for several hours, they possess the properties of tuberculin. We have here the T. V. of von Behring. The bacilli after having been dried in the air are placed in the 9 Tuberk.-Arb. a. d. k. Gsndhtsamte, 1905, H. 3, 30. 10 Behringswerke Mitt., 1907. 11 Arch, de med. exp6r., 1908, 20, 737; Bull. Soc. d’etude scient. de la tuberc., 1911/12, 2. s., 22. 73 TOXINS—EXO AND ENDOTOXINS—TUBERCULINS exhausting chamber of a digestor with automatic intermittent siphon- age, and extracted with absolute alcohol. There are collected: 1. A coloring matter; 2. An alkaloid extracted with phosphotungstic reagent; 3. Some fatty acids extracted with a solution of sodium carbonate; 4. A little lecithin, extracted with acetone or cadmium chloride in alcoholic solution, and recovered later by treatment with hydrogen sulphide. The extraction continued with ether yields: 1. Some neutral fats; 2. A large quantity of lecithin; 3. A little cholesterol. The lecithin is precipitated with cadmium chloride and the choles- terol is isolated by crystallization. Further extraction with chloroform yields: 1. A substantial quantity of cholesterol; 2. Some waxy substances, undetermined and small in amount. The chloroform solution is evaporated to dryness and then again taken up in boiling alcohol-ether. On cooling, the cholesterol crystallizes out in long glistening needles. All of the extractions are carried out in vacuo in order to lower the boiling point of the solvents and to avoid alteration of the extracted products by too high a temperature. These substances make up the lipoid or adipowaxy poisons of the bacillus (ethero-bacilline and chlorof ormo-bacilline). The bacillary mass is treated at a temperature of 80°C. with pure concentrated acetic acid which dissolves the caseins without altering them. The acetic solution is precipitated with sodium hydroxide. The precipitate washed in water, alcohol and ether and dried in a vacuum, constitutes the bacillo-casein. After washing in 90 per cent alcohol it can be preserved in the fresh state in emulsion without glycerin. It loses some of its properties with time. The bacillary mass so treated contains only a small quantity of nuclein, soluble only in caustic potash which modifies its constitu- tion, and of cellulose which can be recognized by its characteristic reactions (quick transformation into hydro-cellulose with sulphuric acid and a later blueing with Lugol’s solution). 74 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS The chemical and biological characteristics of the different toxins to be extracted from the tubercle bacillus are tabulated by Auclair and Paris: TOXINS EXTRACTING FLUIDS CHEMICAL PROPERTIES BIOLOGICAL PROPERTIES Albumoses Traces First class: Water Albumins of nu- • cleo- pro- teids Not well understood Soluble toxins NaCl solution 1 per cent •Globulins at present Fatty acids Alcohol Alkaloids Lecithin Local action: Second class: acids Caseation Lipoids Ether Lecithin Cholesterol Chloroform Cholesterol ' Local action: • Waxes Sclerosis Local action: Nodules, glands, small visceral Third class: Protoplasmic« toxins Acetic acid; tubercles. Neutral salts of Paranucleo-albumin • General action: alkaline reac- (bacillo-casein) Congestion, hema- tion topoietic disturb- ances, cachexia, death Volatile bacillary poisons In addition to the extractible and stable poisons, the tubercle bacillus produces certain others of a volatile nature. They are well known to bacteriologists who have experienced their effects in evapo- rating cultures on a water-bath for the preparation of tuberculins, or in grinding dry tubercle bacilli killed by heat. Auclair called attention to them in his thesis and Armand-Delille12 describes very minutely the train of symptoms which he had the 12 Bull. Soc. d’etude scient. de la tuberc., 1913, Dec. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 75 opportunity of observing in his own case. There is a sensation of muscular pain, at first lumbar, then general, coming on 6 to 8 hours after the manipulation of the bacilli. Then there are one or more severe chills, with nausea and violent headache. The temperature gradually rises to about 39°C. After a period of restless sleep lasting 4 to 5 hours, sweating sets in, the fever falls and there remains only a general lassitude of varying duration. Bacteriologists are not all sensitive to these volatile poisons of the cultures, but those who are susceptible and who suffer brisk reactions do not appear able to immunize themselves by habituation. More- over they also react to tuberculin. B. TUBERCULIN OF ROBERT KOCH—ITS PREPARATION Comparative study of lesions produced by dead tubercle bacilli in healthy animals and in those already rendered tuberculous, led Robert Koch to the discovery of tuberculin.13 He had noted that the effects produced by subcutaneous injection of dead bacilli are quite different in the infected body. Healthy guinea pigs tolerate large doses of the dead bacilli, reacting with a simple local abscess, while tuberculous guinea pigs on the contrary are killed in 6 to 48 hours by very small quantities. If the dose is so small as not to lead to death, a more or less extensive area of cutaneous necrosis is produced about the point of inoculation; and if one continues to inject still smaller doses, the general condition of the tuberculous animal im- proves, the inoculation ulcers become cicatrized the primarily en- gorged lymph nodes diminish in size and the evolution of the disease seems to undergo a period of arrest. Thinking at once that this action was due to a toxic substance liberated by the dead bacilli, Robert Koch attempted its extraction. To this end he first scraped off cultures grown on glycerin agar and took up the material with a 4 per cent solution of glycerin in water: the mixture was evaporated on a water bath to one-tenth of its orig- inal volume and the bacilli removed by filtration. But before long he adopted the technique which is still used in preparing what is today called Koch’s old tuberculin. It was origi- nally prepared as follows: A liter of weakly alkaline beef broth, containing 10 grams of peptone and 40 to 50 gms. of glycerin, is poured into a large flat 13 Deutsch. med. Wchnschr., 1891, 17, 101. 76 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS bottom flask until the latter is about one-third full. This is steri- lized in the autoclave at 120°C. and after cooling a fragment of culture is inoculated upon the surface of the medium. The flask is put away in the incubator at 38°C. After 6 to 8 weeks, when the completely developed film tends to break up on the surface of the liquid, the latter is evaporated on a water-bath to one-tenth of its original volume and filtered through a porous earth (Berkefeld) or porcelain (Chamberland) filter. The tuberculin so obtained is clear, syrupy and of a dark brown color. In most laboratories, the original technique has been improved. The cultures are first sterilized with steam at 100°C. and then evaporated on the water-bath; the concentrated liquid is filtered through two layers of thick paper (Chardin paper), which retain most of the bacterial bodies. A dose of 0.1 to 0.3 cc. of this tuberculin should kill within 6 to 24 hours guinea pigs injected subcutaneously four weeks before with one centigram of a culture of tubercle bacilli, a quantity ordinarily sufficient to cause the death of the animals in 8 to 10 weeks. On autopsying the guinea pigs there is found, at the point of inoculation of the bacilli and round about it, a red edematous infiltration which extends to the corresponding lymph nodes. The spleen, liver, lungs and small intestines show small, dark-red ecchymotic spots due to extravasations of blood. The suprarenal capsules are increased in volume and congested. Von Bergman in 1890, in a case of tumor of the cheek, was the first to use tuberculin as a diagnostic agent in man. C. EXO- AND ENDOTOXINS OF THE BACILLUS The original tuberculin of Robert Koch contains both the exo- toxins or soluble poisons excreted by the bacillus into the culture media, and a portion of the protoplasmic endobacillary poisons. The diffusion of the latter into the liquid is facilitated by the slow concen- tration with heat, the media becoming richer and richer in glycerin. The exotoxins however are particularly abundant since, on the one hand, the tubercle bacilli retained on the filter after maceration are scarcely less toxic than if they had simply been killed by heating during an equal period. On the other hand, a tuberculin,—definitely less toxic to be sure but still active,—can be prepared by concentrat- ing the culture media from which the tubercle bacilli have previously TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 77 been separated. As a matter of fact it is in this way that Denys (of Louvain) prepares his tuberculin for treating patients. It is nothing but culture broth filtered through porous candles. Maragliano14 evaporates this same filtered broth at 30°C. in a vacuum and obtains a liquid of which 1 cc. causes a fall of tempera- ture and kills healthy guinea pigs. This liquid, contrary to what occurs in the case of Koch’s tuberculin and the bacilli, is said to lose its toxicity on being heated to 100°C. (but this has been shown to be incorrect by A. Koeppen).15 Maragliano considers it the toxalbumin of the cultures, as distinct from the toxoproteins of the bacilli. By growing the bacilli on a liver broth to which is added glycerin and a leucotoxic serum, which he secures from animals several times inoculated with emulsions of spleen or of leucocytes, Marmoreck16 pre- pares a soluble toxin which, after simple filtration, is more toxic for healthy than for tuberculous animals. The subcutaneous injection of 5 to 10 cc. suffices to kill a normal guinea pig or rabbit. This soluble toxin is used to inoculate horses for the production of his anti- tuberculosis serum. Koch himself and many later investigators have tried to purify the original tuberculin by eliminating substances contained in the culture media and which themselves might be harmful (peptone, salts, waxes and fats). The method most to be recommended consists in diluting a measured quantity of crude tuberculin in 5 parts of dis- tilled water and pouring the mixture, a little at a time and with constant shaking, into 20 volumes of 95 per cent alcohol. The precipitate is collected on a Berzelius paper filter, redissolved in a quantity of water equal to the original quantity of raw tuber- culin and then precipitated a second time with 20 volumes of alcohol. The second precipitate, when dried in an oven, furnishes a spongy, grayish-white mass, soluble in water and very toxic. One milligram of the powder produces the same effects as 50 mgms. of raw tuber- culin. The quantity obtained varies between 1 and 2 gms. per 100. When redissolved in 50 per cent glycerin water, this precipitated D. PURIFIED TUBERCULINS 14 Berl. klin. Wchnschr., 1899, 36, 385. 16 Ztschr. f. Hyg., 1906, 52, 111. 16 Bull. Acad, de m6d., 1903, 50, 332; 465; 480. 78 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS tuberculin gives clear solutions, easily preserved and which can be sterilized in the autoclave at 120°C. and are very stable. But it still contains many impurities which can be eliminated in large part by repeatedly re-dissolving and re-precipitating. More easily purified products are obtained by starting with cultures on liquid media without peptone, such as the one the composition of which I worked out with L. Massol (see Chapter II). Such a culture is concentrated in a vacuum at 45°C. and only to one-fifth of its original volume. The liquid is precipitated with 20 volumes of a mixture of equal parts of 95 per cent alcohol and sulphuric ether; the precipitate is re-dissolved in a small quantity of water, dialyzed on animal parchment for 6 to 32 hours in cold running distilled water, and precipitated a last time with 10 volumes of absolute alcohol. The white powder (Tuberculin CL) obtained is still 10 times more active than that furnished by precipitating old tuberculin with alcohol. The quantity obtained is about 0.75 gm. per liter of culture. E. CHEMICAL PROPERTIES OF THE TUBERCULINS Koch’s tuberculin purified by precipitation with alcohol gives all the reactions of albuminoid substances. According to Kuhne,17 it contains some deutero-albumoses, a special albumose (acro-albu- mose), some peptones, some tryptophan, and an indol-like body. It is precipitated by ammonium sulphate, iron acetate and tannin; in part by lead acetate. Acetic acid produces first a considerable clouding and even a little sediment which is redissolved in an excess of the acid. Picric acid gives a flocculent precipitate which disap- pears on heating, to reappear again on cooling. Hydrochloric and sulphuric acids, both weak and concentrated, yield no precipitate. Analysis of the ash has given the following results in the hands of various workers: WEIGHT OF SUBSTANCES DRIED AT 100°C. WEIGHT OF ASH OBTAINED PER 100 grams grams grams Brieger 0.4816 0.0802 16.65 Proskauer 0.1410 0.0265 18.46 0.1740 0.0350 20.46 17 Ztschr. f. Biol., 1892, 29, 26. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 79 The ash is made up almost exclusively of potassium and magne- sium phosphates. It contains no chlorides. Elemental analysis of the substance without the ash, shows, according to Brieger: per cent C 47.02 H 7.55 N 14.45 per cent C 48.13 H 7.06 N 14.46 S 1.17 and according to Proskauer: per cent C 47.67 H 7.18 per cent N 14.73 S 1.14 Ruppel18 has studied the chemical composition of the bacillary- endotoxin (bacilli ground and macerated in water). He found a nuclein decomposable into a tuberculo-nucleic acid and a protamin which crystallizes in small hexagonal plaques {tuberculosin) and of which one gram is said to have the same toxicity as 25 to 30 cc. of Koch’s old tuberculin. According to Ruppel, all tuberculous toxin is made up essentially of tuberculosin, which is itself derived from the decomposition of a tuberculothymic acid. But it is not at all certain that this tuberculosamin really repre- sents the tuberculous toxin. Its high degree of toxicity is not a decisive argument, since other protamins, which can be extracted for example from the sperm of the sturgeon (Neufeld), are as toxic for the guinea pig, if inoculated intracerebrally, as is the tuberculosamin of Ruppel. Moreover the latter is not noticeably more toxic for tuberculous animals than for healthy animals. It is said that from 100 gms. of dry bacilli 8.5 gms. of tuberculonucleic acid, 24.5 gms. of nucleoprotamin and 23 gms. of nucleoproteins are to be derived. Andre Jousset19 concluded from his researches, chemical as well as biological, that tuberculin in broth cultures is associated neither with the albumins nor with the peptones, but with the lowest prod- ucts of the culture medium, the amino-acids. The tuberculin ap- pears, according to him, between the first and second week after 18 Ztschr. f. physiol. Chem., 1898, 26, 218; Beitr. z. exp. Therap., 1900, H. 4, 89. 19 Bull. Acad, de med., 1914, 71, 752. 80 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS inoculation of the media and the increase parallels the growth of the film. It is a sort of waste product which can in no way be likened to genuine* toxins. The tuberculins prepared according to Koch’s method (old tuber- culin) have an alkaline reaction when made from cultures of bovine bacilli; the reaction is acid (to phenolphthalein) if made from human bacillus cultures (Th. Smith). F. ESTIMATION OF THE TOXICITY OF TUBERCULINS The toxicity of tuberculins for healthy animals can be measured by the method proposed by von Lingelsheim. It consists in trephin- ing a guinea pig at a point approximately in the middle of a line which would join the posterior commissure of the two eyes (not exactly in the middle but a little to one side in order to avoid opening the superior longitudinal sinus). When the operation is well done there is no bleeding. The needle of the syringe is thrust perpen- dicularly into the cerebral mass of one or the other hemisphere to a depth of 3 to 4 mm. The injection is made gently so that all the liquid,—the volume of which should not exceed 0.5 cc./—shall penetrate during 2 to 3 minutes. Glycerin solutions must not be employed in this inoculation procedure since glycerin itself is toxic for the nerve cells. It is found that normal non-infected guinea pigs succumb in a few minutes to doses of 3 to 4 mgms. of precipitated old tuberculin, while doses 150 and even 200 times larger are innocuous if given by sub- cutaneous or intraperitoneal injection. A. Borrel,20 employing this method, has shown that the bacillary bodies washed and heated to 100°C. kill healthy guinea pigs, under the same conditions, in a dose of 0.5 mgm. In animals artificially infected 30 days before, 1/800 of this dose (0.01 mgm. of precipitated tuberculin) suffices, and even as little as 1/8000 (0.001 mgm. of pre- cipitated tuberculin) is enough to kill guinea pigs infected more than 40 days before. The symptoms of intoxication are always the same (respiratory distress, convulsions, almost immediate asphyxia). In Germany, the official titration of commercial tuberculin is regularly carried out by the Institute of Experimental Medicine at Frankfort, formerly under the direction of P. Ehrlich. The method of Otto21 is used. It consists in taking, for example, 50 guinea pigs 20 Compt. rend. Soc. de biol., 1900, 52, 358. 21 Arb. a. d. k. Inst. f. exper. Therap., 1906. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 81 of the same weight (350 to 400 gms.) and injecting them intraperi- toneally with 0.5 mgm. of a broth culture only 12 to 14 days old evenly suspended in physiological salt solution (1 cc. of liquid for 1 mgm. of the culture weighed in the fresh state). At the end of the third week 2 to 4 guinea pigs are autopsied to be certain that the tuberculosis is well developed and, if this is the case 2 to 4 others are tested out with 0.3 and 0.5 cc. of a standard tuberculin (Standart- tuberkulin). These doses should suffice to kill the animals when injected subcutaneously. If 0.5 cc. does not kill, it is because the tuberculosis is not sufficiently advanced and one must wait a few days before repeating. When the result is positive the guinea pigs are divided into two parallel series, one of which receives 0.05, 0.075, 0.1, 0.2, and 0.3 cc. of standard tuberculin, while the other receives corresponding doses of the tuberculin whose toxicity is to be measured. The death of the guinea pigs should occur in less than 24 hours and at autopsy there should be lesions characteristic of tuberculin intoxi- cation. Under the above conditions the usually fatal dose of stand- ard tuberculin is 0.075 cc. Several investigators, particularly Detre and Spengler, thought that they found differences .of toxicity in tuberculins according as they were prepared from cultures of human or bovine bacilli. Weber and Dieterlen22 carried out many experiments in order to settle this question. Their conclusions agree entirely with those reached at the Pasteur Institute in Paris in 1891; namely, that the two tuberculins may be used indifferently on bovines and small laboratory animals, provided they have an equivalent toxicity measured, as has been said above, by the intracerebral method on tuberculous guinea pigs, which have been infected with the same quantity of human or bovine bacilli and have reached the same stage of the disease (about 4 to 5 weeks after the subcutaneous inoculation of one centigram of bacilli weighed in the fresh state). Avian bacillus tuberculin has exactly the same relative action, but is always somewhat weaker. A. Marie and Tiffeneau23 have made an extensive study of the toxic action of purified tuberculins prepared from peptone-free media. With their product, made from unheated cultures, concentrated 22 Tuberk.-Arb. a. d. k. Gsndhtsamte, 1910, H. 10., 217. 23 Compt. rend. Soc. de biol., 1908, 64, 501; 1909, 66, 206. 82 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS in vacuo, dialyzed and precipitated, the rabbit succumbed after in- tracerebral injection of 0.02 gm. For the normal guinea pig the lethal dose was 0.00075 gm. (intracerebrally), and for the mouse 0.10 gm. (subcutaneously). The guinea pigs tuberculized four weeks before succumbed to 0.0001 gm. (intracerebrally). One must therefore conclude that the tuberculous poisons (endo and exotoxins), extremely toxic for animals infected with tuberculosis, are much less so,—although not entirely harmless,—for normal animals. The extraordinary sensitiveness to tuberculin on the part of tuber- culous subjects, first noted by Koch, has been verified for all animal species, and this discovery has been the basis of innumerable applica- tions (which we shall later discuss) to the diagnosis, prognosis and treatment of tuberculous infection. We shall also consider in other chapters (XXXV, XXXVI and XXXVII) the physiological properties of the tuberculins, as well as their aptitude for serving as antigens and for generating “antibodies,” or defensive substances, in healthy or tuberculous subjects. For the moment we shall consider only how the tuberculins are prepared, their principal characteristics and the different products to be derived from them. G. PRODUCTS DERIVED FROM THE TUBERCULIN OF KOCH Beginning with Robert Koch, repeated efforts have been made to purify tuberculin. Unfortunately however a large number of preparations have also been made for commercial profit and have been offered to physicians and patients under various names, and to them the makers have attributed virtues for which there is too often no scientific basis. I shall here mention only such as are of some interest in research. I. Tuberculin TR. Proposed by R. Koch24 in 1897, this tuber- culin is obtained by drying tubercle bacilli in a vacuum and finely grinding them in a mortar with an agate pestle. The bacilli, after prolonged trituration, are emulsified in distilled water and centri- uiged for 30 to 45 minutes at 4000 revolutions. The supernatant opalescent fluid contains a part of the bacillary endotoxins and con- stitutes the preparation OT. The muddy residue, thrown to the bottom of the tube, is dried, ground, again taken up in water, ccntri- 24 Deutsch. med. Wchnschr., 1897, 23, 209. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 83 fuged and reground several times until centrifugation no longer separates out intact bacilli. The different trituration residues are collected together and 20 per cent of glycerin added. We now have TR, which is an emulsion of the various substances comprising the bacterial cells. One cubic centimeter corresponds to about 2 mgms. of dried substance obtained from 10 mgms. of bacilli also dry. This tuberculin is manufactured at the Hoechst Chemical Manufactory, near Frankfurt on Main. II. Tuberculin BE. In 1901, R. Koch25 proposed to substitute a new preparation for the TR. It was presented under the name of Neutuberkulin-Bazillenemulsion or BE, and was obtained by making a suspension of either human or bovine bacilli previously dried and then very finely ground. The suspension is prepared by adding 1 part of the powder to 200 parts of distilled water glycerinated to 50 per cent and is not centrifuged. Dilutions are made with physio- logical salt solution, the strength being expressed in bacillus content and on the basis of one cubic centimeter of the initial raw product containing 5 mgms. of powdered bacilli. III. Tuberculin AF (Albumosefrei) of R. Koch. This tuberculin is peculiar in that, unlike old tuberculin, it is made from cultures on media containing neither meat, nor peptones. Mineral media containing asparagin or ammoniacal salts are used for its prepara- tion, that of Proskauer and Beck (modified by R. Koch) for example (see Chapter II). When the cultures have developed, which is approximately after two months, they are killed by a two hour heating at 60°C. on two successive days and are then filtered. The filtrate is evaporated in a vacuum to one-tenth of its original volume, and 0.5 per cent of carbolic acid is added for preservation. A number of guinea pigs are inoculated subcutaneously to be certain that no living bacilli remain. After this verification, the product is ready for use in the treatment of patients. It seems, according to those clinicians who have experimented with it, to be tolerated rather better (G. Jochmann and R. Moellers);26 but the serum of patients who receive it is not perceptibly enriched in antibodies, and they continue to be sensitive to ordinary tuberculin, although they tolerate subsequent injections of bacillus emulsions in a better manner. 25 Deutsch. med. Wchnschr., 1901, 27, 829. 26 Deutsch. med. Wchnschr., 1911, 37, 1297; Veroffentl. d It. Koch Stift. z. Bekampf. d Tuberk., 1912, H. 3, 29. 84 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS From the chemical point of view, this tuberculin, derived from an albumin free medium, gives the same albumin reactions, although less intensely, as that prepared with ordinary broth. Therefore albuminous substances are formed during the growth, and a large part of them are derived from autolysis of the bacilli. In their investigations on albumin-free tuberculins, E. Lowenstein and E. Pick27 found that their product did not show all of the charac- teristic reactions of albuminous substances. Their tuberculin, which is prepared from media having as a basis asparagin, ammonium lactate, sodium phosphate, sodium chloride and glycerin, is precipi- tated by alcohol, acetic tannin and by biniodid of mercury and potassium in a hydrochloric acid medium. The product is dialyzable and is destroyed by digestion with pepsin or trypsin. It has there- fore rather the characteristics of a polypeptid. IV. Tuberculocidine of Rlebs. Klebs28 treats tuberculin with alcohol, then redissolves the precipitate in water and treats the latter solution with a mixture of alcohol, chloroform and pure benzol. These solvents remove the impurities in large part. The product thus obtained is desiccated in an oven at 56°C. and again taken up in 100 cc. of glycerin containing 0.5 per cent of phenol. After a final filtration a substance is obtained which is soluble in alcohol, keeps indefinitely and which represents about 5 per cent of the tuberculin originally employed. This preparation, according to Klebs, rid of the alkaloidal princi- ples to which are due the harmful effects of the crude tuberculins, should have a peculiar effect on the bacillus which, in a sense become “vacuolized.” Moreover it is only weakly toxic. Normal guinea pigs and rabbits tolerate 1 cc. without any discomfort and it is well borne by patients in doses 4 to 5 times larger than those of old tuber- culin. Furthermore it is said to have the advantage of possibility of administration by mouth. Turban formerly used it in this way at Davos; his patients took a few drops in the morning on a fasting stomach, and there were some favorable results. Its use has, however, for a long time been discontinued. V. Tuberculin of Maragliano. Even before Beranek, Maragliano29 had proposed making use of the intra- and extracellular toxins to- 27 Biochem. Ztschr., 1911, 31, 142. 28 Deutsch. med. Wchnschr., 1891, 17, 1233. 29 Berl. klin. Wchnschr., 1899, 36, 385. TOXINS— EXO AND ENDOTOXINS—TUBERCULINS 85 gether. His tuberculin is a mixture of toxoproteins derived from whole cultures macerated and concentrated at the boiling point, and of toxalbumins separated from the culture broths by concentration in vacuo at a low temperature. According to Bezangon and Gouget,30 these toxalbumins tend to lower the temperature of both tuberculous and healthy guinea pigs. But A. Koeppen31 has shown that there is in reality no difference between the two toxic substances from the point of view of their temperature lowering action on tuberculous guinea pigs. The active portion of the extracellular toxalbumins is thermostabile, as is that of the intracellular toxoproteins. Bio- chemical reactions are the same in the two groups, so that there is no advantage in separating them. In 189832 Maragliano modified his method of preparation in the following manner: The culture on glycerin peptone broth is filtered and the bacilli remaining on the filter are taken up in order to free them of all traces of glycerin. They are mixed with a quantity of distilled water equal to the original culture volume and the whole is left for 45 hours on the water-bath at 95 to 100°C., loss by evaporation being replaced. The mixture is afterward evaporated to one-tenth of its original volume and filtered. There is thus obtained an aqueous tuberculin of dark brown color and of alkaline reaction. Its effects are the same as those of Koch’s old tuberculin. Maragliano states that this watery tuberculin produces fever in both normal and tuberculous guinea pigs; normal guinea pigs are killed by a dose of 1 cc. per 100 gms. of body weight, while tubercu- lous pigs are killed within 48 hours by 0.1 to 0.2 per 100 gms. It is used chiefly in research, particularly for the preparation and titration of the therapeutic serum which Maragliano used in the treatment of patients at his clinic at Genoa (see Chapter XLI, 1). VI. Oxytoxine of Hirschfelder. Hirschfelder (of San Francisco)33 originally oxidized tuberculin by treating 40 gms. of Koch’s raw prod- uct with 240 cc. of hydrogen peroxide (oxygen 10 volumes) and 936 parts of distilled water. The mixture was left for 96 hours in the autoclave. Later on he modified his procedure by adding the 30 Compt. rend. Soc. de biol., 1899, 51, 521. 31 Ztschr. f. Hyg., 1906, 52, 111. 32 Compt. rend. Soc. de biol., 1898, 50, 94. 33 Lancet, 1898, i, 179. 86 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS peroxide directly to the glycerin veal broth culture in a proportion of one to ten. He sterilizes at 100°. Every twelve hours he adds a fresh quantity of hydrogen peroxide equal to the preceding, heats at 100°C. and continues thus until the volume of added peroxide is equal to the volume of the culture. He neutralizes with sodium hydroxide, adds 5 per cent of boric acid for preservation and filters. The oxytuberculin so prepared should contain no trace of free tuber- culin nor of hydrogen peroxide. It is administered hypodermically, beginning with 5 cc. per day and increasing little by little up to 20 cc. Injections cause neither general nor local reactions; they are innoc- uous. L. Guinard and Mondielli34 tried oxytuberculin upon animals and verified its non-toxicity. It does not appear to have been intro- duced into medical practice, nor have its therapeutic effects been well studied. VII. Tuberculol of Landmann. Landmann35 employed cultures whose virulence had been increased by successive passages through guinea pigs. When the cultures are well developed on glycerin peptone broth, the bacilli are collected by filtration. They are first rid of their fatty waxy envelope by appropriate solvents, then trit- urated and all their extractable products slowly removed at a tem- perature of 40°C. with physiological salt solution, distilled water and dilute glycerin. The supernatant fluid is poured off and the extraction repeated several times at gradually increasing tempera- tures, beginning at 50°C. and then through 60°, 70°, etc., up to 100°C. The extracts obtained at the different temperatures are combined and evaporated to dryness in a vacuum at 37°C. The bulk of the endotoxins is thus collected, without perceptible loss and without alteration, since the bodies of the bacilli which remain are said to be no longer toxic. These endotoxins are redissolved in the culture broth which is filtered and concentrated in a vacuum at 37°C. One cubic centi- meter of the liquid so obtained is sufficient to kill a healthy guinea pig. Filtered through a bougie and supplemented with 0.5 per cent carbolic acid, it constitutes Tuberculol. According to Landmann, the proof that this substance possesses properties distinct from those of tuberculin lies in the fact that its toxicity is much reduced by heating at 100°C., or even by prolonged 34 These, Lyon, 1898. 36 Hyg. Rundschau 1898, No. 10; 1900, No. 8. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 87 preservation in solution. It does not appear however that this toxicity for normal animals is really due to endotoxins derived from the bacilli since, according to the experiments of 0. Bail and those of Lowenstein, one can inject 200 mgms. of living bacilli into guinea pigs, to the extent that their bodies are saturated with them, without any phenomena of intoxication being thus produced. The guinea pigs become tuberculous a little more quickly than those infected with only 2 mgms. of bacilli, and they die two or three weeks earlier. The toxicity of Lanclmann’s preparation for healthy animals results perhaps from the fact that it contains decomposition products of the protoplasmic substances formed during the extraction at low temperature. With tuberculous patients, treatment with this tuberculin should, according to its originator, be begun with a dose of 0.005 mgms. and be thence progressively increased up to 0.1 mgms. Merck of Darmstadt supplies tuberculol in three forms, in order that clinicians may be enabled to utilize separately the endo and the exotoxins. Tuberculol A is the form just described; tuberculol B contains only the extractable products of the bacterial bodies; tuberculol C is made up by simply concentrating the culture media. Tuberculols D, E, F, are also prepared, being derived from cul- tures of bovine bacilli, and are used for general or local tuberculin reactions. But all of these “specialties” are of no scientific interest. VIII. Tuberculin of Beranek (of Neuchatel in Switzerland). This tuberculin, prepared especially for tuberculin therapy, is used chiefly in Switzerland by Sahli (of Bern). It is a mixture of culture broth rid of bacilli by filtration through a Chamberland filter then concentrated in a vacuum at low temperature, together with an extract of the bodies of the bacilli obtained by macerating the latter for two hours at 60°C. in a 1 per cent solution of orthophosphoric acid which is afterward neutralized with sodium hydroxide. Thus is obtained AT (acido-toxin) or endocellular toxin. The culture broth is prepared by macerating veal at room tempera- ture, sterilizing it and then adding, 0.5 per cent of sodium chloride and 5.6 per cent of glycerin. It contains no added peptones. The tuberculin of Beranek 36 is a 1 in 20 dilution of a mixture of 36 Compt. rend. Acad, des sci., 1903, 137, 889; Rev. med. de la Suisse Rom., 1905, 25, 684; 1906, 26, 461; 1907, 27, 444; Internat. Congr. on Tuberc., 6th., Wash., 1908. 88 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS equal parts of AT and of filtered broth (TB). For therapeutic use, there are prepared 15 different solutions, the concentration of which increases by multiples of 2, so that each solution is twice as strong as the preceding. It is not toxic for normal guinea pigs and is scarcely so for tuberculous pigs. According to K. Siegesmund its toxic power is 3.3 times less than that of the tuberculin control (Standart- tuberkulin) of the Institute of Experimental Medicine at Frankfurt, and it can be injected into tuberculous guinea pigs in amounts up to 16 cc. without the death of the animal. IX. Tubolytin of Siebert and P. Romer. Siebert and P. Romer,37 working in Von Behring’s laboratory at Marburg, prepared a tuber- culin without the use of heat or any chemical reagent which might modify or injure the active principle. They named it Tubolytin. This product, like Koch’s old tuberculin, is innocuous for healthy animals and induces the characteristic reaction in the tuberculous animal. For the latter it is much less toxic, since it requires about five times as much tubolytin as tuberculin to cause death. The two products, if inoculated intradermally, are equivalent. The residue after evaporation (dry extract) is one hundred times, the ash content 39 times and the nitrogen content 43 times less than for old tuberculin. The complement fixation reaction of Bordet-Gengou can be per- formed with 1 cc. of tubolytin and the presence of tuberculous anti- bodies be demonstrated in 0.0025 cc. of serum. To obtain the same fixation with old tuberculin, 0.02 cc. of the latter and 0.01 cc. of serum are said to be required. X. Tuberculo-plasmin. E.Buchner38andHahn39appliedto tubercle bacilli a process devised by the former for extracting the zymase from beer yeast. They triturated the bacilli with sand, infusorial earth, 5 per cent of sodium chloride and 20 per cent of glycerin, and sub- mitted them to a pressure of 400 to 500 atmospheres. In this manner they obtained a liquid of a light amber color, which is relatively stabile and is said to be a particularly active tuberculin. The difficulties of preparation are however so great that its use has never been extensive. 37 Beitr. z. klin. d. Tuberk., 1913, 26, 193. 38Miinchen. med. Wchnschr., 1897, 44, 299. 39 Ibid., 1897, 44, 1344. TOXINS —EXO AND ENDOTOXINS—TUBERCULINS 89 XI. Tuberculin of Rosenbach.i0 This tuberculin is a complex ex- tract derived from a mixed culture of tubercle bacilli and of a mold, the Trichophyton holosericum album. It is prepared cold and carbolic acid is added for preservation. It possesses practically no toxicity, even for tuberculous animals which tolerate 5 cc. without difficulty. According to Lesser (Karl) and Koegel41 this tuberculin is in no way superior to the old tuberculin of Koch; it appears to be about 1000 times less active than the latter and offers the disadvantage of con- taining non-specific products of the trichophyton. It has been used in man in doses of 0.01 cc. to 0.1 cc. at the begin- ning of treatment and thence up to 2 cc. H. Schaefer42 considers that, under the influence of the proteolytic diastases of the trichophyton, the tuberculin of this product is simply digested and therefore becomes inactive. XII. Tuberculin of Vaudremer. In studying the action of different microorganisms on the crude tuberculin of Koch, Vaudremer43 ob- served that the proteolytic ferments, such as those of B. pyocyaneus, Aspergillus niger, Aspergillus fumigatus and Penicillium glaucum, destroy the active substance of tuberculin. He noted later that Aspergillus fumigatus renders living tubercle bacilli avirulent, and destroys tuberculin in vitro if the latter is submitted to prolonged maceration (24 hours at 39°) in the juice of ground-up mycelia. He thus obtains a tuberculin analogous to that of Rosenbach and, like the latter, almost devoid of toxicity, since 2 cc. of a 1 to 8 dilution are innocuous where 2 cc. of the same dilution of crude tuberculin in physiological salt solution would kill a tuberculous guinea pig within 24 hours. Vaudremer gives no information what- ever as to the antigenic value of his preparation, so that the question arises as to whether the complete lack of toxicity on the part of his product (as is the case of Rosenbach’s tuberculin) is not quite simply due to the fact that it no longer contains active tuberculin, the latter having been completely disintegrated by the thermostabile proteases of the mold, as is the case with trypsin and pepsin in artificial diges- tions, as we shall see later. 40 Deutsch. med. Wchnschr., 1910, 36, 1513; 1553;-1912, 38, 539; 589. 41 Beitr. z. klin. d. Tuberk., 1913, 27, 103. 42 Ztschr. f. Tuberk., 1911/12, 18, 168. 43 Ann. de l’Inst. Pasteur, 1910, 24, 189; Compt. rend. Soc. de biol., 1912, 73 , 501; 1913, 74, 278; 752. 90 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS XIII. Neurin-tuberculin of Much.u This is a solution of tubercle bacilli in neurin, a substance prepared by Liebreich and which is derived from the decomposition of brain tissue. Its chemical form- ula is said to be that of an hydroxide of trimethylethylammonium (C5H13 NO). Ten to 22 grams of bacilli are dissolved during 24 hours in 100 cc. of a 25 per cent solution of Merck’s neurin. The bacterial cells at first become swollen, then the protoplasm disappears; but the granules continue unaffected for a long time. If heated to 56°C., only a few hours are required for solution. The fatal dose of neurin-tuberculin for a healthy guinea pig of 300 gms. is about 0.1 gm. According to Wilhelm Schlaudraff,45 of the Institute of Pathology of Saint Georges Hospital at Hamburg, the animals succumb with all the typical symptoms of neurin poisoning and the toxicity of neurin-tuberculin is no greater for tuberculous than for healthy animals. There exists therefore no hypersensitive- ness to this product, from which one may conclude that neurin- tuberculin does not contain the specific substance to which the tuber- culin reaction is due. It is said to be capable of binding antibodies and of serving as antigen for the complement fixation test; but repeated injections into animals (goats, rabbits, guinea pigs) are not followed by any antibody formation. XIV. Tuberculo-mucin of Fr. Weleminsky,46 If the film produced by the growth of human tubercle bacilli is frequently immersed in the culture media (glycerine peptone broth), two substances are gradually formed in the fluid. One has all the properties of a coagulable albumin; the other is a mucin occurring in fairly large amount and characterized by its mucilaginous appearance and by its being pre- cipitated with acetic acid. This last permits of its isolation. This tuber culo-mucin, a mixture of proteins and mucin, first dried and then redissolved in 100 times its volume of water, has been utilized clinically for the treatment of tuberculosis. Ernst Guth47 begins with a dose of 1 mgm. (0.1 cc. of the solution) and increases gradually until a slight febrile reaction is obtained. It would seem to have some favorable effects, particularly in gland tuberculosis. 44Mtinchen. med. Wchnschr., 1909, 56, 1985; Centralbl. f. Bakt., 1910, 54, 342; Berl. klin. Wchnschr., 1910, 47, 1933. 45 Ztschr. f. Immunitatsforsch., 1912, 12, 91. 46 Berl. klin. Wchnschr., 1912, 49, 1320; Tuberculosis, 1914 13, 456. 47 Ztschr. f. Tuberk., 1914, 21, 554. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 91 XV. Tuberculin bovine PTO (Perlsucht tuberkulin) of Spengler. This preparation is made exactly like the old tuberculin of Koch, except that bovine bacilli are used. Spengler,48 who recommended it for a long time, believes that it is better tolerated and less toxic for man, and in a rather exaggerated manner he compares the “im- munizing” effects to those of small pox vaccine. As a matter of fact there is no difference whatever between the tuberculin prepared from bovine bacilli and that prepared from bacilli of human origin. The ideas put forth a few years ago by Detre and his pupil V. Gebhardt, relative to the possibility of de- termining the human or bovine origin of a tuberculous infection in man through differences in reaction to human or bovine tuberculins, were based upon errors of observation and are today abandoned. Spengler himself seems to have given it up since employing his immunizing bodies IK which will be taken up in connection with the serotherapy of tuberculosis (Chapter XLI). XVI. Tuberculous endotoxin of Baudran. Baudran49 treats tuber- cle bacilli with 95 per cent alcohol which precipitates the albumins, peptones and albumoses, and dissolves the glycerin. He then filters and to the bacillary bodies adds successively ether, chloroform and toluene. Each solvent is eliminated after its action is completed. The residue is treated with water, which readily dissolves the pep- tones and the albumoses, and then filtered through moist paper. The resulting liquid is concentrated in the presence of a few drops of a 1 per cent solution of sulphuric acid. The extract obtained is again taken up in alcohol, with which only albumoses and a small quantity of peptones are dissolved. He then neutralizes, filters, and gently evaporates. The substance obtained is completely soluble in cold water. It precipitates with saturated ammonium sulphate and acetic ferrocyanide of potassium; it takes on a red color with Millon’s reagent and is not dialyzable. This endotoxin, according to Baudran, is very toxic for the normal guinea pig which is killed by the intraperitoneal injection of 5 mgms. Such a fact would tend to indicate that it does not represent the true bacillary poison, but rather a modification resulting from the chemical reactions undergone. 48 Deutsch. med. Wchnschr., 1904, 30, 1129; 1905, 31, 1228; 1353. 49 Compt. rend. Acad, des sci.. 1909, 149, 941. 92 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS XVIII. Ferruginous tuberculin of Ditthorn and Schultz50 (Eisen- tuberkulin). Ten cc. of old tuberculin are diluted in 50 cc. of sterile water and the whole precipitated with a 12 per cent solution of oxy- chloride of iron. The precipitate is collected upon a filter, washed for two days and redissolved upon the filter by adding a 1 per cent solution of sodium hydroxide, drop by drop. To this 25 per cent glycerin is added to make a volume of 40 cc., which are filtered and sterilized at 100°C. This product is Eisentuberkulin A. An Eisentuberkulin B is prepared by utilizing the bacilli which have already served to produce the A preparation. They are washed several times in hot water and macerated for 24 hours in weakly carbolized water. The mixture is shaken from time to time to facilitate the extraction of the soluble substance. It is centrifugated and filtered, after which the filtrate is treated with 12 per cent oxy- chloride of iron as for the preparation of Eisentuberkulin A. A third Eisentuberkulin E is made by beginning with unheated bacilli which, after being filtered from the culture broth, are washed with sterile water, then rapidly with alcohol to remove the water. The mass is next dried at 37°C. and extracted successively in a Soxhlet apparatus with ether and chloroform to remove the fats soluble in these solvents. The bacilli thus freed of their fats are left to macerate in water for 24 hours, and after centrifugation and fil- tration the solution is precipitated with oxychloride of iron. From here on the procedure is as for the preparation of Eisentuberkulin A. Finally, a fourth Eisentuberkulin S is made from a broth culture 6 to 8 weeks old, from which the bacilli are filtered off as a preliminary step. The broth is concentrated to one-tenth of its original volume and then treated in the same manner as is old tuberculin in obtaining Eisentuberkulin A. Fritz Ditthorn and Werner Schultz have likewise applied their method of precipitation with iron to cultures of bacilli grown on albumin-free media, such as that of Proskauer and Beck (asparagin 0.5; magnesium citrate 0.25; magnesium sulphate 0.06; monopotas- sium sulphate 0.5; glycerin 2 gms. per 100 cc. of water). According to Schultz, subcutaneous injections of these substances rarely provoke any general reactions. They are but weak antigens and have no special practical virtue (Schellenberg) .51 50 Ztschr. f. lmmunitatsforsch., 1909, 2, 567; Deutsch. med. Wchnschr., 1908, 34, 1221. 51 Ztschr. f. Tuberk. 1911/12, 18, 132. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 93 XVIII. Tebean of Levy and Kaenker. Tebean is an emulsion of tubercle bacilli killed by being left for a long time in a 25 per cent solution of galactose at a temperature of 37°C. One gram of tebean powder is equivalent to 50 mgms. of bacilli. Inoculation of this product is often followed by the formation of painful abscesses. A. Fraenkel and Steffen however, claim to have had good results with it among their cases at Badenweiler. The treatment is begun with 0.001 mgms. and the dose is increased pro- gressively up to 4 mgms. XIX. Tuberculo-toxoidin of Ishigami. Ishigami52 makes this prep- aration by treating bacilli previously dried and then washed, with sulphuric acid in concentrated solution. When diluted with 10 volumes of water the fatty and the waxy substances separate out upon the surface, and the insoluble precipitate, collected from the bottom by decanting, is emulsified in a weakly alkaline solution. The product is not toxic, even for tuberculous animals, although it is said to be still capable of provoking antibody formation and conse- quently of serving as an antigen. XX. Tebesapin or Molliment No. 8 (the former Prosperol of Zeuner). In a communication before the Tuberculosis Congress at Washington in 1908, Noguchi called attention to the immunizing properties of tubercle bacilli macerated for 24 hours in sodium oleinate at 37°C. These experiments have not been confirmed; but at the same time Zeuner launched a preparation under the name of Prosperol, later of Tebesapin, and then finally under the name of Molliment No. 8. It is derived as follows: Tubercle bacilli are macerated with continuous shaking for four days at 37°C. in a 2 per cent solution of sodium oleinate. They are next heated for one hour on a water bath at 70° to 72°C. and the shaking at 37°C. is continued for another three days. The emulsion is now concentrated to different strengths corresponding to a given number of milligrams of bacilli per cubic centimeter; No. 2 equals 0.5 mgms. of the bovine type per cubic centimeter; No. 5 to 2 mgms. of the human type; No. 6 to 10 mgms. of the bovine type; No. 9 to 10 mgms. of the human type, etc. It is strongly hemolytic and for this reason cannot be utilized in vitro as an antigen for complement fixation. 52 Philippine J. Sci., 1908, 3, 379. 94 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS According to investigations by R. Moellers and Georg Wolff,53 this preparation possesses no immunizing properties and exerts no favor- able influence on experimental tuberculosis in the rabbit or guinea pig. Zeuner54 claims that it can be advantageously given to patients by mouth or by rectum. The clinical experiments carried out by Weicker on 50 patients at Gorbersdorf are however nothing less than convincing. The com- mercial exploitation of this product is not justified on any scien- tific ground. Unfortunately the same is true of many similar preparations. I shall limit myself to citing a few products which have been pro- posed by their originators for the treatment of tuberculosis without their having been tested experimentally. In the special articles concerning them will be found the indications for their administration. Tuberculinum pururn of Gabrilowitsch55 (Endotin, albumin-free tuberculin). Tuberculin of Marechal (of Brussels). (A mixture of old tuberculin and guaiacol.) Mycolysin-Tuberculin of E. Doyen. (A mixture of tuberculin and yeast juices.) All of these products owe their properties to endo- and exotoxins contained in the original tuberculin of Koch or in purified precipi- tated tuberculin, and it does not appear that any of them replaces to advantage the original which is easily prepared and of perfect stability, provided it is protected from air and is in concentrated glycerin solution. H. ACTION OF PHYSICAL AND CHEMICAL AGENTS ON TUBERCULINS Dilutions of tuberculin are ordinarily made in 0.5 per cent car- bolic water. They should be promptly used since they gradually lose their toxicity;—approximately one half in two weeks. It is therefore preferable, especially for laboratory experiments and even for therapeutic use, to make dilutions with physiological salt solution as need arises. Tuberculin diluted with plain water is very unstable and readily permits the development of microorganisms which modify or destroy its activity. 53 Verdffcntl. d. R. Koch. Stift. z. Bekampf. d. Tuberk., 1913, H. 8/9. 74. 64 Ztschr. f. Tuberk., 1909, 15, 135; 1912, 19, 268. 55 Ztschr. f. Tuberk., 1908, 13, 234; Beitr. z. klin. d, Tuberk., 1911, 19, 485; 21, 235. TOXINS—EXO AND ENDOTOXINS—TUBERCULINS 95 Tuberculin in the crude concentrated state is not affected by light. Hans Jansen56 found it still undamaged after two hours of exposure to the intense bright rays of a Finsen apparatus. Ultra-violet rays render it incapable of producing reaction in tuberculous guinea pigs only after five hours of exposure (A. Jousset, L. Massol,57 M. and Mme. Victor Henri and Baroni).58 On the other hand, the digestive juices, trypsin in alkaline medium, papain in neutral medium, pepsin in acid medium and the digestive juices of insect-absorbing plants (Drosera), destroy it more or less rapidly whether in vitro or in the digestive tract (Carriere, Kinghorn,59 Koehler, Th. Pfeiffer and Persch, Loeffler). Danielopolu60 studied the action of hydrochloric acid and of pepsin separately through artificial digestion of precipitated tuberculin, at a temperature of 37°C. He was able to establish in this way that hydrochloric acid alone has no effect, that pepsin alone attenuates tuberculin quite strikingly and that mixtures of acid and pepsin destroy it after 24 hours. The relative delay of digestion of tuberculin by hydrolyzing fer- ments permits of its partial absorption by the digestive mucosa, so that it produces its toxic effects when ingested in fairly large doses, especially in young animals whether healthy or tuberculous (Calmette and Breton) ,61 or even in tuberculous human beings after alkalization of the stomach with bicarbonate of soda (Freymuth) .62 The toxic effect is still more marked if the tuberculin is ingested in keratin capsules which are dissolved only by the intestinal secretions, (Mollers and Heinemann)63 (see Chapter XLII). H. J. Bing and V. Ellermann64 called attention to the fact that if an emulsion of egg-yolk lipoids (and more particularly of the ether- insoluble fraction, albin, which is a diamidophosphatid) be allowed to act upon tuberculin, the specific action of the latter is increased. The lipoids of caseous tissue (lymph nodes, liver, lungs) exert the same 56 Centralbl. f. Bakt., 1906, 41, 677; 775. 67 Unpublished work. 58 Compt. rend. Acad, des sci., 1910, 151, 724. 69 J. Med. Research, 1904, 12, 213. 60 Compt. rend. Soc. de biol., 1910, 68, 896. 61 Compt. rend. Acad, des sci., 1906, 142, 616. 62 Miinchen. med. Wchnschr., 1905, 52, 62. 63 Deutsch. med. Wchnschr., 1911, 37, 1825. 64 Biochem. Ztschr., 1912, 42, 289. 96 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS activating effect, while cholesterol, oleic acid and its sodium soap produce no effect. Danielopolu,65 Moussu (of Alfort),66 Zieler67 and Haentjens have, on the other hand, shown that tuberculin, raw or precipitated and redissolved, passes through dialyzing bags of viscose, collodion or vegetable parchment; that the dialysis is slow and progresses best at a temperature of 37°C., and that the active substance likewise passes through porous candles introduced into the animal body. 65 Compt. rend. Soc. de biol., 1909, 66, 334. 66 Ibid., 1906, 61, 95. 67 Miinchen. med. Wehnschr., 1908, 55, 1685. CHAPTER VI HISTOGENESIS AND EVOLUTION OF THE TUBERCLE AND OF BACILLARY LESIONS WITHOUT TUBERCLES. A. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE.—ANATOMICAL PROCESS OF HEALING Certain vegetable parasites (.Aspergillus, Actinomyces), or animal parasites (Nematodes, Acari, Sarcosporidia of the skin of the ox) (Ch. Besnoit and V. Robin1), various bacteria (Bacillus mallei, Bacillus leprae), and also certain foreign bodies such as mercury, oil of turpen- tine, euphorbium powder, lycopodium powder or even coal dust impregnated with Bacillus subtilis (Marcel Gamier and A. Chaoul2) are capable of producing within the body tissues cellular reactions which terminate in the development of real tubercles. When however the latter are produced by substances other than living microorganisms they present the essential characteristic of not being inoculable in series. They are pseudo-tubercles. The true tubercle, produced by the tubercle bacillus, appears first in the form of a granulation, a small, hard, barely elevated, grayish nodule, which cannot be enucleated and which is almost always surrounded by a reddish vascular zone. Its diameter varies from 0.1 to 3 mm. In the beginning, this small granulation is transparent and hyaline; then little by little it becomes more opaque at the center which takes on a yellowish tint. It then becomes the tuberculous follicle characterized histologically by a giant cell surrounded by a zone of epithelioid and embryonal cells. The outline of this small mass is irregular. Its size is that of the head of a pin or of a millet seed (miliary tubercle). There may be no further enlargement and the small mass may undergo fibrous degeneration ending in a small hard cartilaginous nodule containing atrophic elements, that is, healing. Several neighbouring tubercles may fuse and form a mass the center of which undergoes hyaline degeneration (Grancher) and caseation. 1 Compt. rend. Soc. de biol., 1913, 75, 442. 2 Ibid., 1912, 72, 1005. 97 98 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS The protoplasm, and next the nuclei of the giant cells become destroyed; so that nothing is to be distinguished except remnants among which the bacilli in relatively large number are irregularly scattered, especially at the periphery, inside the zone of epithelial cells. As the caseation continues the stainable bacilli diminish in number and end by apparently disappearing completely. Neverthe- less a few of them always remain, since the caseous matter, when inoculated into a susceptible animal like the guinea pig, shows itself to be virulent and produces tuberculous infection. In this stage of caseous tubercle, the lesion may still retrograde. In such case the embryonal cells surrounding the small mass become organized into fibrous tissue, forming a dense wall which thickens gradually up to the center, where ultimately only leucocytic debris is to be found in the sclerotic framework. A few granular malformed bacilli persist there for years in a dormant state (Metchnikoff), capable of being revived by experimental inoculation after the envelop- ing substance has been crushed, but ordinarily incapable of multi- plying in situ in the lesion itself. In this manner the process of spontaneous healing of the tubercles is most often accomplished,— an apparent healing, rarely complete, since it is only exceptionally that some vestiges of caseous matter and a few bacilli do not persist indefinitely at the center. It does happen nevertheless in certain cases that the atrophy progresses to a point where nothing remains except a sort of cicatricial tissue which is hornlike in its hardness. Under other circumstances, the sclerosed tubercles become infil- trated with calcareous deposits and transformed into veritable pearls which are hard and opaque and resist the knife like chalk. Inoculation with them, after grinding, proves that they still fre- quently contain bacilli which can be revived (L. Rabinowitsch, Piettre, Lubarsch). After this short resume of the different modes of evolution of the tubercle, the question arises as to the source of the cellular elements which compose it; a subject which has been the object of numerous works and is still in dispute. Until recent years, many histologists agreed with Baumgarten that the tubercle builds itself up at the expense of and through the proliferation of the fixed tissue elements. This opinion was defended by Kostenitsch and Wolkow,3 3 Arch, de m6d. exp4r. 1892. 4, 741. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 99 by Klebs, Thoma, Stieck, Kockel, by I. Straus in his book published in 1895 on tuberculosis and its bacillus, and by Grancher. Baumgarten4 inoculated bits of tuberculous matter into the anterior eye chailiber of the rabbit and was convinced that he saw the fixed cells of the iris divide by karyokinesis to form giant cells and epithe- lioid cells. According to him, it is only secondarily that the leucocytes enter in and invade the small tumor formed by the fixed pre-existing cell elements produced by indirect division. In the lung a similar process is observed. During the first few days after infection with the bacillus, nothing is visible macroscopically although karyokinesis of the various fixed cells in the areas invaded by the bacilli can be seen microscopically. The bacilli become lodged upon the walls of the alveoli and bronchioles and fix themselves, in part upon the endo- thelial cells of the capillaries, in part in the interalveolar connective tissue. There they produce an irritation. The alveolar epithelium detaches itself from the basal membrane and accumulates in the center of the alveolus; the cell protoplasm becomes granular, while the endothelial cells of the vessels preserve their transparency. In the connective tissue of the interlobular septa, in the walls of the vessels and bronchi and in the lymphatic follicles, there is always to be observed, according to Baumgarten, the same process of karyokine- sis provoked by the presence of the bacilli. A few days later the tubercles are visible to the naked eye. After still another few days histological study reveals voluminous tubercles formed from a cluster of alveoli packed with epithelioid cells. Then the tubercles become caseous and the further evolution shows nothing in particular. New tubercles continue to form and to unite them- selves with the old, constituting the caseous masses. When the casea- tion becomes general, it is difficult to distinguish it from tuberculous lobar or lobular pneumonia, regardless of whether the latter conditions are produced spontaneously or artificially by inhalation of bacilli or by direct injection of tuberculous material into the lung. “In this case,” says Baumgarten, “infiltration of the lung is much more rapid. The tuberculous process is provoked immediately in a certain number of lobules; leucocytes appear earlier and in larger num- ber. It is true that karyokinesis has not then been observed, but the identity of structure of the miliary tubercle and of that following 4 Ztschr. f. klin. Med., 1885, 9, 93; 245; — 1886, 10, 24. 100 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS inhalation warrants the supposition that the same histogenic laws are applicable to both cases. ” This theory of the formation of the tubercle through proliferation of fixed tissue elements in reponse to irritation has been actively and successfully opposed by Metchnikoff,5 by Yersin6 and the whole Pasteur school, especially by A. Borrel7 to whom we are indebted for a most important experimental study of the pathological anatomy of the tuberculous process in the lung and in the kidney. By intra- venous injection Borrel infected the rabbit kidney and was able to observe in the absence of traumatism the earliest expression of the infection, the immediate reaction of the animal body, a rapid forma- tion of tuberculous granulations in the vessels themselves and their evolution through to caseation. The study of the process of lung tuberculization led Borrel to observe the fact that the granulations, in this secondary period following infection, are always developed in and from the lymphatic elements. “In the lung as in the serous cavities, where Kiener first called attention to the fact, the site of election of tubercles about the vessels is due to the peculiarity that they develop almost exclusively in the lymphatic system. The latter is the matrix wherein the tubercles are formed, and not the connective tissue as claimed by Virchow. The tuberculous cell is always a lymphatic cell and is not derived in one instance from a lung cell, again from a liver cell, or still again from a kidney cell. “These lymphatic granulations, so easily studied in the lung, constitute the true tubercles of the majority of investigators; they are the granulations of Laennec, the nodular tubercles of Virchow, the miliary granulations of Cruveilhier, the fibro-plastic granulations of Robin, Empis, etc. “They exist in other organs exactly as in the lung. But in the latter because of its structure, the tuberculous matter may take the form of an infiltration, and we are led to the conclusion that the tuberculous pneumonic process is not due to a desquamation of the epithelial cells of the alveoli (as the partisans of the Baumgarten theory believed), but to an exudation, into the interior of these alveoli, 5 Virchow’s Arch., 1888, 113, 63; Ann. de l’lnst. Pasteur, 1888, 2, 505; Legons sur la pathologic comparee de l’inflammation, Paris, 1892, Masson & Cie 6 Ann. de l’lnst. Pasteur, 1888, 2, 245. 7 Ibid., 1893, 7, 593; 1894, 8, 65. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 101 of lymphatic elements analogous to those which we find in the intra- lymphatic tubercles (A. Borrel).” On killing a rabbit immediately after the intravenous injection of an even suspension of tubercle bacilli, Borrel finds first that almost all the bacilli are already contained within the polynuclear leucocytes disseminated more or less everywhere. By the end of 24 hours, how- ever, leucocytes and bacilli are already localized. On the third day, the leucocytes containing bacilli are beginning to undergo degeneration, the nucleus in breaking up becomes more and more homogeneous and cloudy and the chromatin network and bodies are no longer visible. From the fifth day on there is no further trace of polynuclear leuco- cytes. Indeed by the second day, at the points where bacilli and poly- nuclear leucocytes are collected, one observes the influx of large mono- nuclear cells which are vesiculated, stain but poorly, and have an abundant protoplasm, giving off numerous projections. In view of the intravascular location their significance is not open to doubt, they are the large mononuclear leucocytes. Soon, from the third day, they can be observed grouping themselves about the masses of bacilli, fusing together and thus forming the typical giant cells (see plate III). The number of nuclei contained in such protoplasmic masses can at times be very considerable, says Borrel, who was able to count as many as 60 of them. “Quite often, in a single plasmic mass the nuclei are arranged in groups and are almost always at the periphery, as a sort of collar. This marginal disposition of the nuclei is not difficult to understand when it is realized that in every mobile cell the portion deprived of its nucleus is the one which undergoes move- ment, while the cell portion containing the nucleus is always the relatively stationary part. In the presence of a small clump of bacilli, the mononuclear leucocytes situated on the vascular walls can be seen to send out extensions in the direction of the bacilli, the nucleus remaining always at the periphery. These pseudopods are at times very long and, from the progressive confluence of a large number of them, the giant cell results. In certain cases all of the nuclei are concentrated at one pole and the bacilli are situated in the non-nucleated portion of the cell. From the first days after inocula- tion such giant cells can be found forming in the alveoli by this same process.” 102 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS PLATE III Genesis of Giant Cells and Early Stage Tubercles (According to A. Borrel) 1. Section of a large capillary of rabbit’s lung in which are to be seen leu- cocytes, full of bacilli, and well isolated among the red cells (a few minutes after intravenous inoculation). 2. Longitudinal section of a dilated pulmonary capillary containing bacilli and numerous leucocytes which have already ingested bacilli a few minutes after inoculation; a, isolated leucocytes carrying bacilli. 3. A giant cell on the fourth day. The cell is in a capillary; the nuclei are concentrated at one pole; the bacilli are grouped at the opposite pole in the midst of a well defined felt-like mass of protoplasmic filaments; d, polynuclear leucocytes. 4. Formation of a giant cell from dust cells, in a lung alveolus. A mass of bacilli is in the center of the alveolus; all about are grouped a number of dust cells sending out protoplasmic extensions in the direction of the bacilli. On the alveolar walls there can be distinguished isolated dust cells, one of which is joining itself to the principal mass. 5. A mass of dust cells in the interior of the alveolus on the fifteenth day after inoculation. The bacilli have even developed in the interior of the cells; b, a bacillus-containing dust cell in a neighboring alveolus; h, wall of the alveolus; i, intra-alveolar cells. 6. Tuberculous granulation in process of formation on the third day. In the center one can distinguish the lumen of a capillary containing some mono- nuclear leucocytes and bacilli; all about the capillary, in the midst of the infiltration of small round cells, large elements, which are either mononuclear leucocytes or dust cells. In c a giant cell with bacilli; e, a dust cell mitotic figure. A. CALMETTE. Tubetxulosis Plate III. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 103 The phenomena are identical in all the organs when infected by- way of the blood stream. They are likewise identical for all animal species naturally infected by way of the lymphatics, for example in cattle made to live with other tuberculous cattle and which are then slaughtered while infection is still localized in the glandular system (Calmette and Guerin). In tuberculosis of the lung, Borrel has noted that the dust cells (Staubzellen of the Germans) play the same role in the interior of the alveoli as the large mononuclear leucocytes in the vessels. These dust cells, each made up of a very large vesiculated nucleus and dense granular protoplasm, are quite voluminous and have irregular out- lines. They are adherent to the walls of the alveoli and are found spread out on the surface of the bronchial epithelium, moving in some way deep within the ciliated layer. These cells, which have been carefully studied by Tchistowitsch8 in Metchnikoff’s laboratory, are certainly of lymphatic origin. They are contractile, ingest foreign bodies with the greatest ease, and form typical giant cells. If a concentrated emulsion of living or dead bacilli is introduced directly into the trachea, or is inhaled (Calmette and V. Grysez), the alveoli, from the earliest days, are found to be invaded by an enormous number of dust cells. The majority of them contain bacilli and are as though drowned in an effusion of polynuclear leucocytes which soon undergo processes of degeneration. On the fourth or fifth day of the acute pneumonia, the alveolar cells are filled with chromatic granules. The walls of the alveoli are intact, the epithelium is in place, and yet the alveoli are filled to over- flowing. Multiplication of the epithelial cells by karyokinesis cannot be invoked to explain this invasion (Borrel). If one examines the liver of an animal experimentally infected, it is seen, as Metchnikoff has shown, that the tubercle cells, both epithelioid and giant cells, are formed exclusively from the large mononuclear leucocytes and from the Kupfer star cells, which are of endothelial origin. Neither the hepatic nor epithelial cells contribute to the formation of the tubercle. It is true that at times a few such elements are found in process of karyokinetic division, but this proliferation has no direct relationship with the formation of the tubercle and serves only to regenerate the specific elements of the hepatic tissue. 8 Ann. de l’lnst. Pasteur, 1889- 3, 337. 104 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Phagocytosis and Modifications Undergone by the Tubercle Bacillus Within the Giant Cells of the Gerbille PLATE IV According to Metchnikoff 1. Giant cell from the spleen of the gerbille. a, capsule of the bacillus; b, Bacillus of Koch. The spleen, fixed with Flemming’s solution, has been stained by Gram and with eosin. Oc. 3; imm. obj. y1 Zeiss. 2. Giant cell from the spleen of the gerbille, enclosing a calcareous body with a double bacillus. Same magnification. Stained with hematoxylin and Ziehl fuchsin. 3. Another giant cell from the spleen of the gerbille, enclosing a bacillus surrounded by concentric layers. Same staining. Oc. 2; obj. Zeiss. 4. Giant cell with a calcareous body which contains only a trace of the bacillus b. Fuchsin-hematoxylin; same magnification. 5. Another giant cell in which the bacillus a has been transformed into a body stained light pink. 6. A giant cell containing a calcareous body definitely formed. A. CALMETTE. Tuberculosis. Plate IV. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 105 Tubercles of the spleen and those of lymphatic glands develop in like manner following an accumulation of the large phagocytes of these organs, so that the process of giant cell formation is always the same. Bowman, Winternitz and Evans9 demonstrated that it is possible to vitally stain giant cells and tubercles in process of formation, in the rabbit, by means of intravenous injections of about 20 cc. of a 1 per cent solution of trypan blue, provided the injections are made not more than a half hour after the intravenous injection of the infecting bacilli. The tuberculous cells stain an intense blue and appear sharply differentiated upon the vascular walls, especially in hepatic tissue. Metchnikoff noted a very characteristic degeneration of human and avian tubercle bacilli in the epithelioid cells and especially in the giant cells of spermophiles, animals which, in general, are quite resistant to tuberculosis. The bacilli became swollen and gradually lost their capacity to retain anilin dyes. “ Most often it is the central part which first decolorizes; at times it is the peripheral portion. Later the bacillus becomes transformed into a yellowish sausage- shaped body in the interior of which a very narrow canal is to be seen. The bacilli, thus malformed, unite to make up a mass which takes on the characteristic appearance of a piece of amber. At this time they are conspicuous because of their brownish staining. All of these transformations are never observed, either in cultures (even though many dead bacilli are present), or outside of tuberculous cells.” Similar transformations of bacilli have also been demonstrated in the giant cells of rabbits and, very rarely, in those of guinea pigs. They are not found in cattle or in man, although in these species bodily resistance is often very marked. “For along time,” says Metchnikoff,10 “calcification of tubercles as a mode of healing of tuberculosis in man, has been observed. In order to give a more exact idea of this reaction phenomenon I may cite the case of the resistance to the tubercle bacillus on the part of the body of the gerbille11 of Algeria (Meriones shawii). This rodent, which 9 Centralbl. f. Bakt., 1912, 65, 403; J. Exp. Mod., 1914, 19, 283. 10 Legons sur la pathologic comparee de l’inflammation. Paris, 1892, Masson & Cie. 11 Gerbille or Gerbil, the name of a group of small jumping rodents forming a special subfamily of the rat tribe or Muridae. 106 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS is not absolutely refractory to tuberculosis, resists infection much more effectively than most animals of the same species. Meriones inocu- lated subcutaneously, and even into the eye, with a culture of human bacillus, resist infection for several months. Fig. 6. Invasion of the Air Passages in Miliary Tuberculosis Wall of intra-lobular bronchiole being broken through (from a preparation of Professor Letulle). “ If the Meriones are killed 6 to 8 months after inoculation, one finds a large number of tubercles in the abdominal organs, in the lungs and in the lymph nodes. Such tubercles however, in the majority of cases, present none of the phenomena of necrosis and caseation. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 107 “The tuberculous tissue, made up of vigorous live cells, contains bacilli of which the great majority show a very remarkable degenera- tion meriting more detailed description (see plate IV). “The spleen in particular is sown with small tubercles composed of epithelioid and non-necrosed giant cells. The tubercle cells enclose a small number of ordinary tubercle bacilli, while the giant cells contain very characteristic calcareous bodies. On examination under the microscope, these bodies for the most part have the shape of the figure 8, and are very retractile. At times their form is simply round or irregular. Under the influence of acids, the calcareous salt (calcium phosphate) is dissolved, leaving a more or less numerous series of concentric fairly thin layers. “The calcareous bodies bear great resemblance to the formations described by Schueppel12 in scrofulous lymph glands and found by several authors in many cases of glandular tuberculosis in man (Ziegler). But while, in the latter, the origin of the striated calcare- ous bodies is still completely obscure, in the gerbille it can be easily demonstrated. Examination of the preparations spread on slides, or of sections stained by Ziehl, shows at once that these calcareous bodies reveal a state of degeneration of the tubercle bacilli in the interior of the giant cells. In the early stages the bacilli stain nor- mally, but there are found other giant cells in which the bacilli are encapsulated with a fairly thick layer of an amorphous unstained substance. “This secretion becomes more and more abundant, so that the bacilli are found enclosed within several concentric layers. Some- times in the center of a calcareous body a bacillus is seen divided into two parts, one half stainable, the other half no longer so. Through a series of intermediate changes decolorized bacilli are ulti- mately found, traces of them being still present in contour. But eventually the bacilli are no longer to be distinguished from the surrounding tissue and in the final stages they disappear completely. This last stage, which is by far the most frequent, is that of the strati- fied calcareous bodies “The struggle of the two living organisms—the tubercle bacillus and the giant cell of the gerbille—is carried on therefore with the help of secretions. The bacillus defends itself by the secretion of the 12 Untersuckungen uber Lymphdriisen-Tuberkulose sowie uber die damit verwandten und verwechselten Driisenkrankheiten. Tubingen, 1871, Laupp. 108 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS cuticular membranes and probably also by the production of toxins, while the giant cell secretes a calcareous deposit by the aid of which it walls in the bacillus and finally destroys it in a very large number of cases. ” B. PROCESS OF CASEATION OF TUBERCLES The mechanism of caseation of tubercles is not yet fully understood. This caseation, in the opinion of certain workers, is associated with a deficient blood supply, the tubercles being deprived of vessels. According to others, it results from the action of the toxins produced by the bacillus. Auclair13 believes that the specific poisons are of the nature of fats and are soluble in ether, chloroform, benzin and xylol. The extracts from these different solvents, emulsified in water and injected into the subcutaneous tissues, cause the formation of cheesy abscesses. When injected into the trachea of the guinea pig, caseous areas appear. Ethero-bacillin in particular (the preparation of which was described in Chapter IV) seems to be the bacillary poison which causes casea- tion, while the chloroformo-bacillin appears to be the toxin which induces sclerosis. However it is known today that the caseating property attributed by Auclair to the fats of the tubercle bacillus which are extractable with ether, is in no wise specific. The fats extracted by the same method from many other bacteria, such as paratubercle bacilli (tim- othy bacillus, bacilli from dung, maize, etc.), diphtheria bacilli and the Bacillus subtilis itself, are capable of producing absolutely simi- lar necrotizing and caseating effects when introduced into animal tissues. Joest14 found that in miliary tubercles there is no fat, while in older tubercles it is found in the very center of the focus, and fatty degeneration precedes the necrotic changes. In very old lesions fatty degeneration is found only in the zone between the dead (necrosed) and living tissues. The deposit of fat is intra-cellular and is in the giant and epithelioid cells. It is not produced either in the leucocytes or in the intercellu- lar spaces. Excess of fat is the prelude to cell death and its origin lies in the toxic products of the bacillus. The effect of the toxic 13 These, Paris, 1897; Arch, de med. exp6r., 1899, 11, 363; 1900, 12, 189. 14 Ztschr. f. Infektionskr. . d. Haustiere, 1911, 10, 120. 109 HISTOGENESIS AND EVOLUTION OF THE TUBERCLE products depends on their concentration for if present in but small amount they cause a cellular proliferation; if present in large amount, they cause a fat excess and then a necrosis. The investigations of H. Dominici and Ostrovski,15 with regard to the action of the diffusible toxins of the tubercle bacillus upon normal Fig. 7. Lesions of the Pulmonary Vein Tuberculous nodules and conglomerate miliary granulation. Tuberculous peri-phlebitis; a tuberculous focus breaking through a vein wall. Lung of a child (from a preparation of Professor Letulle). tissues, brought proof that the lesions most characteristic of tuber- culosis (tuberculosis undergoing sclerotic or caseous evolution, atrophic sclerosis, dry or liquefying necrosis, caseation) can be produced, remote from the jpoint of inoculation, by injecting into the 15 Compt. rend. Acad, des sci., 1913, 157, 1171. 110 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS cellular tissue of a guinea pig, a fluid obtained by macerating (in distilled water at 42°) living bacilli treated with pure sulphuric ether and then washed to remove all traces of the broth culture. This liquid, filtered through a Chamberland bougie, is an aqueous solution of protein substances, some of them dialyzable and others in a colloidal state. In other experiments, the same workers first treated the bacilli with ether and then macerated them at 70°C. in distilled water. The liquid, filtered through a Chamberland filter and concentrated to one-tenth of its original volume, was subjected to dialysis in order that the dialyzable substances and the colloidal substances might be utilized separately. The injection of this extract was followed by reaction phenomena of cellular hyperplasia, localized particularly in the lymphoid organs and in the lung. With the dialyzable substances there were observed in addition very small dissemi- nated lesions of necrosis, which were lacking when the colloidal sub- stances were employed alone. It is therefore impossible to attribute the origin of the process of caseation to the fats and to the waxes which make up the larger part of the ecto- plasm of the tubercle bacillus. It is much more probable that the cellu- lar ferments play the principal role. In caseous pneumonia, for example, many cells die and disintegrate when the inflammatory exudate, originally composed of dust cells, leucocytes and fibrin, becomes rapidly caseous. Now these cells contain ferments which, according to Jobling and Petersen,16 are inactive because the tubercle bacillus itself contains a substance, probably a lipoid soap, which is inhibitory. It is known in fact that trypsin becomes inactive when exposed to a temperature of 30°C. in the presence of non-saturated fatty acid soaps, and Jobling and Petersen have demonstrated that the bodies of the tubercle bacilli contain non-saturated fatty acids which, when saponified, inhibit in vitro the action of trypsin and leucocytic protease. The absence of autolysis in caseation, as in anaemic infarct, is said to result from substances of the same nature. It should be granted then that caseation of tubercles is the result of the direct and localized toxic action of the bacilli and of their diastatic or toxic secretions upon the giant cells which contain them. It shows 16 J. Expcr. Med., 1914, 19, 251; 383. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 111 itself histologically in a granulo-jatty degeneration and a nuclear fragmentation (Chaussee and L. Pissot).17 The center of the giant cell becomes a culture medium for the devel- opment of a greater or lesser quantity of bacilli. When, with the progressive softening of the caseous material, a perivascular tubercle breaks into a blood vessel or into a neighboring lymphatic space, the enclosed bacilli, swept into the lymphatics or the blood stream, become quickly the prey of new polynuclear leucocytes and then of Fig. 8. Nodular Tuberculosis Destruction of a bronchiolar wall by a caseous tuberculous nodule (from a preparation of Professor Letulle). mononuclear cells which will reproduce other and similar tubercles either near or far from the point of initial infection. If the caseous tubercle ruptures into a lung alveolus or into a small bronchus, it produces at first an isolated area of tuberculous pneu- monia, then of tubercle formation in the alveoli, for which the fixed cells of the lung serve as a passive support until they become dissoci- 17 Compt. rend. Acad, des sci., 1911, 152, 108. 112 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS ated and destroyed in their turn by the progressive extension of the caseous process. “One may state,” says Letulle, “that aside from the examples of primary tuberculous pneumonia of purely respiratory origin, which are in truth very rare, and experimental so to speak, there are innumerable cases in which the vascular origin of pneumonic caseous lesions cannot be questioned. “As an example, miliary tuberculosis of the lungs (the hematogenous or fymphogenous origin of which according to the case, cannot be doubted) by breaking through a neighboring wall, gives rise often prematurely to the complete series of specific obliterating changes in the bronchioles, which are the most important causal elements of caseous pneumonia.” This is shown very well in figures 6, 7 and 8. The tubercle therefore is truly a lymphatic production in all cases. The fixed cells of the different organs where it develops play no active role in its histogenesis. C. TUBERCLE BACILLUS LESIONS WITHOUT TUBERCLE FORMATION Until the last few years it was accepted that in all forms of tuber- culosis described by clinicians and pathologists the presence of tuber- cles was an essential characteristic (miliary tuberculosis, caseous pneumonia, scrofula, etc. ). Contemporary studies however show that infection by the bacillus may exist, and may be produced experi- mentally in animals, without any tubercles being formed. Landouzy18 and his pupils, Leon Bernard, Salomon, Gougerot,19 Laroche, Jean Troisier, must receive the credit for having, since 1882, drawn and held the attention of observers to the previously unrecog- nized forms of tuberculous infection without tubercle formation (■non-follicular bacillosis) which, as we know today, may involve all tissues and all organs (see Chapter VII, B). Proof of this was brought by those whom I have just named, as well as by Darier, Andre Jousset and Braillon, Claude, Oettinger, Triboulet, and the work of Poncet and his pupils, although the latter have gone much too far in asserting, without histological or experi- mental proofs, that “the various disease accidents ocurring in tuber- culous patients should in general be regarded as manifestations of the tuberculosis. ” 18 Legons cliniques de La Charite, 1881-86; de l’Hopital Laennec 1891-1910. 19 These, Paris, 1908. HISTOGENESIS AND EVOLUTION OF THE TUBERCLE 113 Leon Bernard first, and later A. Jousset and Gougerot adduced a number of definite facts to prove the existence of these inflammatory reactions without tubercles which are capable of being produced through tubercle bacillus infection. Anatomically they have been found in the kidney, in the skin, in the endocardium, etc., and no one longer denies that it is possible for them, through an extension to “fluid tissue” such as the blood, to establish a more or less limited septicemia which frequently precedes the primary localization and almost always the secondary localizations (see Chapter XVIII). L. Renon and E. Geraudel20 studied the non-follicular bacilloses from the histological point of view. They clearly brought out the fact that the granulation,21 the tubercle, the nodule, are macroscopic forms but should not be regarded as histological realities. In the different organs these “nodular” lesions are lesions of inflammation and, in the lung, they are lesions of pneumonia, borrowing their special appear- ance from their limited extent and their focal arrangement. “The composite anatomical picture found in a tuberculous lung,” says Itenon, “corresponds to the aggregate of the successive pneu- monic extensions which make up the clinical history of every pul- monary tuberculosis. To these pneumonic lesions tubercle formation may or may not add itself here and there; but the essential element of the lesion is always the pneumonia.” This conception is obviously the only one which enables us to understand why the tuberculous infection, according as it is single or repeated, according as it is caused by bacilli of greater or lesser virulence and in greater or lesser number, in individuals tubercle-free or relatively immunized by one or more previous benign infections, manifests itself at times by small localized foci of inflammation which tend to fibrous transformation and healing; again by foci of severe inflammation resulting promptly in cell necrosis, and afterward in the formation of true tubercles. 20 Compt. rend. Soc. de biol., 1913, 75, 699. 21 See Chap. VI, §A, for explanation of these terms. CHAPTER VII THE PRINCIPAL PATHOLOGICO-ANATOMICAL TYPES OF INFECTION BY THE TUBERCLE BACILLUS A. TUBERCLE BACILLUS SEPTICEMIA.—MILIARY TUBERCULOSIS Infection of the human body by the tubercle bacillus is not always manifested from the outset by the formation of tubercles. It fre- quently happens, especially in the young, that the infection assumes the guise of a general infectious disease similar to typhoid fever, without any localized lesion to cause the formation of giant cells and the later evolution of the tuberculous process, and without any characteristic incubation period. The clinical entity of this form of tubercle bacillus septicemia was established in 1882 by Landouzy,1 who called it “ tyyho-bacillose” by reason of the frequent resemblance of its symptoms to those of true typhoid fever, from which it is differentiated by the absence of intestinal catarrh and rose spots, and by the frequent presence of tubercle bacilli in the circulating blood. When benign and atypical it frequently causes errors of diagnosis, being mistaken for a protracted mild form of typhoid fever or for a manifestation of grippe. When quickly fatal, which is exceptional, there is no ulceration of Peyer’s patches nor any visceral localization, only the diffuse congestive lesions in the different viscera, as in all septicemias, and at times a few small gray translucent granulations “not enough to cause local symptoms and only just sufficient to establish the identity of the disease.’’ “In the great majority of cases,” states Landouzy, “after 3 to 4 weeks of continuous fever, accompanied by more or less marked pros- tration progressing generally to a definite typhoid state, with dry tongue and more or less hypertrophy of the spleen (a condition which, 1 J. de med. et de chir. prat., 1885, 488;—Legons cliniques de La Charite et de l’Hopital Laennec 1885-1891; Semaine mcd., 1891, 11, 225;—Presse M6d., 1908, 2, 681. 114 PRINCIPAL PATHOLOGICO-ANATOMICAL TYPES 115 according to the intensity of the manifestations, is diagnosed as typhoid fever, a mild typhoid, or a febrile gastric disturbance), the patient begins to convalesce. ‘‘Usually however the convalescence is not genuine; the patient does not regain his normal spirits; the keen appetite of the typhoid convalescent fails to appear; the lost weight is not regained. After a few weeks or months, there appear abruptly or stealthily the signs of a localization of the tuberculosis, most frequently pulmonary or pleural; rather often meningeal in the child. “Let us take for example the case of a boy of 7 years, without past history of illness, suddenly taken ill with fever absolutely like a typhoid of moderate intensity, except for the absence of intestinal catarrh and rose spots. At the fourth week the child begins to con- valesce and is taken to the country; he returns looking well but his cheeks are less full and he is less active than is ordinarily the case with a child after typhoid fever. The winter passes without trouble; then one morning, he does not feel well, has a headache, vomits, has some fever, and convulsions. In a few days death occurs from tuber- culous meningitis. “At times the convalescence which follows the “typho-bacillose” appears entirely frank and genuine; apyrexia is complete, cure is all but attained; and yet here again a localized tuberculosis intervenes sooner or later, and more or less abruptly. “These cases which simulate typhoid infection (typhoiques bacil- laires), although cured of their fever, almost always remain in a state of “ gestation ” as regards their tuberculosis; and a few weeks, a few months or several years after the initial acute septicemia they reveal themselves tuberculous. They develop the pathology and symp- tomatology of tuberculosis therefore only after having passed through the stage of bacillosis.” (Landouzy.) Acute tubercle bacillus septicemia was produced experimentally by Yersin in 1888, later by Gougerot in the rabbit and by C. Guerin and myself in 1906 in the bullock. In these animals, as in man, it terminates at times in complete cure without the formation of visible tubercles in any organ. However, except when the animals have been infected with bacilli culturally attenuated or modified in viru- lence, tuberculous localizations are always observed in the various lymph gland groups and in the lungs. The disease then develops in its usual chronic form, or remains in the state of latent tuberculosis. 116 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Present-day clinicians, and particularly the pediatricians, recognize that tubercle bacillus septicemia is extremely frequent, either as a primary manifestation or (probably much more commonly) secondary to a sudden discharge into the circulation from a primary and up to that time latent tuberculous focus (tracheo-bronchial, for example) (Hutinel). Until described by Landouzy, Gougerot, Loederick, Widal, Hutinel, Weil and Mouriquand, Leon Bernard, and others, it passed unnoticed. Undoubtedly it plays a considerable role in the pathogenesis of tuberculosis and is probably very often an important factor in immunization. We shall have occasion to return to this subject. In addition to this last type of infection there occurs fortunately much less frequently, the massive tuberculous infection, apparently primary, which manifests itself in man by a subacute septicemic form with more or less immediate single or multiple localizations. Again to Landouzy and his pupils are we indebted for the best study. The clinical type may first evolve like the preceding and end with a pleurisy, a meningitis or a more or less “galloping” phthisis; or else it may lead from the beginning to an acute milary tuberculosis, the “granulie” of Empis, characterized by the early formation of an immense number of small gray translucent tubercles with caseous opaque centers in most of the viscera, especially in the lungs. In the course of this severe form of infection there is frequently observed, particularly in young children, a specific cutaneous eruption which takes the form of disseminated and discrete papules scarcely larger than the head of a pin. These papules are soon transformed into vesicles which break and dry, leaving a little crust surrounded by a zone of brownish pigmentation. The eruption occurs in several successive outbreaks over a number of days and weeks, on the but- tocks, the genital organs, the inner surfaces of the thighs, the abdomen, and more rarely on the chest. It results from the septicemia and the accumulation of tubercle bacilli in the skin capillaries. Tileston2 was able, in 70 per cent of the many cases which he observed, to demonstrate the presence of the bacillus in the papules, as did Landouzy and Gougerot in a case of erythema nodosum (see Chapter XVII). Acute miliary tuberculosis (“granulie”) results from the irruption into the circulatory system of a large quantity of tubercle bacilli having their source in the softened caseous matter of one or more 2 Arch. Internal Med., 1909, 4, 21. PRINCIPAL PATHOLOGICO-ANATOMICAL TYPES 117 tubercles constituting the true primary localization. An intense febrile reaction is produced, the temperature rises and remains in the neighborhood of 39.5° to 40° C. until the end of the illness, which almost always terminates in death in 2 to 8 weeks, depending upon the severity of the infection and whether the miliary tubercles are particularly abundant in the nervous centers or in the lungs. In some rare instances where the miliary eruption was less intense, the acute symptoms have been seen to improve and the disease to develop into a chronic tuberculization. Such an outcome is however, quite exceptional. B. LATENT TUBERCULOUS INFECTION Infection by the tubercle bacillus manifests itself most commonly, in man as well as in naturally susceptible animals, by the slow for- mation of tubercles in one or more groups of lymph nodes more or less distant from the site where the virus originally penetrated the body. It was formerly believed and is still frequently taught, altogether wrongly, that the infecting element always leaves its mark at the point where it enters. The author of this doctrine, Cohnheim,3 has formulated it into a law in the following terms: At whatever point the tuberculous virus is introduced and remains during a sufficient period of time, a tuberculous or scrofulous lesion is created. As regards the primary localization, the portal of entry of the virus plays the important role in the primary localization. Once introduced into the body the virus propagates itself and is disseminated in accordance with local conditions, following the natural lymphatic and venous channels. In the chapters to follow, in studying the mechanism of tuberculous infection, we shall see that the infection first occurs unobtrusively, and that unless it is effected in massive form (as happens at times accidentally, or as is almost always the case in experimental inocula- tion of animals), it remains latent in the lymph or blood system for a longer or shorter time and discloses its existence only by conferring upon the infected organism the capacity to react to tuberculin. It is only after this varyingly prolonged period of latent blood or lymphatic infection that a primary tuberculous lesion is created, with or without the formation of tubercles, and the localization or fixation of the primary lesions is determined either by certain anatomical relationships 3 Die Tuberkulose vom Standpunkte der Infectionslehre; Leipz., 1880; 2nd edition, 1881, French translation by Musgrave-Clay, 1882. 118 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS within the organs (such is the case with the lungs), or by other circum- stances, mechanical, physiological or accidental, which bring about the arrest, against the endothelial wall of a lymphatic or blood capil- lary, of a polynuclear leucocyte containing one or more phagocyted bacilli. At the present time it is well established that apparently normal lymphatic glands, removed at autopsy from subjects who have died at all ayes whether from acute diseases or following accidents, harbor living and virulent tubercle bacilli whose presence is revealed only by the inocu- lation of these glands into very susceptible animals, such as the guinea pig. As early as 1890-92, Loomis,4 then Pizzini5 had brought proof of this, and the proportion of latent bacillus carriers found by them in the course of their autopsies of non-tuberculous subjects was, for Loomis’, 26.6 per cent; for Pizzini 42 per cent. Proofs of a similar nature were later furnished by Spengler,6 Kalble,7 A. MacFadyen and MaeConkey,8 Harbitz, Weichselbaum and Bartel,9 Rosenberger.10 With C. Guerin and Delearde11 I have myself demonstrated the frequency of latent infection of the tracheo-bronchial nodes, without evidence of pulmonary lesions, in infants dying from non-tuberculous diseases and presenting no clinical signs of tuberculosis. These facts have since been experimentally confirmed by several observers (Goodale,H. Wright and Smith, L. Rabinowitsch,12 Lignieres, Moussu, Vallee, Junack). Particularly abundant proof has been furnished by the veterinarians who, in the abattoirs, have systematically inoculated guinea pigs with the lymphatic glands of calves or swine which showed no evidence of any tuberculous lesion (Joest, Noack and Liebrecht,13 Rievel14 Linnenbrink, Jonske, Nieberle,15 Emshoff and Semmler, Haeutle,16 and others). 4 Studies from the Loomis Lab., 1890, Vol. 1. 6 Ztschr. f. klin. Med., 1892, 21, 329. 6 Ztschr. f. Hyg., 1893, 13, 347. 7 Miinchen. med. Wchnschr., 1899, 46, 622. 8 Brit. Med. J., 1903, ii, 129. 9 Wien. klin. Wchnschr., 1905, 18, 241. Am. J. Med. Sci., 1905, 130, 95. 11 Compt. rend. Acad, des sci., 1906, 142, 1136. 12 Berl. klin. Wchnschr., 1907, 44, 35. 13 Ztschr. f. Infektionskr. . . . d. Haustiere, 1907, 3, 257. 14 Deutsch. tierarztl. Wchnschr., 1909, 17, 347. 15 Ztschr. f. Infektionskr. . . . d. Haustiere, 1913, 13, 59; 141. 16 Centralbl. f. Bakt., 1914, 74, 91. PRINCIPAL PATHOLOGICO-ANATOMICAL TYPES 119 The discrete glandular infections, that is to say those which result from a small amount of virus and are produced by attenuated bacilli, tend usually to remain latent. They then play, as will be seen later, a very important role in immunity against tuberculosis. They are very frequently encountered at autopsy, especially in children and adolescents, occurring most commonly in the peribronchial and mediastinal nodes, then, in diminishing order, in the glands of the neck, of the axilla, of the groin, in the cubital and popliteal, and finally in the retro-auricular and submaxillary glands. In the adult, after the twentieth year, they are found much more rarely or, to be more exact, are less apparent; as the glands are then the seat of sclerotic lesions which cannot always be discovered on macroscopic examination and which are visible only under the micro- scope. In the majority of cases these sclerotic lesions still contain bacilli, at times very few and impossible to disclose in sections, but alive and virulent as shown by animal inoculation. C. PROGRESSIVE TUBERCULOUS INFECTION.—PREDOMINANCE OF PULMONARY LOCALIZATIONS It is a well known fact that in tuberculous infection, in man as well as in cattle, pulmonary localizations are the most frequent. This does not imply that they are the first from the standpoint of priority, for in reality they are almost always of later date and consequent to tubercle formation in a gland of the cervical, tracheo-bronchial or mediastinal groups. But it is the lung localizations which, in the majority of cases, make known their presence by the more or less serious functional disturbances (cough, hemoptysis, attacks of congestion) which char- acterize the early stages of the disease. This predominance results from the fact that in the loose connective tissue surrounding the alveoli and the small bronchi, the lymph spaces and blood capillaries are the seat of a circulation which is less rapid than in any other organ, and this slowing is particularly marked at the apices, which have a lesser degree of elasticity as well as smaller blood vessels. Dust in suspension in lymph or blood, bacteria and degenerated or dead leucocytes have, as a result, a natural tendency to be retained there. Adherence to the walls of the lymph or blood capillaries occurs here with more intensity than elsewhere. The anterior margins of the middle lobes and the circumferences of the lower lobes possess the same character in lesser degree: there too the 120 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS localizations of tuberculosis are often observed and these same parts are the most frequent site of parenchymatous inflammations of bacterial origin (pneumonia in man, glanders in the horse, parapneu- monia in the ox, and various infectious diseases in a large number of animals). At times it happens that the tubercles are but few and are scattered among several lobules; again they form real conglomerate masses round about or in the center of the same lobule or of the same lobe; or again they appear in small isolated groups which tend to coalesce through the large number of smaller younger tubercles encircling the older larger lesions. Caseation then involves a part of a lobe or even an entire lobe, producing large cavities, being surrounded at times by an inflammatory zone of broncho-pneumonia and hepatiza- tion, which is red or gray according to the amount of infiltrating caseous matter or the degree of secondary bacterial infection. All of these lesions may combine to present an infinite variety of aspects. They encroach little by little upon the surrounding tissues, following always the lymphatic channels. Grancher 17 insists quite rightly on this fact which he had carefully noted: “ The blood vessels, ” says he, “and specially the lymphatic vessels, are the true paths of conveyance of the miliary tubercle. In fact, when tuberculous granulations are discrete, they can very easily be seen sowing themselves one by one in the interlobular spaces and thus circumscribing the base of the lobules. Often, where the spaces are very large, three or four tubercles join together and form a sort of small focus, whence they radiate out always along the lymphatic route. The tuberculous matter behaves like an injection fluid; it follows the path of least resistance and there are often found, at a fairly great distance from the focus of infection, tuberculous granu- lations which occupy the perilobular lymphatics long before the lobule itself is involved.” The caseous softening of the tubercles, the inflammatory infiltration of the neighboring tissues, the degeneration, the destruction, the defensive reactions of which these tissues are the seat, remain limited to the lungs but rarely. They frequently extend to other organs. 17 Arch, de physiol, normale et path., 1878, 2. s., 5, 1. PRINCIPAL PATHOLOGICO-ANATOMICAL TYPES 121 D. LOCALIZATIONS IN THE PLEURA One would expect the pleura, whose lymphatic system is in direct communication with that of the lung, to be very frequently involved, and this is indeed the case. In those affected with phthisis the walls of the serous sac are almost never free from the disease. Adhesions or serous effusions more or less rich in leucocytes and bacilli develop (dry pleurisy or pleurisy with effusion). It is now known, through the work of Lanclouzy, Kelsch and Vaillard,18 Netter, Weichselbaum, Gombault and Chauf- fard,19 Le Damany,20 Prinz Ludwig, and others that pleurisies are almost always of tuberculous origin. “Every individual,” says Landouzy, “ who cannot furnish a satisfactory reason for his effusion, whether an infection (scarlet fever, puerperal fever, etc.), a dyscrasia (rheumatism), or a trauma (fracture of the rib, pulmonary infarct) is a tuberculous subject, even though he is plump and well nourished, and declares that he is well and that there is no past history of tuber- culosis in himself or in his family. So-called pleurisy a frigore is ordinarily a manifestation or vicarious expression of tuberculosis.” But tubercles of the pleura, when not concomitant with wide spread lesions of the lung, tend generally to become sclerotic. They represent a benign form of tubercle bacillus infection which, at least to outward appearances, is usually capable of cure (see Chapter XIII). E. OTHER LOCALIZATIONS Along with its localization in the lungs or pleura or at an isolated point, the tuberculosis may, from the beginning, localize or sow itself in any organ provided with a plexus of lymphatic or blood vessels; but it has an evident predilection for serous membranes both visceral and articular, for lymphatic glands both superficial and deep, and for mucous membranes and skin where lymphatics are most abundant (phlyctenular conjunctivitis for example). It never 'primarily invades the highly specialized tissue elements. The latter are involved only secondarily, through extension of the processes of caseation or sclerosis which, as we have already seen, complete the evolution of the tubercle. Whether we have to do with lupus, with tubercles in different parts 18 J. de physiol, normale et pathol., 1906, 8, 162. 19 Semaine Med., 1896, 16, 81. 20 Semaine Med., 1897, 17, 427. 122 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS of the digestive tract, in the peritoneum or abdominal viscera, in the larynx, the eyes, the nose, the ears, the kidneys, the genital organs, the nervous centers, in the bones or in certain muscle masses clinical and experimental evidence are in agreement that this rule is absolute. The chapters to follow will furnish an abundance of proof. Of the autopsy observations, Biedert’s figures on the principal localizations of tuberculous infection will be found sufficiently accurate. They summarize those already published by Simmonds, Rilliet and Barthez, Steiner and Neureuther, Widerhofer, Steffen, and others and are based upon 3104 autopsies of adults and 1346 of children. They may be condensed as follows: LOCALIZATIONS ADULTS CHILDREN Pulmonary lesions in per cent 91.2 per cent 79.6 Intestinal lesions in 40.7 31.6 Glandular lesions in 26.8 88.0* Peritoneal lesions in 18.0 18.3 * 78 per cent bronchial, 10 per cent mesenteric. CHAPTER VIII MECHANISM OF TUBERCULOUS INFECTION Penetration of the Virus into the Body by Way of the Skin and Mucous Membranes A knowledge of the mechanism by which the tuberculous virus penetrates into the body of man and susceptible animals is of extreme importance, since it must serve as the basis for both individual and collective measures of defense against the spread of the disease. Therefore it has been the object of a great deal of study from both the clinical and experimental standpoints. To thoroughly understand this mechanism one must have in mind the essentials of what we know today regarding the anatomy and physiology of the lymphatic circulation. It will not be amiss then to briefly recall these subjects at this point. A. LYMPHATIC CIRCULATION, LYMPH, GLANDS, R6LE OF THE LEUCO- CYTES IN TUBERCULOUS INFECTION All the organs possess a system of lymph capillaries which penetrate and anastomose within the meshes of the connective tissue, establish communications between them and the serous cavities and form small lakes and widened spaces. This capillary network is lined with a single layer of endothelial cells; it drains the lymph (or interstitial fluid) in which the tissues are bathed, and collects it into the lymphatic vessels which conduct it to the lymph nodes. The lymph, produced throughout the body by cell metabolism,— and in particularly large quantity in the abdominal viscera,—contains chiefly, in addition to a basal fluid substance analogous to blood plasma (but more watery and less rich in albuminoid material), white cells or leucocytes. These cells differ in number according to the animal species (about 7200 per cubic millimeter in man; 7500 in the dog; 11,300 in the rabbit and only 180 in the frog). The leucocytes are motile after the manner of the amebae. They put forfh protoplasmic extensions, at times rounded, lobate and broad, 123 124 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS again filiform and slender. By virtue of their motility they are able to penetrate the endothelial cells of the lymphatic capillaries, to enter into the blood capillaries or to leave them and scatter about in the tissues, to migrate to the cutaneous or mucous surfaces and into the interior of the alimentary tract. Like the amebae, they have the faculty of ingesting solid particles and cellular or bacterial debris. They can even attack and absorb degenerated cells. There are several varieties which are found in unequal number in the lymph. Some (the lymphocytes), of which the diameter is equal to or a little less than that of the red blood cell (5 to 8 microns), are round or oval, with a central nucleus filling almost the entire cell and leaving a very narrow margin of protoplasm. These lymphocytes contain no granules and are incapable of ingesting foreign bodies; they are therefore not phagocytes (Metchnikoff). They are only slightly motile. In human blood they constitute about 31 to 25 per cent of the white cells. They are rather more numerous in children and in adults during digestion, less numerous in the aged. A second variety of leucocytes is made up of the medium and large mononuclears (lympholeucocytes of Pappenheim). They differ greatly in diameter, varying from 10 to 25 microns, are round or irregularly oval and contain a large kidney or horse-shoe shaped nucleus, often divided into two lobes. They are markedly phagocytic, ingesting altered leucocytes, degenerated tissue cells and certain bacteria such as the Bacillus leprae. Their proportion in the blood is relatively small: 4 to 8 per cent of the white cells. Their ameboid movements are rather sluggish and they end by becoming fixed in the connective tissue. The leucocytes with neutrophile granules, the so-called polynuclears, or more exactly the polymorphonuclears, constitute a third and more numerous variety (40 to 75 per cent, average GO per cent). They are 10 to 14 microns in diameter and have a polylobed nucleus, of quite variable shape, which is formed of 2 to 4 irregular masses joined one to another by fine filaments, and which takes a dark color with the triacid stain of Ehrlich (mixture of methyl-green, methyl- orange and acid-fuchsin). Their protoplasm is studded with neutro- phile granules which stain violet. These cells, which are very motile, readily phagocytize bacteria and in particular the bacillus of tuberculosis. 125 MECHANISM OF TUBERCULOUS INFECTION The cells with acidophile granules, or eosinophile myelocytes, make up a fourth variety. They also have a polylobed nucleus provided with cytoplasmic granules. Their protoplasm contains an abundance of granules which color intensely with the acid stains, eosin or orange. They are also phagocytic, but to a less marked degree. They con- stitute 2 to 4 per cent of the white cells of the blood and are found to be considerably increased in a wide variety of diseases. A last variety of leucocytes is represented by the cells with basophilic metachromatic granules (mastzellen of Ehrlich). These cells are round, polygonal or tapering, at times even branching. In diameter they are from 8 to 12 microns and are found in only very small numbers in human blood (0.05 per cent). They also contain granules which stain by Gram and Ziehl; at times also small vacuoles. The nucleus, which is always very large in proportion to the protoplasm, takes a pale blue tint with Unna’s polychrom blue. They appear to have no phagocytic function and they become more abundant in the course of certain pathological conditions. These various white cells, carried about by the lymph in the lymphatic capillaries, fill the sinuses of the lymphatic glands. They enter in by the vessels afferent to the convexity of the glands and there undergo multiple transformations. It is in these glands that the lymphocytes, and probably also the large mononuclears, are generally formed. There too the eosinophiles are produced, and perhaps also the polymorphonuclear neutrophiles which accumulate in the glands in great number, bearing innumerable particles and bacteria phago- cyted before their arrival. It is there that the processes of intra- cellular digestion are chiefly affected, and the lymph node, thanks to its delayed circulation, serves not only as a sort of filter but also as a laboratory wherein is elaborated a host of protein, carbohydrate and fat splitting ferments (trypsin, amylase, lipase, etc.). The younger the individual, the more active is this laboratory and the more efficient the filter. Little by little, with increasing age, the glandular tissue undergoes sclerosis, especially at the hilus (Salimbeni and L. Gery1); the capsule thickens, the connective tissue frame- work hypertrophies, and in the pulp the lymphocytes become more and more rare, while the macrophages predominate, many of them containing yellow pigment and having a very acidophile protoplasm. Whenever a tubercle bacillus, deposited upon skin surface or 1 Ann. de l’lnst. Pasteur, 1912, 26, 577. 126 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS mucosa, or introduced into the healthy body in any other way (inhalation, traumatism), finds itself near a motile polymorphonuclear leucocyte, it immediately becomes the prey of the latter which takes it along in the lymphatic or blood circulation. The digestive ferments of the leucocyte do not succeed in digesting the bacillus because, as we have seen before, it is protected by an extremely resistant fatty waxy ectoplasm. The bacillus remains enclosed in its phagocyte or, if rejected by the latter, is immediately ingested by another which in turn transports it for a longer or shorter time in the circulation. This continues until such time, often remote, as the leucocyte host, poisoned and then degenerated and killed by the secretions of its parasite, becomes the prey of one of the large mononuclears which pave the endothelium,—for it is one of the essential functions of the large mononuclears to absorb and digest degenerated and dead cells. A little later, this mononuclear too will be subjected to the effects of the toxins which will have increased in the cell with the multiplication of the original bacillus. These effects will soon manifest themselves by that peculiar pathological formation which we have studied under the name of the giant cell, the first stage of the tuberculous lesion. But between the moment when the virulent germ has affected its entrance into the body and the beginning of the formation of this first tuberculous lesion, an often considerable interval of time may have elapsed, which the infected organism has utilized to set at work its defensive reactions and its means of elimination. Thus the defensive reactions, in a great many cases, come into play with sufficient success to wall in, as it were, the focus of infection by a mass of large mononuclears which become organized into dense connective tissue, either in a lymph node sinus or in a lymphatic or venous capillary of the lung or of some other organ. The lesion then remains latent and may continue so during several years. In other cases, still more, fortunate, one or more bacilli phagocyted by motile polymorphonuclear leucocytes are evacuated from the body through the liver and intestines, as are pigment granules or inert foreign bodies before they have been able to initiate the formation of a giant cell lesion at a point of arrest (lesion diarrU). When the infection is produced by very virulent bacilli, that is to say by those which are perfectly adapted to the invaded host, and when these germs are present in a large number, or cause, though attenuated, a massive infection, the normal processes of defense and MECHANISM OF TUBERCULOUS INFECTION 127 elimination are no longer able to perform their protective function and the fact of this inability is soon manifest in the group of lymphatic glands nearest to the portal of entry of the virus. Then,—and then only,—the law of Cohnheim, of which I spoke in the preceding chap- ter (VII, B), holds good. The leucocytes engorged with bacilli and rapidly poisoned, become arrested in the cavernous sinuses of the first lymph node encountered on their way, and there they succumb. Their debris (and virulent contents) become the prey of mononuclears which quickly organize themselves into giant cells and the same lymph node becomes the seat of several tubercles whose later casea- tion will result in the pouring of a relatively considerable number of bacilli into the efferent lymph channels. Through the latter the bacilli will go elsewhere to create other foci of tuberculosis, near-by or at a distance. Such is the mechanism of infection by the tubercle bacillus. One will readily understand the enormously important role played therein by questions of quantity and source of the virus, and also by the ana- tomical structure of the organs into which the virus is borne by the lymph or by the blood. B. PORTALS OF ENTRY OF THE VIRUS IN LATENT TUBERCULOUS CONDITIONS It seems at once obvious that in cases of latent infection,—the most frequent form, as attested by the very large proportion of apparent healthy human or bovine subjects who react to tuberculin,—it is quite impossible to disclose the route followed by the virus in its original invasion of the body. Whatever may have been the portal of entry of the infection, it closed itself behind the migratory leucocyte which phagocytized the bacillus, without the slightest trace of the passage remaining. The long wanderings through the lymphatic channels, then into the blood stream, remain equally masked, and only when the virus causes somewhere the formation of a tubercle, can its presence, until then unrecognized, be revealed by a positive tuberculin reaction. These cases of latent infection, brought about by very small numbers of bacilli or by those less virulent—that is to say not adapted to the organism invaded,—are observed with extreme frequency in man and in animals spontaneously susceptible to tuberculosis. 128 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Orth had already noted them in 1876; but it was chiefly Loomis2 as the first (1890), followed by Pizzini (1892) and Kalble (1899) who attracted attention to them by proving that bronchial or other glands, even though normal on microscopic examination, and removed from subjects showing no sign of tuberculosis, nevertheless contained bacilli. Since then, the investigations of A. MacFadyen and MacConkey,3 of Harbitz, of Weichselbaum and Bartel, of Rosenberger (1905), those which I published with C. Guerin and A. Delearde in 1908,4 those of L. Rabinowitsch (1907), of S. Arloing (1909)5 and still others, show that in man, in early as well as in adult life, and also in cattle, tubercle bacilli are frequently found in the mesenteric, mediastinal and tracheo-bronchial lymph glands, where there exists no lesion nor any suspicion of tuberculosis. Harbitz systematically inoculated into guinea pigs the different lymph node groups from 91 non-tuberculous children, he obtained positive results 18 times with cervical, bronchial, mesenteric and retroperitoneal nodes. With C. Guerin I have shown that, in the bullock infected simply through living with other animals, tubercle bacilli may remain latent over a very long period (more than 11 months) without pro- ducing any lesion in the lymphatic system.6 Furthermore our experiments, coming after those of Schroeder and Cotton, of Rabino- witsch, of Ravenel, and of Moussu, proved to us that apparently healthy cows, free from udder lesions but reacting to tuberculin, eliminate virulent tubercle bacilli from time to time either in their dejections or in their milk. In human beings and in animals which harbor these latent tubercu- loses, it is therefore never possible- to indicate precisely either the man- ner of infection or the portal of entry of the bacilli, or even the time of their penetration into the body. There is reason to suppose that, in the majority of cases, the penetration is brought about through the medium of intestinal absorption. Indirect proof is furnished by the fact, pointed out some time ago by Nocard, that, during digestion, many microbes emigrate from the intestine with the chyle and are 2 J. Am. Med. Assn., 1891, 16, 98. 2 Brit. M. J., 1903, ii, 129. 4 Compt. rend. Acad, des sci., 1906, 142, 1136. 6 J. de med. vet. et zootech., 1909, 5. s , 13, 193. 6 Compt. rend. Acad, des sci., 1913, 156, 34. MECHANISM OF TUBERCULOUS INFECTION 129 recovered again, first in the mesenteric glands, next in the lymph of the thoracic duct, then in the blood and in the majority of the organs (liver, spleen, kidneys, bone marrow, muscles, etc.). Culture of lymph, or of blood or pulp from these organs, during the hours which follow digestion, frequently gives a growth, whereas none occurs if the material is taken when the animals are fasting. If the organisms have disappeared at the latter time, it is because the bactericidal action of the body fluids and phagocytosis have had time to act. But where the invading organism is the tubercle bacillus we know that these defensive processes, particularly phagocytosis, are relatively inert. To be sure, the intestine is not the only source of physiological infection of the tissues. It may also arise from the lungs. Never- theless the role of this organ is incontestably much less important, first because the respiratory channels are less charged with bacteria than is the digestive tract, and furthermore because the mucus and ciliated epithelium form obstacles which are overcome only with difficulty. Besides, if physiological bacterial infection of the body fluids is commonly effected by way of the lungs, inoculations of blood and different tissues would give positive results aside from dur- ing the period of digestion, and we have seen that such is not the case. Of all the organs of the human body, the skin offers the least favorable conditions for the retention, penetration and multiplication of the bacilli, as clinical observations demonstrate. Moreover, despite the frequency of contact of the integument with infectious material (sputum and bacillus-containing excretions), cutaneous tuberculosis is rare and, when it occurs in one of its forms (lupus, tuberculous ulcer, verrucous tuberculosis, scrofuloderma, anatomic tubercle), it tends generally to remain benign and localized, and to evolve slowly toward a chronic course or toward spontaneous healing {see Chapter XVII). Nevertheless, grave and rapidly progressive forms of infection are sometimes observed following contamination of an accidental or sur- gical wound by excrement or by saliva containing bacilli. In young calves, one encounters, not rarely, a tuberculous ulcera- tion of the navel resulting from contamination of the umbilical wound by infected bedding. C. TUBERCULOUS INFECTION THROUGH THE SKIN 130 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS PLATE V 1. Generalized tuberculous infection in the guinea pig. (Infection via digestive tract with human tubercle bacilli.) Caseous tubercles in the lungs, the liver, the spleen, the tracheo-bronchial and sub-lumbar glands. 2. The lymphatic stage of the infection produced by instillation upon the ocular conjunctiva, in the guinea pig. (Primary adenopathy of the cervical and tracheo-bronchial lymph nodes. Progressive extension of the tuberculous lesions by way of the lymphatics to the lungs and to other viscera). A. CALMETTE. Tuberculosis. Plate V. MECHANISM OF TUBERCULOUS INFECTION 131 Among the poorer Jewish or Mohammedan children, ritual circumci- sion, uncleanly done, becomes from time to time the point of depar- ture for a fatal tuberculosis. L. E. Holt, of New York, was able to collect from the medical literature 41 such cases, 17 of which, to his knowledge, ended fatally. Some of them survived 11 months, to die of meningitis, but the usual termination of this form of infection was generalized glandular and visceral tuberculosis. It has been reported that several children have been infected by one and the same operator as in the 10 cases reported by Lehmann. Experimentally, it is proved that tuberculous infection can be effect- ed through healthy skin after vigorous friction (that is to say under conditions which favor the migration of leucocytes from the super- ficial lymphatic network to the interstices of the epidermal cells), —and particularly skin which has been depilated or newly shaven,— by smearing upon it either bacillus-rich sputum or virulent cultures. Babes and Riegler, J. Courmont and Lesieur, Carl Fraenkel, H. Takeya and Dolcl7 have thus produced tuberculosis in guinea pigs, rabbits and cattle. These authors have established that transcu- taneous infection, usually sluggish, may leave no trace, nor develop local adenitis as a primary symptom, and may go on meanwhile to generalized disease. The proportion of positive results differs according as the skin is broken (75 per cent) or apparently intact (25 per cent). At times, particularly in the rabbit, the virus localizes itself in the lungs without setting up cutaneous or glandular lesions which can be demonstrated either macroscopically or microscopically. Here we have an example of pulmonary tuberculosis which originates at some distance from the respiratory tract, and whose portal of entry can not be discovered. At other times, especially if attenuated bacilli are employed, skin alterations in situ are produced similar to those characteristic of verrucous cutaneous tuberculosis or scrofulo-tuberculosis in man, and having the same tendency to heal spontaneously. D. INFECTION THROUGH THE MUCOUS MEMBRANES (OCULAR, NASAL, BUCCOPHARYNGEAL, GENITAL) At the mucous surfaces, about the points where the glandular acini pour out their secretions, the lymphatic vessels and sinuses 7 Arb. a. d. Geb. d. path. Anat. .. . . Inst, zu Tubing., 1908, 6, 710. 132 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS form a particularly dense network and the wandering leucocytes perform one of their chief functions, namely the sweeping from the mucous membranes of all bacterial or other impurities brought there by the external air. Moreover these mucous membranes are particularly exposed to infection by the tubercle bacillus, or to be more exact, they offer an easy path for the penetration of the virus into the subjacent lymph spaces. a. Ocular mucous membrane As I have demonstrated with C. Guerin and V. Grysez,8 if a particle of tuberculous sputum or a drop of a culture (containing for example 0.01 mgm. of virulent bacilli) be allowed to fall upon the eye-ball of a guinea pig, it suffices to set up a typical glandular tuber- culosis in that animal, without producing the slightest lesion in or about the eye. Such a glandular tuberculosis, beginning in the retro- mastoid gland, quickly invades the two retropharyngeal glands, the two glands of the anterior part of the neck, then the tracheal and bronchial glands, and extends in the course of 4 to 5 weeks to other visceral groups, to the glands of the liver hilus, to those of the mesen- tery, at times to superficial glands like the inguinal, to the spleen and almost constantly to the lungs. This mode of infection is, therefore, very certain and, aside from its severity (due to the virulence of the bacilli employed), the resulting form of tuberculosis is, in its first phase, singularly like that which characterizes scrofula in man, especially in childhood. Such is the resemblance that, on examining experimental animals, one is immedi- ately struck with the idea that human family contagion probably frequently takes the same path and that it then follows the projection, by a tuberculous cougher, of particles of saliva abounding in bacilli upon the ocular conjunctiva of healthy individuals. Infection by ocular instillation, producing as it does a sort of natural infection by the lymphatic channels without destruction of tissue and without lesions at the portal of entry of the bacilli, enables one, much better than does experimental inoculation, to study in the guinea pig the action of tuberculins, sera and chemical substances capable of influencing the progress of the tuberculosis. This pathway 8 Compt. rend. Soc. de biol., 1913, 74, 310. MECHANISM OF TUBERCULOUS INFECTION 133 may also be utilized to advantage for attempts to vaccinate with modified or attenuated bacilli. Spontaneous tuberculous infection of the eye is observed but rarely in man and cattle. It is scarcely ever encountered except in individuals who at the same time have tuberculosis elsewhere and where found it is as a complication of generalized tuberculosis. The choroid, the iris, or the anterior chamber, may be the seat of a tuberculosis of vascular origin. The conjunctiva and the cornea are sometimes infected by the hands soiled with sputum, or through small abrasions, the individual being phthisical. Lesions so produced are always accompanied by swellings of the preauricular glands, those at the angle of the jaw or of the cervical glands, and these engorgements are then the indicators of the nature of the disease. They are observed frequently for example in phlyctenular conjunc- tivitis, which is of tuberculous origin. b. Nasal mucous membrane The mucosa of the nose, despite the richness of its lymphatic and venous supply and in spite of the enormous amount of dust of all kinds which accumulates upon it at each inspiration, is not penetrated by the tubercle bacillus as readily as one might assume. The reason, which is easily understood, lies in the fact that the dusts,—bacterial, vegetable or mineral,—exercise for the most part a positive chemo- taxis toward the leucocytes, which by diapedesis go forth from the submucous lymphatic capillaries and spaces to phagocytize them. Many of the leucocytes however find themselves entrapped at the mucous surface in the folds of the turbinates, by a particular variety of sticky mucous secretion; they then become incapable of again passing through the mucous membrane to reenter the circulation, and the only fate remaining to them is to be expelled from the nose. The proof that events happen thus lies in the fact that bacilli are frequently found in the nasal cavities of perfectly healthy individuals. By 1894 Strauss had already demonstrated the fact by examinations of non-tuberculous patients and students attached to his service at the Charite Hospital at Paris. From these different subjects he collected the dust, the crusts and the mucous discharges contained in the nasal cavities, by means of small previously sterilized cotton swabs. The swabs were stirred about in a test tube of broth and 134 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS sterile water and the very turbid liquid thus obtained was injected into the peritoneum of guinea pigs. Of 29 subjects, there were 9 who harbored virulent bacilli in their nasal passages. Le Noir and J. Camus9 in repeating these experiments collected their material from the nasal cavities of the hospital personnel, physicians, students and nurses, in award in which were 14 tuberculous cases. This was done after the ward had been twice swept, the beds made, the bed tables dusted, etc. The cotton tampons covered with the mucous secretions were inserted under the skin of 9 guinea pigs by means of small incisions. Not one of the pigs became tuberculous. The same tests repeated with nasal mucus from the tuberculous patients themselves, all of whom had bacilli in their sputum, gave only 3 positive results in 13 cases. The nasal cavities constitute without doubt an excellent agent for filtering air and by means of the mucus there secreted in abundance they retain the greater part of the bacteria, prevent them from pene- trating further, and expel them. On this account they are only exceptionally the seat of lesions of primary infection. F. Fraenkel has never seen a single case. They are reported however in the medical literature, although they are extremely rare and are always accompanied by the characteristic engorgement of the glands of the neck. On the other hand, secondary infection of the nose in the form of lupus ulcers or of tuberculous ulceration is encountered fairly fre- quently in phthisical patients, who have probably re-infected them- selves locally with fingers soiled with bacillus-containing sputum. This secondary nasal infection is at times, especially in children, the point of departure for a tuberculous meningitis, through propagation by the lymphatics over the cells of the ethmoid bone (Naunyn and Schreider, Demme, Schwalbe and Flatau). c. Bucco-pharyngeal mucous membranes—Tonsils The mouth and the pharynx oppose the direct penetration of tubercle bacilli into the subjacent lymphatic system with the same natural obstacles as do the nasal cavities. The leucocytes, which migrate by diapedesis from the submucous vessels, sweep the bucco- 9 Gompt. rend. Soc. de biol., 1908, 65, 646. 135 MECHANISM OF TUBEECULOUS INFECTION pharyngeal cavity and are borne with the saliva to the digestive tract by the movements of swallowing, so that they do not reenter into the circulation with the bacilli which they have succeeded in ingesting. Exception must however be made for that region of the naso- pharynx which is occupied by the tonsils. These follicular lymphatic glands, made up of crypts and spaces tightly packed with lymphocytes and polymorphonuclear leucocytes, constitute, by virtue of their situation at the entrance to the digestive and the respiratory tracts, a very important system of defense against infections in general and tuberculous infections in particular. Many bacteria are there destroyed by the lymphoid cells. Tubercle bacilli, as a rule, do not there meet such a fate. In most cases probably they are swallowed or expelled in the efforts of cough- ing, but it nevertheless happens that migratory phagocytes conduct them to the lymphatic circulation. From a massive tonsillar infec- tion there can then result an engorgement of the retro-pharyngeal, sub-parotid or cervical glands. Poirier and, after him, George B. Wood,10 Robertson,11 and A. Edmunds, have demonstrated that the lymphatics of the tonsil anastomose with the whole pharyngeal system tributary to the deep glands of the neck. The collecting vessels proper of the tonsil per- forate the muscular coat of the pharynx a little above the large cornu of the hyoid bone and terminate in the nodes situated upon the inter- nal jugular, just above the posterior belly of the digastric, where this muscle is crossed by the anterior border of the sternocleido- mastoid. This point is located a little behind and below the angle of the jaw. When these glands are infected primarily by bacilli coming from the tonsils, the infection may extend to other cervical gland groups and propagate itself little by little to the tracheo-bronchial nodes, to the lungs and to other organs. The primary lesion of tuberculosis may also, in certain cases, develop in the closed follicles of the tonsil and there progress to casea- tion and ulceration. Pressure with a tongue depressor on the crypt then forces out the caseous matter. The anterior pillars of the gland 10 Am. J. Med. Sci., 1905, 130, 216. 11 J. Am. Med. Assn., 1906, 47, 1725. 136 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS are red, tense, and congested, and the local engorgement extends to the sub-sterno-cleido-mastoid chain of nodes. Tuberculous tonsils are usually neither large nor pedunculate, but on the contrary are small, pallid and closely set against the pillars. They do not occur frequently. John Wright, Hodenpyl, P. Nobe- court and Tixier12 insist that they are rare. The lesions are localized only at the base of the gland, so that if one removes merely the super- ficial portion (an operation to be avoided) the portion most infected is left with the stump. Tonsillar tuberculosis never remains limited to the tonsils which make up the ring of Waldeyer. It always extends almost simul- taneously to the pillars, to the soft palate, and to the posterior walls of the mouth, but it predominates in one gland or another. It is most often encountered, not as a primary lesion, but as a secondary localization in phthisical and other forms of tuberculosis. It tends to form an irregular ulceration, more or less concave, extensive, and granular at the base (A. Hautant).13 In the bullock, primary tuberculosis of the tonsils is exceptional. According to Max Devriendt,14 who has collected important data on the subject at the abattoir at Berlin, infection of these glands appears only secondarily in animals which are already possessed of other glandular or pulmonary lesions and is said to follow upon the repeated ingestion of tubercle bacilli. d. Genito-urinary mucous membranes Tuberculous infection of the genito-urinary tract is most often of hematogenous origin. It may result also from a direct infection of the mucosa (vulvar, vaginal, urethral or bladder) by sexual relations or by the introduction into the genital organs of catheters, irrigation tubes, sounds, fingers or other objects contaminated with bacilli. Verchere, Fernet, and Derville,15 have published several observations relative to wives who contracted pelvic peritonitis from contact with husbands suffering from pulmonary or epididymis tuberculosis. 12 Gaz. des hop de Par., 1908, 81, 1287. 13 Rev. de la tuberc., 1906, 3, 326. 14 Deutsch. tierartz. Wchnschr., 1909, 16, 720; 745. 18 These, Paris, 1887. MECHANISM OF TUBERCULOUS INFECTION 137 Primary localizations in the vagina occur only very exceptionally. According to Cornil, this is due to the fact that the thick stratified pavement epithelium which lines the vagina renders it difficult for bacteria to penetrate as deeply as the encompassing lymphatic spaces, despite the marked development of the latter. But if some leucocytes in their migration do carry in a few bacilli, these microor- ganisms will settle in the sub-pelvic system and there give rise to tubercles among the viscera of the pelvis. The fact that tuberculous salpingitis and metritis have been reported in young virgins, and also tuberculosis of the testis in young boys, is proof that such localizations may result from blood infections. Experimentation shows nevertheless that primary infection of the genital tract may be easily brought about. Thus Cornil and Dobroklowski16 introduced a few drops of a culture of tubercle bacilli into the vagina of a female guinea pig, being careful not to wound the mucous membrane. As early as the fifteenth day thereafter micro- scopic examination revealed the presence of minute tubercles in the uterus, beneath the intact epithelial lining. Gaertner17 inoculated the virus directly into the testicles of guinea pigs and rabbits and demonstrated that a certain number of females impregnated by these animals, became tuberculous, with vaginal and uterine lesions; and Baumgarten produced an ulcerative tuberculosis of the posterior urethra and of the neck of the bladder by instilling bovine bacilli into the urethra of male rabbits. In my laboratory, M. Breton18 readily succeeded in producing primary infection of the guinea pig bladder by directly introducing bacilli from a culture. There was thus produced, on the walls of the cavity, a granular infiltration which ulcerated and extended progres- sively to the sublumbar and retromesenteric glands, then to the tracheo-bronchial nodes and to the lungs. In such cases, the kidney always escaped. Clinically, tuberculosis of the bladder is extremely rare. When observed it is almost always secondary to a renal or prostatic lesion (see Chapter XV). 16 Internat. Congr. on Tuberculosis, Paris, 1888. 17 Ztschr. f. Hyg., 1893, 13, 101. 18 Ann. de l’lnst. Pasteur, 1910, 24, 820. CHAPTER IX TUBERCULOUS INFECTION BY THE RESPIRATORY PASSAGES The extreme frequency of localization of tuberculosis in the lungs and the fact that it often exists to the apparent exclusion of all other localizations in a certain measure justified the opinion which has prevailed until very recent years, that the tubercle bacillus usually penetrates into the body by way of the respiratory passages. Today however we possess more definite knowledge of the various processes of infection by the tubercle bacillus. Careful clinical observation and experimentation have led us to the conviction that the infection most frequently remains from the outset and for a long period,— sometimes even indefinitely,—concealed (occulte) in the lymphatic glandular system before inducing the formation of tubercles capable, by their development and later caseation, of discharging bacilli into the lymphatic or blood circulation. Therefore we are obliged to examine the question more closely and to modify our interpreta- tion of facts which up to now have been incompletely studied. A. MECHANISM OF PRIMARY TUBERCULOUS INFECTION OF THE RESPIRA- TORY PASSAGES Primary tuberculous infection of the lung or of the bronchi may be brought about either by way of the air passages, that is to say directly by bacilli borne by the air entering into these organs, or by the blood stream. In the latter event,—more frequent by far,— some leucocyte, containing its parasitic bacilli recently introduced into the body or derived from a more or less long standing focus of infection, is arrested in the interalveolar or peribronchial capillaries and becomes the point of departure for the formation of a giant cell. The disposition of the bronchial tree and. the character of the mucous membrane with which it is lined from the epiglottis onward presents quite special characteristics which tend to assure the pro- tection of the lung alveoli against so deep a penetration of mineral, vegetable or bacterial dusts suspended in the inspired air. 138 INFECTION BY THE RESPIRATORY PASSAGES 139 Within the larynx, the mucosa is covered in parts with stratified squamous epithelium (anterior surface of the epiglottis, the true vocal cords and small areas in the aryteno-epiglottic folds), and in other parts with a stratified cylindrical ciliated epithelium. The latter is traversed by numerous excretory ducts from mucous glands of alveolar or ramified tubular type, and encloses, especially at the level of the ventricles, true closed lymph follicles (tonsillae basilares) from which wandering leucocytes are constantly issuing forth. These closed follicles communicate at their bases with the lymphatic vessels and spaces of the submucosa, which discharge, either into a gland placed between the greater cornu of the hyoid bone and the superior border of the thyroid cartilage (Treichmann), or into the glands situated beneath the sterno-mastoid muscle at the level of the bifurcation of the common carotid (Sappey), or again into the glands situated at the two sides of the membranous portion of the trachea (Treichmann). The tracheal epithelium and that of the bronchi are also formed of ciliated cylindrical cells, with little islands of flat stratified cells through which the excretions of numerous mucous glands are poured out. The very extensive subjacent lymphatic system is distributed in two layers, one superficial, the other more deep, which pass to- gether toward the chain of glands strung along each side of the trachea and of the esophagus. As for the lung itself, it is known that it can be subdivided into a number of segments or lobules, each one of which in a sense represents the expansion of a bronchial branching divided into bronchioles which terminate in alveoli or vesicles. “The lung,” says Laguesse, “is a hollow tree ramifying almost to infinity, whose numerous branches are the bronchi, whose ultimate twigs or alveolar canals widen themselves, form alveoli and change their structure in order to assume the character of respiratory surfaces and to adapt them- selves to the function of blood aeration. ” The number of alveoli in the human lung is enormous. Aeby says that there are about 404 millions of them in the adult, repre- senting a surface of 79 square meters in average inspiration when resting and of 129 square meters in a state of complete dilatation. Indeed, according to researches of Leon Bernard, A. Le Play and Ch. Mantoux,1 one-sixth of this surface is adequate to support life. 1 J. de Physiol, et de Path. gen.. 1913, 15, 16. 140 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS Each lobule, the average volume of which is about one cubic centimeter, is separated from its neighbors by connective tissue partitions which enclose lymphatic spaces and which are very compact and resistant, particularly in the adult and still more so in the aged. The epithelium of the penetrating bronchus there loses its cilia, becomes cubical, and transforms itself into small nucleated granular cells, flat, rounded or polygonal, and into broad lamellae or plaques without visible nuclei (cells of Koelliker), which cover particularly the alveolar walls over the edges of the separating partitions. These cells, whether small or large, lie upon an ex- tremely thin fibrillar membrane, reinforced by elastic and smooth muscle fibers. Many alveoli communicate with one another by means of pores which Henle, Hansemann, F. E. Schulze, and more recently E. Laguesse and R. Marchand2 have demonstrated. These pores, ordinarily very small, become expanded in certain pathological conditions of the lungs, principally in emphysema. The epithelial cells of the alveolar wall are constantly desquamating and collecting in the cavity, along with many leucocytes which issue by diapedesis from the interior of the vessels. When an inflam- matory process is set up by any irritant foreign body borne in by the air (tubercle bacillus for example), these migratory dust cells (cellules a poussieres, Staubzellen) are found in great abundance throughout the exuded serous fluid which fills the cavity of the alveolus. At times they are carried on toward the bronchi and swept out of the lungs by expiration and the movement of the cilia of the bronchial epithelium, or they may reenter the perialveolar lymphatic circulation and be carried to the nearest glands. In the denser portion of the connective tissue partitions which separate one alveolus from another the blood capillary system is spread out. The latter projects into the cavity and is so dense that the diameter of the acini is scarcely equal to that of a red blood cell; with the result that the surface of each alveolus is overspread with an almost continuous sheet of circulating blood, separated from the air by an extremely thin single layer of epithelial cells. In miliary tuberculosis, the initial lesions take rise precisely in these vessels, the diameter of which is insufficient for the passage of a 2 Compt. rend. Soc. de biol., 1911, 70, 178. INFECTION BY THE RESPIRATORY PASSAGES 141 leucocyte engorged with bacilli and rendered incapable of movement. The resulting irritant embolus then serves as a location for the forma- tion of a giant cell. About each alveolus and each lobule, the abundant lymphatic spaces and vessels drain the lymph and carry it to the glands of the lung hilus by way of the collecting peribronchial and perivascular trunks, some of which are superficial and subpleural, while others are more deeply placed. In the bullock, each lobule is surrounded by wide partitioned lymphatic spaces which are paved with a wavy epithelium (Pierret, Renaut, Sussdorf) and become engorged with serous exudate in certain pathological conditions (peripneumonie). In air-produced tuberculous infection, the wandering leucocytes issue from these peri-alveolar lymphatic spaces and play the principal role. By causing mice or guinea pigs to inhale a small quantity of tubercle bacilli from a finely suspended culture and then killing the animals at intervals of 24 hours, 48 hours, three days, etc., and up to two weeks after the original inhalation, I could easily follow experi- mentally the evolution of the inflammatory process within the alveoli. Thus it was found that bacilli penetrating into an alveolus quickly induce there an influx of polymorphonuclear leucocytes and dust cells, all of which form at the center of the alveolus a mass which organizes into a small tubercle. This rapidly progresses to caseation and breaking down into pus leads to the extension of the caseous matter into the neighboring alveoli. The result is an infiltration which extends more or less to the whole of one lobule and then to several lobules. Meanwhile the wandering leucocytes have collected bacilli from the substance of this infiltration and have transported them through the perilobular or peribronchial lymph spaces and vessels as far as the corresponding glands of the lung hilus which in their turn become considerably swollen and the seat of tuberculosis. The evolution of these primarily alveolar lesions, when produced experimentally, is always rapid, even in the large animals. Very often the pathological picture of caseous pneumonia of children is reproduced. But if the infection is small in amount, the bacilli ingested by the extravasated leucocytes, instead of giving rise to tubercles in the interior itself of the alveoli, reenter with the leucocytes into the lymphatic circulation and the initial lesions are formed only at the periphery of the lobule, or in the subpleural spaces, or again in the peribronchial glands. 142 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS It sometimes happens, and I have seen it occur in guinea pigs, that a discrete airborne infection induces no intraalveolar lesion, brings about the appearance of no primary lesion (chancre d’inoculation), nor of any pulmonary tubercle, and manifests its effects only more or less tardily by a glandular lesion at some point in the body relatively distant from the portal of entry of the bacillus. In such cases it is impossible to detect any difference between the effects of an airborne infection and those which follow the direct penetration of the infecting element through other lymphatic channels. An argument on favor of the predominance of airborne infection has been drawn from the fact that localizations of tuberculosis occur most frequently at the apices of the lungs, more particularly at the right apex. These localizations are encountered almost constantly following intravenous inoculations into animals, and result from purely mechanical causes and chiefly from the particular arrangement of the lymphatic vessels which permit a prolonged lymph stagnation. Bac- meister3 performed some very conclusive experiments on this subject. He placed a metal band about the thorax of young rabbits, at the level of the first ribs, and as the animals developed the band produced a nar- rowing of the transverse diameter of the thorax and a constriction of the pulmonary and pleural tissue. When, afterward, he injected an emulsion of bacilli into the marginal ear vein or into the superior vena cava of the rabbits, he regularly saw the tuberculous lesion first localize itself in and around the depression caused by the band. In the control animals nothing like this was produced; the tubercles were about equally scattered throughout the whole of the lungs. Bacmeister never succeeded in producing this localization in other similar animals when he caused them to inhale a fine spray of tubercle bacilli, despite the fact that the animals wore the band at the level of the first ribs. B. EXPERIMENTAL PULMONARY INFECTION WITH DRIED BACILLI OR WITH DRIED TUBERCULOUS MATERIAL An early as 1869, Villemin4 had shown that dried and pulverized tuberculous sputum can reproduce tuberculosis when insufflated 3 Deutsch. med. Wchnschr., 1911, 37, 1385. 4 Gaz. hebdomadaire, 1869, p. 260. INFECTION BY THE RESPIRATORY PASSAGES 143 into the rabbit’s trachea. Later however, Cadeac and Malet5 showed that dust collected in hospital wards occupied by phthisical patients, and sputa or fragments of tuberculous lungs dried and pul- verized, produce tuberculosis only exceptionally by inhalation. Of 46 animals, rabbits and guinea pigs, which were made to respire several liters of these dusts for one hour each day during several weeks, only two became tuberculous. In three later works (1898, 1905 and 1907),6 the same investigators still insist on the difficulties encountered in communicating tubercu- losis by inhalation of dried sputa. Of 38 guinea pigs and 11 rabbits which they tried to infect in a bell shaped funnel with an insufflator, 5 guinea pigs only developed the disease and two of these showed lesions indicating that the infection had occured through the digestive tract. Other experiments proved to their minds that sputa dried in day- light are harmless and that those dried in the dark are capable of transmitting tuberculosis by inhalation only very exceptionally and then only when administered in massive doses. Practically speaking therefore, it must be admitted that dusts contaminated with dry bacilli offer but little harm to healthy respiratory passages. Experiments were carried out at Pouilly-le-Fort in 1900, by Nocard and Rossignol, under the auspices of the Societe de medecine veteri- naire and of the Societe d’Agriculture de Melun, with a view to establishing the duration of the period of incubation of tuberculosis in cattle. Two heifers were infected by the inhalation of 3 cc. of dried and finely sifted tubercle bacilli from a culture; they reacted to tuberculin, one on the thirty-second day, and the other on the ninetieth day. They were slaughtered soon after the test and their lungs as well as the bronchial and mediastinal glands were found full of miliary tuber- cles. The viscera and mesenteric nodes were normal in appearance; but proof of the absence of infection was not established by guinea pig inoculation. It was noted only that “the bronchi, the bronchioles and the pulmonary alveoli had escaped infection, and that tubercu- lous nodules were located under the pleura at the periphery of the lobules in the interstitial cellular tissue. Presumably, each tubercu- 5 Rev. de m6d., 1887, 7, 337; Internat. Congr. on Tuberculosis, Paris, 1888. 6 Rev. d’Hyg., 1905, 27, 961; J. de med. vet. et zootech., 1907, 5. s., 11, 65. 144 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS lous focus had formed itself about a phagocyte which reentered into the lymphatic circulation after having ingested one or more tubercle bacilli.”7 In order to settle the question as to whether dry virulent dusts can actually infect the alveoli, I performed the following experiment, in collaboration with Vansteenberghe:8 Two guinea pigs were placed under a bell jar inside of which a strong current of air, carrying a large quantity of bacilli, was made to pass for 20 minutes. The bacilli were of bovine origin, dried under a glass jar in the presence of calcium chloride only 24 to 48 hours at the longest, and then finely pulverized. The two guinea pigs were killed and at once autopsied after this single period of inhalation, and different portions of their respiratory organs, the trachea, and the anterior and posterior lobes of the two lungs of each animal were inoculated separately and immediately into other guinea pigs. Only the pigs inoculated with emulsions from the trachea and anterior lobes presented tuberculous lesions and the latter were very discrete. Those inoculated with the emulsion of the posterior lobes remained free from infection. Therefore, despite the extreme method used to infect, only a very small number of bacilli succeeded in penetrating into the trachea and only as far as the- first bronchial ramifications. Numerous other experimenters. (Baumgarten, Sirena, Pernice, Di Toma, Peterson) have vainly attempted to transmit tuberculosis by inhalation of dust containing live bacilli. In the opinion of Kohlisch,9 the guinea pig must inspire 50,000 dried pulverized bacilli in order to become infected. Cornet in Germany and Kuss in Franee, on the other hand, affirm the virulence of dry bacilli. Cornet10 states that he succeeded in infecting a very large number of experimental animals with dust obtained by beating rugs contami- nated with dried sputa. At the first attempt, 46 out of 48 guinea pigs were infected; and, by the thirtieth day, almost all at once, they showed miliary tubercles and lung cavities. The frequency of these 7 Rep. pub. by the Soc. de med. v6t. pratique, 1901. 8 Ann. de l’lnst. Pasteur, 1905, 19, 787. 9 Ztschr. f. Hyg., 1908, 60, 508. 10 Ibid., 1888, 5, 191. INFECTION BY THE RESPIRATORY PASSAGES 145 cavities would seem to indicate that the guinea pigs used were previ- ously tuberculous. Later the same author repeated the test on 392 animals. Half of them died in a short time from septic conditions. Of 196 which remained, 59 became tuberculous and 137 remained uninfected. According to G. Kuss,11 sputa dried in darkness are still virulent after 15 days; but are no longer so after 19 days. In one series of experiments, he caused guinea pigs to breathe a mixture of powdered sputum and talcum for a period of thirty minutes to one hour. All of his animals died within two or three months with pulmonary and glandular lesions and generalized miliary tuberculosis. In another series, the pigs inhaled for one to three hours the dust from a carpet soiled with dried sputum, which was brushed for one minute every quarter or half hour. The animals died in from 40 to 86 days and at autopsy presented pulmonary or mediastinal lesions with extensive miliary tuberculosis of all the organs. Chausse12 thought that bacilli contained in dried sputum are but slightly virulent after 24 hours and absolutely innocuous after 10 days. Five days at room temperature and only 12 hours at 37°C. are sufficient to deprive them of their power to infect. His experiments tend to show that tuberculization of susceptible animals, such as the guinea pig, is better accomplished by brushing or shaking pieces of cloth or soiled linen in a tightly closed box. One to five minutes of breathing are sufficient. C. EXPERIMENTAL INFECTION OF THE LUNGS WITH FRESH BACILLI AND WITH SPRAYS OF TUBERCULOUS MATERIAL Although it is difficult to communicate tuberculosis to animals by making them inhale dry infectious dust, it appears that, on the con- trary, the infection may be easily effected by the inhalation of fine liquid droplets, which contain tubercle bacilli in suspension. As early as 1876 Tappeiner13 had placed dogs, for a few hours each day, in a box wherin he ground up a large amount of sputum of phthisical patients, and in his animals had observed obvious lesions. In 1877 he extended his experiments with a better technique: he 11 Compt. rend. Acad, des sci., 1908, 147, 272; Bull, med., 1908, 22, 709. 12 Ann. de PInst. Pasteur, 1914, 28, 771; 1915, 29, 556; 633; These, Paris, 1916. 13 Virchow’s Arch., 1880, 82, 353; Deutsch. Arch. f. klin. Med., 1881, 29, 595. 146 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS mixed a teaspoonful of sputum from a case of tuberculosis with cavities in 300 to 500 gms. of distilled water and, by means of a steam atomizer, sprayed the liquid into a cage which was open on only one side, and in which was a dog. The animal was subjected each day to one or two inhalations of one hour duration. The number of inhalations and the period of stay in the cage were varied and the experiments continued 24 to 45 days. Twelve dogs were treated in this manner. In all, at autopsy the lungs were found full of tubercles: ten times in the form of miliary tubercles, once as a caseous pneumonia. The kidneys were usually involved; the liver and the spleen less constantly. Evidences of tuberculosis were pres- ent after the third week. Thaon,14 in 1885, exposed rabbits and guinea pigs for a quarter of an hour, morning and evening, during a period of one week, to a spray of ground-up tuberculous sputum suspended in water. By the end of the third week, the lungs of the rabbits were infiltrated with small gray tubercles and the guinea pigs died invariably in 12 or 14 days, with extreme dyspnea. At autopsy their lungs formed a bluish-red solid mass, peppered with yellow points. By killing a few animals successively from the outset of the experiment, it was possible to detect in the slides the arrival of the bacillus by way of the bronchioles, its penetration to the extremity of the respiratory passages and its multiplication in the pulmonary epithelium. Cadaec and Malet15 had also observed that the infecting powTer of droplets when inhaled contrasts singularly with the difficulty encoun- tered in producing infection with dried dusts. Of 45 guinea pigs made to inspire either fresh cultures of bacilli or finely triturated sputa, not one remained uninfected. Nocard and Rossignol, in the already cited experiment at Pouilly- le-Fort, covered the heads of two cows with cloth bags and obliged them, during a period of six minutes, to breathe in 100 cc. of a fine emulsion from a culture of bovine tubercle bacilli. The two animals reacted to tuberculin after 13 and 19 days. At autopsy, their lungs were found infiltrated with a considerable number of miliary tubercles. Another animal, however, into the trachea of which were injected 10 cc. of the same emulsion, did not react until the 38th day. At autopsy, one month later, there was not trace of lung lesions; the 14 Compt. rend. Soc. de biol., 1885, 37, 582. 16 Rev. de Med., 1887, 7, 337. 147 INFECTION BY THE RESPIRATORY PASSAGES raucous membrane of the trachea alone was studded with fine tu- bercles about the point of needle traumatism; the retropharyngeal glands, however, and the bronchial and esophageal glands were full of small tubercles. The bacilli therefore, despite their being intro- duced into the bronchi in enormous quantity, had been eliminated in the expectorated mucous discharges. Vallee (of Alfort)16 met with the same difficulties in infecting the lungs with bovine bacilli when he injected them directly into the trachea and when he sprayed a fine emulsion of ground-up bacilli into the naso-pharynx. Of 12 calves twice sprayed with 2 mgms. of bacilli, only 4 contracted tuberculosis. The lesions were confined exclusively to the retropharyngeal, cervical and tracheal lymph nodes, the lungs and tracheo-bronchial glands not being affected. In order that the bacilli may surely penetrate as deeply as the alveoli, it seems necessary, as I showed with C. Guerin,17 that they be projected as far into the trachea as the bifurcation,—that is to say into a zone not subject to the cough reflex,—by means of a flexible sound and in a fine state of suspension in a large volume of fluid. At once there is obtained a massive broncho-pneumonia, with miliary tubercles in whose formation the lymphatic vessels of the alveolar walls take part. However, these are not at all the lesions which are ordinarily pro- duced by the penetration into the alveoli of a few separate bacilli held in suspension in moist dust. C. Fliigge18 (of Breslau) and his pupil H. Findel produced these lesions with the greatest ease, just as Preyss had already succeeded in doing (1891). And I have repeated their experiments many times with V. Grysez. Our method was to fasten guinea pigs or mice in a Reichenbach apparatus or in a still simpler one, like that shown in figure 9, and to make them inhale during varying periods of time quantities of bacilli which could be counted with sufficient accuracy and which were held suspended in the spray of a Buchner atomizer. 16 Ann. de l’lnst. Pasteur, 1905, 19, 619. 17 Ibid., 1906, 20, 609. 18 Ztschr. f. Hyg., 1907, 57, 104. In the work published by Fliigge in 1908 (von Veit, Leipz.) under the title “Die Verbreitungsweise und Bekampfung der Tubsrkulose” will be found in detail the investigations of Fltigge and his collaborators;. Findel, Laschtchenko, Bruno Heymann, Ziesche, Neisser. Sticker, Beninde, Koliscb, and others. See also P. Chausse, These, Paris ,1916. 148 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS It was found that infection of the animals, accomplished now and then with a few bacilli (15 or 20, perhaps fewer), was more rapid and more intense in proportion to the number of bacteria inhaled. An average of 20 very virulent bovine bacilli suffice to produce in guinea pigs a miliary tuberculosis, with generalized lesions in all the viscera and fatal in 90 to 180 days. The bacterial elements contained in a virulent suspension can be counted without difficulty by the following method: in 10 centigrams of a culture or of tuberculous material, triturated in an agate mortar, there are first carefully mixed 2 or 3 drops of ox bile, and then a Fig. 9. Schema of Apparatus Used to Infect the Guinea pig by Inhalation measured quantity of physiological salt solution, 10 cc. for example. One drop of this dilution is placed upon a special slide which I had made by Stiassnie19 and which contains a ruled area, each side being one half centimeter long and the whole divided into 25 squares. The surface of each square, therefore, represents 1/100 of a square centimeter. The drop is allowed to dry, the slide being kept horizontal. After the smear is fixed and stained withZiehl by the ordinary technique the bacilli deposited upon each of the 25 squares are counted with an oil immersion lens and the total multiplied by 20 (number of drops per 19 204, Boulevard Raspail, Paris. 149 INFECTION BY THE RESPIRATORY PASSAGES cubic centimeter). The result gives the average number of bacilli contained in one cubic centimeter of the dilution. One milligram uj culture weighed in the fresh state contains an average of 40 million bacilli. Although using a technique more or less different from that of Fliigge or myself, numerous investigators of late years have produced direct infection in various animals (guinea pigs, dogs, cats, cattle) by making them inhale moist bacillus-containing dust. Some used sputa from phthisical patients (Hamilton and Young, Weber and Titze, Chausse, in the calf, the dog and the cat; Kuss and Lobstein, in the guinea-pig); others preferred to use cultures of human, bovine or avian strains (Kossel, Weber and Heuss in the calf, Kuss and Lob- stein, Louis Cobbett (of Cambridge) in the guinea pig, Weber and Bofinger in the mouse). In 1907, H. Ziesche,20 in Flugge’s laboratory, was able to satisfy himself that some 30 to 40 per cent of phthisical patients eject in coughing, to a distance of 40 to 80 cm., droplets of saliva which are more or less rich in bacilli. These he collected in properly placed Petri dishes and was able to count the number of bacilli expelled, for example, during a half hour. In 20 per cent of the positive cases the number varied from 400 to 20,000 bacilli, while in 80 per cent fewer than 400 bacilli were found. It is evident that these bacillus- containing particles, which deposit themselves upon all sorts of objects (particularly on foods) and which at times contaminate the face or hands of those who are obliged to live in the immediate vicinity of these coughers and spitters, constitute a continual and most dangerous source of infection. The absorption of virulent elements so disseminated does not, however, in most circumstances, seem to be brought about by the respired air; it occurs much more readily and certainly through the mucous membranes, through the conjunctiva, through the mouth or by the digestive tract. Chausse,21 like Sanger22 before him, studied the conditions,-—purely physical according to him,—which permitted the liquid particles to pentrate into the lung. He proved by some particularly ingenious experiments that it is only the particles from about 2 to 15 microns Ztschr. f. Hyg., 1907, 57, 50. 21 Compt. rend. Acad, des sci., 1913, 156, 954, 1485; 157, 862; 1914, 158, 134;— Ann. de l’lnst. Pasteur, 1914, 28, 608; 720; 1915, 29, 556; 633;—Bull, de l’lnst. Pasteur, 1917, 15, 33; 65. 22 Munchen. med. Wchnschr., 1901, 48, 831. 150 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS in diameter which are respirable, since they remain in suspension in the atmosphere during a long period (up to 7 hours) and because they can be reflected by surfaces which arrest particles of greater volume. According to this investigator, the passing of an air current over the surface of saliva or sputum, even at an initial velocity of 90 meters per second, separates out only a very small number of respirable particles. Of 31 guinea pigs which inhaled air blown under these conditions over the material containing the organisms, 600 litres of air having passed through the inhalation apparatus, but a single animal contracted one primary lung tubercle; but at this high velocity there was ruffling and breaking of bubbles. With initial velocities of 35 meters per second or less, deep blowing of the sputum gave only a single tuberculous animal in 5 experiments on 33 guinea pigs. One guinea pig among twenty-two was infected with the velocity at 36.5 meters. Another experiment with a veloc- ity of 80 meters per second was entirely negative. Chausse concluded from his researches that air contact at velocities of 30 meters per second or less can detach only a very small number of respirable particles from sputum or saliva; that, contrary to Fliigge’s idea, the majority of droplets borne away by the blowing process are too large to be transported and then inhaled, while those which are minute contain usually no bacilli. No one denies, however, that bacteria suspended in very fine drops can, at times, penetrate by inhalation into the pulmonary alveoli. Experiments show that this is possible, that there results a rapid migration of leucocytes which phagocytize the bacilli, and that, according to the number and virulence of the latter, character- istic tuberculous lesions are soon formed, either in loco or in the inter- alveolar or subpleural lymphatic spaces. These lesions, primarily alveolar or peri-alveolar, extend somewhat later to the tracheo- bronchial nodes and very often, by blood or lymphatic channels, to other organs. This metastasis, as well as the fixation of infectious matter in the parenchyma of certain organs such as the lung, are greatly favored by the prolonged absorption of particularly noxious dusts (plaster, cement, mother-of-pearl, emery, etc.) as D. Cesa- Bianchi23 demonstrated experimentally. But it does not follow that this aerogenous path, under normal 23 Ztschr. f. Hyg., 1913, 73,166. INFECTION BY THE RESPIRATORY PASSAGES 151 conditions of natural infection, is the one by which man and suscep- tible animals are the most commonly and frequently infected. D. CONDITIONS AND RELATIVE FREQUENCY OF PRIMARY INFECTION OF THE LUNG BY THE INSPIRED AIR It should not be forgotten that the complexity of the respiratory passages (large filtering surfaces of the nasal cavities, pharynx, glottis, the length of the trachea and bronchi, and their lining ciliated epithelium), their reflex irritability which determines sneezing and the expulsion of offending particles, and the presence finally, throughout their length, of mucous-secreting glands designed to prevent extravas- ated leucocytes from reentering into the lymphatic circulation, render it exceedingly difficult for dust and infectious bacteria to penetrate directly as far as the alveoli, except under quite unusual circumstances where dusts and germs are extremely abundant in the respired air. These conditions are met with, as regards mineral or organic dusts, in certain industries (mining, metal working or textiles, for example) and they then induce anthracoses or pneumoconioses which,however, are usually harmless. They are also met, as regards tubercle bacilli, in the experimental conditions which are described above. But it is altogether improbable that tubercle bacilli, floating isolated in the air, in the fresh state, are often sufficiently numerous for many of them to escape the natural obstacles which the body opposes to their pene- trating as deeply as the alveolus. This improbability is the greater since, as numerous workers (Friedrich Muller,24 Klippstein,25 Bartel,26 Boni,27 Emmerich,28 Quensel29), have shown, healthy lungs are almost always aseptic. And yet Saint-Clair Thomson and Hewlett30 have shown that in London more than 14.000 microorganisms are inhaled per hour, and Hildebrand,31 Lucien Beco,32 and other authors cited in the lat 24Miinchen. med. Wchnschr., 1897, 44, 1382. 25 Ztschr. f. klin. Med., 1898, 34, H.3/4. 26 Centralbl. f. Bakt., 1898, 24, 401; 433. 27 Deutsch. Arch. f. klin. Med., 1901, 69, 542. 23 Miinchen. med. Wchnschr., 1902, 49, 1610. 29 Ztschr. f. Hyg., 1902, 40, 505. 30 Brit. M. J., 1896, i, 137;—Med.-chir. Tr„ Lond., 1895, 78, 239. 31 Beitr. z. path. Anat., 1888, 2, 411. 32 Arch, de med. exper., 1899, 11, 317. 152 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS ter’s work, have proved that the trachea of animals killed in their laboratories almost never contained any bacteria. J. Arlo33 demonstrated the same thing at the Pasteur Institute of Lille, when he cultured the lungs and tracheo-bronchial lymph nodes of a fairly large number of guinea pigs taken at random in the en- closures (where hay and straw were used for bedding, without any special precaution against dust) and killed by decapitation. He thus found that the right lung in 94.03 per cent of cases was sterile after 24 hours of culture and in 78.62 per cent after 48 hours of culture. The left lung was sterile under the same culture conditions in 99.02 and 79.02 per cent of cases respectively. The tracheo-bronchial lymph nodes on the other hand, were found fairly often infected or harboring bacteria which could be revived. They were sterile in only 52.28 per cent of cases after 24 hours and in 20 per cent after 48 hours. J. Arlo concluded from his experiments: 1. That even in the very dusty atmosphere of the guinea pig enclosures, the air bacteria are arrested by the upper respiratory passages and reach only rarely the pulmonary alveoli. 2. That the bacteria which can be recovered in the lungs are present in only small number, since most frequently they appear in culture media only after 48 hours. 3. Finally, that in normal animals living under ordinary condi- tions, the tracheo-bronchial glands retain many bacteria caught by the leucocytes in the bronchi and carried into the lymphatic circula- tion, since, in about 50 per cent of cases, culture of these glands upon artificial media gives growth of saprophytic bacteria not yet digested by the phagocytic processes. It is seen therefore that the defensive and eliminatory functions of the different structures guarding the entrance to the respiratory passages are conducted with wonderful perfection. Furthermore, if primary infection in man were air-borne as fre- quently as many physicians today still think, and as P. Chausse believes he has demonstrated,—without his experiments and argu- ments carrying conviction,—it would be incomprehensible why infection is so rare in animals readily susceptible to the human virus (the cat, the dog, the ass, the rodents of our homes or laboratories) 33 Compt. rend. Soc. de biol., 1914, 76, 292. INFECTION BY THE RESPIRATORY PASSAGES 153 and which breathe the air ruminated,-—as Peter used to say,—by tuberculous individuals who throw off at times, according to B. Fraenkel, more than 7 billion bacilli in their sputum in a single day. If, as Chausse contends, a single bacillus penetrating into an alveolus is sufficient to set up a tuberculosis, it would be still less conceivable, considering the relative abundance and ubiquity of the tubercle bacillus, that one man or one susceptible animal in our hospitals or our cities or even on the face of the earth should escape the disease. The fact is well established moreover that the air expired by phthisical patients contains no infectious elements. Tappeiner, and Grancher tried many times to infect animals by making them inhale, for a longer or shorter period, air which came directly from the respiratory passages of patients. They obtained only negative results. The bacilli are contained only in the particles of saliva ejected in the efforts of coughing, and these are the minute droplets which Fliigge, B. Heymann, and P. Chausse34 regard as the most active agents in the propagation of tuberculosis. Now, according to certain experiments of Chausse, these liquid particles have a diameter of at least 30 microns and are not respirable! In his opinion they become infectious and capable of penetrating into the lungs only when in the dry state, and they dry very quickly in the atmosphere after their emission.35 Should it be said that primary air-borne infection is to be regarded as practically non-existent? That would be contrary to the evidence of facts which, although rare, seem well founded. Among such facts, one of the most striking, because equivalent to a laboratory experiment, is that published by Reich36 and bearing upon 10 new-born infants brought into the world by a phthisical midwife at Neuenburg. All of them died of tuberculous meningitis within 14 months, without there being any similar occurrence among children delivered by other midwives. The phthisical woman had the deplorable habit of practicing insufflation with the mouth, even when the infants showed no sign of asphyxia (Grancher and Hutinel). 34 Ann. de l’lnst. Pasteur, 1916, 30, 613. 36 Soc. centr. de m6d. vet., Mem. pour le prix Trasbot, 1912; Rec. de med. vet., 1912, 89, 555. 36 Berl. klin. Wchnschr., 1878, 15, 551. 154 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS One cannot help thinking that primary pulmonary infection occurs at times in a similar manner in very early life, for example, in the infant whose phthisical mother or nurse coughs before its mouth, the latter held open in search of the breast which is to feed it. Massive aerogenous infection can then produce intra or perialveolar primary foci as in the above-described experimental animals, and the always rapid evolution of the foci soon ends in the caseation and purulent softening which characterize the caseous pneumonia of nurslings. But the problem is much more complex when we have to definitely establish the aerogenous origin of tracheo-bronchial adenopathies which are so frequent in children and which in a very large number of cases are the sole earliest indication of tuberculous infection. Every one knows that Parrot37 believed himself justified in asserting that there exists no tracheo-bronchial adenopathy that does not originate in the lungs. “Whenever a bronchial lymph node is the seat of a tuberculous lesion there is a tuberculous lesion of the lung.” This “loi des adenopathies similaires” or law of similar adenopathies, based upon a large number of incontestably well-proved facts, de- fended by Hutinel, Hervouet, G. Kiiss,38 Albercht, and Anton Ghon,39 is often quite correct if one considers only the findings in infants after death. It frequently happens in fact that when one or more caseous tubercles exist in the bronchial gland group, one or more other tubercles are to be found in the corresponding lung zone. Parrot, Hutinel, and G. Kuss hold these lung tubercles to be the “chancres d’inoculation” or primary foci of infection, but there is nothing to prove that they really mark the portal of entry of the bacillus into the body. The fact that many extremely carefully made autopsies reveal no pulmonary tubercle (Marfan, Weigert, Biedert, Bollinger, 0. Muller, and others) and, on the other hand, the customary absence of any lung lesion in the tracheo-bronchial adenopathy of animals killed three to four weeks after experimental infection by the instillation of cultures or bacillus-containing sputa upon the healthy mucous membrane of the eye (Calmette),—while, if one waits longer, isolated or solitary tubercles identical with what 37 Compt. rend. Soc. de biol., 1876, 28, 308. 38 L’heredile parisitaire de la Tuberculose humaine. Paris, 1898, Asselin & Houzeau. 39 Primdre Lungenherd bei der Tuberkulose der Kinder. Berl., 1912, Urban & Schwanzenberg. INFECTION BY THE RESPIRATORY PASSAGES 155 Parrot calls the primary foci of infection, appear on the surface of the lung,—prove that these latter are but secondary manifestations of the so-called primary tuberculosis of the hilic glands. It can be easily demonstrated that in cases of spontaneous infection, even though massive, no local lesion is produced as a rule at the point of penetration of the bacillus. The supposed “chancre of inocula- tion” therefore implies in no way that the virulent elements from which it develops were deposited by the inspired air at the very place of its appearance. To be convinced one need only repeat the fol- lowing experiment which I performed with V. Grysez.40 Upon the mucous membrane of one of the eyes, in a series of guinea pigs, we instilled one drop of a suspension containing 0.5 mgm. of a culture of bovine bacilli in physiological salt solution. The animals were killed successively after 2,4, 6, 8,12,15 and 18 days and from each one there were removed, separately and aseptically: the retro-mastoid, retro-pharyngeal and cervical glands, the tracheo- bronchial glands and the lungs, spleen and liver. The organs thus removed were completely ground up in physiological salt solution and inoculated into guinea pigs, all of which were sacrificed at the end of three months, none having succumbed in the interval. The results showed that bacilli were present in the lungs as early as the fourth day and that by the sixth day they could, by experimental inoculation, be demonstrated to have been present in the lungs, cervical glands and spleen, before any visible tuberculous lesion was to be made out. On allowing a longer interval to elapse before killing the animals infected by ocular instillation, it is found that at the end of three weeks, only the glands of the neck are swollen and that frequently a few tubercles already exist in the lungs; later other organs become involved, particularly the spleen and the tracheo-bronchial nodes; then those of the hilus of the liver and the mesenteric nodes. It seems evident, therefore, that in the case of a local infection, ocular, pulmonary, intestinal, cutaneous, etc., the lymphatic and blood infection becomes general before manifesting itself in the form of miliary lesions in the glands near the site of penetration of the bacillus. According to a statement which O. Medin was kind enough to send to me, there were in the Stockholm hospital, from 1842 to 1910 inclu- 40 Compt. rend. Acad, des sci., 1913, 157, 981. 156 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS sive, among 7590 infants dying in the first year, 622 who were infected with tuberculosis, although dying of the most varied diseases. Examination of the autopsy protocols of these 622 cases showed that there were 194 cases with tuberculosis localized exclusively in the lungs and bronchial glands, but in only 17 cases were the tubercles limited to the lungs alone. Of 287 children, Medin judged that the primary infection was undoubtedly air-borne and, from his other findings, heconcludedthat a similar origin was to be ascribed to 97.7 per cent of cases of tuber- culosis in infancy. This estimate of the celebrated Swedish clinician is approved by many pediatricians. I do not believe it accurate because it takes account only of the results of autopsies which, as I said before, permit one only exceptionally to identify the true portal of entry of the tuberculous virus, and because it discounts all the experimental results going to prove that a lymphatic infection, through any mucous membrane whatever,—that of the eye for example, or of the pharynx, or the digestive tract,—manifests itself usually by glandular lesions to begin with, and by pulmonary lesions afterward, in circumstances where aerogenous infection is out of the question. It cannot be denied that the tuberculous virus may enter into the body by the pulmonary route; but, for the reasons which I have stated above, penetration by this path is certainly much more diffi- cult than through the mucous membranes of organs which are directly exposed to air contamination or through those whose essential function is absorption (the buccal cavity and the digestive tract). The frequency of primary air-borne infection of the lungs has been much exaggerated, and such is the case because Cohnheim’s law imposed itself as a dogma on the minds of physicians up to the actual moment when there was introduced into science the fruitful idea of occult infections by the tubercle bacillus (typho-bacilloses and infections without tubercle formation). CHAPTER X TUBERCULOUS INFECTION BY ABSORPTION FROM THE DIGESTIVE TRACT A. THE MECHANISM OF DIGESTIVE ABSORPTION OF TUBERCLE BACILLI. THEIR MIGRATIONS IN THE BODY In the lower animals, such as the fresh water hydra, which possess a digestive sac, the cells of the entoderm send forth into the interior of the cavity pseudopods which are like those of the amebae and which ingest solid particles capable of serving as food. In the higher animals, the enormous absorbing surface of the digestive tube is covered almost throughout with fixed epithelial cells which do not possess the same properties. The stratified pavement epithelium which covers the esophagus, the dense layer of cylindrical, prismatic or pyramidal cells which line the stomach, the great number of glands which empty their abundant secretions upon the surface of these structures, as also the very nature of the secretions, all of these elements do not normally permit the passage of leucocytes through their walls. These migrations can take place only in consequence of some sort of an irritation lesion or through a traumatism. Moreover, the so-called primary tuberculous localizations occur there only very exceptionally. Tuberculous lesions of the esophagus and stomach in man, and of the true stomach and rumen in cattle are cited as pathological curiosities. On the other hand, the phenomena of absorption are accomplished with increasing intensity from the duodenum to the termination of the small intestine, and with diminishing intensity from the ileo- caecal valve to the rectum. They do not consist solely in a simple penetration by osmosis of the food substances dissolved by the diges- tive juices. It is known today that the intestinal epithelium permits the transit of a variety of protein substances and fats previously split into fatty acids and glycerins. It is also known,-—and this fact is all important for the question which concerns us,—that bacterial 157 158 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS bodies and certain mineral particles in fine suspension in the chyme are constantly transported during digestion from the interior of the intestine into the central chyle-bearing vessels of the villi. This passage outward is accomplished through the intermediary of the wandering cells which penetrate between the cylindrical epithelial cells of the intestinal epithelium or, as Renault puts it, into the very interior of the latter which they transform into true fenestrated cells (temporary stomata). It is specially active at the level of Peyer’s patches which represent in a certain sense veritable lymphatic sponges and which are particularly abundant in the ileum. In order to understand the mechanism of absorption of bacteria from the intestine, it should be recalled that the food mass, which contains the microorganisms in relatively large number, does not pass through the digestive tract as though gently pushed along in the interior of a cylindrical tube of uniform diameter. It progresses very slowly, and intermittently, under the control of peristaltic and antiperistaltic movements which cause it to penetrate into one after another of the deep pocket-like depressions between the valvulae conniventes and between the villi. If one admits with Sappey that the small intestine alone has a capacity of 6 to 8.8 liters and a surface area of 10.125 square meters, not including that of the villi, one can form an idea of the very favorable conditions there existing for the introduction of infectious elements into the lymphatic circulation of which the highly developed superficial network collects all the chyle produced during digestion. When a bacterium (tubercle bacillus or other variety), conveyed by a wandering leucocyte, has penetrated into a chyle-bearing vessel, it follows the current of the lymph which floats it first through the filtering lymph nodes (Schaltdrusen) and then into the cavernous sinuses of the mesenteric gland corresponding to the region whence it came. The glands of the mesentery, numbering from 130 to 150 in man (Quain), vary greatly in size, from the volume of a millet seed to that of an olive. They are larger in children than in the aged. In cattle they form an almost unbroken beaded chain of flattened drawn-out strands, between the layers of the peritoneum, throughout the length of the suspensory ligament of the intestine. In these nodes the lymph current is considerably retarded by numerous obstacles opposed to it by the gland sinuses and the con- nective tissue partitions which separate them. If the bacilli conveyed INFECTION BY ABSORPTION FROM DIGESTIVE TRACT 159 are in large number (after a massive infection for example), or if they have had time to multiply in the wandering cells which have phagocy- tized them and to kill these cells, endothelial reactions, leading to the formation in loco of giant cells, immediately supervene and tuber- culosis of the gland will make its appearance. A little later, as caseous softening occurs in the earliest tubercles, the infection will by degrees involve other gland groups along the course of the efferent lymphatics, or it will set up a general infection of the body by disseminating a greater or lesser number of bacterial elements into the lymph stream, the latter in man being poured into the blood mass by the thoracic duct at its confluence with the left subclavian vein. But if it happens that the infecting bacilli are isolated, few in number, or not very virulent,-the leucocytes ingesting them remain unharmed, despite the presence of these undigested parasites in the cell protoplasm; they preserve their motility and continue their migrations in the lymph or blood circulation of the various organs up to the time when, sooner or later, they end by undergoing death. Then, at the point where the dead cells form a capillary embolus, perhaps far removed from an obscure portal of entry of the bacilli a tuberculous lesion develops. Thus, following a non-massive infection, of no matter what origin, whether produced through an excoriation of the skin, through a healthy mucous membrane, by way of the respiratory passages or by the intestine, there appears occasionally an isolated localization of tubercle bacilli in some organ, be it lung, pleura, joint or other serous membrane, bone, testicle or ovary, larynx, etc. But the lung is the most liable to be the seat of this localization, by reason of the immense surface which it presents for the development of a blood and lymph capillary system which are here more extensive and delicate than anywhere else. The great frequency of so-called, primary tuberculosis of the lung is therefore due to the fad that it represents the first manifestation of a bacillary infection, which may have occurred by way of any lymphatic or blood route, often long before the appearance of disease, and which may have remained latent possibly for years. If the reader will keep well in mind what has already been said, he will now be convinced without difficulty of the fact that most tubercu- losis, in earliest infancy as well as in adult life, has its origin in intesti- nal infection, at times very remote or again very recent, massive or 160 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS discrete, more virulent or less virulent, and which leaves no trace in the glands adjacent to the portal of entry (Law of Cohnheim) except when the number and character of simultaneously penetrating bacilli oblige the body to immediately expel them by setting its defensive cellular reactions at work. The latter event takes place, as we shall see later, only in subjects already previously infected and rendered partially immune (phenomenon of Koch). B. EXPERIMENTAL DEMONSTRATION OF THE PASSAGE OF TUBERCLE BACILLI THROUGH NORMAL DIGESTIVE MUCOUS MEMBRANE.— THE COURSE WHICH THEY FOLLOW IN INFECTING THE LUNGS OR OTHER ORGANS Long before any experimental research had been undertaken as to the transmissibility of tuberculosis by the ingestion of infectious material, Malin1 had published the very interesting observation of two dogs belonging to a phthisical woman of 58 years. Both dogs had greedily swallowed her sputa and had died one after the other with enormous suppurative lesions of the two lungs. But without doubt we owe the demonstration of transmissibility of tuberculosis by the digestive tract to the excellent experiments carried out and published by Chauveau from 1868 to 1872.2 This scientist furnished the first examples of tuberculosis of the lungs and of the bronchial and mediastinal glands, of certain intesti- nal origin, in which no trace of lesions at the portal of entry of the virus could he detected. It should be acknowledged also that we are indebted to him for the conception,—which today appears so important,—that, following the ingestion of tuberculous material, primary pulmonary tuber- culosis, with or without tracheo-bronchial adenopathy, may occur in calves, regardless of whether the infecting virus be of human or bovine origin. From 1868 Chauveau was writing, “it is evident that the natural and spontaneous contagiousness of tuberculosis cannot be attributed exclusively to infection of the surrounding atmosphere by air expelled from the lungs of phthisical subjects. Animals confined in the same 1 Gaz. medicale de Paris, 1839, p. 634. 2 Bull. Acad, med., 1868, 33, 1007; (Letter to Villemin) Gaz. hebdomadaire, 1872, 215; Assoc, pour l’avanc. des sci., Lyon, 1873, p. 727; Lille, 1874, p. 943; Compt. rend. Acad, des sci., 1907, 144, 777; 817. INFECTION BY ABSORPTION FROM DIGESTIVE TRACT 161 stable or in the same pastures, drinking from a common spring, tank or receptacle, have there the opportunity to be constantly swallowing mucous discharges from the noses of other animals. Now, if these excretions are from phthisical animals, they cause tuberculous infection. This is likewise true for the human race.” And the following conclusion, the truth of which becomes more obvious every day, stands out from these researches: The digestive tract, in man as in cattle, constitutes one path of tuber- culosis infection, and it may indeed come into play more frequently than the pulmonary. By a long series of experiments first at Hanover and then at Berlin, Gerlach3 confirmed in 1870 the facts announced by Chauveau and was the first to demonstrate that tuberculosis could be transmitted to healthy animals by the ingestion of milk from tuberculous cows. Meanwhile Villemin,4 Parrot,5 and a little later Klebs, Gunther and Harns, Saint-Cyr, Viseur,6 Toussaint,7 Peuch,8 Baumgarten, Wesener, Perroncito, Sydney Martin, Schottelius,Nocard and Rossig- nol, de Haan, Vallee,9 and others published a large number of facts proving that the rabbit, guinea pig, cat, dog, pig, sheep, goat, ox, and monkey contract tuberculosis following the ingestion of various tuberculous products (milk, sputa, ground-up organs); that these different animals show a variable species susceptibility, being greater in ruminants than in carnivores; and finally that infection is more easily produced in the young than in the old. And yet, in many cases, attempts at infection were unsuccessful, even though repeated, and the reason could not be discovered. (Colin, Semmer). Straus and Wurtz10 supposed that the gastric juice intervened to destroy the bacilli, but experiment proved to them that this was not the case. At that time the great differences in the virulence of tuberculous material, according as it was of bovine or human origin, were not understood. It was supposed therefore that the resistance offered in certain cases to penetration of the bacilli 3 Jahresb. der Tierarztl., 1869, p. 6. 4 Gaz. hebdomadaire, 1869, p. 260. 6 Soc. med. des h6p., 1869, March 12. 8 Bull. Acad, med., 1874, 38, 890. 7 Compt. rend. Acad, des sci., 1880, 20, 754. 8 Ibid., 1880, 20, 1581. 9 Internat. Congr. on Tuberculosis, Paris, 1905. 10 Arch, de med. exp6r., 1889, 1, 370. 162 TUBERCLE BACILLUS INFECTION AND TUBERCULOSIS through the digestive tract was due to the protection of the epithelial covering of the mucous membrane although Chauveau, Wesener,11 and later Dobroklowski,12 had already insisted upon the ease with which the tuberculous virus could pass through the normal epithelial lining of the intestine, without producing any apparent lesion. But this idea was meeting with opposition. Robert Koch, and Baumgarten believed that the bacillus always left its trace in the form of a visible lesion at the point of penetration. Meanwhile Desoubry and Porcher,13 in Nocard’s laboratory at Alfort, had established the fact that many bacteria of all sorts pass through the intestinal mucosa during the digestion of fatty sub- stances and are found for several hours in the chyle and in the blood. This finding has been verified so often since then that it is now a rule in all institutes of serotherapy to bleed the horses only when fasting, if sterile sera are to be obtained. Furthermore, Nicolas and Dercas14 demonstrated the same fact as regards the tubercle bacillus. They incorporated virulent cultures into a fatty soup which was fed to dogs. When these animals were killed at the height of digestion, 3 hours after the infecting meal, and chyle was collected from the receptaculum chyli and inoculated into guinea pigs in doses of 5 to 10 cc. tubercle bacilli were found with doses which were very small in comparison with the whole amount of fluid absorbed by the chyle-bearing vessels during the three hour interval. Maz. Ravenel, Von Behring and Roemer, Bisanti and Panisset,15 Ficker, Oberwarth and Lydia Rabinowitsch16 repeated these experi- ments and found that after a meal of infectious material, not only the lymph but also the heart-blood frequently contained bacilli, and that the results were the more constant the younger the experimental animals employed, the intestinal wall of suckling animals permitting the passage with the greatest facility. In the new-born in fact,'—as Disse17 has shown (and this is a fact 11 Habilitationschrift, Freiburg, 1885. 12 Arch, de med. exp